Free roaming dog population, community perception and control of dog related rabies: the Indian story
Harish Kumar Tiwari
A thesis presented in fulfilment of the requirement for the Degree of Doctor of Philosophy
School of Veterinary Medicine Murdoch University Western Australia September 2019
Declaration
I declare this thesis is my own account of my research and contains as its main content work which has not previously been submitted for a degree at any tertiary education institution.
Harish Kumar Tiwari
ii
Abstract
Most human deaths from rabies in India are caused by the bites of free roaming dogs
(FRD), however studies on the demography of FRD and the community perception of rabies and FRD are virtually lacking in the country. This study was conducted in rural and urban India to: recommend a reliable enumeration method for FRD; describe the demography of FRD; assess the knowledge, attitudes and practices (KAP) of communities towards rabies and FRD; and assess the KAP of rural para-medical staff on dog-bite wound management.
The Application SuperDuplicates online tool was found to reliably enumerate the FRD population size with minimal resources. In the rural site fewer dogs were sighted within
20 metres of garbage points (OR 0.3) than more distant; while in the urban site more FRD were sighted near garbage points (OR 1.6) than away from these sites. The re-sight probability (β = 0.3) and de-sexing status (β = -0.07) of FRD had a positive and negative influence, respectively on urban FRD forming groups. The tendency to form groups in the rural FRD was influenced by frequency of being re-sighted (β = -0.1) and presence of garbage within 20m (β = 0.2). The FRD in the rural setting that were sighted in groups had a larger home-range (>0.11 ha) than those sighted alone (≤0.11 ha).
Rural respondents with a smaller family size (OR 2.1) were more knowledgeable about rabies, than those with bigger families and older respondents (OR 2.6) had a more positive attitude towards FRD than did younger respondents (<35 years). Urban respondents from high/middle socio-economic sections were more knowledgeable (OR 3.03) with positive attitudes and practices (OR 3.4) towards rabies than those from lower socio-economic sections. Urban households containing children (≤ 14 years) (OR 0.5) had a lower level
iii
of knowledge about rabies compared to households with older or no children.
Experienced and graduate paramedical staff were more aware (OR 3.4) and adopted adequate practices (OR 5.6) regarding the management of dog-bite wounds than less experienced or non-graduate staff.
It is recommended that control of dog-related rabies in India requires: enumeration and interpretation of the demographic characteristics such as tendency to form groups and the spread of home ranges of FRD to strategically adopt mass-immunisation; concerted efforts to promote knowledge and adoption of healthier practices in rural communities; educational outreach directed towards the lower socio-economic sections in the urban community; and the development and implementation of compulsory training modules for rural paramedical staff on dog-bite wound management.
iv
Table of Contents
Declaration ...... ii
Abstract ...... iii
Acknowledgements ...... xiv
Abbreviations ...... xvi
List of Figures ...... xx
List of Tables ...... xxii
Chapter One ...... 1
Literature review ...... 1
Prologue ...... 2
1.1 Introduction ...... 2
1.1.1 A history of a shared environment ...... 2
1.1.2 Domestication of dogs ...... 3
1.1.3 Dogs and zoonoses ...... 4
1.2 Rabies ...... 5
1.2.1 Definition and Aetiology ...... 5
1.2.2 Rhabdovirus: structure and characteristics...... 5
1.2.3 Transmission ...... 6
1.2.4 Incubation ...... 7
1.2.5 Epidemiology ...... 8
1.2.6 Diagnosis ...... 9
1.2.7 Prevention and control ...... 11
1.3 Burden of Rabies in India ...... 14
1.3.1 Mortality ...... 14
1.3.2 Economic burden of rabies in India ...... 15
1.3.3 Burden in DALYs ...... 16
v
1.3.4 Burden due to the loss of tourism ...... 17
1.4 Population dynamics of FRD in India ...... 19
1.4.1 Domestication of dogs in India compared with Western countries 19
1.4.2 Carrying capacity and estimates of the dog population in India .... 20
1.5 Population size estimation of dogs ...... 24
1.5.1 Enumeration of owned dogs...... 24
1.5.2 Enumeration of unowned or FRD ...... 25
1.6 FRD enumeration techniques ...... 26
1.6.1 Direct counts ...... 26
1.6.2 Capture – recapture studies ...... 28
1.6.3 Regression method ...... 30
1.6.4 Beck’s method ...... 31
1.6.5 Methods based on variations/modifications of capture-recapture procedures ...... 32
1.6.6 Method based on distances ...... 36
1.7 Dog demographics and rabies control ...... 39
1.8 Home ranges and Social behaviour of FRD ...... 42
1.9 Studies on Knowledge, attitudes and practices (KAP) of rural and urban communities on rabies, free roaming dogs (FRD) and their impact ...... 45
1.10 Strategies for the prevention and control of rabies in India ...... 48
1.11 Aims, scope and relevance of the present study ...... 49
1.12 Objectives of the current study ...... 51
1.14 The layout and format of this thesis ...... 51
Chapter Two ...... 54
A comparative study of enumeration techniques for free roaming dogs in rural Baramati, District Pune, India ...... 54
Preface ...... 55
Abstract ...... 59
vi
2.1 Introduction ...... 60
2.2 Materials and Methods ...... 62
2.2.1 Study area ...... 62
2.2.2 Field methodology ...... 62
2.2.3 Animal identification and capture histories ...... 64
2.2.4 Data analysis ...... 65
2.2.5 Population estimation methods ...... 65
2.2.6 Capture-recapture (C-R) techniques ...... 65
2.3 Ethical approval ...... 70
2.4 Results ...... 70
2.4.1 Sighting variability between sessions ...... 70
2.4.2 Test for equal catchability and closure of population ...... 71
2.4.3 Regression method ...... 71
2.4.4 Lincoln-Petersen Index (L-P) and Chapman’s corrected (C) estimator ...... 72
2.4.5 Beck’s method (Schnabel’s multi-capture method) ...... 72
2.4.6 Schumacher-Eschmeyer method ...... 73
2.4.7 Logit-normal mark-resight method ...... 73
2.4.8 Huggin’s heterogeneity models using Program CAPTURE ...... 73
2.4.9 Estimation using the Good-Turing frequency formula using AS tool ...... 73
2.5 Discussion ...... 74
2.5.1 Sighting variations ...... 77
2.5.2 Capture-Recapture (C-R) techniques ...... 77
2.5.3 Types of C-R estimation techniques ...... 78
2.5.4 Comparison of the methods used in this study...... 83
Author contributions ...... 89
vii
Funding ...... 89
Acknowledgements ...... 89
Conflict of Interest Statement ...... 90
Chapter Three ...... 95
Validation of Application SuperDuplicates (AS) enumeration tool for free roaming dogs (FRD) in urban settings of Panchkula Municipal Corporation in north India ...... 95
Preface ...... 96
Abstract ...... 100
3.1 Introduction ...... 101
3.2 Materials and Methods ...... 103
3.2.1 Study area ...... 103
3.2.2 Field methodology ...... 104
3.2.3 Data entry and analyses ...... 104
3.3 Results ...... 105
3.4 Discussion ...... 109
Author contributions ...... 113
Funding ...... 113
Ethical approval ...... 113
Acknowledgements ...... 113
Chapter Four ...... 120
Demographic characteristics of free roaming dogs in rural and urban India following a photographic sight-resight survey ...... 120
Preface ...... 121
Abstract ...... 124
4.1 Introduction ...... 125
4.2 Results ...... 127
viii
4.2.1 Sighting variability and the demographic characteristics of free roaming dogs ...... 127
4.2.2 Variations in the composition of the free-roaming dog population within the urban region ...... 134
4.3 Discussion ...... 136
4.3.1 Influence of temperature on FRD sightings ...... 136
4.3.2 Gender ratio ...... 137
4.3.3 Age composition of FRD ...... 138
4.3.4 Body condition and the sightings near garbage points ...... 139
4.3.5 Activity ...... 140
4.3.6 Free-roaming dog demography and relationships to ABC programmes ...... 141
4. 4 Materials and methods ...... 143
4.4.1 Study area ...... 143
4.4.2 Field methodology ...... 145
4.4.3 Data entry and analysis ...... 146
4.4.4 Ethical approval ...... 147
Author contributions ...... 147
Funding ...... 148
Acknowledgements ...... 148
Conflict of Interest Statement ...... 148
Chapter Five ...... 154
Utilising group-size and home-range characteristics of free roaming dogs (FRD) to guide mass vaccination campaigns against rabies in India ...... 154
Preface ...... 155
Abstract ...... 158
5.1 Introduction ...... 159
5.2 Materials and methods ...... 161
ix
5.2.1 Study area and field methodology ...... 161
5.2.2 Analysing distribution of different sized FRD groups across time and space ...... 161
5.2.3 Analyses of factors that influence the number of FRD sighted together ...... 163
5.2.4 Analysis of home-range ...... 163
5.2.5 Data entry, storage and cleaning and analyses ...... 164
5.3 Results ...... 165
5.3.1 Sightings of the FRD in rural and urban settings as solitary or in groups ...... 165
5.3.2 Determinants of number of FRD sighted together or alone ...... 167
5.3.3 Home-range of FRD and its determinants ...... 168
5.4 Discussion ...... 173
5.5 Ethical approval ...... 179
Author contributions ...... 179
Funding ...... 179
Acknowledgements ...... 180
Conflict of Interest Statement ...... 180
Chapter Six ...... 185
Knowledge, attitudes and practices (KAP) towards rabies and free roaming dogs (FRD) in Shirsuphal village in western India: A community based cross-sectional study ...... 185
Preface ...... 186
Abstract ...... 190
6.1 Introduction ...... 190
6.2 Materials and methods ...... 192
6.2.1 Sample size...... 192
6.2.2 Study area and survey procedure ...... 192
6.2.3 Questionnaire design ...... 193
x
6.2.4 Data management and analysis ...... 194
6.2.5 Univariable and multivariable analyses...... 195
6.3 Results ...... 196
6.3.1 Demographic and socio-demographic characteristics of the respondents ...... 196
6.3.2 Respondent’s knowledge, attitudes and practices regarding rabies ...... 196
6.3.3 Respondent’s attitudes and practices towards rabies...... 199
6.3.4 Respondent’s attitudes and practices towards free roaming dogs (FRD) ...... 201
6.3.5 Characteristics of dog owners ...... 204
6.4 Discussion ...... 205
Ethical approval ...... 213
Acknowledgments ...... 213
Chapter Seven ...... 221
Knowledge, Attitudes and Practices (KAP) towards rabies and Free Roaming Dogs (FRD) in Panchkula District of north India: A cross-sectional study of urban residents ...... 221
Preface ...... 222
Statement of Contribution ...... 223
Abstract ...... 225
7.1 Introduction ...... 225
7.2 Materials and Methods ...... 228
7.2.1 Study area ...... 228
7.2.2 Sample size...... 229
7.2.3 Sampling procedure ...... 230
7.2.4 Questionnaire design ...... 230
7.2.5 Data management and analysis ...... 231
7.3 Results ...... 232
xi
7.4 Discussion ...... 241
7.4.1 Community knowledge and awareness of rabies ...... 241
7.4.2 Community attitudes and practices towards rabies ...... 244
7.4.3 Community attitudes and practices towards free roaming dogs .. 246
7.4.4 Characteristics of urban dog owners ...... 246
Ethical approval ...... 248
Acknowledgements ...... 249
Chapter Eight ...... 256
Knowledge, attitudes and practices towards dog-bite related rabies in para-medical staff at rural primary health centres in Baramati, western India ...... 256
Preface ...... 257
Abstract ...... 260
8.1 Introduction ...... 260
8.2 Materials and Methods ...... 263
8.2.1 Study area, sampling procedure and sample size ...... 263
8.2.2 Questionnaire design ...... 263
8.2.3 Data management and analysis ...... 264
8.3 Results ...... 265
8.4 Discussion ...... 271
Author contributions ...... 276
Funding ...... 276
Acknowledgements ...... 276
Conflict of Interest Statement ...... 276
Chapter Nine ...... 277
General Discussion ...... 277
Prologue ...... 278
9.1 Estimation of the population size of free roaming dogs ...... 279
xii
9.2 Free roaming dog demography, group behaviour and home-ranges ...... 281
9.3 Knowledge, attitudes and practices (KAP) of the people towards rabies and FRD ...... 284
9.4 Responsible ownership of dogs ...... 286
9.5 Paramedical staff and rabies control ...... 288
9.6 Control of dog-mediated rabies in India ...... 289
9.7 Conclusions ...... 293
Epilogue ...... 294
References ...... 295
Appendices ...... 326
Appendix I ...... 327
Appendix II ...... 333
Appendix III ...... 335
Appendix IV ...... 336
xiii
Acknowledgements
Heartfelt gratitude to my parents and teachers, who nurtured, guided and bestowed upon me the confidence to pursue the highest level of learning in education.
Sincere gratefulness to Emeritus Professor Ian Robertson, my Principal Supervisor, who risked accepting me as his pupil well aware that I was returning to academics after 15 years of drill practice in Indian Army, who patiently bore my incompetence resulting from long hiatus from academics, who painstakingly went through my writings and provided valuable insights, who offered succour to my academic woes, who encouraged and instilled confidence to my sometimes-sagging optimism, who helped me finish this thesis, who changed the course of my life.
Genuine thanks to Dr Abi Tamim Vanak, my adjunct Supervisor, who helped with man and material resources during the field study, who constantly challenged my critical thinking to bring out the best, provided valuable comments on the manuscript of the articles that were published, without whose involvement this research would not have been conducted and without whom I was a boat without oars.
Profound gratitude to Dr Mark O’Dea, my Co-supervisor, who was ever encouraging of my research pursuits, who was always approachable to render advice, who was ever so quick with the turn arounds of manuscripts edited with invaluable suggestions, and who supported me beyond the realms of the researcher-supervisor relationship during the course of this arduous journey.
This study wouldn’t have reached its fruition had it not been for the support I received from Dr Mieghan Bruce, who in spite of her busy schedule spared time to discuss, suggest, support and encourage.
xiv
I am indebted to my friend and a very noble human being, Dr Rajinder Singh Bajwa for the selfless help during my study in Panchkula.
Many thanks to the Abhijit, Pranav, Reetika, Pradeep, students of Government PG
College and Government Girls College, Panchkula, the Village Panchayat of Shirsuphal, and Municipal Corporation, Panchkula for their help in conducting the field studies.
Love and thanks to Karma, Jiangyong, Zixian, Petrus, Tuyet, Yibo, Elvina, Telleasha,
Arash, Mohnad, Ali, Ali Zahedi, Emad, Sylvie, Jessica, David, Idy, and the galaxy of friends who shared the journey or part thereof with me, who exchanged notes while sharing the coffee at the library, who shared a laugh when chips were down, who helped me learn some tricks of the trade passing ‘R’ codes, ArcGIS tips and Excel shortcuts, who are now indelible imprints in my memory of the time spent at the University.
Lexicon fails to deliver appropriate words to thank my classmate, best friend, valued critic, my soulmate, my wife, Dr Jully Gogoi Tiwari, who stood like a rock, shouldered the familial responsibilities without complaining ever, and believing in me. Jully, you are the reason I took it up; you are the reason I finished it.
The list will remain incomplete without the mention of my beautiful, talented and loving daughter, who bore my absence during the field work, who helped me better my syntaxes and most importantly who tolerated my idiosyncrasies. Thank you, Nishtha!
Heartfelt thanks to Ausvet family, who during my association with them for last one year never let me lose sight of mission PhD.
Finally, I am grateful to Murdoch University for the Murdoch International Postgraduate
Scholarship without whose financial help this feat was impossible.
xv
Abbreviations
AAEC ATREE Animal Ethics Committee
ABC Animal Birth Control
AIC Akaike's Information Criteria
ARV Antirabies Vaccine
AS Application SuperDuplicates
ATREE Ashoka Trust for research in Ecology and the Environment
BCS Body Condition Score
CCV Chicken Cell Vaccine
CDC Centre for Disease Control
CI Confidence Interval
CNVR Capture Neuter Vaccinate Release
C-R Capture-Recapture Techniques
CSF Cerebrospinal Fluid
CV Coefficient of Variation
DALY Disability Adjusted Life Years
xvi
ELISA Enzyme Linked Immunosorbent Assay
EC Estimate by Chapman’s correction
EL-P Estimate by Lincoln – Petersen f1 Number of singletons
FAT Fluorescent Antibody Test
FRD Free Roaming Dog
GNM General Nursing Midwives
GPS Global Positioning System
ICEID International Conference on Emerging Infectious Diseases
IEC Information, Education and Communication
ISVEE International Symposium of Veterinary Epidemiology and Economics
KAP Knowledge Attitudes and Practices
LHV Ladies Health Visitor
Mo Huggins closed capture model with null factor
Mh Huggins closed capture model with heterogeneity factor
Mt Huggins closed capture model with temporal factor
Mth Huggins closed capture model with heterogeneity and temporal factor
xvii
MCP Minimum Convex Polygon
MIT Mouse Inoculation Test
MLE Maximum Likelihood Estimate
MPW Multi-purpose Worker
OJT On Job Trained
OR Odds Ratio
ORV Oral rabies Vaccine
PCECV Purified Chicken Embryo Cell Vaccine
PDEV Purified Duck Embryo Vaccine
PEP Post Exposure Prophylaxis
PHC Public Health Centre
Q1 Number of uniques
RFFIT Rapid Fluorescent Focus Inhibition Test
RIG Rabies Immunoglobulins
RNA Ribo Nucleic Acid
RREI Rapid Rabies Enzyme Immuno-diagnosis
RTCIT Rapid Tissue Culture Inhibition Test
xviii
SECR Spatially Explicit Capture Recapture
Sobs Single observations
WHO World Health Organisation
WSPA World Society for Protection of Animals
YLD Years of Life lived in Disability
YLL Years of Life Lost
ZTPD Zero Truncated Poisson Distribution
xix
List of Figures
Chapter One
Figure 1.1 A matrix to categorise a dog’s dependency on humans with respect to restrictions placed on their movement (Cliquet and Barrat 2012) ...... 22
Chapter Two
Figure 2.1 Google earth imagery (www.googleearth.com) of the village landscape and the various tracks used by the observation teams for survey (accessed on 22/07/2016).64
Figure 2.2 Prediction of population size of free roaming dogs by regression method for all sessions of FRD survey in Shirsuphal ...... 72
Figure 2.3 Graphical representation of the trend of population estimates using Huggin’s models and Application SuperDuplicates (AS) with the number of survey sessions ..... 87
Supplementary Figure 2.1 Graphical representation of estimates of free roaming dog population size in Shirsuphal village by different methods ...... 91
Supplementary Figure 2.2 The devices used and a glimpse of the field observation process ...... 91
Supplementary Figure 2.3 Six examples (A-E) of descriptions of free roaming dogs as recorded by the observer team...... 92
Supplementary Figure 2.4 The sightings of the free roaming dogs and garbage points in Shirsuphal village during the dog enumeration survey ...... 93
Supplementary Figure 2.5 Testing the fit of the Schumacher-Eschmeyer method by plotting number of previously marked FRD against proportion of marked FRD sighted during each survey effort during the enumeration survey in Shirsuphal ...... 94
Chapter Three
Supplementary Figure 3.1 Geographical location and surveyed sectors of Panchkula that were selected for the enumeration survey during September-October 2016...... 117
Supplementary Figure 3.2 Flow chart of the sampling strategy for selection of the survey- tracks in 15 sectors of the Municipal Corporation, Panchkula for the free roaming dog enumeration survey during September-October 2016 ...... 117
xx
Supplementary Figure 3.3 Comparative estimates of the free roaming dog population size obtained by Huggin’s closed capture heterogeneity model Mh-JK (after 5 surveys) and Application SuperDuplicates (after 2 surveys) ...... 118
Supplementary Figure 3.4 Comparative estimates of the free roaming dog population size obtained by Huggin’s closed capture heterogeneity model Mth-Chao (after 5 surveys) and Application SuperDuplicates (after 2 surveys) ...... 119
Chapter Five
Supplementary Figure 5.1 Boxplots for the univariable analyses of regression of the various intrinsic and extrinsic factors on the group size of free roaming dogs sighted during the enumeration survey in Shirsuphal village of western India in June 2016 ... 181
Supplementary Figure 5.2 Boxplots for the univariable analyses of regression of the various intrinsic factors on the group size of free roaming dogs sighted during the enumeration survey in Municipal Corporation Panchkula in north India during September – October 2016 ...... 182
Supplementary Figure 5.3 Boxplot graphics for the univariable analyses of regression of the various extrinsic factors on the group size of free roaming dogs sighted during the enumeration survey in rural and urban setting ...... 183
Supplementary Figure 5.4 Boxplots for the univariable analyses of regression of resight probability on the group size of free roaming dogs sighted during the enumeration survey in rural and urban settings ...... 184
Supplementary Figure 5.5 Boxplots for the univariable analyses of de-sexed status probability on the group size of free roaming dogs sighted during the enumeration surveys in the urban setting in Panchkula ...... 184
Chapter Seven
Figure 7.1 Study area in Panchkula Municipal Corporation, Haryana state, India with the number of households interviewed from each ward (total households interviewed = 204) ...... 229
xxi
List of Tables
Chapter One
Table 1.1 List of scholarly articles on knowledge, attitudes and practices of urban and rural communities towards dog - bites and rabies in India published during 2011-2015 ...... 48
Chapter Two
Table 2.1 Details of climatic characteristics and the number of free roaming dogs sighted at each survey session during the photographic capture-recapture survey at Shirsuphal Village during 5 - 13 June 2016...... 71
Table 2.2 Size of the free roaming dog population estimated by the Lincoln–Petersen index and Chapman’s correction (EC) with counts on successive days during photographic capture-recapture survey in Shirsuphal ...... 75
Table 2.3 Comparison of the models run using the Logit-normal mark-resight method on the basis of the Akaike Information Criteria (AIC) for FRD enumeration in Shirsuphal ...... 76
Table 2.4 Population estimates and calculated capture probability as obtained by available estimators* under Program CAPTURE for FRD at Shirsuphal ...... 76
Table 2.5 Population estimates of free roaming dogs using Application SuperDuplicates for sampling occasions ranging from 2 to 7 ...... 79
Table 2.6 Summary of the population estimates obtained using 11 different methods until saturation (7 survey occasions spread over 9 days) ...... 80
Table 2.7 The population estimates obtained by the Huggin’s heterogeneity models compared with Application SuperDuplicates (AS) online tool based on Good-Turing frequency formula on successive reduction of sampling efforts ...... 86
Chapter Three
Table 3.1 Results of Leslie’s test for equal catchability of free roaming dogs in the different sectors of Panchkula during the enumeration survey carried out during September-October 2016 ...... 106
Table 3.2 The Coefficient of Variation, population coverage and corresponding difference in the estimate by Maximum Likelihood and Application SuperDuplicates shinyapp tool during the free roaming dog enumeration surveys in 14 survey tracks in Panchkula... 107
xxii
Table 3.3 Estimates of the free roaming dog population size by Maximum Likelihood (5- 6 surveys) and Application SuperDuplicates shinyapp (first 2 surveys) using sight-resight techniques in different sectors of Panchkula following enumeration surveys carried out during September-October 2016 ...... 108
Supplementary Table 3.1 Details of the number of free roaming dogs sighted during the sight-resight surveys conducted in fourteen sectors of Panchkula Municipal Corporation administrated areas of Haryana state, India along with the meteorological data for each survey session...... 115
Chapter Four
Table 4.1 Demographic details (gender, age distribution and body condition) of the free roaming dogs sighted on each survey occasion during the enumeration survey (7 occasions) in the rural survey (Shirsuphal village, Baramati, Pune)...... 128
Table 4.2 Details of the activity and sightings within 20 m of garbage of free roaming dogs sighted on each survey occasion during the rural survey (Shirsuphal village, Baramati, India)...... 129
Table 4.3 Demographic details (gender, age distribution, neutering status and body condition), their respective distribution for the FRD sighted in the different sectors of the urban location (Panchkula Municipal Corporation administrated sectors)...... 131
Table 4.4 Details of activity and sightings within 20 m of garbage point for free roaming dogs sighted across 14 sectors of the urban location (Panchkula Municipal Corporation administrated area)...... 132
Supplementary Table 4.1 Details of the number of free roaming dogs sighted during the sight-resight surveys undertaken in Shirsuphal village and fourteen survey tracks of Panchkula Municipal Corporation along with the meteorological data for each survey session...... 149
Supplementary Table 4.2 Details of the track lengths, duration of the photographic sight- resight surveys, and population estimates of free roaming dogs during the enumeration surveys carried out in Shirsuphal village and Panchkula Municipal Corporation administered sectors ...... 151
Supplementary Table 4.3 Comparative analyses of the characteristics of free roaming dogs from the rural (Shirsuphal village) and urban (Panchkula Municipal Corporation administrated area) locations...... 152
Supplementary Table 4.4 The characteristics (gender, age, body condition and reproductive status) of free roaming dogs in the different localities of the urban survey (Municipal Corporation, Panchkula) presented as odd ratios (OR) and their 95% confidence intervals (CI)...... 153
xxiii
Chapter Five
Table 5.1 The number and likelihood of free roaming dogs being sighted alone or with other dogs during the photographic sight-resight survey carried out in Shirsuphal (rural) and selected sectors of Panchkula (urban), India ...... 165
Table 5.2 The odds of various categories of FRD being sighted in groups (≥ 2) in rural (Shirsuphal) and urban (Panchkula) settings...... 169
Table 5.3 The coefficients, standard error and p values of the univariable generalised linear mixed models for the regression of predictor variables on the actual group size of FRD* in rural Shirsuphal# and urban Panchkula† in India ...... 170
Table 5.4 Final multivariable model of predictors that significantly influenced the number of FRD sighted together (1 to 5) in a rural setting* ...... 171
Table 5.5 Final multivariable model of predictors that significantly influenced the number of FRD sighted together (1 to 12) in an urban setting*...... 171
Table 5.6 Test of association between dichotomised home-range* and predictor variables for the rural (Shirsuphal) and urban (Panchkula) locations...... 172
Table 5.7 Final multivariable logistic regression model with regression coefficient values for predictors yielding significant p- values for dichotomised home-range area for free roaming dogs in rural Shirsuphal, India ...... 173
Chapter Six
Table 6.1 Demographic characteristics of the respondents (n=127) of the Knowledge, attitudes and practices survey in the Shirsuphal village ...... 197
Table 6.2 Association of the knowledge of the participants about rabies with various descriptive variables ...... 198
Table 6.3 Final multivariable logistic regression model of factors associated with respondent’s knowledge of rabies ...... 199
Table 6.4 Association of the attitudes and practices of the respondent’s about rabies with various descriptive variables ...... 200
Table 6.5 Responses to various questions pertaining to attitudes and practices regarding free roaming dogs ...... 202
Table 6.6 Association of the attitudes and practices of the community regarding free roaming dogs with various descriptive variables ...... 203
Table 6.7 Final multivariable logistic regression model of factors associated with respondent’s attitudes and practices towards free roaming dogs ...... 204
xxiv
Table 6.8 Characteristics of the owned dogs and owner’s perceptions and practices about their pets ...... 205
Supplementary Table 6.1 The matrix developed to categorise the respondents into high, middle and low socio-economic groups (www.praja.org) ...... 214
Supplementary Table 6.2 Bivariate analyses of responses to the questions pertaining to knowledge of rabies for various descriptive determinants * ...... 215
Supplementary Table 6.3 Bivariate analyses of responses to the questions pertaining to attitudes and practices regarding rabies for various descriptive determinants ...... 217
Supplementary Table 6.4 Bivariate analyses of the response by various categories of respondents to the questions regarding their attitudes and practices against free roaming dogs ...... 219
Chapter Seven
Table 7.1 Demographic characteristics of respondents in Panchkula, India, 2016...... 232
Table 7.2 Test of association (χ2) between knowledge about rabies and various predictor variables in Panchkula, India, 2016 ...... 234
Table 7.3 Multivariable logistic regression model of factors associated with the participants’ knowledge of rabies in Panchkula, India, 2016 ...... 234
Table 7.4 Test of association (χ2) between the respondents’ attitudes and practices towards better control and prevention of rabies and their attitudes towards free-roaming dogs, and various predictor variables in Panchkula, India, 2016 ...... 236
Table 7.5 Multivariable logistic regression model of factors associated with the respondents’ attitudes and practices towards rabies control and prevention in Panchkula, India, 2016 ...... 237
Table 7.6 Respondents’ responses to various questions pertaining to attitudes and practices relevant to free roaming dogs in Panchkula, India, 2016 ...... 239
Table 7.7 Characteristics of owned pet dogs and the owner’s perceptions and practices about their pets in Panchkula, India, 2016 ...... 240
Supplementary Table 7.1 Descriptive and bivariate analyses (χ2) of the responses to the individual questions relating to knowledge about rabies amongst various predictor variables in the residents of Panchkula Municipal Corporation ...... 250
Supplementary Table 7.2 Descriptive and bivariate analyses (χ2) of the responses to the individual questions related to attitudes and practices about rabies amongst various predictor variables in the residents of Panchkula Municipal Corporation ...... 253
xxv
Supplementary Table 7.3 Descriptive and bivariate analyses (χ2) of the responses to the questions relating to attitudes and practices towards free roaming dogs amongst various predictor variables in the residents of Panchkula Municipal Corporation ...... 254
Chapter Eight
Table 8.1 List of Primary Health Centres and Sub-centres around Baramati and the number of staff interviewed along with their positions ...... 265
Table 8.2 Bivariate analyses of the individual questions pertaining to knowledge about rabies for health workers belonging to different categories (n=54) ...... 267
Table 8.3 Bivariate analyses of the individual questions pertaining to practices of health workers towards management of dog-bite wounds belonging to different categories (n=54) that help control rabies ...... 269
Table 8.4 Test of association (χ2) of the independent variables (experience, education and appointment) with the dependent variables (knowledge about rabies and practices regarding management of dog-bite patients)...... 270
Table 8.5 Final multivariable model showing the influence of various independent factors over the knowledge about rabies and practices pertaining to rabies that help its control ...... 271
xxvi
Chapter One
Literature review
“…the most lavish prophylaxis against hydrophobia in the hunting hounds was carried out, fittingly, by the kings of France. In the hunting accounts of the French palace, historians have found annual outlays for all the king’s hounds to undergo a special ceremony. They were transported to the church of St Menier les Moret, in order to have a mass sung in the presence of said hounds, and to offer candles in their sight, for fear of the “mal de rage”- that is the disease of rabies. One wonders whether the hounds howled along…”
Bill Wasik
in
“Rabid: A cultural history of the world’s most diabolical virus”
1
Prologue
The origin of rabies as a disease is arguably as old as the first encounter of humans with carnivores. While it can be contested that these encounters were only accidents at the dawn of prehistoric times, their frequency would have multiplied manyfold as humans started domesticating dogs. Having existed for an eon, rabies has been a subject of interest for academia and researchers resulting in a plethora of available scholastic literature on its various aspects from earliest to modern times. It is not only a very old disease; but also has a history of emerging in geographical areas where it was never present, as well as remerging in places where it was controlled or eradicated from. In accordance with the aims of the research work outlined in this thesis, which focuses on measures leading to the control of rabies in the developing world, this literature review highlights relevant domains of the disease. While various aspects of rabies research are reported in this review, an attempt has been made to focus on vector (free roaming dogs - FRD) population demographics and dynamics, studies on awareness levels of affected human populations and health workers, along with associated factors that contribute towards the failure to control the disease in India.
1.1 Introduction
1.1.1 A history of a shared environment
Canis lupus familiaris, the domestic dog, has shared a common environment with humans for many years (Clutton-Brock 1999), with archaeological findings of ritualistic burials of dogs and dog figurines providing evidence that they have been domesticated for more than 10,000 years (Morey 2006, Somvanshi 2006, Udell and Wynne 2008, Majumder,
Chatterjee, and Bhadra 2014). Juliet Clutton-Brock (1995) in the book “The Domestic
2
Dog: Its Evolution, Behaviour and Interactions with People” reasoned that early humans occasionally reared captured wolf pups. The domesticated dog evolved from such tamed wolves after the breeding out of their wild temperament through selection over multiple generations. However in contrast, Hiby (2013) argued that the association started with the less aggressive wolves accompanying humans due to the scavenging benefits arising from this association. Irrespective of the original method of association, humans have had a profound impact on dogs as a species, resulting in selective breeding for a wide range of behavioural traits and physical attributes. In particular their acute senses of smell and hearing has enabled humans to employ them for multiple roles in society (Hiby 2013). It has been demonstrated that dogs possess a natural ability and willingness to help humans
(Bräuer, Schönefeld, and Call 2013), and even today in times of tremendous technological advances they are being gainfully used as rescue dogs, guide dogs, police dogs and drug, mines, termite and cancer detecting dogs (Udell and Wynne 2008). This long history of cohabitation with humans has enabled dogs to adapt to living with humans, either in their homes or as free ranging dogs in the vicinity of human settlements (Majumder, Chatterjee, and Bhadra 2014).
1.1.2 Domestication of dogs
The attitudes of people towards dogs vary, with communities treating dogs differently according to the roles they play in their lives (Hiby 2013). The socio-cultural norms, income status and level of literacy of the human community determines the density of the dog population and the extent of their contact with humans within an area (Otolorin,
Umoh, and Dzikwi 2014). Religious beliefs also affect the scope of human-dog interactions (Aiyedun and Olugasa 2012, Macpherson 2012). In developed countries the domestication process has seen dogs become pets, where the owner is totally responsible
3
for their welfare, but in developing countries, such as India, their welfare has not been responsibly shared (Majumder, Chatterjee, and Bhadra 2014). In rural India, dogs are usually associated with several households that provide food and shelter, but no single individual or household is solely responsible for them (Dutta 2002, Wandeler et al. 1988).
While the Western influence has led to the introduction of pedigree dogs to many homes in the urban Indian society, the indigenous Indian native dog thrives on the streets, living on garbage and food provided by households (Majumder, Chatterjee, and Bhadra 2014).
1.1.3 Dogs and zoonoses
An ecological imbalance has been created in the developing countries between humans and dogs by unchecked breeding of stray dogs and consequently increased numbers on the streets (Krishna 2009). This indiscriminate increase in the dog population in proximity to humans in both rural and urban areas has resulted in many public health concerns
(Otranto, Dantas-Torres, and Breitschwerdt 2009). Stray dogs form packs which have an adverse effect on the community through their scavenging behaviour and aggression towards humans (Beck 1975). As a competent reservoir host of many zoonotic agents, dogs are also a potential threat to public health and animal welfare (Otranto, Dantas-
Torres, and Breitschwerdt 2009). As carriers of pathogens such as rabies virus, parvovirus, adenovirus, distemper virus and Leptospira spp., stray dogs are not only harming other domestic animals (Vanak and Gompper 2009b) but have also been documented to cause declines of wildlife populations through the transmission of pathogens (Belsare and Gompper 2013). Many zoonotic diseases find their niche in countries where a large population of stray dogs share a common environment with constantly enlarging densities of human populations (Traub et al. 2005). Of all the canine zoonotic diseases, rabies has had the greatest impact in developing countries, both in
4
terms of human deaths and economic burden (Ratsitorahina et al. 2009). Rabies has been known to exist in India for at least 4,000 years, as indicated by its reference in the earliest
Vedic texts (Suraweera, Morris, Kumar, Warrell, Warrell, and Jha 2012, Menezes 2008).
The principle aim of this study is to assess the demographic characteristics of the FRD population in India, which is largely responsible for the high prevalence of rabies in the country. However, before studies relating to dog population are discussed, it is pertinent to provide information about rabies: its cause; modes of transmission; epidemiology; and preventive, control and eradication methods adopted.
1.2 Rabies
1.2.1 Definition and Aetiology
The second edition of the WHO expert consultation on rabies held in 2013 defined the disease as “an acute encephalitis or meningoencephalitis due to a lyssavirus infection”
(WHO 2013). Rabies virus is one of the 16 identified variants of lyssaviruses, all of which affect mammalian hosts and are zoonotic (Crowcroft and Thampi 2015). They belong to the Genus Lyssavirus, Family Rhabdoviridae, order Mononegavirales.
1.2.2 Rhabdovirus: structure and characteristics
The genome of the rabies virus is an unsegmented negative strand RNA, encoding for five viral proteins: a nucleoprotein (N); a phosphoprotein (P); a matrix protein (M); a glycoprotein (G); and an RNA dependent RNA polymerase (L) (Baron and Rupprecht
1996). The virus comprises of two units, an internal helical nucleocapsid and an external envelope. The nucleocapsid consists of a ribonucleoprotein complex comprising the genomic RNA and tightly bound N protein, together with the L and P proteins. The ribonucleoprotein unit is active for transcription and replication. The N-RNA template is
5
processed by the L protein, which contains most of the RNA polymerase activities, with its cofactor, the P protein. The lipid envelope is derived from the host cytoplasmic membrane. The glycoprotein (G) spikes bind the virion to host cell receptors. The M protein is comprised of oligomers that bind to the outside of the nucleocapsid, giving rigidity to the virion structure and providing a binding platform for the viral glycoprotein and the host envelope membrane (Ge et al. 2010).
The biochemical and physical properties of ‘fixed' rabies viruses were described by
Turner and Kaplan (1967). The ‘fixed’ rabies viruses are the laboratory passaged attenuated viruses with fixed incubation periods in contrast to the ‘street’ rabies viruses that are isolated from naturally infected animals and have unpredictable incubation periods, varying from 2 weeks to 1 year (Gupta et al. 2005). It was demonstrated that the fixed virus is inactivated by extreme pH (< 5.0 and >10.0). Being highly thermo-sensitive, the virus remains stable only up to 2 hours and loses 90% of its infectivity after 4 hours at 370 C. The virus is inactivated in less than 10 minutes at 600 C. It is also sensitive to the commonly used organic solvents, such as methanol, ethanol and acetone, being inactivated in less than 10 min at 25% concentration (Turner and Kaplan 1967). The virus is also inactivated by exposure to UV light and detergents (Wandeler and Bingham 2013).
1.2.3 Transmission
Rabies viruses have a wide host range covering all mammals. The common hosts include bats, foxes, raccoons and all feral canids, but rabid dogs are known to be the greatest hazard to humans (Rupprecht, Hanlon, and Hemachudha 2002). The usual natural route of transmission occurs through a bite from a reservoir animal (Dietzschold et al. 2008).
Although the viruses cannot cross intact skin, they may gain ingress through open wounds or direct exposure to mucous membranes (Crowcroft and Thampi 2015). Other reported
6
possible routes of infection include consumption of affected carcasses by arctic foxes
(Vulpes lagopus) or through aerosolised rabies viruses accidently released in a laboratory setting (Crandell 1991). Aerosol transmission may also occur in caves with large population of bats (Dietzschold et al. 2008). Humans are generally dead-end hosts, although the viruses may be present in body fluids, such as the lacrimal gland and tracheal secretions, and tissues such as the pharynx, skeletal muscles, liver, kidney and skin during the first five weeks of illness (Helmick, Tauxe, and Vernon 1987). There are also four reported cases of human-to-human transmission through corneal transplant recipients
(Crowcroft and Thampi 2015, Helmick, Tauxe, and Vernon 1987). Another case of human-to-human transmission was reported from Texas in 2004 when four recipients of organs from a common donor died of rabies (Srinivasan et al. 2005). The donor had died from a subarachnoid haemorrhage; however, it was later discovered that he had previously been bitten by a bat.
1.2.4 Incubation
Rabies virus has a more variable incubation period than any other infection (Rupprecht
1996), varying from five days to several years. The incubation period is influenced by three factors: the amount of virus in the inoculum; the nerve motor endplate density at the wound site; and the proximity of the virus entry to the central nervous system (Ugolini
2007, Hemachudha et al. 2013). In one study the mean incubation time was found to be
273.6 days (median 80 days, range 12 days to 10 years) (Carrara et al. 2013). After introduction the virus may or may not replicate at the entry site, but enters the peripheral nervous system during the incubation period (Rupprecht 1996). On entering the unmyelinated axon terminals, virus is transported into the nerve cell centripetally in a retrograde manner. The virus replicates in the nerve cell and reaches the brain after
7
spreading through the chain of neurons connected by synaptic junctions. After infecting the central nervous system the virus moves centrifugally to the peripheral and autonomic nervous systems and finally reaches the salivary glands where it can be shed (Dietzschold et al. 2008). The speed of the migration depends on the mode of propagation as centripetal retrograde movement is faster than the centrifugal mode (Hemachudha et al. 2013).
1.2.5 Epidemiology
Rabies in humans follows two different epidemiological cycles. One is the urban cycle which requires dog populations to maintain the infection and the other, the sylvatic form, involves wildlife. There are documented evidence of the virus moving from dogs to wildlife and vice-versa (Gongal and Wright 2011). In a presentation on “Global perspective of rabies,” during the Global Conference on Rabies Control in Incheon-Seoul,
Republic of Korea (7-9 September 2011), an eminent scientist in the field, A.I. Wandeler
(2011), explained that classical rabies virus was the most important of all the lyssaviruses that could infect a multitude of hosts (all warm blooded animals) with the disease not following any one epidemiological pattern. The disease has been eliminated from large parts of Europe and North America through immunisation of the principal hosts, such as dogs and foxes, along with population control of wildlife reservoirs such as raccoon dogs, wolves and skunks (Finnegan et al. 2002), however it still has a dominating presence in many developing countries (Wandeler 2011). A number of Pacific Island nations and
Australia have never reported the presence of carnivore rabies, while a few countries in
Asia, such as Japan and Malaysia, have eradicated the disease (Mani and Madhusudana
2013). Most of the human deaths due to rabies are reported from Asia and Africa, and the majority of these are due to the transmission of variants of dog rabies virus (Nagarajan et al. 2006).
8
1.2.6 Diagnosis
Although the focus of this thesis is on the epidemiologic aspects of rabies, a brief summary of the tools used for the laboratory diagnosis of the disease will be made as they contribute to the effectiveness of surveillance for the disease. However, countries where rabies is endemic, such as India, often lack robust laboratory diagnostic capabilities and surveillance infrastructure and consequently the disease burden and its impact on public health remains underestimated (Mani and Madhusudana 2013). Knowledge of the rabies status of an animal that bites a human can assist physicians on deciding whether or not to use post-bite prophylaxis, resulting in a major saving to the economy from overuse of this form of prophylaxis. However, generally the only tools available in such situations are an assessment of the presenting clinical signs and symptoms, along with a history of an animal bite (Acharya, Kaur, and Lakra 2012). The common symptoms of dog-bite transmitted rabies include hydrophobia, encephalopathy and aerophobia (Udow, Marrie, and Jackson 2013). Clinically suspected cases of rabies, however, can be confirmed by isolation of the virus, or the detection of anti-rabies antibody, viral protein or RNA. Brief information on the main diagnostic tests available is provided below.
(a) Histological identification
Histological detection of intracytoplasmic inclusion bodies, “Negri-bodies”, composed of aggregates of viral particles present in infected neuronal cells of the affected brain tissue in impression smears (Seller’s technique) is considered to be a confirmatory diagnosis.
This technique, however, is suitable only for fresh specimens and generally has a low sensitivity (Mani and Madhusudana 2013).
9
(b) Demonstration of viral antigens
The Fluorescent Antibody Technique (FAT) and Rapid Rabies Enzyme Immuno- diagnosis (RREI), which detect viral antigens in specimens, are also regarded as confirmatory tests for rabies. The FAT is the gold standard for rabies diagnosis post- mortem where the rabies virus nucleoprotein antigen in fresh brain smears is demonstrated using immunofluorescent techniques (Dean and Abelseth 1973). However, as it requires an expensive fluorescent microscope and skilled personnel, its use is restricted in developing countries (Mani and Madhusudana 2013). The RREI is an ELISA based test which captures rabies N protein (Perring, Rolling, and Sureau 1986) and has been reported to be as sensitive and specific as the FAT. A major shortcoming of both of these tests is the need to use brain tissue, thus precluding their use ante-mortem.
A significant development in the field of the diagngosis of rabies has been the development of the Direct Rapid Immuno-histochemical test (dRIT). This test is not only cost effective but also easy to interpret and uses ordinary light microscopy (Reeta Mani and Madhusudana 2013, Niezgoda and Rupprecht 2006). It was developed by the Centre for Disease Control (CDC), Atlanta, USA, and detects rabies N protein in specimens.
(c) Virus isolation
Virus isolation can be performed by the Mouse Inoculation Test (MIT) and the Rapid
Tissue Culture Infection Test (RTCIT). While the MIT, if positive, may be beneficial to harvest large quantities of the infective virus for strain identification and can be performed in situations where skills and facilities for cell culture are not available, it is a time- consuming test (as long as 28 days) and requires live animals. On the other hand, although
RTCIT can be performed only in laboratories with cell culture facilities and a fluorescent
10
microscope, it is faster and has a higher specificity and sensitivity than the MIT (Mani and Madhusudana 2013, Madhusudana, Sundaramoorthy, and Ullas 2010).
(d) Demonstration of Antibodies
The presence of antibodies in serum or cerebrospinal fluid (CSF) offers indirect evidence of rabies infection, even though there may be a varied pattern of host immune responses.
Serological testing is seldom used for ante-mortem diagnosis due to seroconversion only occurring late in the disease and the high mortality of the disease. The Rapid Fluorescent
Focus Inhibition test (RFFIT) is considered a gold standard for detecting antibodies in
CSF; however, the test requires skilled personnel and facilities for cell culture and fluorescent microscopy. This test also has the disadvantage that it can yield false positive results (Mani and Madhusudana 2013).
However in India few centres have rabies diagnostic facilities and this lack of diagnostic laboratories is a serious impediment for disease surveillance (Acharya, Kaur, and Lakra
2012).
1.2.7 Prevention and control
Rabies is one of the earliest diseases for which a vaccination was introduced and recently there have been significant advances in the development of rabies vaccine from ‘Pasteur- treatment’ to Cell Culture Vaccines (CCV)/Purified Chicken Embryo Cell Vaccines
(PCECV) and Purified Duck Embryo vaccines (PDEV) (Briggs 2012). The first mass vaccination of dogs was reportedly undertaken in Japan in 1920 (Briggs 2012).
Vaccination of people against rabies involves pre-exposure and post-exposure schedules.
Pre-exposure vaccinations are recommended for individuals who are at a higher risk of infection, such as wildlife professionals and veterinarians dealing with stray dog populations. Post-exposure vaccination is administered to individuals after they have had
11
potential exposure to a rabid animal (Briggs 2012). A long incubation period associated with rabies infection provides a window for post-exposure prophylaxis. This comprises washing and adequate flushing of the bite wound, followed by administration of a series of post-bite vaccines and infiltration of Rabies Immunoglobulins (RIG) into/around the site of the bite, immediately after the exposure (WHO 2018a).
Human mortality arising from dog-related rabies can be minimised through the following strategies: firstly by administering Post-exposure Prophylaxis (PEP) to exposed individuals; and secondly by vaccinating a sufficient number of dogs to interrupt the transmission cycle; or by a combination of both of these strategies (Taylor and Nel 2015).
Dogs are the largest reservoir host for rabies virus and require immunisation of at least
70% of the population to effectively break the transmission cycle of the rabies virus leading to elimination of canine rabies (Coleman and Dye 1996, Franka et al. 2013). It has also been suggested that owing to a low basic reproduction number (R0) for rabies, a vaccination coverage of even 35% may be sufficient to eliminate the disease (Fitzpatrick et al. 2016). Nonetheless, besides the level of immunisation cover for FRD against rabies, the accessibility of the dogs for vaccination and routes of administration of the vaccines constitute crucial factors that require deliberation.
Parenteral vaccination, when undertaken effectively (ruling out vaccine failures due to faulty inoculation or a break in the cold chain process), induces a robust immune response and is the first choice for vaccination of FRD (WHO 2018). However accessibility of
FRD for vaccination is difficult due to the challenge of restraining them, which can result in the handlers being bitten or injured, or the dogs themselves suffering an injury (Cliquet et al. 2007). Thus, the target of achieving 70% “herd-immunity” remains elusive in many areas that have large FRD populations. The latest report of the WHO consultations on
12
rabies advocates the use of oral rabies vaccinations (ORV) to complement parenteral coverage (WHO 2018). Oral rabies vaccination has contributed to the successful elimination of fox-rabies from Europe (Freuling et al. 2013), although its efficacy to evoke adequate immune response in FRD is not assured owing to immunological and delivery concerns. Besides the selection of appropriate baiting requirements, the dogs are required to chew the sachet/blister containing the vaccine for the latter to be correctly deposited into the oral mucosa. Furthermore, replication of the modified live or recombinant construct vaccine is essential to evoke an immune response in the host. In spite of such challenges to the use of ORV, the benefits of using ORV to immunise FRD that are not accessible to parenteral vaccination due to difficulties in sighting and catching them cannot be discounted. The use of ORV has been recommended for FRD, albeit under supervision of the people responsible for vaccine administration (Cleaveland et al. 2006,
WHO 2018b).
Rabies has survived the history of human civilisation and continues to take human lives, particularly in Africa and Asia and especially in India where the highest number of deaths have been reported annually since 1985. No successful removal of the rabies virus from its enzootic environment (animal host) has been achieved throughout the world to date
(Wunner and Briggs 2010). As most of the Indian deaths due to rabies (95%) are attributable to transmission from dogs (Sudarshan et al. 2007), it is important that a comprehensive study of the burden of rabies in India is undertaken to establish the magnitude of the problem due to the large numbers of uncontrolled dogs in the country.
13
1.3 Burden of Rabies in India
1.3.1 Mortality
Ninety-nine percent of all human deaths due to rabies are in developing nations (Knobel et al. 2005). Rabies is endemic in Asia and Africa and the disease poses a significant public health threat to poverty-stricken nations in these continents. Asia has the highest number of human deaths from rabies, with India accounting for the majority of these
(WHO 2013). However, insufficient financial resources, a weak health care infrastructure and inadequate reporting systems potentially leads to underestimation of the real burden of the disease in India (Banyard et al. 2013). Diseases that affect children and can be prevented through vaccination, such as diphtheria, pertussis, tetanus, measles and poliomyelitis, take priority over rabies and other zoonotic diseases in India because of the general national perception, until recently, that maternal and child health programmes are more important aspects of health management (John et al. 2011). Rabies is not a notifiable disease in India (Kole, Roy, and Kole 2014), which further results in under-reporting and less emphasis placed on the disease. Other infectious diseases also impose a heavy burden and while rabies is a major public health issue, it remains a low priority disease due to a lack of systematic control and surveillance initiatives (Banyard et al. 2013). The absence of an integrated national programme for rabies also contributes towards 60% of the global human mortality from rabies occurring in India (Sudarshan et al. 2007).
The annual number of human deaths in India caused by canine rabies has been estimated to be 19,713 (95% CI: 4,192 – 39,733) (Knobel et al. 2005). Another study estimated its annual incidence at 2 per 100,000 population, giving a total of more than 20,000 deaths per year (Burki 2008). However, in the absence of any nationwide epidemiological survey, the estimates of rabies mortality are based on extrapolation of data collected from regional hospitals. In a multi-centre study conducted in 2003, the annual number of
14
human deaths due to rabies was 20,565, of which more than 95% had street dogs as the vectors of infection (Sudarshan et al. 2007). Another study involving a large scale verbal autopsy of the rural population in 2005 put the figure conservatively at 12,700, although that study excluded atypical cases (Suraweera, Morris, Kumar, Warrell, Warrell, Jha, et al. 2012). In the absence of an effective reporting system, the number of human deaths due to rabies in India remains uncertain, however rabies ranks as the most important of all the recognised and documented infectious diseases transmitted by dogs (Traub et al.
2005).
1.3.2 Economic burden of rabies in India
The annual expenditure on rabies throughout Asia, taking into account the direct and indirect cost of PEP, was predicted to be US$ 563 million (Gongal and Wright 2011) and that for Africa US$ 20.5 million based on a model published by Knobel et al. (2005). In a study conducted in 2004 to evaluate the burden of rabies in Asia, the economic burden of each individual bitten by a suspected rabies carrier animal and administered PEP was estimated to be US$ 49.41 (Knobel et al. 2005). The study observed uneven distribution of the disease across different socio-economic groups and inferred that the prevalence was influenced by both age-related and socio-economic factors, concluding that rabies caused five times more deaths in rural localities than in urban regions.
In contrast to other regions, the burden of rabies in India has mainly been determined through extrapolation of indirect estimates (Maroof 2013). In a report on the burden of rabies specific to India in 2014 it was estimated that at least one person was bitten by a dog every two seconds, implying 15 million bite wounds annually, resulting in a substantial impact on the Indian economy amounting to approximately US$25 million from treatment of bite-wounds and PEP (Gamble 2014).
15
A recent study estimated the annual global economic burden of endemic canine rabies at
US$ 8.6 billion (95% CIs : 2.9-21.5 billion), of which India accounted for US$ 2.4 billion
(Hampson et al. 2015) . In contrast to earlier estimates, this study incorporated the economic losses arising from the death of livestock, thus potentially arriving at a more plausible estimate. In India, the only study to forecast the cost of launching a rabies control programme through interventions for both human and animal rabies was based in
Tamil Nadu. This is a southern state of India with a human population of 72 million and the study estimated the cost to be US$ 2.9 million and US$78 million over a 20-year period for the control of human and animal rabies, respectively (Abbas, Kakkar, and
Rogawski 2014). However there are limited studies investigating the economic burden or future economic projections of the impact of rabies in India, a reflection of the limited focus on the public health importance of this disease within the country (Kakkar et al.
2012).
1.3.3 Burden in DALYs
Disability Adjusted Life Years (DALY), a measure of the impact of a disease across different backgrounds and stages of economic and public health development, was not estimated for rabies by the WHO until 2004 (Coleman, Fèvre, and Cleaveland 2004). The
DALY for a disease is an estimate of the total number of years of life lost due to premature death (YLL) and the number of years lived with a disability (YLD). The reason for not previously estimating DALYs for rabies in India was related to the quality of the data due to potential disease under-reporting. The first estimate of DALYs for rabies was calculated by Coleman, Fèvre, and Cleaveland (2004) to compare it with estimates of
DALYs for other diseases of global significance identified by the United Nations
Development Programme (UNDP). The estimated DALYs was based on the 1996 WHO
16
estimate of 35,000 human deaths annually due to rabies and was estimated as 1,160,000 using the age and sex distribution of the affected patients (Coleman, Fèvre, and
Cleaveland 2004). This was a crude estimate as it did not consider the disability component (YLD) of DALYs losses. In spite of this omission the estimated DALYs for rabies was still much higher than estimates for other tropical diseases, such as leprosy and dengue. An estimated DALYs of 1,039,119 (95% CI: 302,324 – 1,983,646) was arrived at in a study from 2004 for rabies in Asia, which included vaccine reactions as a component of YLD (Knobel et al. 2005). A recent study reported a global DALYs of
3,700,000 (95% CIs: 1,600,000-10,400,000), of which India alone contributed 1,301,865
(95% CIs: 377,000-3,436,000) (Hampson et al. 2015). However, the number of studies estimating the DALYs due to rabies are few. Coleman et al. (2004) emphasised that quantitative estimates of the impact of rabies should not be underestimated, even if there are inaccuracies in the calculations. They believed that such estimates demonstrate the public health importance of the disease, the control of which is generally seen as the responsibility of veterinary authorities (Coleman, Fèvre, and Cleaveland 2004).
1.3.4 Burden due to the loss of tourism
From 1980 to 2011 there has been a 17% increase in international travel to countries with emerging economies, such as India (Lankau et al. 2014). However, for tourists to such countries rabies is a serious travel related disease (Raghav, Bhardwaj, and Saxena 2014).
Rabies acquired from travel-and-trade related exposures in India also demonstrates the potential for translocation of lyssaviruses globally. The translocation of rabies is also known to occur through the adoption of pets incubating the disease or through travelling with an unvaccinated pet in countries where the disease is endemic (David et al. 2012).
17
As rabies is a fatal disease, travel related exposures may result in a serious impact to the local tourism industry (Lankau et al. 2014).
In an investigation into 42 deaths resulting from rabies in travellers from Europe, Japan and the USA from 1990 to 2010, six were found to have contracted the disease in India through exposure to dog licks or bites (Malerczyk, DeTora, and Gniel 2011). It was estimated that 0.2-0.4% of travellers to developing countries have at least one incident of an animal bite, with an overall estimated incidence of rabies exposure being 16-200 per
100,000 travellers per year (Neilson and Mayer 2010). In a study conducted in the
Republic of Korea on rabies in travellers during 2006 to 2012 it was found that travel to
India accounted for 12.2% of exposures, placing India in the high risk nation category behind Thailand (34.9%) and the People’s Republic of China (18%) (Park et al. 2014).
There are no published studies that have measured the impact of rabies risk to prospective travellers to India. The additional cost of prophylaxis and pre-travel instructions to travellers to countries where the disease is endemic have also not been studied.
Impediments to effective rabies control in India are multifactorial, with the presence of large numbers of stray/unowned/semi-owned/FRD in urban and rural India being recognised as the major factor associated with the disease (Davlin and VonVille 2012,
Butcher 1999). However a lack of: awareness about the disease among the economically depressed sections of society; coordinated and on-going efforts on the part of physicians, veterinarians and social workers towards sustained rabies control campaigns; and adequate administrative support from the government to support rabies control programmes, all result in persistence of the disease within the country (WHO 2016).
Although some research attention has been directed towards assessing the knowledge and attitudes of the general population and health workers towards rabies (Kalaivani, Raja,
18
and Geetha 2014, Acharya, Kaur, and Lakra 2012), studies on the population dynamics and management of FRD and the role of unrestricted dogs in the transmission of the disease, and investigations into efforts to control dog-rabies are lacking in India
(Cleaveland, Lankester, et al. 2014, Dalla Villa et al. 2010, Abbas and Kakkar 2015,
2013).
1.4 Population dynamics of FRD in India
1.4.1 Domestication of dogs in India compared with Western countries
The combined effects of artificial and natural selection led to a similar worldwide transformation of the wild wolf into a community companion dog. However, the assimilation of dogs from being a social community carnivore to becoming a family member following adaption of its physical and behavioural characteristics for desired aesthetic, economic or ritualistic functions was not uniform across the world (Clutton-
Brock 1999, Hiby 2013, Majumder, Chatterjee, and Bhadra 2014, Udell and Wynne
2008). A varied perception of dog-keeping, dog-ownership and associated responsibilities, welfare, and utility in developed countries has facilitated the gradual transition of the village/community dog into becoming a “resident” member of the family
– a change that has not been universally witnessed in developing countries, such as India.
In addition to such cultural differences that guide the attitudes of people towards FRD, the high fecundity of dogs and the availability of adequate carrying capacity (food and shelter) in developing countries has resulted in a rapidly expanding dog population that poses a serious health risk to human populations in these countries (Wandeler et al. 1988).
Besides being the largest reservoir host for rabies, FRD are also responsible for attacks on humans and other animals, damage to property, road accidents, contaminating the
19
environment with faeces, spreading garbage waste and causing noise pollution (Beck
1973, Rinzin, Tenzin, and Robertson 2016).
1.4.2 Carrying capacity and estimates of the dog population in India
The most important needs for any species to breed and survive are food, water and shelter, and these define the carrying capacity of a habitat (Hiby 2013, Wandeler et al. 1988).
Hiby (2013) reported that the carrying capacity of a specific location for domesticated dogs was controlled by humans. She further elaborated that rapid urbanisation has led to abandoning of pet dogs resulting in dogs being allowed to roam freely in private and public areas where inadequate management of edible waste results in sufficient resources
(food) for the survival of these dogs. The carrying capacity of habitats for dogs is also influenced by the socio-cultural traits, religious practices, and economic level of the urban and rural human populations (Wandeler et al. 1988). The population density of unowned dogs is also dependent upon the rate of abandonment of unwanted litters and the propensity of the inhabitants to adopt or own dogs (Amaku, Dias, and Ferreira 2010).
The unchecked increase in the population size of street dogs has resulted in an ecological imbalance in the developing world, including India (Krishna 2009). Factors, including the absence of sustained and successful population control programmes for dogs on the streets and the presence of large amounts of edible litter, contribute to the high density of street dogs within the country (Butcher 1999). An understanding of the potential harm to human health and other community issues related with over-population of dogs is essential when implementing control programmes, such as dog population management or mass vaccination of dogs against rabies. An educated and aware public, that accepts responsible ownership of dogs, makes implementation of such control measures easier and more effective (Selby et al. 1979).
20
A report published in 2013 estimated that, in India, there was a dog for every 36 people, with only 20% of these dogs being pets (Chatterjee and Riaz 2013). With the population of India estimated to be 1.28 billion in 2013 (www.populationpyramid.net/india/2013/), this equates to 36 million dogs. In contrast a livestock survey conducted in India in 2012 estimated that the combined population of dogs, including stray dogs, in rural and urban
India was 11.67 million (Ministry of Agriculture 2015). Interestingly, this census found that the population had reportedly decreased by nearly 40% from 19.08 million in 2007.
This is also in contrast to a business survey monitoring the marketing of dog feed and accessories which claimed that the owned dog population in India grew by 58% between
2007 and 2012, making it the fastest growing canine population globally (Bradley and
King 2012). This may indicate that the reported 40% drop in the total dog population was a result of fewer stray dogs, although there are no studies to substantiate this estimate.
Bradley and King (2012) also pointed out that India has one of the lowest rates of dog ownership (4 dogs owned per 1,000 people), attributing this low figure to the large rural and poor population.
The classification of owned and unowned dogs is largely blurred in most Indian societies, both rural and urban. For rabies to establish and flourish, there needs to be ample interaction between canine populations (Amaral, Ward, and Freitas da Costa 2014).
However to assume that an owned dog will not interact with an unowned or free roaming one would be incorrect as owned dogs can also be found roaming freely (Slater 2001).
The various circumstances under which an owned/semi-owned or a completely unowned dog could potentially be infected and transmit the rabies virus depends on the level of restriction imposed on it (Cliquet and Barrat 2012). It is important to define what a FRD is, before the aspects of their demography in the context of rabies control can be discussed. A dog can be categorised as a FRD depending on the level of restriction placed
21
on its movement. The matrix displayed in Figure 1.1 categorises dogs on the basis of their dependency on humans and the restraints that humans place on their movement. Owned dogs could be family dogs that are fully dependent on their owners for food and shelter, but which could still have opportunities to interact with other dogs that are neither restricted nor directly dependent upon humans for their subsistence. The community or neighbourhood dogs can have partial restrictions placed on their movements (for example they may be restricted during the daytime but may be let free to roam during night-time or vice versa) and are semi-dependent upon humans for their food and shelter. The dogs that are always under supervision, kept indoors all the time and are always accompanied/restrained (on leash) when outdoors are the ones that are truly owned and least likely to interact with other dogs (Cliquet and Barrat 2012).
Figure 1.1 A matrix to categorise a dog’s dependency on humans with respect to restrictions placed on their movement (Cliquet and Barrat 2012)
22
A description of the various categories of dogs was given by Vanak and Gompper
(2009b). Dogs were classified as urban, rural or wild depending upon their habitation with their movement range being categorised as limited, wide ranging or feral, respectively.
However, this categorisation fails to meet the practical circumstances in urban environs in India as an abandoned dog could roam widely in search of food and/or shelter.
Similarly, a feral or rural dog could have a limited range of movement if the habitat has a high carrying capacity. The broad categories that were constructed by the authors on the basis of the feeding and ranging habits of dogs were: owned, urban free ranging, rural free ranging, village, feral, and wild dogs. A feral dog refers to an abandoned dog that has turned wild, while a wild dog is one that inhabits the fringes of forests and human habitats
(Vanak and Gompper 2009b). A dog could be claimed to be owned without any obligation from the owners for its health (including immunisation) or veterinary care. A rural free- ranging and a village dog may be similar because both may have equal probability of interacting with an infected dog. A definition given by Beck (1973) encompasses the true circumstances when a dog can be called a FRD as “Any dog observed without human supervision on public property or on private property with immediate unrestrained access to public property”. This definition incorporates all the circumstances which allow the circulation of rabies virus within a dog population.
While there has been some pioneering work, predominantly ecological, undertaken relating to the interface of dogs and wildlife, there are few studies in India that have focused on enumeration of FRD with the aim to controlling their population or assessed the limitations to achieve adequate vaccination coverage against rabies.
23
Prior to discussing the need for, and impact of, conducting studies on the population dynamics of dogs, it is pertinent to review the procedures used to quantify the number of
FRD within a particular habitat and to elaborate on their application in India.
1.5 Population size estimation of dogs
Study of the FRD population is important to those involved with management of the dog population and/or rabies control as it is essential to: appreciate the magnitude of the problem of the uncontrolled FRD population; understand the demographics and welfare status of FRD; enable the design and implementation of dog population control, and rabies control interventions; and assess the outcomes of such interventions.
Internationally, studies conducted on canine demographics have primarily been undertaken to assess rabies control (Slater 2001). Many authors have also emphasised the need to know the size, dynamics and demographics of the target population before planning any intervention approach (Meslin, Fishbein, and Matter 1994, Robertson,
Wilks, and Williamson 1993).
1.5.1 Enumeration of owned dogs
Enumeration of domesticated owned dogs has been undertaken by many researchers and includes: public surveys (Patronek and Rowan 1995, Butler and Bingham 2000); surveys of local veterinary practices (Perrin 2009); data sourced from a pet travel scheme (Asher et al. 2011) and registrations with pet insurance companies (Egenvall et al. 2009), kennel pounds, shelter homes, rescue groups (Chua, Rand, and Morton 2017, Rowan and Kartal
2018) and identification (microchip or ear tag) enterprises (Morley 2002); and surveys implemented by government agencies/city councils (Ministry of Agriculture 2015).While these methods may have biases, they help contribute to the evolution of strategies for improved veterinary and animal health coverage for dogs (Downes et al. 2013). However,
24
it is the presence of unrestrained, ownerless FRD on the streets of urban and rural India that accounts for the high incidence of human and animal mortality due to rabies (Morters et al. 2013, Kitala et al. 2002, Sudarshan et al. 2007). Because of this large population of
FRD it becomes critical to compare methods used to estimate the abundance of the dogs within the affected area prior to implementing any disease or population control measure to ensure adequate resources are allocated. Estimation of abundance of FRD has been attempted through the administration of questionnaire surveys to the residents of the area, however this method is more appropriate for owned dogs (Mustiana et al. 2015, Otolorin,
Umoh, and Dzikwi 2014, Ratsitorahina et al. 2009). This method also results in a large margin of error in estimating the population size of FRD (Fei et al. 2012). In contrast, it has been reported that methods used to estimate the population size of wild animals can yield reliable estimates of the population size of FRD (Belo et al. 2015).
1.5.2 Enumeration of unowned or FRD
Formulating an enumeration methodology for dogs, including FRD, is very challenging in areas where registration and licensing of dogs is not mandatory (Özen, Böhning, and
Gürcan 2016). “Guidelines for Dog Population Management”, a joint publication in 1990 authored by the World Society for Protection of Animals (WSPA 2010) and the World
Health Organisation (WHO), described four wildlife enumeration techniques that could be applied to estimate the population size of FRD: total or indirect counts through counting the number of dogs sighted in one or more transects and extrapolating the population from this number; estimates from the rate of capture (regression method); estimates from physical recaptures; and estimates from photographic recaptures or Beck’s method (Fei et al. 2012). Estimations of the FRD population by these methods are valid only when a number of assumptions are applied, and the estimates vary with each
25
enumeration method. The document specified two critical assumptions. The first was that the search effort on all counting occasions should be same and the second was that the likelihood of each dog being encountered on each occasion should also be the same.
While the first assumption is easy to satisfy, the second assumption has a high chance of failing as the likelihood of encountering an individual dog may vary depending on the time of sampling, the prevailing weather conditions and the individual behaviour and demographic characteristics of that dog (Chao, Lee, and Jeng 1992, Fei et al. 2012). The impact of these assumptions can be reduced by increasing the number of count sequences or by reducing the factors which affect the behavioural response of the FRD, such as individuals showing varied response to their capture/recapture (Amstrup, McDonald, and
Manly 2010, Chao 1987). The pursuit to develop a more accurate enumeration method has led to the use of other methods, including distance methods, extensive direct counts in the chosen areas and mark-resight surveys (Belo et al. 2015). However, all methods have some limiting features and these benefits and limitations are summarised in the following sections.
1.6 FRD enumeration techniques
1.6.1 Direct counts
A direct count is the simplest counting technique and involves counting the dogs seen in a defined geographical area during an assigned period of time. The methodology has been explained in detail in the report published by the WSPA (WSPA 2010). Application of this technique can be used in two ways: either by collecting an indicator count (also called an index of abundance) or by calculating a population estimate. Indicator counts employ minimal resources and, while they do not derive the exact number of dogs in a locality, periodic repeats of such counts indicate an increase or decrease in the population size
26
during the period under observation. This population trend can be depicted through constructing a line graph drawn over a number of counts spanning the period under observation. Another advantage of this technique is the possibility of splitting the measure into various groups such as age, sex, reproductive status and health score categories. The method has been used in India where counts conducted every six months from 1997 to
2002 were used to assess the effectiveness of a sterilisation programme on the number of
FRD in the northern city of Jaipur (Reece and Chawla 2006). Direct counts have also been employed during a WSPA effort in 2006 while conducting stray dog counts in Cairo and Colombo (WSPA 2010). As well as not being an ideal method to calculate the density of dogs, an important drawback of this method is that it gives an estimate of the population and its characteristics which may not be a true representation of the total population in a city, state or district. Another disadvantage is that it is very sensitive to sampling biases and confounding factors. The time of the day, prevailing weather conditions and level of garbage management can also act as confounding factors. Any changes in the topography of the area, especially due to rapid urbanisation, also makes this method unreliable
(WSPA 2010). There are few studies that have assessed the utility of the direct count method, however Hossain et al. (2013) commented that extrapolation of the direct count to estimate the population size of FRD is unreliable and should be based on dog densities rather than the human to dog ratios resulting from such counts. Unfortunately, other studies have not assessed the reliability of this method.
When counts are conducted in various representative regions, the counts can be combined, and the population of an entire city estimated. This method allows for calculating the density of dogs. However, the accuracy of such estimates can be questionable, where, as the counting exercise is repeated across many sample counts, the magnitude of variance tends to be large. It also requires significant amounts of resources,
27
such as personnel and time. As a result, the confounding factors can increase and thus some assumptions are necessary to augment the accuracy of the method. To start with sampling of the regions selected should be random and the individuals performing the count should not add to the bias. This means that the search effort on each occasion and in each location, should be equal. The other assumption, which is often difficult to satisfy and constitutes a major drawback to this method, is the assumption that each dog has an equal probability of being counted (WSPA 2010). Despite its shortcomings, this method has been used by many ecologists to study the abundance of FRD in urban and rural settings (Tenzin, Ahmed, et al. 2015, Aiyedun and Olugasa 2012, Hossain et al. 2013,
Otolorin, Umoh, and Dzikwi 2014).
1.6.2 Capture – recapture studies
Capture-recapture methodology yields unbiased estimates of animal populations in a closed population when executed over a short period of time, provided there are no counting errors due to loss of the “markers” used to identify animals (Belo et al. 2015).
The method is commonly known as the ‘Petersen–Jackson’, ‘Lincoln’ or ‘Lincoln-
Petersen’ index and was first known to have been used by Pierre Simon Laplace in 1786 to estimate the size of the human population (Goldsmith and Sutherland 2006, Belo et al.
2015). However, it was only after Petersen in 1896 and Lincoln in 1930 used the technique to estimate the population size of flatfish and aquatic birds, respectively, that it found wider use in ecology and biology (Fei et al. 2012, Belo et al. 2015, WSPA 2010).
The approach involves two waves of capturing effort with individuals caught on the first capturing being marked or tagged and released back into the population. The population size estimate is derived from the simple ratio of the product of individuals captured, marked and released on the first instance and animals captured on the second time to the
28
number of marked animals recaptured on the second capturing effort. The population sampled at the second sampling then contains both marked and unmarked animals. The estimate of the population is calculated as below:
푴 퐱 풏 퐍 = where, 풎
N = Estimate of population size,
M = Number of animals marked and released at the first capture event, m = Number of marked or tagged animals recaptured in the subsequent effort, n = Number of animals captured at the second attempt.
(Caughley 1977, Davis 1982, WSPA 2010).
The capture-recapture estimation is reliable under the following assumptions: (i) the population being estimated is a closed population; the time period between the two capture efforts is very short; and (iii) the capture probabilities for each individual remains unchanged throughout the capture-recapture session (Jeremy and Robert 2006, Krebs
1999). Any compromise to these assumptions introduces bias in the estimate which is pronounced with smaller or no recaptures (Amstrup, McDonald, and Manly 2010). To account for this bias, Chapman (1951) proposed a correction which is calculated as follows:
[(풏₁ + ퟏ) 퐱 (풏₂ + ퟏ)] 푵 = − ퟏ (풎 + ퟏ) where,
푵 = Total population size,
풏₁ = Number of individuals captured on the first attempt,
29
풏₂ = Number of individuals captured on the subsequent attempt,
풎 = Number of marked or tagged individuals recaptured on the second attempt (Chapman
1951).
This correction has been used by some researchers, although the limitation of the estimation still lies with the difficulty in verifying the assumptions (Belo et al. 2015).
There are also other issues/disadvantages when the corrected method is used for estimating the size of a dog population. As it requires physical trapping and re-trapping of individuals, capture probabilities are influenced by subjects being either trap-shy or trap-happy (Williams, Nichols, and Conroy 2002). Other factors, such as the activity pattern of the dogs and variations due to observer’ fatigue induced by long duration of observation surveys, also potentially can have a major influence on the estimate (Fei et al. 2012, Belsare and Gompper 2013, McCallum 2005).
1.6.3 Regression method
This method utilises the rate of capture of the dogs over a number of capture trials and has remained one of the more popular methods used by researchers involved in studies of dog populations and rabies management (Fei et al. 2012). The method uses visual captures using photographic equipment, rather than requiring the physical capture of the dogs. The dog population in an area is calculated by extrapolation of the regression line obtained by plotting the number of dogs “captured” on each capture effort against the accumulated number of distinct (unique) dogs captured until the present trial. That is, if a total of ‘t’ capture trials are conducted, where Mj and mj are the number of dogs captured and
th recaptured on the j occasion, respectively, and 풙풋 is the total number of unique dogs captured until the j-1th trial, then the number of new dogs captured at the jth trial is
30
풚풋 and is defined as 풚풋 = 푴j−mj. When 풚풋 is plotted against 풙풋 the regression line obtained can be extrapolated to estimate the dog population.
Mathematically,
−풂 푵 = 푹 풃 where,
풕 ∑풋=ퟏ(풙풊풋 − 풙¯) 퐱 (풚풊풋 − 풚¯ ) 풃 = 풕 ∑풋=ퟏ(풙풋 − 풙¯) ² and,
풂 = (풚¯ − 풃풙¯)
is the slope and the intercept estimates of the simple linear regression line of 푦 against 푥 ;
푥¯ and 푦¯ are the sample means of 푥 and 푦, and 푁푅is the total number by the regression method (Fei et al. 2012).
Although the regression method is not difficult to perform, only one study was found where this method has been applied to enumerate dogs. Tung et al. (2010), applied the technique to estimate the dog population in Taiwan; however, the requirement for multiple surveys (sampling efforts) has limited its use. Moreover, the method is an extension of the Beck’s method which is discussed in the subsequent section.
1.6.4 Beck’s method
Beck’s method, first described by Beck in 1973, was devised especially for enumeration of dogs without the requirement to physically mark the dogs or to capture-recapture dogs.
Instead, it was argued that since dogs are generally individually distinguishable, a record of their natural markings on two or multiple occasions will generate sufficient data to be used to estimate “recapture” proportions (WSPA 2010, Beck 1973). In a review
31
evaluating different enumeration studies in dogs, Beck’s method was described as an extension of the Lincoln-Petersen’s approach from two capture efforts to multiple closed population captures (Belo et al. 2015). Fei et al. (2012), claimed that the Beck’s method provided better estimates of the population size with smaller variance compared to the regression–extrapolation method. Mathematically, for j number of capture attempts where j = 2, 3…, t; the capture rates would be
풎풋 풎풋 , and, 푴풋풙풋 = 푴풋 ÷ ( ) 풙풋 풙풋
And the weighted average would be:
풕 풕 푵 = ∑풋=ퟏ(푴풋풙풋) ÷ ∑풋=ퟏ(풎풋) , where,
푁= Population estimate,
푀푗 = Number of animals observed in the first capture attempt,
푥푗= Total number of animals observed in each of the subsequent capture attempts,
푚푗 = Number of marked animals caught in the subsequent capture attempts.
푡 ∑푗=1 = Summation of t capture attempts, j denoting capture attempts from 1 to t.
1.6.5 Methods based on variations/modifications of capture-recapture procedures
An important premise of the capture-recapture method is that the catchability of all animals is similar. Schnabel (1938) and Schumacher and Eschmeyer (1943) developed variations of the Petersen’s estimate to derive relatively robust population estimates compared to the Lincoln–Petersen estimation technique.
32
(a) Schnabel method
A series of Petersen’s surveys are treated as multiple sessions of the survey and the estimate is obtained as a weighted average of Petersen’s estimates (Krebs 1999).
Mathematically, if ‘t’ sessions are run then,
푵 = ∑풕( 푪풕푴풕) ÷ ∑풕(푹풕 ), where,
푁 = Population estimate,
퐶푡 = Number of individuals caught in sample t,
푀푡 = Number of marked animals just before sample t was run,
푅푡= Number of marked animals in the sample t.
Simply stated, the Schnabel method proposes subsequent samplings to the point when no more unmarked individuals are observed. At this point it infers that all animals in the population have been captured at least once. The Beck’s method is a variant of this technique.
(b) Schumacher and Eschmeyer Method
It was observed that if Mt (the number of animals marked before sample t was undertaken) is plotted on the 푥 axis against the proportion of marked individuals from the t th sample i.e. (Rt ÷ Ct), the points should lie on a straight line of slope 1 ÷ N which passes through the intersection of the 푥 and 푦 axes. The estimate can be obtained by:
2 N = ∑ (Ctx Mt ) ÷ ∑ (Rt x Mt) where,
N = Population estimate,
33
Ct = Number of individuals caught in sample t,
Mt = Number of marked animals just before sample t was run,
Rt = Number of marked animals in the sample t.
Estimation of the slope (1 ÷ N) and the population size can also be obtained using linear regression (Krebs 1999, Totton, Wandeler, Zinsstag, et al. 2010). This method was employed by Totton, Wandeler, Zinsstag, et al. (2010) to enumerate the number of dogs in a northern district of Jodhpur in urban settings. Belo et al. (2015), however, found the use of this approach for their chosen premises difficult to verify and observed that the method was not appropriate for use in FRD as it was initially designed for aquatic environments where saturation (when no new individual is captured) is difficult to achieve.
(c) Mark-resight logit normal method
Punjabi, Athreya, and Linnell (2012) used a mark-resight technique to estimate the size of the FRD population in Mumbai, India. Their estimate was based on employing natural marks of individual dogs in the capture-recapture procedure in conjunction with several assumptions. The first assumption was that the naturally marked animals were representative of the unmarked population in terms of their likelihood of being sighted.
Other assumptions included that, the marks were permanent or not likely to be removed during the study and the population was closed demographically and geographically. The bias due to heterogeneity resulting from different sighting probabilities of dogs was corrected by using a logit normal estimator.
The surveys required for employing the Mark-resight logit normal method consist of an initial survey, where all individuals are recorded. After the initial survey, a series of
34
evenly spaced primary surveys, consisting of three or more secondary surveys, are conducted. Subsequently, this approach utilised additional information obtained from observing the population that was not sighted during the initial survey but was sighted in the subsequent surveys. The software program MARK (Cooch and White 2006) could be used to run the different analytical models based on the nature of the data collected. Belo et al. (2015) pointed out that, although the logit normal model took care of the heterogeneity in the capture probabilities, in the case of Punjabi, Athreya, and Linnell
(2012), the authors fitted a rather simpler model (citing unspecified convergence problems) that did not allow for inclusion of impacts of the parameters of heterogeneity.
Further, Belo et al. (2015) felt that it also prolonged the time period between the first and last capture, thereby failing to conform to an important assumption of the capture- recapture studies regarding the surveys being conducted over a short time period.
Essentially, this basic flaw meant this method was not suited for FRD as it was not possible to conduct more than one primary survey, without failing the assumption of rapid surveys.
Hiby et al. (2011), applied a different approach to the mark-resight surveys when they conducted enumeration of FRD in three cities in the northern state of Rajasthan, India utilising recorded information on sterilised dogs. Dogs that are sterilised under an Animal
Birth Control (ABC) programme in India usually have their ears notched for identification purposes to minimise re-catching of dogs that had already been de-sexed. Using a time constant model to derive the estimate, Otis et al. (1978) argued that sighting probabilities might vary with time, but would apply equally to marked and unmarked dogs. The benefit of this method is that the notched ears used for identification purposes would not be lost or removed, as is the case when artificial marks are used for identification. Also the method did not require the assumption of equal rates of mortality for marked and
35
unmarked animals because the estimate of marked animals was explicitly used to estimate the surviving unmarked population as well (Hiby et al. 2011). However, the authors claimed that, as the population was closed, a shift due to immigration/emigration could not be completely excluded. Additionally, they also highlighted that the veracity of the sterilisation records could affect the robustness of the method (Belo et al. 2015).
(d) Estimation using Bayesian analysis
Using the principles of Bayesian analysis, “prior” distribution data consisting of information on canine abundance and their accessibility for vaccination are compiled using household surveys, counts of collared dogs, transect line observations and counts from vaccination points during a mass vaccination programme. The “posterior” distribution for the parameters needed is obtained by relying on the binomial model of capture probabilities of unmarked dogs and the Markov Chain Monte Carlo (MCMC) approach (Matter et al. 2000, Belo et al. 2015). A similar approach was adopted by Gsell et al. (2012), where they used the population estimate updates available from the Iringa
Municipal Veterinary office in Tanzania to arrive at a more precise estimate of the actual proportions of stray/feral/ownerless dogs. While this approach would seem credible, its dependence on “prior” distribution inputs is a concern as its choice may result in bias
(Belo et al. 2015). For example, a count or distribution of collared dogs in a survey to be used as “prior” may be completely different from the count of dogs brought to a central vaccination point and would yield different population estimates for the same area of interest.
1.6.6 Method based on distances
The methods elaborated above, which form the guidelines from the WSPA and WHO for surveying dog populations, essentially estimate the number of dogs in an area, but do not
36
give any robust indication of their density (WSPA 2010, Belo et al. 2015). Childs et al.
(1998), estimated the density of FRD by recording their presence along randomly selected transects in the area of interest. This technique was devised primarily on the basis of the need to estimate population-density for designing oral bait distribution for effective vaccination coverage against rabies in heterogeneous localities (Childs et al. 1998). The method is believed to provide a reliable density estimate with few assumptions (Buckland et al. 1993). The primary assumptions are that the population is randomly distributed across the area, the probability to detect an individual dog along a transect line is 1, and the distances to individual dogs from the transect line is measured exactly. The perpendicular distance of an individual from the transect line is measured and these measurements collated. These distance data are then categorised into discrete groups and displayed graphically in a histogram. A line of best fit on the resultant histograms is obtained, which is usually Poisson distributed as the number of dogs decreases with increasing distance from the transect line. The detection function then estimates the number of undetected individuals (visibility diminishes as distance from the transect line increases). This estimation is followed by density calculations and confidence level determination as the number of dogs per unit distance along the transect line. These steps are simplified using Distance sampling software (Laake et al., 1994).
The distance method has many advantages over other enumeration methods, especially the capture-recapture methods, as its assumptions are relatively easily met, it does not require capturing or marking the animals and it involves a single time effort (Childs et al.
1998). It has been claimed that this method does not require photographs and can produce accurate repeatable estimates, even when different observers are used (Buckland 1993).
It is surprising that only one study could be found where this method has been used for
FRD, in spite of the significance of density estimation for various interventions, such as
37
mass vaccination against rabies (Anderson, Jackson, and Smith 1981). However, Belo et al. (2015) did point out that this method may not be suitable for a locality where unrestricted dogs live in close association with humans because such a situation will increase the proportion of dogs that may not be observed as they may be “indoors” at the time of the trial. It may be argued that the knowledge of the density may not be beneficial for managing resources, such as procurement of sufficient doses of vaccines, for which knowing the number of FRD appears a better option.
Although there are numerous methods available to quantify the number of FRD in any particular area, it is difficult to identify one methodology which is superior. This is because all methods have potential limitations and biases and the economics of implementing different methods also varies (Belo et al. 2015). However, the importance of quantifying the FRD population to evaluate interventions cannot be understated (Belo et al. 2015, Belsare and Gompper 2013, WSPA 2010), as it is key to the success for any intervention to control rabies. Although there is a perception that the strategy that resulted in the eradication of canine rabies from North America and Western Europe may also be applicable in the developing world, the increasing incidence of canine rabies in Africa and Asia suggests otherwise (Hampson et al. 2009). Elimination of canine rabies also requires a thorough understanding of the FRD population demographics (Hampson et al.
2009). The patterns of contact rates between dogs and between humans and dogs, and the resources that contribute to the carrying capacity of the environment can be understood by assessing the demographics of dogs (Wandeler 1985). An accelerated increase in the rate of urbanisation in the developing world has contributed to increased dog-to-dog and human-dog interactions, further highlighting the need for studies on the demographics of dogs in urban communities (Gsell et al. 2012).
38
1.7 Dog demographics and rabies control
Demographic data refers to the statistical data relating to a whole population or groups within a population (Banerjee and Chaudhury 2010). When applied in the context of the role of the dog as a vector for the spread of rabies, it is important to understand the dynamics and demographics of the canine population with respect to: estimates of the population size; the turnover and change (increase or decrease) in the population; sources and numbers of FRD and ownerless dogs in the total population; and assessing the level of supervision of owned dogs and their distribution and accessibility to vaccination
(Kitala et al. 2001). Thus, the implementation of an intervention, either for population management or mass vaccination, warrants an understanding of the demographic characteristic of these dogs (Knobel et al. 2005, Morters, McKinley, Restif, et al. 2014).
In areas where rabies is endemic, such as India, control of FRD can substantially reduce the incidence of rabies in humans. To achieve this it is important to study the available habitat that provides the resources for survival of FRD i.e. the available food and shelter in the environment (Perry 1993). Furthermore, epidemiological investigations into the demographics of FRD are important from the point of view of separating sub-populations: pups from adults; males from females; intact from de-sexed; and the tendency to form different sized groups or remain solitary wanderers (Slater 2001).
Studies have been undertaken to determine the usefulness of information on the demographics of canines for the control of diseases, including rabies, while others have been conducted to predict the usage of veterinary services and to forecast the future canine population size (Downes, Canty, and More 2009, Matter et al. 2000). One of the earliest studies conducted on dog population dynamics and their demographics was in Manhattan,
USA where the rate of change of the population was demonstrated to be related to age dependent birth and death rates (Nassar and Mosier 1980). The demography of dogs and
39
the dog-human relationship were explored in a study in Zimbabwe in the context of rabies control and it was found that higher human densities resulted in higher dog densities
(Butler and Bingham 2000). In that study, it was concluded that all dogs were owned, although the level of restriction place on them varied and there were no feral dogs. In one of the first studies on the characteristics of dog population dynamics and demography conducted in India by Hiby et al. (2011) in cities of the northern state of Rajasthan from existing ABC intervention data of twelve years (1996-2008), it was concluded that a high rate of castration of males resulted in a reduced population reproductive rate. The estimate of the FRD population reported in that study exceeded the observed number of surviving sterilised dogs, which the authors reasoned was due to low survivability of young dogs.
Notwithstanding the conclusion that such overestimation tends to rise as an ABC program progresses (Hiby et al. 2011), it highlights a lack of techniques for enumeration. Another similar study to demonstrate the benefits of ABC on non-vaccinated, unsterilized dogs found that even the sexually intact dogs in areas where the intervention was applied were relatively disease-free compared to those in areas where the program had not been implemented (Yoak et al. 2014). This study, along with Totton, Wandeler, Zinsstag, et al.
(2010), demonstrated that the Body Condition Score (BCS) of sexually intact dogs was significantly lower than that of the sterilised dogs in the areas of intervention and there was a lower prevalence of several diseases in the sterilised dogs due to enhanced capability to resist infections. Totton, Wandeler, Zinsstag, et al. (2010), also assessed the stray dog population in Jodhpur, India where ABC interventions were adopted and predicted the trend of population decline using models constructed based on these demographic studies. The population size and demographics of the FRD population have been studied in India’s neighbour, Bhutan, and it was demonstrated that the key to the success of Capture, neuter, vaccinate, release (CNVR) programmes lay in: targeting
40
female dogs, especially in the breeding season; regular monitoring and assessment of the programme; encouraging community participation; and implementation of relevant legislation (Rinzin, Tenzin, and Robertson 2016).
Studies have also been conducted in Africa on the ecology and demographics of stray dogs to estimate the size and characteristics of that population (Ratsitorahina et al. 2009,
Kitala et al. 2001). Interestingly, although control of dog-related rabies in Africa is as equally challenging as it is in Asia, these studies revealed the difference in the demographic characteristics of FRD in Africa vis-à-vis Asia. Dog ownership was high
(63% in Machakos, Kenya; 88.9 % in Antananviro, Madagascar), although an equally high percentage of owned dogs but with unrestricted movements were reported (69% in
Kenya and 79.1% in Madagascar). However, the proportions of dogs that were vaccinated against rabies were much lower than the required 70% (29% in Kenya and 3-17% in
Madagascar). In another study, Perry (1993) observed a decline in the vaccination coverage against rabies in the dogs of eastern and southern Africa and recommended generating data on the dog population size and structure to effectively target those populations that were instrumental in transmitting rabies. In a systematic review carried out in 2012, the characteristics of the dog population in developing nations(defined as
“low income” or “low-middle income” countries by World Bank or as developing and emerging economies by the International Monetary Fund) were examined, including ownership rates, male–female ratios, life expectancy and accessibility to vaccination and its coverage for rabies control (Davlin and VonVille 2012). The review concluded that vaccination of dog populations was a feasible option to combat rabies, even in the remotest of places, provided that the national governments and local health authorities make rabies control a priority. Among other recommendations, the review stressed the need for educating dog–owners and making available affordable veterinary services. The
41
southeast Asian region, which saw emergence of rabies on previously rabies-free islands, such as in Bali (2008), Nias and Larat (2010) and Kisar (2012), Indonesia (Putra et al.
2013, Mustiana et al. 2015), led to the implementation of a demographic study in 2014 on the island of Lombok. That study found that ownership of dogs varied between ethnic groups inhabiting Lombok. Chinese households were more likely to own a pedigreed dog compared to other ethnicities, which preferred local dog breeds. The study also reported that the number of unowned FRD was higher in urban than in rural communities, primarily due to the quantity of edible garbage (fish waste, left over meals and kitchen refuse) in urban areas (Mustiana et al. 2015).
1.8 Home ranges and Social behaviour of FRD
An important aspect of demographic studies of FRD involves evaluating their behaviour in relation to potential human exposures (Meslin and Briggs 2013). Such investigations are important due to the prevailing ambiguity in the status of FRD in human settlements as being either partially restricted, free-roaming or truly feral animals (Boitani and Ciucci
1995). Apart from a general lack of information about the existing FRD population, the failure to achieve the required 70% vaccination coverage of the population against rabies in mass vaccination programmes, especially in India, has been attributed primarily to the inaccessibility of these dogs for vaccination (Cliquet et al. 2007). Most recently, in the third report of the World Health Organisation expert consultation on rabies, it was recommended that supervised oral vaccination should be adopted for semi-restricted, non-restricted and unapproachable canines, in conjunction with traditional parenteral applications to the dogs that are easily approachable, to achieve adequate vaccination coverage (WHO 2018).
42
The transmission of rabies is also influenced by the territoriality and movement of FRD and knowledge of their demographic characteristics has been used to model the spread of the disease based on the contact rates between dogs (Hampson et al. 2009). Understanding the sub-populations and their propensity to form social groups forms an important part of epidemiological studies of the dog population and rabies control (Slater 2001). While the
FRD in groups are likely to have a higher rate of mutual contact by merely staying together, the likelihood of transmission is increased due to enhanced agonistic behaviour towards other groups of FRD and humans (Fox, Beck, and Blackman 1975, Pal, Ghosh, and Roy 1998a). The FRD in groups also present a challenge for the vaccinators for parenteral inoculation due to safety concerns arising from dog-bites (Jibat, Hogeveen, and
Mourits 2015, Bögel and Joshi 1990).
Investigation of the home-ranges of FRD generates data on the likely spread of rabies through assessment of their size and if these ranges overlap with potentially rabies- infected areas. Overlapping of home-ranges of FRD results in altered behaviours, such as territorial rivalry or pack fights during mating season (Laffan, Wang, and Ward 2011,
Pal, Ghosh, and Roy 1998a), which increase the likely transmission of rabies. Dürr and
Ward (2014), studied the home ranges of FRD in remote Aboriginal and Torres Strait
Islander communities in northern Australia to investigate the contact rates between dogs within and between communities. Although the study primarily was designed to compare four different methods to assess the home-ranges and utilisation-distribution of the dogs in the two communities, it shed light on FRD that had unusually larger home-ranges. They reasoned that these dogs could be the ones that accompanied their owners for activities such as hunting. As most of the dogs were owned, exchange of dogs between relatives and family friends living in different areas was frequent and resulted in some dogs having larger home-range than others. The study observed that such dogs could play a crucial
43
role in the transmission of rabies virus, if it was introduced into the population. They considered theirs and similar studies as being essential for modelling the transmission of rabies, especially if disease was introduced into a previously free population.
Although few studies on the home ranges of FRD have been published and those that have, have primarily been undertaken in countries where dog-related rabies is not a serious issue, they impart crucial information regarding the social behaviour and roaming tendencies of this population (Boitani, Ciucci, and Ortolani 2007, Daniels and Bekoff
1989, Beck 1973, Font 1987). A novel and interesting development in the study of home ranges of FRD is the recent investigations of the influences of extrinsic and intrinsic factors that influence the size of the home range (Durr et al. 2017, Pérez et al. 2018). The social behaviour of FRD have been studied in Kolkata and Katwa in the state of West
Bengal in India, albeit reporting the behaviour and home ranges from the point of view of estimating their dispersal, denning habits and agonistic behaviour towards other FRD
(Pal, Ghosh, and Roy 1998a, Majumder et al. 2014, Majumder et al. 2016).
The population dynamics and the behavioural traits of FRD are affected by the socio- cultural and religious practices of the human societies they are associated with (Wandeler
2012). Thus the data generated through demographic studies are useful when combined with information obtained from the affected communities regarding public knowledge, their attitudes and practices with respect to dog ownership, and attitudes towards feeding
FRD and rabies control, as this enables development and implementation of comprehensive rabies control programmes (WHO 1987). Implementation of a
Knowledge, Attitudes and Practices (KAP) survey is important to assess the level of awareness of the affected communities and is an effective tool to design awareness campaigns (Rinzin, Robertson, and Mahat 2017, Sambo et al. 2014).
44
1.9 Studies on Knowledge, attitudes and practices (KAP) of rural and urban communities on rabies, free roaming dogs (FRD) and their impact
Studies on the KAP towards diseases have been conducted worldwide and are based on the principle that increasing knowledge among communities can help minimise the disease burden by bridging the knowledge gap between different communities about the disease, and to bring a change in their attitudes and practices towards the disease (Sambo et al. 2014, Hlongwana et al. 2009, Govere et al. 2000). Such surveys have been widely used in the medical sciences to understand the awareness level of the affected communities regarding particular diseases (Dhand et al. 2012). The communities’ understanding of the transmission, recognition of signs and symptoms, perception of the causes, patterns of seeking treatment, inclination towards usefulness of traditional healing methods and attitudes towards putative preventive measures are identified during such investigative surveys (Herbert, Basha, and Thangaraj 2012, Sambo et al. 2014, Matibag et al. 2009). KAP surveys on rabies have been conducted in most rabies endemic regions in the world and have helped to: identify knowledge gaps, cultural beliefs and behavioural patterns of affected communities; design relevant public health awareness programmes; and provide baseline data to plan, implement and evaluate national control programmes for the disease (Sambo et al. 2014).
India contributes 4.4% of the total global research output on rabies, however there is a demonstrable gap between the laboratory based pathogen driven research agenda and studies focussing on the vector demography, risk factors, epidemiological studies and economic evaluations (Kakkar et al. 2012). In a study to examine the gap between research and policy regarding rabies in India, a lack of awareness about the rabies control programmes in the rural population was highlighted (Abbas and Kakkar 2015).
Epidemiological studies on the awareness levels and practices of people, along with
45
research on the implementation of control strategies, are important aspects of developing control programmes for rabies (Abbas and Kakkar 2015, Kole, Roy, and Kole 2014, Burki
2008). It has also been highlighted that, despite increased recognition of rabies as a prioritized zoonosis in India, a lack of knowledge translation activities is a key concern for policy makers (Abbas and Kakkar 2013). There are very few truly representative studies involving a KAP survey of the general community in India regarding rabies. In contrast, there are numerous hospital based studies which mainly focus on the knowledge of dog-bite victims but very few community based studies, thus confirming the findings of Davlin and VonVille (2012) who reported that research on rabies in India focussed on reporting and investigating the epidemiology of dog-bites. The lack of cross-sectional
KAP studies on rabies in the country was a key driver to a part of the research conducted and reported in this thesis.
A systematic literature search was performed as part of this literature review to collate data generated by hospital based KAP studies in rural and urban India during the five year period of 2011 to 2016 (Tiwari et al. 2016) (Table 1.1). It was found that most studies profiled bite victims or assessed their post-bite treatment seeking behaviours when they reported to primary health centres, rural hospitals or Infectious Disease hospitals
(Meshram, Thakre, and Khamgaokar 2016, Ganasva, Bariya, and Shringarpure 2015).
Some studies have also explored the reasons for the delay in reporting to hospitals after a bite event and the initial bite wound management practices adopted (Bharadva et al. 2015,
Shridevi et al. 2014), while other studies have investigated the various traditional bite wound dressing options undertaken (Varsharani, Chinte, and Jadhav 2014, Salve et al.
2015). Although much of the epidemiological work on rabies in India involves retrospective hospital based surveys, a small number of KAP surveys have been
46
performed as cross-sectional studies (Sudarshan et al. 2006, Agarwal and Reddajah 2004,
Herbert, Basha, and Thangaraj 2012).
There have been very few community-based door-to-door KAP surveys undertaken in
India to assess the awareness, attitudes and practices of the residents towards rabies, and virtually none on the perception of urban/rural residents towards FRD. Importantly some recent studies have highlighted critical gaps in the awareness about rabies in urban and rural residents of Delhi (Sharma et al. 2016) and the urban residents of Dehradun district of north India (Ohri et al. 2016). Surveys undertaken in rural areas and among the communities of lower economic status have not only revealed a lack of knowledge about rabies, but have also highlighted inadequate post-dog-bite measures generally adopted by this group (Kamble et al. 2016, Ichhpujani et al. 2006). An important aspect of rabies arising from dog-bites is the perception of the residents towards the FRD population, and although studies exploring this aspect have been undertaken in the neighbouring nation of Nepal (Massei et al. 2017), there is a dearth of such surveys in India. This knowledge gap needs to be addressed to provide empirical basis for the development of effective interventions for India.
47
Table 1.1 List of scholarly articles on knowledge, attitudes and practices of urban and rural communities towards dog-bites and rabies in India published during 2011- 2015
Year of Duration of the study Region Urban/Rural study (months) Sample size Citation 2011 West Urban 3 200 Prakash et al. (2013) 2011 North Urban 10 3525 Mushtaq et al. (2013) 2011 South Urban 4 110 Shridevi et al. (2014) 2012 South Urban 1 100 Patnaik (2013) 2012 West Urban 4 382 Umrigar et al. (2013) 2012 West Rural 3 318 Wankhede et al. (2013) 2012 West Urban 6 223 Varsharani et al. (2014) 2011 -12 North Rural 24 619 Salve et al. (2015) 2012 West Urban 4 410 Ganasva et al. (2017) 2012 West Urban 12 3548 Patil et al. (2017) 2012 -13 West Urban 3 580 Patel et al. (2014) 2012 -13 West Urban 2 357 Patil et al. (2014) 2012 -13 East Rural 12 308 Samanta et al. (2015) 2013 East Urban 1 871 Ghosh and Pal (2014) 2013 North Urban 3 250 Jain and Jain (2014) 2013 East Rural 1 119 Chaudhari (2015) 2013 East Rural 1 102 Santra et al. (2015) 2013 North Urban 12 390 Sahu et al. (2015) 2013 West Urban 3 100 Meshram et al. (2016) Patle and MKhakse 2013 West Urban 6 434 (2014) 2014 West Urban 3 119 Bharadva et al. (2015) 2014 North Urban/ Rural 8 1427/1460 Sharma et al. (2016b) Marathe and Kumar 2014 -15 North Urban 12 406 (2016) Williams and Logaraj * South Rural 3 50 (2012) * South Rural 12 1471 Venkatesan (2014)
* The year of study was not mentioned in the published manuscript
1.10 Strategies for the prevention and control of rabies in India
The control of rabies in a country, such as India, is possible only when interventions are planned and executed based on sound scientific methodology involving sustained operational activities. Such activities are possible only through significant support and input from the national and provincial/state Governments (Meslin and Briggs 2013).
Successful interventions for the elimination of rabies require the immunisation of at least
70% of the FRD to break the transmission cycle of the virus (Franka et al. 2013, Hampson et al. 2009), along with implementing ABC measures to limit the population of FRD
48
(Reese 2005, Totton, Wandeler, Zinsstag, et al. 2010). As well as these operative measures, increasing awareness regarding the disease, its transmission and simple methods to manage dog-bites wounds in people are essential for the successful implementation of a disease control intervention (Lembo et al. 2011, Kole, Roy, and Kole
2014). Availability of post-bite prophylactic measures is another aspect that requires attention from policy makers (Dodet, Goswami, et al. 2008, Gongal and Wright 2011). It has been reported that the hospital staff in India, especially those located in rural areas, lack comprehensive knowledge and are often not trained to deal with dog-bite wounds effectively, which, if improved, could potentially reduce the number of dog-bite related human deaths due to rabies by at least 65% (Burki 2008). Regular training of these para- medical staff to increase their knowledge and practices also forms an effective part of a rabies control strategy (Kishore, Singh, and Ravi 2015, Nguyen et al. 2015, Salahuddin et al. 2011). Implementation of all these measures requires input and expertise in a range of areas including political, administrative, human health and animal health. Rabies is an excellent example of a disease that requires a multi-sectorial “One – Health” approach
(Fitzpatrick et al. 2016, Meslin and Briggs 2013). Control strategies developed for rabies must bring together the findings of research undertaken in different fields, such as laboratory based immunological research to improve the diagnostic capability of the country, and epidemiological studies that investigate factors influencing transmission rates and the FRD population size and structure.
1.11 Aims, scope and relevance of the present study
Rabies is a fatal disease which is not treatable once clinical signs develop; however it is preventable with prophylaxis and timely PEP (WHO 2018). India has the highest number of human deaths globally due to rabies (Kalaivani, Raja, and Geetha 2014) and
49
consequently suffers great losses due to the economic burden imposed by the disease
(Maroof 2013). For rabies to be controlled in India it is important to understand the demographic characteristics of the FRD population, which is a key reservoir for the disease in the country. It is now established that removal of the rabies virus from the most important vector, the dogs, is the first step towards achieving the goal of eliminating rabies from India (Davlin and VonVille 2012, WHO 2005, 2018). Mass vaccination of the dog population against rabies and managing dog populations are largely accepted as effective measures for control of rabies (Conan et al. 2015, Morters, McKinley, Horton, et al. 2014). While such measures cannot be effectively implemented if the population estimate of FRD is erroneous, involvement and cooperation of the local human community can only be possible when correct information about the disease is disseminated so that the community’s knowledge level is accurate and high. Hence, there is a need for studies in India that: investigate the awareness level of urban and rural communities on rabies and FRD; assess FRD demographics and estimate their population size; and explore factors that restrict the accessibility of FRD for vaccination campaigns
(Briggs 2012). The studies reported in this thesis were designed to address these needs and specifically were planned to: compare and recommend a reliable technique to estimate the size of the FRD population in an urban and a rural location of India; record the awareness levels of the rural and urban populations towards rabies and FRD; critically assess the demographic characteristics of the canine population, including their social behaviour; and assess the knowledge, attitudes and practices of frontline medical staff towards rabies and dog-bite wound management.
50
1.12 Objectives
The present study was undertaken in rural and urban localities in western and northern
India with the following aims:
(a) Compare and recommend a reliable enumeration technique for estimating
the size of the FRD population in Shirsuphal village in rural Baramati town in
western India.
(b) Apply the recommended enumeration technique from the rural study
conducted in Shirsuphal village to estimate the FRD population in the urban
municipal area of Panchkula, north India.
(c) Investigate and compare the demographic characteristics of FRD in a
rural (Shirsuphal) and an urban (Panchkula) location in India.
(d) Study the group behaviour and home ranges of FRD in rural (Shirsuphal)
and urban (Panchkula) India.
(e) Assess the knowledge, attitudes and practices (KAP) of the residents of
Shirsuphal village and Panchkula towards rabies and FRD through administering
a questionnaire to a cross-section of the adult general population.
(f) Assess the KAP of para-medical staff employed in Public Health Centres
of Baramati, western India towards dog-bite related rabies.
1.14 The layout and format of this thesis
Rabies is a disease that can be best controlled using the principles of a One Health approach which necessitates adopting a multi-disciplinary approach (Fitzpatrick et al.
51
2016, Cleaveland and Hampson 2017). The work summarised in this thesis incorporates the fields of ecology (population estimation), animal behaviour (FRD home ranges and grouping tendency), epidemiology (KAP analyses) and public health (KAP assessment of medical staff). The thesis comprises a series of manuscripts that have already been published or have been submitted for publication to international peer reviewed journals.
The thesis is presented as an integrated body of related topics united towards the purpose of formulating a strategy for the control of dog-related rabies in India. To maintain the consistency of style and content, each chapter (manuscript) is linked to the previous one through a preface that introduces and maintains the relevance within the overall aims and framework of the thesis. Finally, the last chapter integrates and summarises the findings of the complete work. All chapters, with the exception of the first (Literature review) and the ninth (General discussion), contain their own abstract, introduction, materials and methods, results, and discussion sections as required by the relevant journals the articles have been published in or submitted to. The references for each chapter have been combined into a single reference list and included at the end of the thesis.
In Chapter One, scholarly publications relevant to the objectives of the research work carried out in this thesis have been reviewed. The focus of the review was restricted to general information on the rabies virus, FRD enumeration techniques, KAP studies on rabies and literature on FRD home-ranges and group behaviour.
In Chapters Two and Three the results of surveys comparing the various available enumeration techniques that have been applied for FRD in India and elsewhere are presented. In these two chapters a method to estimate the FRD population size which provides a balance between accuracy and resource inputs is outlined.
52
In Chapter Four the demographics and characteristics of FRD in a rural and an urban setting in India and their relevance to rabies control are outlined. In Chapter Five the demographic details of urban and rural FRD in India relevant to group formation are investigated to allow recommendation of a suitable vaccination strategy (parenteral and/or oral) to achieve adequate vaccination coverage of the FRD population.
In Chapters Six and Seven the KAP of rural and urban communities, respectively, are reviewed towards rabies and FRD. The results of a KAP study of the para–medical staff of rural Primary Health Centres towards rabies and PEP are reported and discussed in
Chapter Eight.
Finally, in Chapter Nine the findings of the entire project, and their limitations and need for future studies are discussed and summarised.
53
Chapter Two
A comparative study of enumeration techniques for free roaming dogs in rural Baramati, District Pune, India
“Not everything that counts can be counted, and not everything that can be counted counts”
Albert Einstein
54
Preface
The majority of human mortality due to rabies infection in India is attributed to dog-bites.
The incidence of human rabies in the country can only be prevented if the disease is controlled in dogs which are the major reservoir host of the rabies virus, however India has the largest canine population in the world of which a large proportion are free roaming dogs (FRD). Animal birth control or capture, neuter, vaccinate and release (CNVR) programmes and mass vaccination of FRD against rabies are the most popular and ethically accepted means to control dog rabies. These approaches have unfortunately failed to achieve success in India because mostly they are applied without a prior knowledge of the population size of FRD. A number of methods, such as simple direct counts to complex capture - recapture techniques, have been used to enumerate FRD, although no one method is considered a gold standard. This study reports on the comparison of eight enumeration methods that used capture-recapture techniques in a rural setting in India. All eight enumeration methods were applied using the same capture- recapture data to obtain and recommend a method that could provide the most reliable estimate of the FRD population to achieve the recommended 70% vaccination coverage using minimal utilisation of resources.
The text of this chapter is the same as the manuscript of the published paper in ‘Frontiers in Veterinary Science’ except for the reference list which has been combined with references of other chapters and incorporated as one list at the end of the thesis.
In addition, Supplementary Figures 2.1 - 2.6 have been added at the end of the chapter to further support the manuscript.
55
The results of this chapter were also delivered as an oral presentation at:
The 15th International Symposium of Veterinary Epidemiology and Economics (ISVEE
15), held in Chiang Mai, Thailand, November 12-16, 2018.
The results of this chapter were also displayed as a poster presentation at:
International Conference on Emerging Infectious Diseases (ICEID 2018), held in Atlanta, Georgia, August 24-27, 2018.
This chapter can be found published as:
Tiwari HK, Vanak AT, O'Dea M, Gogoi-Tiwari J and Robertson ID (2018). A
Comparative Study of Enumeration Techniques for Free roaming dogs in Rural Baramati,
District Pune, India. Frontiers in Veterinary Science 5:104. https://doi.org/10.3389/fvets.2018.00104
56
Statement of Contribution
A Comparative Study of Enumeration Techniques for Free roaming dogs in Rural Baramati, District Title of Paper Pune, India
Publication Status
Tiwari HK, Vanak AT, O'Dea M, Gogoi-Tiwari J and Robertson ID (2018). A Comparative Study of Enumeration Techniques for Free roaming dogs in Publication Details Rural Baramati, District Pune, India. Frontiers in Veterinary Science 5:104. https://doi.org/10.3389/fvets.2018.00104
Principal Author
Name of Principal Author Harish Kumar Tiwari (Candidate) Harish Kumar Tiwari conceptualised and developed the study, planned and conducted the field study, Contribution to the Paper collected and analysed the data, interpreted the results and wrote the paper. Overall percentage (%) 60 % Signature
Date: 15/08/2019
Co-Author Contributions
By signing the Statement of Contribution, each author certifies that: i. the candidate’s stated contribution to the publication is accurate (as detailed above); ii. permission is granted for the candidate to include the publication in the thesis; and iii. the sum of all the co-author contributions is equal to 100% less the candidate’s stated contribution
57
Name of Co-Author Emeritus Professor Ian Robertson Supervised the study and provided critical Contribution to the Paper comments to improve the interpretation of results, edited and revised the manuscript. Overall percentage (%) 15% Signature
Date: 12/09/2019
Name of Co-Author Dr Abi Tamim Vanak Provided critical comments to improve the Contribution to the Paper interpretation of results, edited and revised the manuscript. Overall percentage (%) 10%
Date:15/08/2019 Signature Name of Co-Author Dr Mark O’Dea Provided critical comments to improve the Contribution to the Paper manuscript. Overall percentage (%) 10% Signature
Date: 15/08/2019 Name of Co-Author Dr Jully Gogoi Tiwari Contribution to the Paper Helped editing of the manuscript. Overall percentage (%) 5% Signature
Date:15/08/2019
58
Abstract
The presence of unvaccinated free roaming dogs (FRD) amidst human settlements is a major contributor to the high incidence of rabies in countries where the disease is endemic, such as India. Estimating FRD population size is crucial to the planning and evaluation of interventions, such as mass immunisation, against rabies. Enumeration techniques for FRD are resource intensive and can vary from simple direct counts to statistically complex capture-recapture techniques primarily developed for ecological studies. In this study we compared eight capture-recapture enumeration methods
(Lincoln–Petersen’s index, Chapman’s correction estimate, Beck’s method, Schumacher-
Eschmeyer method, Regression method, Mark-resight logit normal method, Huggin’s closed capture models and Application SuperDuplicates on-line tool) using direct count data collected from Shirsuphal village of Baramati town in Western India, to recommend a method which yields a reasonably accurate count to use for effective vaccination coverage against rabies with minimal resource inputs. A total of 263 unique dogs were sighted at least once over 6 observation occasions with no new dogs sighted on the 7th occasion. The methods that do not account for individual heterogeneity yielded population estimates in the range of 248-270, which potentially underestimate the real
FRD population size. The highest estimates were obtained with the Huggin’s Mh-
Jackknife (437±33), Huggin’s Mth-Chao (391±26), Huggin’s Mh-Chao (385±30), models and Application “SuperDuplicates” tool (392±20). When the sampling effort was reduced to only two surveys, the Application SuperDuplicates online tool gave the closest estimate of 349±36, which is 74% of the estimated highest population of free roaming dogs in
Shirsuphal village. This method is considered to be the most reliable method for estimating the FRD population with minimal inputs (two surveys conducted on consecutive days).
59
2.1 Introduction
Free roaming dogs (FRD) are responsible for attacks on humans and other animals, damage to property, road accidents, contaminating the environment with faeces, spreading garbage waste and causing noise pollution (Beck 1973, Rinzin, Robertson, and
Mahat 2017). There has been a rapid increase in the number of dogs during the last decade in India, with a concurrent increase in the number of dog-bites of humans (Bradley and
King 2012, Davlin and VonVille 2012). The large number of unrestricted, unowned, free roaming dogs within the country is responsible for 99% of all dog-bite transmitted rabies in humans (Knobel et al. 2005, Meslin and Briggs 2013). A large uncontrolled population of free-roaming canines is also damaging to their own welfare (Butcher 1999, Rinzin,
Robertson, and Mahat 2017), as a lack of veterinary care leaves these dogs malnourished and often suffering from diseases and injuries (Totton et al. 2011, Butcher 1999).
Interventions for rabies control are feasible for household pets as they generally receive adequate veterinary attention, however, such care is difficult for FRD (Morters et al. 2013,
Kitala et al. 2002). The interventions usually applied to control rabies and to decrease the
FRD population include culling, mass vaccination and sterilisation (Morters et al. 2013,
Høgåsen et al. 2013). However culling does not result in a sustained reduction in the number of FRD (Beran and Frith 1988, Windiyaningsih et al. 2004), and the efficacy of sterilisation on population control remains debatable (WHO 2006, Franka et al. 2013).
There is a growing unanimity among researchers that mass vaccination is the best way to eradicate dog-bite related rabies (Franka et al. 2013, Cleaveland, Beyer, et al. 2014) and it is generally agreed that successful mass vaccination campaigns require 70% coverage of the dog population to achieve critical herd immunity against the disease (WHO 2013,
Knobel et al. 2005). However, a lack of information about the true population size of FRD raises doubts about the coverage of mass vaccination campaigns in many locations
60
(Wallace, Etheart, et al. 2017) and restricts critical assessment of disease intervention and population control measures and welfare issues relating to FRD (Conan et al. 2015).
Although knowing the size, dynamics and demographics of the target FRD population prior to the implementation of an intervention and for post-intervention assessment is crucial (Slater 2001, Meslin, Fishbein, and Matter 1994), there is no accepted standardised enumeration technique.
Formulating an enumeration methodology for FRD is very challenging not only in countries where registration and licensing of dogs is not mandatory (Özen, Böhning, and
Gürcan 2016), such as India, but even in countries where registration is mandatory, e.g. estimating population of free-ranging dogs in Australian indigenous communities.
Various studies have used rate of capture (regression method), Beck’s method, (Fei et al.
2012), distance methods (Childs et al. 1998), extensive counts in the chosen areas and extrapolation of this number (WSPA 2010, Tenzin, Ahmed, et al. 2015, Tenzin,
McKenzie, et al. 2015), mark-resight surveys (Punjabi, Athreya, and Linnell 2012,
Tenzin, McKenzie, et al. 2015), Huggin’s closed capture techniques (Belsare and
Gompper 2015) and Schumacher-Eschmeyer method (Totton, Wandeler, Zinsstag, et al.
2010) to estimate the FRD population. There is also growing acceptance that methods for estimating the population size of wild animals yield reliable results when applied to FRD
(Belo et al. 2015, Belsare and Gompper 2013). However, no researchers have critically evaluated and compared the different evaluation methods. As the main purpose to know the FRD population is to achieve effective vaccination coverage to eliminate rabies in
India, rather than to accurately enumerate the population per se, the methods used should consider the time and monetary constraints involved, while still being reliable.
61
This study was undertaken in a rural setting of Shirsuphal village of Baramati Town in western India to (1) compare the estimates of the FRD population obtained with different analytical methods; (2) study the impact of extrinsic abiotic factors including temperature, humidity and wind velocity on FRD counts; and (3) recommend an enumeration technique that allows for rapid, yet robust population estimates to determine the number of FRD requiring vaccination against rabies to achieve the 70% vaccination coverage.
2.2 Materials and Methods
2.2.1 Study area
The study was conducted in the Shirsuphal village of Baramati town located in Pune
District of Maharashtra State, India in June 2016. The village comprises patches of human settlements interspersed with farmlands (Figure 1) that are connected through 16 km of roads, of which 12 are bitumen. In June, the temperatures in Baramati ranges vary from
230C to 320C, with an average humidity of 72%
(https://www.timeanddate.com/weather/india/baramati). Agriculture is the mainstay of the economy of the village with a number of poultry farms around the village that have been established over the last five years. The major land cover categories consist of agricultural fields, grazing land and protected reserve forests. No prior dog population control campaign had been undertaken in the sampled area (personal communication with village administrative head).
2.2.2 Field methodology
Beck's definition of an FRD, “Any dog observed without human supervision on public property or on private property with immediate unrestrained access to public property”
(Beck 1973), was used in this investigation (Berman and Dunbar 1983). Any dogs that were restrained or confined were excluded from the study.
62
The study was conducted from June 5th to June 13th, 2016 with surveys undertaken during the early mornings and late afternoons of alternate days. As photography was used to identify dogs, to ensure adequate light the surveys were conducted between 7 and 9 am and 4 to 6 pm. No surveys were conducted on the 10th and 11th June due to heavy rainfall.
Surveys alternated between mornings and afternoons on five consecutive days (5th - 9th
June) and again on the 12th and 13th June.
Two teams of two individuals each were trained to carry out the surveys on motorcycles.
They were assigned separate predetermined routes covering all the human settlements in the village. Team A rode a track of 7.52 km divided into two sub-tracks (A1 and A2) while Team B covered 5.93 km on four sub–tracks (B1, B2, B3 and B4) (Figure 1). The rider was trained to take a photograph and record the GPS waypoints while the pillion passenger completed the data sheets to record various characteristics of the encountered dog and its corresponding camera picture number. The individuals, their duties and the route ridden remained the same throughout the study. The survey was ceased subsequent to no new dogs sighted.
Each team was equipped with a motorcycle, a Garmin eTrex20 GPS device
(www.garmin.com), a digital camera and a clipboard with datasheets and writing materials. Both teams started the surveys at exactly the same time on each sampling occasion and travelled on the pre-determined tracks at a speed of ~20 km/h. During the counting sessions, teams attempted not to disturb the natural behaviour of dogs by not driving too close to the animals while still maintaining their pre-set route. The teams recorded waypoints and the sex (male/female/not verifiable), age (pup/young/adult/old), size (small/medium/large), coat pattern (solid/bicoloured, tricoloured/mixed), primary and secondary colours of the coat, coat condition (good/average/poor), reproductive
63
status (lactating/pregnant/oestrus), and overall health assessment (good/average/poor, presence of lameness, dermatitis or any other disability) of observed dogs.
Figure 2.1 Google earth imagery (www.googleearth.com) of the village landscape and the various tracks used by the observation teams for survey (accessed on 22/07/2016).
A1 (5.88km) A2 (1.64 km) B1 (1.23 km) B2 (3.2km) B3 (1km) B4 (0.5km) The light-yellow lines depict the roads in areas of no human settlements. Depicts the border of Shirsuphal village.
2.2.3 Animal identification and capture histories
In order to avoid recording the same dog by both teams each counting session was followed by ruling out any double counts that may have occurred due to the movement of dogs across the tracks. Individual animal photographs were examined and tallied with the physical attributes recorded in the datasheet. Overall, the individual identity of 98.2
% (617 out of total 628 sightings from both routes) of the capture events was agreed upon by the teams to be included in constructing capture histories. Each animal was given an identity number depending on the route and the date of capture. The sighting or absence
64
of a previously or subsequently sighted dog was recorded as either 1 or 0, respectively for each session of the survey conducted.
2.2.4 Data analysis
Data were recorded in Microsoft Excel (Microsoft Excel, 2013, Redmond, USA).
Program MARK (www.phidot.org/software/mark/docs/book/) was used to estimate the population size of the closed population using Huggin’s heterogeneity models. The
“appropriate” option was selected from the Program CAPTURE option to select the suitable estimator (White 2008, Belsare and Gompper 2013). The same software was also used to estimate the population size using a Mark-resight logit normal model (McClintock
2011). Regression analyses, Pearson’s correlation tests and χ2 tests were performed in R
(R Development Core Team 2013).
2.2.5 Population estimation methods
Eight capture-recapture probability techniques were used to estimate the FRD population in this study. In addition, a direct cumulative visual count of all dogs encountered during all sessions was the naïve estimate or direct count. A multiple linear regression analysis was performed to examine the effect of temperature, humidity and wind velocity at the time of the surveys on the number of sightings.
2.2.6 Capture-recapture (C-R) techniques
All the methods used in this study used the capture-recapture technique where the animals were not marked but were photographed and matched with the photographs taken on other sampling days. Most of the basic assumptions of C-R techniques, such as the interval period between surveys and complete mixing of the surveyed population, were met (Fei et al. 2012, Belsare and Gompper 2013). Further, the data generated in this study were subjected to test of equal catchability and test of closure (Southwood and Henderson
65
2009, Page 76-83). Equal catchability of animals was assessed by calculating the G- statistic from the observed and expected number sighted during the sampling period and comparing this with the critical value of a Chi-squared distribution and a two-tailed ranked correlation test between the percentage of FRD re-sighted and the sequence of sampling occasions was used to assess if there was any trend in the numbers sighted across sessions (Jeremy and Robert 2006). Leslie’s test for equal catchability was used to calculate the expected variance (σ2) and the Chi-square value (χ2) to check for the probability of occurrence at p<0.05 for FRD known to be in the population during the survey period (Orians and Leslie 1958, Southwood and Henderson 2009). The testing for the closure of the population was based on the logic that the proportion of animals re- sighted on successive occasions would decline if a population was not closed. The
Spearman’s rank correlation with the null hypothesis (H0 ) that, such a decline occurs in the observed data set was used to test for the closure of the population (Jeremy and Robert
2006). The following paragraphs describe the methods followed in this study.
2.2.6.1 Regression method
The linear regression of the number of dogs captured on each survey session on the estimates of the total number of distinct dogs captured until the previous session yields the population estimate (Fei et al. 2012, Jeremy and Robert 2006, Tenzin, McKenzie, et al. 2015). The catchability/detectability (k) on each occasion is taken as the absolute value of the intercept (b).
2.2.6.2 Lincoln-Petersen Index and Chapman’s correction method
The Lincoln-Petersen Index and Chapman’s correction capture-recapture methodology estimate the population size based on the principle that the proportion of animals re- sighted in a subsequent sample are a proportion of the marked population as a whole
66
(Sekar and Deming 1949, Tenzin, McKenzie, et al. 2015, Jeremy and Robert 2006).
Chapman (1951) correction was applied to remove the bias resulting from using the
Lincoln-Petersen’s estimate (Tenzin, McKenzie, et al. 2015). Six estimates were obtained each for the Lincoln-Petersen Index and Chapman’s correction (EC) from each successive pair of sightings and re-sightings. To study the temporal variation, a set of two estimates for morning surveys and three for late afternoon surveys were calculated. The two methods were then compared using a two-sample independent t-test.
2.2.6.3 Beck’s method
Beck’s method (Beck 1973) is an extension of the Lincoln-Petersen’s approach to multiple captures which takes into account successive recaptures following an initial effort (Amstrup, McDonald, and Manly 2010). The estimate is obtained by dividing the summation of the product of total sighted and the cumulative total marked animals at large by the total number of re-sighted on each occasion (Fei et al. 2012, Krebs 1999).
2.2.6.4 Schumacher - Eschmeyer method
The Schumacher-Eschmeyer method states that if the total number of marked individuals is plotted against the proportion of marked samples in the tth sample, the graph should be a straight line passing through the origin (푥=0, 푦=0) with a slope of 1÷N, where N is the total population (Schumacher and Eschmeyer 1943). A failure in linearity of the plotted lines implies that one or more assumptions of the closed capture method have been violated. However, if fulfilled, the N can be estimated using linear regression techniques
(Totton, Wandeler, Zinsstag, et al. 2010, Southwood and Henderson 2009).
67
2.2.6.5 Mark – resight logit normal method
The mark-resight method takes into account the individuals that remain undetected due to individual heterogeneity and thus constitute slightly different data than for traditional methods of mark-recapture (McClintock 2011). Amongst the various models available, the logit-normal mark-resight estimator for individually identifiable animals with replacement was used (McClintock 2011, page 18-8). This method is suitable for FRD as marks are individually identifiable and the number of individuals of the primary subset are known and sighting is done with replacement (Punjabi, Athreya, and Linnell 2012).
The Program MARK software with logit—normal estimator was used with models derived from a combination of available or fixed parameters. Time constant models with and without individual heterogeneity (p=pij σij=σ N(t) and p=pij σij=0 N(t)) were run. The sin link function was used for all model runs. The model yielding the smallest Akaike’s
Information Criteria (AIC) was chosen from the available model-run options to obtain the estimate (Punjabi, Athreya, and Linnell 2012, Tenzin, McKenzie, et al. 2015, McClintock
2011).
2.2.6.6 The Closed capture Huggin’s heterogeneity model
The encounter histories were analysed using the feature CAPTURE available within the
Program MARK software. The sampling design approach was similar to Horvitz-
Thompson’s model as individual dogs have an unequal probability of being re-sighted
(Belsare and Gompper 2013, Huggins 2001). The software allows the use of Huggin’s p
(initial capture probability) and c (recapture probability) data type to obtain heterogeneity models which are then read with a suitable estimator (White 2008). The “appropriate” option was selected to obtain the most suitable estimator for the population size (Alho
1990, Huggins 1989, Huggins 2001, White, Burnham, and Anderson 2001). However, in
68
order to fulfil the aim of the study of drawing comparisons between estimates the Program
CAPTURE was run with all possible model-estimators (Mh-Jackknife, Mh-Chao, Mth-
Chao and M0), besides the one recommended by the program.
2.2.6.7 Application SuperDuplicates (AS)
The Application SuperDuplicates (AS)is a tool derived from the formula developed by
Alan Turing and his colleague IJ Good that the number of uniques/singletons (individuals that appear only once during the whole sampling exercise) holds all the information required about the undetected individuals and it was adapted by Chao et al. (2017) to assess the species richness in a given area. It is also based on probability of an individuals’ re-sighting and we extended this technique to enumerate the population size of FRD. Each individual dog sighted at least once was counted as a unique species (Sobs – species observed). The tool utilises two kinds of data to estimate the population size: “incidence data” which are a record of the presence or absence of each observed individual in repeated samples (count and frequency), and “abundance data” which are a record of an individual observed in a single sample (Bunge 2013). Both abundance data and the incidence data can be used to estimate the population size. In abundance data nomenclature, the individual observed in only one sampling unit is called a singleton, one that is seen in exactly two sampling units is called a doubleton and an individual seen in more than two sampling units is called a super-doubleton. The corresponding terms for the incidence data are unique, duplicates and super-duplicates. The inputs required for the abundance data are the total number of individual observations (Sobs); and the number of singletons (f1). The corresponding input requirements for incidence data are the total number of individual observations (Sobs); the number of uniques (Q1); and the number of sampling units conducted. The input was entered into the online tool
69
https://chao.shinyapps.io/SuperDuplicates/ to estimate the population size and the percentage and number of undetected individuals (Chao et al. 2017).
2.3 Ethical approval
Ethics approval for this study was granted by ATREE (Ashoka Trust for Research in
Ecology and the Environment) Animal Ethics committee (AAEC) via their approval letter number AAEC/101/2016.
2.4 Results
2.4.1 Sighting variability between sessions
A total of 617 reliable sightings of FRD consisting of 263 unique dogs were recorded during seven surveys undertaken over the nine-day study period. The number of unique
FRD reached saturation on the 6th session with no new dogs sighted on the 7th session.
The lowest count (52) was observed on the last day of the study (7th session). Wind velocity during the time of the survey had a strong negative correlation (r = -0.92, p<0.01) with the number of dogs sighted in a counting session. Other meteorological variables, including temperature at the time of survey (r= -0.07, p= 0.39) and humidity (r= +0.07, p=0.42), did not have a significant impact on the count, irrespective of the time of the survey (Table 2.1).
70
Table 2.1 Details of climatic characteristics and the number of free roaming dogs sighted at each survey session during the photographic capture-recapture survey at Shirsuphal Village during 5 - 13 June 2016.
Total number of Time of Temperature Humidity Wind velocity Weather dogs Date count (*C) (%) (Km/h) condition sighted 5/06/2016 Evening 32 55 7 Sunny 93 6/06/2016 Morning 26 80 2 Overcast 106 7/06/2016 Evening 32 55 6 Overcast 103 8/06/2016 Morning 27 78 6 Overcast 91 9/06/2016 Evening 35 42 4 passing clouds 90 12/06/2016 Evening 30 59 13 passing clouds 82 13/06/2016 Morning 30 70 19 passing clouds 52
(Source: https://www.timeanddate.com/weather/india/baramati )
2.4.2 Test for equal catchability and closure of population
The catchability rate was not the same across the survey period as the G statistic value
(24) was significantly higher than the critical χ2 value of 12.6 (p = 0.0013). The data also failed the two-tailed Spearman's rank correlation test for equal catchability as the rcritical value (0.886) was less than the Spearman’s rank coefficient (0.48; p= 0.32); andLeslie’s test was significant (p = 0.0005). However, the population was verified to be closed as the data passed the test for the closure of population (Spearman's rank correlation coefficient, r = 0.35 was smaller than rcritical value for a one-tail test (0.829, p=0.44).
2.4.3 Regression method
The dog population was estimated to be 282 (95%CI 265-304, p<0.001) using the regression method (푦 = -0.3287푥 + 92.792, R2= 0.987). The overall detection probability
(P) was found to be 0.33. The estimate when only the morning data were used was 267
(95%CI 200-1752, p = 0.04, P=0.36) compared with 278 (95%CI 274-323, p < 0.001,
P=0.31) when only the afternoon data were used (Figure 2.2).
71
Figure 2.2 Prediction of population size of free roaming dogs by regression method for all sessions of FRD survey in Shirsuphal
2.4.4 Lincoln-Petersen Index (L-P) and Chapman’s corrected (C) estimator
The estimates by the Lincoln–Petersen’s index (EL-P) and Chapman’s correction estimator
(EC) for each session are shown in Table 2.2. The Pearson’s correlation test between the capture probability (P) for each set of counts and the estimate demonstrated a non- significant, weak negative correlation (r = -0.27, p=0.6). The estimates from the two techniques were similar (two-sample t-test, t=0.18, p=0.85).
2.4.5 Beck’s method (Schnabel’s multi-capture method)
The Beck’s estimate of the population was 276 (95%CI 244-317) when all 6 multiple re- sighting sessions subsequent to the initial sighting session were used. When influence of the temporal factor on sampling during a fixed time of the day was assessed, morning surveys resulted in a population estimate of 259 (95%CI 193-392), compared with 290
(95%CI 236-375) from the afternoon sessions (p=0.67).
72
2.4.6 Schumacher-Eschmeyer method
The test for the Schumacher-Eschmeyer’s method demonstrated data validity as the line obtained by plotting the number of marked (Mt) dogs to the ratio of the total dogs sighted:
2 re-sighted (Ct/Rt) was linear (푦=0.0036푥, R =0.9623) and passed through the intercept.The population size was estimated at 270 (95% CI 235-317).
2.4.7 Logit-normal mark-resight method
The logit-normal mark-resight method was applied with one primary and six secondary sessions. The 93 dogs sighted and photographed on the 1st day were taken as the initially marked and identified individuals. During the subsequent 6 secondary sampling sessions
61, 59, 50, 50, 49 and 20 individuals were counted as unmarked but sighted (total = 289), with 45, 44, 41, 40, 33 and 32 marked individuals sighted on each session, respectively.The estimates along with Akaike Information Criteria (AIC) scores and the mean overall sighting probability (µ) for both models are summarised in Table 2.3.
2.4.8 Huggin’s heterogeneity models using Program CAPTURE
All data types selected under Huggin’s closed capture models yielded exactly the same results. No estimators and models were suggested for any of the data types and the models with the highest weights were Mth (0.99), Mbh (0.47) and Mh (0.40) when data from all seven days of the survey were used. The estimates after conducting the survey until saturation as derived from various estimators, along with the measure of p (capture probability) for all possible model-estimator combinations, are outlined in Table 2.4.
2.4.9 Estimation using the Good-Turing frequency formula using AS tool
Estimation by Good-Turing frequency theory with singleton observations (abundance data) yielded a value of 392 ± 20 (95%CI 358-437) when data from all seven sessions were considered (Sobs=263 and singletons, f1=118, undetected number = 129 (32.85%),
73
duplicates=53). The estimate for seven days of counting using the uniques observation
(incidence data) was 375±18 (95%CI 344-416) with an undetected percentage of 29.8%
(112) (Table 2.5).
2.5 Discussion
In this study, we compared techniques used in ecological studies for enumeration of FRD in a rural setting, using the same operators, materials, and temporal and geographical settings to identify a robust method for estimating the population size. A reliable estimate of population size is a vital requirement for effective implementation of control measures of diseases such as rabies.A protocol that was comparable to other studies (Belsare and
Gompper 2013, Tenzin, McKenzie, et al. 2015) was used that standardised the efforts across the survey period. A direct count of the FRD, along with documentation of their characteristics, was found to be an effective and simple method to individually identify the FRD within the selected area. The counting of dogs along frequently used routes is important from resident’s point of view as these routes /roads are also locations where people are more likely to be bitten (Hiby and Hiby 2017). In our study data collection was discontinued after no new FRD were sighted following seven days of observations conducted over nine days to enable comparisons of population estimates with different sampling efforts and methods.
74
Table 2.2 Size of the free roaming dog population estimated by the Lincoln–Petersen index and Chapman’s correction (EC) with counts on successive days during photographic capture-recapture survey in Shirsuphal
Days 1&2 Days 2&3 Days 3&4 Days 4&5 Days 5&6 Days 6&7 µ* All surveys ELP(95%CI) 219 (184-254) 254 (209-299) 247 (199-294) 248 (194-302) 194 (160-229) 124 (106-144) µ = 215 ± 49 EC(95%CI) 214 (180-247) 248 (205-291) 241 (195-286) 242 (190-293) 189 (156-222) 121 (102-139) µ = 209 ± 49 P*= 0.48P*= 0.41P*= 0.37 P*= 0.36 P*= 0.42 P*= 0.41 Late afternoon surveys Morning surveys Days 1&3 Days 3&5 Days 5&6 ELP(95%CI) 234 (192-275) 211 (177-244) 194 (160-229) Days 2&4 Days 4&7 EC(95%CI) 228 (188-268) 205 (173-238) 189 (156-222) µ = 186 ± 54 224 (192-262) 148 (122-173) µ = 246 ± 20 P*= 0.44 P*= 0.43 P*= 0.42 µ = 181 ± 54 219 (182-255) 142 (118-167) µ = 207 ± 20
µ* is the mean of the estimates. P* is the re-sighting probability of each session which is exactly the same for EC and ELP
75
Table 2.3 Comparison of the models run using the Logit-normal mark-resight method on the basis of the Akaike Information Criteria (AIC) for FRD enumeration in Shirsuphal
Parameters Model used AIC score* N ± SE (95% CI) µ
Time constant with heterogeneity 334±18 N, µ [ pij=p, σij=σ, N(t)] 936.6 (307-379) 0.16
Time constant without heterogeneity 334±9 N, µ [ pij=p, σij=0, N(t)] 1336.9 (318-354) 0.16
Time constant without heterogeneity and with fix capture probability 326±89 N, µ [pij=p=0, σij=σ=0, N(t)] 110486.29 (271-755) -
* AIC score for Time constant model with heterogeneity is smaller and hence this represents the best model. µ is the overall mean sighting probability across the primary session and it remains the same even when heterogeneity is fixed. When capture probability is fixed to be constant for all secondary sessions and heterogeneity (σ) is assumed to be not present, then the estimate (N) was a plausible value but not acceptable as the AIC score was high.
Table 2.4 Population estimates and calculated capture probability as obtained by available estimators* under Program CAPTURE for FRD at Shirsuphal
Model Estimator Estimate ± SE (95% CI) Capture probability (p) #
Mth Chao 391± 25.79 (350-452) 0.24, 0.27, 0.26, 0.23, 0.23, 0.21 Mh Jackknife 437± 32.57 (385-513) 0.20 Mh Chao 385±29.83 (295-340) 0.23 M0 Null 283 ± 5.48 (274-295) 0.31
* Mbh model was not considered as behavioural variation was mitigated by photographic capture- nd th recapture. # Chao’s estimator for model Mth presents the capture probability for the 2 to 7 session respectively.
76
2.5.1 Sighting variations
The decline in the number of sightings over the seven survey days was influenced by climatic (rains preceding the surveys) and local factors (community event in the adjacent village) during the last two days of the survey (Table 2.1). Unexpectedly, Daniels (1983) found that heavy rains had no noticeable effect on the behaviour of FRD. The heavy downpour in the current study could be a confounding factor as it was also accompanied by strong winds that appear to reduce dog activity as shown by lower counts on days when high wind velocity was recorded. While there was no correlation between the dog count and the ambient temperature or humidity at the time of the survey, more dogs were counted on overcast days than on clear days. Others have also reported an increase in dog activity with increasing cloud cover (Daniels 1983, Dias et al. 2013). The occurrence of rains and accompanying stormy weather prior to the last two surveys and a local religious community feast in the adjacent village (as informed by the village head), a day prior to the last survey may have resulted in reduced activity of dogs in the survey location, as
FRD moved away. This finding highlights the need to consider forecasted weather, as well as human events, during the planning of enumeration surveys. The attentiveness of the observer has also been highlighted as a factor that affects detectability (McCallum
2005). Conducting more surveys may introduce fatigue amongst observers, resulting in a decrease in counts as the study progresses.
2.5.2 Capture-Recapture (C-R) techniques
The assumption of a closed population was validated in this study as proportion of re- sightings across the sessions did not vary significantly. This confirmed the suitability of
C-R techniques for estimating the FRD population due to the short duration of this survey.
However, the catchability rate did vary significantly across the survey sessions due to
77
differences in the recaptures (re-sightings) arising from weather conditions and sociological factors (organised community events) may influence the count on a particular day. Conducting counts during the mating season, when males may move greater distances, could also result in differences in counts. As the methodologies used in this study considered the re-sighted number without accounting for such extrinsic factors, the assumption of equal catchability across the survey sessions was not unexpectedly violated.
2.5.3 Types of C-R estimation techniques
2.5.3.1 Methods not based on individual identity
The Lincoln-Petersen’s index and the Chapman’s corrected method have advantage over other methods as they require just two sighting sessions. The estimates of both NL-P and
NC were lower than NB, NS-E and NR which is supported by the findings of a comparative study in Brazil (Shimozako 2012). We established that the estimate is influenced by counts from each day but not by re-sighting probabilities (r = -0.27, p = 0.6), implying that the estimate may still vary irrespective of identical sighting probabilities (Table 2.2).
The number of re-sights oscillates with the total sightings of the day, so that if extrinsic factors, such as weather conditions, result in a large drop in the total count, the corresponding re-sights will also decrease, and thus their reliability for determining the population size for adopting a mass vaccination program is questionable. Thus, we do not recommend the Lincoln-Petersen’s index or the Chapman’s corrected estimator for FRD.
78
Table 2.5 Population estimates of free roaming dogs using Application SuperDuplicates for sampling occasions ranging from 2 to 7
Number of Number of Abundance data Incidence data sample units singletons/uniques Estimate ± SE (95%CI) Undetected % (N) Estimate ± SE (95%CI) Undetected% (N) 7 118 392 ± 20 (358-437) 33 (129) 375 ± 18 (344-416) 30 (112) 6 123 404 ± 20 (370-450) 35 (141) 380 ± 20 (347-426) 31 (117) 5 125 390 ± 21 (354-437) 36 (144) 357 ±18 (326-400) 31 (111) 4 121 374 ± 24 (334-428) 40 (150) 328 ± 20 (296-375) 32 (104) 3 116 354 ± 27 (309-415) 45 (158) 285 ± 18 (255-328) 31 (89) 2 109 349 ± 39 (287-441) 56 (195) 220 ± 19 (192-268) 30 (66)
79
Table 2.6 Summary of the population estimates obtained using 11 different methods until saturation (7 survey occasions spread over 9 days)
Method Estimate ± SE (95%CI) Direct method 263 Lincoln-Petersen’s estimate 254 (209-299) Chapman’s correction 248 (205-291) Beck’s method 276 (244-317) Schumacher-Eschmeyer’s estimate 270 (236-317) Regression method 282 ± 94 (265-304) Log-Normal Mark Re-sight method 326 ± 15 (303-364) Huggins Mth (Chao estimator) 391 ± 26 (350-452) Huggins's Mh (Jackknife estimator) 437 ± 33 (385-513) Huggin's Mh (Chao estimator) 385 ± 30 (340-458) Good-Turing (Application SuperDuplicates) 392 ± 19 (358-437)
In comparison, the regression method was found to be robust with a higher count (NR) than other multi-capture methods yielding higher precision and smaller SE than NB (Table
2.6). This is contrary to a study by Fei et al. (2012), where simulations for NR and NB were compared and NB was found to be a better estimator as NR failed to deliver reliable estimates at low capture probabilities, even when the number of survey occasions was higher. However, in this study, the NR was higher than NB, even when the capture probability (k= 0.382) was much smaller than the mean P for Beck’s method (0.629) with the same sampling effort. Beck’s and the Schumacher-Eschmeyer’s methods resulted in a comparable estimate of N (NB = 276, NS-E = 270) (Table 2.6). This was not unexpected as both methods rely on individuals re-sighted on successive sessions.
The Schumacher-Eschmeyer’s method applied by Totton, Wandeler, Zinsstag, et al.
(2010) to study the effect of sterilisation on the population size of FRD in Jodhpur, India, emphasised the graphical test to ensure non-violation of assumptions without describing them. Assuming similar assumptions as for closed populations, the straight line plotted signifies the relationship between the fractions of identified animals on each sampling to the total number of animals identified prior to that sampling and the slope of the line of
80
best fit gives NS-E. A practical aspect of capture-recapture in FRD, especially in countries with large dog populations such as India, is that there would seldom be a sampling occasion with no or a very small fraction of re-sightings. As a consequence, the line of best fit terminates sooner (when y=0, or when saturation is reached) as compared to other species (e.g. fish) where the point of saturation is difficult to meet. The inappropriateness of the Schumacher-Eschmeyer method was highlighted by Belo et al. (2015) who considered this method was more appropriate for aquatic species rather than terrestrial ones, especially dogs. Beck’s method has been used for estimation of feral dog populations in Brazil, Mexico and India (Belsare and Gompper 2013, Belo et al. 2015,
Dias et al. 2013). As Beck’s estimator also relies on the ratio of marked animals sighted on the sampling day to the cumulative number of animals sighted before the start of that sampling, predictability is limited by a strong plausibility of reaching saturation. Thus, both these methods tend to underestimate the population size of FRD as dogs are not difficult to resight in rabies endemic areas. This is also supported by the finding that the estimates using this method are similar to the total number of FRD actually sighted at least once during the survey (N=263). As the primary aim of this study was to arrive at an estimate for effective vaccination coverage (70%) to keep R0 <1, the difference between the estimates is not high enough to instil any confidence on deciding the number for effective vaccination coverage.
2.5.3.2 Methods using individual identity
We applied the Logit-normal mark-resight model as the sightings were accomplished without replacement, the intervals between surveys were small but uniform and repeat sightings of an individual during a particular sampling session were negligible. The survey design used in this study allowed only one primary sighting survey. If the survey
81
continued long enough to start another primary, the interval may have violated the assumption of a closed population. The estimate was found to be comparatively precise
(small SE) compared to the Huggin’s heterogeneity models but its accuracy may still be debatable, as only one primary was conducted. The estimate was much lower than the
Huggin’s model (Table 2.7), thus raising concerns about it being an under-estimate.
In the case of the CAPTURE feature of Program MARK, abundance is predicted by the conditional estimator model, which uses capture histories of individuals seen at least once, was selected (Huggins 1989, Lukacs 2009). This method was used by Belsare and
Gompper (2013) usingthe Mh model to estimate the population size of FRD in a nearby locality in India during a mass vaccination programme. Using the Jackknife estimator that study concluded that Huggin’s closed capture models yielded higher population estimates than the Beck’s estimator as the latter was found to be even lower than the number of identified dogs that were vaccinated.
In the current study, however, the feature CAPTURE could not suggest an appropriate estimator, which was likely due to the nature of the input data, e.g. the time variation is evident for the last session when the count was the lowest, but reasons were not temporal but environmental (heavy rains) and sociological (community feast). The Program
CAPTURE algorithm nevertheless identified it as a temporal and/or behavioural factor as apparent from the rank and weight of the models (Mth =0.99, Mbh=0.47, Mh= 0.4). The second-best weight was given to model Mbh indicating that model selection was influenced by the sharp drop in the count on the last day. As the behavioural variation
(generally used for trap-shy or trap-happy behaviour) is more of a capturing effort attribute (White 2008) and it is negated by using photographic capture-recapture, considering the Mbh model for population estimation would be misleading. The best-
82
suggested model, Mth (weight= 0.99), and Mh (weight = 0.4), were run using Chao and
Jackknife estimators, respectively. The Mh model with Jackknife estimator obtained a higher estimate (437±33) than Mth with Chao (391±26), however, the capture probability of Mh –Jackknife (p = 0.20) was lower than the average capture probability of Mth-Chao
(p=0.24). This implies that, implying that Mh-Jackknife overestimates the population size if nearly all animals are captured (Amstrup, McDonald, and Manly 2010, page 63), which could be the case in this study where most animals were captured at least once. However, if the survey was continued after the 7th survey, it would result in increasing the number of re-sights without adding to the total unique individuals sighted and tend to reduce the estimate. Hence, we recommend stopping the survey, once saturation is reached.
The “abundance data” estimates from the Good – Turing formula using the Application
SuperDuplicates (Chao et al. 2017) after seven sampling surveys were identical with the estimates resulting from the Huggin’s model Mth with estimator Chao (Table 2.7). The
“incidence data” may appear to be a better model as it considers the number of sampling units and this study found that the duplicate estimates were closer to the actual figures than the doubleton estimates; the population size estimate, however, was less than the
“abundance data”. Although the difference between the estimates may not appear substantial assuming them to indicate the true population, the difference widens with reduction of sampling efforts (Table 2.5) which doesn't augur well if we want to get a reliable estimate with minimal efforts (counts). This suggests that “abundance data” is a better option over “incidence data” to generate a reliable estimate of FRD.
2.5.4 Comparison of the methods used in this study
The primary aim of this study was to recommend an enumeration method that could provide a reliable estimate of the population size of FRD to achieve effective vaccination
83
coverage against rabies. The methods that do not include heterogeneity and assume equal catchability across resight surveys are not reliable estimators and thus are not recommended for FRD population size estimation. Among the C-R methods which include the influence of heterogeneity, the Logit-normal mark-resight method could only be run with one primary and provided an estimate lower than other C-R methods, and thus is unsuitable for FRD estimation for rabies control. Thus, Huggin’s heterogeneity models and the AS tool are considered acceptable methods for estimating the FRD population. As these methods were run on data using exactly the same resource input, a comparative study of their respective estimates with diminishing resource input helps in the selection of the method that would give a reliable population estimate (Table 2.7).
The estimates derived from the various methods (Table 2.7), clearly show that the estimates decrease across all methods on the 7th survey. This is because the saturation point was reached on the 6th survey and further re-sights do not add any new information.
While this keeps the pi (individuals not seen at all) part of the data unchanged while reducing p (seen once), the overall estimate tends towards stabilisation. It can, however, be inferred that the estimate after the 6th survey is the highest possible, albeit a likely slight overestimation.
Further examining the trends with reduced inputs (Table 2.7), we find the estimates decrease when data from fewer surveys are included, except for model Mth-Chao (due to reduced temporal information available to the Program CAPTURE algorithm with two/three surveys). This helps remove Mth model from the choice of methods to use as our endeavour is to obtain a reliable estimate using a minimum number of surveys.
The remaining methods (Mh-Jackknife, Mh-Chao and Application SuperDuplicates) also show diminishing estimates with a reduced number of surveys, however, there is a wide
84
variation in the rate at which they fall (Figure 2.3). The steepest fall is seen by the Mh –
Jackknife model with an estimate of 207±9 compared to Mh –Chao (286±34) and AS tool
(349±39) at minimum input effort. A smaller estimate is not unexpected as the CAPTURE algorithm has less data that adversely affects its accuracy. The negative bias of the model
Mh- Jackknife was explained by Chao (1987) who reported that smaller sampling efforts
(<5) reduced its precision. Otis et al. (1978, page 34)proved that the Mh-Jackknife has a tolerable bias when trapping occasions are sufficiently large (>5), indicating that the estimate by this method after six surveys (467±34) may be closer to the true population.
This rationale was used by Belsare and Gompper (2013) to recommend Mh-Jackknife as the most appropriate model-estimator to assess the population size of FRD in India.
The Mh –Chao yielded an estimate of 286±34, which, although better than Mh-Jackknife combination, is still a smaller number if we consider the Mh-Jackknife with six surveys as approximating the true population. It has been mentioned that Chao’s estimator works well with sufficient capture occasions (usually >5) (Amstrup, McDonald, and Manly
2010, page 69-73, Chao, Lee, and Jeng 1992). Consequently, this estimate does not provide us with a reliable estimate to achieve 70% herd immunity and thus the Mh-Chao is also not a recommended method in this situation.
85
Table 2.7 The population estimates obtained by the Huggin’s heterogeneity models compared with Application SuperDuplicates (AS) online tool based on Good-Turing frequency formula on successive reduction of sampling efforts
ESTIMATES (numbers) Number of survey effort Mh-Jackknife± SE 95% CI Mh-Chao± SE 95% CI #Mth-Chao±SE 95% CI *AS±SE 95% CI 2 207±9 193-228 286±34 235-371 349±39 287-441 3 302±15 277-335 321±31 274-396 493±103 347-772 354±27 309-415 4 371±21 336-418 352±31 305-428 371±34 318-455 374±24 334-428 5 429±28 383-492 384±33 333-465 390±27 343-460 390±21 354-437 6 467±34 410-546 400±27 356-464 400±33 350-480 404±20 370-450 7 437±33 385-513 385±30 340-458 391±26 350-452 392±20 358-437
# The Mth-Chao model could not project any estimates after single re-sight survey due to lack of temporal data *Application SuperDuplicates
86
Figure 2.3 Graphical representation of the trend of population estimates using Huggin’s models and Application SuperDuplicates (AS) with the number of survey sessions
550
467 437 450 429 404 374 390 392 371 Mh, JK 349 354 350 Mh,Chao 302
size Mth,Chao
roamingdogs population - M0 250 207 AppSup
150 Estimate Estimate ofFree 1 2 3 4 5 6 7 8 Number of Survey Efforts
Mh, JK= model Mh-Jackknife, Mh, Chao = model Mh-Chao, Mth, Chao= model Mth-Chao, M0 = model M0 and AppSup= Application SuperDuplicates
The AS tool developed by Chao et al. (2017) based on Good-Turing theory, however, obtains a sufficiently large estimate (349±39) with only two surveys or with one set of sight-resight data. Considering Mh-Jackknife estimate after six surveys (467±34) to be close to the true population, the “abundance data” estimate by AS tool is 74% of the true population. However, before recommending this as the best method to estimate the population size of FRD, it is pertinent that we crosscheck the reliability of estimates by the AS tool in comparison to other model-estimator combinations that are known to be accurate with sufficient (>5) sampling events. We found that the AS-tool estimate after seven surveys (392±20) was similar to the Mth-Chao (391±26) and Mh-Chao (385±30), although lower than Mh-Jackknife (437±33). It is interesting to note that barring Mh-
87
Jackknife, other estimates are similar after six surveys as well (Table 2.7). However, we can infer that the AS-tool estimate, after a single set of a sight-resight exercise, is a robust estimator of at least 70% of the true population size of the FRD in Shirsuphal village. The introduction of AS tool by Chao et al. (2017) based on the Good-Turing theory, fortunately, solves most of the complexities of choosing the datatypes, model and estimator selection associated with program CAPTURE and obtains a dependable number to work towards achieving 70% vaccination coverage in FRD against rabies. Nonetheless, there is an inherent shortcoming of the AS software, namely that the output generated for the SE and 95% CIs differ slightly on repeats, even when exactly the same data are entered
(Chao et al. 2017); this, however, does not affect the total estimate.
The major limitations of this study are firstly the absence of a gold-standard estimate against which the estimates can be compared with and thus we had to be content with the largest estimate, which could be an overestimate of the true population. Secondly, as these techniques have been developed and perfected for wildlife enumeration, we surmise the sighting probability for FRD in human frequented habitat is influenced by factors probably not applicable to other ecological studies. Finally, we admit that the study is based on limited tracks as the inclusion of the interior tracks of the village was not possible due to limited resources and time constraints. However, we recommend extensive surveys should be carried out in the future to generate a gold-standard estimate to compare and validate the findings.
In this study, we compared most of the available enumeration methods that can be applied for estimating populations of FRD, except for the Spatially Explicit capture-recapture
(SECR) and methods based on Bayesian models. The former could not be used as it requires intensive spatial preparation of the data in a GIS, and the latter needs use of
88
reliable priors which are not present for FRD in the study location. The distance method was not used as it is more appropriate for density estimates rather than total population size and because dogs are not distributed randomly in relation to the transect lines (Belo et al. 2015). The importance of the various models of Huggin’s closed capture models cannot be overlooked for future studies where empirical priors can be evolved for population size estimations based on Bayesian models. We recommend further testing the applicability of the Application SuperDuplicates software for FRD enumeration study in different locations and conditions.
Author contributions
All authors have contributed and approve the contents of this article. HT developed the study, collected and analysed the data. HT, IR, AV, MO and JT wrote the article, provided critical revision and helped interpretation of contents and implications.
Funding
The work was funded by the Wellcome Trust-DBT India Alliance Program through a
Fellowship to ATV (Grant number: IA/CPHI/15/1/502028) and the Research grant to HT from Murdoch University, Western Australia, Australia.
Acknowledgements
Murdoch University International Postgraduate Scholarship (MIPS) to HT is gratefully acknowledged. The authors are grateful to Abhijeet Kulkarni, Pranav Panwalkar, Reetika
Maheshwari, and Pradeep Satpute for helping with the data collection and the residents of Shirshuphal village for allowing us to conduct this survey.
89
Conflict of Interest Statement
The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
90
Supplementary Figure 2.1 Graphical representation of estimates of free roaming dog population size in Shirsuphal village by different methods
Supplementary Figure 2.2 The devices used and a glimpse of the field observation process
A B C
A&B GPS devices used to record the way points. C – Survey team recording the sighting of a free roaming dog
91
Supplementary Figure 2.3 Six examples (A-E) of descriptions of free roaming dogs as recorded by the observer team.
A B
C D
E F
A- sex-NV (not verified), pup small, solid colour (cream), body condition poor, coat condition fair, active and no garbage in vicinity; B- male, adult, large, solid colour (brown), body condition good, coat condition poor (dermatitis), active and no garbage in vicinity; C- female, adult (lactating), large, bicoloured (primary-black, secondary-white), body condition good, coat condition good, active, no garbage in vicinity; D- male, adult, medium, solid colour (brown, muzzle black), body condition fair, coat condition good, active, garbage in vicinity; E- male, old, large, solid colour (black), body condition fair, coat condition fair, not active, no garbage in vicinity; F- A closer picture of the same dog as E, male, old, large, grey hairs on muzzle, solid colour (black), body condition fair, coat condition fair, not active, no garbage in vicinity
92
Supplementary Figure 2.4 The sightings of the free roaming dogs and garbage points in Shirsuphal village during the dog enumeration survey
A E
B F
C G
D H
A –G represent the locations of the Free Roaming Dogs sighted from the first to the seventh occasion of the survey. A& B have the end points of the selected routes depicted as H - Shows the locations of the points where garbage was found littered more frequently. In A and B represent the end points of the tracks
93
Supplementary Figure 2.5 Testing the fit of the Schumacher-Eschmeyer method by plotting number of previously marked FRD against proportion of marked FRD sighted during each survey effort during the enumeration survey in Shirsuphal
1.2 y = 0.0036x R² = 0.9623 1
0.8
0.6
0.4
sightings sightings (Rt/Ct) 0.2
Proportionofmarked individuals 0 0 50 100 150 200 250 300 350 Number of previously marked FRD (Mt)
94
Chapter Three
Validation of Application SuperDuplicates (AS) enumeration tool for free roaming dogs (FRD) in urban settings of Panchkula Municipal Corporation in north India
“Correct and improve by validation alone.”
Meir Ezra
95
Preface
An estimate of the size of the FRD population for mass vaccination against rabies is required to evaluate the potential effectiveness of the intervention. To achieve adequate herd immunity that can break the transmission cycle of the rabies virus, at least 70% of the dogs should be vaccinated. In the preceding chapter probabilistic models (Huggins’s closed capture heterogeneity models) that incorporated the variability in the re-sighting of a FRD due to temporal, behavioural and individual heterogeneity factors were found to obtain realistic estimates of the population size, although they are resource intensive.
Such huge resource investment in the form of conducting at least five photographic sight– resight surveys is mitigated by using the Application SuperDuplicates shinyapp which requires only two surveys to obtain a reliable estimate of at least 70% of the FRD population in the surveyed area. Although the study compared all available methods from the sight-resight data obtained from the same rural location to recommend Application
SuperDuplicates as the method of choice to estimate the FRD population size with minimal resource investment, it could be argued that results may differ between settings.
Validation of the Application SuperDuplicates method requires repeating the survey in more than one location and further comparison with the other probabilistic models. In this chapter, photographic sight-resight surveys were conducted on a number of survey tracks in the sectors administered by the Municipal Corporation, Panchkula in the Haryana state of north India to evaluate the suitability of the Application SuperDuplicates method to enumerate FRD in an urban setting.
The text of this chapter is the same as the manuscript of the published paper in ‘Frontiers in Veterinary Science’ except for the reference list which has been combined with references of other chapters and incorporated as one list at the end of the thesis. In
96
addition, Supplementary Figures 3.3 and 3.4 have been added at the end of the chapter to further support the manuscript.
This manuscript has been selected as a poster presentation at:
3rd International Conference on Humane Dog Population Management, 18-20 September
2019, Mombasa, Kenya.
This chapter is published as:
Tiwari HK, Robertson ID, O'Dea M, Gogoi-Tiwari J, Panvalkar P, Bajwa RS and Vanak
AT (2019). Validation of Application SuperDuplicates (AS) Enumeration Tool for Free roaming dogs (FRD) in Urban Settings of Panchkula Municipal Corporation in North
India. Frontiers in Veterinary Science 6:173. https://doi.org/10.3389/fvets.2019.00173
97
Statement of Contribution
Validation of Application SuperDuplicates (AS) Title of Paper Enumeration Tool for Free roaming dogs (FRD) in Urban Settings of Panchkula Municipal Corporation in North India.
Publication Status
Tiwari HK, Robertson ID, O'Dea M, Gogoi-Tiwari J, Panvalkar P, Bajwa RS and Vanak AT. (2019) Validation of Application SuperDuplicates (AS) Enumeration Tool for Free roaming dogs (FRD) in Publication Details Urban Settings of Panchkula Municipal Corporation in North India. Frontiers in Veterinary Sciences 6:173. https://doi.org/10.3389/fvets.2019.00173
Principal Author
Name of Principal Author Harish Kumar Tiwari (Candidate) Harish Kumar Tiwari conceptualized and developed the study, planned and conducted the field study, Contribution to the Paper collected and analysed the data, interpreted the results and wrote the paper. Overall percentage (%) 60 %
Signature Date: 08/08/2019
Co-Author Contributions
By signing the Statement of Contribution, each author certifies that: i. the candidate’s stated contribution to the publication is accurate (as detailed above); ii. permission is granted for the candidate to include the publication in the thesis; and iii. the sum of all the co-author contributions is equal to 100% less the candidate’s stated contribution.
98
Name of Co-Author Emeritus Professor Ian Robertson Supervised the study and provided critical Contribution to the Paper comments to improve the interpretation of results, edited and revised the manuscript. Overall percentage (%) 15% Signature
Date: 12/09/2019
Name of Co-Author Dr Abi Tamim Vanak Provided critical comments to improve the Contribution to the Paper interpretation of results, edited and revised the manuscript. Overall percentage (%) 10%
Date: 12/08/2019 Signature Name of Co-Author Dr Mark O’Dea Contribution to the Paper Provided critical comments to improve the manuscript. Overall percentage (%) 7.5% Signature
Date: 24/08/2019 Name of Co-Author Dr Jully Gogoi Tiwari Contribution to the Paper Helped editing of the manuscript. Overall percentage (%) 2.5% Signature
Date: 12/08/2019 Name of Co-Author Dr Rajinder Singh Bajwa Contribution to the Paper Contributed during field study to collect data. Overall percentage (%) 2.5 %
Signature Date: 18/08/2019
Name of Co-Author Mr Pranav Panvalkar Contribution to the Paper Contributed to collect and collate the field data. Overall percentage (%) 2.5 %
Signature Date:20/8/2019
99
Abstract
A cost-effective estimation of the number of free roaming dogs is an essential prerequisite for the control of rabies in countries where the disease is endemic, as vaccination of at least 70% of the population is recommended to effectively control the disease. Although estimating the population size through sight-resight based maximum likelihood methodology generates an estimate closest to the actual size, it requires at least five survey efforts to achieve this. In a rural setting in India a reliable estimate of >70% of the population of free roaming dogs was obtained with the Application SuperDuplicates shinyapp online tool using a photographic sight-resight technique through just two surveys. We tested the wider applicability of this method by validating its use in urban settings in India. Sight-resight surveys of free roaming dogs were conducted in 15 sectors of the Panchkula Municipal Corporation in north India during September- October 2016.
A total of 1408 unique dogs were identified through 3465 sightings on 14 survey tracks.
The estimates obtained by the Application SuperDuplicates shinyapp online tool after two surveys were compared with the maximum likelihood estimates and it was found that the former, after two surveys, provided an estimate that was >70% than that obtained by the latter after 5-6 surveys. Thus, the Application SuperDuplicates shinyapp online tool provides an efficient means for estimating the minimum number of free roaming dogs to vaccinate with a considerably lower effort than the traditional mark-resight based methods. We recommend use of this tool for estimating the vaccination target of free roaming dogs prior to undertaking mass vaccination efforts against rabies.
100
3.1 Introduction
Free roaming dogs (FRD) are a serious public health problem in most urban societies of the developing world (Beck 1973, Berman and Dunbar 1983, Rinzin 2015) and play an important role in the spread of dog-bite related rabies in countries where the disease is endemic (WHO 2018). Local governments apply several strategies, including fertility control programmes (Jackman and Rowan 2007) and removal of shelters, to limit the size of the FRD population. However, due to infrequent and/or sporadic implementation, these methods have had little impact on the population size of FRD (Taylor et al. 2017). Mass vaccination of FRD against rabies has been advocated as a practical and effective intervention to prevent dog-bite related rabies in countries where it is endemic (Franka et al. 2013, Hampson et al. 2009, WHO 2018). However, mass vaccination campaigns require 70% coverage to develop critical herd immunity against the virus in the target
FRD population (WHO 2013, Knobel et al. 2005, Taylor et al. 2017), and a lack of information about the true population size of FRD casts uncertainty over the efficacy of such campaigns (Wallace, Etheart, et al. 2017).
The significance of reliably estimating the FRD population size is important to: assess the impacts of rabies control interventions, dog population management and effect on dog welfare; and to reduce the threat to wildlife (Vanak and Gompper 2009b, Slater 2001, Fei et al. 2012, Belo et al. 2015). Enumeration of the FRD population, however, is a time and labour-intensive exercise. An accurate estimate of the number of FRD at any given time depends on a number of factors, such as socio-economic status, cultural and social beliefs of the human population and characteristics of the habitat that influence the dynamics of the FRD population (WHO 2005, Butcher 1999, Wandeler 1985). Many enumeration methods have been employed by researchers to estimate the FRD population in different
101
parts of the world, although the accuracy of such estimates can be questionable
(Shimozako 2012, Belo et al. 2015). In the absence of a gold standard, population estimation methods used for wildlife that are based on the maximum likelihood estimates
(MLE) using photographic sight-resight surveys have found acceptance for enumeration of FRD (Belsare and Gompper 2013). However, these methods require multiple surveys to obtain robust estimates. Methodology that can provide a reliable estimate of the FRD population in an area using a minimum of resources is better suited to the effective implementation of a mass vaccination programme against canine rabies (Tiwari et al.
2018).
In an earlier study, Tiwari et al. (2018) compared eight methods to enumerate FRD and found that the methods that do not take into account the individual heterogeneity of the dogs potentially underestimate the population size. The Huggin’s heterogeneity models
(Mh or Mth) with suitable estimators (Jackknife/Chao) were shown to yield robust estimates depending upon if the surveys had reached saturation or not. The number of enumeration surveys conducted is central to the robustness of the estimates, with at least five surveys required to obtain an estimate close to the true population (Amstrup,
McDonald, and Manly 2010, 69). However, conducting surveys for 5-6 occasions is not only challenging and resource intensive, but may also result in bias from surveyor fatigue
(McCallum 2005). To overcome these challenges, an online tool based on the Good-
Turing formula to assess species richness, called Application SuperDuplicates (AS), has been shown to provide a robust estimate that is equal to or greater than 70% of the FRD population size with only two consecutive surveys (Chao et al. 2017). Tiwari et al. (2018) concluded that the AS tool (https://Chao.shinyapps.io/SuperDuplicates/) could be used to estimate the minimum target FRD population requiring vaccinating with minimal resource implications to control dog-bite related rabies. However, that study only
102
assessed the AS tool in a single rural location. In this study, we test if the AS tool is as effective in urban areas where there are more complex environments and higher dog densities. Assuming that the MLE using heterogeneity models with suitable estimators provides an estimate closest to the true population size, we compare this method with the
AS tool to estimate the FRD population size in the urban sectors of the Municipal
Corporation of Panchkula in northern India.
3.2 Materials and Methods
3.2.1 Study area
The study was conducted in the wards administrated by the Municipal Corporation,
Panchkula in Haryana state in north India during September-October 2016. The municipal area is divided into 20 wards that are subdivided into sectors which contain residential, administrative or industrial sectors or unorganised colonies (slums) (Duggal 2004). The selected sectors/colonies were surveyed through predetermined roads/tracks frequented by humans and FRD. A total of 15 sectors were selected from 7 wards using purposive
(industrial, administrative and mixed sectors) and random (residential sectors including unorganised colonies or slums) sampling. Two of the administrative sectors were surveyed together (sectors 1&5) resulting in a total of 14 survey tracks. Eight residential sectors (sectors 2, 7, 8, 9, 12, 16, 17 and 18) were randomly selected from the 16 available in the 7 selected wards. A mixed residential sector (sector 6) that contained public amenities including a public park, an educational institute and a hospital was also included. All the industrial sectors (IAP I and IAP II) and the unorganised colonies
(Budhanpur - BP, Rajeev Colony - RC, Indira Colony - IC) in the Panchkula Municipal
Corporation area were included in the study. The administrative and industrial sectors and the unorganised colonies do not have walled perimeter boundaries; however, the
103
movement of FRD between the residential sectors is restricted by solid brick perimeter walls. To account for the presence of FRD on the roads connecting to the residential sectors, we also surveyed the surrounding roads of one of the sectors (Sector 8(P)).
3.2.2 Field methodology
The field methodology is described in detail in Tiwari et al. (2018). Briefly, the selected sectors were traversed by a team of two observers riding a motorcycle at a constant speed of ~20 Km/hour following a predetermined route (survey-track) on alternate mornings and evenings for five or six occasions. The teams were equipped with a GPS device
(Garmin eTrex20 GPS device, www.garmin.com), a digital camera and writing materials.
The survey-tracks and the teams remained unchanged throughout the survey period. The
FRD count survey was conducted during mornings (6 - 8 AM) and afternoons (4 - 6 PM), except for the three unorganised colonies (BP, RC and IC) where only morning counts were undertaken. The survey was conducted in each sector along the connecting bitumen roads.
3.2.3 Data entry and analyses
The sighting/re-sighting of a FRD was recorded as a “1” and if that FRD was not sighted on a subsequent survey it was recorded as a “0” in the latter survey(s). The capture history was constructed for all FRD sighted during the survey. Leslie’s test was used to test equal catchability of FRD by calculating the G-statistic from the observed and expected number sighted during the sampling period and comparing this with the critical value of a Chi- squared distribution (Orians and Leslie 1958). Program MARK software
(www.phidot.org/software/mark/docs/book) was used to analyse the recorded capture histories that were run with Huggin’s closed capture heterogeneity models (Mh and Mth) using Jackknife and Chao estimators. The re-sighting probabilities obtained from
104
Program MARK were used to calculate the coefficient of variation (CV) of the re-sighting probability across the surveys in a sector. The survey data were assumed to be saturated if no new FRD were sighted during the last survey. The capture histories of the first two consecutive surveys were used for obtaining the population estimate using the online AS shinyapp tool (https://Chao.shinyapps.io/SuperDuplicates/) (Chao et al. 2017). As the estimate of the true FRD population size was not known, the percentage coverage for each of the survey-tracks for first two surveys was calculated on the basis of MLE (Mth-
Chao and Mh-JK estimate for the survey tracks with and without saturation, respectively).
Odds ratios and their 95% confidence intervals were calculated using the “odds ratio” package in R (Patrick 2017).
3.3 Results
A total of 1408 unique FRD were sighted at least once across the 14 survey-tracks in
Panchkula (total of 3465 sightings). The estimates of the FRD population in the various sectors through the MLE method after 5-6 surveys ranged from 44 to 212, while the estimates after two surveys using the AS shinyapp tool were 49 – 342. Details of the sightings in each sector, along with the climatic conditions on the day of the survey, are presented in the Supplementary Table 3.1. All the assumptions for capture-recapture techniques were met, except the test of equal catchability for which the G-statistic value was greater than the critical χ2 value at p=0.05 for k-1 degrees of freedom in 6 of the sectors included in the survey (Table 3.1). The odds of failing the assumption of equal catchability was significantly higher for survey tracks on which lower FRD counts were observed in the afternoon compared to the morning (OR 1.18, 95% CI 1.03-1.36, p= 0.02).
The CV of the re-sighting probability ranged from 0.11 to 0.36 and the survey-tracks with
CV values ≥ 0.27 failed to conform to the assumption of equal catchability for these
105
surveys (Table 3.1). The percentage population coverage by the two surveys used to estimate population size of the FRD through the AS shinyapp tool for each survey-track is presented in Table 3.2.
A point of saturation (when no new FRD were sighted on subsequent surveys) was reached in 7 of the 14 survey tracks. The estimates of the FRD population size in the surveyed sectors by the using Huggin’s heterogeneity models (Mh and Mth with estimators Jackknife and Chao, respectively) and with the online AS shinyapp tool
(https://Chao.shinyapps.io/SuperDuplicates/) are presented in Table 3.3.
Table 3.1 Results of Leslie’s test for equal catchability of free roaming dogs in the different sectors of Panchkula during the enumeration survey carried out during September-October 2016
Survey- Number Coefficient of tracks G value Critical χ2 value of surveys P value Variation (CV)** 8 7.08 9.49 5 0.13 0.19 2 3.69 9.49 5 0.45 0.12 12 12.55 9.49 5 0.01^ 0.28 IAP* 1 4.08 11.07 6 0.53 0.11 IAP* 2 6.43 11.07 6 0.27 0.13 BP,IC,RC# 1.36 9.49 5 0.85 0.11 9 31.35 11.07 6 0.0001^ 0.36 17 4.06 11.07 6 0.54 0.19 16 16.07 11.07 6 0.03^ 0.27 1&5 6.93 9.49 5 0.13 0.17 8 (P)@ 1.68 11.07 6 0.9 0.13 18 15.92 9.49 5 0.0031^ 0.32 6 12.79 11.07 6 0.02^ 0.30 7 12.29 11.07 6 0.03^ 0.27
*IAP = Industrial Area Part; # Budhanpur, Indira Colony, Rajeev Colony; @ Sector 8 perimeter; ^Significant p values indicate that the catchability was not equal between the sampling sessions, ** Coefficient of variation for the re-sighting probability
106
Table 3.2 The Coefficient of Variation, population coverage and corresponding difference in the estimate by Maximum Likelihood and Application SuperDuplicates shinyapp tool during the free roaming dog enumeration surveys in 14 survey tracks in Panchkula
Population Number of unique Population size based on Population size based on coverage after first 2 Difference in the dogs identified after Maximum Likelihood Application SuperDuplicates surveys estimates Survey - tracks CV† 2 surveys Estimate (MLE) estimate (AS) (% of the MLE) (AS – MLE) 1,5 0.17 89 212 221 42 9 2 0.12 97 152 186 64 34 6 0.3 56 142 184 39 42 8 0.19 77 146 190 53 44 8(p) * 0.13 28 44 49 64 5 IAP@2 0.13 108 172 208 63 36 BP, RC, IC # 0.11 45 131 125 34 -6 7 0.27 39 89 85 44 -4 9 0.36 82 114 130 72 16 12 0.28 67 100 145 67 45 16 0.27 82 142 342 58 200 17 0.19 41 78 185 53 107 18 0.32 62 135 162 46 27 IAP@1 0.11 99 190 193 52 3
* Sector 8 perimeter; @ IAP - Industrial Area Part; # Budhanpur, Indira Colony, Rajeev Colony; † Coefficient of variation for the re-sighting probability
107
Table 3.3 Estimates of the free roaming dog population size by Maximum Likelihood (5-6 surveys) and Application SuperDuplicates shinyapp (first 2 surveys) using sight-resight techniques in different sectors of Panchkula following enumeration surveys carried out during September-October 2016
Total number of Number of Huggin's Mh (JK) Huggin's Mth (Chao) AS shiny app tool unique dogs singletons (5-6 surveys) (5-6 surveys) (2 surveys) observed observed Survey tracks (estimate± SE) † 95% CI (estimate± SE) † 95% CI (estimate± SE) †† 95% CI (2 surveys) (2 surveys) Survey tracks where new FRD were sighted during the last survey (recommended model + estimator is Mh -JK) 1,5 212±16 188-253 198±15 176-235 221±36 166-313 89 66 2 152±7 142-170 146±8 137-168 186±22 152-240 97 61 6 142±16 119-182 140±19 114-193 184±51 99-279 56 46 8 146±11 131-174 141±11 127-170 190±32 121-199 77 57 8 (P)* 44±4 40-59 44±4 40-58 49±11 36-81 28 16 IAP@ 2 172±9 160-196 164±7 154-183 208±24 170-267 108 68 BP, RC, IC# 131±15 108-168 122±19 96-173 125±70 79-233 45 35 Survey tracks where no new FRD were sighted during the last survey (recommended model + estimator is Mth -Chao) 7 100±12 84-134 89±8 79-113 85±21 64-114 39 27 9 117±8 107-141 114±7 105-133 130±12 111-161 82 41 12 106±6 97-122 100±7 92-122 145±23 107-172 67 46 16 154±13 136-188 142±10 129-169 342±97 149-868 82 71 17 78±7 69-96 78±9 68-104 185±130 73-697 41 36 18 140±14 120-176 135±15 114-176 162±29 118-238 62 47 IAP@ 1 201±8 188-221 190±8 180-211 193±23 158-249 99 63
* Sector 8 perimeter; @IAP = Industrial Area Part; # Budhanpur, Indira Colony, Rajeev Colony; † Estimates after 5-6 surveys; †† Estimate after two surveys
108
3.4 Discussion
We estimated the population size of FRD in 15 sectors of the Municipal Corporation,
Panchkula using photographic capture-recapture through MLE and multiple sight-resight surveys and compared the estimates with those obtained after only two surveys using the online AS shinyapp tool. Since individual dogs were not physically captured but photographed from a distance without disturbing their natural behaviour, any variation in the count due to behavioural attributes, such as being “trap-happy” or “trap-shy” is excluded in this study. The MLE were obtained using the Huggin’s heterogeneity models, namely Mh and Mth with most appropriate estimators (Jackknife/Chao) in Program
MARK after 5-6 surveys (Cooch and White 2006). However, after only two surveys the online AS shinyapp tool consistently obtained estimates of the population for all sectors that were ≥ 70% of the MLE estimates.
We sighted more dogs in the mornings than in the afternoons (Supplementary Table 3.1).
This temporal variation is not unexpected because FRD appear to avoid the heavy motor traffic and enhanced human activity during the times when the afternoon sessions were undertaken (Dias et al. 2013). In contrast no temporal variation was observed in the study of Tiwari et al. (2018) in the rural settings as there was no apparent change in the level of human activity or traffic between the morning and afternoon sessions.
The count-data of the FRD in this urban sight-resight survey conformed to all the requisite assumptions of the capture-recapture technique, except for the “assumption of equal catchability” in 6 of the 14 survey-tracks (Table 3.1). This could be due to the significant difference between counts in the morning and the afternoon sessions. The likelihood of obtaining a decreased FRD-count in the afternoon was significantly higher (OR 1.18) in
109
the survey tracks that failed the equal catchability assumption than those where the assumption of equal catchability held true. As well as the presence of less vehicular traffic in the mornings than in the afternoons of residential sectors, residents in the urban areas were observed to feed the FRD during the morning surveys but not during the afternoon surveys. This is not surprising because during a knowledge, attitudes and practices (KAP) survey of the urban residents towards FRD conducted in parallel to this study, 72% of the respondents advised they would feed a FRD (unpublished data) with most of these (77%) fed leftover food to the FRD. Such practices would attract FRD, resulting in increased sightings during the morning surveys. However, no temporal variation in FRD sightings was observed in sectors 2, 8 and 17 (Table 3.1). These sectors are located near highways and active market areas, where FRD are frequently sighted, irrespective of the time of the day.
The survey-tracks in six sectors (residential - 7, 9, 12, 16, 18 and mixed - 6) with lower p values for the Leslie’s test imply unequal individual catchability during survey sessions and not surprisingly also had higher CV indicating a wider spread of the re-sighting probability (Table 3.1). These tracks have a noticeable temporal difference of FRD counts between morning and afternoon sessions (Supplementary Table 3.1) compared to the rural survey in Shirsuphal (Tiwari et al. 2018).
The key finding of this study, however, is the applicability of the online AS shinyapp tool to quickly and reliably obtain a minimum target for vaccination coverage after just two surveys (Table 3.3). Interestingly, the online AS shinyapp tool estimates exceeded the
MLE of the Huggin’s heterogeneity models (except for survey track 7; and the unorganised colonies (BP, RC, IC) for which only morning sessions were run) which is not unexpected as the temporal variation of the FRD counts was wider in Panchkula.
110
Another factor that might influence the performance of the AS shinyapp tool is the population coverage during the two surveys used for calculating the estimate. Given that the true population size of FRD is unknown, the MLE is the closest approximation of true population size, albeit with bias. Assuming that the population size of FRD in the survey areas is fixed, the bias in the MLE should decrease as the coverage increases with successive surveys. The MLE by the Mth model overestimates the population size when the CV < 0.4 and underestimates when the CV ≥ 0.4 (Chao, Lee, and Jeng 1992). As none of the sight-resight data from surveyed sectors used for the Huggin’s models had a CV value ≥ 0.4, we infer that the true population size of FRD may actually be less than the
MLE (Table 2.3). This finding is of concern for at least two of the survey tracks, namely
16 and 17 where the AS shinyapp tool estimate exceeded the MLE estimate by a large margin (Table 2.3) and may lead to over-estimation of the vaccination effort required than is actually needed in the area.
Unfortunately, our investigation did not observe any pattern that could relate the relationship of the CV between counts (implying catchability of FRD), percentage of population coverage (after two surveys) and the margin of difference of MLE with the
AS shinyapp tool estimate (Table 3.2). A plausible cause for such a large margin of difference in the estimates could be the pattern in which new animals were sighted in these survey tracks. It is therefore necessary to also examine the drivers that affect the detection probability of FRD in an area to elucidate the reasons for such large variations between sectors.
As observed by Tiwari et al, (2018) in rural areas, Huggin’s heterogeneity model with
Jackknife estimator (Mh - JK) potentially overestimated the population size of FRD when the count reaches saturation (no new FRD sighted in the last count). Also, the estimates
111
obtained by the online AS shinyapp tool were always >70% of the MLE of the FRD population size from two consecutive surveys (morning and afternoon sessions).
However, we found that the performance of the online AS shinyapp tool depends on the
CV of the resight probability.
We restricted the number of surveys to 5 or 6 in the urban FRD enumeration as recommended by Chao, Lee, and Jeng (1992), Otis et al. (1978). We followed the recommendation for using Huggin’s heterogeneity models where the sampling efforts should not be less than 5 for closed populations if the capture probabilities are relatively small (Chao 1987). Nonetheless, half the surveyed tracks reached saturation and thus we compared the AS estimates of those sectors with estimates of model-estimator combination of Mth-Chao.
The population estimate by the AS tool depends on the first two counts and if the number of singletons is large after two surveys, or conversely, if the number of individuals re- sighted on the second survey is small, the estimate tends to be larger. We conclude that the AS estimates that can be obtained through only two capture-recapture surveys provides an estimate that is greater than 70% of the highest MLE estimate of the FRD population size. This can thus be used as a surrogate for determining the minimum number of dogs to vaccinate to ensure effective vaccination coverage against rabies.
In addition to accessibility of the FRD for immunisation (Ratsitorahina et al. 2009) and a shortage of resources (Wallace, Undurraga, et al. 2017), many vaccination campaigns fail to achieve the required 70% annual immunisation coverage due to a lack of information regarding the true population size of FRD in the area (Wallace, Etheart, et al. 2017). Based on Tiwari et al. (2018) in a rural area, and this study in an urban area, we thus recommend
112
that the AS online shinyapp can be used to reliably obtain a minimum estimate of FRD populations for planning mass vaccination programmes in India and other countries.
Author contributions
All authors have contributed and approve the contents of this article. HT conceived and developed the study; HT, RB and PP collected and analysed the data. HT, IR, AV, JT and
MO wrote the article, provided critical revision and helped interpret the results and implications.
Funding
The work was funded by the Wellcome Trust-DBT India Alliance Program through a
Fellowship to ATV (Grant number: IA/CPHI/15/1/502028).
Ethical approval
This study involved survey of free roaming dogs in Panchkula Municipal Corporation administrated wards in the state of Haryana, India. The administrative approval of
Panchkula Municipal Corporation was obtained for the study while the ethical approval was granted by ATREE (Ashoka Trust for Research in Ecology and the Environment)
Animal Ethics committee (AAEC) vide their approval letter number AAEC/101/2016.
Acknowledgements
Murdoch University International Postgraduate Scholarship (MIPS) to HT is gratefully acknowledged. The authors are grateful to the Municipal Commissioner, the Principal and students of Government Girls College Sector-14, Panchkula for all the help to
113
conduct the study. The help and support rendered by Mr Surender Dhiman is also gratefully acknowledged.
114
Supplementary Table 3.1 Details of the number of free roaming dogs sighted during the sight-resight surveys conducted in fourteen sectors of Panchkula Municipal Corporation administrated areas of Haryana state, India along with the meteorological data for each survey session.
Unique Count Wind Survey Total Temp Humidity dogs Date Session each velocity Climate track sightings (0C) (%) sighted day (km/hr) 8 250 112 16/09/2016 M 53 26 85 4 Clear 18/09/2016 A 44 34 57 no wind passing clouds 21/09/2016 M 58 28 88 no wind Sunny 22/09/2016 A 37 29 67 4 passing clouds 24/09/2016 M 58 26 82 4 Sunny 2 313 127 6/10/2016 M 58 24 85 no wind Clear 6/10/2016 A 55 32 56 no wind Sunny 7/10/2016 M 75 25 85 no wind Fog 7/10/2016 A 61 31 61 no wind Sunny 8/10/2016 M 64 24 85 no wind Fog 12 199 86 12/10/2016 A 36 31 45 2 Sunny 13/10/2016 M 52 22 71 no wind Clear 14/10/2016 M 51 21 72 no wind clear 14/10/2016 A 33 30 44 4 Sunny 15/10/2016 A 27 30 43 2 Sunny IAP 1 452 168 3/10/2016 M 64 28 88 no wind Fog 3/10/2016 A 71 33 63 no wind Haze 4/10/2016 A 76 32 68 no wind Fog 5/10/2016 M 83 27 79 no wind Fog 5/10/2016 A 85 32 66 4 Scattered clouds 6/10/2016 M 73 24 85 no wind Clear IAP 2 451 144 8/10/2016 M 68 24 85 no wind Fog 9/10/2016 M 65 23 82 no wind Fog 9/10/2016 A 76 31 53 no wind Sunny 10/10/2016 M 88 24 78 no wind Scattered clouds 10/10/2016 A 68 31 53 no wind Clear 11/10/2016 M 86 22 82 no wind Fog BP,IC,RC 140 69 21/09/2016 M 29 28 88 no wind Sunny 22/09/2016 M 26 25 82 7 Partly sunny 23/09/2016 M 32 24 85 no wind Clear 24/09/2016 M 29 26 82 4 Sunny 25/09/2016 M 24 26 87 4 Sunny 9 280 97 15/09/2016 M 65 26 84 no wind Sunny 15/09/2016 A 30 32 63 no wind passing clouds 16/09/2016 A 46 33 58 2 sunny 18/09/2016 M 58 27 85 no wind Fog 20/09/2016 M 57 27 89 no wind Fog 20/09/2016 A 24 33 74 no wind passing clouds
Continued/-
115
Unique Count Wind Survey Total Temp Humidity dogs Date Session each velocity Climate track sightings (0C) (%) sighted day (km/hr) 17 136 60 16/09/2016 M 25 26 85 4 Clear 16/09/2016 A 21 33 58 2 sunny 17/09/2016 M 28 27 87 no wind Sunny Scattered 17/09/2016 A 18 33 62 2 clouds 18/09/2016 M 26 26 85 6 passing clouds 18/09/2016 A 18 34 57 no wind passing clouds 16 277 114 25/09/2016 M 60 27 89 no wind passing clouds 25/09/2016 A 33 31 64 2 passing clouds 26/09/2016 M 54 26 87 4 Fog 27/09/2016 M 57 26 85 no wind Fog 30/09/2016 A 38 33 64 no wind Sunny 1/10/2016 A 35 33 69 no wind Sunny 1,5 308 148 12/09/2016 M 63 32 58 4 Sunny 14/09/2016 A 49 32 64 4 passing clouds 15/09/2016 A 63 32 63 no wind passing clouds 16/09/2016 M 77 26 88 no wind Clear 16/09/2016 A 56 26 85 4 Clear 8 (P) 120 37 8/09/2016 M 18 28 87 no wind Clear 8/09/2016 A 22 30 78 no wind Clear 9/09/2016 M 19 28 85 no wind passing clouds 10/09/2016 M 17 26 65 2 Sunny 12/09/2016 A 20 32 58 4 Sunny 15/09/2016 A 24 32 63 no wind passing clouds 18 183 92 10/09/2016 M 31 28 73 4 passing clouds 10/09/2016 A 28 33 59 2 Sunny 11/09/2016 M 46 28 70 no wind Clear 11/09/2016 A 53 33 58 6 passing clouds 12/09/2016 M 25 32 58 4 Sunny 6 164 85 9/09/2016 M 26 28 85 no wind passing clouds 10/09/2016 M 40 28 73 4 passing clouds 10/09/2016 A 18 33 59 2 Sunny 11/09/2016 M 33 28 70 no wind Clear 13/09/2016 A 19 29 64 4 Partly sunny 14/09/2016 M 28 32 64 4 passing clouds 7 192 69 7/09/2016 A 19 32 62 4 passing clouds 8/09/2016 M 32 28 87 no wind Clear 8/09/2016 A 35 30 78 no wind Clear 9/09/2016 M 40 28 85 no wind passing clouds 10/09/2016 M 41 26 65 2 Sunny 11/09/2016 A 25 33 58 6 passing clouds
†Session (A=Afternoon, M=Morning); ^ Temp= Temperature. *Industrial Area Part 1, # Industrial Area Part 2, @ Budhanpur, Indira Colony, Rajeev Colony, ^ Sector 8 perimeter
116
Supplementary Figure 3.1 Geographical location and surveyed sectors of Panchkula that were selected for the enumeration survey during September-October 2016.
A. The geographical location of the study site relative to India and the state of Haryana. B. Imagery of site of urban survey (Source: “Panchkula, Haryana state, India.” 30038’58.58” N and 76049’52.73” E. Google Earth. May 5, 2018. July 21, 2018). The sectors on which survey-tracks were traversed are in yellow and black font. Sectors 1&5 were covered by a single survey-track. The red perimeter around sector 8 formed the perimeter survey-track 8(P).
Supplementary Figure 3.2 Flow chart of the sampling strategy for selection of the survey-tracks in 15 sectors of the Municipal Corporation, Panchkula for the free roaming dog enumeration survey during September-October 2016
117
Supplementary Figure 3.3 Comparative estimates of the free roaming dog population size obtained by Huggin’s closed capture heterogeneity model Mh-JK (after 5 surveys) and Application SuperDuplicates (after 2 surveys)
Mh-JK AS
118
Supplementary Figure 3.4 Comparative estimates of the free roaming dog population size obtained by Huggin’s closed capture heterogeneity model Mth-Chao (after 5 surveys) and Application SuperDuplicates (after 2 surveys)
Mth-Chao AS
119
Chapter Four
Demographic characteristics of free roaming dogs in rural and urban India following a photographic sight- resight survey
“There are only the pursued, the pursuing, the busy and the tired.”
F. Scott Fitzgerald
120
Preface
In the previous chapters it was established that having a method that could reliably estimate the population size of the FRD in an area targeted for a mass immunisation programme against dog mediated rabies is paramount to achieve 70% vaccination coverage. The probabilistic models, including the Application SuperDuplicates, requires sightings and re-sightings of FRD which are likely to vary from one setting to another and may be influenced by characteristics of the respective FRD populations including their composition. The gender, age structure and body condition of the dogs, along with the distribution of resources that sustain the FRD population in the environment, such as food and shelter, potentially influence the frequency with which an FRD may be sighted during a survey. A study of the FRD demography is thus imperative for the effective implementation of rabies control measures. In this chapter the demographic characteristic of FRD in rural and urban locations are described and compared.
The text and the supplementary tables of this chapter are the same as the manuscript of the published paper in ‘Scientific Reports’ except for the reference list which has been combined with references of other chapters and incorporated as one list at the end of the thesis.
This chapter can be found published as:
Tiwari, HK, ID Robertson, M O’Dea, AT Vanak (2019). Demographic characteristics of free-roaming dogs (FRD) in rural and urban India following a photographic sight-resight survey. Scientific Reports 9,16562. https://doi.org/10.1038/s41598-019-52992-y
121
Statement of Contribution
Demographic characteristics of free roaming dogs in Title of Paper rural and urban India following a photographic sight-resight survey.
Publication Status
Tiwari, HK, ID Robertson, M O’Dea, AT Vanak. (2019). Demographic characteristics of free- roaming dogs (FRD) in rural and urban India Publication Details following a photographic sight-resight survey. Scientific Reports 9, 16562. https://doi.org/10.1038/s41598-019-52992-y
Principal Author
Name of Principal Author Harish Kumar Tiwari (Candidate) Harish Kumar Tiwari conceptualised and developed the study, planned and conducted the field study, Contribution to the Paper collected and analysed the data, interpreted the results and wrote the paper. Overall percentage (%) 60 %
Signature Date: 08/08/2019
Co-Author Contributions
By signing the Statement of Contribution, each author certifies that: i. the candidate’s stated contribution to the publication is accurate (as detailed above); ii. permission is granted for the candidate to include the publication in the thesis; and iii. the sum of all the co-author contributions is equal to 100% less the candidate’s stated contribution.
122
Name of Co-Author Emeritus Professor Ian Robertson Supervised the study and provided critical comments Contribution to the Paper to improve the interpretation of results, edited and revised the manuscript. Overall percentage (%) 20 % Signature
Date: 12/09/2019
Name of Co-Author Dr Abi Tamim Vanak Provided critical comments to improve the Contribution to the Paper interpretation of results, edited and revised the manuscript. Overall percentage (%) 10%
Date: 08/08/2019 Signature Name of Co-Author Dr Mark O’Dea Provided critical comments to improve the Contribution to the Paper manuscript. Overall percentage (%) 10% Signature
Date: 24/08/2019
123
Abstract
An understanding of the core demographic characteristics of the sub-populations of FRD is essential to effectively implement both rabies control interventions through mass vaccination of FRD, and dog population control programmes. This study compares the data obtained following photographic sight-resight surveys in rural (Shirsuphal village in west India) and urban (Municipal Corporation Panchkula in north India) locations of
India. A total of 263 and 1260 FRD were seen at least once through 617 and 3465 sightings in the rural and urban sites, respectively. The rural location had a lower proportion of females (OR 0.5, 95% CI 0.4-0.7) and a higher proportion of poor and fair conditioned dogs (OR 1.8, 95% CI 1.3-2.3) compared to the urban setting. The rural site also had fewer active FRD (OR 0.6, 95% CI 0.5-0.7) and FRD were less likely to be sighted within 20 m of garbage points (OR 0.3, 95% CI 0.2-0.3) compared to the urban site. The demographic composition of the FRD was found to vary within the urban location, with the odds of sighting a de-sexed dog being significantly higher in residential areas compared to other areas. The study underlines the importance of knowing the demographic composition of FRD for implementation of effective interventions against rabies. Fewer female dogs in the rural location indicate that spaying could be an effective tool for dog population management in a rural setting, while presence of dogs within 20m of garbage in urban settings highlights that an improved garbage management may reduce the carrying capacity of the urban locality that may lead to better management of FRD population. It is concluded that quick and lowcost surveys can generate useful demographic data for FRD in urban and rural settings which can be useful to understand epidemiology of rabies and its control.
124
4.1 Introduction
Free-roaming dogs (FRD) pose a serious threat to human health in countries where dog- bite related rabies is endemic, as well as causing environmental contamination with faeces, spreading garbage waste, damage to property and noise pollution (Tiwari et al.
2018, Rinzin, Robertson, and Mahat 2017, Ratsitorahina et al. 2009). The epidemiology of human rabies is intrinsically connected with the presence of rabies virus in FRD, and thus understanding the ecology of these dogs is imperative when developing and implementing control programmes for rabies as well as other zoonotic diseases (Meslin,
Fishbein, and Matter 1994, Otolorin, Umoh, and Dzikwi 2014, Butler and Bingham 2000,
Wandeler et al. 1993, Davlin and VonVille 2012). The structure and turnover of the FRD population is based on the characteristics of the dog’s demography including gender and age ratios, body condition, birth rates, success of rearing, mortality and survival rates
(WHO 2005).
The presence of FRD on the streets of urban and rural areas in rabies endemic countries is maintained by a combination of factors, namely indiscriminate breeding of unowned dogs, unrestricted movement of semi-owned dogs and abandonment of owned dogs by irresponsible owners (WSPA 2010). The World Health Organisation (WHO) recommends control of rabies through annual mass vaccination of FRD, with coverage of at least 70% of the population required to break the disease’s transmission cycle (WHO
2013). This percentage accounts for the loss of herd immunity levels resulting from the turnover of the dog population due to deaths, births and migrations (Conan et al. 2015).
Knowledge of the core demographic characteristics of sub-populations of FRD, such as male-female ratios, age composition of the population, social behaviour with respect to their dependence on edible litter/garbage and their activity level, is important to
125
effectively implement both rabies control interventions and dog population control programmes (Slater 2001).
Animal Birth Control programmes have been implemented in some urban localities in
India, although at many places the efforts have been irregular and sporadic (Krishna 2009,
Totton, Wandeler, Zinsstag, et al. 2010). These efforts are often implemented without considering the demographic composition of the FRD in the area of application, resulting in little or no reduction in the population. Furthermore, there are few epidemiological studies on the demographic composition of FRD in India, where rabies is endemic and the majority of human mortality from rabies is linked to dog bites (Davlin and VonVille
2012). Although some studies have been conducted on the demographics of FRD in
Eastern India (West Bengal) in urban settings (Pal 2001, Oppenheimer and Oppenheimer
1975), few studies from rural areas have been undertaken (Belsare and Gompper 2013).
In this study we present the demographic details of FRD in Shirsuphal village in western
India (rural location) and compare these with various residential and industrial sectors of the urban municipality of Panchkula in north India, through a series of photographic capture-recapture surveys of individually identifiable FRD undertaken on 5-7 occasions on predetermined tracks. We also discuss the various factors that possibly influence the
FRD demography in rural and urban settings and the implications of such data for implementing effective rabies control interventions.
126
4.2 Results
4.2.1 Sighting variability and the demographic characteristics of free roaming dogs
At the rural site a total of 263 distinct dogs were identified through 617 sightings during the seven surveys conducted over the nine-day survey period. The demographic details of the FRD seen at least once are presented in Table 4.1. Variations in the number of dogs sighted on each survey, along with meteorological data, are presented in the
Supplementary Table 4.1. The number of dogs sighted each day ranged from 106 to 52 with a declining trend as the survey progressed. This decrease followed a linear relationship with a negative slope (R2 = 0.63, 푦 = - 6.57푥 + 114.43). The number of active
FRD during the survey period remained similar across survey days but a significant variation was observed in the number of FRD with respect to their proximity (≤ 20m) to garbage points (Table 4.2).
127
Table 4.1 Demographic details (gender, age distribution and body condition) of the free roaming dogs sighted on each survey occasion during the enumeration survey (7 occasions) in the rural survey (Shirsuphal village, Baramati, Pune).
Gender* Age @ Body condition@
Survey Number of Young Poor occasion dogs sighted Male (%) Female (%) Pup (%) (%) Adult (%) Old (%) Good (%) Fair (%) (%) 1 93 56 (62) 34 (38) 8 (9) 17 (18) 63 (68) 5 (5) 57 (61) 22 (24) 14 (15) 2 106 70 (75) 23 (25) 11 (10) 11 (10) 74 (70) 10 (9) 54 (51) 41 (39) 10 (10) 3 103 67 (68) 31 (32) 8 (8) 21 (20) 63 (61) 11 (11) 56 (54) 35 (34) 13 (13) 4 91 66 (78) 19 (22) 2 (2) 18 (20) 64 (70) 7 (8) 53 (58) 31 (34) 7 (8) 5 90 58 (68) 27 (32) 9 (10) 17 (19) 54 (60) 10 (11) 39 (43) 37 (41) 15 (16) 6 82 45 (60) 30 (40) 4 (5) 11 (13) 59 (72) 8 (10) 53 (62) 27 (32) 5 (6) 7 52 33 (65) 18 (35) 2 (4) 10 (19) 35 (67) 5 (10) 43 (83) 8 (15) 1 (2) 170# (71) 70 (29) 17 (6) 34 (13) 197 (75) 15 (6) 143 (54) 89 (34) 31 (12) Test for independence over the seven surveys χ²=10.33, p=0.17 χ²=23.7, p =0.31 χ²=32.8, p =0.003
Rows indicate the number of confirmed unique animals in the relevant category; *Gender of 23 FRD could not be verified; @ Age and Body condition was assessed based on visual characteristics as: Pup (< 6 months), Young (6 months to 1 year), Adult (≥1 to 7 years), and Old (> 7years); Good, Fair and Poor, respectively.
128
Table 4.2 Details of the activity and sightings within 20 m of garbage of free roaming dogs sighted on each survey occasion during the rural survey (Shirsuphal village, Baramati, India).
Survey Number of dogs Activity (%) Proximity to garbage (%) occasion sighted Active Not active ≤ 20 m > 20 m 1 93 40 (43) 53 (57) 32 (34) 61 (66) 2 106 42 (39) 64 (61) 34 (32) 72 (68) 3 103 42 (41) 61 (59) 25 (24) 78 (76) 4 91 36 (40) 55 (60) 15 (16) 73 (84) 5 90 41 (46) 49 (54) 17 (19) 73 (81) 6 82 40 (49) 42 (51) 16 (20) 66 (80) 7 52 22 (42) 30 (58) 9 (17) 43 (83) Total 617 263 (43) 354 (57) 148 (24) 469 (76) Test for independence over the χ²=1.004, p =0.99 χ²=15.98, p =0.025 seven surveys
At the urban site a total of 1408 unique FRD were identified through 3465 reliable sightings in the 14 sectors of Panchkula Municipal Corporation administrated areas during September-October 2016. The demographic details of these FRD are displayed in
Table 4.3. The details of the number of FRD sighted each day, along with the meteorological details for the urban survey, are presented in the Supplementary Table
4.2.
The proportion of dogs sighted (of the total unique FRD identified in each sector of
Panchkula) on each day of the survey varied from 21 to 67%. The proportion of animals sighted on each day of the survey mostly followed a linear relationship with a positive slope, except for the residential sectors (9, 12, 16 and 17), where the trend was negative.
Wind velocity and ambient temperature during the survey had a negative correlation (r =
-0.3, p=0.01; r = -0.5, p=0.001, respectively) with the number of dogs sighted. Although no significant variation was observed in the gender composition of the FRD (χ2 =7.9, df
= 13, p=0.8), a significant variation was found in the age and body condition composition
129
of the FRD observed in each of the surveyed sectors (χ2 = 146.19, df = 39, p<0.001 and
χ2 = 160.6, df = 26, p<0.001, respectively).
The number of dogs observed as active in Panchkula was negatively correlated with the ambient temperature at the time of the survey (r= -0.4, p=0.000056) and varied significantly across the surveyed sectors (χ2 = 27.6, df = 13, p = 0.01). Similarly, the number of dogs sighted within 20 meters of garbage dumps/accumulated litter differed significantly between the sectors (χ2= 287.5, df = 13, p<0.001). Of the total 3465 sightings of FRD in all urban sectors, 1884 (54%) involved animals that were active and 1678
(48%) were sightings of dogs within 20m of garbage points/dumps (Table 4.4).
The comparison of the demographic characteristics of rural and urban FRD is presented in Supplementary Table 4.3. The odds of sighting a female dog were significantly lower in the rural area (Shirsuphal) compared to the urban area (Panchkula) (OR 0.5, 95% CI
0.4-0.7, p<0.001). The likelihood of sighting a FRD with a poor or fair body condition was higher in the rural setting compared to the urban setting (OR 1.8, 95% CI 1.3-2.3, p<0.001). Rural FRD were less active (OR 0.6, 95% CI 0.5-0.7, p<0.001) and less likely to be sighted within 20m of a garbage point (OR 0.3, 95% CI 0.2-0.3, p<0.001) than urban
FRD (Figure 4.1).
130
Table 4.3 Demographic details (gender, age distribution, neutering status and body condition), their respective distribution for the FRD sighted in the different sectors of the urban location (Panchkula Municipal Corporation administrated sectors).
Unique Gender$ Age† Neutered Body condition† dogs Young Sector identified Male (%) Female (%) Pup (%) (%) Adult (%) Old (%) Yes (%) No (%) Good (%) Fair (%) Poor (%) 1&5 148 82 (55) 66 (45) 13 (9) 1 (1) 134 (90) 0 11 (7) 137 (93) 123 (83) 18 (12) 7 (5) 2 127 57 (51) 55 (49) 13 (10) 12 (9) 94 (75) 8 (6) 45 (35) 82 (65) 89 (70) 25 (20) 13 (10) 6 85 51 (61) 32 (39) 5 (6) 11 (13) 69 (79) 2 (2) 10 (12) 75 (88) 63 (74) 17 (20) 5 (6) 7 69 40 (58) 29 (42) 11 (16) 17 (25) 40 (58) 1 (1) 29 (42) 40 (58) 58 (84) 7 (10) 4 (6) 8 112 54 (55) 44 (45) 15 (13) 8 (7) 81 (73) 8 (7) 18 (16) 94 (84) 70 (62) 31 (28) 11 (10) 9 97 54 (57) 40 (43) 3 (3) 14 (14) 67 (70) 12 (13) 31 (32) 66 (68) 58 (60) 29 (30) 10 (10) 12 86 42 (52) 39 (48) 3 (3) 4 (5) 74 (86) 5 (6) 32 (37) 54 (63) 65 (75) 13 (15) 8 (10) 16 114 65 (61) 42 (39) 8 (7) 13 (11) 91 (80) 2 (2) 36 (31) 78 (69) 79 (69) 18 (16) 17 (15) 17 60 39 (65) 21 (35) 2 (3) 8 (13) 49 (82) 1 (2) 16 (27) 44 (73) 45 (75) 11 (18) 4 (7) 18 92 59 (64) 33 (36) 5 (5) 25 (27) 56 (61) 6 (7) 17 (18) 75 (82) 71 (77) 3 (3) 18 (20) 8 (P)^ 37 22 (59) 15 (41) 3 (8) 0 33 (89) 1 (3) 6 (16) 31 (84) 29 (78) 8 (22) 0 IAP 1* 168 86 (55) 69 (45) 12 (6) 17 (10) 135 (82) 4 (2) 44 (26) 124 (74) 90 (53) 55 (33) 23 (14) IAP 2# 144 74 (57) 55 (33) 15 (11) 5 (3) 118 (82) 5 (4) 26 (18) 118 (82) 76 (53) 40 (28) 28 (19) BP,IC,RC@ 69 37 (56) 29 (44) 1 (1) 7 (11) 61 (88) 0 20 (29) 49 (71) 39 (56) 3 (4) 27 (40) 1408 762 (57) 569 (43) 109 (8) 142 (11) 1102 (77) 55 (4) 330 (23) 1078 (77) 955 (68) 278 (20) 175 (12) Test for independence over fourteen survey tracks χ2 =7.9, p=0.8 χ2 = 146.2, p<0.001 χ2 =76.3, p<0.001 χ2 = 160.6,p<0.001
^ Sector 8 perimeter, *Industrial Area Part 1, # Industrial Area Part 2, @ Budhanpur, Indira Colony, Rajeev Colony; $ Gender of 77 FRD could not be verified; † Age was recorded by visual appreciation as: Pup (< 6 months), Young (6 months to 1 year), Adult (≥1 to 7 years), and Old (> 7years); and Good, Fair and Poor, respectively.
131
Table 4.4 Details of activity and sightings within 20 m of garbage point for free roaming dogs sighted across 14 sectors of the urban location (Panchkula Municipal Corporation administrated area).
Activity Proximity to garbage Not active Sector Sightings (n) Active (%) (%) ≤20 m (%) > 20m (%) 1&5 308 181 (59) 127 (41) 175 (57) 133 (43) 2 313 163 (52) 150 (48) 105 (34) 208 (66) 6 164 93 (57) 71 (43) 73 (45) 91 (55) 7 192 107 (56) 85 (44) 29 (15) 163 (85) 8 250 109 (44) 141 (56) 61 (24) 189 (76) 9 280 119 (43) 161 (58) 94 (34) 186 (66) 12 199 111 (56) 88 (44) 70 (35) 129 (65) 16 277 128 (46) 149 (54) 135 (49) 142 (51) 17 136 85 (63) 51 (38) 12 (9) 124 (91) 18 183 121 (66) 62 (34) 36 (20) 147 (80) 8(P)^ 120 59 (49) 61 (51) 70 (58) 50 (42) IAP* 1 452 223 (49) 229 (51) 433 (96) 19 (4) IAP* 2 451 278 (62) 173 (38) 285 (63) 166 (37) BP,IC,RC# 140 107 (76) 33 (24) 100 (71) 40 (29) Total 3465 1884 (54) 1581 (46) 1678 (48) 1787 (52) Test for independence over fourteen survey tracks χ2 = 27.6, p = 0.01 χ2= 287.5, p<0.001
^ Sector 8 perimeter, *IAP = Industrial Area Part, # Budhanpur, Indira Colony, Rajeev Colony.
132
Figure 4.1 Graphical representation of the likely sightings of FRD according to gender, age, body condition, activity level and proximity to garbage (≤ 20 m) in the rural (Shirsuphal village) and urban (sectors of Panchkula Municipal Corporation) study sites during the enumeration survey carried out in September - October 2016
*The dots represent the odds ratio and bars represent the 95% confidence limits.
133
4.2.2 Variations in the composition of the free-roaming dog population within the urban region
Comparison of the composition of the FRD population observed in different localities during the urban survey is presented in Supplementary Table 4.4. The proportion of adult and old dogs in the administrative sector was significantly higher (OR 2.0, 95%CI 1.1-
3.7, p = 0.01) but significantly lower in the urban villages in the organised sectors (OR
0.3, 95%CI 0.2-0.4, p<0.001) compared to the residential sectors. The proportion of the
FRD with a good body condition was significantly higher in the administrative areas (OR
2.8, 95%CI 1.7-4.9, p=0.01), and the urban village (OR 1.9, 95%CI 1.2-3.0, p<0.001), but lower in the industrial sectors (OR 0.5, 95%CI 0.4-0.7, p<0.001) compared to the residential sectors. The residential areas had a significantly higher proportion of de-sexed
FRD (ear-notched) compared with the industrial, administrative, and the part residential, part administrative sectors (Figure 4.2).
134
Figure 4.2 Graphical representation of the likely sightings of FRD (odds ratios and 95% confidence intervals) according to age, body condition and de-sexing status during the enumeration survey carried out in the residential, industrial, administrative and mixed sectors of Panchkula Municipal Corporation during September - October 2016*
*The dots represent the odds ratio and bars represent the 95% confidence limits.
135
4.3 Discussion
A series of photographic sight-resight surveys of individually identifiable FRD was conducted in an urban and rural setting in India to estimate the population size of FRD and the data obtained were analysed to describe their demographic characteristics to further our understanding of the implications of the demographics of FRD in intervention programmes against rabies or dog population management measures.
Although, the enumeration surveys were conducted at different times of the year, and only one village could be included from each location due to resource constraints, the methodology followed was consistent between surveys and each survey covered the entire selected village or sector. Nonetheless, there could be an inherent difference in the level of detectability of urban and rural FRD which may be a potential limitation of this study. Further, only one survey could be undertaken on the perimeter area separating the residential sectors in the urban location due to resource constraints. It is possible that dogs that were sighted in that perimeter area were in transit and may have actually resided in one of the neighbouring sectors, but we could not confirm this. The influence of extrinsic factors and demographic composition of FRD in urban and rural locations is described in the following sections.
4.3.1 Influence of temperature on FRD sightings
The ambient temperature of the survey day was found to have a significant negative effect on the number of FRD sighted, an observation that was more pronounced in the urban survey, where a larger spatial and temporal dataset were available. Similar results were reported in a study in Berkeley, USA by Berman and Dunbar (1983), where the sightings of FRD declined when temperatures rose above 24 0C. We observed higher counts in morning sessions that were cooler (260C on average), compared to the warmer (320C
136
average) afternoon sessions. Increased number of sightings during the cooler periods of the day is likely associated with increased movements of the dogs to seek food or company. In contrast with increasing temperatures dogs would seek shaded or cooler shelters resulting in less frequent sightings on the roads. Similar findings were also reported in West Bengal, India by Oppenheimer and Oppenheimer (1975). The influence of overcast conditions and rainfall on the sightings of FRD in the rural location is discussed in more detail in Tiwari et al. (2018).
4.3.2 Gender ratio
The gender ratio in Shirsuphal village (rural) was heavily skewed in favour of males (1:
2.45) (71% males), which was similar to that reported from villages in the vicinity of the
Great Indian Bustard Sanctuary in the neighbouring district of Solapur, as well as from
Bangladesh, Chile, Indonesia, and the Republic of South Africa (Mustiana et al. 2015,
Conan et al. 2015, Belsare and Gompper 2013, Hossain et al. 2013, Acosta-Jamett,
Cleaveland, and Cunningham 2010). This disparity could be attributed to the preference of farming communities for male dogs (Margawani and Robertson 1995, Hossain et al.
2013) or high female dog mortality (Kitala et al. 2002, Conan et al. 2015). In contrast a closer male-female parity was observed in the urban study (Panchkula) (1.34: 1) (57% males), which is consistent with estimates from other urban studies in West Bengal, India
(1.37: 1) and Bhutan (1.31: 1) (Pal 2001, Rinzin, Tenzin, and Robertson 2016). The variations in the gender ratios between different sectors (1.07 - 1.85: 1) in Panchkula could be attributed to the varying degree of human influence on FRD, which would be expected to be higher in residential sectors than in sectors that comprised open areas
(parks, school playgrounds, markets, industrial areas) (Boitani, Ciucci, and Ortolani
2007). The ABC programme in Panchkula may have also contributed to gender parity in
137
FRD (Totton, Wandeler, Zinsstag, et al. 2010), but with only 23% of dogs identified as de-sexed (ear-notched), the impact would not be expected to fully explain the gender ratio in the urban location.
4.3.3 Age composition of FRD
No significant difference was observed in the age composition of FRD in the rural and urban surveys, which was similar to findings reported in other countries (Belsare and
Gompper 2013, Aiyedun and Olugasa 2012, Mustiana et al. 2015, Conan et al. 2015,
Rinzin, Tenzin, and Robertson 2016, Tenzin, Ahmed, et al. 2015). The low percentage of pups and young dogs in both rural (17%) and urban (18%) settings could be due to high early mortality (Conan et al. 2015). It may be argued that the time of the survey in rural
Shirsuphal was an influencing factor for the infrequent sightings of puppies and young dogs because the survey was undertaken before (early June) the whelping season in
September-October (Pal 2001), but a low percentage was also observed in urban
Panchkula where the study was undertaken in September-October supporting the hypothesis of early-age mortality in FRD irrespective of the location. Human influences, such as motor-vehicle accidents and deliberate poisoning, also contribute to the death of
FRD puppies (Acosta-Jamett, Cleaveland, and Cunningham 2010, Faleke 2003, Paul et al. 2016, Davlin and VonVille 2012). The stress caused by biannual breeding in reproductive females has also been cited as a potential reason for early juvenile mortality
(Butler and Bingham 2000, Daniels and Bekoff 1989). Furthermore, the absence of any communal or group care of pups by other bitches witnessed in FRD (as opposed to wild canids) also contributes to the low survival of juveniles (Boitani and Ciucci 1995). Butler and Bingham (2000) , observed that extra nutritional pressure exerted on reproductive free-roaming bitches by the sympatric semi-owned dogs may also result in higher
138
mortality of puppies. Others have reported that dogs after the age of 4 to 6 months may move to neighbouring villages with less competition for food/shelter resulting in a lower percentage of young dogs, a possibility which cannot be ruled out in the present study
(Ivanter and Sedova 2008, Pal, Ghosh, and Roy 1998b, Paul, Majumder, and Bhadra
2014). However, as the population size is driven by a large number of reproductively active animals coupled with large litter sizes (Conan et al. 2015, Gsell et al. 2012), a high proportion of reproductively active animals is an indicator of the high fecundity in the population in both the urban and rural survey sites.
4.3.4 Body condition and the sightings near garbage points
The likelihood of sighting a FRD of poor or fair condition in rural Shirsuphal was significantly higher (OR 1.8) than in urban Panchkula. This may be related to the availability of food as FRD were less likely to be sighted within 20m of garbage points in the rural study (OR 0.3) than in the urban study. Tenzin, McKenzie, et al. (2015), also reported a high proportion of FRD with a good body condition in Bhutan and attributed this to ready access to food and the local community’s responsibility to feed FRD. In this study, besides FRD in urban Panchkula sourcing feed from the garbage points, a higher percentage of urban (72%) than rural (39%) residents fed the FRD due to compassionate or religious reasons (Tiwari, Vanak, et al. 2019, Tiwari, Robertson, O’Dea, and Vanak
2019). Although only 24% of the FRD were found within a 20m radius of garbage dumps/sites in rural Shirsuphal, it is possible that they scavenge from such sites in the evenings (Vanak and Gompper 2009a, b, Boitani, Ciucci, and Ortolani 2007, Boitani and
Ciucci 1995). The good body condition of more than half of the sighted FRD and a lack of congregation around garbage sources provides evidence that the FRD in Shirsuphal are not typical of feral dogs, but are more likely to have some level of human association and
139
at least some of them could be loosely categorised as owned as it is highly unlikely for a
FRD population to remain in a state of good health without some form of human intervention (Morters, Bharadwaj, et al. 2014).
Totton et al. (2011), cited that ABC programmes were key drivers of better health conditions for the FRD in urban Jaipur, north India. The benefits of conducting sustained
ABC programmes to improve the body condition of unowned dogs is also confirmed by the results of studies in Dhaka, Bangladesh and in urban Jodhpur, India (Yoak et al. 2014,
Tenzin, Ahmed, et al. 2015). Potentially the better body condition observed in the urban survey could also be a result of the ongoing sterilisation programme in Panchkula, although only 23% of FRD dogs were identified as neutered (ear-notched).
4.3.5 Activity
The sightings of FRD in urban Panchkula that were involved in some kind of activity, such as walking, running, foraging or playing, was significantly higher (OR 1.6) than that in rural Shirsuphal. The percentage of active FRD in urban settings (54%) was comparable to that observed in California, USA (Berman and Dunbar 1983) (56%).
Although some studies suggest that the activity of FRD varies according to the time of the day (Pérez et al., Berman and Dunbar 1983), in this study there was no apparent temporal pattern in their activity classification, although the proportion of FRD undergoing an activity was negatively correlated with the ambient temperature at the time of the survey.
In the urban survey at Panchkula most of the FRD categorised as “not-active” were sighted under parked cars, even when other places of shade were available. Although availability of food is considered the primary cause for the high number of FRD (Kato et
140
al. 2003, Beran and Frith 1988), the presence of these “shelters” could also be an important factor contributing to the higher “carrying capacity” of an urban environment.
Construction of dog-proof enclosed parking lots may contribute to the control of the FRD population in urban environs as they would deny the FRD temporary shelters - an essential component for their survival.
4.3.6 Free-roaming dog demography and relationships to ABC programmes
The number of female dogs in Shirsuphal village was much lower indicating that de- sexing of females is potentially an economically viable option for the control of the FRD population in this location. The same, however, may not be true in urban settings where there were comparable numbers of males and females, and as such de-sexing of both males and females may yield a faster population control. A salient finding of FRD in urban settings is that the number of notched (de-sexed) FRD was significantly higher in the residential sectors than in other sectors (Figure 4.2). This is most likely because the
FRD are more easily accessible for capture in residential areas compared to other sectors.
The process of neutering a FRD involves capture of the dog, a procedure which is very challenging and the dogs that are wary of human interaction often prove difficult, if not impossible, to catch. The challenges of catching a FRD to administer interventions is a serious impediment towards the control of rabies in canines (Massei and Miller 2013,
Cleaveland et al. 2006) and is a major cause for its persistence in countries, such as India, where the disease is endemic. Nevertheless, the efficacy of ABC measures in Panchkula remains doubtful in light of the wide disparity between the proportions of neutered FRD in the different sectors/areas. Demographic surveys, such as the one described in this study, thus also help assess the efficacy of ABC programmes.
141
The assumption that sterilisation can reduce the density of FRD resulting in fewer dog- bites and thus control the incidence of rabies (Totton, Wandeler, Gartley, et al. 2010) can only be verified by studying the dynamics of population through a longitudinal observational study. In Panchkula, where ABC programmes are reportedly irregular and sporadic (Executive Officer, Municipal Corporation, Panchkula), it is likely that FRD do not live long as no aged dogs were sighted among the sterilised dogs. Reece and Chawla
(2006), while discussing control of rabies in Jaipur, India, argued that sterilisation followed by vaccination against rabies results in life long immunity in stray dogs presumably due to their short lifespan, reported to average 2.6 years (Pal 2001).
Nevertheless, the possible reduction in the density of the dogs due to sterilisation is likely to be mitigated by high population turnover and immigration of FRD from neighbouring sectors. Moreover, host density does not appear to affect the transmission dynamics for dog-related rabies due to a low reproductive number (R0) for the disease (Morters et al.
2013).
In contrast, the dynamics of disease transmission depends largely on population size and the factors that sustain high numbers of the reservoir hosts (Morters, McKinley, Restif, et al. 2014). Although mass vaccination of FRD has been widely recommended for the control of rabies (Cleaveland, Beyer, et al. 2014, Morters, McKinley, Horton, et al. 2014), this strategy has had very little success in India. This is likely due to the tendency of FRD to group around accumulated garbage is enhanced when there is poor garbage management, which it turn makes it difficult for vaccine administrators to catch dogs for parenteral vaccination (Pal, Ghosh, and Roy 1998a). Besides inaccessibility of dogs for mass vaccination (Bögel and Joshi 1990), a factor that works against mass vaccination is the inability of malnourished or poor body conditioned FRD to sero-convert and sustain the population immunity at critical levels (Hampson et al. 2009) (Morters, Bharadwaj, et
142
al. 2014). Such immune response constraints may not apply to areas such as Panchkula, where the majority of the FRD (68%) were found to be in good body condition. However, as cities improve their solid waste management, leading to a reduction of the resources that sustain dog populations in cities such as Panchkula, it may in the short-term create cohorts of under-nourished dogs.
We have demonstrated that quick and relatively low-cost surveys such as described in
Tiwari et al. (2018) and Tiwari, Robertson, O’Dea, Gogoi-Tiwari, et al. (2019), can not only provide robust population estimates for FRD, but can also be used to generate useful demographic data for dogs in urban and rural areas of a rabies endemic country such as
India. Such data provide useful insights into the various factors that need to be considered for understanding the epidemiology of rabies and its control.
4. 4 Materials and methods
4.4.1 Study area
(a) Rural
The rural surveys were conducted in the village of Shirsuphal in Baramati town of Pune
District in Maharashtra state of western India (18018’49.08” N and 74034’44.40” E) from the 4th to 13th June 2016. The village has human settlements interspersed with farmland.
Sixteen km of roads, of which 12 are bitumen, connect the various settlements. The village comprises of 1161 households (www.censusofindia.gov.in, as accessed in July
2016). The study excluded FRD on farmlands sighted at a distance of 200 meters or more from the roads. In the month of June, the average humidity in Baramati is 72% with an ambient temperature range of 230C to 320C
(https://www.timeanddate.com/weather/india/baramati). The survey was conducted
143
using the roads and trails along the human habitats that connect the various settlements
(Tiwari et al. 2018).
(b) Urban:
The residential and industrial areas under the administrative control of the Municipal
Corporation, Panchkula (30038’58.58” N and 76049’52.73” E) were surveyed during the months of September and October 2016. Panchkula is one of the highly urbanised and planned districts in India (Bala 2014) and the Municipal Corporation administered area is comprised of wards that are further divided into sectors. The number of sectors in each ward varies from 1 to 6, and includes highly organised residential, administrative and industrial sectors interspersed by unorganised slums and villages (Duggal 2004). Fifteen sectors were selected through purposive (industrial, unorganised and mix sectors) and random (residential sectors) selection and included residential (sectors 2, 8, 9, 12, 16 and
17); administrative (sectors 1, 5); industrial (sectors IAP I and IAP II); and part– residential part-public areas such as hospitals, colleges and parks (sector 6). Two of the surveyed sectors (7 and 18) included an urban village where, although the survey route was bitumen, many alleys branched out into densely unorganised settlements (slums).
Such alleys were not included in the survey due to the resistance of the residents to participation. However, an unorganised area included in the Municipal Corporation limits, comprising three colonies (Budhanpur, Rajeev Colony, Indira Colony - referred to as BP, RC, IC, respectively throughout this paper), was surveyed. While the possibility of FRD moving from one sector to another could not be completely excluded, their movement was restricted within residential areas by solid brick walls surrounding the areas. The counts of FRD were undertaken in the perimeter area of sector 8 on the roads connecting the residential sectors. In contrast the industrial areas did not have such
144
defined walled restrictions. The survey was conducted inside each sector along the connecting bitumen roads. The surveyed area comprised 6337 households
(www.censusofindia.gov.in, as accessed in July 2016). The average temperature observed in September/October 2016 was 290C/260C with an average humidity of 75%/69%, respectively (www.timeanddate.com/weather/india/panchkula).
4.4.2 Field methodology
A consistent methodology through a series of photographic sight-resight surveys of the
FRD was followed in rural (Shirsuphal village) and urban (sectors under administrative control of Panchkula Municipal Corporation) locations (Tiwari et al. 2018, Tiwari,
Robertson, O’Dea, Gogoi-Tiwari, et al. 2019). Briefly, the selected areas were traversed by a team of two observers riding a motorcycle at a constant speed of ~20 Km/hour following a predetermined route (survey-track) on alternate mornings and evenings/afternoons for five to seven occasions. The survey-tracks and the teams remained unchanged throughout the survey period for a particular survey track and were equipped with a GPS device (Garmin eTrex20 GPS device, www.garmin.com), a digital camera (Nikon COOLPIXA900) and writing materials. The surveys were conducted during mornings (6.30 – 8.30 for the rural and 6.00 – 8.00 for the urban locations) and evenings/afternoons (17.00 – 19.00 for the rural and 16.00 – 18.00 for the urban locations), except for the three unorganised colonies (BP, RC and IC) in the urban settings where the residents resisted the survey during the afternoon sessions. The duration of the surveys was constant for a particular survey track and lasted for 1 to 2 hours, depending on the track length and number of FRD sightings on the day of the survey. The surveys covered all bitumen routes in the selected areas that were frequented by human and FRD alike (Hiby and Hiby 2017) and covered the entire village/sector. The lengths of the two
145
survey tracks in Shirsuphal were 7.5 and 6 km, while those in the urban location of
Panchkula varied from 4.2 to 14.7 km (Supplementary Table 4.2).
Each FRD sighted during the survey was photographed and details recorded on its gender
(male/female/not verified), age (pups/young/adult/old), body condition (poor/fair/good), reproductive status (pregnant/in-oestrous/lactating/notched), and details of the coat colour (single coloured/bicoloured/tricoloured/mixed/striped) along with its location
(GPS waypoints). The photographs of all FRD were matched with details on the datasheet after each survey to ascertain if a FRD had been seen for the first time or if it was a resighted one. A list of FRD sighted at least once during the entire survey period for each location was compiled and used for analysing the attributes of the dogs.
Besides the assigned sites of garbage disposals (garbage dumps/points), many temporary accumulations of litter along the roads were witnessed, especially in the rural settings.
The proximity of the FRD to such garbage dumps/points or accumulated litter (presence
≤ 20m of such sites), the activity of the FRD at the time of the survey (active/inactive), the reproductive status of adult female dogs (if lactating or in oestrus), and their de-sexed status (left ear-notched indicating de-sexed/not notched indicating entire) were also recorded. The FRD was recorded as active if it was found walking, running, playing or foraging and as inactive if observed sitting, lying or sleeping.
4.4.3 Data entry and analysis
All population survey data were entered and organised in a spreadsheet (Microsoft Excel
2013, Redmond, USA) after each survey. Every dog sighted during the complete survey was marked as ‘1’ or ‘0’ as having been ‘sighted’ or ‘not sighted’ on a particular survey.
Chi-square tests for independence were used to examine variation in the counts of different categories (gender, age groups, body condition score, proximity to garbage)
146
observed between the surveys. Regression analyses and χ2 tests were performed in R
Programming Environment (R Development Core Team 2013). The R package “epitools” was used to calculate the odd ratios (Tomas 2017 ). Bonferroni adjusted p values were calculated by dividing the critical level of significance ( =0.05) by the number of dependent variables compared (six).
4.4.4 Ethical approval
Ethics approval for the observation of the FRD in the rural and urban areas was obtained from ATREE (Ashoka Trust for Research in Ecology and the Environment)
(AAEC/101/2016).
Author contributions
All authors have contributed and approve the contents of this article. HT developed the study, collected and analysed the data, and wrote the article. AV supervised the field work and IR, AV, MO supervised, provided critical revision and helped interpretation of contents and implications.
147
Funding
The work was funded by the Wellcome Trust-DBT India Alliance Program through a
Fellowship to AV (Grant number: IA/CPHI/15/1/502028) and the Research grant to HT from Murdoch University, Western Australia, Australia.
Acknowledgements
Murdoch University International Postgraduate Scholarship (MIPS) to HT is gratefully acknowledged. The authors are grateful to Pranav Panwalkar, Pradeep Satpute, Reetika
Maheshwari, and the students of Government College Sector 14, Panchkula for helping with the data collection. Gratitude is also due to the residents of Shirsuphal and Municipal
Corporation, Panchkula to have allowed to carry out the study.
Conflict of Interest Statement
The authors declare that the research was conducted in the absence of any commercial, financial or non-financial relationships that could be construed as a potential conflict of interest.
148
Supplementary Table 4.1 Details of the number of free roaming dogs sighted during the sight-resight surveys undertaken in Shirsuphal village and fourteen survey tracks of Panchkula Municipal Corporation along with the meteorological data for each survey session.
Unique Count Wind Survey Total Temp Humidity dogs Date Session each velocity Climate track sightings (0C) (%) sighted day (km/hr) Rural (Shirsuphal) A & B 617 263 5/06/2016 E 93 32 55 7 Sunny 6/06/2016 M 106 26 80 2 Overcast 7/06/2016 E 103 32 55 6 Overcast 8/06/2016 M 91 27 78 6 Overcast 9/06/2016 E 90 35 42 4 passing clouds 12/06/2016 E 82 30 59 13 passing clouds 13/06/2016 M 52 30 70 19 passing clouds Urban (Panchkula) 8 250 112 16/09/2016 M 53 26 85 4 Clear 18/09/2016 A 44 34 57 no wind passing clouds 21/09/2016 M 58 28 88 no wind Sunny 22/09/2016 A 37 29 67 4 passing clouds 24/09/2016 M 58 26 82 4 Sunny 2 313 127 6/10/2016 M 58 24 85 no wind Clear 6/10/2016 A 55 32 56 no wind Sunny 7/10/2016 M 75 25 85 no wind Fog 7/10/2016 A 61 31 61 no wind Sunny 8/10/2016 M 64 24 85 no wind Fog 12 199 86 12/10/2016 A 36 31 45 2 Sunny 13/10/2016 M 52 22 71 no wind Clear 14/10/2016 M 51 21 72 no wind Clear 14/10/2016 A 33 30 44 4 Sunny 15/10/2016 A 27 30 43 2 Sunny IAP 1 452 168 3/10/2016 M 64 28 88 no wind Fog 3/10/2016 A 71 33 63 no wind Haze 4/10/2016 A 76 32 68 no wind Fog 5/10/2016 M 83 27 79 no wind Fog 5/10/2016 A 85 32 66 4 Scattered clouds 6/10/2016 M 73 24 85 no wind Clear IAP 2 451 144 8/10/2016 M 68 24 85 no wind Fog 9/10/2016 M 65 23 82 no wind Fog 9/10/2016 A 76 31 53 no wind Sunny 10/10/2016 M 88 24 78 no wind Scattered clouds 10/10/2016 A 68 31 53 no wind Clear 11/10/2016 M 86 22 82 no wind Fog 9 280 97 15/09/2016 M 65 26 84 no wind Sunny 15/09/2016 A 30 32 63 no wind passing clouds 16/09/2016 A 46 33 58 2 sunny 18/09/2016 M 58 27 85 no wind Fog 20/09/2016 M 57 27 89 no wind Fog 20/09/2016 A 24 33 74 no wind passing clouds
Continued/-
149
Unique Count Wind Survey Total Temp Humidity dogs Date Session each velocity Climate track sightings (0C) (%) sighted day (km/hr) BP,IC,RC 140 69 21/09/2016 M 29 28 88 no wind Sunny 22/09/2016 M 26 25 82 7 Partly sunny 23/09/2016 M 32 24 85 no wind Clear 24/09/2016 M 29 26 82 4 Sunny 25/09/2016 M 24 26 87 4 Sunny 17 136 60 16/09/2016 M 25 26 85 4 Clear 16/09/2016 A 21 33 58 2 sunny 17/09/2016 M 28 27 87 no wind Sunny 17/09/2016 A 18 33 62 2 Scattered clouds 18/09/2016 M 26 26 85 6 passing clouds 18/09/2016 A 18 34 57 no wind passing clouds 16 277 114 25/09/2016 M 60 27 89 no wind passing clouds 25/09/2016 A 33 31 64 2 passing clouds 26/09/2016 M 54 26 87 4 Fog 27/09/2016 M 57 26 85 no wind Fog 30/09/2016 A 38 33 64 no wind Sunny 1/10/2016 A 35 33 69 no wind Sunny 1,5 308 148 12/09/2016 M 63 32 58 4 Sunny 14/09/2016 A 49 32 64 4 passing clouds 15/09/2016 A 63 32 63 no wind passing clouds 16/09/2016 M 77 26 88 no wind Clear 16/09/2016 A 56 26 85 4 Clear 8 (P) 120 37 8/09/2016 M 18 28 87 no wind Clear 8/09/2016 A 22 30 78 no wind Clear 9/09/2016 M 19 28 85 no wind passing clouds 10/09/2016 M 17 26 65 2 Sunny 12/09/2016 A 20 32 58 4 Sunny 15/09/2016 A 24 32 63 no wind passing clouds 18 183 92 10/09/2016 M 31 28 73 4 passing clouds 10/09/2016 A 28 33 59 2 Sunny 11/09/2016 M 46 28 70 no wind Clear 11/09/2016 A 53 33 58 6 passing clouds 12/09/2016 M 25 32 58 4 Sunny 6 164 85 9/09/2016 M 26 28 85 no wind passing clouds 10/09/2016 M 40 28 73 4 passing clouds 10/09/2016 A 18 33 59 2 Sunny 11/09/2016 M 33 28 70 no wind Clear 13/09/2016 A 19 29 64 4 Partly sunny 14/09/2016 M 28 32 64 4 passing clouds 7 192 69 7/09/2016 A 19 32 62 4 passing clouds 8/09/2016 M 32 28 87 no wind Clear 8/09/2016 A 35 30 78 no wind Clear 9/09/2016 M 40 28 85 no wind passing clouds 10/09/2016 M 41 26 65 2 Sunny 11/09/2016 A 25 33 58 6 passing clouds
†Session (A=Afternoon, E= Evening, M=Morning); ^ Temp= Temperature. *Industrial Area Part 1, # Industrial Area Part 2, @ Budhanpur, Indira Colony, Rajeev Colony, ^ Sector 8 perimeter
150
Supplementary Table 4.2 Details of the track lengths, duration of the photographic sight-resight surveys, and population estimates of free roaming dogs during the enumeration surveys carried out in Shirsuphal village and Panchkula Municipal Corporation administered sectors
Survey Survey Date Start time End time Date Start time End time track track
Shirsuphal *Population Estimate 391 ± 26 Track A Track length: 7.5km Track B Track length: 6km 5/06/2016 0500pm 0655pm 5/06/2016 0500pm 1841pm 6/06/2016 0630am 0814am 6/06/2016 0630am 0812am 7/06/2016 0500pm 0657pm 7/06/2016 0500pm 0700pm 8/06/2016 0630am 0810am 8/06/2016 0630am 0805am 9/06/2016 0500pm 0641pm 9/06/2016 0500pm 1842pm 12/06/2016 0500pm 0642pm 12/06/2016 0500pm 1843pm 13/06/2016 0630am 0758am 13/06/2016 0630am 0802am Panchkula 2 Track length: 13.8km; Population estimate 146 ± 8 8 Track length: 12.7km; Population estimate 14 ±11 6/10/2016 0605am 0800am 16/09/2016 0608am 0811am 6/10/2016 1612pm 1810pm 18/09/2016 0430pm 0612pm 7/10/2016 0615am 0815am 21/09/2016 0600am 0750am 7/10/2016 0419pm 0615pm 22/09/2016 0430pm 0623pm 8/10/2016 0626am 0810am 24/09/2016 0610am 0810am 6 Track length: 8.2km; Population estimate 140 ± 19 7 Track length: 6.7km; Population estimate 89 ± 8 9/09/2016 0622am 0738am 7/09/2016 0420pm 0530pm 10/09/2016 0630am 0748am 8/09/2016 0600am 0736am 10/09/2016 0410pm 0542pm 8/09/2016 0430pm 0542pm 11/09/2016 0605am 0737am 9/09/2016 0600am 0746am 13/09/2016 0412pm 0545pm 10/09/2016 0600am 0736am 14/09/2016 0606am 0735am 11/09/2016 0436pm 0614am 9 Track length: 8.4 km; Population estimate 114 ± 7 16 Track length:10.5km; Population estimate 142 ± 10 15/09/2016 0553am 0826am 25/09/2016 0620am 0820am 15/09/2016 0400pm 0536pm 25/09/2016 0435pm 0614pm 16/09/2016 0450pm 0643pm 26/09/2016 0610am 0755am 18/09/2016 0555am 0805am 27/09/2016 0620am 0818am 20/09/2016 0555am 0800am 30/09/2016 0436pm 0614pm
20/09/2016 0430pm 0615pm 1/10/2016 0426pm 0612pm 8 (P) Track length: 5.8 km; Population estimate 44 ± 4 17 Track length: 5.7 km; Population estimate 78 ± 9 8/09/2016 0600am 0700am 16/09/2016 0558am 0733am 8/09/2016 0420pm 0517pm 16/09/2016 0432pm 0535pm 9/09/2016 0600am 0700am 17/09/2016 0602am 0724am 10/09/2016 0606am 0702am 17/09/2016 0438pm 0539pm 12/09/2016 0425pm 0532pm 18/09/2016 0601am 0705am 15/09/2016 0555am 0653am 18/09/2016 0440pm 0543pm IAP 1 Track length:12.8km; Population estimate 190 ± 8 IAP 2 Track length: 9.0 km; Population estimate 164 ± 7 3/10/2016 0607am 0806am 8/10/2016 0406pm 0548pm 3/10/2016 0415pm 0610pm 9/10/2016 0610am 0733am 4/10/2016 0400pm 0612pm 9/10/2016 0419pm 0610pm 5/10/2016 0604am 0812am 10/10/2016 0606am 0752am 5/10/2016 0413pm 0618pm 10/10/2016 0412pm 0545pm 6/10/2016 0605am 0819am 11/10/2016 0620am 0809am BP,IC,RC Track length: 4.2 km; Population estimate 122 ± 9 1,5 Track length:14.7km; Population estimate 198 ± 15 21/09/2016 0559am 0712am 12/09/2016 0554am 0758am 22/09/2016 0610am 0710am 14/09/2016 0543am 0659am 23/09/2016 0602am 0715am 15/09/2016 0414pm 0607pm 24/09/2016 0553am 0654am 16/09/2016 0540am 0804am 25/09/2016 0559am 0654am 16/09/2016 0430pm 0652pm 12 Track length: 7.6 km; Population estimate 100 ± 7 18 Track length:10.4km; Population estimate 135 ± 15 12/10/2016 0411pm 0526pm 10/09/2016 0606am 0803am 13/10/2016 0640am 0800am 10/09/2016 0424pm 0608pm 14/10/2016 0624am 0742am 11/09/2016 0556am 0801am 14/10/2016 0430pm 0538pm 11/09/2016 0430pm 0600pm 15/10/2016 0418pm 0524pm 12/09/2016 0604am 0815am
151
Supplementary Table 4.3 Comparative analyses of the characteristics of free roaming dogs from the rural (Shirsuphal village) and urban (Panchkula Municipal Corporation administrated area) locations
Demographic Characteristic Rural Urban OR p Gender * Male 170 762 1 Female 70 569 0.5 (0.4-0.7) <0.001 Age Pup + Young 51 251 1 Adult + Old 212 1157 0.9 (0.6-1.2) 0.5 Body condition Good 143 955 1 Poor + fair 120 453 1.8 (1.3 - 2.3) <0.001 Activity^ Not active 354 1581 1 Active 263 1884 0.6 (0.5 - 0.7) <0.001 Proximity to garbage (≤ 20 m) ^ No 469 1787 1 Yes 148 1678 0.3 (0.2 - 0.4) <0.001
*Gender of 23 FRD in rural and 77 FRD in urban areas could not be identified and were not included in the table; ^Based on number of sightings, i.e. 617 for rural and 3465 for urban areas, respectively.
152
Supplementary Table 4.4 The characteristics (gender, age, body condition and reproductive status) of free roaming dogs in the different localities of the urban survey (Municipal Corporation, Panchkula) presented as odd ratios (OR) and their 95% confidence intervals (CI).
Gender# Age# Body condition# Whether de-sexed? #
Type of p Adult Young p poor OR p OR p locality Male Female OR (95%CI) value + Old +pups OR (95%CI) value good + fair (95%CI) value Yes No (95%CI) value
Residential 311 241 1.0 (referent) 492 104 1.0 (referent) 406 181 1.0 (referent) 178 418 1.0 (referent) Industrial 160 123 1.0 (0.7-1.3) 0.9 263 49 1.1 (0.7-1.6) 0.5 166 140 0.5 (0.4-0.7) <0.001 70 242 0.7 (0.5-0.9) 0.01 Res+Adm* 51 32 1.2 (0.8-1.9) 0.4 71 16 0.9 (0.5-1.7) 0.8 63 19 1.5 (0.8-2.5) 0.8 10 75 0.3 (0.1-0.6) <0.001 Urban village 99 62 1.2 (0.8-1.8) 0.2 103 72 0.3 (0.2-0.4) <0.001 129 30 1.9 (1.2-3.0) <0.001 46 115 0.9 (0.6-1.4) 0.7 Admin** 82 66 0.9 (0.6-1.4) 0.8 134 14 2.0 (1.1-3.7) 0.01 123 19 2.8 (1.7-4.9) 0.01 11 137 0.2 (0.1-0.3) <0.001 Perimeter 22 15 1.1 (0.6-2.2) 0.7 34 5 1.4 (0.6-4.2) 0.5 29 8 1.6 (0.7-3.8) 0.5 6 31 0.5 (0.2-1.1) 0.07 Unorganised*** 37 29 0.9 (0.6-1.6) 0.9 61 8 1.5 (0.8-3.7) 0.2 39 25 0.7 (0.4-1.2) 0.2 20 49 0.9 (0.5-1.6) 0.8
* Sectors comprising some residential and some administrative buildings. ** Administrative areas. ***The area comprised of Budhanpur, Indira colony and Rajiv colony which included unorganised houses and slums. # Bonferroni adjusted p value is 0.008
153
Chapter Five
Utilising group-size and home-range characteristics of free roaming dogs (FRD) to guide mass vaccination campaigns against rabies in India
“Rabies’ residence in people is also, by these standards, accidental, though its inability to spread through humans largely boils down to issues of anatomy and behaviour: although the virus does express itself in human saliva, humans lack a propensity to bite and the sharpened teeth with which to do it effectively.”
Bill Wasik
154
Preface
The importance of reliable population estimates of FRD and the demographic composition of these populations for interventions against dog rabies has been elaborated in the preceding chapters. However, despite the knowledge of the population size and the characteristics of the FRD subpopulations, mass parenteral vaccination campaigns fail to achieve 70% coverage due to inaccessibility of the dogs. In this chapter options for adopting oral and parenteral vaccination strategies are explored in light of the group forming tendencies exhibited by rural and urban FRD in India.
This manuscript has been selected for poster presentation at:
The text and the supplementary figures included in this chapter are the same as the manuscript published in ‘Vaccines’ except for the reference list which has been combined with references of other chapters and incorporated as one list at the end of the thesis.
This chapter can be found published as:
Tiwari, HK, M Bruce, M O’Dea, ID Robertson (2019). Utilising Group-Size and Home-
Range Characteristics of Free-Roaming Dogs (FRD) to Guide Mass Vaccination
Campaigns against Rabies in India. Vaccines, 7, 136. https://doi.org/10.3390/vaccines7040136
This manuscript was selected for oral presentation at:
21st National Conference of Association for Prevention and Control of Rabies in India,
(APCRICON) Ranchi, Jharkhand, India, 6-7 July 2019.
155
Statement of Contribution Utilising group-size and home-range characteristics Title of Paper of free roaming dogs (FRD) to guide mass vaccination campaigns against rabies in India
Publication Status
Tiwari, HK, M Bruce, M O’Dea, ID Robertson. 2019. Utilising Group-Size and Home-Range Characteristics of Free-Roaming Dogs (FRD) to Publication Details Guide Mass Vaccination Campaigns against Rabies in India. Vaccines. 7, 136. https://doi.org/10.3390/vaccines7040136
Principal Author
Name of Principal Author Harish Kumar Tiwari (Candidate) Harish Kumar Tiwari conceptualized and developed the study, planned and conducted the field study, Contribution to the Paper collected and analysed the data, interpreted the results and wrote the paper. Overall percentage (%) 60 %
Signature Date: 08/08/2019
Co-Author Contribution
By signing the Statement of Contribution, each author certifies that: i. the candidate’s stated contribution to the publication is accurate (as detailed above); ii. permission is granted for the candidate to include the publication in the thesis; and iii. the sum of all the co-author contributions is equal to 100% less the candidate’s stated contribution.
156
Name of Co-Author Emeritus Professor Ian Robertson Supervised the study and provided critical comments Contribution to the Paper to improve the interpretation of results, edited and revised the manuscript. Overall percentage (%) 20 % Signature
Date: 12/09/2019
Name of Co-Author Dr Mieghan Bruce Provided critical comments to improve the Contribution to the Paper interpretation of results, edited and revised the manuscript. Overall percentage (%) 10%
Signature Date: 24/08/2019 Name of Co-Author Dr Mark O’Dea Contribution to the Paper Provided critical comments to improve the manuscript. Overall percentage (%) 10% Signature
Date: 24/08/2019
157
Abstract
Adequate vaccination coverage of free roaming dogs (FRD) against canine rabies is not achieved due to difficulties in administering parenteral vaccinations to this population.
One factor associated with this difficulty is the tendency of FRD to form groups which increases their aggressive behaviour resulting in a significant risk of dog-bites for the vaccinators. This study investigated factors that influenced FRD forming groups and their home-ranges, using data obtained from photographic capture-recapture/sight-resight surveys conducted in rural Shirsuphal (584 sightings) and urban Panchkula (3208 sightings), India.
In the rural site older dogs (OR 0.5, 95% CI 0.2-0.9, p=0.03) and FRD sighted within
20m of garbage (OR 0.6, 95%CI 0.4-0.9, p=0.02) were less likely to be in groups. The number of dogs sighted with a FRD decreased with increased resight-probability of that dog (β= -1.0, p < 0.001). The rural FRD with smaller home-ranges were more likely to be sighted alone (OR 2.3, 95%CI 1.0-95.0, p=0.04) than those with larger home-ranges.
In the urban site, females (OR 1.3, 95%CI 1.1-1.5, p=0.002) and older dogs (OR 1.5,
95%CI 1.1-2.1, p=0.07) were more likely to be found in groups, and groups of dogs were more likely to be seen within 20 metres of garbage sites (OR 1.7, 95%CI 1.5-
2.0,p<0.001). in the urban study location. The distribution of urban FRD sighted alone, in pairs, triads, and in packs of ≥4 dogs were not random in the administrative (p = 0.02), and the two industrial (p = 0.03 & 0.01) survey tracks of the urban site, implying stable groups.The re-sighting-probability of a dog (β = 0.3, p<0.0001) and presence of garbage within 20m (β=0.2, p<0.0001) in the urban site increased the likelihood of sighting a FRD with other dogs.
158
It is concluded that data on the re-sighting probability, presence of garbage points, and home-ranges can be utilised to guide the selection of parenteral or oral rabies vaccination to achieve a population vaccination coverage of 70% to break the transmission cycle of rabies virus in FRD in India.
5.1 Introduction
The high incidence of dog-bite related rabies in India is attributed to the overwhelming presence of free roaming dogs (FRD) (Jackman and Rowan 2007). The FRD population can contain two categories of dogs: those that depend on human settlements for food and shelter; and those that are bereft of any human association and are often classified as feral dogs (Hughes and Macdonald 2013). The behavioural traits and demography of FRD are influenced by the socio-cultural and economic features of the human societies they are connected to (Serpell 2016) and their location is influenced by the availability of food and shelter, making them potentially responsible for the spread of zoonotic diseases (Dürr and Ward 2014, Traub et al. 2005). In particular, the territoriality and movement of FRD greatly influences rabies transmission and its spread can be modelled on the contact rates between FRD, which in turn depends upon their grouping behaviour and home-ranges
(Hampson et al. 2009).
The home-range of a free ranging animal is defined as the space it commonly uses for normal activities, such as foraging, hunting and whelping (Burt 1943, Calenge 2011). The availability of food and shelter, and hence the home-range of a FRD, is strongly influenced by the attitudes of the human population towards them (Macdonald and Carr
2016). Free roaming dogs are known to display altered social behaviour and travel further for activities such as mating (Boitani, Ciucci, and Ortolani 2007, Pal, Ghosh, and Roy
1998a, Meek 1999), and are also known to take isolated forays into neighbouring
159
villages/localities due to human community or climatic events (Tiwari et al. 2018), which increases the likelihood of becoming infected with rabies virus due to potential contact with a larger population of dogs or other potentially infected wildlife. Although studies on the behaviour and home-ranges of FRD have been conducted elsewhere (Boitani,
Ciucci, and Ortolani 2007, Berman and Dunbar 1983, Beck 1973, Dürr and Ward 2014,
Meek 1999) and in urban areas in India (Majumder et al. 2014, Malerczyk, DeTora, and
Gniel 2011, Pal, Ghosh, and Roy 1998b, a, Pal 2001), there is a lack of such information regarding FRD in rural areas of India.
A major impediment against successful immunisation programmes against canine rabies is the failure to achieve the prerequisite vaccination coverage (Cliquet et al. 2007), primarily due to difficulties in catching FRD and their frequent inaccessibility(Bögel and
Joshi 1990, Jibat, Hogeveen, and Mourits 2015). Nets have been used to capture and restrain dogs in India during mass parenteral immunisation campaigns against rabies
(Gibson et al. 2015); however there are significant occupational risks for those involved.
In spite of being expensive, Oral Rabies Vaccination (ORV) can be judiciously used in areas where poor catchability/accessibility of FRD precludes achieving 70% population immunisation coverage against rabies by parenteral inoculation (Cleaveland et al. 2006,
WHO 2018).
The planning and implementation of successful mass vaccination campaigns against rabies requires an understanding of the demographic characteristics of the FRD population and their propensity to form social groups (Knobel et al. 2005, Morters,
McKinley, Restif, et al. 2014, Slater 2001). Investigation of the determinants that promote
FRD to be sighted together, along with their home-ranges, can help design effective vaccination efforts including both parenteral and oral vaccination programmes. In the
160
present study, we explore whether mathematical interpretation of FRD grouping behaviour and home-ranges can inform decision-making of the most effective intervention or combination of interventions to adopt against rabies. This study was undertaken in urban and rural locations of western and northern India to: 1) investigate the predictors of group forming behaviours of FRD; 2) evaluate the home-ranges of frequently sighted FRD and their determinants; and, 3) compare and contrast the home- ranges and tendency to form groups between urban and rural FRD to make an informed decision for adopting suitable interventions against rabies in these localities.
5.2 Materials and methods
5.2.1 Study area and field methodology
This study utilised data generated during our studies on enumeration of FRD in a rural location of Shirshupal village, Maharashtra state (Tiwari et al. 2018), and 15 urban sectors administered by the Municipal Corporation, Panchkula, Haryana state, India. The details of the study area and methodology followed can be found in Tiwari, Robertson, O’Dea,
Gogoi-Tiwari, et al. (2019) and Tiwari et al. (2018).
5.2.2 Analysing distribution of different sized FRD groups across time and space
An individual FRD was classified as being sighted alone (solitary) or in a group (presence of another dog within 10m). Groups were further categorised as a pair, triad, or a pack (≥ four dogs).
To analyse the distribution of the different sized groups, we developed a model for random distribution and tested it against the data obtained from the rural and urban surveys. Assuming that 푋i is the group-size (solitary, pair, triad or pack) of the FRD, Oi is the frequency with which the group-size 푋i is sighted during the survey, and P(푥) is the
161
probability of sighting the group-size 푋i then, if FRD are randomly distributed in the village, the probability distribution, P(푥) follows a Poisson distribution, where the composition of the FRD group is independent of any particular dog in the group. As the distribution of the response variable (observed group-size of the FRD) should ideally follow a Zero-truncated Poisson distribution (ZTPD) with no chance of the frequency being 0 (every sighting comprises the detection of at least one dog), we tested the observed frequencies with expected frequencies for goodness of fit of the data to a ZTPD using a χ2 test (Majumder et al. 2014, Font 1987, Berman and Dunbar 1983).
Mathematically this equates to the following formula:
푒−휆 휆푥 1 푃(푥) = × 푥! 1 − 푒−휆
Where λ is obtained from the mean of the distribution μ, and is equal to λ÷(1-e-λ) (obtained from Molina’s table for the Poisson function) (Cohen 1960). If E풾 is the expected frequency of group-size, Χ풾 , then, E풾 = 푁 × Ρ (Χ풾 ), where 훮 is the total number of dogs sighted on each survey and is equal to ∑푖 퐸푖 푂푖. The mean of the distribution (μ) was calculated by dividing 훮 by the number of sightings of different sized groups and the values of λ for the observed and expected frequencies for the Poisson distribution were obtained from Molina’s table. In other words, if the distribution of different group-sizes were random in space and time, we could infer that the FRD sighted together were together only by chance, and not due to the existence of stable social associations (Font
1987, Majumder et al. 2014).
The distributions of the different group-sizes in rural Shirsuphal and urban Panchkula were tested for goodness of fit to the ZTPD-process for the temporal and spatial data.
162
5.2.3 Analyses of factors that influence the number of FRD sighted together
Intrinsic (age, body condition, gender, de-sexing) and extrinsic (ambient temperature, humidity, wind velocity and the dog’s proximity to garbage) influences on the tendencies of FRD to form groups were analysed on the basis of the sightings of the individually identifiable dogs during a series of photographic sight-resight surveys. The gender of some dogs could not be confirmed, and these were removed from the final analyses. The regression analyses of various predictors (categorical - age, gender, within 20m of a garbage site/bin, body condition; and numerical - temperature, humidity, wind velocity on the day of the survey and resight probability of a FRD) on the numerical response variable (actual number of FRD sighted together) were performed using generalised linear mixed model analyses (Majumder et al. 2016). Initially, the regression of the response variable (number of FRD sighted together) on the predictors were tested with univariable models. A saturated multivariable model was then constructed by including all variables yielding a p value < 0.25 on the univariable analyses. A reduced subset model was subsequently achieved by using a backward elimination approach with predictors excluded based on their AIC scores and p values. The predictors were tested for confounding by excluding them from the final model and observing the change in the coefficient values of the fixed effects. If the intercept changed by more than 20% and the
AIC score decreased by more than two points, the predictor was retained in the model, otherwise it was excluded (Richards, Whittingham, and Stephens 2011).
5.2.4 Analysis of home-range
All dogs that were sighted on more than four occasions were selected for determining home-ranges using the Maximum Convex Polygon (MCP) method. A logistic regression model was developed to investigate the determinants that affect the home-ranges, and for this purpose the two variables were dichotomised at the median score; the response
163
variable, home-range area, (≤0.11 ha and >0.11 ha in rural Shirsuphal, and ≤1.07 ha and
>1.07 ha in urban Panchkula); and the predictor variable, probability of sighted alone (≤
0.42 and >0.42 in rural Shirsuphal and ≤0.2 and >0.2 in urban Panchkula). Initially, the putative variables were tested for association with the dependent variable by a χ2 or
Fisher’s exact test (univariable analyses) and only those variables with p values of < 0.25 were selected for inclusion in the final multivariable logistic regression model with backward elimination approach.
5.2.5 Data entry, storage and cleaning and analyses
All population survey data were entered and organised in Microsoft excel (Microsoft
Excel 2007, Redmond, USA). The tracks and waypoints recorded on the GPS were converted into Excel files using the online tool, www.mygeodata
(https://mygeodata.cloud/converter/gpx-to-xlsx). The data were then cleaned by removing unwanted data records, except for individual identity and location coordinates, and converted to CSV files for use in R.
The R package “glmmTMB” was used to construct the regression model of factors that influenced the number of FRD sighted together. This package caters for the random effects (individual heterogeneity, survey-track and survey occasion) for truncated Poisson distributions with the predictors as fixed effects and the outcome as a numerical response
(Brooks et al. 2017 ). The validity of the model was tested using the ‘DHARMa’ package by visual appreciation of the distribution of the simulated residuals (number of simulations = 10,000) using Q-Q plots, along with checking the homoscedasticity by plotting fitted model predicted values vs. standardised residuals (Florian 2018 ). The normality of the residuals distribution was verified by using a one-sample Kolmogorov-
Smirnov (KS) test. The R package ‘adehabitatHR’ (Calenge 2011) was used to estimate
164
home-ranges. A logistic regression model was constructed in R and the validity of the final logistic model was tested with ‘general hoslem’ (Matthew 2017) and an ANOVA
(LRT). All statistical analyses were performed using R (R Core Team 2017).
5.3 Results
5.3.1 Sightings of the FRD in rural and urban settings as solitary or in groups
There were total of 584 and 3208 sightings of FRD with a mean re-sighting probability of 0.51 and 0.62 in the rural and urban settings, respectively. The details of the number of FRD sighted alone or in groups are presented in Table 5.1. The FRD in urban
Panchkula were more likely to be sighted in groups (OR 2.7, p < 0.001) than in rural
Shirsuphal.
Table 5.1 The number and likelihood of free roaming dogs being sighted alone or with other dogs during the photographic sight-resight survey carried out in Shirsuphal (rural) and selected sectors of Panchkula (urban), India
Number of FRD sighted; n (%) OR (95% CI) Rural Urban Total Sighted 584 (100) 3208 (100) Sighted alone 288 (49) 844 (26) 1.0 Sighted with other dogs (≥2) 296 (51) 2364 (74) 2.7 (2.3 - 3.3) * Sighted alone 288 (49) 844 (26) 1.0 Sighted in a pair 130 (22) 759 (24) 2.0 (1.6 - 2.5) * Sighted in a triad 67 (12) 580 (18) 2.9 (2.2 -3.9) * Sighted in a pack (≥4 dogs) 99 (17) 1025 (32) 3.5 (2.7 - 4.5) * Mean resight probability 0.51 (0.14-1.0) 0.62 (0.2-1.0) Mean group size† 1.98 3.03
*p value < 0.001; † FRD sighted alone were included as group size 1 in the calculation
165
No temporal variation was observed in the number of solitaries, pairs, triads or packs (≥4) of dogs sighted across the seven survey days in rural Shirsuphal (χ2 = 14.35, df =18, p =
0.7) and the distribution of the different sized groups of FRD followed a ZTPD (p = 0.09).
The temporal frequency distribution of the count of the different sized groups along all survey tracks in urban Panchkula across five days of survey followed a ZTPD, except for the perimeter and the administrative survey tracks (χ2 = 25.4, df =12, p = 0.01 and χ2 =
22.5, df =12, p = 0.03, respectively).
The spatial distribution of the solitaries, pairs, triads or packs (≥4) of FRD sighted across four major survey-tracks (A1, A2, B1, B2) in rural Shirsuphal varied significantly (χ2 =
25.42, df =9, p value = 0.0005), with thepattern for two tracks (A1 and B1) not following a ZTPD. The distribution of different group sizes across the survey tracks in urban
Panchkula also differed significantly (χ2 = 119.97, df = 26, p<0.0001), but no significant difference was observed within the tracks of similar settings(residential: χ2 = 20.7, df =
15, p = 0.14; industrial: χ2 = 0.9, df = 3, p = 0.8; part-residential part-public facilities: χ2
= 5.5, df = 6, p = 0.5; and sectors comprising unplanned colonies/villages/slums: χ2 = 5.9, df = 3, p = 0.11). However, the frequencies of the counts of different sized groups for survey tracks in Parts I (χ2 = 8.6, df = 3, p = 0.03) and II (χ2 = 10.9, df = 3, p = 0.01) of the industrial sectors, the administrative sector (χ2 = 10.1, df = 3, p = 0.02) and the perimeter track (χ2 = 10.9, df = 3, p = 0.01) did not follow a ZTPD.
Female FRD was more likely to be sighted with other dogs than male FRD in rural (OR
1.8, 95%CI 1.3-2.7, p=0.0008) and urban settings (OR 1.3, 95%CI 1.3,1.1-1.5, p=0.002).
Old FRD were more likely to be sighted with other dogs in urban Panchkula (OR 1.5,
95%CI 1.1-2.1 p=0.007) than adult dogs, whereas in rural Shirsuphal old FRD were less likely to be sighted with other dogs (OR 0.5, 95%CI 0.2-0.9, p=0.03). Similarly in urban
166
PanchkulaFRD were more likely to be sighted in a group (≥ 2 dogs) within 20m of garbage sites than alone (OR 1.7, 95%CI 1.5-2.0, p<0.001). In contrast in rural Shirsuphal
FRD were less likely to be sighted in a group within 20m of garbage sites (OR 0.6, 95%CI
0.4-0.9, p=0.02). The odds for different categories of FRD (gender, age, body condition, de-sexed and within 20m of a garbage site) being sighted alone or together with other dogs are presented in Table 5.2.
5.3.2 Determinants of number of FRD sighted together or alone
The univariable analyses of factors that may influence the number of FRD which were sighted together in rural and urban settings are presented in Table 5.3. In the final multivariable model for dogs in rural settings, resight probability (β= -1.0, SE = 0.2, p <
0.001), fair body condition (β= -0.3, SE = 0.1, p = 0.008) and within 20m of garbage (β
= 0.2, SE = 0.1, p = 0.03) were found to have a significant effect on the actual number of
FRD sighted together (Table 5.4). The model was found to fit the data well with the residuals on 10,000 simulations being normally distributed (Kolmogorov-Smirnov test,
D = 0.05, p = 0.15).
The final multivariable model for urban settings determined that the re-sighting probability (β = 0.3, SE=0.06, p<0.0001) and the sighting of a FRD within 20m of garbage
(β = 0.2, SE = 0.03, p<0.0001) positively influenced the actual number of FRD sighted together (Table 5.5). A negative influence was observed for adult (β = -0.2, SE = 0.07, p
= 0.03) and old (β = 0.3, SE = 0.09, p = 0.02) FRD. De-sexed animals were less likely to be sighted with other FRD than entire dogs (β= -0.07, SE = 0.03, p=0.05). The influence of temperature and humidity was significant but inconsequential considering the per unit change on the number of FRD sighted together. The model could not be evaluated for goodness of fit by testing the simulated residuals as it failed to converge after 10,000
167
iterations. However, the model did fit the data well when random subsets (150 to 250 observations) of the data were tested for normal distribution of the residuals.
5.3.3 Home-range of FRD and its determinants
The mean home-ranges of dogs with four or more sightings were similar (p=0.3); rural
FRD (n=29) had a median home-range of 2.8 ha (minimum = 0.0021 ha, maximum=
50.48 ha) and urban FRD (n=74) had a mean home-range of 3.7 ha (minimum 0.0065 ha, maximum= 40.6 ha) (Supplementary Table 5.1). The test of associations between the home-ranges and the predictor variables for rural Shirsuphal and urban Panchkula are presented in Table 5.6. Although, the variable ‘gender’ yielded a p value of 0.26, being a variable of consequence, it was included for building the multivariable regression model.
In the final multivariable logistic regression model for the factors influencing the dichotomised median home-range (>0.42, ≤ 0.42 ha) of FRD in rural Shirsuphal, gender and the probability of being sighted alone were retained (Table 5.7). The Hosmer-
Lemeshow test for goodness of fit could not be applied to evaluate the model due to the small number of observations but it was found to be weakly stable on the ANOVA (LRT) test (p = 0.02). The association between the two predictors in the final model was not significant (p = 0.11). A final multivariable logistic regression model for the influences on the dichotomised median home-range of urban FRD could not be produced as only one factor (body condition) significantly influenced home-range size.
168
Table 5.2 The odd ratios of various categories of FRD being sighted in groups (≥ 2) in rural (Shirsuphal) and urban (Panchkula) settings.
RURAL (Shirsuphal) URBAN (Panchkula) Number sighted in Number sighted in Factors Number sighted a group* (%) OR (95% CI) p - value Number sighted a group* (%) OR (95% CI) p - value Gender Gender Male 415 192 (46) 1.0 - 1791 1282 (72) 1.0 - Female 169 104 (62) 1.8 (1.3-2.7) 0.0008 1417 1082 (76) 1.3 (1.1-1.5) 0.002 Age† Age† Pup 30 16 (53) 1.0 (0.5-2.3) 0.88 118 99 (84) 1.9 (1.2-3.2) 0.01 Young 100 51 (51) 1.0 (0.6-1.5) 0.86 328 257 (78) 1.3 (1.0-1.7) 0.04 Adult 416 216 (52) 1.0 - 2584 1894 (73) 1.0 - Old 38 13 (34) 0.5 (0.2-0.9) 0.03 178 114 (64) 1.5 (1.1 -2.1) 0.007 Body condition Body condition Good 361 210 (58) 1.0 - 2212 1635 (74) 1.0 - Fair 199 119 (60) 1.06 (0.7-1.5) 0.7 521 376 (72) 0.9 (0.7-1.1) 0.4 Poor 24 18 (75) 2.1 (0.8-6.0) 0.1 475 353 (74) 1.02 (0.8-1.3) 0.8 Proximity to garbage (within 20m) Proximity to garbage (within 20m) Yes 140 50 (36) 1.0 - 1571 1241 (79) 1.0 - No 444 206 (46) 0.6 (0.4-0.9) 0.02 1637 1124 (69) 1.7 (1.5-2.0) <0.0001 De-sexed De-sexed Yes 0 0 849 595 (70) 1.0 - No 584 296 (51) 2359 1769 (75) 1.3 (1.1-1.5) 0.005 *Group refers to ≥ 2 dogs together. † Age was assessed based on visual characteristics as: Pup (≤ 6 months), Young (6 months to 1 year), Adult (> 1 to 7 years), and Old (≥ 7years).
169
Table 5.3 The coefficients, standard error and p values of the univariable generalised linear mixed models for the regression of predictor variables on the actual group size of FRD* in rural Shirsuphal# and urban Panchkula† in India
RURAL (Shirshuphal) URBAN (Panchkula) Factors Coefficient (β) SE p value Coefficient (β) SE p value Resight Probability -0.9 0.2 0.0004 0.3 0.07 <0.001 Temperature 0.02 0.02 0.3 -0.02 0.003 <0.001 Humidity -0.004 0.004 0.3 0.006 0.001 <0.001 Wind velocity 0.01 0.008 0.2 -0.03 0.01 0.003 Gender Female 1.0 - - 1.0 Male -0.14 0.14 0.3 -0.03 0.03 0.5 Age@ Pup 1.0 - - 1.0 - - Young -0.18 0.33 0.6 -0.01 0.1 0.9 Adult -0.04 0.30 0.9 -0.2 0.1 0.01 Old -0.28 0.41 0.5 -0.3 0.1 0.03 Body condition Good 1.0 - - 1.0 - - Fair -0.3 0.14 0.02 -0.05 0.04 0.3 Poor -0.8 0.4 0.05 0.006 -0.05 0.9 Proximity to garbage site (within 20m) No 1.0 - - 1.0 Yes 0.2 0.1 0.03 0.2 0.03 <0.001 De-sexed NA NA NA No 1.0 - - Yes -0.08 0.04 0.03
* Size of groups in rural Shirsuphal ranged between 1 (single dog) to 5 FRD per group and and between 1 to 12 FRDin urban Panchkula; #Variance from the random effects due to repeat identity of FRD on successive surveys and day of the survey was negligible (<0.05 and <1.5x 10-9 , respectively); †Variance ± SD from the random effects due to repeat identity of FRD on successive surveys and survey-tracks was 0.2±0.4 and, 0.04±0.2, respectively for all variables; @Age was assessed based on visual characteristics as: Pup (<6 months), Young (6 months to 1 year), Adult (>1 to 7 years), and Old (≥ 7years); NA- Not Applicable
170
Table 5.4 Final multivariable model of predictors that significantly influenced the number of FRD sighted together (1 to 5) in a rural setting*
Factors Coefficient (β) SE p value Constant 0.6 0.1 Resight Probability -1.0 0.2 <0.001 Wind velocity 0.02 0.01 0.07† Body condition Good 1.0 - - Fair -0.3 0.1 0.008 Poor -0.7 0.4 0.06 Proximity to garbage site (≤20m) No 1.0 - Yes 0.2 0.1 0.03
*Variance from the random effect due to individual heterogeneity = 0.4 ± 0.6; day of survey = 0.0001 ± 0.01; † removal of the factor enhanced the AIC (1509.4) of the model, consequently a more stable model was selected including this factor (AIC 1507.9)
Table 5.5 Final multivariable model of predictors that significantly influenced the number of FRD sighted together (1 to 12) in an urban setting*
Factors Coefficient (β) SE p value Constant -0.7 0.4 Resight Probability 0.3 0.06 <0.0001 Temperature 0.02 0.01 0.03 Humidity 0.01 0.003 <0.0001 Age@ Pup 1.0 - - Young -0.01 0.08 0.9 Adult -0.2 0.07 0.03 Old -0.3 0.09 0.02 De-sexed? No 1.0 - - Yes -0.07 0.03 0.05† Proximity to garbage site (≤20m) No 1.0 - - Yes 0.2 0.03 <0.0001
*Variance from the random effect due to individual heterogeneity = 0.2 ± 0.4; survey-track = 0.03 ± 0.2; † removal of the factor enhanced the AIC (11569.8) of the model, consequently a more stable model was selected including this factor (AIC 11568.1); @Age was assessed based on visual characteristics as: Pup (< 6 months), Young (6 months to 1 year), Adult (>1 to 7 years), and Old (≥ 7 years) .
171
Table 5.6 Test of association between dichotomised home-range* and predictor variables for the rural (Shirsuphal) and urban (Panchkula) locations
Rural* Urban* Variable N = 29 n (%) OR (95%CI ) p value N = 74 n (%) OR (95%CI ) p value Gender Female 11 4 (36) 1.0 - 33 17 (52) 1.0 - Male 18 11 (61) 2.6 (0.5-17.5) 0.26 41 21 (51) 1.0 (0.4-2.5) 0.9 Age† Pup 1 1 (100) - - 4 3 (75) 1.0 - Young 7 3 (43) 1.0 0.3 9 6 (67) 0.7 (0.01 -14.3) 1 Adult 18 9 (50) 1.3 (0.2-7.7) 0.3 54 25 (46) 0.3 (0.005-3.9) 0.7 Old 3 2 (67) 2.7 (0.2-45.1) 0.5 7 4 (57) 0.5 (0.006-10.5) 0.5 Body condition Fair 10 4 (40) 1.0 - 19 4 (21) 1.0 - Good 17 9 (53) 1.7 (0.4-8.2) 0.5 44 28 (63) 6.3 (1.6-30.9) 0.002 Poor 2 2 (100) - 1 11 6 (54) 4.2 (0.7-30.8) 0.06
Probability of being sighted alone# High 16 6 (37) 1.0 - 9 5 (56) 1.0 - Low 13 9 (69) 3.5 (0.7-19.0) 0.13 65 33 (45) 1.2 (0.2-6.7) 0.9 Proximity to garbage@ No 17 9 (53) 1.0 - 25 11 (44) 1.0 - Yes 12 6 (50) 0.9 (0.2-4.1) 0.99 49 27 (55) 1.5 (0.6-4.2) 0.36 De-sexed NA NA NA NA No 49 28 (57) 1.0 - Yes 25 10 (40) 0.5 (0.2-1.3) 0.16
*Home-range dichotomised as ≤0.11 ha and >0.11 ha (the median home-range of the sampled population in Shirsuphal) and ≤1.07 ha and >1.07 ha (the median home-range of the sampled population in Panchkula); †Age was assessed based on visual characteristics as: Pup (< 6 months), Young (6 months to 1 year), Adult (>1 to 7 years), and Old (≥ 7years).# the probability of a free roaming dog sighted solitary was dichotomised as high or low ( >0.42 or ≤ 0.42 for rural Shirsuphal, and >0.2 or ≤ 0.2 for urban Panchkula); @ at least one sighting within 20m of a garbage dump across the survey period.
172
Table 5.7 Final multivariable logistic regression model with regression coefficient values for predictors yielding significant p- values for dichotomised home-range area for free roaming dogs in rural Shirsuphal, India
OR (95% Factors N=29 Coefficient (β) SE p value CI) Constant -2.2 1.2 Probability of being sighted alone High (> 0.42) 16 - - 1.0 Low (≤ 0.42) 13 2.3 1.1 0.04 2.3 (1.0-95.0) Gender Female 11 - - 1.0 Male 18 2.1 1.2 0.07 8.1 (0.8-81.3)
Likelihood ratio test (LRT): Deviance 7.4, df =2, p value 0.02 *Home-range dichotomised as ≤0.11 ha and >0.11 ha (the median home-range of the sampled population in Shirsuphal)
5.4 Discussion
In the current study the likelihood of sighting a dog in a group was higher in the urban study site than in the rural site. However, irrespective of the study site, FRD were sighted in groups close to a food resource, similar to that reported by Berman and Dunbar (1983) in the city of Berkeley, USA. The proportion of FRD sighted in groups in urban Panchkula
(73.7%) was higher than in Berkeley (17.8%) because in Panchkula, FRD were observed to group in public places, such as outside of temples and community markets where they potentially received food from visitors and had access to shelter (shaded areas) which may not have been the case in the Berkeley study. The difference in the proportion of
FRD forming groups in rural and urban sites in our study is possibly due to a lack of organised garbage management in rural Shirsuphal that results in edible litter being thrown indiscriminately throughout the location, including lanes (roadways), reducing the need for dogs to congregate at specific garbage refuse points. In urban Baltimore, USA a higher proportion of FRD were sighted in groups which was believed to be due to the
173
presence of garbage in the alleyways (Beck 1973, Berman and Dunbar 1983). In contrast in urban Panchkula edible waste was usually deposited at the assigned waste disposal points, such as garbage bins and dumps, although these were frequently overflowing with litter, accounting for the likely congregation of FRD at these locations.
The observations regarding the categories (gender, age, body condition, proximity to garbage, and de-sexing) likely to be sighted in groups were consistent for both rural
Shirsuphal and urban Panchkula, except for the proximity to garbage points (Table 5.2).
The temporal distribution of the groups across the days of the sight-resight surveys was random in rural Shirsuphal and for most of the survey tracks in urban Panchkula.
However, the spatial frequency distribution of various sized groups in three of the tracks in rural Shirsuphal was not random, implying the tendency of FRD to form stable groups at these sites. A possible explanation for stable groups in rural Shirsuphal could be the presence of lactating bitches (13%) and bitches in oestrous (6%) on these tracks
(Majumder et al. 2016, Daniels and Bekoff 1989, Boitani, Ciucci, and Ortolani 2007,
Ivanter and Sedova 2008). The FRD groupings in rural Shirsuphal can be attributed to individual preferences of the rural FRD to stay together rather than the need to form stable groups for hunting/sourcing food (Majumder et al. 2014). This observation is supported by Boitani, Ciucci, and Ortolani (2007), who speculated that FRD have a lower tendency to form stable social groups in locations where food is readily available. Nevertheless, it is possible that factors other than lactation and oestrous, such as level of human interaction, may be important and further studies are warranted in similar locations to identify these factors which restrict group formation for rural FRD.
In urban Panchkula, the evidence of temporally and spatially stable groups (p = 0.01 and p = 0.02, respectively) in the perimeter survey track in urban Panchkula is not unexpected
174
as FRD are known to form temporary groups while transiting between their usual locations (Berman and Dunbar 1983), such as between adjacent sectors in this study.
However, the distribution of the different sized groups also did not follow a ZTPD in the industrial (p = 0.03 and p = 0.01 for Part I and Part II, respectively) and administrative sectors (p = 0.02), which suggests the presence of stable hierarchical social groups, similar to that reported in some international studies (Bonanni et al. 2010, Cafazzo et al.
2010) and elsewhere in India (Pal, Ghosh, and Roy 1998a). Daniels and Bekoff (1989), who similarly found that the group sizes of FRD in Mexico did not follow a ZTPD, believed that the dog-ownership practices of the resident community in the area influences the level of social organisation of FRD. The authors claimed that the dogs that were cared for by owners did not form groups with conspecifics and thus the pattern of their social organisation would differ depending upon their response to the amount and provision of food resources. In our study, although the dogs were not owned in rural
Shirsuphal, they were provided food by the rural residents (Tiwari, Vanak, et al. 2019), hence limiting the tendency of rural FRD to form groups.
FRD that had been neutered were less likely to be part of a group than entire FRDs (β =
-0.08, p=0.05) its likelihood of being part of a group (Table 5.5). This finding is supported by the presence of stable groups in the industrial and administrative sectors in urban
Panchkula where more dogs had not been neutered (Section 5.3.1). We previously reported that the odds of sighting a de-sexed FRD in industrial (OR 0.7, p value=0.01) and administrative (OR 0.2, p value <0.001) survey tracks was lower compared to residential sectors (Sections 4.2.2 & 4.3.6). Interestingly, this finding is similar to the study in the city of Petrozavodsk, Russia, where the number of stray dogs forming groups in industrial zones was higher than in residential areas where most dogs were solitary
(Ivanter and Sedova 2008). This was believed to be linked with the presence of secluded
175
breeding dens and fewer instances of human interference restricting the formation of packs in these zones. Another plausible explanation of sighting fewer de-sexed FRD could be that they are harder to catch, being frequently sighted in groups. The inaccessibility for parenteral vaccination is pronounced in stable FRD groups, possibly due to their enhanced agonistic behaviour towards humans and the tendency to avoid humans when in groups (Fox, Beck, and Blackman 1975). An anecdotal admission by the dog catchers of Panchkula Municipal Corporation regarding the heightened risk of dog- bites in the industrial and administrative sectors also corroborated the higher likelihood of stable social groups in these sectors.
The average home-ranges (2.8 ha and 3.7 ha for rural and urban settings, respectively) are comparable to those reported in USA, Australia, Russia and India for urban FRD (Pal,
Ghosh, and Roy 1998b, Berman and Dunbar 1983, Ivanter and Sedova 2008), although were much lower than that reported in a recent study in Chile (65 ha) (Pérez et al. 2018) and Aboriginal and Torres Strait Islander communities in Northern Australia (40-104 ha)
(Dürr and Ward 2014). Although our study limited the inclusion of FRD to those with a high resight probability (≥ 0.7), most of these individuals were sighted alone. In rural
Shirsuphal, dogs that were most likely sighted in a group were likely to have larger home- ranges (OR 2.3, 95%CI 1.0-95.0, p=0.04, Table 5.7) than those mostly sighted alone. This finding, in conjunction with that of the influence of resight probability on grouping tendency (β= -1.0, SE = 0.2, p < 0.001, Table 5.4), imply that rural FRD find the resources
(food, shelter) for survival within a small area and hence do not need to wander. More than half of the rural FRD (59%) for whom the home-range calculation was possible were never seen in the vicinity of garbage points, and the remaining were sighted near garbage
(< 20m) only once or twice during the survey period. This strongly supports the belief
176
that the dogs with smaller home-ranges have higher human affinity in rural Shirsuphal and thus may be more amenable for administering parenteral vaccination against rabies.
Notwithstanding the comparable home-ranges in rural and urban settings, the grouping tendency of FRD in urban Panchkula needs to be examined in relation to the resight probability. An increase in resight probability implied smaller sized groups in rural
Shirsuphal (β = -1.0, SE = 0.2, p value< 0.001), whilst, in contrast, in urban Panchkula it indicated a larger group size (β = 0.3, SE=0.06, p<0.0001) implying that visibility of the urban FRD that are in groups and around a food resource (garbage) is high (Tables 5.1 and 5.2).Even though we failed to produce a multivariable model that could have shed light on the influence of re-sight probability on home-ranges on FRD in urban Panchkula, we assert that re-sighting probability is positively related to the tendency of urban dogs to form groups (Table 5.5).
These observations can be corroborated with other findings where stray/semi-owned dogs remain in proximity to people who provide food to them, even when those people do not claim ownership of the dogs (Rubin and Beck 1982, Durr et al. 2017, Pérez et al. 2018).
Consequently, we conclude that FRD in rural Shirsuphal are more accepting of human proximity than those in urban Panchkula. Human influence over the home-ranges of FRD was also demonstrated by Ivanter and Sedova (2008) and Boitani, Ciucci, and Ortolani
(2007) who reported smaller home-ranges for dogs in areas close to humans. This finding, in conjunction with that of group-size, implies that if a photographic capture-recapture survey reveals solitary dogs with high re-sighting probability and small home-ranges, then more dogs will be accessible for parenteral vaccination against rabies. We speculate this may largely be true for rural India, although it is recommended that more studies are conducted to support this recommendation. In contrast, ORV should be adopted in areas
177
where FRD are more likely to be sighted in groups. Consequently, based on our findings of urban Panchkula we also suggest that, irrespective of the measure of the home-ranges, a high re-sight probability of FRD is indicative of a higher proportion of groups and hence
ORV should be implemented to achieve adequate herd immunity. Oral rabies vaccination has been recommended where catchability of FRD for parenteral immunisation is difficult
(Ratsitorahina et al. 2009). Although ORV has the disadvantages of high cost and the need for strict supervision, its potential use to augment parenteral inoculation to achieve
70% coverage in FRD has been advocated (WHO 2018). Once the required number to vaccinate 70% of the total population is estimated through a reliable enumeration technique (Tiwari et al. 2018), it is recommended that a combination of parenteral vaccination and ORV is adopted to achieve adequate mass immunisation against canine rabies in India.
There were some limitations encountered during the current investigation. Firstly, we used GPS fixes of the FRD at pre-set survey times in the multiple sight-resight survey sessions. The use of GPS collars would have given more frequent GPS fixes and hence a better assessment of home range, although this would have been offset by a smaller sample size due to significantly higher costs associated with this methodology. However, the need for multiple survey sessions can be off-set by using GPS-collars on randomly selected FRD from different locations of the study sites to allow estimation of average home ranges. Secondly, only one rural village was included in this study and more villages should be studied to improve the robustness of the findings. Thirdly, we were unable to develop a multivariable logistic model for factors influencing the home-range of the urban FRD, and consequently data on other influences need to be collected. Finally, error due to observer’ fatigue towards the end of multiple sight-resight surveys cannot be completely ruled out (McCallum 2005). In spite of these limitations, this investigation
178
linked the behaviour of FRD and home-ranges with their accessibility for future vaccination programmes to achieve the necessary herd immunity to control canine rabies.
We recommend more studies are conducted at different urban and rural locations in India to model the group-size of FRD, based upon easily measurable predictors, including home-ranges to make informed decisions on the FRD mass vaccination approach to adopt to control rabies.
5.5 Ethical approval
Ethics approval for the observation of the FRD in the rural and urban areas was obtained from ATREE (Ashoka Trust for Research in Ecology and the Environment)
(AAEC/101/2016).
Author contributions
All authors have contributed and approve the contents of this article. HT developed the study, collected and analysed the data; HT and IR wrote the article; MB, MO, AV, IR provided critical revision and helped interpretation of contents and implications.
Funding
The work was funded by Research grant to HT from Murdoch University, Western
Australia, Australia.
179
Acknowledgements
Murdoch University International Postgraduate Scholarship (MIPS) to HT is gratefully acknowledged. The authors are grateful to Pranav Panwalkar, Pradeep Satpute, Reetika
Maheshwari for their assistance during the conduct of the study. The help of the Principals and the students of Government PG College, Sector 1, and Government College of Girls,
Sector 14, Panchkula is also duly acknowledged. The authors also express their gratitude to the residents of Shirsuphal and Municipal Corporation, Panchkula.
Conflict of Interest Statement
The authors declare that the research was conducted in the absence of any commercial, financial or non-financial relationships that could be construed as a potential conflict of interest.
180
Supplementary Figure 5.1 Boxplots for the univariable analyses of regression of the various intrinsic and extrinsic factors on the group size of free roaming dogs sighted during the enumeration survey in Shirsuphal village of western India in June 2016
181
Supplementary Figure 5.2 Boxplots for the univariable analyses of regression of the various intrinsic factors on the group size of free roaming dogs sighted during the enumeration survey in Municipal Corporation Panchkula in north India during September – October 2016
182
Supplementary Figure 5.3 Boxplot graphics for the univariable analyses of regression of the various extrinsic factors on the group size of free roaming dogs sighted during the enumeration survey in rural and urban setting
Urban Rural
183
Supplementary Figure 5.4 Boxplots for the univariable analyses of regression of resight probability on the group size of free roaming dogs sighted during the enumeration survey in rural and urban settings
Supplementary Figure 5.5 Boxplots for the univariable analyses of de-sexed status probability on the group size of free roaming dogs sighted during the enumeration surveys in the urban setting in Panchkula
184
Chapter Six
Knowledge, attitudes and practices (KAP) towards rabies and free roaming dogs (FRD) in Shirsuphal village in western India: A community based cross- sectional study
“Almost half of the population of the world lives in rural regions and mostly in a state of poverty. Such inequalities in human development have been one of the primary reasons for unrest and, in some parts of the world, even violence.”
APJ Abdul Kalam
185
Preface
The predominant cause of human mortality due to rabies in India arises from transmission of the rabies virus by dog-bites. The uncontrolled presence of dogs, especially the free roaming ones on the streets, not only ensures circulation of this virus among the FRD population, but also is a serious public nuisance resulting in road accidents, contamination of public places with dog faeces and dispersal of garbage. A tolerance by the general public towards the presence of FRD in public spaces and a lack of responsible ownership of dogs among communities vulnerable to dog-bites are major hindrances towards instituting control measures against rabies in India. A general lack of awareness and knowledge about the disease and inadequate practices towards management of dog-bites tend to multiply the challenges for the prevention of dog-bite related rabies in the country.
The elimination of dog rabies requires a methodological approach and the enumeration of FRD and knowledge of their demography and behavioural characteristics are important aspects to achieve this, however, such approaches can seldom succeed if they are not supported by the affected communities. This chapter deals with exploring the knowledge, attitudes and practices of the rural community in the location where the dog enumeration surveys (Chapter Two) were conducted.
The text of this chapter is identical to that in the manuscript published in ‘PLoS Neglected
Tropical Diseases’ except for the reference list which has been combined with references of other chapters and incorporated as one list at the end of the thesis. In addition,
Supplementary Tables 6.1 – 6.4 have been added at the end of the chapter to further support the manuscript.
186
This chapter can be found published as:
Tiwari HK, O’Dea M, Robertson ID, Vanak AT (2019). Knowledge, attitudes and practices (KAP) towards rabies and free roaming dogs (FRD) in Shirsuphal village in western India: A community based cross-sectional study. PLoS Neglected Tropical
Diseases 13(1): e0007120 https://doi.org/10.1371/journal.pntd.0007120 .
187
Statement of Contribution
Knowledge, attitudes and practices (KAP) towards rabies and free roaming dogs (FRD) in Shirsuphal Title of Paper village in western India: A community based cross- sectional study
Publication Status
Tiwari HK, O’Dea M, Robertson ID, Vanak AT (2019) Knowledge, attitudes and practices (KAP) towards rabies and free roaming dogs (FRD) in Shirsuphal village in western India: A community Publication Details based cross-sectional study. PLoS Neglected Tropical Diseases 13(1): e0007120. https://doi.org/10.1371/journal.pntd.0007120
Principal Author
Name of Principal Author Harish Kumar Tiwari (Candidate) Harish Kumar Tiwari conceptualised and developed the study, planned and conducted the Contribution to the Paper field study, collected and analysed the data, interpreted the results and wrote the paper. Overall percentage (%) 60 % Signature Date: 08/08/2019
Co-Author Contribution
By signing the Statement of Contribution, each author certifies that: i. the candidate’s stated contribution to the publication is accurate (as detailed above); ii. permission is granted for the candidate to include the publication in the thesis; and iii. the sum of all the co-author contributions is equal to 100% less the candidate’s stated contribution.
188
Name of Co-Author Emeritus Professor Ian Robertson Supervised the study and provided critical Contribution to the Paper comments to improve the interpretation of results, edited and revised the manuscript. Overall percentage (%) 20 % Signature
Date: 12/09/2019
Name of Co-Author Dr Abi Tamim Vanak Provided critical comments to improve the Contribution to the Paper interpretation of results, edited and revised the manuscript. Overall percentage (%) 10%
Signature Date:08/08/2019 Name of Co-Author Dr Mark O’Dea Provided critical comments to improve the Contribution to the Paper manuscript. Overall percentage (%) 10% Signature
Date: 24/08/2019
189
Abstract
The lack of awareness about dog-bite related rabies in the rural population of developing countries, including India, is a major impediment to controlling the incidence of disease in humans. A survey of 127 rural residents was undertaken in Shirsuphal village in western India using a structured questionnaire to assess the influence of demographic and pet/livestock owning characteristics on the knowledge, attitudes and practices of the respondents towards rabies and free roaming dogs (FRD). Multivariable logistic regression models were constructed and the knowledge of the rural residents of
Shirsuphal village was found to be significantly influenced by family size (OR 2.1,
95%CI 1.0-4.6, p=0.04) and poultry ownership (OR 2.3, 95%CI 1.1-4.9, p= 0.03), while their attitudes towards FRD was significantly influenced by age of the respondents (OR
2.6, 95 % CI 1.2-5.8) and ownership of cattle/buffalo (OR 2.2, 95% CI 1.1-5.5). Although the knowledge score about rabies was high, a comprehensive understanding of the disease was lacking. Concerted efforts to widen the knowledge about rabies and promote healthier practices towards FRD are recommended.
6.1 Introduction
India has the world’s highest number of dog-bite related rabies deaths, most of whom are people of low socio-economic background from rural areas (WHO 2013, Acharya, Kaur, and Lakra 2012). A gross lack of awareness about rabies in rural India is one of the factors that leads to high human mortality from the disease (Sudarshan et al. 2007). Although mortality can be prevented through prompt washing of bite wounds with soap and water
(Muriuki 2016, Singh and Choudhary 2005), along with timely administration of rabies immunoglobulins (RIG) and anti-rabies vaccines (ARV) (Tschopp, Bekele, and Aseffa
2016, Dhiman, Thakur, and Mazta 2017), these practices are potentially undermined by
190
widespread traditional healing practices, such as application of chilli/turmeric powder to bite wounds (Kamble et al. 2016, Patil et al. 2014). Policy makers and the general population lack awareness about theimpact of rabies (Maroof 2013, Dodet, Adjogoua, et al. 2008) which results in insufficient vaccination coverage of dogs, poor knowledge of post-exposure prophylaxis (PEP) amongst medical professionals and unreliable supply of
ARV and RIG (Kole, Roy, and Kole 2014). Also, insufficient financial resources, poor health care infrastructure and inadequate reporting systems leads to an underestimation of the true public health impact of rabies in India (Maroof 2013, Banyard et al. 2013).
Free roaming dogs (FRD), which are responsible for 96% of all human rabies deaths in
India, are ubiquitous in both rural and urban localities/communities (Acharya, Kaur, and
Lakra 2012). Management of the FRD population, along with responsible ownership of dogs, are key strategies to minimise human deaths from dog-bite related rabies (Taylor et al. 2017). Although studies in India have assessed the knowledge, attitudes and practices
(KAP) of communities towards rabies (Tripathy, Satapathy, and Karmee 2017, Chandan and Kotrabasappa 2016, Herbert, Basha, and Thangaraj 2012, Prakash, Bhatti, and
Venkatesh 2013), studies on the community’s attitudes and understanding of FRD are lacking.
Although India contributes 4.4% of the total global research output on rabies, there is a lack of studies focussing on the vector demography, risk factors, epidemiological studies and economic evaluations of the disease (Kakkar et al. 2012). There is also a lack of awareness by the rural population about rabies control programmes (Abbas and Kakkar
2015). Paucity of activities that can transfer knowledge about the disease to the rural population is a key concern for policy makers (Abbas and Kakkar 2013) and the importance of epidemiological studies to assess the awareness level and practices of
191
people regarding aspects of rabies control is paramount in this context (Abbas and Kakkar
2015, Kole, Roy, and Kole 2014, Burki 2008). While there have been a number of hospital based studies that have assessed KAP about rabies involving dog-bite victims, community based studies are virtually lacking in India (Davlin and VonVille 2012). In view of this deficiency, a cross-sectional community study was designed in rural Baramati, western
India, in the village of Shirsuphal to assess the: (1) KAP of the rural community towards rabies; (2) KAP of the rural community towards FRD population management; and (3)
KAP of rural dog owners on responsible ownership of dogs.
6.2 Materials and methods
6.2.1 Sample size
Recent surveys conducted in rural areas in India near to the present location formed the basis for calculating the sample size for this study. A weighted measure (93.6%) of respondents having heard of rabies from four community based cross-sectional studies carried out in neighbouring states was used to calculate the target sample size with 95% confidence and 5% error rate (Prakash, Bhatti, and Venkatesh 2013, Singh and Choudhary
2005, Joice, Singh, and Datta 2016, Chandan and Kotrabasappa 2016). With 1161 households in Shirsuphal, the required sample size was estimated to be 86. However, we had sufficient resources to administer the questionnaire to 132 respondents, of which five failed to complete the survey. Consequently, the responses of 127 participants were included in the survey analysis, thus achieving a confidence level of 98% at 2% error.
6.2.2 Study area and survey procedure
A cross-sectional household survey was undertaken in the Shirsuphal village of Baramati
Town of Pune District in Maharashtra state in western India from 13th – 21st June 2016.
The village has a population of 5512 in 1161 households (www.censusindia.gov.in,
192
accessed on 08 October 2015). The majority of the villagers are farmers, although there are some professionals and small business owners. Some farmers have also taken up poultry farming in recent years. No rabies awareness campaign or dog population control measures had been conducted in the area prior to this survey.
The houses are divided into four clusters of a similar population size in the village; however, they are not numbered. Although the total number of households in each cluster was known, there was limited information available regarding the number of households in each lane,consequently a door-to-door survey method was followed using a rolling sample method where the first randomly selected household provided information about the next available household within the cluster (Dhand et al. 2012, Kish 1998).To avoid potential bias being introduced by the respondents nominating relatives or friends, they were requested to nominate a household in a different direction to that of their friends and relatives within the cluster.A total of 33 households from each cluster were included for the questionnaire survey. The household head was approached to complete the questionnaire and if he/she were not available or not willing then a household member who was older than 18 years of age was invited to complete the questionnaire. A document outlining informed consent was read out to them in the local language (Marathi) and verbal consent obtained before administering the questionnaire. In the event of the household declining to participate in the survey then the adjacent house was selected for inclusion in the study.
6.2.3 Questionnaire design
The KAP survey was designed to: identify gaps in awareness about rabies; assess the practices that potentially contributed to the persistence of the disease in the village; assess the attitudes of the community towards FRD; and assess the attitudes of dog owners
193
towards their pets. The questionnaire consisted of closed questions on: (1) household information to assess the socio-economic status and resident profile (age, education, occupation, religion, family size, number of children below 14 years of age, and pet and livestock ownership); (2) knowledge, attitudes and practices regarding rabies (a total of
16 questions - 11 pertaining to knowledge and five pertaining to attitudes and practices towards rabies, respectively); (3) attitudes and practices towards FRD (seven questions); and (4) pet care practices adopted by dog-owners (15 questions asked only to respondents who owned pet dogs). The questions were read out to the respondents in their local language (Marathi) by the interviewer and their answers were recorded in English.
6.2.4 Data management and analysis
Answers to the questions were tabulated in a spreadsheet (Microsoft Excel, Microsoft
Corp., Redmond, WA, USA). “Not sure” responses were combined with the “No” option and “NA (not applicable)” responses were removed from the study prior to subsequent statistical analysis using the R Programming Environment (R Development Core Team
2013).
A matrix was developed to categorise the respondents into high, medium and low socio- economic status on the basis of their educational qualification and occupation on a design based on www.praja.org (accessed 18 March 2016). Subsequently, the high and medium categories were merged to obtain a binomial distribution of respondents into two socio- economic divisions: low and medium/high. The age of the respondents and the family size of the households was dichotomised into two age groups based on the median age/family size (Supplementary Table 6.1).
Analyses of the responses to the individual questions. Bivariate analyses were performed by using the chi-squared (χ2) test to compare the responses to each question relating to
194
Sections 2 and 3 of the questionnaire. Section 2 comprised of two categories: (a) knowledge of rabies; and (b) attitudes and practices towards preventing and controlling rabies. Only dog-owners were asked the questions in Section 4. The characteristics of the owned dogs and their owners’ perceptions and practices towards FRD were analysed using test of association (odds ratio). Odds ratios were calculated using the “odds ratio” package in R (Patrick 2017).
6.2.5 Univariable and multivariable analyses.
Three separate multivariable models were developed to investigate the association of various factors with the socio-demographic characteristics of the respondents. Initially, the cumulative score obtained for questions pertaining to the three response criteria
(knowledge and awareness about rabies; attitudes and practices towards rabies; and attitudes and practices towards FRD, respectively) was converted into binomial outcomes by categorising the respondents as having scored above or below the median score for each response criteria. The association between this outcome variable and various demographic and household factors was then evaluated using a χ2 test or Fisher’s exact test. All explanatory variables with a p ≤ 0.25 were offered to the multivariable logistic regression models. The reduced subset models were developed using backward elimination based on the AIC (Akaike Information Criteria) score for each model. The final multivariable logistic regression models were evaluated using Pearson’s and
Deviance residuals and its goodness–of-fit was assessed by the Hosmer-Lemeshow test
(Matthew 2017). Variables with p < 0.05 were retained in the final model.
195
6.3 Results
6.3.1 Demographic and socio-demographic characteristics of the respondents
The demographic and household characteristics of the respondents are presented in Table
6.1. The age groups and the family-size were dichotomised at the median age, i.e. 35 years (≤34 years and ≥35 years of age) and the median family size, i.e. 6 members (<6 and ≥6), respectively.
6.3.2 Respondent’s knowledge and awareness of rabies
Most respondents (97%) had heard of rabies. Of these, 98.4% knew that rabies could be transmitted through animal bites, although less than half (50, 40.7%) were aware that it could also be transmitted through licks/scratches. All respondents knew that dogs were capable of transmitting rabies but only approximately a quarter recognised that cats (22%) could also transmit the virus. Most people (86%) knew that rabies was fatal once acquired and 80% were aware that it could be prevented. Of this latter group, a similar proportion
(73%) knew that it could be prevented by administering PEP to dog-bite victims or by vaccinating dogs against rabies. Respondents from smaller households (<6 members) were more aware that rabies could be prevented by vaccination (OR 2.8, 95%CI 1.2-6.7, p=0.01) than respondents from larger households.
196
Table 6.1 Demographic characteristics of the respondents (n=127) of the Knowledge, attitudes and practices survey in the Shirsuphal village
Variable/Category n (%) Gender Male 89 (70) Female 38 (30) Age (years) (range 18-72, average 39.4) 18-34 64 (50) ≥ 35 63 (50) Socio-economic status High/middle 52 (41) Low 75 (59) Family size (range 2-16, average 6.2) <6 64 (50) ≥6 63 (50) Children (≤ 14 years) are present in the family Yes 74 (58) No 53 (42) Own a pet(s) Yes 83 (65) No 44 (35) Own a dog(s) Yes 67 (53) No 60 (47) Own livestock Yes 93 (73) No 34 (27) Own cattle/buffalo Yes 67 (53) No 60 (47)
Own sheep/goats Yes 76 (60) No 51(40) Own poultry Yes 71 (56) No 56 (44)
The association between the descriptive characteristics and the knowledge of the participants is presented at Table 6.2. Smaller family size (<6 members) and poultry ownership were found to influence knowledge about rabies in the final multivariable model (OR 2.1; 95%CI 1.0, 4.6; 2.3 95%CI 1.1, 4.9, respectively) (Table 6.3). The model was a good fit of the data with a Likelihood ratio (χ²) test value of 8.7 (p=0.04) and a
Hosmer–Lemeshow goodness of fit test result of 0.78 (p=0.37).
197
Table 6.2 Association of the knowledge of the participants about rabies with various descriptive variables
Number knowledgeable P value Variable/category (n) (%) (χ2 test) OR (95% CI) Gender Female (38) 19 (50) 1 Male (89) 57 (64) 0.14* 1.8 (0.8-3.9) Age(years) ≤ 34 (64) 36 (56) 1 ≥35 (63) 40 (63) 0.4 1.3 (0.7 - 2.8) Social status Low (75) 43 (48) 1 High/middle (52) 32(61) 0.5 1.2 (0.6 – 2.7) Family size <6 (64) 33 (52) 1 ≥6 (63) 43 (68) 0.05* 2.0 (1.0 -4.2) Children (≤ 14 years) present in the household No (53) 33 (62) 1 Yes (74) 43 (58) 0.63 0.8 (0.4 – 1.7) Pet ownership No (44) 25 (57) 1 Yes (83) 51 (61) 0.61 1.2 (0.6-2.5) Dog ownership No (60) 32 (53) 1 Yes (67) 44 (66) 0.15* 1.7 (0.8 -3.4) Livestock ownership No (34) 19 (56) 1 Yes (93) 57 (61) 0.58 1.2 (0.5 -2.8) Cattle/buffalo ownership No (59) 32 (54) 1 Yes (68) 44 (65) 0.23* 1.5 (0.7 -3.1) Sheep/goat ownership No (50) 26 (52) 1 Yes (77) 50(65) 0.14* 1.7 (0.8-3.5) Poultry ownership No (53) 26 (49) 1 Yes (74) 50 (68) 0.03* 2.1 (1.0-4.5)
*Variables offered to the multivariable model
198
Table 6.3 Final multivariable logistic regression model of factors associated with respondent’s knowledge of rabies
Variable Coefficients (β) SE p value OR (95%CI) Constant -0.45 0.5 - - Family size ≥6 1.0 <6 0.77 0.38 0.04 2.1 (1.0-4.6) Poultry ownership No 1.0 Yes 0.84 0.38 0.03 2.3 (1.1-4.9)
Likelihood ratio (χ²) test =8.7; p=0.04; Hosmer –Lemeshow goodness of fit test=0.78; p=0.37)
6.3.3 Respondent’s attitudes and practices towards rabies.
A majority (87%) of respondents were aware of the ineffectiveness of traditional applications, such as chilli /turmeric powder. Less than half (42%) of respondents believed that washing bite wounds with soap and water was beneficial. Nearly all respondents (97%) would recommend a dog-bite victim attend a hospital. Although most
(92%) respondents believed that restricting the FRD population could help control rabies, only 73% would report the presence of a rabid dog to the municipal authorities. The sheep/goat owners were more likely to perceive wound cleaning as useful (OR 2.4 95%
CI 1.2-5.4, p=0.01), while households that owned poultry were found less likely to report the presence of a rabid dog to the authorities (OR 0.3, 95%CI 0.13-0.7, p=0.01).
The association between the various descriptive variables and the attitudes and practices of the participants towards rabies are presented in Table 6.4. Pet ownership (p=0.11), dog ownership (p=0.1), and poultry ownership (p=0.19) were offered to the multivariable logistic regression model. A stable multivariate logistic regression model with significant p values could not be generated.
199
Table 6.4 Association of the attitudes and practices of the respondent’s about rabies with various descriptive variables
Respondents with positive attitudes and P value Variable/category practices (%) (χ2test) OR (95% CI) Gender Female (38) 30 (79) 1 Male (89) 63 (71) 0.38 0.6 (0.2-1.5) Age(years) 18 - 34 (64) 49 (77) 1 ≥35 (63) 44 (70) 0.39 0.71 (0.32-1.57) Social status Low (75) 54 (73) 1 High/middle (52) 39 (75) 0.7 1.16 (0.52-2.66) Family size <6 (64) 46 (72) 1 ≥6 (63) 47 (75) 0.74 1.14 (0.5-2.5) Presence of children (≤ 14 years) in the household No (53) 38 (72) 1 Yes (74) 55 (74) 0.74 1.14 (0.5-2.5) Pet ownership No (44) 36 (82) 1 Yes (83) 57 (67) 0.11* 0.49 (0.17-1.2) Dog ownership No (60) 48 (80) 1 Yes (67) 45 (67) 0.1* 0.5 (0.22-1.15) Livestockownership No (34) 24 (71) 1 Yes (93) 69 (74) 0.68 1.2 (0.48-2.8) Cattle/buffaloownership No (59) 43 (73) 1 Yes (68) 50 (73) 0.93 1.03 (0.46-2.3) Sheep/goat ownership No (50) 36 (72) 1 Yes (77) 57 (74) 0.8 1.1 (0.48-2.4) Poultry ownership No (53) 42 (79) 1 Yes (74) 51 (69) 0.19* 0.5 (0.24-1.3)
*Variables offered to the multivariable model
200
6.3.4 Respondent’s attitudes and practices towards free roaming dogs (FRD)
The responses of the participants to the questions pertaining to attitudes towards FRD are presented in Table 6.5.The younger respondents (≤34 years) did not consider FRD a threat to human health (OR 0.2, 95% CI 0.04-0.97, p=0.05), and were more likely to feed them
(OR 2.2, 95%CI 1.1-4.5, p=0.04) than older participants. Participants from the high/middle socio-economic level considered FRD were useful (OR 3.09, 95%CI 1.06-
8.97, p=0.03), were likely to feed them (OR2.81, 95%CI 1.34-5.88, p=0.005), and would take an injured stray dog to a veterinarian (OR 2.33 95%CI 1.0-5.48, p=0.04). The respondents from the low socio-economic level believed that the responsibility of the health and vaccination of FRD was with the households that fed/sheltered them (OR 2.3,
95%CI 1.04-5.1, p=0.03), a perception similar to dog owners (OR 2.9, 95%CI 1.3-6.6, p=0.04). A significant number of poultry owners reported that FRD attacked their backyard poultry for food (OR 3.2, 95%CI 1.07-12.1, p=0.02). The association between the various descriptive variables and the attitudes and practices of the participants towards
FRD is presented in Table 6.6. The respondent’s age and ownership of cattle/buffalo (OR
2.6, 95%CI 1.2-5.8; OR 2.2, 95%CI 1.1-5.5, respectively) had a positive influence on their attitudes towards FRD in the final multivariable logistic regression model (Table
6.7). The model was a good fit of the data with a Likelihood ratio (χ²) test value of 10.33
(p=0.006) and a Hosmer–Lemeshow goodness of fit test result of 0.008 (p=0.927).
201
Table 6.5 Responses to various questions pertaining to attitudes and practices regarding free roaming dogs
Criteria n (%) Are there FRDs in your locality? 127 (100) What are the source of FRD? Breeding of local FRDs 99 (78) Nearby villages 16 (13) Pets abandoned by villagers 11 (9) Are FRD useful for the society? (Yes) 18 (14) For guarding premises 4 (23) Keep away wild animals 3 (18) Keep away thieves 11 (71) Are FRD a nuisance to the society? (Yes) 109 (86) Are FRD a threat to human health? (Yes) 116 (91) Where do the FRD get their food? Garbage dumps 79 (62) Meat shop/Poultry farm waste 49 (39) Fed by residents 12 (9) Do you ever feed a FRD? (Yes) 50 (39) Reasons for feeding FRD Religious reasons 41 (84) Compassion 45 (90) Better than wasting the left-over food 42 (84) How would you rank the health of FRD in your locality? Good 51 (40) Average 44 (35) Poor 32 (25) Would take an injured FRD to a veterinarian? (Yes) 28 (22) Should residents who feed/shelter FRDs be responsible Yes 34 (27) for their health/vaccination? No 93 (73) Is health/vaccination of the FRDs a responsibility of Yes 119 (94) the government? No 8 (6) In your opinion which is the best way to control the FRD population? Culling 16 (13) Impounding 43 (34) ABC 52 (41) Garbage management 13 (10) Not sure/others 3 (2)
202
Table 6.6 Association of the attitudes and practices of the community regarding free roaming dogs with various descriptive variables
Respondents with positive attitudes p value Variable/category towards free roaming dogs (%) (χ2test) OR (95% CI) Gender of participant Female (38) 27 (71) 1 Male (89) 58 (65) 0.52 0.77 (0.32-1.73) Age(years) of participant ≤ 34 (64) 37 (58) 1 ≥35 (63) 48 (76) 0.03* 2.3 (1.08-5.07)
Social status of household 56 (75) Low (75) 1 High/middle (52) 29 (56) 0.02* 0.43 (0.2-0.91)
Family size 42 (66) < 6 (64) 1 ≥6 (63) 43 (68) 0.75 1.12 (0.53-2.38) Children (≤ 14 years) present in household 35 (66) No (53) 1 Yes (74) 50 (68) 0.85 1.07 (0.5-2.27)
Pet ownership 27 (61) No (44) 1 Yes (83) 58 (70) 0.33 1.4 (0.67-3.15)
Dog ownership 40 (67) No (60) 1 Yes (67) 45 (67) 0.75 1.02 (0.48-2.15)
Livestock ownership 20 (59) No (34) 1 Yes (93) 65 (70) 0.24* 1.6 (0.7-3.6) Cattle/buffalo ownership 34 (58) No (59) 1 Yes (68) 51 (75) 0.04* 2.2 (1.0-4.7)
Sheep/goat ownership 31 (62) No (50) 1 Yes (77) 54 (70) 0.34 1.4 (0.67-3.06)
Poultry ownership 34 (64) No (53) 1 Yes (74) 51 (69) 0.57 1.2 (0.58-2.62)
*Variables included for building the multivariable model
203
Table 6.7 Final multivariable logistic regression model of factors associated with respondent’s attitudes and practices towards free roaming dogs
Variable/category Coefficients (β) SE p value OR (95% CI) Constant -0.2 0.34 Age(years) ≤ 34 1 ≥35 0.96 0.4 0.02 2.6 (1.2-5.8) Cattle/buffalo ownership No 1 Yes 0.91 0.4 0.02 2.2 (1.1-5.5)
Likelihood ratio (χ²) test =10.33; p=0.006; Hosmer-Lemeshow goodness of fit test=0.008; p=0.9
6.3.5 Characteristics of dog owners
The characteristics of the dog owners in this study are presented in Table 6.8. Dog-owners who had a negative perception of FRD were less likely to seek veterinary attention (OR
0.3, 95%CI 0.1-1, p=0.047) for their pets. Dog owners who adopted their pets “off the street” were less likely to get their dogs vaccinated than those who either purchased them or were given them (OR 0.08, 95%CI 0.01-0.4, p=0.001). Dog owners who had an adequate knowledge about rabies (76, 59.8%) or possessed a perception that controlling
FRD would help control rabies (85, 66.9%) were not significantly different from those who didn’t own dogs (p=0.29, p=0.75).
204
Table 6.8 Characteristics of the owned dogs and owner’s perceptions and practices about their pets
Criteria Number of respondents (%) Breed of dog owned Local 56 (84) Pedigree 8* (12) Mixed 3** (4) Source of dog owned Adopted 46 (69) Gifted 11 (16) Purchased 6 (9) Offspring of owned bitches 4 (6) Preference for pedigreed breeds 25 (37) Reasons for preferring a pedigreed dog Intelligence 17 (40) Cleanliness 3 (7) Social status 3 (7) No specific reasons 2 (5) Are the pet dog(s) confined and restricted? Yes 56 (84) No 11 (16) Is your dog(s) supervised when not restricted? Respondents saying ‘always’ 26 (46) Respondents saying ‘sometimes’ 12 (22) Respondents saying ‘rarely’ 5 (9) Respondents saying ‘never’ 13 (23) Have you ever visited a veterinarian? 19 (28) Has your dog ever been vaccinated against rabies? 9 (12) Is your dog(s) sterilised? 4 (6) If not sterilised why it hasn’t been sterilised? Unaware of the procedure 15 (24) Unavailability of service 9 (14) Consider it a cruel practice 5 (8) Pet reared for breeding 5 (8) Cost of the procedure 1 (1.5) Reduces aggressive nature of the guard dogs 1 (1.5) Too young for procedure 20 (32) No specific reason 7 (11)
*5 Labradors, 2 Pomeranians and 1 Pug **Cross between pedigreed and FRD
6.4 Discussion
This questionnaire study was undertaken to assess the KAP of a rural community in
Shirsuphal village to better understand the challenges facing rural India in the quest to reduce the incidence of dog-bite rabies. Although many studies have identified a lack of
205
awareness about rabies in the rural, socio-economically poor communities of India
(Abbas and Kakkar 2013, Acharya, Kaur, and Lakra 2012, Chatterjee 2009, Chaudhuri
2015), this is one of the first studies that relates the KAP of a rural community towards rabies with FRD.
The proportion of respondents who had heard about rabies (96.4%) was higher in this survey than in previous studies from South India and from in and around Delhi (Herbert,
Basha, and Thangaraj 2012, Lai et al. 2005, Kamble et al. 2016), although it was similar to studies conducted close to the present location in rural areas of Pune, Gujarat,
Karnataka and Tamil Nadu (Prakash, Bhatti, and Venkatesh 2013, Singh and Choudhary
2005, Chandan and Kotrabasappa 2016, Joice, Singh, and Datta 2016). This increase in awareness about rabies could be attributed to prioritising the disease as important by the
Government of India in the 11th five-year-plan (2007-2012) which included creating awareness of the disease as a key focus (Kalaivani, Raja, and Geetha 2014). There is also an improved availability of PEP at local Public Health centres (personal communication,
Medical officer at Public health centre at Shirsuphal) which could also have contributed to this improved awareness. In spite of the increased proportion of respondents having heard about the disease, comprehensive understanding about the disease was lacking amongst most participants. Although the majority of the respondents were aware that rabies: could be transmitted through dog-bites, is fatal once clinical signs develop and can be prevented through post-bite anti-rabies vaccination or prophylactic vaccination of dogs, and most were unaware it could be transmitted through licks/scratches from a rabid animal or through rabid cats. Furthermore, the fact that 29% of the participants were not aware of PEP or prophylactic vaccines is of concern, particularly if they or a family member is bitten by a rabid dog. As no rabies awareness campaign had ever been delivered in the village prior to the current study, this study highlights the need to expand
206
existing programmes and to develop a structured awareness campaign in this and potentially other locations.
In this study respondents from families of a smaller size (<6) were more knowledgeable about rabies (OR 2.15, p=0.04), which is likely associated with the make-up of the families as they largely comprised adults older than 35 years of age (OR 2.08, p=0.04) who had a higher overall knowledge on rabies. This would most likely arise from more opportunity to have read, seen or heard about the disease and how it can be controlled through the media or through discussions with other community members than younger individuals. This finding is consistent with another recent KAP study in south western
Ethiopia that showed that older age groups were more knowledgeable about rabies than younger age groups (Abdela et al. 2017) but was contrary to the findings of Herbert,
Basha, and Thangaraj (2012) where older age groups had an overall lower literacy than younger age groups. However, in the latter survey the participants were “slum dwellers” who are likely to be less literate than the participants of this survey in Shirsuphal village.
Rabies causes economic losses in rural communities through deaths of livestock and poultry (Hampson et al. 2015, Shwiff, Hampson, and Anderson 2013). Although there have been no studies conducted in India to evaluate livestock losses from rabies in India, dog-bites have been reported in almost all domesticated species (Dar et al. 2014). In the current study, 20% of the participants reported that packs of FRD often attacked backyard poultry for food. This is probably linked with the fact that poultry owners were more likely to consider FRD a nuisance and consequentially they may have a better understanding of the diseases transmitted by them.
Unexpectedly more people with a good overall understanding of the practices to adopt to control rabies did not consider wound cleaning beneficial (OR 3.6, p=0.009) as reported
207
in other studies also undertaken in India (Ichhpujani et al. 2006, Menezes 2008, Santra,
Lahiri, and Ray 2015). This finding highlights a lack of understanding about important measures to prevent infection. There is a Public Health Centre (PHC) within the village and PEP is available (information obtained from the Village head and PHC staff), but a lack of awareness by the general community about preventive practices hinders seeking prompt medical attention (Ganasva, Bariya, and Shringarpure 2015, Hemachudha et al.
2013). A concerted effort to improve the knowledge of rural communities is paramount to raising the awareness of practices to adopt to counter rabies.
The current study found that the majority (86%) of participants felt that FRD were a nuisance and a similar proportion (91%) considered that they were a threat to human health. Similar perceptions have been reported from the neighbouring country of Bhutan
(Rinzin 2015, page 97). Garbage sites and waste from meat/poultry shops were identified as the main sources of food for the FRD, as also reported by El Berbri et al. (2015) in
Morocco. Effective management of garbage can help reduce the nuisance arising from
FRDs and the incidence of dog-bites (Krishna 2009, Raymond et al. 2015). Animal Birth
Control or sterilisation programs were the most favoured means for controlling the dog population, which was not unexpected as non-government organisations (NGO) and
Government agencies have promoted ABC programs over the last few years (Menezes
2008, Yoak et al. 2014, Totton et al. 2011), even though none had been conducted in the village.
Respondents from a low socio-economic level were less likely to consider FRD useful, feed them or take an injured dog to a veterinarian than participants from high/middle levels. Conversely, respondents from the high/middle socio-economic levels had more sympathetic and positive attitudes towards FRD. This could be exploited for the
208
promotion of responsible dog ownership, including adopting FRD (Taylor et al. 2017).
This is also corroborated by our finding that respondents from high/middle socio- economic class (OR=2.9, p=0.02), and those from smaller families (OR =2.1, p=0.07) were more likely to take an injured dog to a veterinarian. The smaller families in
Shirsuphal usually comprised adults and elderly people who may be more compassionate with stronger religious beliefs on the importance of feeding and/or caring for dogs than younger members of larger families. Not surprising, respondents from the high/middle socio-economic class were also more likely to feed a FRD (OR= 2.8) as a result of potentially having more disposable income than those from a lower class. Most respondents believed that it was the government’s responsibility to mass vaccinate FRD.
The cost of the rabies vaccine and the difficulty in administering prophylactic vaccination to dogs is a plausible explanation why people of a low socio-economic status from rural areas are unable or unwilling to provide these services (Kayali et al. 2003, Cliquet and
Barrat 2012).
Households with children were more likely to consider FRD a threat to human health, most likely arising from the higher incidence of dog-bites in children reported in developing countries (Sudarshan et al. 2007, Davlin and VonVille 2012). Livestock owners were less likely to feed a stray dog (OR 0.39, p=0.02) than non-livestock owners and this perception may arise from the direct economic impact of injuries/deaths or dog- bite rabies in livestock as a result of dog attacks from FRDs.
Some unexpected findings also arose from the current survey. For example, although only
9% of respondents believed that feed provided by villagers was a major source of food for FRD, a larger proportion (39%) actually admitted to feeding them. Similarly, the respondents who fed them surprisingly had a lower acceptance for FRD (OR 0.08). It is
209
likely that feeding stray dogs is strongly influenced by religious convictions, as most of the respondents were followers of Hinduism, a religion which is tolerant of animals and caring for them (Widyastuti et al. 2015, Rohlf et al. 2012), but, apparently it did not motivate them sufficiently to accept responsibility for immunisations, probably also due to issues of availability of vaccines and their cost.
The “age of the respondents” and “cattle/buffalo ownership” significantly influenced the perception of the respondents towards FRD as older respondents (> 35 years of age) had a better understanding of the problems posed by these dogs (OR 2.6, p=0.016), which is probably linked with their corresponding more awareness of rabies. Cattle/buffalo owners were found to have an overall positive attitude (OR 2.18, p=0.04) towards FRD.
Cattle/buffalo owners were more likely to own a dog (OR 4.3, p <0.001) than non- cattle/buffalo owners, which may justify their positive attitudes towards practices that help manage the FRD population.
Most dog owners (72%) had adopted offspring of FRD, indicating that the rural population is not averse to adopting local breeds. Similar results were reported in a KAP study in Nepal where 65% of pet dogs were the progeny of street dogs (Massei et al.
2017). Responsible ownership is an essential tool for population management of dogs and rabies awareness campaigns should incorporate aspects of responsible dog ownership to reduce the number of FRD (Taylor et al. 2017). Although 84% of dog owners said that they confined their dog(s), 61% also said that there were occasions when their pets wandered freely unsupervised. This would increase the risk of owned dogs contracting rabies. This is a significant health risk particularly as only 13% of dog owners had their dogs immunised against rabies. Not surprisingly dog owners with a positive perception on rabies and FRD were more likely to get their dogs vaccinated and sterilised,
210
demonstrating the importance of educational campaigns in disease control and population management.
As rabies is an ancient disease, and India has the largest population of FRD in the world, it is not surprising that most participants had heard of the disease. However even though one has heard about a disease, one’s knowledge about the practices that can truly reduce the incidence may be minimal. We suggest that instead of determining the sample size based upon “have you heard about the disease”, it may be more appropriate to base it on
“can rabies be cured?” or “can rabies be prevented?” to obtain more robust results of the level of knowledge and practices to reduce infection in humans and other animals.
Another limitation of this study was that only adults (> 18 years) were interviewed.
Understanding the knowledge level of children would be useful information as targeted education of this group can result in long-term community changes (Auplish et al. 2017).
This is even more important for rabies because younger people are more likely to interact with FRD and be at greater risk of contracting the disease. As a household was represented by only one respondent, we also accept that the opinion of one person may not indeed be a true representation of overall knowledge, attitudes and practices of all members in a family. Finally, none of the variables analysed in this study were found to significantly influence the practices considered favourable for the elimination of rabies.
This could mean that variables/factors other than those included in the current questionnaire study were associated with the villagers’ attitudes on practices regarding rabies. Potentially, religion may play a role; however, in this study the majority of participants were Hindus, precluding a comparison between different religions.
Expanding the size of the study may help overcome this limitation. This study analysed data using binary logistic regression analyses which required the KAP scores of the respondents to be dichotomised at the median resulting in inclusion or exclusion of some
211
respondents with scores near to the median. We believe that with such categorisation some factors with significant difference in scores might have been missed. Also, it may have given rise to type 1 errors such as the high knowledge scores observed for poultry owners. We also analysed the data using Poisson regression (results not shown) and found that no factors were significant on the univariable analyses. However, it is possible that this may not be the case with a larger sample size and hence we suggest that alternative regression modelling techniques may be considered in future studies to overcome this limitation.
The fight against rabies in India is unlikely to be successful unless there is concerted effort to improve the knowledge of the community about the disease. Most villagers believe that only dogs can transmit rabies; however the dangers for the rural population from cattle, buffalo, sheep or goats which contract rabies from dog-bites can still be significant
(Cleaveland, Lankester, et al. 2014). The usefulness of thorough cleaning of dog-bites was not known by the majority of respondents highlighting the need to emphasise this basic, yet important fact. A structured and sustained information campaign to raise the level of awareness and improve practices against rabies is an important tool which requires serious consideration, especially in rural areas. Focussed campaigns to advise healthy practices to adopt to care for a pet are recommended for dog owners. The villagers were receptive to the conducting of mass vaccination and sterilisation campaigns and this sentiment should be utilised to immunise FRD and control their numbers in the future, although sourcing sufficient funds for on-going programs could be challenging. A major finding of this study was the positive attitudes of residents from the high/middle socio- economic level towards FRD, which could be used to encourage adoption of dogs and responsible dog ownership. In spite of the study limitations, this study highlighted that people lack comprehensive knowledge about rabies and their perception of FRD is
212
influenced by incomplete or incorrect information. Further research is recommended to analyse the effect of predictors, such as distance from PHCs or willingness to pay for vaccination of dogs. It is also recommended to explore the use (utility) of FRD which will positively influence peoples’ attitudes towards indigenous FRD on Indian streets.
Interventions for rabies control should be seen as a key component of dog health and population management.
Ethical approval
This study involved survey of rural residents of Shirsuphal village in Baramati town of
Pune district, Maharashtra, India and the ethics approval was obtained from the Murdoch
University Human Ethics Committee (permission number: 20/2016). Oral consent was obtained from all respondents prior to their participation (Appendix).
Acknowledgments
The authors are grateful to Pranav Panvalkar, Reetika Maheshwari, and Pradeep Satpute for helping with the data collection and the residents of Shirsuphal village for allowing us to conduct this survey.
213
Supplementary Table 6.1 The matrix developed to categorise the respondents into high, middle and low socio-economic groups (www.praja.org)
Primary College Graduate Postgraduate Education level & No Formal Secondary Matriculate (year (Bachelor’s (Master’s/ Occupation/Trade Education education education (year 10) 12) degree) PhD) Unskilled F F F E E E D Skilled F F E E E D D Small traders F E E E D D D Shop owner E E E D D D C Business (including professional) E E D D D C C Clerical E D D D C C C Supervisor D D D C C C B Officer/executive D D C C C B B Senior officer D C C C B B B Professional (on job) C C C B B B A
A/B – High socio-economic group; C/D – Middle socio-economic group; E/F – Low socio-economic group
214
Supplementary Table 6.2 Bivariate analyses of responses to the questions pertaining to knowledge of rabies for various descriptive determinants *
Families with Age of respondents Family size children<14years Social status p p p p Variable n (%) ** ≤34years ≥35 years value ≤5 ≥6 value Yes No value High/middle Low value Can cats transmit rabies? Yes 27 (22) 17 10 15 12 14 13 15 12 No 96 (78) 43 53 0.09 45 51 0.42 59 37 0.37 56 40 0.8 Can rats transmit rabies? Yes 22 (18) 12 10 13 9 14 8 13 9 No 101 (82) 53 48 0.5 47 54 0.28 59 42 0.65 58 43 0.88 Can rabies be transmitted through animal bites? Yes 121 (98) 60 61 61 60 72 49 71 50 No 2 (2) 0 2 0.5# 2 0 0.49# 1 1 0.79# 0 2 0.18# Can rabies be transmitted through licks/scratches? Yes 50 (41) 26 24 24 26 28 22 34 16 No 73 (59) 34 39 0.5 36 37 0.88 45 28 0.29 37 36 0.99 Is rabies fatal? Yes 106 (86) 53 53 55 51 65 41 62 44 No 17 (14) 7 10 0.5# 5 12 0.08# 8 9 0.26 9 8 0.67 Can rabies be prevented? Yes 98 (80) 46 52 48 50 56 42 54 44 No 25 (20) 14 11 0.4 12 13 0.9 17 8 0.32 16 9 0.42 Can rabies be prevented by post-bite anti-rabies vaccines (ARV)? Yes 90 (73) 44 46 47 43 52 38 54 36 No 33 (27) 16 17 0.96 16 17 0.71 21 12 0.7 17 16 0.4 Can rabies be prevented by vaccinating dogs against the disease? Yes 90 (73) 44 46 50 40 50 40 49 41 No 33 (27) 16 17 0.96 10 23 0.01 23 10 0.22 22 11 0.22
Continued/
215
Cattle/buffalo Sheep/goat Dog ownership ownership ownership Poultry ownership p p p p Variable n (%) ** Yes No value Yes No value Yes No value Yes No value
Can cats transmit rabies? 0.9 0.57 0.81 0.6
Yes 27 (22) 14 13 16 11 17 10 17 10
No 96 (78) 51 45 51 45 58 38 15 41
Can rats transmit rabies? 0.99 0.63 0.81 0.15#
Yes 22 (18) 12 10 13 9 14 8 16 6
No 101 (82) 53 48 54 47 61 40 56 45
Can rabies be transmitted through animal bites? 0.22# 0.99# 0.52# 0.16# Yes 121 (98) 65 56 66 55 73 48 72 49 No 2 (2) 0 2 1 1 2 0 0 2 Can rabies be transmitted through licks/scratches? 0.09 0.004 0.004 0.03 Yes 50 (41) 31 19 35 15 38 12 35 15 No 73 (59) 34 39 32 41 12 36 37 36 Is rabies fatal? 0.6 0.99 0.99# 0.6 Yes 106 (86) 57 49 58 48 65 41 63 43 No 17 (14) 8 9 9 8 10 7 9 8 Can rabies be prevented? 0.31 0.72 0.78 0.77 Yes 98 (80) 54 44 59 39 54 44 58 40 No 25 (20) 11 14 16 9 13 12 14 11 Can rabies be prevented by post-bite anti-rabies vaccines (ARV)? 0.5 0.22 0.03 0.17 Yes 90 (73) 49 41 52 38 60 30 56 34 No 33(27) 16 17 15 18 15 18 16 17 Can rabies be prevented by vaccinating dogs against rabies? 0.07 0.99 0.22 0.26 Yes 90 (73) 52 38 46 44 52 38 50 40 No 33 (27) 13 20 21 12 23 10 22 11
*Only those respondents who answered yes (n=123) to “have you heard of rabies?” were administered the rest of the questionnaire. All (n=123) respondents answered “yes” to “Is rabies transmitted through dogs?”; and answered “Dogs”, when asked “Which is the most common animal responsible for bite wounds?” ** The bivariate analyses were based on the responses from 123 participants who had heard of rabies; # results from Fisher’s exact test.
216
Supplementary Table 6.3 Bivariate analyses of responses to the questions pertaining to attitudes and practices regarding rabies for various descriptive determinants
Families with Age of p p Children p Social p n (%) * respondents value Family size value <14years value status value Variable ≤34years ≥35years ≤5 ≥6 Yes No High/middle Low In your opinion will the application of local treatments, like chilli powder and turmeric, on animal-bite wounds prevent rabies? Yes 20 (16) 12 8 0.3 9 11 0.65 11 9 0.7 8 12 0.99 No 107 (84) 51 56 54 53 63 44 44 63 In your opinion should animal bite wounds be washed with soap and water to reduce the chances of rabies infection? Yes 52 (41) 21 31 0.08 24 28 0.51 33 19 0.32 23 29 0.53 No 75 (59) 42 33 39 36 41 34 29 46 In your opinion is it necessary to go to a hospital if someone is bitten by a dog, even if the injury is not severe? Yes 119 (94) 57 62 0.16# 58 61 0.5# 70 49 0.71# 50 69 0.46# No 8 (6) 6 2 5 3 4 4 2 6 Can rabies be controlled by restricting the size of the stray dog population? Yes 113 (89) 57 56 0.77# 57 56 0.59# 68 45 0.21# 46 67 0.99# No 14 (11) 6 8 6 8 6 8 6 8 If you saw a dog with signs of rabies would you inform the municipal authorities? Yes 90 (71) 41 49 0.15 40 50 0.19 54 36 0.53 40 50 0.21 No 37 (29) 22 15 12 25 20 17 12 25
Continued/
217
Dog Cattle/buffalo Sheep/goat Poultry ownership ownership ownership ownership p p p Variable n (%) * Yes No value Yes No p value Yes No value Yes No value In your opinion will the application of local treatments, like chilli powder and turmeric, on animal bites wounds prevent rabies? Yes 20 (16) 12 8 0.62 10 10 0.72 13 7 0.8# 9 11 0.18 No 107 (84) 55 52 58 49 64 43 65 42 In your opinion should animal bite wounds be washed with soap and water to reduce chances of rabies infection? Yes 52 (41) 23 29 0.1 32 20 0.13 38 14 0.01 32 42 0.53 No 75 (59) 44 31 36 39 39 36 20 33 In your opinion is it necessary to go to hospital if someone is bitten by a dog, even if the injury is not severe? Yes 119 (94) 63 56 0.99# 65 54 0.34# 74 45 0.16# 69 50 0.99# No 8 (6) 4 4 3 5 3 5 5 3
Can rabies be controlled by restricting the size of the stray dog population? Yes 113 (89) 59 54 0.8# 62 51 0.41# 71 42 0.16# 69 44 0.08# No 14 (11) 8 6 6 8 6 8 5 9 If you saw a dog with signs of rabies would you inform the municipal authorities? Yes 90 (71) 47 43 0.85 46 44 0.4 51 39 0.15 46 44 0.01 No 37 (29) 20 17 22 15 26 11 28 9
* The bivariate analysis was based on responses from all (n=127) respondents. # results from Fisher’s exact test.
218
Supplementary Table 6.4 Bivariate analyses of the response by various categories of respondents to the questions regarding their attitudes and practices against free roaming dogs
Age p Family p Families with p Social p of respondents value size value Children<14years value status value
Variable n (%) * ≤34years ≥35years ≤5 ≥6 Yes No High/middle Low Do you feel FRD in your locality are useful to society? Yes 17 (13) 9 8 0.8 6 11 0.2# 7 10 0.99# 11 6 0.03# No 110 (87) 55 55 58 52 65 45 41 69 Do you think that the FRD in your locality are a nuisance or a problem for the society? Yes 109 (86) 53 56 0.44# 57 52 0.32# 46 63 0.99# 68 41 0.07# No 18 (14) 11 7 7 11 7 11 7 11 Do you think that the FRD are a threat to human health? Yes 116 (91) 55 61 0.05# 57 52 0.75# 69 47 0.5# 68 48 0.33# No 11 (9) 9 2 7 11 5 6 7 4 Do you ever feed FRD? Yes 50 (39) 31 19 0.04 28 22 0.2 26 24 0.24 28 22 0.005 No 77 (61) 33 44 35 42 48 29 24 53 If you see an injured FRD would you take it to a veterinarian? Yes 28 (22) 17 11 18 10 14 14 16 12 No 99 (78) 47 52 0.21 45 54 0.07 60 39 0.31 36 63 0.04 In your opinion should people who feed / shelter these dogs take responsibility for their health and vaccination? Yes 34 (27) 16 18 18 16 21 13 15 19 No 93 (73) 48 45 0.64 45 48 0.64 53 40 0.62 60 33 0.03 In your opinion is it the responsibility of the government to take care of the health of FRD? Yes 119 (94) 61 58 0.49# 59 60 0.99# 70 49 0.71# 50 69 0.34# No 8 (6) 3 5 4 4 4 4 2 6
Continued/-
219
Dog Cattle/buffalo Sheep/goat Poultry ownership p ownership ownership ownership p Variable n* (%) Yes No value Yes No p value Yes No p value Yes No value Do you feel FRD in your locality are useful to society? Yes 17 (13) 11 6 0.31# 6 11 0.12# 10 7 0.1# 8 9 0.79 No 110 (87) 56 54 62 48 40 70 45 65 Do you think that the FRD in your locality are a nuisance or a problem for the society? Yes 109 (86) 55 54 0.3# 61 48 0.2# 71 38 0.01# 66 43 0.2 No 18 (14) 12 6 7 11 6 12 8 10 Do you think that the FRD are a threat to human health? Yes 116 (91) 59 57 0.21# 64 52 0.34# 72 44 0.33# 69 47 0.52# No 11 (9) 8 3 4 7 5 6 5 6 Do you ever feed FRD? Yes 50 (39) 24 26 0.38 17 33 0.003 27 23 0.003 27 23 0.4 No 77 (61) 43 34 51 26 50 27 47 30 If you see an injured FRD would you take it to a veterinarian? Yes 28 (22) 17 11 0.33 16 12 0.66 16 12 0.66 16 12 0.89 No 99 (78) 50 49 52 47 61 38 58 41 In your opinion should people who feed / shelter these dogs take responsibility for their health and vaccination? Yes 34 (27) 13 21 0.04 14 20 0.09 17 17 0.13 18 16 0.46 No 93 (73) 54 39 54 39 60 33 56 37 In your opinion is it the responsibility of the government to take care of the health of FRD? Yes 119 (94) 62 57 0.72# 64 55 0.9# 74 45 0.16# 70 49 0.71# No 8 (6) 5 3 4 4 3 5 4 4
* The bivariate analyses was based on responses from all (n=127) respondents. # Results from a Fisher’s exact test
220
Chapter Seven
Knowledge, Attitudes and Practices (KAP) towards rabies and Free Roaming Dogs (FRD) in Panchkula District of north India: A cross-sectional study of urban residents
“The art and science of asking questions is the source of all knowledge”
Thomas Berger
221
Preface
In Chapters Four and Five it was found that the patterns of sighting, demographic characteristics, and the group forming behaviour of FRD differed between rural and urban areas. These differences are influenced by the attitudes and practices of the respective communities towards FRD and the efficiency of garbage management. In the preceding chapter the KAP of the rural community towards rabies in the Shirsuphal village of western India was assessed. However, the attitudes and practices of communities that reside in urban environs may not be the same as rural residents. Consequently, in this chapter, the KAP for the urban residents in the sectors of Municipal Corporation
Panchkula in north India (Chapter Three) are assessed.
The text and the Supplementary Tables included in this chapter are the same as the manuscript published in ‘PLoS Neglected Tropical Diseases’ except for the reference list which has been combined with references of other chapters and incorporated as one list at the end of the thesis.
This chapter can be found published as:
Tiwari HK, O’Dea M, Robertson ID, Vanak AT (2019). Knowledge, attitudes and practices (KAP) towards rabies and free roaming dogs (FRD) in Panchkula district of north India: A cross-sectional study of urban residents PLoS Neglected Tropical Diseases
13(4): e0007384. https://doi.org/10.1371/journal.pntd.0007384
222
Statement of Contribution
Knowledge, attitudes and practices (KAP) towards rabies and free roaming dogs (FRD) in Panchkula Title of Paper district of north India: A cross-sectional study of urban residents
Publication Status
Tiwari HK, Robertson ID, O’Dea M, Vanak AT (2019) Knowledge, attitudes and practices (KAP) towards rabies and free roaming dogs (FRD) in Panchkula district of north India: A cross-sectional Publication Details study of urban residents. PLoS Neglected Tropical Diseases 13(4): e0007384. https://doi.org/10.1371/journal.pntd.0007384
Principal Author
Name of Principal Author Harish Kumar Tiwari (Candidate) Harish Kumar Tiwari conceptualized and developed the study, planned and conducted the Contribution to the Paper field study, collected and analysed the data, interpreted the results and wrote the paper. Overall percentage (%) 60 %
Date: 08/08/2019 Signature
Co-Author Contributions
By signing the Statement of Contribution, each author certifies that: i. the candidate’s stated contribution to the publication is accurate (as detailed above); ii. permission is granted for the candidate to include the publication in the thesis; and iii. the sum of all the co-author contributions is equal to 100% less the candidate’s stated contribution.
223
Name of Co-Author Emeritus Professor Ian Robertson Supervised the study and provided critical comments Contribution to the Paper to improve the interpretation of results, edited and revised the manuscript. Overall percentage (%) 20 % Signature
Date: 12/09/2019
Name of Co-Author Dr Abi Tamim Vanak Provided critical comments to improve the Contribution to the Paper interpretation of results, edited and revised the manuscript. Overall percentage (%) 10%
Date:08/08/2019 Signature Name of Co-Author Dr Mark O’Dea Provided critical comments to improve the Contribution to the Paper manuscript. Overall percentage (%) 10% Signature
Date: 24/08/2019
224
Abstract
Canine rabies is endemic in urban India. A questionnaire was administered to 204 residents of the urbanised municipality of Panchkula in north India to assess the influence of gender, age, family size, social status and dog ownership, over the knowledge, attitudes and practices (KAP) towards rabies control and free-roaming dogs (FRD) in their locality.
Bivariate analyses revealed significant knowledge gaps regarding crucial information on the control and transmission of rabies. Multivariable logistic regression models found that the respondents with a high/middle socio-economic status were likely to be more knowledgeable than those from low socio-economic levels (OR 3.03, 95%CI 1.5-6.0, p=0.001). Households with children ≤14 years of age were likely to be lacking in knowledge about rabies compared to households with older or no children (OR 0.5,
95%CI 0.3-0.9, p= 0.04). The attitudes and practices of the respondents towards rabies control was positive in households with a high/middle socio-economic status (OR 3.4,
95%CI 1.7-7.2, p=0.0008) but poor in older (≥ 35 years) participants (OR 0.4, 95%CI
0.2-0.7, p= 0.001). It is concluded that rabies awareness campaigns should be developed and conducted to target sectors of the urban community such as those belonging to lower socio-economic sections and schools to improve the residents’ knowledge and practices towards rabies. Educating dog owners about sterilising their pets is also recommended to alter the attitudes of the residents towards FRD population control.
7.1 Introduction
An estimated 59,000 human deaths occur annually due to rabies in the world, and 99% of this global mortality is attributed to the transmission of the virus through dog-bites
(Knobel et al. 2005). The disease is endemic in Asia with India reporting the highest number of human deaths within the region, primarily amongst people from rural areas
225
with poor socioeconomic backgrounds (WHO 2013, Meslin, Fishbein, and Matter 1994,
Acharya, Kaur, and Lakra 2012). However, the true public health impact of rabies in India is unknown due to a lack of accurate data (Maroof 2013). A gross lack of awareness about the disease is one of the prime factors that leads to under-reporting of human mortality due to rabies (Sudarshan et al. 2007). In a key epidemiological study on human rabies in
India Sudarshan et al. (2007), demonstrated the wide rural-urban divide in the incidence of rabies deaths and identified that adults are at a higher risk of rabies infection in urban than in the rural population.
As most rabies prevention centres in India are located in urban areas (Kalaivani, Raja, and Geetha 2014), one would expect lower exposure and higher treatment seeking behaviour against rabies in the urban compared to the rural population. However, rapid urbanisation and inadequate garbage management systems within urban environments facilitates the emergence of rabies through the indiscriminate breeding of free-roaming dogs (FRD) on the city streets (Castillo-Neyra et al. 2017, Butcher 1999, Krishna 2009).
These FRD in the urban localities might range from semi-owned dogs that maintain some level of human interaction through the supply of food/shelter, to being completely unrestricted feral dogs that solely depend on scavenging for their existence. The dog population in India grew the fastest in the world during 2007-2012 (Reese 2005, Bradley and King 2012) and the human-FRD conflict in urban areas is evident through the increasing incidence of unprovoked dog-bites (Jain and Jain 2014, Davlin and VonVille
2012, Ganasva, Bariya, and Shringarpure 2015, Umrigar et al. 2013, Menezes 2008, Kole,
Roy, and Kole 2014). Vanak (2017), stated that there were fewer cases of rabies in humans in urban India than in rural India due to better availability to post-exposure prophylaxis. However, as rabies is endemic in the FRD population of Indian cities, there
226
remains potential risk of rabies exposure to urban residents through dog-bites (Marathe and Kumar 2016, Bharadva et al. 2015).
An efficient rabies control programme in urban areas is characterised by application of measures that would involve: mass vaccinations and control of the movements of FRD; control of their reproduction; initiating habitat control measures such as better garbage management; and remove unsupervised dogs (WSPA 2010, WHO 2013). None of these are possible without the active participation of the people whose knowledge, attitudes and practices regarding rabies and FRD may be largely influenced by their religious, cultural or traditional beliefs. The availability and economics of preventive measures, such as anti- rabies vaccine (ARV), rabies immunoglobulins (RIG) and canine vaccines, also influence the uptake of control programmes.
In the wake of the increased FRD population in urban India it is important to assess the
KAP of urban communities, not only towards rabies, but also towards FRD, before the disease can be effectively controlled (Dalla Villa et al. 2010). KAP surveys can help to point out the inadequacies of the existing disease control programmes and help improve their effectiveness by managing the shortcomings. In the current study the KAP of a sample of residents of Panchkula Municipal Corporation in Haryana state, India was undertaken through a cross-sectional survey to assess: (1) the KAP of an urban community towards rabies and its control; (2) the KAP of an urban community towards
FRD population management; and (3) the attitudes and practices of an urban dog owning population towards responsible ownership of dogs.
227
7.2 Materials and Methods
7.2.1 Study area
A questionnaire was developed and administered to a sample of residents of Panchkula
Municipal Corporation, Panchkula district, Haryana state in north India during
September-October 2016. In the greater Panchkula district, 54.87% of the population is described as urban, of whom most reside in the wards under the administrative control of
Panchkula Municipal Corporation (Bala 2014). The total number of households in wards approved for conducting the study was 13,627, with a population of 59,306 people
(www.censusofindia.gov.in, as accessed in July 2016). The wards of the Panchkula
Municipal Corporation administrated area are numbered 9 to 16 and comprise of highly organised residential, administrative and industrial habitats interspersed with unorganised slums and villages (Duggal 2004). Ward 11 comprises of slums and urban villages with non-numbered houses, while the houses in the other wards are numbered. The number of households in the wards varies from 1,016 (ward 13) to 2,322 (ward 16) (Figure 7.1).
Administrative approval for conducting the study in the eight wards was obtained from the Panchkula Municipal Corporation.
228
Figure 7.1 Study area in Panchkula Municipal Corporation, Haryana state, India with the number of households interviewed from each ward (total households interviewed = 204)
7.2.2 Sample size
The target sample size for this study was based on the weighted average of three previous
KAP surveys conducted in urban India by calculating the average number of participants with awareness of rabies from these studies (Prakash, Bhatti, and Venkatesh 2013, Ohri et al. 2016, Herbert, Basha, and Thangaraj 2012). 87.8% of the households were assumed to be aware of rabies, and for a 95% confidence and 5% error rate, a sample size of 172 was determined from the 13,627 present in Panchkula (http://epitools.ausvet.com.au, accessed 23 April 2016). We approached slightly more households (204) to cater for refusals (2) or partially completed surveys (7).
229
7.2.3 Sampling procedure
The number of households selected from each ward was in direct proportion to the total number of households in that ward and ranged from 18 to 38. The households were randomly selected from sectors of each ward using a random number generator, except for ward 11 where the houses were not numbered, and from where 20 households were selected through a rolling sampling method (Dhand et al. 2012).
The head of the household was invited to participate; however, if the household head was not present, the oldest adult member of the family/household (> 18 years of age) was surveyed in a face-to-face situation. Prior to the commencement of the survey, the study was explained to the participants, the confidentiality of their answers confirmed and oral consent to participate obtained.
7.2.4 Questionnaire design
The aim of the KAP questionnaire was to: identify gaps in the knowledge about rabies; assess the practices of the urban residents towards the disease that potentially contributes to the persistence of rabies; evaluate the attitudes of the respondents towards FRD; and assess the attitudes and practices of urban dog owners towards their pets. The questionnaire consisted of closed questions on: (a) the demographic characteristics of the household; (b) KAP regarding rabies (16 questions of which 11 pertained to knowledge and five to attitudes and practices towards rabies, respectively); (c) attitudes and practices towards FRD (seven questions); and (d) pet care practices adopted by the dog owners (15 questions). The questions were read out to the respondents in their local language (Hindi) by the interviewer and their answers were recorded in English. The questionnaire was approved by the Murdoch University Human Ethics Committee (2016/20).
230
7.2.5 Data management and analysis
The responses on the questionnaire sheet were transferred onto an EXCEL spreadsheet
(Microsoft Excel, Microsoft Corp., Redmond, WA, USA) and made compatible for subsequent analysis using the software R (R Development Core Team 2013). A matrix was developed to categorise the respondents into high, middle, and low socio-economic status on the basis of their educational qualification and occupation on a design based on www.praja.org (accessed 18 March 2016). Subsequently, the high and middle categories were merged to obtain a binomial distribution of respondents into two socio-economic divisions: low and high/middle (Supplementary Table 6.1). The age of the respondents and the family size of the households were each dichotomised into two groups based on the median age/family size.
A bivariate analysis of the responses of the participants to the individual questions pertaining to knowledge of rabies, attitudes and practices regarding rabies control and attitudes towards FRD was carried out using a χ2 or Fisher’s exact test. The residents were categorised as having adequate or inadequate knowledge of rabies; positive or negative attitudes and practices towards controlling rabies; and positive or negative attitudes towards FRD based on the median score to the responses to the questions pertaining to the relevant sections of the questionnaire. The associations between this outcome and the various descriptive variables were initially evaluated with the χ2 or Fisher’s exact test. All descriptive variables with a p ≤ 0.25 were then offered to multivariable logistic regression models. Reduced subset models were developed using backward elimination based on the
AIC (Akaike Information Criteria) score. The final multivariable logistic regression models were evaluated using Pearson’s and Deviance residuals and goodness-of-fit was assessed by the Hosmer-Lemeshow test (Matthew 2017).
231
7.3 Results
A total of 204 respondents completed the questionnaire and their descriptive characteristics are summarised in Table 7.1.
Table 7.1 Demographic characteristics of respondents in Panchkula, India, 2016
Variable/Category n (%) Gender Male 120 (59) Female 84 (41)
Age (years) 18-34 73 (36) ≥ 35 131 (64) Socio-economic status High/middle 153 (75) Low 51 (25) Family size ≤5 147 (70) ≥6 57 (30) Family contains children ≤ 14 years Yes 121 (59) No 83 (41) Dog ownership Yes 74 (36) No 130 (64)
The median score for correct responses towards knowledge of rabies; attitudes and practices towards rabies control; and attitudes towards FRD were 8, 3 and 3 respectively.
The univariable analyses (χ2 test) of responses pertaining to knowledge of rabies that was asked of 195 (96%) respondents are displayed in Table 7.2. The variables gender (p=0.2), socio-economic status (p=0.0003), families with and without children ≤ 14 years of age
(p=0.01) and dog ownership (p=0.14) were offered to the multivariable logistic regression model to assess the participant’s knowledge (Table 7.3). The model was shown to
232
adequately fit the data with a Likelihood ratio (χ²) test of 17.6 (p=0.00015) and a Hosmer–
Lemeshow goodness of fit test value of 0.01 (p=0.91).
In the bivariate analyses of responses about knowledge on rabies (Supplementary Table
7.1), respondents from the low socio-economic group were less likely to: have heard of rabies (OR 0.1 95%CI 0.02 - 0.4, p=0.001); and know that the disease was fatal (OR 0.5,
95%CI 0.2 – 0.9, p=0.02) or preventable (OR 0.2 95%CI 0.1 - 0.5, p=0.001). They also were less knowledgeable about: the role dogs played in rabies transmission (OR 0.15,
95%CI 0.04-0.6, p=0.008); the importance of administering human prophylaxis, such as
PEP (OR 0.2, 95%CI 0.1-0.5, p=0.001); and the control of rabies through vaccination of dogs (OR 0.3, 95%CI 0.1-0.5, p=0.001). Younger respondents (18-34 years of age) were more likely to know of the possible transmission through licks/scratches from infected animals (OR 1.9, 95%CI 1.1-3.5, p=0.02). Households with children ≤ 14 years of age were less aware of rabies transmission through animal bites or through licks or scratches from infected animals (OR 0.27, 95%CI 0.05-0.9, p=0.04 and OR 0.57, 95%CI 0.3-1.0, p=0.05, respectively). Dog-owners were more aware of the usefulness of vaccination of dogs to prevent rabies (OR=3.0, 95%CI 1.4-7.1, p=0.003) than non-dog owners.
233
Table 7.2 Test of association (χ2) between knowledge about rabies and various predictor variables in Panchkula, India, 2016
Variable/category N (%) Number knowledgeable p value OR (95% CI) n (%) Gender Female 84 (41) 51 (61) 1.0 Male 120 (59) 62 (52) 0.2* 0.7 (0.4 -1.2) Age (years) ≤ 34 73 (36) 38 (52) 1.0 ≥35 131 (64) 75 (57) 0.47 1.2 (0.7 -2.2) Socio-economic status Low 51 (25) 17 (33) 1.0 High/middle 153 (75) 96 (63) 0.0003* 3.3 (1.7-6.6) Family size
≤5 147 (70) 82 (56) 1.0 ≥6 57 (30) 31 (54) 0.86 1.05 (0.6-2.0) Children ≤ 14 years No 83 (41) 55 (66) 1.0 Yes 121 (59) 58 (48) 0.01* 0.5 (0.3-0.8) Dog ownership No 130 (64) 67 (51) 1.0 Yes 74 (46) 46 (62) 0.14* 1.5 (0.9-2.8)
N = total respondents, n = respondents having a knowledge score of ≥8, * Variables offered to the initial saturated model.
Table 7.3 Multivariable logistic regression model of factors associated with the participants’ knowledge of rabies in Panchkula, India, 2016
Variable/category Coefficient (b) SE p value OR (95% CI) Constant -0.24
Socio-economic status Low 1.0 High/middle 1.11 0.34 0.001 3.03 (1.5-6.0) Households with Children ≤ 14 years No 1.0 Yes -0.6 0.3 0.04 0.5 (0.3-0.9)
Likelihood ratio (χ²) test =17.6; p=0.0001; Hosmer–Lemeshow goodness of fit test=0.01 (p=0.91)
The variables age of the respondents (p=0.003), socio-economic status (p=0.001), family size (p=0.24) and households having children ≤ 14 years of age (p= 0.06) were offered to the multivariable logistic regression to assess participant’s attitudes and practices towards
234
rabies control based on the univariable analysis (Table 7.4). The model was shown to be a good fit of the data (Hosmer-Lemeshow goodness of fit test=0.04; p=0.83) (Table 7.5).
The bivariate analyses of the responses regarding attitudes and practices of the respondents towards rabies that help its control are presented in Supplementary Table 7.2.
Younger respondents (≤ 34 years) were more likely to use soap and water to wash dog- bite wounds (OR 2.7, 95%CI 1.4 - 5.1, p=0.001) and inform the municipal authorities if they came across a dog displaying rabies-like signs (OR 2.7, 95%CI 2.7 - 5.0, p= 0.001) than older respondents (≥ 35 years). Respondents from the high/middle socio-economic class were more likely to: approach a hospital in the event of a dog bite (OR 3.8, 95%CI
1.7-8.6, p=0.01); inform municipal authorities about sighting a dog showing rabies-like- clinical signs (OR 3.8, 95%CI 1.9-7.4, p=0.001); and believe that restricting the population of FRD would help control rabies (OR 3.8, 95%CI 1.9-7.5, p=0.001) than those from a low socio-economic class.
In the univariable analysis of the attitudes and practices towards FRD, none of the explanatory variables, when tested for association with the response variable, had p values
< 0.25 (Table 7.4), hence developing a multivariable logistic regression model was not attempted.
235
Table 7.4 Test of association (χ2) between the respondents’ attitudes and practices towards better control and prevention of rabies and their attitudes towards free-roaming dogs, and various predictor variables in Panchkula, India, 2016
Respondents with positive Respondents with positive attitudes Variable/category N (%) attitudes and practices towards p value OR (95% CI) p value OR (95% CI) towards free-roaming dogs, n (%) rabies, n (%) Gender Female 84 (41) 41 (49) 1.0 14 (17) 1.0 Male 120 (59) 51 (42.5) 0.37 0.8 (0.4-1.3) 19 (16) 0.87 0.9 (0.4-2.0) Age (years) ≤ 34 73 (36) 43 (59) 1.0 9 (12) 1.0 ≥ 35 131 (64) 49 (37) 0.003* 0.4 (0.2-0.7) 24 (18) 0.26 1.5 (0.7-3.8)
Socio-economic status Low 51 (25) 13 (25) 1.0 9 (18) 1.0 High/middle 153 (75) 79 (52) 0.001* 3.08 (1.5-6.4) 24 (16) 0.74 0.8 (0.4-2.1)
Family size ≥6 147 (70) 70 (48) 1.0 22 (15) 1.0 22 (39) ≤ 5 57 (30) 0.24* 1.4 (0.8-2.7) 11 (19) 0.45 0.73 (0.3-1.7)
Children ≤ 14 years No 83 (41) 44 (53) 1.0 11 (13) 1.0 Yes 121 (60) 48 (40) 0.06* 0.6 (0.3-1.0) 22 (18) 0.34 1.4 (0.7-3.3)
Dog ownership No 130 (64) 60 (46) 1.0 23 (18) 1.0 Yes 74 (46) 32 (43) 0.69 0.9 (0.5-1.6) 10 (13) 0.43 0.7 (0.3-1.6)
N= total respondents, n= respondents scoring knowledge score of >3, * Variables offered to the initial saturated model.
236
Table 7.5 Multivariable logistic regression model of factors associated with the respondents’ attitudes and practices towards rabies control and prevention in Panchkula, India, 2016
Variable/category Intercept(b) SE p value OR (95% CI)
Constant -0.53 Age (years) ≤ 34 1.0 ≥35 -0.98 0.31 0.001 0.4 (0.2-0.7)
Socio-economic status Low 1.0 High/middle 1.24 0.37 0.0008 3.4 (1.7-7.2)
Likelihood ratio (χ²) test = 21.2; p<0.0001; Hosmer–Lemeshow goodness of fit test = 0.04; p=0.83
The bivariate analyses of responses regarding attitudes and practices of the respondents towards FRD are summarized in Supplementary Table 7.2. Dog-owners (OR 2.3, 95% CI
1.2-4.4, p=0.008) and older residents (≥ 35 years) (OR 2.4, 95%CI 1.3-4.5, p=0.006) believed that FRD were useful to society while younger respondents (≤ 34 years) (OR
2.01, 95%CI 1.1-3.7, p=0.02), households containing children ≤ 14 years (OR 1.9, 95%CI
1.06-3.5, p=0.03), and households with a low socio-economic level (OR 2.7, 95%CI 1.3-
6.4, p=0.01) considered that FRD were a problem to their society. Younger respondents
(≤ 34 years) were less likely to take an injured dog to a veterinarian (OR 0.3, 95%CI 0.2-
0.6, p=0.002) than older respondents. More male (OR 2.2, 95%CI 1.1-4.4, p=0.02) than female respondents considered FRD a threat to human health. In contrast, respondents from the high/middle socio-economic group (OR 0.3, 95%CI 0.1-0.9, p=0.02) and dog owners (OR 0.5, 95%CI 0.2-0.9, p=0.03) did not feel that FRD were a threat to human health. Details of the attitudes and practices of the urban residents towards FRD are presented in Table 7.6. A significant association was found between respondents who
237
have good knowledge and those who have positive attitudes about rabies control (OR 3.2,
95% CI 1.8- 5.8, p<0.001).
Seventy-four (36%) of the survey participants owned one or more dogs (total of 91 dogs owned). Most (67) owned one dog, three residents owned two, three residents owned four, and one resident owned 6 dogs. The details of the owned dogs are presented in Table 7.7.
Respondents from the low socio-economic status were less likely to own a dog (OR 0.3,
95%CI 0.1-0.7, p=0.004) and if they did it was less likely to be a pedigree dog (OR 0.2,
95%CI 0.07-0.5, p=0.008).
238
Table 7.6 Respondents’ responses to various questions pertaining to attitudes and practices relevant to free roaming dogs in Panchkula, India, 2016
Criteria n (%) Are there FRD in your locality? 204 (100) Source of FRD Breeding of local FRD 115 (56) Nearby villages 74 (36) Pets abandoned by villagers 15 (8) FRD are useful for the society 55 (27) For guarding premises* 46 (84) Keep away wild animals* 5 (9) Keep away thieves* 14 (25) FRD are a nuisance to the society 138 (68) FRD are neither useful nor a problem 11(5) FRD are a threat to human health 160 (78) Food source for FRD* Garbage dumps 106 (52) Edible street litter 43 (21) Fed by residents 109 (53) Respondents who would feed a FRD 148 (72) Reasons why respondents would feed a FRD* Religious reasons 57 (37) Compassion 118 (80) Better than wasting the left-over food 114 (77) Condition of the FRD# Good 39 (19) Average 108 (53) Poor 57 (28) Would take an injured FRD to a veterinarian 78 (38) Residents who feed/shelter FRD should Yes 111 (54) be responsible for their health/vaccination No 93 (46) Health/vaccination of the FRD is the responsibility Yes 182 (89) of the government? No 22 (11) Best way to control FRD population* Culling 8 (4) Impounding 53 (26) Animal Birth Control 139 (68) Garbage management 50 (24) *Respondents could choose more than one option
239
Table 7.7 Characteristics of owned pet dogs and the owner’s perceptions and practices about their pets in Panchkula, India, 2016
Criteria Number (%) Gender of the dog(s) owned ^ Male 75 (82) Female 16 (18) Breeds of dog(s) owned ^ Pedigreed 62* (68) Local 22 (24) Mixed breed 7# (8) Source of owned dogs ^ Purchased 48 (53) Gifted 12 (13) Adopted 26 (29) Offspring of owned pet 5 (5) Respondents who preferred pedigree dog to local breed 60 (81) Reasons for preferring a pedigree dog Intelligence~ 24 (40) Cleanliness~ 12 (20) Social status~ 11 (18) Combination of above reasons 13 (22) Are the dog(s) registered ^ Yes 52@ (70) No 39 (30) Are the dog(s) confined and restricted? $ Yes 65 (88) No 9 (12) Are the dog(s) supervised when not confined/restricted? $ Always 49 (66) Sometimes 15 (20) Rarely 10 (14) Respondents who visited a veterinarian in the last year 67 (90) Number of dogs vaccinated against rabies ^ 65 (71) Number of households with sterilised dogs $ 7 (8) Respondents’ reasons for not sterilising an owned dog $ Unaware of the procedure 1 (2) Unavailability of the service 2 (3) Consider it a cruel practice 8 (11) Pet reared for breeding 14 (19) Cost of the procedure 8 (11) Pet too young for the procedure 2 (3) No specific reason 32 (43)
*19 Labradors, 15 German Shepherd dogs, 9 Pugs, 7 Pomeranians, 3 each of Rottweilers and Cocker Spaniels and one each of Bull Terrier, Dobermann Pinscher, German Spitz, Irish setter, Chihuahua and Himalayan Gaddi dog; # cross between pedigreed and FRD; ~ Respondents chose more than one option; @ Registered with Kennel Club of India; ^Total number of owned dogs = 91; $ Total number of dog-owning respondents = 74
240
7.4 Discussion
We identified a number of factors of interest regarding the KAP of respondents in
Panchkula, in particular: (a) the rabies awareness level of households from the low socio- economic level and those with children ≤14 years is significantly low; (b) the respondents in the higher age group (≥35 years) and households from the low socio-economic level have gaps in the attitudes and practices towards rabies control; and (c) dog-owning residents prefer a pedigree dog than a FRD, however, they would provide food and shelter to FRD due to compassion for them.
7.4.1 Community knowledge and awareness of rabies
A high proportion (96%) of respondents had heard of rabies, an increase from that reported in urban localities in India (69 and 74%) by Ichhpujani et al. (2006) and Herbert,
Basha, and Thangaraj (2012). More recent studies, report a higher proportion (84%), albeit lower to the present study (Tripathy, Satapathy, and Karmee 2017, Ohri et al. 2016).
Elsewhere, international studies have reported comparable levels of awareness viz. Sri
Lanka (90%); Bali, Indonesia (97%); and 94% in the Bohol Province, Philippines
(Matibag et al. 2007, Davlin et al. 2014, Widyastuti et al. 2015). This rise has been largely attributed to wide dispersion of information via television and radio sources (Widyastuti et al. 2015). While these reasons apply to increased cognisancein India as well, it may also be linked to the Government of India prioritizing rabies as a disease of importance and its inclusion in the Nation’s recent five year plan (Kalaivani, Raja, and Geetha 2014).
In Panchkula, information on rabies is spread by the Municipal Corporation through awareness campaigns, rallies and awareness quizzes in schools (personal communication,
Municipal Commissioner, Panchkula). Although these measures are intended to increase the awareness about rabies, they are also instrumental in more residents taking notice of
241
the disease, even when their knowledge about various aspects of the disease remains incomplete.
In spite of Panchkula being one of the well planned and organised municipal towns in
India (Bala 2014), with a literacy rate higher than the national average
(www.schooleducationharyana.gov.in, accessed 30 June 2017); the respondents lacked understanding of the disease, especially regarding its mode of transmission, methods to prevent infection after dog-bites, and the possibility of disease transmission by animals other than dogs or by licks and scratches from a rabid animal (Supplementary Table 7.1).
Such knowledge gaps not only contradicts the presumption that urban dwellers are well informed about the disease but also reflects the inconsistent reach of the awareness programmes in different sections of the population. Similar findings have also been reported in Dehradun (23.7%) and Delhi (42%) (Ohri et al. 2016, Sharma et al. 2016).
The respondents in Panchkula were, however, better informed about the prophylactic rabies immunisation for dogs and PEP for humans compared to studies from elsewhere in the country (Sharma et al. 2016, Ohri et al. 2016, Prakash, Bhatti, and Venkatesh 2013).
Nonetheless, a lack of vital information on transmission of rabies virus in the residents may also be due to the prime focus of measures initiated by the Municipal authorities towards control of the FRD population rather than concerted efforts to enhance general awareness of the community about rabies as a disease (Executive officer, Panchkula
Municipal Corporation). A shift in focus is recommended to make the residents aware of the routes of transmission and also simple measures that can possibly prevent rabies, such as washing of dog-bite wounds with soap and water. Awareness of the disease and processes to adopt can be improved by increasing the visibility of information through posters, print and mass media as reported in villages near Bangalore in southern India
242
following use of Information, Education and Communication (IEC) material to enhance knowledge about rabies (Sudarshan et al. 2013). Introduction of rabies information sessions in schools also helps improve knowledge and awareness about the disease as has been demonstrated in Sikkim, India (Auplish et al. 2017). As the present study reinforced the role of the socio-economic status on a participant’s knowledge (Table 7.2), awareness campaigns need adjusting to target disadvantaged groups (Sudarshan et al. 2006, Herbert,
Basha, and Thangaraj 2012). A smaller number of respondents (16) recalled the running of awareness campaigns on rabies, of which onlytwo were from low socio-economic sections of the society implying that the awareness campaigns are neither far-reaching, nor targeting the low socio-economic sector.
Conversely, there are some positive outcomes from this study, such as no significant difference between the knowledge about rabies in males and females in Panchkula
(Supplementary Table 7.1). This may be due to equal opportunity of males and females to acquire information on rabies, as opposed to the better opportunities offered to males to gather knowledge regarding rabies as reported in some studies such as in Ethiopia
(Abdela et al. 2017, Guadu et al. 2014). The prospect of equality of knowledge dissemination among genders in Panchkula could be capitalised to spread information amongst women with children as it was found that families with children of vulnerable ages (≤14 years) lacked adequate understanding of rabies. As the currently employed tool for spreading awareness in Panchkula Municipality is primarily mass media, which has similar exposure opportunities to all groups in the community, irrespective of gender or age, a targeted approach to enhance knowledge, such as in educational institutions, is recommended to educate school children about the methods of rabies virus transmission
(Auplish et al. 2017).
243
Dog ownership was found to be a trait of the economically well-off members of the urban society. The likelihood of a dog-owner belonging to a low socio-economic level in
Panchkula was found to be low (OR 0.34, p=0.04), and this explains why dog ownership is not an influencing factor for having a high knowledge score in the multivariable model.
7.4.2 Community attitudes and practices towards rabies
Failure to wash dog-bite wounds with soap and water by a large proportion of the respondents (39%) reinforces the observation by others (Burki 2008, Sambo et al. 2014) that a large section of society is unaware of a simple procedure that can help reduce the incidence of rabies substantially. It is not surprising that more than half of the respondents
(113, 55.4%) favoured the use of traditional healing applications, such as chilli powder and turmeric, similar to studies reported elsewhere in urban India including Delhi (51%)
(Kamble et al. 2016) and Dehradun (57%) (Ohri et al. 2016). It is important that awareness campaigns should emphasise that, although turmeric may have antiseptic properties (Srimal 1997), it is not able to kill the rabies virus, which has higher chances of being destroyed if wounds are properly washed with soap and water.
The urban respondents favoured restricting the FRD population but lacked genuine concern for controlling rabies as demonstrated by their negative perception to questions pertaining to practices regarding the disease (Supplementary Table 7.2). In contrast, a
KAP study in Bhutan by Dhand et al. (2012) found that most respondents who favoured
FRD population control (99.7%) also reported cases of canine rabies to the authorities
(98.8%) and practiced washing dog-bite wounds with soap and water (85.4%). In another recent study in Bhutan, the public demand for formulating legislation that could control the FRD population also implied a high level of awareness and a responsible attitude towards controlling rabies (Rinzin 2015, page 99), which unfortunately was lacking in
244
Panchkula. A positive aspect of the respondents in the present study, however, was their
PEP seeking behaviour (85%), which was similar to a survey in Eluru district (85.5%) in
Andhra Pradesh and far higher than the findings of participants from Delhi slums (27.6% and 26.5%), Pune (24%), and Dehradun (55%) (Kamble et al. 2016, Ohri et al. 2016,
Prakash, Bhatti, and Venkatesh 2013, Sharma et al. 2016, Shridevi et al. 2014). This response is largely driven by easy accessibility to hospitals/clinics in the vicinity of
Panchkula, compared to the previously mentioned areas that do not have close access to hospitals.
The socio-economically better off respondents were more likely to seek hospital treatment (OR 3.83, p<0.001), alert the municipal authorities of the presence of a rabid dog (OR 3.79, p<0.001) and support measures restricting the stray dog population (OR
2.49, p=0.01, Supplementary Table 7.3). This is not unexpected, as better knowledge about the disease should translate into adoption of better practices, as we found a significant association between respondents with good knowledge about rabies and those with positive attitudes about rabies control (OR 3.21, p<0.001).
The older respondents (≥ 35 years of age) were found to have inadequate knowledge regarding the measures that can control/prevent rabies (OR= 0.37, p=0.001). The low literacy level in the older generation and limited access to information about health and diseases via the internet may be reasons for this lack of knowledge and positive attitudes.
This, once again, warrants targeting older people by formulating a structured and sustained information campaign in urban centres, if the knowledge and practices of older residents are to be improved to reduce the incidence of rabies throughout the country.
245
7.4.3 Community attitudes and practices towards free roaming dogs
The majority of the respondents (68%) in Panchkula considered FRD a problem, which was consistent with recent findings in Bhutan (Rinzin 2015, 70%) but was lower than that reported in Abruzzo, Italy (90%) by Slater et al. (2008). This difference is most likely because of the low socio-economic standing of participants included in the current study.
However, in this study, over a quarter of the respondents felt that FRD were useful (27%) which may be due to the higher sense of security that the urban residents apparently desire, as guarding of the house premises was the most quoted utility (83.6%) by the respondents for the usefulness of FRD.
A prominent finding in this study was that 72.5% of the respondents admitted to feeding
FRD, with most (80%) doing this out of compassion, probably linked with the perceived poor welfare of these dogs as only 19% of the respondents felt that FRD were of good health. We feel that factors, other than those included in this study, influence the attitudes of the residents towards FRD, as none of the predictors included in this study were significant. The spatial distribution of the FRD was concentrated in the vicinity of community shopping centres and it is likely that the distance of the households from the shopping centres could be a significant factor that may influence the residents’ attitudes.
The city has seen many rabies awareness campaigns and dog population control interventions over the past decade, however, the programmes are often interrupted owing to failure of contracts or insufficient personnel or other resources (personal communication, Executive officer, Panchkula Municipal Corporation).
7.4.4 Characteristics of urban dog owners
Most dog owners (68%) kept pedigree dogs and not surprisingly, more owned dogs were purchased (53%) than were adopted (29%) off the streets. The rise in dog ownership in
246
India has resulted from the increased income of urban Indian residents (Bradley and King
2012). The current study found that dog owners preferred to register their dogs with the
Kennel Club India (KCI) but not their local municipality implying that registration is considered important for commercial purposes and eligibility for entering in dog shows rather than enforcing responsible dog ownership. As most dogs were pedigree and purchased, it was expected that the number of dog owners supervising their dogs when they were not restricted/confined would be high (66%). Although a majority (71%) of owned dogs were vaccinated against rabies, only 8% of them were sterilised and many dog owners (43%) could not cite any specific reasons for not having their dog sterilised.
This highlights the need for spreading information on dog population control, as well as rabies, in Panchkula.
This KAP survey in Panchkula highlights that contrary to the expected belief that urban residents are better informed about rabies, significant gaps persist in their knowledge towards the disease, especially regarding the means of transmission through licks and scratches of rabid animals in residents of low socio-economic level and in families with children of vulnerable age (≤ 14 years). Inadequate practices regarding rabies prevention were found in the older urban respondents (≥35 years of age) and those from the low socio-economic status. We recommend that in addition to the holistic efforts to spread awareness about rabies, a targeted focus on the sections of the society such as slum residents, primary schools and unskilled workers of industrial sectors should be adopted by the Municipal Corporation to improve the knowledge of these community sectors.
Implementing compulsory registration of pet dogs by Municipal Corporation will help in the monitoring of vaccination coverage of owned dogs. An incentive of free vaccination against rabies could also be started by the Municipal Corporation to reduce infection in dogs and hence the human community. Apart from wider participation of educational
247
institutes, it is suggested that the frequency of awareness campaigns should be also increased to alter the perception of the wider community, including dog owners, towards rabies and its control and management of the dog population.
This KAP survey, however, had some shortcomings. We suspect that some of the responses may not reflect the true picture, as people sometimes do not practice what they say, e.g. we feel a higher percentage of residents feed a FRD but did not acknowledge this due to the prevailing feeling by the community of their nuisance in this locality. Many dog-owners from high/middle socio-economic sections usually asked the household servants to answer the questionnaire on the pretext that in most cases it is the servant who takes care of their dog. This potentially introduced a bias as the servants do not necessarily represent the awareness and knowledge level of the dog owners. Also, the questionnaire did not explore the reasons for the 29% of dog owners failing to have their pets vaccinated against rabies which is an important aspect that should be included in future surveys. It is recommended that follow-up surveys are conducted including sampling more people per household to confirm the perceptions of urban residents towards FRD. Also, we accept that comparisons made in this study with other surveys must be interpreted with caution as the questionnaires and statistical methods vary between studies. Enlarging the sample size and repeating the surveys in other urban areas may overcome such limitations for future KAP surveys.
Ethical approval
This study involved survey of residents of Panchkula Municipal Corporation administrated wards (ward 9 to 17) in the state of Haryana, India and the administrative approval of Panchkula Municipal Corporation was obtained for the study while ethical
248
approval was obtained from the Murdoch University Human Ethics Committee
(permission number: 20/2016).
Acknowledgements
Murdoch University International Postgraduate Scholarship (MIPS) to HKT is gratefully acknowledged. The authors are grateful to the Municipal Commissioner, Panchkula for the permission to conduct the study and help by Mr Pranav Panvalker conducting the interviews is also duly acknowledged.
249
Supplementary Table 7.1 Descriptive and bivariate analyses (χ2) of the responses to the individual questions relating to knowledge about rabies amongst various predictor variables in the residents of Panchkula Municipal Corporation
Gender of respondents Age of respondents Variable N= 204 Male Female pvalue ≤34 years ≥35 years p value Have you heard of rabies? Yes 195 116 79 0.49 127 68 0.28 No 9 4 5 4 5 Can dogs transmit rabies? Yes 195 114 80 0.74 127 68 0.28 No 9 6 3 4 5 Can cats transmit rabies? Yes 67 37 30 0.45 41 26 0.82 No 137 83 54 90 47 Can rats transmit rabies? Yes 67 38 29 0.62 46 21 0.35 No 137 82 54 85 52 Can rabies be transmitted through animal bites? Yes 186 108 78 0.61 117 69 0.5 No 18 12 6 14 4 Can rabies be transmitted through licks/scratches? Yes 94 55 38 0.99 68 26 0.02 No 110 65 45 63 47 Is rabies fatal? Yes 142 80 62 0.7 95 47 0.22 No 62 40 22 36 26 Can rabies be prevented? Yes 162 94 68 0.86 102 60 0.76 No 42 26 16 29 13 Can rabies be prevented by post-bite anti-rabies vaccines (ARV)? Yes 155 90 65 0.87 101 54 0.61 No 49 30 19 30 19 Can rabies be prevented by vaccinating dogs against rabies? Yes 156 93 63 0.67 99 57 0.68 No 48 27 21 32 16
Continued/-
250
Family size Families with (members) Children ≤ 14years Variable N=204 ≤5 ≥6 p value Yes No p value Have you heard of rabies? Yes 195 139 56 0.44 116 79 0.99 No 9 8 1 5 4 Can dogs transmit rabies? Yes 195 140 55 0.72 114 81 0.31 No 9 7 2 7 2 Can cats transmit rabies? Yes 67 46 21 0.7 37 30 0.4 No 137 101 36 84 53 Can rats transmit rabies? Yes 67 47 20 0.74 34 33 0.08 No 137 100 37 87 50 Can rabies be transmitted through animal bites? Yes 186 136 50 0.28 106 80 0.04 No 18 11 7 15 3 Can rabies be transmitted through licks/scratches? Yes 94 67 27 0.9 49 45 0.05 No 110 80 30 72 38 Is rabies fatal? Yes 142 101 41 0.65 85 57 0.81 No 62 46 16 36 26 Can rabies be prevented Yes 162 117 45 0.99 93 69 0.67 No 42 30 12 28 14 Can rabies be prevented by post-bite anti-rabies vaccines (ARV)? Yes 155 108 47 0.59 90 65 0.8 No 49 39 10 31 26 Can rabies be prevented by vaccinating dogs against rabies? Yes 156 113 43 0.82 91 65 0.84 No 48 34 14 30 18
Continued/-
251
Social status Dog ownership High/ Variable N=204 middle Low p value Yes No p value Have you heard of rabies? Yes 195 151 44 0.001 73 122 0.16 No 9 2 7 1 8 Can dogs transmit rabies? Yes 195 150 45 0.001 73 122 0.16 No 9 3 6 1 8 Can cats transmit rabies? Yes 67 51 16 0.8 23 44 0.75 No 137 102 35 51 86 Can rats transmit rabies? Yes 67 44 23 0.03 23 44 0.75 No 137 109 28 51 86 Can rabies be transmitted through animal bites? Yes 186 143 43 0.05 67 119 0.8 No 18 10 8 7 11 Can rabies be transmitted through licks/scratches? Yes 94 75 19 0.14 38 56 0.3 No 110 78 32 36 74 Is rabies fatal? Yes 142 113 29 0.02 49 93 0.42 No 62 40 22 25 37 Can rabies be prevented? Yes 162 132 30 0.001 62 100 0.24 No 42 21 21 12 30 Can rabies be prevented by post-bite anti-rabies vaccines (ARV)? Yes 155 127 28 0.25 59 96 0.34 No 49 26 23 15 34 Can rabies be prevented by vaccinating dogs against rabies? Yes 156 127 29 0.001 65 91 0.003 No 48 26 22 9 39
252
Supplementary Table 7.2 Descriptive and bivariate analyses (χ2) of the responses to the individual questions related to attitudes and practices about rabies amongst various predictor variables in the residents of Panchkula Municipal Corporation
N Gender of respondents p Age of respondents p Variables =204 Male Female value ≤34years ≥35 years value In your opinion will application of local treatments, like chilli powder and turmeric, on animal bite wounds prevent rabies? 0.47 0.67 Yes 113 64 49 74 39 No 91 56 35 57 34 In your opinion should animal bite wounds be washed with soap and water to reduce chances of rabies infection? 0.39 0.001 Yes 124 70 54 69 55 No 80 50 30 62 18 In your opinion is it necessary to go to hospital if someone is bitten by a dog, even if the injury is not severe? 0.58 0.91 Yes 174 101 73 112 62 No 30 19 11 19 11 Can rabies be controlled by restricting the size of the stray dog population? 0.11 0.7 Yes 143 79 64 93 50 No 61 41 20 38 23 If you saw a dog with signs of rabies would you inform the municipal authorities? 0.75 0.001 Yes 113 63 50 62 51 No 91 57 34 69 22
Families with N Family size p children ≤ 14 years p Variables =204 ≤5 ≥6 value Yes No value In your opinion will application of local treatments, like chilli powder and turmeric, on animal bite wounds prevent rabies? 0.16 0.39 Yes 113 77 36 70 43 No 91 70 21 51 40 In your opinion should animal bite wounds be washed with soap and water to reduce chances of rabies infection? 0.39 0.51 Yes 124 92 32 66 58 No 80 55 25 55 25 In your opinion is it necessary to go to hospital if someone is bitten by a dog even if the injury is not severe? 0.47 0.62 Yes 174 127 47 102 72 No 30 20 10 19 11 Can rabies be controlled by restricting the size of the stray dog population? 0.48 0.13 Yes 143 101 42 80 63 No 61 46 15 41 20 If you saw a dog with signs of rabies would you inform the municipal authorities? 0.2 0.99 Yes 113 85 28 67 46 No 91 62 29 54 37
Continued/-
253
Social status p Dog ownership p Variables N=204 High/middle Low value Yes No value In your opinion will application of local treatments, like chilli powder and turmeric, on animal bite wounds prevent rabies? 0.12 0.55 Yes 113 80 33 39 74 No 91 73 18 35 56 In your opinion should animal bite wounds be washed with soap and water to reduce chances of rabies infection? 0.58 0.91 Yes 124 26 98 44 80 No 80 25 55 30 50 In your opinion is it necessary to go to hospital if someone is bitten by a dog even if the injury is not severe? 0.01 0.38 Yes 174 138 36 61 113 No 30 15 15 13 17 Can rabies be controlled by restricting the size of the stray dog population? 0.01 0.5 Yes 143 115 28 50 93 No 61 38 23 24 37 If you saw a dog with signs of rabies would you inform the municipal authorities? 0.001 0.24 Yes 113 97 16 45 68 No 91 56 35 29 62
Supplementary Table 7.3 Descriptive and bivariate analyses (χ2) of the responses to the questions relating to attitudes and practices towards free roaming dogs amongst various predictor variables in the residents of Panchkula Municipal Corporation Gender of respondents p value Age of respondents p value Variable N=204 Male Female ≤34 years ≥35 years Do you feel FRD in your locality are useful to society? Yes 55 27 28 0.08 27 28 0.006* No 149 93 56 104 45 Do you think that the FRD in your locality are a nuisance or a problem for the society? Yes 138 85 53 0.24 96 42 0.02* No 66 35 31 35 31 Do you think that the FRD are a threat to human health? Yes 160 101 59 0.02* 103 57 0.9 No 44 19 25 28 16 Do you ever feed FRD? Yes 148 89 59 0.53 91 57 0.18 No 56 31 25 40 16 If you see an injured FRD would you take it to a veterinarian? Yes 78 52 26 0.07 38 40 0.002* No 126 68 58 93 33 In your opinion should people who feed / shelter these dogs take responsibility for their health and vaccination? Yes 111 64 47 0.9 68 43 0.7 No 93 57 36 63 30 In your opinion is it the responsibility of the government to take care of the health of FRD? Yes 182 109 73 0.38 115 67 0.37 No 22 11 11 16 6
Continued/-
254
Families with Family size children ≤ 14years Variable N=204 ≤5 ≥6 p value Yes No p value Do you feel FRD in your locality are useful to society? Yes 55 40 15 0.89 27 28 0.56 No 149 107 42 94 55 Do you think that the FRD in your locality are a nuisance or a problem for the society? Yes 138 97 41 0.41 89 49 0.02 No 66 50 16 32 34 Do you think that the FRD are a threat to human health? Yes 160 118 42 0.3 98 62 0.28 No 44 29 15 23 21 Do you ever feed FRD? Yes 148 108 40 0.63 83 65 0.12 No 56 39 17 38 18 If you see an injured FRD would you take it to a veterinarian? Yes 78 60 18 0.22 45 33 0.71 No 126 87 39 76 50 In your opinion should people who feed / shelter these dogs take responsibility for their health and vaccination? Yes 111 81 30 0.75 64 47 0.6 No 93 66 27 57 36 In your opinion is it the responsibility of the government to take care of the health of the FRD? Yes 182 129 53 0.28 106 76 0.37 No 22 18 4 15 7
Social status Dog ownership Variable N=204 High/middle Low p value Yes No p value Do you feel FRD in your locality are useful to society? Yes 55 46 9 0.08 28 27 0.008 No 149 107 42 46 103 Do you think that the FRD in your locality are a nuisance or a problem for the society? Yes 138 96 42 0.01 45 93 0.11 No 66 57 9 29 37 Do you think that the FRD are a threat to human health? Yes 160 114 46 0.02 52 108 0.03 No 44 39 5 22 22 Do you ever feed FRD? Yes 148 107 41 0.15 54 94 0.91 No 56 46 10 20 26 If you see an injured FRD would you take it to a veterinarian? Yes 78 53 25 0.07 26 52 0.81 No 126 100 26 48 78 In your opinion should people who feed / shelter these dogs take responsibility for their health and vaccination? Yes 111 84 27 0.81 33 78 0.03 No 93 69 24 41 52 In your opinion is it the responsibility of the government to take care of the health of FRD? Yes 182 137 45 0.79 62 120 0.06 No 22 16 6 12 10
255
Chapter Eight
Knowledge, attitudes and practices towards dog-bite related rabies in para-medical staff at rural primary health centres in Baramati, western India
“Knowledge is of no value unless you put it into practice”
Anton Chekhov
256
Preface
There was a general lack of awareness about rabies and simple measures, such as washing of dog-bite wounds, which could effectively prevent infection among the surveyed rural populations. This makes them highly vulnerable to the disease. In addition, in many instances the para-medical staff of the rural Primary Health Centres (PHC) and sub- centres, who constitute the first point for rural residents seeking medical attention, also lack awareness of this vital information. Unavailability of PEP/RIG at such centres and inadequate expertise of staff on their application are also crucial factors resulting in human mortality due to rabies in the rural areas. In this chapter, the KAP of the para- medical staff in rural PHC towards dog-bite related rabies is explored.
The text of this chapter is the same as the manuscript published in ‘PLoS one’ except for the reference list which has been combined with references of other chapters and incorporated as one list at the end of the thesis.
This chapter can be found published as:
Tiwari HK, Vanak AT, O’Dea M, Robertson ID (2018) Knowledge, attitudes and practices towards dog-bite related rabies in para-medical staff at rural primary health centres in Baramati, western India. PLoS one 13(11): e0207025. https://doi.org/10.1371/journal.pone.0207025
257
Statement of Contribution
Knowledge, attitudes and practices towards dog-bite related rabies in para-medical staff at rural primary Title of Paper health centres in Baramati, western India
Publication Status
Tiwari HK, Vanak AT, O’Dea M, Robertson ID (2018) Knowledge, attitudes and practices towards dog-bite related rabies in para-medical Publication Details staff at rural primary health centres in Baramati, western India. PLoS one 13(11): e0207025. https://doi.org/10.1371/journal.pone.0207025
Principal Author
Name of Principal Author Harish Kumar Tiwari (Candidate) Harish Kumar Tiwari conceptualized and developed the study, planned and conducted the Contribution to the Paper field study, collected and analysed the data, interpreted the results and wrote the paper. Overall percentage (%) 60 %
Signatur Date: 08/08/2019
Co-Author Contributions
By signing the Statement of Contribution, each author certifies that: i. the candidate’s stated contribution to the publication is accurate (as detailed above); ii. permission is granted for the candidate to include the publication in the thesis; and iii. the sum of all the co-author contributions is equal to 100% less the candidate’s stated contribution.
258
Name of Co-Author Emeritus Professor Ian Robertson Supervised the study and provided critical Contribution to the Paper comments to improve the interpretation of results, edited and revised the manuscript. Overall percentage (%) 20 % Signature
Date: 12/09/2019
Name of Co-Author Dr Abi Tamim Vanak Provided critical comments to improve the Contribution to the Paper interpretation of results, edited and revised the manuscript. Overall percentage (%) 10%
Date:12/08/2019 Signature Name of Co-Author Dr Mark O’Dea Provided critical comments to improve the Contribution to the Paper manuscript. Overall percentage (%) 10% Signature
Date: 24/08/2019
259
Abstract
The lack of awareness regarding rabies amongst rural primary care health staff and their adverse practices towards the management of dog-bite wounds is a major contributor to the high incidence of rabies infection and subsequent human mortality in India. A
Knowledge, Attitudes and Practices survey was carried out involving 54 nursing and non- nursing staff working in eighteen rural Primary Health centres and sub-centres around
Baramati town of Pune district in Western India. Multivariable logistic regression models were constructed to assess factors that influenced knowledge of rabies and practices towards management of dog-bite related wounds. The more experienced and better- educated workers were found to have a good awareness of rabies (OR 3.4, 95%CI 1.0-
12.1) and good practices towards dog-bite wound management (OR 5.6, 95%CI 1.2-
27.0). Surprisingly, non-nursing staff were significantly more knowledgeable about rabies (OR 3.5, 95%CI 1.0-12.3), but their practices towards dog-bite wound management were inadequate (OR 0.18, 95%CI 0.04-0.8) compared to the nursing staff. A mandatory training module for primary care health staff is recommended to improve their knowledge regarding rabies and management of dog-bite wounds to reduce the incidence of human rabies in rural India.
8.1 Introduction
Rabies is a viral zoonosis transmitted through the bite of a rabid animal and affects all warm blooded animals (Crowcroft and Thampi 2015). Although it is present in most countries of the world (Shankaraiah et al. 2013), the incidence in developing nations is higher, with India contributing more than 36%of global deaths each year (Hampson et al.
2015, WHO 2018), of which the majority are as a result of bites from free roaming dogs.
However the impact of rabies in India is likely to be even larger due to an inadequate
260
reporting system (Banyard et al. 2013). The disease primarily affects disadvantaged groups, in both rural and urban areas, due to a lack of awareness of the disease, insufficient financial resources to seek medical help, poor health care infrastructure, unavailability of prophylactic and therapeutic measures and an overemphasis on the use of traditional practices for treatment and wound healing (Maroof 2013, Singh and
Choudhary 2005). The rural population is more likely to suffer higher mortalities due to a lack of infrastructure and staff to provide timely first aid in the form of wound cleaning and administration of post-exposure prophylaxis (PEP) (Knobel et al. 2005).
The role of primary healthcare staff, who are the first point of contact for dog-bite victims seeking medical intervention, is crucial for the prevention of rabies (Kole, Roy, and Kole
2014). Although the number of Primary Health Centres (PHC) in rural areas of India is increasing, the presence of sufficient adequately qualified personnel to staff these centres remains a challenge for the Indian government. Although the focus of these health centres is control of preventable diseases of children, such as diphtheria, pertussis, tetanus, measles and poliomyelitis (John et al. 2011), they are also responsible for administering anti-rabies PEP and providing first aid measures for dog-bite victims. Consequently, assessment of the knowledge, attitudes and practices (KAP) of staff towards rabies and animal bites is a key factor in the effective control of rabies (Kishore, Singh, and Ravi
2015).
There is a reported lack of adequate knowledge about the preventive measures adopted, including PEP, of primary health care professionals, especially in rural India (Kole, Roy, and Kole 2014). Furthermore, there is evidence that some physicians know little about the correct prophylactic measures to adopt to prevent rabies (Chowdhury et al. 2013), and this lack of adequately trained medical and paramedical staff contributes towards the
261
failure of the rabies control strategy adopted in India (Burki 2008). Knowledge, skills and motivation of health care providers are essential for effective prevention and control of diseases; however, India has been unable to meet set targets for endemic diseases
(Satpathy and Venkatesh 2006). In the case of rabies, it is important to understand the level of knowledge and preparedness of health workers, particularly those from rural areas, to deal with patients who suffer dog-bites.
In rural areas, there is often a failure to provide PEP to dog-bite victims in a timely manner due to unavailability at PHC or privately-owned pharmacies in these areas. Consequently
PEP or Rabies Immunoglobulins (RIG) must be acquired from adjacent urban centres, an option which is often not undertaken or is delayed by patients due to distance, time and/or cost, increasing the likelihood of progression to clinical rabies (Joseph et al. 2013). These cases could potentially be prevented if the PEP/RIG was either readily available at rural
PHC or if health workers at PHC where there was no PEP/RIG available could convince dog-bite victims of the importance of obtaining these products.
We conducted a KAP survey in PHC and sub-centres in the rural areas around the town of Baramati to assess: 1) the KAP of the paramedical staff towards rabies; 2) the availability of PEP/RIG in rural PHC; and 3) the awareness of the rural paramedical staff on the use of PEP/RIG.
262
8.2 Materials and Methods
8.2.1 Study area, sampling procedure and sample size
All 18 PHC and sub-centres located within a 20-kilometre radius of Baramati town of
Pune district in western India were included in this study. These cater to the primary health needs of approximately 360,000 rural residents (Economic Survey of Maharashtra,
2017-18, http://admin.indiaenvironmentportal.org.in). The centres were visited during
15th to 19th December 2016 and the paramedical staff present at the time of the visit included in the survey. A total of 54 staff members were interviewed from the 18 centres.
The respondents comprised of 31 nursing staff (GNM/ANM) and 23 non-nursing staff
(Laboratory technicians; Pharmacists; Multi-purpose workers (MPW); Ladies’ Health visitors (LHV); Health workers (HW); On Job Trained workers (OJT)).
8.2.2 Questionnaire design
The KAP survey was designed to determine the awareness of the paramedical staff regarding rabies and to assess practices adopted to prevent and control the disease. The questionnaire consisted of closed questions in two sections: Section one explored demographic details regarding information about the individual pertaining to their medical qualifications, years of experience and if the individual had undergone any training to manage animal bite injuries. The second section was comprised of questions to assess the knowledge, attitudes and practices of the individuals with respect to rabies.
The permission of the Taluka health officer, Panchayat Samiti Baramati was obtained to conduct this study. Ethics approval was obtained from the Murdoch University Human
Ethics Committee (permission number: 20/2016).
Prior to administering the questionnaire, the study was explained to the participants, the confidentiality of their answers confirmed and their consent to participate obtained. No
263
personal details of the participants, including their name, were recorded. The questionnaire was read out to the participant in their local language (Marathi) and the answers were recorded in English.
8.2.3 Data management and analysis
The answers to the questions were tabulated in a spreadsheet (Microsoft Excel, Microsoft
Corp., Redmond, WA, USA). Before submitting for statistical analysis, data were made compatible for analyses in the R programming environment (R Development Core Team
2013), which included converting all “Not sure” responses as incorrect and removal of
“NA” (not applicable) responses.
A participant’s “knowledge” about rabies and “attitudes and practices” about rabies were the dependent variables in the analyses, while the “years of experience”, “educational level” and the “professional appointment” held at the PHC were the explanatory variables for these analyses. Univariable analyses were initially performed using the Chi-squared test of independence or the Fisher’s exact test. Variables with a p value ≤ 0.25 were then offered to a multivariable logistic regression model and a final model generated using a backward stepwise process. Variables with a p < 0.05 were retained in the final model.
The final model was evaluated with the Hosmer-Lemeshow goodness-of-fit test (Matthew
2017). Odds ratios were calculated using the “odds ratio” package in R (Patrick 2017).
The answers to six questions were used to assess a participant’s knowledge about rabies and the answers to 10 used to assess their practices. For these questions, a correct answer was scored as 1 and an incorrect answer as 0. The results for these questions were summed and the median score for all participants calculated. People with a knowledge score higher than 4 (median score) were categorised as having good knowledge and those ≤ 4 as having poor knowledge. Similarly, people with a practices score > 6 (median) were categorised
264
as having positive practices. The division of respondents on the basis of median score on knowledge about rabies and on the practices regarding dog-bite wound management was based on recommendations of Likert type scale (Dane 2006).
8.3 Results
A total of 54 staff employees were interviewed from 14 PHCs and 4 sub-centres located with 20 km of Baramati (Table 8.1).
Table 8.1 List of Primary Health Centres and Sub-centres around Baramati and the number of staff interviewed along with their positions
Location ANM* GNM** PHARMACIST OTHERS# TOTAL PRIMARY HEALTH CENTRES Shirsuphal 1 1 2 Lasurne 1 1 1 3 Kalas 1 1 Sanasar 2 1 3 Dorlewadi 1 1 2 4 Hol 1 2 2 1 6 Murti 2 2 Parandare 1 1 1 3 Kedgaon 3 3 Sangavi 2 1 3 Moregoan 1 2 1 1 5 Katewadi 1 1 1 1 4 Loni Bhapker 2 1 3 Parwadi 2 1 1 4 SUB-CENTRES Gojubawi 1 1 Dalaj 2 2 Katphal 1 1 1 3 Pimpli 1 1 2 TOTAL 13 18 13 10 54
*ANM- Auxiliary Nursing Midwives, **General Nursing Midwives # others comprised MPW (Multi- purpose worker), Laboratory Technicians, Pharmacists, LHV (Ladies’ Health visitor), HW (Health worker) and OJT (On Job Trained) worker.
The number of years in service for the employees ranged from one to 34 years (average
14 years, median 11 years). Twelve staff members were educated up to university graduate level or higher. None of the participants had received any formal training for the
265
management of injuries sustained from animal bites. All respondents identified dogs as the most common source of animal-bite cases presented to the health centres. Most respondents (53, 98%) said that their centres were equipped to provide PEP treatments to dog-bite victims. Forty-nine (91%) health workers said that their centres were supplied with Anti Rabies Vaccines (ARV), 4 with ARV and RIG, while one respondent said their centre received none. The majority of the participants (51, 94%) reported that ARV was readily available from chemist stores close to their clinic. In contrast only 6 (11%) respondents said that RIG was available in nearby chemist stores.
The bivariate analyses of the responses to the questions on knowledge and the attitudes and practices are presented in Tables 8.2 and 8.3, respectively.
266
Table 8.2 Bivariate analyses of the individual questions pertaining to knowledge about rabies for health workers belonging to different categories (n=54)
Years in service Education Appointment p value Not p value Non- p value >11 years ≤ 11 years Graduate Graduate Nursing nursing Questions n=54 (%) (%) (%) (%) (%) (%) (%) Have you heard about rabies? Yes 52 (96) 26 (50) 26 (50) 0.5 12 (23) 40 (77) 0.9 30 (57) 22 (43) 0.9 No 2 (4) 0 2 (100) 0 2 (100) 1 (50) 1 (50) Do you think rabies can spread to another human from a human patient with rabies? Yes 19 (35) 4 (21) 15 (79) 0.004* 7 (37) 12 (63) 0.1 9 (47) 10 (53) 0.3 No 35 (65) 22 (63) 13 (37) 5 (14) 30 (86) 22 (63) 13 (37) Do you think rabies can be spread through licks/scratches from an animal? Yes 29 (54) 12 (41) 17 (59) 0.3 6 (21) 23 (79) 0.9 21 (72) 8 (28) 0.02* No 25 (46) 14 (56) 11(44) 6 (24) 19 (76) 10 (40) 15 (60) Do you think rabies can be spread through contaminated food or water? Yes 13 (24) 6 (46) 7 (54) 0.9 4 (9) 9 (91) 0.4 11 (85) 2 (15) 0.02* No 41 (76) 20 (49) 21(51) 8 (19) 33 (81) 20 (49) 21 (51) Do you think death is inevitable if a person bitten by a rabid animal develops signs of the disease? Yes 47 (87) 23 (49) 24 (51) 0.9 11 (48) 36 (52) 0.9 28 (59) 19 (41) 0.4 No 7 (13) 3 (43) 4 (57) 1 (14) 6 (86) 3 (43) 4 (57)
*Significant p values
267
The reasons given for the failure to control rabies included a lack of control of free roaming dogs (FRD) (24, 44%); lack of awareness about the disease by residents (19,
35%); a combination of both lack of awareness and lack of control of FRD (10, 19%); or the non-availability of PEP (1, 2%). Staff with ≤ 11 years of service were less informed about human-to-human transmission of rabies (OR 0.2, 95%CI 0.03-0.6, p=0.004) than more experienced staff. In contrast, the years of service did not affect responses to other knowledge questions (Table 8.2). Nursing staff were more aware that rabies could be transmitted by licks and scratches from a rabid animal (OR 3.8, 95%CI 1.0-14.4) and that it could not be transmitted through contaminated food and water (OR 5.6, 95%CI 1.0-
58.2, p = 0.02) (Table 8.2) than the non-nursing staff. Less experienced staff (≤ 11 years of medical service) were more likely to suture a dog-bite wound (OR 16, 95%CI 2.0-755) and graduates were less likely to know the schedule of administration of PEP (OR 0.2,
95%CI 0.02-1.0) (Table 8.3).
The results of the univariable analyses for the independent variables (years in service, education and appointment of the respondents) with the knowledge of rabies and the practices of the primary health staff to control rabies are summarised in Table 8.4. The predictors, years in service and appointment, yielded p values ≤ 0.25 (0.1 and 0.14 respectively) and were offered into the multivariable logistic regression model for assessing the knowledge of the respondents (Table 8.5). The final model was a good fit of the data (Hosmer-Lemeshow goodness of fit test χ2=0.71, df =1, p value=0.39).
268
Table 8.3 Bivariate analyses of the individual questions pertaining to practices of health workers towards management of dog-bite wounds belonging to different categories (n=54) that help control rabies
Variable n=54 (%) Years in service Education Appointment >11 years ≤11 years p Graduate Not p Nursing Non p (%) (%) value (%) Graduate (%) value (%) Nursing (%) value Do you think traditional treatment is useful? Yes 2 (4) 1 (50) 1 (50) 0.9 1 (50) 1 (50) 0.4 1 (50) 1 (50) 0.9 No 52 (96) 25 (48) 27 (52) 11 (21) 41 (79) 30 (57) 22 (43) Is washing the dog-bite wound with soap water useful? Yes 53 (98) 25 (47) 28 (53) 0.5 12 (23) 41 (77) 0.9 31 (58) 22 (42) 0.4 No 1 (2) 1 (100) 0 0 1 (100) 0 1 (100) How long do you wash the wound with soap-water? >10-15 min 9 (17) 7 (78) 2 (22) 0.1 2 (22) 7(78) 0.9 6 (67) 3 (33) 0.7 <10-15 min 45 (83) 19 (42) 26 (58) 10 (28) 35 (72) 25 (55) 20 (45) Would you suture a dog-bite wound? Yes 11 (20) 10 (91) 1 (9) 0.001* 0 11 (100) 0.1 8 (73) 3 (27) 0.3 No 43 (80) 16 (37) 27 (63) 12 (28) 31 (72) 23 (53) 20 (47) Do you think it is important to observe a dog that has bitten someone? Yes 53 (98) 26 (49) 27 (51) 0.9 12 (23) 41 (77) 0.9 31 (58) 22 (42) 0.4 No 1 (2) 0 1 (100) 0 1 (100) 0 For how many days should the dog that has bitten someone be observed for? ^ ≥10days 42 (78) 22 (52) 20 (48) 0.5 9 (26) 31 (74) 0.7 18 (41) 25 (59) 0.9 <10days 11 (22) 4 (36) 7 (67) 3 (27) 8 (73) 5 (45) 6 (55) Are you aware of the schedule of the ARV followed at your clinic? Yes 47 (87) 23 (49) 24 (51) 8 (17) 39 (83) 28 (59) 19 (41) 0.4 No 7 (13) 3 (43) 4 (57) 0.9 4 (57) 3 (43) 0.04* 3 (43) 4 (57) Should RIG be administered immediately after the dog-bite? Yes 14 (26) 7 (50) 7 (50) 3 (21) 11 (79) 11 (78) 3 (22) 0.1 No 40 (74) 19 (48) 21 (52) 0.9 9 (23) 31 (77) 0.9 20 (50) 20 (50) Can RIG be administered 7 days after the dog-bite exposure? Yes 41 (76) 17 (41) 24 (59) 10 (24) 31 (76) 21 (51) 20 (49) 0.1 No 13 (24) 9 (69) 4 (31) 0.1 2 (15) 11 (85) 0.7 10 (77) 3 (23)
*Significant p values, ^ – answers only included for those respondents who believed it was necessary to observe a dog that had bitten someone
269
Table 8.4 Test of association (χ2) of the independent variables (experience, education and appointment) with the dependent variables (knowledge about rabies and practices regarding management of dog-bite patients)
Criteria/variable n=54 (%) Knowledgeable respondents (%) OR p value Years in service ≤ median# 28 (52) 21 (75) 1 > median 26 (48) 14 (54) 0.4 (0.1-1.2) 0.1* Educational qualifications Graduate 12 (22) 9 (75) 1 Non-graduate 42 (78) 26 (62) 0.5 (0.1 -2.3) 0.5 Appointment Nursing 31 (57) 23 (74) 1 Non-nursing 23 (43) 12 (52) 0.4 (0.1-1.2) 0.15* Respondents with positive Criteria/variable n=54 (%) practices (%) OR p value Years in service ≤ median 28 (52) 16 (57) 1 > median 26 (48) 10 (38) 0.5 (0.2-1.4) 0.2* Educational qualifications Non-Graduates 42 (78) 23 (55) 1 Graduates 12 (22) 3 (25) 0.4 (0.07-1.2) 0.1* Appointment Nursing 31 (57) 17 (55) 1 Non-nursing 23 (43) 9 (39) 0.5 (0.2-1.6) 0.2*
*Variables offered to the multivariable models # Median years of service = 11 years
The practices towards dog-bite wound management were assessed by developing a multivariable logistic regression model that initially included all the three predictor variables namely, years in service (p= 0.2), appointment (p= 0.2) and educational qualifications (p=0.1). The final model indicates that more experienced staff and those with higher education adopt better practices (OR 3.2 and 5.6 respectively, Table 8.5). The final model was found to be a good fit of the data (Hosmer-Lemeshow goodness of fit test χ2=1.66, df =1, p value=0.19).
270
Table 8.5 Final multivariable model showing the influence of various independent factors over the knowledge about rabies and practices pertaining to rabies that help its control
Multivariable model (Knowledge) Intercept (b) SE p value OR Constant -1.8 0.61 _ _ Years in Service* ≤ median _ _ _ 1 > median 1.2 0.64 0.05 3.4 (1.0-12.1) Appointment Nursing _ _ _ 1 Others 1.2 0.64 0.05 3.5 (1.0-12.3) Likelihood ratio (χ²) test =6.7; p=0.03; Hosmer-Lemeshow goodness of fit test- χ2=0.71, df=1, p value=0.39 Multivariable model (Practices) Intercept (b) SE p value OR Constant 2.05 0.86 _ _ Years in Service* ≤ median _ _ _ 1 > median 1.2 0.62 0.06 3.2 (1.0-11.0) Educational qualifications Non-graduate _ _ _ 1 Graduate 1.7 0.8 0.03 5.6 (1.2-27.0) Likelihood ratio (χ²) test =7.3; p=0.02; Hosmer-Lemeshow goodness of fit test- χ2=1.66, df=1, p value=0.19
* Median years of service = 11 years
8.4 Discussion
Paramedical staff are usually the first point of interaction for a dog-bite victim in rural
India. However, few studies have been conducted in India focusing on the KAP of these staff at PHC and sub-centres. Although one would hope that the frontline health staff in rural areas are specifically trained on rabies and animal bite management, this study found that none of the health staff surveyed had received formal training on the management of dog-bite related injuries, hence, it is not surprising that awareness about some aspects of rabies-control and treatment of dog-bites was low.
Most respondents (96%) were aware of the disease, but this did not equate to knowledge on the best method of prevention or the correct protocol for administering PEP/RIG. In
271
contrast to a recent study conducted in the rural areas of Himachal Pradesh in northern
India, reporting irregular supply or total lack of ARV (Kumar, Thakur, and Mazta 2016), we found that ARV was available at all centres surveyed and RIG at three of the 18 centres. It was apparent that not all staff surveyed were aware of the availability of preventive measures against rabies in their clinic as one respondent reported that they had neither ARV nor RIG in their centre, even though the other respondent from the same centre reported receiving regular supplies. The improved availability of PEP in areas around Baramati may be attributed to effective implementation of the government’s initiatives after the 12th Five year plan and better practical application of these measures in the area of study compared to some other parts of India (Kalaivani, Raja, and Geetha
2014). Five year plans are regular economic policy roll-outs in India and rabies was included as a disease of economic consequence for the first time in 2012. Additionally, easier accessibility to big commercial cities viz. Pune compared to some other parts of
India that are located further from commercial centres, could also be a factor for availability of PEP in the area of this study. The high percentage (94%) of respondents confirming the availability of ARV in nearby private medical stores supports the increased awareness of PEP in this study area. However, it also indicates that there are a high number of dog-bite injuries in the area (Vishwanath et al. 2018).
In this study, 24% of staff believed that rabies could be transmitted through contaminated food or water or between humans (an extremely rare possibility (Crowcroft, 2015 #73)).
Although only approximately half (54%) of the participants were not aware that rabies could be transmitted through licks and scratches of a rabid animal, this was an improvement from a similar study conducted in north India, where 80% of nursing staff and 73% of non-nursing staff did not know other modes of transmission of rabies other than animal bites (Kishore, Singh, and Ravi 2015). Another study in Vietnam of public
272
health workers similarly reported poor awareness of the potential risk from licks/scratches of rabid animals (Nguyen et al. 2015). These findings highlight the need for a prescribed training module on the management of animal bite cases and rabies for frontline health staff.
Although most health staff were aware of the need to wash a bite wound with soap and water, a practice that can reduce human rabies cases by at least 65% (Burki 2008), a majority (83%) would wash the wound for less than ten minutes, while the recommended duration is more than 15 minutes (Dodet 2007, WHO 2018). Similarly, although almost all the health staff were aware that a dog that had bitten someone should be observed, only half of them correctly reported that the observation should be for 10 days (Rupprecht and Gibbons 2004). These features are important in the transmission of rabies and if information about them are correctly disseminated amongst the PHC staff, the incidence of rabies could be substantially reduced.
While all the staff interviewed said that they were aware of the treatment to be given to a dog-bite victim, only 87% could actually correctly state the schedule of vaccination to be followed. It has also been reported that even physicians practicing in rural India are not adequately informed about the importance of administering vaccinations and immunoglobulins in cases of any dog-bite injury, regardless of severity (Shankaraiah et al. 2013). However, more than half of the staff that were unaware of the ARV schedule
(57%) in the current study belonged to the non-nursing appointments. Not unexpectedly, given the low availability of RIG in the area, few health staff surveyed were aware of
RIG and its administration.
Generally, the more experienced workers and non-nursing staff were more knowledgeable about the practices (OR 3.4 and 3.5, respectively) required to control rabies, as indicated by the multivariable model (Table 8.5), although contrasting results
273
were obtained in the univariable analyses (Table 8.4). Similarly, the multivariable model for practices indicates that more experienced workers and better-educated staff adopt better practices towards dog-bite injury management (OR 3.2 and 5.6) in contrast to the findings of univariable analyses. This dichotomy may be peculiar to this study because half of the respondents who had less experience were comprised of non-nursing staff (who were better educated) thus having a better knowledge score. Although not significant at the 95% confidence level, the odds of an experienced worker being a nursing staff was higher (OR 1.8) and the odds of an experienced worker being better educated was also significant at the 90% confidence level (OR 3.5, p = 0.07). We surmise that such opposing results may be atypical of this data set and can be reduced in studies with larger sample sizes. However, knowledge gaps were identified even in experienced staff as they were more likely to favour the suturing of an animal bite wound (OR 16.1, p=0.001), contrary to the recommended practice of avoiding suturing animal-bite wounds or to suture only after instilling RIG (Wilde et al. 1996). Such inconsistencies could be overcome by mandatory training of health care staff at the commencement of their service, in conjunction with periodic training updates during their career.
In contrast, the knowledge about rabies of the non-nursing staff was found to be comparatively better (OR 3.5) than the nursing staff. This can be explained by the fact that most of the non-nursing staff were pharmacists and laboratory technicians and the majority of them (68%) were graduates with presumably more theoretical knowledge about rabies. The non-graduate nursing staff (Auxiliary Nursing Midwives and General
Nursing Midwives) who are trained to perform nursing procedures, had a lower level of knowledge on animal-bite injuries and rabies. The poorer practices adopted by non- graduates (predominantly nursing staff) could be improved by initial and regular training.
274
In summary, we found that staff in the early period of their service and those who are not graduates lacked comprehensive knowledge on rabies and reflect poor perception regarding practices that can reduce the incidence of dog-bite related rabies (Table 8.5).
As some non-nursing staff were found to treat dog-bite injuries in the absence of nursing staff, it is similarly important that these staff should also be included in training on rabies and dog-bite wound management.
The ubiquitous presence of unrestricted FRD was cited by 44% of the respondents for the presence of rabies in the area of study, followed by a lack of awareness about rabies in the general population (35%). This reiterates the WHO policy for control of rabies in developing countries where emphasis has been placed on managing the FRD population, along with mass vaccination of FRD (WHO 2013). The role of education in the control of zoonotic diseases has been emphasised by Robertson et al. (2000) and Slater (2001), and rabies is no exception. There is a need to develop educational campaigns for the general public as well, and these campaigns should include information on the need to seek immediate treatment for dog-bites and to not rely on traditional methods of treating dog-bite wounds. This is also an indicator that knowledge about rabies among health staff in rural PHC and sub-centres is not complete, as noted by the lack of a stipulated observation period of dogs. Education campaigns should also emphasise that there is no correlation between the period of observation of a dog involved in a biting incident, and administration of PEP, and PEP should be sought immediately after any animal bite.
It is likely that more remote centres would have different outcomes than that found in this study (Kishore, Singh, and Ravi 2015). It would be useful to expand the study to all staff in the centres and to sample further rural centres to ensure findings were consistent across a larger area of rural India when developing control measures and educational packages.
275
Our study highlights the need for programs to ensure that staff dealing with dog-bite victims have correct knowledge about rabies and know how to correctly treat such injuries to reduce the incidence of rabies in rural India.
Author contributions
All authors have contributed and approve the contents of this article. HT developed the study, collected and analysed the data. HT, IR, AV and MO wrote the article, provided critical revision and helped interpretation of contents and implications.
Funding
The work was funded by the Wellcome Trust-DBT India Alliance Program through a
Fellowship to AV (Grant number: IA/CPHI/15/1/502028) and the Research grant to HT from Murdoch University, Western Australia, Australia.
Acknowledgements
Murdoch University International Postgraduate Scholarship (MIPS) to HT is gratefully acknowledged. The authors are grateful to Pranav Panwalkar and Vinayak Shitole for helping with the data collection and the respondents of PHCs for their participation.
Conflict of Interest Statement
The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
276
Chapter Nine
General Discussion
“The serpent, the king, the tiger, the stinging wasp, the small child, the dog owned by other people, and the fool: these seven ought not to be awakened from sleep”
Chanakya
277
Prologue
India has the highest recorded human mortality due to dog - related rabies in the world
(Baxter 2012), and the economic losses from the disease are enormous (Hampson et al.
2015). Educational outreach, post-exposure prophylaxis, dog population management and the need for diagnostic laboratories have been recognised by the Government of India as essential tools for rabies control in humans and formed a component of the countries
11th five-year-plan (2007- 2012) (Kalaivani, Raja, and Geetha 2014). For human fatalities due to rabies to be effectively controlled, the disease needs to be eradicated from dogs as these are the most common host reservoir for the rabies virus in both rural and urban settings (Beran 2017, Lembo et al. 2008, Rupprecht, Hanlon, and Hemachudha 2002).
However, rabies control programmes in India have primarily focused on humans, owing to a lack of studies investigating the difficulties in achieving a suitable level of immunity in dogs.
Control of rabies requires a One Health approach and a strategy which integrates participation from physicians, veterinarians, public health workers, animal welfare workers, educational institutes and social scientists is essential to combat the disease
(Anderson and Shwiff 2015, Buchy et al. 2017, Lavan et al. 2017). Although bites from dogs are the commonest cause of human rabies in India, the veterinary component of rabies control is only partially implemented or not at all. A lack of understanding of the disease and varied attitudes of people towards rabies and FRD further complicates the situation in India. Finally, improper dog-bite wound management practices by para- medical staff at PHC is partially responsible for the high rate of human mortality due to rabies in the country.
278
This thesis addresses the difficulties of estimating the population size of FRD, their home ranges, and their group forming behaviour that impede achieving adequate vaccination coverage against rabies in this population, as well as exploring the knowledge gaps towards rabies and the attitudes of the rural and urban communities towards rabies and
FRD. Finally, it evaluates the level of preparedness of the rural para-medical staff to efficiently manage dog-bite wounds in human patients.
9.1 Estimation of the population size of free roaming dogs
The lack of information regarding the population size of FRD is a key factor in the failure of mass vaccination campaigns against rabies (Wallace, Etheart, et al. 2017). A number of methods, including the probabilistic models that are commonly used in wildlife studies, have been used to estimate the FRD population but the absence of a gold standard method to reliably estimate the number of dogs poses a substantial challenge for agencies involved in the mass vaccination of FRD against rabies (Belo et al. 2015).
On comparing different enumeration techniques for FRD it was found that probabilistic models that yield maximum likelihood estimates (MLE) were nearest to the likely true population but required a minimum of five surveys (Chapters Two and Three), hence are expensive and time consuming to implement in most situations. A freely accessible online calculator tool called the Application SuperDuplicates shinyapp
(https://chao.shinyapps.io/SuperDuplicates/) developed by Chao et al. (2017), based on the Good-Turing formula, obtained an estimate that may not be accurate of the true population size, but the projected number always exceeded 70% of the MLE estimates arising from five or more surveys (Chapters Two and Three). The Application
SuperDuplicates could be used as a surrogate method, not only to obtain a reliable estimate of the FRD population size in the targeted area of intervention, but the data
279
generated through the two surveys can also provide information on the composition of the FRD population (Chapter Four). The demographic characteristics of the FRD, such as male to female ratios, proportion of adults to puppies, and the proportion of good/poor body conditioned dogs didn’t vary across surveys in this study. Using the AS method may encourage intervention implementing agencies to carry out frequent enumerations to understand the FRD population dynamics and to evaluate the effectiveness of dog population management interventions, such as ABC.
The influence of survey time on the FRD counts was highlighted in the urban survey where the morning counts exceeded those conducted in the afternoons (Chapter Three).
In future the FRD sight-resight data for estimating the population-size should be generated by using data from morning surveys in urban settings to counter the large SE arising from temporal variation in the dog-counts when data from both morning and afternoon counts are combined. However, in contrast to the urban setting, in the rural setting no temporal variation was witnessed and hence surveys conducted in rural locations to estimate the FRD population using the AS online tool could be undertaken either in the morning or afternoon or both (Chapter Two).
The temporal variation in the FRD counts in the urban settings is potentially attributable to the increased vehicular traffic in the afternoon surveys (Chapter Three). Although temporal variation was not significant in the rural study, heavy rains preceding the survey sessions resulted in a diminished FRD count. The influence of other climatic factors on the day of the survey, such as temperature, wind velocity and humidity, did not significantly influence the count, but this could be due to minimal variation in these meteorological parameters during the survey period. Nonetheless, the rural survey witnessed higher counts on overcast days, which is similar to findings in Sao Palo, Brazil
280
(Dias et al. 2013) and the city of Newark, New Jersey, USA (Daniels 1983). However, effect of such climatic influences over FRD re-sight counts cannot be ruled out summarily and modelling influences of such factors over the FRD count could form the basis of future investigations. Furthermore, societal events, such as a community feast in a neighbouring village which possibly attracted FRD through the presence of edible leftovers (as communicated by the village chief), may have led to the reduction in the
FRD count in Shirsuphal (Chapter Two). Others have similarly reported sighting an abundance of FRD in the proximity of “relatively undisturbed food source(s)” (Boitani et al. 1995). As such events influence the sightings of FRD in the targeted areas, it is recommended that climatic conditions and local events that may generate temporary accumulated food resources should be taken into consideration prior to conducting enumeration surveys.
9.2 Free roaming dog demography, group behaviour and home-ranges
Information on the population size of FRD alone is not sufficient to ensure effective implementation of adopted interventions, be they mass vaccination campaigns against rabies or ABC programmes designed to manage the population size. An understanding of the demographic characteristics, roaming behaviour, social organisation and home ranges of the reservoir host is also paramount for effective intervention (Slater 2001, Hudson et al. 2019). The population dynamics of FRD is influenced by the carrying capacity of the habitat, the level of human interaction and the life span of FRD (Conan et al. 2015). A number of studies have emphasised the role of the sex-ratios, age composition and neuter status of FRD and their relationship with communities on the successful implementation of mass vaccination programs (Slater 2001, Robertson, Wilks, and Williamson 1993,
Morters, McKinley, Restif, et al. 2014, Gsell et al. 2012). In the Indian context,
281
inaccessibility of the FRD for parenteral vaccination is often cited as a major impediment for failure to reach herd-immunity levels (minimum 70% protection) (Gibson et al. 2019,
Jackman and Rowan 2007). The inaccessibility of FRD may result from reduced sightings due to the influence of extrinsic factors, such as climatic and societal ones (Chapter One,
Section 9.1) (Ciucci and Boitani 1998). Additionally, the FRD are reportedly sighted more often in groups during the mating season (Boitani, Ciucci, and Ortolani 2007); and in locations where ABC programmes have been conducted, entire dogs have a higher tendency to form groups compared to de-sexed dogs (Pal, Ghosh, and Roy 1998b). Free roaming dogs that are in a group are harder to catch than individual animals with a concurrent heightened occupational risk for the catchers (Chapter Five) (Jibat, Hogeveen, and Mourits 2015).
The analyses of the demographic data (gender, age, body condition, proximity to garbage and level of activity) of FRD during this project demonstrated significant differences between the rural and urban settings (Chapter Four). The findings are consistent with the majority of other demographic studies on FRD conducted elsewhere that highlight a higher percentage of males than females in rural settings (Oppenheimer and Oppenheimer
1975, Berman and Dunbar 1983, Dias et al. 2013, Daniels and Bekoff 1989). In contrast, in the urban location the gender ratio was close to parity, suggesting higher human influence over the FRD in the rural setting examined in the current study. This may have resulted from a preference for male dogs to females in rural locations as the former were potentially more useful as guard dogs (Chapter Six), as opposed to the reported killing of females to avoid unwanted litters in neighbouring Nepal (Massei et al. 2017). The indirect influence of humans over FRD in the rural setting was also demonstrated through demographic findings such as: a lower level of activity of the dogs (the FRD were more likely to be sighted resting than foraging, walking, running or playing with other dogs); a
282
better ratio of good to poor/fair condition dogs, and fewer groups of dogs within 20m of garbage in the rural FRD population than in the urban setting (Chapter Four). An unexpected finding reported in Chapter Five was the tendency of FRD to form groups in urban settings, especially around garbage points, and to have larger home ranges, possibly due to the requirement to travel further to find food, than for FRD in the rural setting. The rural FRD presumably could find food in a comparatively smaller area due to indiscriminate scatterings of edible litter. In contrast, in the urban settings the FRD would be required to move to neighbouring lanes to scavenge from the garbage bins located at fixed locations (Chapter Five). Broadly, the findings of the study presented in Chapter
Five underline the need to assess the demographic characteristics and grouping tendencies of FRD prior to implementing any vaccination programme. The accessibility of FRD in urban areas for parenteral vaccination is likely to be comparatively challenging and hence it is recommended that mass vaccination of FRD in urban settings would require a larger proportion of the FRD population to receive ORV compared to the rural settings (Chapter
Five). Conversely, door-to-door parenteral vaccinations may be comparatively successful in rural areas where FRD are less wary of human interaction.
A highlight of the findings of this study relates to the management of garbage and the dog population. In urban Panchkula, efficient management of garbage points, such as provision of closed garbage bins, their frequent emptying/recycling, and prevention of the accumulation of litter around garbage points, especially in the administrative and industrial sectors, would possibly reduce the carrying capacity of FRD and potentially diminish the tendency of FRD to congregate around garbage points (Taylor et al. 2017).
This may be implemented in conjunction with enhanced public outreach using media, involving focus groups of targeted communities and the educational institutions, and by augmenting the Municipal public services through efficient garbage management. The
283
reduction in the carrying capacity would, in turn, result in a larger dependence of the dogs on humans for their subsistence, and potentially their accessibility for parenteral vaccination could be enhanced (Chapters Four and Five). It is also recommended that dog pounds/shelters for FRD be established by the Municipal Corporations for the dogs currently surviving on scavenging, thereby reducing the number of dogs feeding at the garbage points. A reduction in competition for food will possibly mitigate the enhanced aggressive behaviour of FRD to other dogs and humans (Scott, Bronson, and Trattner
1968).
9.3 Knowledge, attitudes and practices (KAP) of the people towards rabies and FRD
The first half of this thesis was devoted towards mitigating the complications of FRD enumeration, and their accessibility for parenteral and/or oral vaccinations. The overarching objectives of this thesis, however, involve an assessment of the KAP of rural and urban communities towards rabies and the FRD. Mere knowledge of the population size, demographics, home ranges and group forming behaviour of FRD in the targeted area cannot alone achieve effective mass vaccination against rabies unless the communities support such a campaign. Rabies is primarily a disease associated with the poor socio-economic sections of society (Sudarshan et al. 2007, Baxter 2012). A large population of this group reside in Indian villages that do not have access to PEP, such as
ARV or RIG, either due to its unavailability in remote locations or their expense (Joseph et al. 2013). The rural population is also reportedly more vulnerable to the disease due to a general lack of awareness about the disease (WHO 2018).
Unfortunately, most KAP studies previously undertaken on rabies in India have focused on hospital based retrospective surveys which obviously exclude communities that do not
284
have access to hospitals, or those who do not think reporting dog-bites to hospitals is important. A general lack of awareness about the disease, few diagnostic facilities, and poorly coordinated efforts on the part of physicians, veterinarians and social workers towards sustained rabies control campaigns makes rabies one of the most underreported and under-controlled diseases in India (Chapter One). The lack of community based studies on rabies in India, especially in the veterinary sector, has been highlighted by
Kakkar et al. (2012).
Since 2012 the Government of India has initiated measures to: improve the availability of PEP at no cost to the general public at PHC; and enhance support of awareness campaigns conducted in rural areas (Kalaivani, Raja, and Geetha 2014). Although the results of the current KAP study in rural Shirsuphal and urban Panchkula corroborate increased awareness about the disease, gaps were identified in the knowledge of rabies by the participants, including that rabies could be spread by licks from a rabid animal
(Chapters Six and Seven). The published literature on dog mediated rabies does not differentiate between the infective potential of dog licks from dog-bites (Jemberu et al.
2013, Herbert, Basha, and Thangaraj 2012, Sudarshan et al. 2007, Meslin 2005), but ignoring licks as a potential route of infection by the victims may prove fatal. Knowledge about simple practices to control rabies, such as washing the bite wound with soap and water, was generally lacking in respondents from the lower socio-economic sections of urban Panchkula. This indicates that awareness campaigns, even when conducted in cities, possibly do not reach the vulnerable residents or are ineffective in imparting specific relevant details. In Chapter Seven the results showed that urban households with children ≤ 14 years of age were less knowledgeable of the disease compared to households with either no children or with older children. This is of concern as others have reported that the ≤ 14 years of age, age group were a high risk group for dog-bites
285
(Ichhpujani et al. 2008, Chatterjee 2009). Such knowledge gaps could be overcome by specifically including information about rabies in the primary/elementary school teaching curriculum and impressing upon children the need to avoid contact with FRD, and to promptly report any bites or licks to family elders or teachers who then hopefully could undertake suitable actions to reduce the risk of the child acquiring rabies. The positive impacts of integrating educating about rabies in the elementary school curriculum has been demonstrated in the Indian state of Sikkim, as well as other places in Asia including the Philippines and Sri Lanka (Auplish et al. 2017, Amparo et al. 2019, Kanda et al. 2014).
The attitudes and practices of the urban and rural population towards FRD were investigated, in addition to assessment of their KAP towards rabies. A comparatively higher percentage of rural respondents (86%) considered FRD were a nuisance compared to the urban population (68%) and this sentiment is reflected in their practices, as only half of the rural respondents provided food to FRD compared to 72% of the urban respondents. This perception is possibly a result of the campaigns by animal welfare organisations that advocate compassion for the FRD, promote their feeding and insist that dogs should not be removed from the residential localities in urban locations (Majumder,
Chatterjee, and Bhadra 2014).
9.4 Responsible ownership of dogs
India is reported to have the lowest proportion of dog owning households in the world
(Davlin and VonVille 2012). However, ‘dog ownership’ can have a wide meaning that varies from mere provision of food or shelter to discharging responsibilities beyond provision of food and/or shelter, such as its veterinary care, vaccination against infections and restricting the area an animal is free to move in. The reason of a higher claim of dog ownership in rural Shirsuphal (53%) than in urban Panchkula (36%) could be due to the
286
perception of rural respondents of ownership of dogs without the associated responsibilities listed above. Fewer respondents who owned dogs in urban Panchkula adopted a FRD (29%) compared to rural Shirsuphal (69%), where most owned dogs were free roaming (Chapters Six and Seven). However, fewer rural owners (12%) had their pets vaccinated against rabies (Chapter Six) compared to urban Panchkula (71%)
(Chapter Seven). An indirect finding of the analyses was that the owners who paid to procure their pets further invested in the care of their pets, while those who adopted dogs off the streets refrained from spending money on their pets’ health and felt that the responsibility to vaccinate dogs lay with the government (Chapter Six). To maximise the vaccination coverage of dogs, free canine anti-rabies vaccines could be made available from the rural government veterinary hospitals, in conjunction with the regular conduct of mass vaccination programmes (Lavan et al. 2017).
A strategy to control dog-related rabies and dog population management is to examine ways to increase the adoptability of FRD, and also develop and implement regulations for dog owners to have their pets vaccinated against rabies. Others have suggested the need to promote responsible ownership of dogs (Wandeler et al. 1988), however, this will have little uptake because FRDs (or the street dogs) are considered of no or little value by both the rural and urban populations. There is little research on the trainability of the
FRD in India, although there is evidence to suggest that such measures could enhance their adoptability (Demirbas et al. 2017). It is recommended that the utility of the FRD for the society, such as guard dogs, sniffer dogs or companion dogs, be explored so that people are encouraged to develop responsible ownership of dogs adopted off the streets.
Responsible ownership of FRD could also be promoted by their free registration and by organising dog-shows exclusively for non-pedigree dogs, thus encouraging people to value these types of dogs similar to the pedigree dogs.
287
9.5 Paramedical staff and rabies control
Although, the PHC surveyed in this study were found to have adequate supply of ARV,
RIG was not available in the majority of rural PHC included in the study (Chapter Eight).
Furthermore, the paramedical staff, especially the less experienced ones and those with lower qualifications (non-graduates), lacked knowledge about rabies and were not acquainted with simple practices, such as wound washing, that would reduce the disease’s incidence. This is reflective of the similar findings of the KAP of rural residents regarding rabies (Chapter Six) and one of the possible reasons for the low knowledge level of paramedical staff could be that they are primarily drawn from the local population. Most rabies awareness efforts are focussed in urban areas, with only few awareness programmes conducted in rural areas. Although it did not form a part of the questionnaire administered to PHC staff, it was anecdotally learnt from them and the villagers that no rabies awareness campaign was ever held in any of the 18 villages where the surveys were conducted. Organising frequent rabies awareness programmes and use of Information, education & communication (IEC) materials, such as pamphlets and posters, could be used to spread awareness regarding simple procedures, such as dog-bite wound washing, to reduce rabies infection following dog-bites (Sudarshan et al. 2013). It is recommended that implementation of routine training programmes for paramedical staff on rabies and dog-bite wound management could be an important measure to enhance knowledge about the management of dog-bite wounds, thereby reducing human mortality due to dog related rabies, as has been demonstrated with Resident doctors and casualty medical officers of
G.R. Medical College and Hospital, Gwalior, India (Agarwal, Pathak, and Mahore 2018).
288
9.6 Control of dog-mediated rabies in India
The reservoir species of rabies virus in the low and middle income group countries of the world are domestic dogs (Cleaveland and Hampson 2017). In countries where rabies elimination has been achieved, interventions have focused on the main reservoir hosts.
For example the strategy of vaccinating wildlife (red foxes) through the use of ORV during the 1980s and 90s resulted in the elimination of the disease from Eastern Europe
(King et al. 2004, Freuling et al. 2013). Similarly canine rabies in North America was controlled through vaccination focused on the terrestrial wildlife carnivore reservoirs including raccoons (Procyon lotor), wild foxes (Vulpes vulpes) and skunks (Mephitis mephitis) (Rosatte 2013). Although persistence of the rabies virus in wildlife reservoirs, such as bats or wild canids, is a threat in all places which are not free of rabies, there are few human mortalities due to exposure to sylvatic or bat rabies (Léchenne et al. 2019). In contrast, the wildlife population in rabies endemic countries is threatened by the incursion of rabies virus from domestic dogs in the fringe areas of human habitation (Viana et al.
2015, Vanak and Gompper 2009b, Young et al. 2011, Gongal and Wright 2011). In contrast to the success stories of Europe and North America, rabies persists in the countries of Asia and Africa due to a deficiency of data on the demography of FRD and limited research interest to derive potential solutions to eliminate the virus from the reservoir host (Molyneux et al. 2011). Nonetheless, recently a number of researchers have produced a body of research that demonstrates the feasibility of eradication of canine rabies from such countries, including India (Cleaveland, Beyer, et al. 2014).
Rabies in India has been viewed as a manifestation of the overwhelming dog population and consequently control strategies applied over the last two decades have advocated
ABC or CNVR programmes (Reece and Chawla 2006) as these appeared plausible compared to the enormity of mass vaccination coverage (70%) of dogs (Fitzpatrick et al.
289
2016). Recent models, based on the low basic reproductive number (R0 – average number of dogs that can be infected by a rabid dog) of the disease, however indicate that the disease could be effectively controlled by mass vaccination rather than by reducing dog densities (Cleaveland and Hampson 2017). Contesting the widely accepted 70% vaccination coverage required to control the disease, Fitzpatrick et al. (2016) advocated that 7 - 35% coverage would be sufficient. This lower coverage was based on the low R0 value (1.41, 95%CI 1.39-1.45) for rabies transmission. They claimed that vaccinating
25% of the population annually could effectively eradicate the disease. However, the study was based on data obtained from only one state in India and may not be representative for other states which may potentially differ in peoples’ attitudes towards the disease as well as their perception of FRD, as demonstrated in this study (Chapters
Six and Seven). Interestingly, instead of altogether rejecting the sterilisation strategy
Fitzpatrick et al. (2016) recommended a combination of sterilisation and mass vaccination in countries where stray dogs constitute more than 70% of the total dog population. This is a more realistic approach for countries, such as India, where besides the rabies threat, uncontrolled FRD population is a rampant cause of road accidents, damage to property, environmental pollution and public nuisance (Chapters One, Six and Seven) (Kakrani et al. 2013, Kale et al. 2006). However, Fitzpatrick et al. (2016) models were based upon the assumption from the Government livestock census that 58% of the canine population was owned and 42% were stray (FRD). These values are, however, debatable, as discussed in Chapter One (Section 1.4.2). It was demonstrated in Chapters Six and Seven of this thesis that the attitudes of the dog owning population towards their pets differs between rural and urban settings. Fitzpatrick et al. (2016) found that vaccination at a central point was successful in achieving 25% vaccination coverage in the state of Tamil
Nadu, however such an attempt in rural Shirsuphal failed to attract less than 10% of the
290
estimated FRD population. It was observed that although FRD were friendly towards the residents who fed them, they resisted being leashed by them to be brought to a central vaccination point. Alternatively, it is recommended that adopting a door to door vaccination programme, where households that provide food and shelter to FRD could restrain them for parenteral vaccination would yield better results in rural Shirsuphal
(Chapter Five).
Given the enormity of the FRD population size and the perception of the local communities towards them in India, a combination of all the intervention strategies, namely: mass vaccination and ABC for dogs; educational outreach programmes for children; and, awareness campaigns for communities under the umbrella of One Health would be the best approach to control and eventually eradicate rabies from India. Such an approach, where interventions applied to the animals will help reduce the disease incidence in humans, should also include provision and accessibility of PEP, along with capacity building of medical and nursing staff involved in the provision of primary health care in rural and urban areas (Lavan et al. 2017).
It is pertinent to mention the two states in India that are leading the way towards control of rabies through adopting a One Health approach. The north eastern state of Sikkim commenced the Sikkim Anti-Rabies and Animal Health (SARAH) programme in 2006 that comprised canine rabies vaccination, dog population management and rabies prevention education. This programme has resulted in the elimination of dog-mediated rabies in humans from the state (Byrnes, Britton, and Bhutia 2017). Although two animal deaths due to rabies were reported in 2015, nine years after the programme was launched in Sikkim, the state is claimed to be “on the road” to declare total freedom from the disease
(Byrnes, Britton, and Bhutia 2017). Moreover, only 18% of the dogs in the state were
291
claimed to be free roaming which completely contrasts the findings from the current study in Shirsuphal or Panchkula where FRD outnumbered owned dogs by a large margin
(Chapters Six and Seven). One important aspect that is lacking or at least underdeveloped in the SARAH programme in Sikkim is a robust surveillance mechanism to accompany the sterilisation and mass vaccination of FRD. Such a mechanism would ensure that the
FRD population in the targeted area achieves adequate herd immunity and instills confidence that the intervention efforts are restricting the transmission cycle of the rabies virus. Practically, it requires that every mass-vaccination campaign is preceded by FRD enumeration to enable subsequent evaluation of the efficacy of the intervention (Chapter
Two).
In the western state of Goa, approximately 100,000 dogs (owned and FRD) are claimed to be vaccinated annually by Mission Rabies; an ongoing education programme on the disease for school children has been running since 2015; and infrastructure and expertise skills for surveillance of canine-rabies has been expanded (Rupprecht et al. 2019).
However, in a presentation outlining the challenges to rabies control programme in Goa during the 21st National conference of Association of control of rabies in India (APCRI), it was pointed out that the enormous resources required for parenteral rabies vaccination of FRD is a major hindrance for such measures to be adopted in other states. It was recommended a combination of parenteral and ORV may potentially make such campaigns feasible in other states (Gibson 2019), a strategy which was outlined in
Chapter Five of this thesis.
Another aspect for the control of rabies in India that is currently not part of the strategies being applied in Sikkim or Goa is the lack of efforts to evaluate disease freedom in the regions where eradication efforts are being undertaken. A methodical approach to
292
evaluate the level of protective immunity in the reservoir host and simultaneous proof of disease freedom by diagnosing rabies deaths in canines is required as part of disease surveillance. Disease freedom is recommended to be supported by diagnosis of every
FRD death (a targeted surveillance may be misleading as many FRD deaths may not be supported by a reliable history of exhibiting signs of rabies prior to death) to rule out rabies using inexpensive user friendly Direct Rapid Immunohistochemical tests (dRIT) developed by the Centre for Disease Control (CDC), Atlanta, USA (Chapter 1, Section
1.2.6 (e)).
9.7 Conclusions
The key findings from this study are summarised as follows:
(a) The probabilistic models that incorporate individual heterogeneity, such
as Huggin’s closed capture models, obtain the estimates closest to the real FRD
population size.
(b) Application SuperDuplicates online shinyapp provides a reliable estimate
of 70% FRD population size with minimal resource application.
(c) A knowledge of the FRD demography, sighting patterns, grouping
behaviour and home-ranges of FRD can help formulate a vaccination strategy to
achieve 70% coverage.
(d) There are knowledge gaps in the awareness regarding rabies in rural and
urban communities in India along with inadequate dog-bite wound management
practices.
293
(e) There is a need to develop a training module on rabies, PEP/RIG
administration, and dog-bite wound management practices for the para-medical
staff in rural PHC and sub-centres.
Epilogue
India, as a nation, is losing the battle against rabies due to a lack of nation-wide concerted efforts to devise and implement a comprehensive intervention strategy that follows the
One Health approach. Such an approach should include: mass vaccination of FRD; management of the FRD population; garbage management; sustained educational outreach to children; increased scope and frequency of awareness campaigns; incentives to encourage people to adopt FRD and develop responsible ownership behaviours; and increasing the availability of PEP, such as ARV and RIG, to all members of the community. A pilot project that incorporates the One Health approach is recommended to be applied in rural and urban India that can obtain empirical evidence through constant intervention and surveillance to demonstrate to policy makers and the community that eradication of rabies is achievable. Such an eradication programme would require: mass vaccination of FRD based on reliable population estimates; regular awareness campaigns that includes targeting the lower socio-economic sections of society, as well as the front- line medical staff; sustained ABC intervention; reporting and testing of FRD deaths from rabies through the use of the low-cost diagnostic kits (dRIT); incentivise adoption of
FRD; and responsible ownership of dogs. The path has been shown by Sikkim and a start has been made in Goa, but such efforts need to be mirrored in other states, including lessons learnt from these campaigns, and be applied in rural and urban areas alike. Rabies eradication may be a difficult and long drawn battle, but it is neither an impossible one, nor one that India can afford to lose.
294
References
Abbas, SS, and M Kakkar. 2013. "Research & policy disconnect: The case of rabies research in India." The Indian Journal of Medical Research 138 (4):131-133.
Abbas, SS, and M Kakkar. 2015. "Rabies control in India: a need to close the gap between research and policy." Bulletin of World Health Organisation 93:131–132. doi: 10.2471/BLT.14.140723.
Abbas, SS, M Kakkar, and ET Rogawski. 2014. "Costs analysis of a population level rabies control programme in Tamil Nadu, India." PLoS Neglected Tropical Diseases 8 (2):1-7. doi: 10.1371/journal.pntd.0002721.
Abdela, N, B Midekso, J Jabir, and W Abdela. 2017. "Knowledge, attitudes and practices towards rabies in Dedo district of Jimma zone, southwestern Ethiopia: A community based cross-sectional study." International Journal of Medicine and Medical Sciences 9 (5):61-71. doi: 10.5897/IJMMS2017.1302.
Acharya, AS, R Kaur, and K Lakra. 2012. "Rabies epidemiology and control in India: A review." Journal of Communicable Diseases 44 (2):59-69.
Acosta-Jamett, G, S Cleaveland, and AA Cunningham. 2010. "Demography of domestic dogs in rural and urban areas of the Coquimbo region of Chile and implications for disease transmission." Preventive Veterinary Medicine 94 (3-4):272-281. doi: 10.1016/j.prevetmed.2010.01.002.
Agarwal, AK, A Pathak, and RN Mahore. 2018. "Study of knowledge status and effectiveness of training about rabies prophylaxis among resident's doctors." British Journal Pharmacological Medical Research 03 (03):1015-1023. doi: 10.24942/bjpmr.2018.254.
Agarwal, N, and VP Reddajah. 2004. "Epidemiology of dog bites: a community-based study in India." Tropical Doctor 34 (2):76-78. doi: 10.1177/004947550403400205.
Aiyedun, JO, and BO Olugasa. 2012. "Use of aerial photograph to enhance dog population census in Ilorin, Nigeria." Sokoto Journal of Veterinary Sciences 10 (1):22- 27.
Alho, JM. 1990. "Logistic regression in capture-recapture models." Biometrics 46 (3):623-635.
Amaku, M, RA Dias, and F Ferreira. 2010. "Dynamics and control of stray dog populations." Mathematical Population Studies 17 (2):69-78.
295
Amaral, AC, MP Ward, and J Freitas da Costa. 2014. "Estimation of roaming dog populations in Timor Leste." Preventive Veterinary Medicine 113 (4):608-613.
Amparo, ACB, ECB Mendoza, DA Licuan, LM Valenzuela, JD Madalipay, SI Jayme, and LH Taylor. 2019. "Impact of integrating rabies education into the curriculum of public elementary schools in Ilocos Norte, Philippines on rabies knowledge, and animal bite incidence." Frontiers in Public Health 7:119. doi: 10.3389/fpubh.2019.00119.
Amstrup, SC, TL McDonald, and BFJ Manly. 2010. Handbook of capture-recapture analysis. Princeton and Oxford Princeton University Press.
Anderson, A, and SA Shwiff. 2015. "The cost of canine rabies on four continents." Transboundary and Emerging Diseases 62 (4):446-452.
Anderson, RM, HC Jackson, and AM Smith. 1981. "Population dynamics of fox rabies in Europe." Nature 289:765-771.
Asher, L, EL Buckland, C Phylactopoulos, MC Whiting, SM Abeyesinghe, and CM Wathes. 2011. "Estimation of the number and demographics of companion dogs in the UK." BMC Veterinary Research 7 (1):74-86. doi: 10.1186/1746-6148-7-74.
Auplish, A, AS Clarke, T Van Zanten, K Abel, C Tham, TN Bhutia, CR Wilks, MA Stevenson, and SM Firestone. 2017. "Estimating the intra-cluster correlation coefficient for evaluating an educational intervention program to improve rabies awareness and dog bite prevention among children in Sikkim, India: A pilot study." Acta tropica 169:62-68.
Bala, A. 2014. "Trends and levels of urbanization in Haryana." An International Journal: Geography XII (12) (February-2014):115-117.
Banerjee, A, and S Chaudhury. 2010. "Statistics without tears: Populations and samples." Industrial Psychiatry Journal 19 (1):60-65. doi: 10.4103/0972-6748.77642.
Banyard, AC, DL Horton, C Freuling, T Müller, and AR Fooks. 2013. "Control and prevention of canine rabies: The need for building laboratory-based surveillance capacity." Antiviral Research 98 (3):357-364. doi: 10.1016/j.antiviral.2013.04.004.
Baxter, JM. 2012. "One in a million, or one in thousand: What is the morbidity of rabies in India?" Journal of Global Health 2 (1):010303 doi: 10.7189/jogh.02.010303.
Beck, AM. 1973. The Ecology of Stray Dogs: A Study of Free-ranging Urban Animals. West Lafayette, Indiana: Purdue University Press.
Beck, AM. 1975. "The public health implications of urban dogs." American Journal of Public Health 65 (12):1315-1318.
296
Belo, VS, GL Werneck, ES da Silva, DS Barbosa, and CJ Struchiner. 2015. "Population estimation methods for free-ranging dogs: a systematic review." PloS one 10 (12):e0144830.
Belsare, AV, and ME Gompper. 2013. "Assessing demographic and epidemiologic parameters of rural dog populations in India during mass vaccination campaigns." Preventive Veterinary Medicine 111 (1):139-146.
Belsare, AV, and ME Gompper. 2015. "A model-based approach for investigation and mitigation of disease spillover risks to wildlife: Dogs, foxes and canine distemper in central India." Ecological Modelling 296:102-112. doi: 10.1016/j.ecolmodel.2014.10.031.
Beran, GW. 2017. "Urban rabies." In The Natural History of Rabies, edited by George M Baer, 447-464. USA: CRC Press, Florida.
Beran, GW, and M Frith. 1988. "Domestic animal rabies control: an overview." Review of Infectious Diseases 10 (Supplement 4):S672-S677.
Berman, M, and I Dunbar. 1983. "The social behaviour of free-ranging suburban dogs." Applied Animal Ethology 10 (1-2):5-17.
Bharadva, N, SR Mehta, P Yerpude, K Jogdand, and KN Trivedi. 2015. "Epidemiology of Animal bite cases attending tertiary health care centre of Bhuj City of India: A cross- sectional study." International Journal of Interdisciplinary and Multidisciplinary Studies 99:98-102.
Bögel, K, and DD Joshi. 1990. "Accessibility of dog populations for rabies control in Kathmandu valley, Nepal." Bulletin of the World Health Organization 68 (5):611.
Boitani, L, and P Ciucci. 1995. "Comparative social ecology of feral dogs and wolves." Ethology Ecology & Evolution 7 (1):49-72. doi: 10.1080/08927014.1995.9522969.
Boitani, L, P Ciucci, and A Ortolani. 2007. "Behaviour and social ecology of free-ranging dogs." In The behavioural biology of dogs, edited by Per Jensen, 147-165. Wallingford, UK: CAB International.
Boitani, L, F Francisci, P Ciucci, and G Andreoli. 1995. Population biology and ecology of feral dogs in central Italy. Edited by James Serpell. 2nd Edition ed. Vol. 217-244, The domestic dog: Its evolution, behaviour, and interactions with people. United Kingdom: Cambridge University Press.
297
Bonanni, R, S Cafazzo, P Valsecchi, and E Natoli. 2010. "Effect of affiliative and agonistic relationships on leadership behaviour in free-ranging dogs." Animal Behaviour 79 (5):981-991.
Bradley, T, and R King. 2012. "The Dog Economy Is Global—but What Is the World's True Canine Capital?" The Atlantic, November 13.
Bräuer, J, K Schönefeld, and J Call. 2013. "When do dogs help humans?" Applied Animal Behaviour Science 148 (1–2):138-149. doi: 10.1016/j.applanim.2013.07.009.
Briggs, DJ. 2012. "The role of vaccination in rabies prevention." Current Opinion in Virology 2 (3):309-314. doi: 10.1016/j.coviro.2012.03.007.
Brooks, ME, K Kristensen, K J. van Benthem, A Magnusson, CW Berg, A Nielsen, HJ Skaug, M Maechler, and BM Bolker. 2017 "glmmTMB balances speed and flexibility among packages for Zero-inflated Generalized Linear Mixed Modeling." The R Journal 9 (2):378-400.
Buchy, P, S Preiss, V Singh, and P Mukherjee. 2017. "Heterogeneity of rabies vaccination recommendations across Asia." Tropical Medicine and Infectious disease 2 (3):23.
Buckland, ST, KP Burnham, DR Anderson, and JL Laake. 1993. "Density estimation using distance sampling." Chapman Hall, London, Inglaterra: 133-134.
Bunge, JA. 2013. "A survey of software for fitting capture–recapture models." Wiley Interdisciplinary Reviews: Computational Statistics 5 (2):114-120.
Burki, T. 2008. "The global fight against rabies." Lancet 372 (9644):1135-1136. doi: 10.1016/S0140-6736(08)61462-2.
Burt, WH. 1943. "Territoriality and home range concepts as applied to mammals." Journal of Mammalogy 24 (3):346-352.
Butcher, R. 1999. "Stray dogs - a worldwide problem." The Journal of Small Animal Practice 40 (9):458-459.
Butler, JR, and J Bingham. 2000. "Demography and dog-human relationships of the dog population in Zimbabwean communal lands." The Veterinary Record 147 (16):442-446.
Byrnes, H, A Britton, and T Bhutia. 2017. "Eliminating Dog-Mediated Rabies in Sikkim, India: A 10-Year Pathway to Success for the SARAH Program." Frontiers in Veterinary Science 4 (28). doi: 10.3389/fvets.2017.00028.
298
Cafazzo, S, P Valsecchi, R Bonanni, and E Natoli. 2010. "Dominance in relation to age, sex, and competitive contexts in a group of free-ranging domestic dogs." Behavioral Ecology 21 (3):443-455.
Calenge, C. 2011. "Home range estimation in R: the adehabitatHR package." Office national de la classe et de la faune sauvage: Saint Benoist, Auffargis, France.
Carrara, P, P Parola, P Brouqui, and P Gautret. 2013. "Imported human rabies cases worldwide, 1990–2012." PLoS Neglected Tropical Diseases 7 (5):e2209.
Castillo-Neyra, R, J Brown, K Borrini, C Arevalo, MZ Levy, A Buttenheim, GC Hunter, V Becerra, J Behrman, and VA Paz-Soldan. 2017. "Barriers to dog rabies vaccination during an urban rabies outbreak: Qualitative findings from Arequipa, Peru." PLoS Neglected Tropical Diseases 11 (3):e0005460.
Caughley, G. 1977. Analysis of Vertebrate Populations: John Wiley & Sons.
Chandan, N, and K Kotrabasappa. 2016. "Awareness of animal bite and rabies among agricultural workers in rural Dharwad, Karnataka, India." International Journal of Community Medicine and Public Health 3 (7):1851-1855.
Chao, A. 1987. "Estimating the population size for capture-recapture data with unequal catchability." Biometrics 43 (4):783-791.
Chao, A, RK Colwell, C-H Chiu, and D Townsend. 2017. "Seen once or more than once: applying Good–Turing theory to estimate species richness using only unique observations and a species list." Methods in Ecology and Evolution: 8 (10):1221-1232. doi:10.1111/2041-210X.12768
Chao, A, SM Lee, and SL Jeng. 1992. "Estimating population size for capture-recapture data when capture probabilities vary by time and individual animal." Biometrics: 201- 216.
Chapman, DG. 1951. Some properties of the hypergeometric distribution with applications to zoological sample censuses. Vol. 1. Calfornia: University of California Press.
Chatterjee, P. 2009. "India's ongoing war against rabies." Bulletin of the World Health Organisation 87(12):890-1. doi: 10.2471/BLT.09.021209.
Chatterjee, S, and H Riaz. 2013. "Rabies: beware of the dog." British Medical Journal 13:347-349. doi: 10.1136/bmj.f5912.
299
Chaudhuri, S. 2015. "Knowledge, attitude and practice about animal bite and rabies among victims attending a rural hospital in eastern India." Global Journal of Medicine and Public Health 4 (1).
Childs, JE, LE Robinson, R Sadek, A Madden, ME Miranda, and NL Miranda. 1998. "Density estimates of rural dog populations and an assessment of marking methods during a rabies vaccination campaign in the Philippines." Preventive Veterinary Medicine 33 (1- 4):207-18.
Chowdhury, R, A Mukherjee, S Naskar, and SK Lahiri. 2013. "A study on knowledge of animal bite management and rabies immunization among interns of a government medical college in Kolkata." International Journal of Medicine and Public Health 3 (1):17.
Chua, D, J Rand, and J Morton. 2017. "Surrendered and Stray Dogs in Australia— Estimation of Numbers Entering Municipal Pounds, Shelters and Rescue Groups and Their Outcomes." Animals 7 (7):50-78. doi: 10.3390/ani7070050.
Ciucci, P, and L Boitani. 1998. "Wolf and dog depredation on livestock in central Italy." Wildlife Society Bulletin: 504-514.
Cleaveland, S, H Beyer, K Hampson, D Haydon, F Lankester, T Lembo, F-X Meslin, M Morters, Z Mtema, M Sambo, and S Townsend. 2014. "The changing landscape of rabies epidemiology and control." Onderstepoort Journal of Veterinary Research 81 (2, Suppl. 4):731. doi: 10.4102/ojvr.v81i2.731.
Cleaveland, S, and K Hampson. 2017. "Rabies elimination research: juxtaposing optimism, pragmatism and realism." Proceedings of the Royal Society B: Biological Sciences 284 (1869):20171880.
Cleaveland, S, M Kaare, D Knobel, and MK Laurenson. 2006. "Canine vaccination— providing broader benefits for disease control." Veterinary Microbiology 117 (1):43-50.
Cleaveland, S, F Lankester, S Townsend, T Lembo, and K Hampson. 2014. "Rabies control and elimination: a test case for One Health." Veterinary Record 175 (8):188-193.
Cliquet, F, and J Barrat. 2012. "Assuring the quality and sustainability of a rabies dog vaccination programme: vaccination, rabies surveillance and post-vaccination monitoring." Compendium of the OIE Global Conference on Rabies Control 7–9 September 2011 Incheon-Seoul (Republic of Korea).
Cliquet, F, JP Gurbuxani, HK Pradhan, B Pattnaik, SS Patil, A Regnault, H Begouen, AL Guiot, R Sood, and P Mahl. 2007. "The safety and efficacy of the oral rabies vaccine SAG2 in Indian stray dogs." Vaccine 25 (17):3409-3418.
300
Clutton-Brock, J. 1995. Origins of the dog: domestication and early history. Edited by James Serpell, The domestic dog: Its evolution, behaviour and interactions with people. London, UK: The press syndicate of the University of Cambridge.
Clutton-Brock, J. 1999. A natural history of domesticated mammals: Cambridge University Press 29-41.
Cohen, AC. 1960. "Estimating the parameter in a conditional Poisson distribution." Biometrics 16 (2):203-211.
Coleman, PG, and C Dye. 1996. "Immunization coverage required to prevent outbreaks of dog rabies." Vaccine 14 (3):185-186.
Coleman, PG, EM Fèvre, and S Cleaveland. 2004. "Estimating the Public Health Impact of Rabies." Emerging Infectious Diseases 10 (1):140-142. doi: 10.3201/eid1001.020744.
Conan, A, O Akerele, G Simpson, B Reininghaus, J van Rooyen, and D. Knobel. 2015. "Population dynamics of owned, free-roaming dogs: implications for rabies control." PLoS Neglected Tropical Diseases 9 (11):e0004177.
Cooch, E, and G White. 2006. "Program MARK: a gentle introduction." Available in. pdf format for free download at http://www.phidot.org/software/mark/docs/book.
Crandell, RA 1991. "Arctic fox rabies." In The Natural History of Rabies, edited by GM Baer. Boca Raton,FL,USA: CRC Press.
Crowcroft, NS, and N Thampi. 2015. "The prevention and management of rabies." British Medical Journal 350:g7827.
Dalla Villa, P, S Kahn, L Stuardo, L Iannetti, A Di Nardo, and JA Serpell. 2010. "Free- roaming dog control among OIE-member countries." Preventive Veterinary Medicine 97 (1):58-63. doi: 10.1016/j.prevetmed.2010.07.001.
Dane, B. 2006. "Likert Scales are the meaning of life." Topic Report. CPSC681:1-10.
Daniels, TJ. 1983. "The social organization of free-ranging urban dogs. I. Non-estrous social behavior." Applied Animal Ethology 10 (4):341-363.
Daniels, TJ, and M Bekoff. 1989. "Population and social biology of free-ranging dogs, Canis familiaris." Journal of Mammalogy 70 (4):754-762. doi: 10.2307/1381709.
Dar, KH, MM Ansari, MM Bhat, SH Dar, and Hakim Ather. 2014. "Studies on dog bites of domestic animals and avian in Kashmir Valley." International Journal of Veterinary Science 3 (3):151-154.
301
David, D, N Dveres, I Davidson, J Yagil, Z Dvorkin, and Z Oved. 2012. "Geographic Translocation of Dog Rabies by Tourism." Israel Journal of Veterinary Medicine 67 (3):139-141, 23.
Davis, DE. 1982. CRC handbook of census methods for terrestrial vertebrates: CRC Press.
Davlin, SL, SM Lapiz, ME Miranda, and KO Murray. 2014. "Knowledge, attitudes, and practices regarding rabies in Filipinos following implementation of the Bohol Rabies Prevention and Elimination Programme." Epidemiology & Infection 142 (7):1476-1485.
Davlin, SL, and HM VonVille. 2012. "Canine rabies vaccination and domestic dog population characteristics in the developing world: A systematic review." Vaccine 30 (24):3492-3502. doi: 10.1016/j.vaccine.2012.03.069.
Dean, DJ, and MK Abelseth. 1973. "Laboratory techniques in rabies: the fluorescent antibody test." Monograph Series. World Health Organization (23):73.
Demirbas, YS, E Safak, B Emre, İ Piskin, H Ozturk, and G Da Graça Pereira. 2017. "Rehabilitation program for urban free–ranging dogs in a shelter environment can improve behavior and welfare." Journal of Veterinary Behavior 18:1-6.
Dhand, NK, BD Rai, S Tenzin, K Tsheten, P Ugyen, K Singye, and MP Ward. 2012. "Community-based study on knowledge, attitudes and perception of rabies in Gelephu, south-central Bhutan." International Health 4 (3):210-219.
Dhiman, AK, A Thakur, and SR Mazta. 2017. "Treatment seeking behavior of the dog bite patients in Himachal Pradesh, India: a qualitative study." International Journal Of Community Medicine And Public Health 3 (8):2064-2069.
Dias, RA, AGA Guilloux, MR Borba, MC de Lourdes Guarnieri, R Prist, F Ferreira, M Amaku, JSF Neto, and M Stevenson. 2013. "Size and spatial distribution of stray dog population in the University of São Paulo campus, Brazil." Preventive Veterinary Medicine 110 (2):263-273.
Dietzschold, B, J Li, M Faber, and M Schnell. 2008. "Concepts in the pathogenesis of rabies." Future Virology 3 (5):481-490. doi: 10.2217/17460794.3.5.481.
Dodet, B. 2007. "An important date in rabies history." Vaccine 25 (52):8647-8650.
Dodet, B, EV Adjogoua, AR Aguemon, OH Amadou, AL Atipo, BA Baba, S Bara Ada, P Boumandouki, and H Bourhy. 2008. "Fighting rabies in Africa: the Africa Rabies Expert Bureau (AfroREB)." Vaccine 26 (50):6295-6298.
302
Dodet, B., A. Goswami, A. Gunasekera, F. de Guzman, S. Jamali, C. Montalban, W. Purba, B. Quiambao, N. Salahuddin, G. Sampath, Q. Tang, T. Tantawichien, O. Wimalaratne, and A. Ziauddin. 2008. "Rabies awareness in eight Asian countries." Vaccine 26 (50):6344-6348. doi: 10.1016/j.vaccine.2008.09.003.
Downes, M, MJ Canty, and SJ More. 2009. "Demography of the pet dog and cat population on the island of Ireland and human factors influencing pet ownership." Preventive Veterinary Medicine 92 (1–2):140-149. doi: 10.1016/j.prevetmed.2009.07.005.
Downes, MJ, RS Dean, JH Stavisky, VJ Adams, DJC Grindlay, and ML Brennan. 2013. "Methods used to estimate the size of the owned cat and dog population: a systematic review." BMC Veterinary Research 9 (1):121.
Duggal, B. 2004. "The slums of Panchkula." Social Change 34 (4):38-46. doi: 10.1177/004908570403400403.
Durr, S, NK Dhand, C Bombara, S Molloy, and MP Ward. 2017. "What influences the home range size of free-roaming domestic dogs?" Epidemiology & Infection 145 (7):1339-1350. doi: 10.1017/S095026881700022X.
Dürr, S, and MP Ward. 2014. "Roaming behaviour and home range estimation of domestic dogs in Aboriginal and Torres Strait Islander communities in northern Australia using four different methods." Preventive Veterinary Medicine 117 (2):340-357. doi: 10.1016/j.prevetmed.2014.07.008.
Dutta, JK. 2002. "Disastrous results of indigenous methods of rabies prevention in developing countries." International Journal of Infectious Diseases 6 (3):236-237.
Egenvall, A, A Nødtvedt, J Penell, L Gunnarsson, and BN Bonnett. 2009. "Insurance data for research in companion animals: benefits and limitations." Acta Veterinaria Scandinavica 51 (1):42. doi: 10.1186/1751-0147-51-42.
El Berbri, I, MJ Ducrotoy, A-F Petavy, O Fassifihri, AP Shaw, M Bouslikhane, F Boue, SC Welburn, and A Dakkak. 2015. "Knowledge, attitudes and practices with regard to the presence, transmission, impact, and control of cystic echinococcosis in Sidi Kacem Province, Morocco." Infectious Diseases of Poverty 4 (1):48.
Faleke, OO. 2003. "Studies on dog population and its implication for rabies control." Nigerian Journal of Animal Production 30 (2):242-245.
303
Fei, S-Y, J-T Chiang, C-Y Fei, C-H Chou, and M-C Tung. 2012. "Estimating stray dog populations with the regression method versus Beck’s method: a comparison." Environmental and Ecological Statistics 19 (4):485-498. doi: 10.1007/s10651-012-0197- 0.
Finnegan, CJ, SM Brookes, N Johnson, J Smith, KL Mansfield, VL Keene, LM McElhinney, and AR Fooks. 2002. "Rabies in North America and Europe." Journal of the Royal Society of Medicine 95 (1):9-13.
Fitzpatrick, MC, HA Shah, A Pandey, AM Bilinski, M Kakkar, AD Clark, JP Townsend, SS Abbas, and AP Galvani. 2016. "One Health approach to cost-effective rabies control in India." Proceedings of the National Academy of Sciences 113 (51):14574-14581.
Florian, H. 2018 "DHARMa: Residual Diagnostics for Hierarchical (Multi-Level / Mixed) Regression Models.”
Font, E. 1987. "Spacing and social organization: urban stray dogs revisited." Applied Animal Behaviour Science 17 (3):319-328.
Fox, MW, AM Beck, and E Blackman. 1975. "Behavior and ecology of a small group of urban dogs (Canis familiaris)." Applied Animal Ethology 1 (2):119-137.
Franka, R, TG Smith, JL Dyer, X Wu, M Niezgoda, and CE Rupprecht. 2013. "Current and future tools for global canine rabies elimination." Antiviral Research 100 (1):220- 225. doi: 10.1016/j.antiviral.2013.07.004.
Freuling, CM, K Hampson, T Selhorst, R Schröder, F-X Meslin, TC Mettenleiter, and T Müller. 2013. "The elimination of fox rabies from Europe: determinants of success and lessons for the future." Philosophical Transactions of Royal Society B: Biological Sciences 368 (1623):20120142. doi: 10.1098/rstb.2012.0142.
Gamble, L. 2014. "The mission continues." The Veterinary Record 174 (12):292-294.
Ganasva, A, B Bariya, and K Shringarpure. 2015. "Perceptions and treatment seeking behaviour of dog bite patients attending regional tertiary care hospital of central Gujarat, India." Journal of Research in Medical and Dental Science 3 (1):60-64.
Ge, P, J Tsao, S Schein, TJ Green, M Luo, and ZH Zhou. 2010. "Cryo-EM model of the bullet-shaped vesicular stomatitis virus." Science 327 (5966):689-693.
Gibson, AD. 2019. "Success in Goa rabies control, but what about other states of India?" 21st National Conference of Association for Prevention and Control of Rabies in India, Ranchi, Jharkhand, India, 6-7 July 2019.
304
Gibson, AD, P Ohal, K Shervell, IG Handel, BM Bronsvoort, RJ Mellanby, and L Gamble. 2015. "Vaccinate-assess-move method of mass canine rabies vaccination utilising mobile technology data collection in Ranchi, India." BMC Infectious Diseases 15:589. doi: 10.1186/s12879-015-1320-2.
Gibson, AD, G Yale, A Vos, J Corfmat, I Otter, A King, R Wallace, L Gamble, I Handel, R Mellanby, BM de C Bronsvoort, and S Mazeri. 2019. "Oral bait handout as a method to access roaming dogs for rabies vaccination in Goa, India: a proof of principle study." Vaccine: X: 100015. doi: 10.1016/j.jvacx.2019.100015.
Goldsmith, FB, and WJ Sutherland. 2006. Ecological Census Techniques - A Handbook. 2nd ed. University of East Anglia: Cambridge University Press.
Gongal, G, and AE Wright. 2011. "Human Rabies in the WHO Southeast Asia Region: Forward Steps for Elimination." Advances in Preventive Medicine 2011:383870. doi: 10.4061/2011/383870.
Govere, J, D Durrheim, KI La Grange, A Mabuza, and M Booman. 2000. "Community knowledge and perceptions about malaria and practices influencing malaria control in Mpumalanga province, South Africa." South African Medical Journal 90 (6):611-6.
Gsell, AS, DL Knobel, S Cleaveland, RR Kazwala, P Vounatsou, and J Zinsstag. 2012. "Domestic dog demographic structure and dynamics relevant to rabies control planning in urban areas in Africa: the case of Iringa, Tanzania." BMC Veterinary Research 8 (1):236. doi: 10.1186/1746-6148-8-236.
Guadu, T, A Shite, M Chanie, B Bogale, and T Fentahum. 2014. "Assessment of knowledge, attitude and practices about rabies and associated factors: in the case of Bahir Dar town." Global Veterinaria 13 (3):348-54. doi: 10.5829/idosi.gv.2014.13.03.8579.
Gupta, PK, VK Chaturvedi, PC Verma, and KD Pandey. 2005. "Differentiation of rabies fixed and street viruses using RT-PCR coupled with restriction endonuclease analysis." Indian Journal of Biotechnology 4:284-286.
Hampson, K, L Coudeville, T Lembo, M Sambo, A Kieffer, M Attlan, J Barrat, JD Blanton, DJ Briggs, S Cleaveland, P Costa, CM Freuling, E Hiby, L Knopf, F Leanes, F- X Meslin, A Metlin, ME Miranda, T Mueller, LH Nel, S Recuenco, CE Rupprecht, C Schumacher, L Taylor, M Antonio, N Vigilato, J Zinsstag, J Dushoff, and Global Alliance for Rabies Control Partners for Rabies Prevention. 2015. "Estimating the Global Burden of Endemic Canine Rabies” PloS Neglected Tropical Diseases 9 (5):e0003786. doi: 10.1371/journal.pntd.0003786.
Hampson, K, J Dushoff, S Cleaveland, DT Haydon, M Kaare, C Packer, and A Dobson. 2009. "Transmission dynamics and prospects for the elimination of canine rabies." PLoS Biology 7 (3):e1000053. doi: 10.1371/journal.pbio.1000053.
305
Helmick, CG, RV Tauxe, and AA Vernon. 1987. "Is there a risk to contacts of patients with rabies?" Review of Infectious Diseases 9 (3):511-518.
Hemachudha, T, G Ugolini, S Wacharapluesadee, W Sungkarat, S Shuangshoti, and J Laothamatas. 2013. "Human rabies: neuropathogenesis, diagnosis, and management." The Lancet Neurology 12 (5):498-513.
Herbert, M, R Basha, and S Thangaraj. 2012. "Community perception regarding rabies prevention and stray dog control in urban slums in India." Journal of Infection and Public Health 5 (6):374-380.
Hiby, E. 2013. Dog Population Management, In Dogs, Zoonoses and Public Health edited by Calum NL Macpherson. F-X Meslin and AI Wandeler, 178-204, London, CABI.
Hiby, E, and L Hiby. 2017. "Direct observation of dog density and composition during street counts as a resource efficient method of measuring variation in roaming dog population over time and between locations." Animals 7 (8): E57 doi: doi:10.3390/ani7080057.
Hiby, LR, JF Reece, R Wright, R Jaisinghani, B Singh, and EF Hiby. 2011. "A mark- resight survey method to estimate the roaming dog population in three cities in Rajasthan, India." BMC Veterinary Research 7:46. doi: 10.1186/1746-6148-7-46.
Hlongwana, KW, MLH Mabaso, S Kunene, D Govender, and R Maharaj. 2009. "Community knowledge, attitudes and practices (KAP) on malaria in Swaziland: A country earmarked for malaria elimination." Malaria Journal 8 (1):29. doi: 10.1186/1475-2875-8-29.
Høgåsen, HR, C Er, A Di Nardo, and P Dalla Villa. 2013. "Free-roaming dog populations: a cost-benefit model for different management options, applied to Abruzzo, Italy." Preventive Veterinary Medicine 112 (3):401-413.
Hossain, M, K Ahmed, ASP Marma, S Hossain, MA Ali, AKM Shamsuzzaman, and A Nishizono. 2013. "A survey of the dog population in rural Bangladesh." Preventive Veterinary Medicine 111 (1):134-138.
Hudson, EG, VJ Brookes, MP Ward, and S Durr. 2019. "Using roaming behaviours of dogs to estimate contact rates: the predicted effect on rabies spread." Epidemiology & Infection 147:e135. doi: 10.1017/S0950268819000189.
Huggins, R. 2001. "A note on the difficulties associated with the analysis of capture– recapture experiments with heterogeneous capture probabilities." Statistics & Probability Letters 54 (2):147-152.
306
Huggins, RM. 1989. "On the Statistical Analysis of Capture Experiments." Biometrika 76 (1):133-140. doi: 10.2307/2336377.
Hughes, J, and DW Macdonald. 2013. "A review of the interactions between free-roaming domestic dogs and wildlife." Biological Conservation 157:341-351.
Ichhpujani, RL, M Chhabra, V Mittal, D Bhattacharya, J Singh, M Bhardwaj, SK Pattanaik, N Balakrishnan, AK Reddy, and G Samnpath. 2008. "Epidemiology of animal bites and rabies cases in India. A multicentric study." The Journal of Communicable Diseases 40 (1):27-36.
Ichhpujani, RL, M Chhabra, V Mittal, D Bhattacharya, J Singh, and Shiv Lal. 2006. "Knowledge, attitude and practices about animal bites and rabies in general community- -a multi-centric study." The Journal of Communicable Diseases 38 (4):355-361.
Ivanter, EV, and NA Sedova. 2008. "Ecological monitoring of urban groups of stray dogs: An example of the city of Petrozavodsk." Russian Journal of Ecology 39 (2):105-110. doi: 10.1134/S1067413608020057.
Jackman, J, and AN Rowan. 2007. "Free-roaming dogs in developing countries: The benefits of capture, neuter, and return programs." In The State of the Animals, edited by DJ Salem & AN Rowan, 55 - 78. Washington, DC: Human Society Press.
Jain, P, and G Jain. 2014. "Study of general awareness, attitude, behavior, and practice study on dog bites and its management in the context of prevention of rabies among the victims of dog bite attending the OPD services of CHC Muradnagar." Journal of Family Medicine and Primary Care 3 (4):355.
Jemberu, WT, W Molla, G Almaw, and S Alemu. 2013. "Incidence of Rabies in Humans and Domestic Animals and People's Awareness in North Gondar Zone, Ethiopia." PloS Neglected Tropical Diseases 7 (5):e2216. doi: 10.1371/journal.pntd.0002216.
Jeremy, JDG, and AR Robert. 2006. Ecological census techniques. Edited by William J Sutherland, General Census methods. Cambridge: Cambridge University Press Chapter.
Jibat, T, H Hogeveen, and MCM Mourits. 2015. "Review on Dog Rabies Vaccination Coverage in Africa: A Question of Dog Accessibility or Cost Recovery?" PloS Neglected Tropical Diseases 9 (2):e0003447. doi: 10.1371/journal.pntd.0003447.
John, TJ, L Dandona, VP Sharma, and M Kakkar. 2011. "Continuing challenge of infectious diseases in India." The Lancet 377 (9761):252-269.
307
Joice, YS, Z Singh, and SS Datta. 2016. "Knowledge, Attitude and Practices Regarding Dog Bite and its Management Among Adults in Rural Tamil Nadu." International Journal of Scientific Research 5 (5):586-589.
Joseph, J, N Sangeetha, AM Khan, and OP Rajoura. 2013. "Determinants of delay in initiating post-exposure prophylaxis for rabies prevention among animal bite cases: hospital based study." Vaccine 32 (1):74-77.
Kakkar, M, V Venkataramanan, S Krishnan, RS Chauhan, and SS Abbas. 2012. "Moving from rabies research to rabies control: lessons from India." PLoS Neglected Tropical Diseases 6 (8):e1748.
Kakrani, VA, S Jethani, J Bhawalkar, A Dhone, and K Ratwani. 2013. "Awareness about dog bite management in rural population." Indian Journal of Community Health 25 (3):304-308.
Kalaivani, A, D Raja, and M Geetha. 2014. "Rabies in India: A Relook at the Neglected Rampant Disease." The Journal of Neurobehavioral Sciences 1 (3):88-91.
Kale, KM, SK Wadhva, NR Aswar, and ND Vasudeo. 2006. "Dog bites in children." Indian Journal of Community Medicine 31 (1):24.
Kamble, B, S Panesar, A Das, N Roy, G Yadav, A Khokhar, and J Kishore. 2016. "Knowledge, attitude and practices related to animal bites among the residents of an urbanized village in South Delhi." International Journal of Research and Development in Pharmacy and Life Sciences 5 (3):2164-8.
Kanda, K, Y Obayashi, A Jayasinghe, GSP de S Gunawardena, NY Delpitiya, NGW Priyadarshani, Chandika D Gamage, A Arai, and H Tamashiro. 2014. "Outcomes of a school-based intervention on rabies prevention among school children in rural Sri Lanka." International Health 7 (5):348-353.
Kato, M, H Yamamoto, Y Inukai, and S Kira. 2003. "Survey of the stray dog population and the health education program on the prevention of dog bites and dog-acquired infections: a comparative study in Nepal and Okayama Prefecture, Japan." Acta Medica Okayama 57 (5):261-266.
Kayali, U, R Mindekem, N Yemadji, P Vounatsou, Y Kaninga, AG Ndoutamia, and J Zinsstag. 2003. "Coverage of pilot parenteral vaccination campaign against canine rabies in N'Djamena, Chad." Bulletin of the World Health Organization 81:739-744.
King, AA, AR Fooks, M Aubert, and AI Wandeler. 2004. Historical perspective of rabies in Europe and the Mediterranean Basin: OIE Paris, France.
308
Kish, L. 1998. "Space/time variations and rolling samples." Journal of Official Statistics 14 (1):31.
Kishore, S, R Singh, and SK Ravi. 2015. "Knowledge, attitude and practice assessment in health workers regarding rabies disease and its prevention in district Dehradun of Uttarakhand." Indian Journal of Community Health 27 (3):381-385.
Kitala, P, J McDermott, M Kyule, J Gathuma, B Perry, and AI Wandeler. 2001. "Dog ecology and demography information to support the planning of rabies control in Machakos District, Kenya." Acta Tropica 78 (3):217-230.
Kitala, PM, JJ McDermott, PG Coleman, and C Dye. 2002. "Comparison of vaccination strategies for the control of dog rabies in Machakos District, Kenya." Epidemiology & Infection 129 (01):215-222. doi: 10.1017/s0950268802006957.
Knobel, DL, S Cleaveland, PG Coleman, EM Fèvre, MI Meltzer, MEG Miranda, A Shaw, J Zinsstag, and F-X Meslin. 2005. "Re-evaluating the burden of rabies in Africa and Asia." Bulletin of the World health Organization 83 (5):360-368.
Kole, AK, R Roy, and DC Kole. 2014. "Human rabies in India: a problem needing more attention." Bulletin of the World Health Organization 92 (4):230-230. doi: 10.2471/BLT.14.136044.
Krebs, CJ. 1999. Ecological methodology. Vol. 620. California: Benjamin/Cummings Menlo Park.
Krishna, SC. 2009. "The Success of the ABC-AR* Programme in India." FAO.
Kumar, A, A Thakur, and SR Mazta. 2016. "Management of dog bite cases in the health institutions of Himachal Pradesh, India- A qualitative study." International Journal of Recent Scientific Research Vol 7 (Issue 6):11566-11569.
Laffan, SW, Z Wang, and MP Ward. 2011. "The effect of neighbourhood definitions on spatio-temporal models of disease outbreaks: Separation distance versus range overlap." Preventive Veterinary Medicine 102 (3):218-229.
Lai, P, A Rawat, A Sagar, and KN Tiwari. 2005. "Prevalence of dog-bites in Delhi: Knowledge and practices of residents regarding prevention and control of rabies." Health and Population Perspectives and Issues 28 (2):50-57.
Lankau, EW, NJ Cohen, ES Jentes, LE Adams, TR Bell, JD Blanton, D Buttke, GG Galland, AM Maxted, and DM Tack. 2014. "Prevention and Control of Rabies in an Age of Global Travel: A Review of Travel ‐ and Trade ‐ Associated Rabies Events–United States, 1986–2012." Zoonoses and Public Health 61 (5):305-316.
309
Lavan, RP, AIMG King, DJ Sutton, and K Tunceli. 2017. "Rationale and support for a One Health program for canine vaccination as the most cost-effective means of controlling zoonotic rabies in endemic settings." Vaccine 35 (13):1668-1674.
Léchenne, M, T Tenzin, ME Miranda, and J Zinsstag. 2019. "Rabies in East and Southeast Asia: A Mirror of the Global Situation." In Neglected Tropical Diseases - East Asia, edited by Jürg Utzinger, Peiling Yap, Martin Bratschi and Peter Steinmann, 105-127. Cham: Springer International Publishing.
Lembo, T, M Attlan, H Bourhy, S Cleaveland, P Costa, K De Balogh, B Dodet, AR Fooks, E Hiby, and F Leanes. 2011. "Renewed global partnerships and redesigned roadmaps for rabies prevention and control." Veterinary Medicine International 2011: Article ID 923149 doi :10.4061/2011/923149
Lembo, T, K Hampson, DT Haydon, M Craft, A Dobson, J Dushoff, E Ernest, R Hoare, M Kaare, and T Mlengeya. 2008. "Exploring reservoir dynamics: a case study of rabies in the Serengeti ecosystem." Journal of Applied Ecology 45 (4):1246-1257.
Lukacs, P. 2009. "Closed population capture-recapture models." In Program MARK: a gentle introduction, Chapter 14, 14:1-60. Available in pdf format for free download at http://www.phidot.org/software/mark/docs/book. Ithaca, New York, USA.
Macdonald, DW, and GM Carr. 2016. "Variation in dog society: between resource dispersion and social flux." In The domestic dog: its evolution, behavior and interactions with people, edited by James Serpell, 199-216, Cambridge, Cambridge Unversity Press.
Macpherson, CNL. 2012. Dogs, zoonoses and public health, 2nd Edition, edited by Calum NL Macpherson. F-X Meslin and AI Wandeler, London, CABI.
Madhusudana, SN, S Sundaramoorthy, and PT Ullas. 2010. "Utility of human embryonic kidney cell line HEK-293 for rapid isolation of fixed and street rabies viruses: comparison with Neuro-2a and BHK-21 cell lines." International Journal of Infectious Diseases 14 (12):e1067-71. doi: 10.1016/j.ijid.2010.07.004.
Majumder, SS, A Bhadra, A Ghosh, S Mitra, D Bhattacharjee, J Chatterjee, AK Nandi, and A Bhadra. 2014. "To be or not to be social: foraging associations of free-ranging dogs in an urban ecosystem." Acta Ethologica 17 (1):1-8.
Majumder, SS, A Chatterjee, and A Bhadra. 2014. "A dog’s day with humans–time activity budget of free-ranging dogs in India." Current Science 106 (6):874.
Majumder, SS, M Paul, S Sau, and A Bhadra. 2016. "Denning habits of free-ranging dogs reveal preference for human proximity." Scientific Reports 6:32014.
310
Malerczyk, C, L DeTora, and D Gniel. 2011. "Imported human rabies cases in Europe, the United States, and Japan, 1990 to 2010." Journal of Travel Medicine 18 (6):402-407.
Marathe, N, and S Kumar. 2016. "Epidemiological study of animal bite victims in Central India: a cross sectional institutional study." International Journal of Community Medicine and Public Health 3 (1):78-82.
Margawani, KR, and ID Robertson. 1995. "A survey of urban pet ownership in Bali." The Veterinary Record 137 (19):486-488. doi: 10.1136/vr.137.19.486.
Maroof, KA. 2013. "Burden of rabies in India: the need for a reliable reassessment." Indian Journal of Community Health 25 (4):488-491.
Massei, G, AR Fooks, DL Horton, R Callaby, K Sharma, IP Dhakal, and U Dahal. 2017. "Free‐ Roaming Dogs in Nepal: Demographics, Health and Public Knowledge, Attitudes and Practices." Zoonoses and Public Health 64 (1):29-40.
Massei, G, and LA Miller. 2013. "Nonsurgical fertility control for managing free-roaming dog populations: a review of products and criteria for field applications." Theriogenology 80 (8):829-838. doi: 10.1016/j.theriogenology.2013.07.016.
Matibag, GC, T Kamigaki, PVR Kumarasiri, TG Wijewardana, AW Kalupahana, DRA Dissanayake, DDN De Silva, GS De S Gunawardena, Y Obayashi, and K Kanda. 2007. "Knowledge, attitudes, and practices survey of rabies in a community in Sri Lanka." Environmental Health and Preventive Medicine 12 (2):84-89.
Matibag, GC, Y Ohbayashi, K Kanda, H Yamashina, WRB Kumara, ING Perera, DDN De Silva, GS De S Gunawardena, A Jayasinghe, and RA Ditangco. 2009. "A pilot study on the usefulness of information and education campaign materials in enhancing the knowledge, attitude and practice on rabies in rural Sri Lanka." The Journal of Infection in Developing Countries 3 (01):055-064.
Matter, HC, AI Wandeler, BE Neuenschwander, LPA Harischandra, and F-X Meslin. 2000. "Study of the dog population and the rabies control activities in the Mirigama area of Sri Lanka." Acta Tropica 75 (1):95-108.
Matthew, J. 2017. "generalhoslem: Goodness of Fit Tests for Logistic Regression Models. R package version 1.3.2. ."
McCallum, DA. 2005. "A conceptual guide to detection probability for point counts and other count-based survey methods." USDA Forest Service General Tchnical Report PSW-GTR-191:754-761.
311
McClintock, B. 2011. "Mark-resight models." In A gentle introduction to program MARK. Chapter18. 18:1-34. Available in pdf format for free download at http://www. phidot.org/software/mark/docs/book. Ithaca, New York, USA.
Meek, PD. 1999. "The movement, roaming behaviour and home range of free-roaming domestic dogs, Canis lupus familiaris, in coastal New South Wales." Wildlife Research 26 (6):847-855.
Menezes, R. 2008. "Rabies in India." Canadian Medical Association Journal 178 (5):564- 566.
Meshram, HM, SS Thakre, and MB Khamgaokar. 2016. "Socio-demographic and treatment profile of outdoor patients attending anti-rabies vaccination clinic." International Journal of Interdisciplinary and Multidisciplinary Studies 3(2):1-5
Meslin, F-X. 2005. "Rabies as a traveler's risk, especially in high-endemicity areas." Journal of Travel Medicine 12 (SUPPL. 1):S30-S40.
Meslin, F-X, and DJ Briggs. 2013. "Eliminating canine rabies, the principal source of human infection: What will it take?" Antiviral Research 98 (2):291-296. doi: 10.1016/j.antiviral.2013.03.011.
Meslin, F-X, DB Fishbein, and HC Matter. 1994. "Rationale and prospects for rabies elimination in developing countries." Current Topics in Microbiology and Immunology 187:1.
Ministry of Agriculture, Govt. of India. 2015. 19th Livestock census.
Molyneux, D, Z Hallaj, GT Keusch, DP McManus, H Ngowi, S Cleaveland, P Ramos- Jimenez, E Gotuzzo, K Kar, and A Sanchez. 2011. "Zoonoses and marginalised infectious diseases of poverty: where do we stand?" Parasites & Vectors 4 (1):106.
Morey, DF. 2006. "Burying key evidence: the social bond between dogs and people." Journal of Archaeological Science 33 (2):158-175. doi: 10.1016/j.jas.2005.07.009.
Morley, CG. 2002. "Evaluating the performance of PIT tags and ear tags in a capture‐ recapture experiment." New Zealand Journal of Zoology 29 (2):143-148.
Morters, MK, S Bharadwaj, HR Whay, S Cleaveland, I M Damriyasa, and JLN Wood. 2014. "Participatory methods for the assessment of the ownership status of free-roaming dogs in Bali, Indonesia, for disease control and animal welfare." Preventive Veterinary Medicine 116 (1):203-208. doi: 10.1016/j.prevetmed.2014.04.012.
312
Morters, MK, TJ McKinley, DL Horton, S Cleaveland, JP Schoeman, O Restif, HR Whay, A Goddard, AR Fooks, IM Damriyasa, and JLN Wood. 2014. "Achieving population-level immunity to rabies in free-roaming dogs in Africa and Asia." PLoS Neglected Tropical Diseases 8 (11):e3260. doi:10.1371/journal.pntd.0003160.
Morters, MK, TJ McKinley, O Restif, AJK Conlan, S Cleaveland, K Hampson, HR Whay, IM Damriyasa, and JLN Wood. 2014. "The demography of free-roaming dog populations and applications to disease and population control." Journal of Applied Ecology 51 (4):1096-1106. doi: 10.1111/1365-2664.12279.
Morters, MK, O Restif, K Hampson, S Cleaveland, JL Wood, and AJ Conlan. 2013. "Evidence-based control of canine rabies: a critical review of population density reduction." Journal of Animal Ecology 82 (1):6-14. doi: 10.1111/j.1365- 2656.2012.02033.x.
Muriuki, BJ. 2016. "Knowledge, Attitude and Practices on Rabies in Kisumu and Siaya Counties, Kenya." Master's thesis University of Nairobi: 52-5. http://erepository.uonbi.ac.ke/bitstream/handle/11295/100128/Muriuki
Mustiana, A, J-A Toribio, M Abdurrahman, IW Suadnya, M Hernandez-Jover, AAG Putra, and MP Ward. 2015. "Owned and Unowned Dog Population Estimation, Dog Management and Dog Bites to Inform Rabies Prevention and Response on Lombok Island, Indonesia." PLoS One 10 (5):e0124092. doi: 10.1371/journal.pone.0124092.
Nagarajan, T, B Mohanasubramanian, EV Seshagiri, SB Nagendrakumar, MR Saseendranath, ML Satyanarayana, D Thiagarajan, PN Rangarajan, and VA Srinivasan. 2006. "Molecular Epidemiology of Rabies Virus Isolates in India." Journal of Clinical Microbiology 44 (9):3218-3224. doi: 10.1128/JCM.00801-06.
Nassar, R, and JE Mosier. 1980. "Canine population dynamics: a study of the Manhattan, Kansas, canine population." American Journal of Veterinary Research 41 (11):1798- 1803.
Neilson, AA, and CA Mayer. 2010. "Rabies-prevention in travellers." Australian Family Physician 39 (9):641-645.
Nguyen, AK, HT Nguyen, TN Pham, TV Hoang, and B Olowokure. 2015. "Awareness of rabies prevention and control measures among public health workers in Northern Vietnam." Public Health 129 (12):1591-6. doi: 10.1016/j.puhe.2015.07.019.
Niezgoda, M, and CE Rupprecht. 2006. "Standard operating procedure for the direct rapid immunohistochemistry test for the detection of rabies virus antigen." National Laboratory Training Network Course. Atlanta: US Department of Health and Human Services, Centers for Disease Control and Prevention: 1-16.
313
Ohri, P, K Jain, R Kumari, and SK Gupta. 2016. "A study about perception of people regarding animal bite in urban area of Dehradun." Journal of Evolution of Medical and Dental Sciences, 5 (17): 846-853
Oppenheimer, EC, and JR Oppenheimer. 1975. "Certain behavioral features in the pariah dog (Canis familiaris) in West Bengal." Applied Animal Ethology 2 (1):81-92.
Orians, GH, and PH Leslie. 1958. "A Capture-Recapture Analysis of a Shearwater Population: With a Statistical Appendix." The Journal of Animal Ecology: 27 (1):71-86.
Otis, DL, KP Burnham, GC White, and DR Anderson. 1978. "Statistical inference from capture data on closed animal populations." Wildlife monographs (62):3-135.
Otolorin, GR, JU Umoh, and AA Dzikwi. 2014. "Demographic and Ecological Survey of Dog Population in Aba, Abia State, Nigeria." International Scholarly Research Notices Veterinary Science 2014: (Article ID 806849). doi: 10.1155/2014/806849.
Otranto, D, F Dantas-Torres, and EB Breitschwerdt. 2009. "Managing canine vector- borne diseases of zoonotic concern: part one." Trends in Parasitology 25 (4):157-163. doi: 10.1016/j.pt.2009.01.003.
Özen, D, D Böhning, and İS Gürcan. 2016. "Estimation of stray dog and cat populations in metropolitan Ankara, Turkey." Turkish Journal of Veterinary and Animal Sciences 40 (1):7-12.
Pal, SK. 2001. "Population ecology of free-ranging urban dogs in West Bengal, India." Acta Theriologica 46 (1):69-78. doi: 10.1007/BF03192418.
Pal, SK, B Ghosh, and S Roy. 1998a. "Agonistic behaviour of free-ranging dogs (Canis familiaris) in relation to season, sex and age." Applied Animal Behaviour Science 59 (4):331-348.
Pal, SK, B Ghosh, and S Roy. 1998b. "Dispersal behaviour of free-ranging dogs (Canis familiaris) in relation to age, sex, season and dispersal distance." Applied Animal Behaviour Science 61 (2):123-132.
Park, JH, CH Lee, YK Won, BS Chin, HS Shin, and JY Kim. 2014. "Rabies Post- Exposure Prophylaxis of Overseas Travelers in the International Travel Clinic of the National Medical Center from 2006 to 2012, Korea." Infection & Chemotherapy 46 (1):13-20.
314
Patil, SP, VS Singh, SS Chavan, K Khandare, and RM Chaturvedi. 2014. "Study of pre- treatment practices and some of the epidemiological factors associated among dog bite cases attending outpatient department in tertiary care hospital." International Journal of Health Sciences Research 4 (4):34-39.
Patrick, S. 2017. "Oddsratio: Odds Ratio Calculation for GAM (M) s & GLM (M) s. R package version 1.0.0.”
Patronek, GJ, and AN Rowan. 1995. "Determining Dog and Cat Numbers and Population Dynamics." Anthrozoös 8 (4):199-205. doi: 10.2752/089279395787156590.
Paul, M, SS Majumder, and A Bhadra. 2014. "Grandmotherly care: a case study in Indian free-ranging dogs." Journal of Ethology 32 (2):75-82.
Paul, M, SS Majumder, S Sau, AK Nandi, and A Bhadra. 2016. "High early life mortality in free-ranging dogs is largely influenced by humans." Scientific Reports 6:19641. doi: 10.1038/srep19641.
Pérez, GE, A Conte, EJ Garde, S Messori, R Vanderstichel, and J Serpell. 2018. "Movement and home range of owned free-roaming male dogs in Puerto Natales, Chile." Applied Animal Behaviour Science. 205:74-82 doi: 10.1016/j.applanim.2018.05.022.
Perrin, T. 2009. "The Business Of Urban Animals Survey: the facts and statistics on companion animals in Canada." The Canadian Veterinary Journal 50 (1):48-52.
Perring, P, PE Rolling, and P Sureau. 1986. "A rapid rabies enzyme immuno-diagnosis (RREID): a useful and simple technique for the routine diagnosis of rabies." Journal of Biological Standardization 14 (3):217-222.
Perry, BD. 1993. "Dog ecology in eastern and southern Africa-implications for rabies control." The Onderstepoort Journal of Veterinary Research 60(4):429-36.
Prakash, M, VK Bhatti, and G Venkatesh. 2013. "Rabies menace and control–An insight into knowledge, attitude and practices." Medical Journal of Armed Forces India 69 (1):57-60.
Punjabi, GA, V Athreya, and JDC Linnell. 2012. "Using natural marks to estimate free- ranging dog Canis familiaris abundance in a MARK-RESIGHT framework in suburban Mumbai, India." Tropical Conservation Science 5 (4):510-520.
Putra, AAG, K Hampson, J Girardi, E Hiby, D Knobel, IW Mardiana, S Townsend, and H Scott-Orr. 2013. "Response to a Rabies Epidemic, Bali, Indonesia, 2008-2011." Emerging Infectious Diseases 19 (4):648-651. doi: 10.3201/eid1904.120380.
315
R Development Core Team. 2013. R: A language and environment for statistical computing. R Foundation for Statistical Computing Vienna, Austria
Raghav, PR, P Bhardwaj, and DB Saxena. 2014. "Insight for rabies prevention among travellers." International Journal of Medical Science and Public Health 3 (1):1-2.
Ratsitorahina, M, JH Rasambainarivo, S Raharimanana, H Rakotonandrasana, M-P Andriamiarisoa, FA Rakalomanana, and V Richard. 2009. "Dog ecology and demography in Antananarivo, 2007." BMC Veterinary Research 5 (1):21.
Raymond, TN, ME Roland, KMM Françoise, Z Francis, EF Livo, and STH Clovis. 2015. "Do open garbage dumps play a role in canine rabies transmission in Biyem-Assi health district in Cameroon?" Infection Ecology & Epidemiology 5 (1):26055.
Reece, JF, and SK Chawla. 2006. "Control of rabies in Jaipur, India, by the sterilisation and vaccination of neighbourhood dogs." The Veterinary Record 159:379-383.
Reese, JF. 2005. "Dogs and dog control in developing countries." In The state of the animals, edited by D.J. Salem & A.N. Rowan, 55-64. Washington, DC: Humane Society Press.
Reeta Mani, S, and SN Madhusudana. 2013. "Laboratory diagnosis of human rabies: recent advances." The Scientific World Journal 2013 (Article ID 569712) 10.1155/2013/569712
Richards, Shane A, Mark J Whittingham, and Philip A Stephens. 2011. "Model selection and model averaging in behavioural ecology: the utility of the IT-AIC framework." Behavioral Ecology and Sociobiology 65 (1):77-89.
Rinzin, K. 2015. "Population dynamics and health status of free-roaming dogs in Bhutan." PhD thesis, College of Veterinary Medicine, Murdoch University. https://researchrepository.murdoch.edu.au/id/eprint/27867/
Rinzin, K, I Robertson, and H Mahat. 2017. "Roaming dogs in Bhutan: a review on dog population management." Bhutan Journal of Animal Science: 1:75-78.
Rinzin, K, T Tenzin, and I Robertson. 2016. "Size and demography pattern of the domestic dog population in Bhutan: Implications for dog population management and disease control." Preventive Veterinary Medicine 126:39-47.
Robertson, ID, PJ Irwin, AJ Lymbery, and RCA Thompson. 2000. "The role of companion animals in the emergence of parasitic zoonoses." International Journal for Parasitology 30 (12):1369-1377.
316
Robertson, ID, KM Wilks, and P Williamson. 1993. "The Epidemiological Characteristics and Dynamics of Dog population in Aboriginal communities." Dog Health in lndigenous Communities, Cairns, Queensland.
Rohlf, VI, PC Bennett, S Toukhsati, and G Coleman. 2012. "Beliefs underlying dog owners' health care behaviors: results from a large, self-selected, internet sample." Anthrozoös 25 (2):171-185.
Rosatte, RC. 2013. "Rabies control in wild carnivores." In Rabies, edited by Alan C Jackson, 617-670. University of Manitoba, Winnipeg, MB, Canada: Academic Press, Elsevier.
Rowan, A, and T Kartal. 2018. "Dog population & dog sheltering trends in the United States of America." Animals 8 (5):68.
Rubin, HD, and AM Beck. 1982. "Ecological behavior of free-ranging urban pet dogs." Applied Animal Ethology 8 (1-2):161-168.
Rupprecht, CE. 1996. "Rhabdoviruses: rabies virus. Chapter 61" In Medical Microbiology edited by S Baron. Galveston (TX): The University of Texas Medical Branch.
Rupprecht, CE, and RV Gibbons. 2004. "Prophylaxis against rabies." The New England Journal of Medicine 351 (25):2626-2635.
Rupprecht, CE, CA Hanlon, and T Hemachudha. 2002. "Rabies re-examined." The Lancet Infectious Diseases 2 (6):327-343. doi: 10.1016/S1473-3099(02)00287-6.
Rupprecht, CE, IV Kuzmin, G Yale, T Nagarajan, and F-X Meslin. 2019. "Priorities in applied research to ensure programmatic success in the global elimination of canine rabies." Vaccine. pii: S0264-410X(19)30068-4 doi: 10.1016/j.vaccine.2019.01.015.
Salahuddin, N, S Jamali, K Ibraheem, and S Sardar. 2011. "Awareness about rabies post exposure prophylaxis in Pakistan among patients and health care workers: results from an Asian Rabies Expert Bureau study." Journal of College of Physicians and Surgeons — Pakistan 21 (8):491-4. doi: 08.2011/JCPSP.491494.
Salve, H, SA Rizwan, S Kant, SK Rai, P Kharya, and S Kumar. 2015. "Pre-treatment practices among patients attending an Animal Bite Management clinic at a primary health centre in Haryana, North India." Tropical Doctor 45 (2):123-125.
317
Sambo, M, T Lembo, S Cleaveland, HM Ferguson, L Sikana, C Simon, H Urassa, and K Hampson. 2014. "Knowledge, attitudes and practices (KAP) about rabies prevention and control: a community survey in Tanzania." PLoS Neglected Tropical Diseases 8 (12):e3310.
Santra, S, SK Lahiri, and TG Ray. 2015. "Determinants of Animal Bite and Practices Following Bite Among the Victims Attending a Block Primary Health Centre of West Bengal." International Journal of Health Sciences and Research (IJHSR) 5 (8):24-27.
Satpathy, SK, and S Venkatesh. 2006. "Human resources for health in India’s national rural health mission: dimension and challenges." Regional Health Forum. 10 (1):29-37
Schumacher, FX, and RW Eschmeyer. 1943. "The estimation of fish populations in lakes and ponds." Journal of the Tennessee Academy of Science 18:228 -249.
Scott, JP, F Bronson, and A Trattner. 1968. "Differential human handling and the development of agonistic behavior in Basenji and Shetland sheep dogs." Developmental Psychobiology: The Journal of the International Society for Developmental Psychobiology 1 (2):133-140.
Sekar, CC, and WE Deming. 1949. "On a method of estimating birth and death rates and the extent of registration." Journal of the American Statistical Association 44 (245):101- 115.
Selby, LA, JD Rhoades, JE Hewett, and JA Irvin. 1979. "A survey of attitudes toward responsible pet ownership." Public Health Reports 94 (4):380.
Serpell, J. 2016. The domestic dog edited by James Serpell, Cambridge: Cambridge University Press. doi: 10.1017/9781139161800
Shankaraiah, RH, G Bilagumba, D Narayana, A Hanumanthapa, R Annadani, and V Vijayashankar. 2013. "Knowledge, attitude, and practice of rabies prophylaxis among physicians at Indian animal bite clinics." Asian Biomedicine 7 (2):237-42. doi: 10.5372/1905-7415.0702.171.
Sharma, S, A Agarwal, AM Khan, and GK Ingle. 2016. "Prevalence of Dog Bites in Rural and Urban Slums of Delhi: A Community‐based Study." Annals of Medical and Health Sciences Research 6 (2):115-119.
318
Shimozako, HJ. 2012. "Otimização da técnica de captura-recaptura fotográfica para estimação da população canina livre em vias públicas. [Optimization of photographic capture-recapture technique for estimation of free dog population on public roads] “Dissertação Mestre em Ciencias Universidade de São Paulo, Sao Paulo.http://www.teses.usp.br/teses/disponiveis/5/5160/tde-05112008- 101348/publico/HelioJunjiShimozako.pdf
Shridevi, K, A Sridhar, K C Sekhar, and PG Deotale. 2014. "Assessment of Knowledge about Dog Bite Infield Practice Area of Alluri Sitaramaraju Academy of Medical Sciences, Eluru, Andhra Pradesh." Indian Journal of Public Health Research & Development 5 (2):26.
Shwiff, Stephanie, Katie Hampson, and Aaron Anderson. 2013. "Potential economic benefits of eliminating canine rabies." Antiviral Research 98 (2):352-356. doi: 10.1016/j.antiviral.2013.03.004.
Singh, US, and SK Choudhary. 2005. "Knowledge, attitude, behavior and practice study on dog-bites and its management in the context of prevention of rabies in a rural community of Gujarat." Indian Journal of Community Medicine 30 (3):81.
Slater, MR. 2001. "The role of veterinary epidemiology in the study of free-roaming dogs and cats." Preventive Veterinary Medicine 48 (4):273-286.
Slater, MR, A Di Nardo, O Pediconi, PD Villa, L Candeloro, B Alessandrini, and S Del Papa. 2008. "Free-roaming dogs and cats in central Italy: Public perceptions of the problem." Preventive Veterinary Medicine 84 (1–2):27-47. doi: 10.1016/j.prevetmed.2007.10.002.
Somvanshi, R. 2006. "Veterinary medicine and animal keeping in ancient India." Asian Agri-History 10 (2):133-146.
Southwood, Thomas Richard Edmund, and Peter A Henderson. 2009. Ecological methods: John Wiley & Sons.
Srimal, RC. 1997. "Turmeric: a brief review of medicinal properties." Fitoterapia 68:483- 493.
Srinivasan, Arjun, Elizabeth C Burton, Matthew J Kuehnert, Charles Rupprecht, William L Sutker, Thomas G Ksiazek, Christopher D Paddock, Jeannette Guarner, Wun-Ju Shieh, and Cynthia Goldsmith. 2005. "Transmission of rabies virus from an organ donor to four transplant recipients." New England Journal of Medicine 352 (11):1103-1111.
319
Sudarshan, MK, SN Madhusudana, BJ Mahendra, NSN Rao, DH Ashwath Narayana, S Abdul Rahman, F-X Meslin, D Lobo, K. Ravikumar, and Gangaboraiah. 2007. "Assessing the burden of human rabies in India: results of a national multi-center epidemiological survey." International Journal of Infectious Diseases 11 (1):29-35. doi: 10.1016/j.ijid.2005.10.007.
Sudarshan, MK, BJ Mahendra, SN Madhusudana, DH Ashwoath Narayana, Abdul Rahman, NSN Rao, F X-Meslin, D Lobo, and K Ravikumar. 2006. "An epidemiological study of animal bites in India: results of a WHO sponsored national multi-centric rabies survey." Journal of Communicable Diseases 38 (1):32.
Sudarshan, MK, DH Ashwath Narayana, NR Ramesh Masthi, ML Satyanarayana, Praveen Kulkarni, and SN Madhusudana. 2013. "Rural Rabies Prevention Project-A ‘One Health’Experiment in India: An Overview." International Journal of Tropical Disease & Health 3 (2):104-13.
Suraweera, W, SK Morris, R Kumar, DA Warrell, MJ Warrell, P Jha, and et al. 2012. "Deaths from Symptomatically Identifiable Furious Rabies in India: A Nationally Representative Mortality Survey." Plos Neglected Tropical Diseases 6 (10):e1847. doi: 10.1371/journal.pntd.0001847.
Suraweera, Wilson, Shaun K. Morris, Rajesh Kumar, David A. Warrell, Mary J. Warrell, and Prabhat Jha. 2012. "Deaths from symptomatically identifiable furious rabies in India: a nationally representative mortality survey." PLoS Neglected Tropical Diseases 6(10) e1847. doi:10.1371/journal.pntd.0001847.
Taylor, LH, and LH Nel. 2015. "Global epidemiology of canine rabies: past, present, and future prospects." Veterinary Medical Reseach Reports 6:361-371.
Taylor, LH, RM Wallace, D Balaram, JM Lindenmayer, DC Eckery, B Mutonono- Watkiss, E Parravani, and LH Nel. 2017. "The Role of Dog Population Management in Rabies Elimination—A Review of Current Approaches and Future Opportunities." Frontiers in Veterinary Science 4 (109). doi: 10.3389/fvets.2017.00109.
Tenzin, T, R Ahmed, NC Debnath, G Ahmed, and M Yamage. 2015. "Free-roaming dog population estimation and status of the dog population management and rabies control program in Dhaka City, Bangladesh." PLoS Neglected Tropical Diseases 9 (5):e0003784.
Tenzin, T, JS McKenzie, R Vanderstichel, BD Rai, K Rinzin, Y Tshering, R Pem, C Tshering, N Dahal, and K Dukpa. 2015. "Comparison of mark-resight methods to estimate abundance and rabies vaccination coverage of free-roaming dogs in two urban areas of south Bhutan." Preventive Veterinary Medicine 118 (4):436-448.
320
Tiwari, HK, ID Robertson, M O’Dea, J Gogoi-Tiwari, P Panvalkar, RS Bajwa, and AT Vanak. 2019. "Validation of Application SuperDuplicates (AS) enumeration tool for free- roaming dogs (FRD) in urban settings of Panchkula Municipal Corporation in north India." Frontiers in Veterinary Science 6:173. doi: 10.3389/fvets.2019.00173.
Tiwari, HK, ID Robertson, M O’Dea, and AT Vanak. 2019. "Knowledge, attitudes and practices (KAP) towards rabies and free roaming dogs (FRD) in Panchkula district of north India: A cross-sectional study of urban residents." PLoS Neglected Tropical Diseases 13 (4):e0007384. doi: 10.1371/journal.pntd.0007384.
Tiwari, HK, AT Vanak, M O'Dea, J Gogoi-Tiwari, and ID Robertson. 2018. "A comparative study of enumeration techniques for Free-roaming dogs in rural Baramati, District Pune, India." Frontiers in Veterinary Science 5:104. doi: 10.3389/fvets.2018.00104. eCollection 2018.
Tiwari, HK, AT Vanak, M O’Dea, and ID Robertson. 2016. "Meta-analyses on the risk factors of dog bite cases in India based on epidemiological studies " 18th National Conference of Association for Prevention and Control of Rabies in India Bangalore, Karnataka, India 9th-10th July 2016.
Tiwari, HK, AT Vanak, M O’Dea, and ID Robertson. 2019. "Knowledge, attitudes and practices (KAP) towards rabies and free-roaming dogs (FRD) in Shirsuphal village in western India: A community based cross-sectional study." PLoS Neglected Tropical Diseases 13 (1):e0007120. doi: 10.1371/journal.pntd.0007120.
Tomas, JA. 2017 "Epitools: Epidemiology Tools. R package version 0.5-10” https://CRAN.R-project.org/package=epitools.
Totton, SC, AI Wandeler, CJ Gartley, S Kachhawaha, M Suman, CS Ribble, RC Rosatte, and SA McEwen. 2010. "Assessing reproductive patterns and disorders in free-ranging dogs in Jodhpur, India to optimize a population control program." Theriogenology 74 (7):1115-1120. doi: 10.1016/j.theriogenology.2010.05.008.
Totton, SC, AI Wandeler, CS Ribble, RC Rosatte, and SA McEwen. 2011. "Stray dog population health in Jodhpur, India in the wake of an animal birth control (ABC) program." Preventive Veterinary Medicine 98 (2):215-220.
Totton, SC, AI Wandeler, J Zinsstag, CT Bauch, CS Ribble, RC Rosatte, and SA McEwen. 2010. "Stray dog population demographics in Jodhpur, India following a population control/rabies vaccination program." Preventive Veterinary Medicine 97 (1):51-57.
Traub, RJ, ID Robertson, PJ Irwin, N Mencke, and RCAA Thompson. 2005. "Canine gastrointestinal parasitic zoonoses in India." Trends in Parasitology 21 (1):42-48. doi: 10.1016/j.pt.2004.10.011.
321
Tripathy, RM, SP Satapathy, and N Karmee. 2017. "Assessment of knowledge, attitude and practice regarding rabies and its prevention among construction workers: a cross- sectional study in Berhampur, Odisha." International Journal of Research in Medical Sciences 5 (9):3971-3975.
Tschopp, R, S Bekele, and A Aseffa. 2016. "Dog demography, animal bite management and rabies knowledge-attitude and practices in the Awash Basin, Eastern Ethiopia." PLoS Neglected Tropical Diseases 10 (2):e0004471.
Tung, M-C, C-Y Fei, J Chiang, C-H Chou, C-Y Liao, Y-C Su, J-C Chang, and K-C Tung. 2010. "Surveys of dog populations in Taiwan from 1999 to 2009." Journal of the Chinese Society of Animal Science 39 (3):175-188.
Turner, GS, and C Kaplan. 1967. "Some properties of fixed rabies virus." Journal of General Virology 1 (4):537-551.
Udell, MAR, and CDL Wynne. 2008. "A review of domestic dogs’ (Canis familiaris) human‐ like behaviors: or why behavior analysts should stop worrying and love their dogs." Journal of the Experimental Analysis of Behavior 89 (2):247-261.
Udow, Sean J., Ruth Ann Marrie, and Alan C. Jackson. 2013. "Clinical Features of Dog- and Bat-Acquired Rabies in Humans." Clinical Infectious Diseases 57 (5):689-696. doi: 10.1093/cid/cit372.
Ugolini, G. 2007. "Use of rabies virus as a transneuronal tracer of neuronal connections: implications for the understanding of rabies pathogenesis." Developments in Biologicals 131:493-506.
Umrigar, P, GB Parmar, PB Patel, and RK Bansal. 2013. "Epidemiology of animal bite cases attending municipal tertiary care centres in Surat city: a cross-sectional study." National Journal Community Medicine 4 (1):153-7.
Vanak, AT. 2017. "Why does rabies still plague India in the 21st century?" India Alliance Newsletter (14):11-12.
Vanak, AT, and ME Gompper. 2009a. "Dietary niche separation between sympatric free- ranging domestic dogs and Indian foxes in central India." Journal of Mammalogy 90 (5):1058-1065.
Vanak, AT, and ME Gompper. 2009b. "Dogs Canis familiaris as carnivores: their role and function in intraguild competition." Mammal Review 39 (4):265-283. doi: 10.1111/j.1365-2907.2009.00148.x.
322
Varsharani, KV, LT Chinte, and YU Jadhav. 2014. "Cultural practices among animal bite cases of government medical college, Latur." International Medical Journal 1 (2):61-4.
Viana, M, S Cleaveland, J Matthiopoulos, J Halliday, C Packer, ME Craft, K Hampson, A Czupryna, AP Dobson, and EJ Dubovi. 2015. "Dynamics of a morbillivirus at the domestic–wildlife interface: Canine distemper virus in domestic dogs and lions." Proceedings of the National Academy of Sciences 112 (5):1464-1469.
Vishwanath, GR, SS Rajderkar, TV Sangrulkar, SK Sharma, and RI Gajbhiye. 2018. "Animal bite cases in western Maharashtra, India: a retrospective study 2010-2015." International Journal of Community Medicine and Public Health 5 (4):1610-1612.
Wallace, R, M Etheart, F Ludder, P Augustin, N Fenelon, R Franka, K Crowdis, P Dely, P Adrien, and J Pierre-Louis. 2017. "The Health Impact of Rabies in Haiti and Recent Developments on the Path toward Elimination, 2010–2015." The American Journal of Tropical Medicine and Hygiene 97 (4_Suppl):76-83.
Wallace, RM, EA Undurraga, JD Blanton, J Cleaton, and R Franka. 2017. "Elimination of Dog-Mediated Human Rabies Deaths by 2030: Needs Assessment and Alternatives for Progress Based on Dog Vaccination." Frontiers in Veterinary Science 4:9. doi: 10.3389/fvets.2017.00009.
Wandeler, AI. 1985. "Ecological and epidemiological data requirements for the planning of dog rabies control." In Rabies in the Tropics, edited by Mérieux C Kuwert E, Koprowski H, Bögel K, 657-661. Berlin, Heidelberg: Springer.
Wandeler, AI. 2011. "Global Perspective of Rabies". Power point of Global Conference on Rabies Control 7 - 9 September 2011 - Incheon-Seoul, Korea.
Wandeler, AI. 2012. Ecological and epidemiological data requirements for planning of dog rabies control. Edited by Ernst Kuwert, Rabies in the tropics. Berlin: Springer Science & Business Media.
Wandeler, AI, and J Bingham. 2013. "Dogs and rabies." In Dogs, Zoonoses and Public Health, edited by Calum NL Macpherson. F-X Meslin and AI Wandeler, 43-67, London, CABI.
Wandeler, AI, A Budde, S Capt, A Kappeler, and H Matter. 1988. "Dog Ecology and Dog Rabies Control." Reviews of Infectious Diseases 10:S684-S688. doi: 10.2307/4454718.
Wandeler, AI, HC Matter, A Kappeler, and A Budde. 1993. "The ecology of dogs and canine rabies: a selective review." Revue scientifique et technique (International Office of Epizootics) 12 (1):51-71.
323
White, GC. 2008. "Closed population estimation models and their extensions in Program MARK." Environmental and Ecological Statistics 15 (1):89-99.
White, GC, KP Burnham, and DR Anderson. 2001. "Advanced features of program MARK." Wildlife, land, and people: priorities for the 21st century. Proceedings of the second international wildlife management congress. The Wildlife Society, Bethesda, Maryland, USA.
WHO. 1987. "Guidelines for dog rabies control." World Health Organization, Geneva (VPH/83.43). https://www.who.int/rabies/en/Guidelines_for_dog_rabies_control.pdf
WHO. 2005. WHO Expert Consultation on Rabies: first report. In WHO Technical Report Series 931. Geneva. https://apps.who.int/iris/handle/10665/43262
WHO. 2006. "The control of neglected zoonotic diseases: a route to poverty alleviation" : report of a joint WHO/DFID-AHP meeting, 20 and 21 September 2005, WHO Headquarters, Geneva, with the participation of FAO and OIE. https://www.who.int/zoonoses/Report_Sept06.pdf
WHO. 2013. WHO Expert Consultation on Rabies: Second report. In World Health Organization technical report series, Geneva, Switzerland.https://apps.who.int/iris/handle/10665/85346.
WHO. 2016. Global elimination of dog-mediated human rabies: report of the rabies global conference. Geneva, Switzerland, 10–11 December 2015 https://www.who.int/rabies/resources/who_htm_ntd_nzd_2016.02/en/
WHO.2018a. Weekly Epidemiological record. Geneva, Switzerland WHO.
WHO.2018b. WHO expert consultation on rabies: third report. In WHO Technical Report Series 1012. https://apps.who.int/iris/handle/10665/272364.
Widyastuti, MDW, KL Bardosh, Sunandar, C Basri, E Basuno, A Jatikusumah, RA Arief, AAG Putra, A Rukmantara, ATS Estoepangestie, I Willyanto, IK Natakesuma, IP Sumantra , D Grace, F Unger, and J Gilbert . 2015. "On dogs, people, and a rabies epidemic: results from a sociocultural study in Bali, Indonesia." Infectious Diseases of Poverty 4 (1):1-18.
Wilde, H, S Sirikawin, A Sabcharoen, D Kingnate, T Tantawichien, PAL Harischandra, N Chaiyabutr, DGH De Silva, L Fernando, and JB Liyanage. 1996. "Failure of postexposure treatment of rabies in children." Clinical Infectious Diseases 22 (2):228- 232.
324
Williams, BK, JD Nichols, and MJ Conroy. 2002. Analysis and management of animal populations: Academic Press.
Windiyaningsih, C, H Wilde, F-X Meslin, T Suroso, and HS Widarso. 2004. "The rabies epidemic on Flores Island, Indonesia (1998-2003)." Journal of the Medical Association of Thailand 87 (11):1389-1393.
WSPA. 2010. Surveying roaming dog populations: guidelines on methodology. London: World Society for Protection of Animals.
Wunner, WH, and DJ Briggs. 2010. "Rabies in the 21 st century." PLoS Neglected Tropical Diseases 4 (3):e591. doi: 10.1371/journal.pntd.0000591.
Yoak, AJ, JF Reece, SD Gehrt, and IM Hamilton. 2014. "Disease control through fertility control: Secondary benefits of animal birth control in Indian street dogs." Preventive Veterinary Medicine 113 (1):152-156. doi: 10.1016/j.prevetmed.2013.09.005.
Young, JK, KA Olson, RP Reading, S Amgalanbaatar, and J Berger. 2011. "Is wildlife going to the dogs? Impacts of feral and free-roaming dogs on wildlife populations." BioScience 61 (2):125-132.
325
Appendices
326
Appendix I
Survey questionnaire for urban and rural respondents including pet dog owners to determine knowledge, attitudes and practices of the community towards dogs and the control of rabies
1. Household Information
1.1 Name of the respondent 1.2 Gender Male Female 1.3 Age 1.4 How many people including yourself live in this household? 1.5 How many children aged 14 years or less live in this household? 1.6 What is your highest educational qualification? No formal education Please mark ‘×’ against the level of education. Primary school Secondary school Matriculation College Graduate Post- Graduation 1.7 How would you describe your occupation? Unskilled work Please mark ‘x’ against suitable option. Skilled work Small trading Shop owner Business as self- employed professional Businessmen Clerical/salesman Supervisory level Officer/executives Senior officers Others, please specify
1.6 What religion are you? Hindu ( ) Islam ( ) Buddhist ( ) Christian ( ) Prefer not to say ( ) No religion ( ) 1.9 Do you own any pet animal(s)? If no, proceed to 1.11 Yes ( ) No ( ) 1.10 What pets do you own? Dog ( ) Cat ( ) Other, please specify ( ) 1.11 Do you own any livestock? If no, proceed to section 2 Yes ( ) No ( )
327
1.12 What kind and number of livestock do you own? Cattle/buffaloes ( ) Horses ( ) Pigs ( ) Sheep/goats ( ) Poultry ()
2. Details of ward and garbage disposal
2.1 Would you rank the garbage disposal system of Satisfactory ( ) your ward as satisfactory or unsatisfactory? unsatisfactory ( ) 2.2 Would you rank the general cleanliness of open Satisfactory ( ) areas and streets of your ward as satisfactory or unsatisfactory ( ) unsatisfactory?
3. Knowledge of rabies 3.1 Have you ever heard of the disease called rabies? Yes ( ) No () If no proceed to section 4. 3.2 In your opinion can dogs transmit rabies? Yes () No () Not sure ( ) 3.3 In your opinion can cats transmit rabies? Yes () No ( ) Not sure ( ) 3.4 In your opinion can rats transmit rabies? Yes ( ) No ( ) Not sure ( ) 3.5 In your opinion which animal most commonly Don’t know ( ) Dogs () causes rabies? Cats ( ) Rats ( ) Monkeys ( ) wild life ( )Others, please specify ( )
3.6 In your opinion can animal bites transmit rabies? Yes ( ) No ( ) Not sure ( ) 3.7 In your opinion can licks/scratches from animals Yes ( ) No ( ) transmit rabies? Not sure ( ) 3.8 Do you think rabies is fatal if symptoms appear? Yes ( ) No ( ) Not sure ( ) 3.9 In your opinion can rabies be prevented? Yes ( ) No ( ) Not sure ( ) 3.10 In your opinion will application of local Yes ( ) No ( ) Not treatments, like chilli powder and turmeric, on sure ( ) animal bite wounds prevent rabies? 3.11 In your opinion should an animal bite wound Yes ( ) No ( ) Not be washed with soap and water to reduce sure ( ) chances of rabies infection? 3.12 In your opinion is it necessary to go to hospital Yes ( ) No ( ) Not if someone is bitten by a dog, even if the injury sure ( ) is not severe? 3.13 Are you aware that post-bite anti-rabies vaccines Yes ( ) No ( ) can prevent rabies in humans? Not sure ( ) 3.14 Can rabies be prevented by vaccinating dogs Yes ( ) No ( ) against the disease? Not sure ( ) 3.15 Can rabies be controlled by restricting the size of Yes ( ) No ( ) the stray dog population? Not sure ( )
328
3.16 If you saw a dog with signs of rabies would Yes ( ) No ( ) Not you inform the municipal authorities? sure ( )
3.17 Have any awareness campaigns been Yes ( )No ( ) Not sure organised in your locality during the last two ( ) years about rabies and how to control it?
329
4. Free roaming dogs
4.1 Are there any free roaming dogs in your Yes ( ) No() locality? If yes proceed to next question. Not sure ( ) If no and you have pet dogs, please proceed to section 5. If no and you do not own pet dogs, the questionnaire for you ends here. Thanks for your time to complete this questionnaire. 4.2 Where do you think these free roaming dogs From nearby localities ( ) come from? Breeding of local dogs ( ) Abandoned pet dogs ( ) Others, please specify ( ) Not Sure ( ) 4.3 Approximately how many free roaming dogs do you think are present in your locality? 4.4 Do you feel the free roaming dogs in your locality Yes ( ) No ( ) are useful to society? If no, proceed to 4.6 Not sure ( ) 4.5 What do you think are the benefits of free Guarding ( ) Keep away roaming dogs? wild animals ( ) Keep away thieves ( ) Other, please specify ( ) 4.6 Do you believe the free roaming dogs in your Yes ( ) No ( ) locality are a nuisance or a problem for the Not sure ( ) society? 4.7 Do you think that the free roaming dogs in your Yes ( ) No ( ) locality are a threat to human health? Not sure ( ) 4.8 Where do you think these dogs get their food? Fed by residents ( ) Garbage dumps ( ) Litter from streets ( ) Other, please specify ( ) Not sure ( ) 4.9 Do you ever feed free roaming stray dogs? If yes, Yes ( ) No ( ) proceed to next question, else go to 4.13 Not Sure ( ) 4.10 Do you think feeding of free roaming dogs is part Yes ( ) No ( ) of your religious duties? Not Sure ( ) 4.11 Do you think feeding of free roaming dogs is an Yes ( ) No ( ) act of love/ compassion towards these animals? Not sure ( ) 4.12 Do you think feeding of free roaming dogs is Yes ( ) No ( ) better than wasting the food? Not sure ( ) 4.13 Would you rank the health of the free roaming Good health ( ) dogs in your locality as good, average or poor? Average health( ) Poor health ( ) 4.14 If you see an injured free roaming stray dog, Yes ( ) No ( ) would you take it to a veterinarian? Not sure ( ) 4.15 In your opinion should people who feed / shelter Yes ( ) No ( ) these dogs take responsibility for their health Not sure ( ) and vaccination? 4.16 In your opinion is it the responsibility of Yes ( ) No ( ) Not the Government to take care of the sure ( ) health of free roaming dogs?
330
4.17 Are you aware of any programmes Yes ( ) No ( ) undertaken in the last two years in your area Not sure ( ) to control the dog population? 4.18 In your opinion which of the following is the Culling ( ) best way to control the free roaming free dog Impounding ( ) population? Culling; Impounding; birth birth control operations ( ) control operations; better garbage better garbage management; if any other, please specify management ( ) if any other, please specify () Not sure ( )
5. Dog owning population (only for pet dog owners)
5.1 How many dogs do you own? Male Female Total 5.2 What is their age /ages? 5.3 What is the breed of your pet dogs? Local ( ) Pedigreed ( ), specify
Mixed ( )
5.4 Where did you get your pet from? Purchased Gifted ( ) Adopted ( ) Offspring of owned bitch ( ) Other, please specify() 5.5 Would you prefer to own a pedigreed pup to a Yes ( ) No ( ) local Indian Native dog? If yes, proceed to next Not sure ( ) question, else to 5.7 5.6 What are the reasons for you to prefer pedigreed Social status ( ) dogs rather than local Indian Native dogs? Intelligence of dogs ( ) Cleanliness ( ) Other reasons, please specify ( ) No specific reason ( ) 5.7 Is your dog always confined to your home Yes ( ) No ( ) premises? If yes, proceed to next question, else to 5.9 5.8 When away from your house premises, does your dog accompany you or members of your household (a) always (b) sometimes or (c) rarely (d) never 5.9 Is your pet registered? Yes ( ) No ( ) Applied for ( ) Not sure ( ) 5.10 In the last year have you taken your dog to a Yes ( ) veterinarian? If no proceed to 5.13 No ( ) 5.11 How many times was your dog taken to the veterinarian in the last year? 5.12 Has your dog ever been vaccinated against rabies? Yes ( ) No ( ) If not vaccinated, proceed to 5.14 Not sure ( )
331
5.13 If vaccinated, when was the last vaccine given? 5.14 Has your dog been operated upon so that it cannot Yes ( ) No ( ) breed? Don’t Know ( ) 5.15 If your dog has not been operated to stop Sterilising Cost breeding, is there a reason for this? ( ) Lack of service ( ) Not aware of such procedures ( ) Breeding purpose ( ) Cruel practice ( ) Religion ( ) Other, please specify () No reason ( )
Thank you for your time to complete this questionnaire. This information will help understand the role of dogs in the community and the control of rabies.
332
Appendix II
Rabies: Knowledge, attitudes and practices at primary health centres
1. Demographic details 1.1 What is your highest medical qualification? 1.2 How many years have you been working in the health profession? 1.3 Have you undergone any special training Yes ( ) No ( ) pertaining to the management of animal bite cases?
2. Knowledge, attitudes and practices about rabies
2.1 Which animal is responsible for the most cases of bite injuries presented to your practice/clinic? 2.2 Have you heard of rabies? If yes, please proceed Yes ( ) No ( ) to next question. If no, please proceed to 2.9 2.3 Do you think rabies can be spread to another Yes ( ) No ( ) human from a human patient with rabies? Not Sure ( ) 2.4 Do you think rabies can be spread through the Yes ( ) No ( ) bite of an animal? Not Sure ( ) 2.5 Do you think rabies can be spread through licks Yes ( ) No ( ) or scratches from an animal? Not Sure ( ) 2.6 Do you think rabies can be spread through Yes ( ) No ( ) contaminated food or water? Not Sure ( ) 2.7 Do you think death is inevitable if a person Yes ( ) No ( ) bitten by a rabid animal develops signs of Not sure ( ) rabies? 2.8 Do you think a person bitten by a rabid animal Yes ( ) No ( ) can be saved from developing rabies? Not Sure ( ) 2.9 What is the most common treatment Apply local treatment like given to a patient bitten by a dog at your chilli powder/ turmeric () clinic? Apply antiseptic cream/ powder ( ) Wash with soap/ detergent () Apply antibiotics () Other, please specify, ( ) 2.10 Do you think the treatment of an animal bite Yes ( ) No ( ) wound with chilli/ turmeric powder is useful? Not sure ( ) 2.11 Do you think washing an animal bite wound Yes ( ) No ( ) with soap/ detergent and water is useful? If Not sure ( ) yes, proceed to 2.12, else to 2.13 2.12 When the wound is washed with soap/ < 2 Minutes ( ) 2 – 5Minutes ( ) detergent and water, how long do you think it 6 – 10 Minutes ( )11 – should take to complete the procedure? 15Minutes()>15 Minutes() 2.13 In your clinic would you suture a wound caused Yes ( ) No ( ) by a dog bite? Sometimes ( ) 2.14 Do you think it is important to observe a dog Yes ( ) No ( ) that has bitten someone? If no,proceed to Not sure ( ) 2.16
333
2.15 How many days do you think that the dog that has bitten a person or animal should be observed for? 2.16 Do you know of treatments that can be Yes ( ) No ( ) given to a person bitten by a rabid Not Sure ( ) animal to prevent rabies? If yes, proceed to next question, else thanks for your time to complete this questionnaire. 2.17 Are such treatments that can prevent Yes ( ) No ( ) rabies after an animal bite administered Not Sure ( ) at your clinic? If yes, proceed to next question, else to 2.19 2.18 What PEP (Post exposure Anti-rabies vaccine (ARV) RIG prophylaxis) is available at your (rabies immunoglobulins)
clinic? Both ( ) None ( )
2.19 Is anti-rabies vaccine (ARV) readily Yes ( ) No ( ) available from the medical stores around Not sure ( ) this area when required? 2.20 Is rabies immunoglobulin (RIG) readily Yes ( ) No ( ) available from the medical stores when Not sure ( ) required? 2.21 Are you aware of the schedule of ARV to Yes ( ) No ( ) follow? If no, Not sure ( ) please proceed to 2.23 2.22 What is the schedule of ARV you follow in your clinic? 2.23 Do you think RIG should be administered Yes ( ) No ( ) immediately after a person is bitten by an Not sure ( ) animal? 2.24 Do you think RIG can be administered up Yes ( ) No ( ) to 7 days after exposure? Not sure ( ) 2.25 In your opinion which is the most Non availability of ARV/ PEP/RIG ( important factor that results in ) failure to control rabies in humans? Lack of awareness among people ( ) Lack of control over stray dog population ( ) Others please specify ( )
Thank you for your time to complete this questionnaire. This information will help understand rabies and assist in its control in India.
334
Appendix III Consent form for the survey of community members
Project Title: A survey to determine the knowledge, attitudes and practices of people on dogs and dog-related rabies.
Harish Kumar Tiwari, a PhD student at Murdoch University is conducting a survey to determine the knowledge, attitudes and behavioural practices of people from different communities towards the dog population and its role in the spread of rabies. The study is designed to identify factors that may influence the occurrence of rabies in India and the level of awareness about the disease. This study is a collaborative study between the College of Veterinary Medicine, Murdoch University and Ashoka Trust for Research on Environment and the the Ecology.The questionnaire will take approximately 30 minutes to complete. The first part of the questionnaire covers general household information and the latter part deals with awareness about dogs and rabies. The study will help evaluate the existing strategies to control rabies and the dog population in India.
Your participation in this survey is greatly appreciated. All information collected from you will be kept strictly confidential and no information that may identify you will be used in any report or publication. You may withdraw from the questionnaire at any stage if you wish without any impact on you.
Consent:
Do you have any questions about this study? Yes No Do you understand the purpose of the study and your involvement? Yes No Would you like to participate in the study? Yes No
Please sign to record your consent to participate in this study ______OR This consent form with the requirements and conditions associated with this questionnaire survey has been read out to me in the language I understand, and I consent to participate in this survey. (Left Thumb Impression)
If you have any queries or concerns regarding this survey you can contact the Human ethics office at Murdoch University or email at [email protected]. Thank you for your assistance with this project.
Sincerely Harish Kumar Tiwari PhD Student School of Veterinary and Life Sciences Murdoch University South Street, Murdoch 6150 Western Australia Phone: 0426842710 Email: [email protected]
335
Appendix IV
Consent form for the Survey of Primary Health Staff
Project Title: A survey to determine the knowledge, attitudes and practices (KAP) of health workers employed with Primary Health Centres/ Clinics/ hospitals towards canine related rabies.
Harish Kumar Tiwari, a PhD student at Murdoch University is studying the knowledge, attitudes and practices of primary health workers towards free roaming dogs, dog associated rabies and the first aid measures adopted. This study is a collaborative study between the College of Veterinary Medicine, Murdoch University and Ashoka Trust for Research on Environment and the the Ecology. The main objective of the study is to ascertain the level of awareness among primary health officials who are generally the first point of contact for a dog-bite victim. It will help identify better control measures for dog related rabies in India and determine the role of the primary health workers in the fight against this disease. Your participation in this survey will be greatly appreciated. The survey will take about 30 minutes to complete. As part of the survey some questions will relate to information regarding your place of practice and professional details followed by questions relating to your knowledge, attitudes and practices towards rabies. All information collected from you will be kept strictly confidential. No information that may identify you will be used in any report or publication. Your participation is entirely voluntary. You may withdraw at any time from the questionnaire without any impact on you. Consent:
Do you have any questions about this study? Yes No Do you understand the purpose of the study and your involvement? Yes No Would you like to participate in the study? Yes No
Please sign so that we record your consent to participate in this study
If you have any queries or concerns regarding this survey you can contact Human ethics office at Murdoch University or email at [email protected].
Thank you for your assistance with this project. Sincerely Harish Kumar Tiwari PhD Student School of Veterinary and Life Sciences Murdoch University South Street, Murdoch 6150 Western Australia Phone: 0426842710 Email: [email protected]
336