Social Determinants of Unmet Hospitalisation
Total Page:16
File Type:pdf, Size:1020Kb
Nagulapalli, S 2013 Data from “Social determinants of unmet hospitalisation need amongst the poor in Andhra Pradesh, India: A cross-sectional study.” Journal of Open Public Health Data 1(1):e6, DOI: http://dx.doi.org/10.5334/jophd.af DATA PAPER Data from “Social determinants of unmet hospitalisation need amongst the poor in Andhra Pradesh, India: A cross- sectional study.” Srikant Nagulapalli1 2 1 Government of Andhra Pradesh, India 2 Collector and District Magistrate of Nellore, Andhra Pradesh, India The dataset is of a health survey amongst the 21.5 million poor families of the Indian state of Andhra Pradesh conducted during April and May 2013. The dataset captures individual characteristics and household characteristics of the past 365 days. Data was collected by 2022 trained field staff of Aarogyasri Health Care Trust (AHCT) of Government of Andhra Pradesh using a questionnaire mod- elled after that used for the health surveys by National Sample Survey Organisation of India. Keywords: health survey of Andhra Pradesh; Rajiv Aarogyasri; unmet hospitalisation need Funding statement The data is not the result of any funded project. (1) Overview (2) Methods Context Steps a) Design of questionnaire: National Sample Survey Spatial coverage Organisation (NSSO), Ministry of Statistics, Government Description: 23 districts Adilabad, Nizamabad, Karimna- of India conducts yearly consumer expenditure surveys as gar, Warangal, Hyderabad, Rangareddy, Medak, Mahbub- well as focussed surveys on health care. The health survey nagar, Nalgonda, Anantapur, Kurnool, Kadapa, Chittoor, questionnaire and the consumer expenditure question- Nellore, Prakasam, Guntur, Krishna, Khammam, West naire used for the 60th round (2004-05) of NSSO were Godavari, East Godavari, Visakhapatnam, Vizianagaram integrated by suitably abridging the consumer expendi- and Srikakulam of Andhra Pradesh state, India were cov- ture details and used for this survey. The questionnaire ered. for this survey collected the following information: the Boundaries of Andhra Pradesh state: identification of sample household, household character- • Northern boundary: Latitude 19.91671, Longitude istics, household member characteristics, consumption 78.315655 expenditure details of household with break-up of food • Southern boundary: Latitude 13.388455, Longitude and non-food items, chronically ailing household mem- 80.312420 bers, details and expenses of in-patient episodes, and • Eastern boundary: Latitude 17.59052, Longitude details and expenses of out-patient episodes. 83.254010 b) Pilot survey: Changes suggested by external experts • Western boundary: Latitude 14.614320, Longitude were carried out in the questionnaire. Two pilot surveys 76.750103 were conducted with the questionnaire in the districts of Temporal coverage Medak and Rangareddy and the questionnaire frozen. The 1 April 2012 - 31 May 2013 questionnaire was administered in Telugu language and all the responses were given codes based on the results of Species the pilots. Homo sapiens c) Training of surveyors and supervisors: The 2022 field staff (aarogyamitras) of Aarogyasri Health Care Trust (AHCT) of Government of Andhra Pradesh under the supervision of 97 team leaders conducted the survey. An intensive one week training program was given to the Art. e6, p. 2 of 4 Nagulapalli supervisors who in turn conducted a one week training to random samples from each PSU for the different strata. all the field staff under the supervision of 23 district man- The sampling from the RAS stratum was done from the agers. Training was given on how to administer the sur- AHCT data repository. In case the sampled household fell vey questionnaire following the methodology laid down into a stratum from which a household was already sur- by Goode and Hatt (2006)[1]. veyed, resampling was done till a household from the un- d) Conduct of survey: After obtaining an informed con- surveyed stratum was identified. One SSU or household sent, surveyors conducted the survey in the mornings was identified for survey from each of the strata. Three before the household head left for work. households (SSUs) one from each strata were surveyed e) Collection of filled up questionnaires: All the ques- from each of the selected village/ward (PSU). tionnaires were collected at a centralised location in Hyderabad by AHCT. Quality Control f) Design and development of Information Technology (IT) After the survey each questionnaire was scrutinised by application: An IT application was developed by AHCT for the supervisor. Resurveys of the same household were data entry of the questionnaire with required validations conducted by the supervisor along with the surveyor in for data entry. all cases where the surveyed form failed the inspection. g) Data entry: The data entry of all the questionnaires was A total of 6696 households were surveyed, with 2232 done by a trained team of 38 data entry personnel under 3 households from each stratum. A team of 38 trained data supervisors through the custom built IT application. entry persons entered the data in a custom built IT appli- h) Data checking: The entered data was sample checked cation. The primary data in the survey forms was entered for accuracy of data entry. manually while the rest of the data in the forms which involved simple arithmetic operations was auto populated Sampling Strategy by the IT system. Government of Andhra Pradesh launched a fully funded demand-side financed cashless inpatient health care Constraints scheme covering 21.5 million poor families in the state, A limitation of the survey was that it was confined to the with the objective of offering protection against cata- poor families which possessed a BPL card issued by GoAP. strophic health care expenditure, starting from April Though 87.7% of population were in possession of BPL 2007. This scheme, known as Rajiv Aarogyasri Scheme cards, there is a likelihood that some of the eligible poor and implemented by Aarogyasri Health Care Trust (AHCT) families were not in possession of the BPL cards and ineli- of Government of Andhra Pradesh, is the oldest and larg- gible APL families were in possession of BPL cards. est demand-side financed inpatient health care scheme in Privacy India. An informed consent was obtained by the surveyor before A cross-sectional study was done through a multi-stage starting the survey. Permission was given by Government stratified survey conducted during April and May 2013 in of Andhra Pradesh to the author for publication of the all the 23 districts of Andhra Pradesh. Urban and rural data[2]. We have de-identified and anonymised the data- mandals form the grass-root level administrative units set to the maximum extent possible[3]. Publication of the within a district. All the 1074 rural mandals and 268 dataset may not likely present any risk to confidentiality urban mandals in the state were surveyed. Villages in of study participants. rural mandals and wards in urban mandals were consid- The following steps were taken to protect privacy: ered as Primary Stage Units (PSU). Two PSUs were selected • Characteristics such as age, land possessed and family through simple random sampling with replacement from size were classified into intervals. each mandal in the state. The households within a village • The BPL card number which identifies the family was or ward formed the Second Stage Units (SSU). Three strata not captured in the questionnaire. Each family in the were formed for the SSUs viz., families who underwent dataset was given a running serial number. at least one in-patient treatment under Rajiv Aarogyasri • The village to which the family belongs was replaced by Scheme during the last 365 days (RAS stratum), families a serial number. who underwent at least one in-patient treatment but all • No provision was made to enter the names of family of them outside Rajiv Aarogyasri Scheme during the last members in the data entry screen of the IT application, 365 days (non-RAS stratum), and families who did not instead family members were assigned running serial have any member admitted in a hospital as an in-patient numbers in the dataset. during the last 365 days (Out patient stratum). The details such as age and family size will be provided on The SSUs in the three strata were selected through a case by case basis. Simple Random Sampling With Replacement (SRSWR) from the list of households. The Civil Supplies Depart- (3) Dataset Description ment of Government of Andhra Pradesh (GoAP) maintains a database of all households in the State which totalled Object Name 245,57,811 consisting of 29,83,106 Above Poverty Line • healthsurveyquestionnaire_english.doc contains the (APL) and 215,74,705 Below Poverty Line (BPL) families questionnaire used for the survey as on 1st March 2013. The BPL family list of Civil Sup- plies Department was utilised for the purpose of drawing Nagulapalli Art. e6, p. 3 of 4 • metadata_unmetneed1_1.xlsx contains references for from Government of Andhra Pradesh. This work was each variable name to the item number in the health done by SN for fulfilling the requirements of Ph.D. survey questionnaire degree at Andhra University, Visakhapatnam, India, and • data_unmetneed1_1.csv contains data under each vari- was not the asking of Government of Andhra Pradesh; able name • No financial support was received for this work from anyone other than Aarogyasri Health Care Trust; Data Type • No family members or friends of SN with commercial Primary data, processed data. entities have an interest in this