BMJ Open Is Committed to Open Peer Review. As Part of This Commitment We Make the Peer Review History of Every Article We Publish Publicly Available

BMJ Open: first published as 10.1136/bmjopen-2019-030823 on 31 March 2020. Downloaded from

BMJ Open is committed to open peer review. As part of this commitment we make the peer review history of every article we publish publicly available.

When an article is published we post the peer reviewers’ comments and the authors’ responses online. We also post the versions of the paper that were used during peer review. These are the versions that the peer review comments apply to.

The versions of the paper that follow are the versions that were submitted during the peer review process. They are not the versions of record or the final published versions. They should not be cited or distributed as the published version of this manuscript.

BMJ Open is an open access journal and the full, final, typeset and author-corrected version of record of the manuscript is available on our site with no access controls, subscription charges or pay-per-view fees (http://bmjopen.bmj.com).

If you have any questions on BMJ Open’s open peer review process please email [email protected] http://bmjopen.bmj.com/ on September 28, 2021 by guest. Protected copyright. BMJ Open BMJ Open: first published as 10.1136/bmjopen-2019-030823 on 31 March 2020. Downloaded from

Combating Antibiotic Resistance using Guidelines and Enhanced Stewardship in Kenya: An Implementation Science Approach ForJournal: peerBMJ Open review only Manuscript ID bmjopen-2019-030823

Article Type: Protocol

Date Submitted by the 04-Apr-2019 Author:

Complete List of Authors: Gitaka, Jesse; Mount Kenya University, Research and Innovation Mureithi, Dominic; Maasai Mara University, Department of Animal Health and Production Ndegwa, Davies; Kenya Medical Training College Masika, Moses; University of Nairobi College of Health Sciences, Department of Medical Microbiology, Omuse, Geoffrey; Aga Khan University - Kenya Ngari, Moses; KEMRI/Wellcome Trust, Clinical Trial Facility Makokha, Francis; Mount Kenya University, Research and Innovation Mwaura, Peter; Mount Kenya University, Research and Innovation Kamita, Moses; Mount Kenya University, Research and Innovation Mathai, Ronald; Mount Kenya University, Research and Innovation Muregi, Francis; Mount Kenya University, Research and Innovation http://bmjopen.bmj.com/ Mwau, Matilu; Kenya Medical Research Institute

Antimicrobial stewardship, Antimicrobial resistance, Implementation Keywords: science

on September 28, 2021 by guest. Protected copyright.

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 1 of 29 BMJ Open

1 2 3 4 BMJ Open: first published as 10.1136/bmjopen-2019-030823 on 31 March 2020. Downloaded from 5 6 7 8 9 I, the Submitting Author has the right to grant and does grant on behalf of all authors of the Work (as defined 10 in the below author licence), an exclusive licence and/or a non-exclusive licence for contributions from authors 11 who are: i) UK Crown employees; ii) where BMJ has agreed a CC-BY licence shall apply, and/or iii) in accordance 12 with the terms applicable for US Federal Government officers or employees acting as part of their official 13 duties; on a worldwide, perpetual, irrevocable, royalty-free basis to BMJ Publishing Group Ltd (“BMJ”) its 14 licensees and where the relevant Journal is co-owned by BMJ to the co-owners of the Journal, to publish the 15 Work in this journal and any other BMJ products and to exploit all rights, as set out in our licence. 16 17 The Submitting Author accepts and understands that any supply made under these terms is made by BMJ to 18 the Submitting Author Forunless you peer are acting as review an employee on behalf only of your employer or a postgraduate 19 student of an affiliated institution which is paying any applicable article publishing charge (“APC”) for Open 20 Access articles. Where the Submitting Author wishes to make the Work available on an Open Access basis (and 21 intends to pay the relevant APC), the terms of reuse of such Open Access shall be governed by a Creative 22 Commons licence – details of these licences and which Creative Commons licence will apply to this Work are set 23 out in our licence referred to above. 24 25 Other than as permitted in any relevant BMJ Author’s Self Archiving Policies, I confirm this Work has not been 26 accepted for publication elsewhere, is not being considered for publication elsewhere and does not duplicate 27 material already published. I confirm all authors consent to publication of this Work and authorise the granting 28 of this licence. 29 30 31 32 33 34 35 36 37 http://bmjopen.bmj.com/ 38 39 40 41 42 43 44 45 on September 28, 2021 by guest. Protected copyright. 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 2 of 29

1 2 3

Combating Antibiotic Resistance using Guidelines and Enhanced Stewardship in BMJ Open: first published as 10.1136/bmjopen-2019-030823 on 31 March 2020. Downloaded from 4 Kenya: An Implementation Science Approach 5 6 1 2 3 4 5 7 Jesse Gitaka , Dominic Mureithi , Davies Ndegwa , Moses Masika , Geoffrey Omuse , 6 1 1 1 1 8 Moses Ngari , Francis Makokha , Peter Mwaura , Moses Kamita , Ronald Mathai , Francis 9 Muregi1, Matilu Mwau7 10 11 Affiliations 12 1. Directorate of Research and Innovation, Mount Kenya University, P.O. Box 342-01000, 13 Thika, Kenya 14 2. Department of Animal Health and Production, Maasai Mara University, P.O. Box 861- 15 16 20500, Narok, Kenya. 17 3. Department of Medical Laboratory Sciences, Kenya Medical Training College P.O. Box 18 30195-00100, Nairobi,For Kenya peer review only 19 4. Department of Medical Microbiology, College of Health Sciences, University of Nairobi, 20 P.O. Box 30197-00100, Nairobi, Kenya. 21 5. Department of Pathology, Aga Khan University Hospital, P.O. Box 30270-00100, 22 23 Nairobi, Kenya 24 6. KEMRI/Wellcome Trust Research Programme, Centre for Geographic Medicine 25 Research - Coast, P.O. Box 230-80108, Kilifi, Kenya 26 7. Centre for Infectious and Parasitic Diseases Control Research, Kenya Medical Research 27 Institute, P.O. Box 3-50400, Busia, Kenya 28 29 Correspondence: Jesse Gitaka 30 31 [email protected] 32 Directorate of Research and Innovation, Mount Kenya University, P.O. Box 342-01000, 33 Thika, Kenya 34 35 Word count: 2708 words 36 37 http://bmjopen.bmj.com/ 38 39 40 41 42 43 44 45 on September 28, 2021 by guest. Protected copyright. 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 3 of 29 BMJ Open

1 2 3

Abstract BMJ Open: first published as 10.1136/bmjopen-2019-030823 on 31 March 2020. Downloaded from 4 5 Introduction 6 7 8 Antibiotic resistance (AMR) is a growing problem globally especially in Sub-Saharan Africa 9 including Kenya. Without any intervention, developing countries will be most affected due to 10 the high burden of diseases. Studies have consistently shown that inappropriate use of 11 antimicrobials is the major drivers of AMR. To address this challenge hospitals are now 12 implementing antibiotic stewardship programs (ASPs), which have been observed to reduced 13 antibiotic usage, decrease the prevalence of resistance and lead to significant economic 14 benefits. However, the implementation of the guideline is highly dependent on settings in 15 16 which they are rolled out. This study, employing an implementation science approach aims to 17 address the knowledge and research gap in this area and provide critical data and experiences 18 in using antibiotic guidelinesFor peer and stewardship review programs in onlythe public health sector. This will 19 provide evidence of ASP performance and potentially contribute to the county, national and 20 regional policies on antibiotics use. 21 22 Methods 23 24 25 The study will be conducted in three geographically diverse regions each represented by two 26 hospitals. A baseline study on antibiotic usage, resistance and de-escalation, duration of 27 hospital stay, rates of readmission and costs will be carried out in the pre-implementation 28 phase. The intervention, that is, the use of antibiotic guidelines and antibiotic stewardship 29 programs will be instituted for 18 months using a stepwise implementation strategy that will 30 facilitate learning and continuous improvement of stewardship activities and updating of 31 guidelines to reflect the evolving antibiotic needs. 32 33 34 Ethics and dissemination 35 36 The proposal has been approved by the Mount Kenya University Ethics Review Committee 37 and the National Commission for Science Technology and Innovation. Findings from this http://bmjopen.bmj.com/ 38 protocol will provide antibiotic guidelines informed by local bacteria susceptibility patterns, 39 the establishment of Antibiotic Stewardship Committees (ASCs) in 6 County hospitals in 40 Kenya, and data on antibiotics usage. 41 42 Key Words 43 44 45 Antimicrobial stewardship; Implementation science; Antimicrobial resistance on September 28, 2021 by guest. Protected copyright. 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 4 of 29

1 2 3

Article Summary BMJ Open: first published as 10.1136/bmjopen-2019-030823 on 31 March 2020. Downloaded from 4 5 This protocol describes a proposed approach to be applied in order to develop antibiotic 6 7 guidelines and antibiotic stewardship programs (ASPs) in different hospitals in Kenya. The 8 guidelines as well as the ASPs will be developed based on local studies to understand the 9 implementation challenges, derive lessons, understand the knowledge and research gap in this 10 area. 11 12 Strengths and limitations of this study 13 14  First study aimed at rolling out antimicrobial stewardship committees in multiple 15 hospitals in Kenya concurrently. 16 17  Use of implementation approach to support implement suggested guidelines for 18 antimicrobial resistanceFor surveillance.peer review only 19 20  First hand evidence on the antimicrobial resistance in three diverse counties in Kenya. 21 22  The study is limited to only three counties of the 47 counties in Kenya 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 http://bmjopen.bmj.com/ 38 39 40 41 42 43 44 45 on September 28, 2021 by guest. Protected copyright. 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 5 of 29 BMJ Open

1 2 3

Introduction BMJ Open: first published as 10.1136/bmjopen-2019-030823 on 31 March 2020. Downloaded from 4 5 Antibiotic resistance is a serious public health concern globally and locally and fears of running 6 1–3 7 out of antibiotics in the near future have been expressed. In 2016, the World Health 8 Organization (WHO), called for immediate and concerted efforts to mitigate this threat to 9 global health that was estimated to contribute to 700,000 deaths in 2014 and projected to cause 10 10 million deaths in 2050 if inadequately mitigated.4 The driving force for escalating rates of 11 resistance is the injudicious use of antibiotics in patients and in livestock or release into the 12 environment. These forces exert selective pressure for the rise and spread of resistant pathogens 13 that may emerge by mutations or acquisition of plasmids carrying resistance genes.5 In one 14 15 study, prior antibiotic exposure was the key independent risk factor for the acquisition of multi- 16 resistant.6 Broad-spectrum antibiotics have the unintended consequence of selecting multidrug- 17 resistant pathogens and increasing the likelihood of infection by fungi and Clostridium 18 difficile.7 For peer review only 19 20 Nonetheless, the injudicious use of antibiotics is not unusual. In Africa, many patients do not 21 receive treatment from the conventional health care system. Of those who receive antibiotics, 22 23 31.7 % of them do not consult a doctor for a prescription and a further 26.4% obtain the 8 24 antibiotics from over-counter. A study in South Africa found that 54.9% of antibiotics were 25 inappropriately prescribed in intensive care unit settings while in the US, 20-50% of prescribed 26 antibiotics are unnecessary or unwarranted.9–11 27 28 The rate of antimicrobial resistance in Kenya is worrying and rising. In one study, the 29 prevalence of Salmonella typhi resistant to two or more antimicrobials was observed to have 30 increased from 50% in 1998 to 78% in 2004 at Kenyatta National Hospital.12 The Global 31 32 Antibiotic Resistance Partnership – Kenya Working Group Report of 2011 identified antibiotic 33 resistance as a key issue in Kenya and made bespoke recommendations to curtail the spread. 34 These recommendations included the use of antibiotic guidelines that took into consideration 35 local resistance surveillance data and enhanced antibiotic stewardship programs (ASPs).13 36 Even so, these ASPs have not been instituted at county hospitals, and key implementation data http://bmjopen.bmj.com/ 37 and experiences are lacking in their roll out. 38 39 Antibiotic stewardship is defined as the optimum selection, dosage, and duration of 40 41 antimicrobial treatment that yields the best clinical outcomes for the treatment or prevention of 42 infection with the least toxicity to the patient and minimal impact on subsequent resistance.14 43 It has the potential to lower treatment costs and realize economic benefits to the patient, health 44 care system and the country at large.15,16 Moreover, optimizing antibiotic use by minimizing on September 28, 2021 by guest. Protected copyright. 45 exposure, fine-tuning dosage and reducing superfluous therapy and focusing treatment to the 46 likely culprit pathogens is a strategy that boosts patient safety17 and ultimately safeguards 47 48 against antibiotic resistance. 49 50 Justification 51 52 Antibiotic resistance is a major health challenge globally and concerted efforts have been called 53 for to limit the phenomena. Studies in Kenya have shown rising antibiotic resistance over the 54 last 3 decades. Nonetheless, antibiotic guidelines and antibiotic stewardship programs which 55 have been observed to lead to significant economic benefits, reduce antibiotic usage and lower 56 the prevalence of resistance especially in Europe, North America, Japan, and South Africa have 57 58 not been employed to tackle the challenge in the public health sector in Kenya. The GARP 59 report of 2011, recognized use of guidelines and stewardship programs as a potential strategy 60 in ‘saving antibiotics’ but noted the need for local studies to understand implementation

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 6 of 29

1 2 3

challenges, derive lessons and embed the strategy within the Kenyan health care system. This BMJ Open: first published as 10.1136/bmjopen-2019-030823 on 31 March 2020. Downloaded from 4 study, employing an implementation science approach aims to address the knowledge and 5 6 research gap in this area and provide critical data and experiences in using antibiotic guidelines 7 and stewardship programs in the public health sector. 8 9 Objectives of the Study 10 11 General Objectives 12 13 The overall aim of this project is to evaluate the impact of antibiotic guidelines for empirical 14 treatment of urinary tract infections, community-acquired pneumonia, bacteremia and 15 meningitis and antibiotic stewardship program on reducing usage of broad-spectrum 16 17 antibiotics, antibiotic de-escalation, duration of hospital stays, rates of readmission and 18 prevalence of antibioticFor resistance peer in 6 county review hospitals in onlyKenya. 19 20 Specific Objectives 21 22 1. To develop guidelines for antibiotic use for common infections. 23 2. To set up antibiotic steward committees (ASCs) in 6 county hospitals. 24 3. To train the ASCs to be able to perform their mandate. 25 26 4. To measure the usage of broad-spectrum antibiotics, antibiotic de-escalation, duration 27 of hospital stays and rates of readmission in hospitals using the antibiotic stewardship 28 strategy. 29 5. To ascertain antibiotic resistance patterns using culture, sensitivity, and genetic 30 markers. 31 6. To establish the health care workers Knowledge, Attitudes and prescription Practices 32 regarding antibiotics resistance, use of guidelines and ASPs. 33 34 7. To evaluate the economic benefits of using guidelines for antibiotics use and ASPs. 35 36 Expected Outputs of the research 37 http://bmjopen.bmj.com/ 38 1. Antibiotic guidelines informed by local bacteria susceptibility patterns. 39 2. Antibiotic Stewardship Committees (ASCs) in 6 County hospitals in Kenya. 40 3. Trained ASCs 41 4. Data on antibiotics usage, antibiotic de-escalation, duration of hospital stay and rates 42 of readmission in hospitals using the antibiotic stewardship strategy 43 44 5. Map and data of antibiotic resistance pattern before and after implementation of the 45 strategy. on September 28, 2021 by guest. Protected copyright. 46 6. Qualitative data on health care workers’ knowledge, attitudes, and practices on the 47 antibiotic usage and stewardship programs 48 7. Publication on cost-benefit analysis for using antibiotic guidelines and ASPs 49 50 Methodology 51 52 53 Setting 54 55 To be able to evaluate the antibiotic stewardship strategy across Kenya using an 56 implementation science approach, the study will be conducted in 6 county hospitals (Kiambu 57 County Referral Hospital, Bungoma County Hospital, Webuye Sub-County Hospital, Nakuru 58 County Teaching and Referral Hospital, Thika Level 5 Hospital and Naivasha Sub-County 59 Hospital). 60

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 7 of 29 BMJ Open

1 2 3

Design BMJ Open: first published as 10.1136/bmjopen-2019-030823 on 31 March 2020. Downloaded from 4 5 The study design will encompass a pre-implementation phase, a stepwise implementation 6 7 phase, and an endline study to measure the changes in outcomes between the phases. A baseline 8 study on antibiotic usage (Defined Daily Doses per 1000), antibiotic resistance (culture and 9 genetic markers), antibiotic de-escalation, duration of hospital stays, rates of readmission, 10 prescription patterns, and costs analysis will be carried out in the pre-implementation phase. 11 The intervention, which is the use of antibiotic guidelines and antibiotic stewardship programs 12 will be instituted in the 6 county hospitals for 18 months. The stepwise implementation strategy 13 will facilitate learning cycles every 4-6 weeks and continuous improvement of stewardship 14 15 activities and updating of guidelines to reflect evolving antibiotic needs in diverse settings. The 16 end line study will be conducted and differences between the 2 phases evaluated as below 17 (Statistical analysis). 18 For peer review only 19 In the baseline and endline studies, multidisciplinary strategies will be employed as 20 follows 21 i. Health care workers knowledge, attitudes, and practices on antibiotic resistance, 22 23 guidelines, and ASP will be studied qualitatively and quantitatively. 24 ii. Basic science approaches encompassing antibiotics culture sensitivity and molecular 25 biology-genetic markers of resistance will be analyzed as detailed below. 26 iii. Clinical- patient outcomes will be studied to evaluate the guidelines. 27 iv. Health economics-cost savings on using guidelines and ASP will be evaluated as below. 28 29 Prevalence of inappropriate use of broad-spectrum antibiotics 30 31 The sample size is based on the prevalence of inappropriate use of broad-spectrum 32 9 33 antibiotics of ~50% in South Africa. To detect a reduction of the 50% inappropriate use of 34 broad-spectrum antibiotics by 20% to 30% with a power of 90% and α of 0.05 in a two-sided 35 test, a sample size of 410 in each county hospital would be adequate. In order to detect a 36 reduction (20%) but with a power 80% the sample size would be 320 as shown in Figure 1 37 below. http://bmjopen.bmj.com/ 38 With a 15% allowance for loss to follow up, a total of about 500 participants would be enough 39 in each of the selected county hospitals. Therefore, we shall evaluate the prevalence 40 41 of inappropriate use of broad-spectrum antibiotics in 500 x 6= 3,000 patients for the 6 facilities. 42 43 Prevalence of antibiotic resistance 44 45 The sample size is based on the prevalence of antibiotic resistance of 78%12 in 2004 at Kenyatta on September 28, 2021 by guest. Protected copyright. 46 National Hospital, Kenya. To detect a reduction of the 78% antibiotic resistance by 10% to 47 68% with a power of 90% and α of 0.05 two-sided test, a sample size of 500 in each county 48 hospital would be adequate. While to achieve a reduction (10%) but with power, 80% the 49 50 sample size would be 220 as shown in Figure 2 below. 51 52 With 15% allowance for loss to follow up, a total of 600 participants would be enough in each 53 of the selected county hospitals. Therefore, we shall evaluate the prevalence of resistance in 54 600 x 6= 3,600 patients for the 6 facilities. 55 56 Antibiotic guidelines and ASPs: Development 57 58 Antibiotic guidelines will be formulated in consultation with senior clinicians in the study 59 hospitals taking into account each hospital’s antimicrobial resistance patterns. ASPs committee 60

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 8 of 29

1 2 3

will comprise the hospital physician, microbiologist, and pharmacist. Broad spectrum BMJ Open: first published as 10.1136/bmjopen-2019-030823 on 31 March 2020. Downloaded from 4 antibiotic prescriptions will be brought to the attention of the ASP committees who will also 5 6 perform regular ward rounds 3 times a week in the initial stages and later once a week to 7 optimize adherence to antibiotic guidelines. Furthermore, the guidelines will be promoted 8 through teaching sessions, provision of pocket-size guideline cards to clinicians and 9 pharmacists, large poster displays in the wards and through hospital and project websites. 10 11 Data collection 12 13 Laboratory assessment tools will be used to determine the preparedness of the laboratories to 14 perform antibiotic sensitivity tests (see Additional file 1). In addition, knowledge, attitude and 15 16 practices (KAP) about antibiotic prescribing and resistance among medical practitioners will 17 be assessed using a KAP tool (See Additional file 2). A health system assessment tool will be 18 used to gather baselineFor information peer on the review antimicrobial stewardship only activities in each hospital 19 (see Additional file 3). Information on each hospital bed occupancy, antibiotic usage, and 20 antibiotic resistance data will be collected from hospital information systems, pharmacy 21 management systems and laboratory reporting systems respectively. Antibiotic usage will 22 23 analyze the impact of the intervention on antibiotic usage comparing the intervention and 24 control arms. Data on antibiotics issued to adult and children inpatients only will be factored 25 excluding discharge and outpatient supplies. 26 27 To consistently compare antibiotic usage, the defined daily doses (DDD) will be used and 28 expressed per 1000 occupied bed days (OBDs) to account for fluctuations in activity. The 29 OBDs will be obtained from the hospital information systems. 30 31 Culture sensitivity and genetic analyses 32 33 34 Nasal swabs or tracheal aspirate, urine, wound swabs and blood samples will be taken from 35 patients who have consented to the study for bacteriological analysis. Samples will be subjected 36 to standard bacteriological analysis to isolate the culprit bacteria. Bacterial species will be 37 confirmed by use biochemical test and analytical profiles index API strips (Bio Merieux http://bmjopen.bmj.com/ 38 France). Antimicrobial susceptibility test will be performed on isolated bacteria as per the 39 Kirby-Bauer Method following manufacturer’s instruction. Results will be interpreted using 40 the Clinical and Laboratory Standards Institute (CLSI) tables.18 41 42 43 Any bacteria isolate found to be resistant to third generation cephalosporin’s will be tested for 44 production of extended spectrum Beta-lactamase (ESBLs) using the synergy disk diffusion 45 test. Vancomycin Resistance Enterococci (VRE) will be identified using disc diffusion tests. on September 28, 2021 by guest. Protected copyright. 46 Methicillin resistance in S. aureus (MRSA) will be detected by testing isolates resistant to 47 cefoxitin by E test (AB Biodisk, Solna, Sweden) on Mueller–Hinton agar supplemented with 48 2% NaCl and incubated at 37°C for 24 h. The identified ESBLs, VRE and MRSA will be 49 50 analyzed by PCR and sequencing to further identify the resistance genotype. In 51 vitro conjugation tests will be performed to determine if resistance in bacteria is transferable. 52 53 Cost-benefit analysis for the use of antibiotic guidelines and Antibiotic Stewardship 54 Program 55 56 A cost-benefit analysis (CBA) from a health facility and a national perspective will be 57 performed. For health facilities, three cost drivers will be considered: pharmacy spending, 58 length of stay, and antimicrobial stewardship interventions (training, infection control 59 60 measures, etc.). For the country, we will consider, Disability-Adjusted Life years (DALYS),

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 9 of 29 BMJ Open

1 2 3 19–21

work-days lost, and cost of treatment. This will be done by collecting and analyzing data BMJ Open: first published as 10.1136/bmjopen-2019-030823 on 31 March 2020. Downloaded from 4 on patient income, length of hospital stay, death or disability occasioned by drug-resistant 5 6 pathogens, hospital pharmacy expenditure and cost of training/rolling out antimicrobial 7 stewardship guidelines. This data will be collected before and after antimicrobial stewardship 8 interventions. 9 10 Data Management 11 12 Data will be recorded on a standardized Case Report Form (CRF) in use at every hospital by 13 the trained study staff. The data will be held locally and uploaded to the secure central study 14 server hosted at the Mount Kenya University main campus in Thika. This will be overseen by 15 16 the Data Manager/statistician who will run regular reconciliation to derive the final study 17 database. Access to the study database will be restricted and password protected. 18 For peer review only 19 Statistical analyses 20 21 The data will be analyzed using qualitative and quantitative methods. Qualitative data will be 22 analyzed by subjecting the information to content analysis and presenting it in different 23 emerging themes. The summaries of the data emanating from these themes will then be 24 22 25 arranged on a case by case basis through the use of an Excel spreadsheet and the analyses 26 done by using NVIVO software. The quantitative data analyses will be done using Stata version 27 14 (Stata, Inc.). The differences between baseline and endline on all the study outcomes will 28 be compared using appropriate statistical methods like McNemar’s test, paired t-test and the 29 Wilcoxon signed-rank test non-parametric test accounting for pairing and clusters (hospitals). 30 Multivariable log-binomial regression analysis will be used to get risk factors for antibiotic 31 resistance. 32 33 34 Ethical considerations 35 36 The proposal has been approved by the Mount Kenya University Ethics Review Committee 37 (MKU/ERC/0764) and National Commission for Science Technology and Innovation http://bmjopen.bmj.com/ 38 (NACOSTI/P/18/33304/25986). The study will follow all provisions of the Declaration of 39 Helsinki. Participants in the study will not incur any cost in the transport nor processing of the 40 samples, neither will they receive any monetary inducements to participate in the study. 41 Material transfer to laboratories outside of Kenya shall not be undertaken in this study. 42 43 Informed written consent will be sought from the participants enrolled in the study. 44 45 Patient and Public Involvement on September 28, 2021 by guest. Protected copyright. 46 47 Patients and the public were not involved in designing this protocol. 48 49 Exit strategy and stakeholder involvement 50 51 In the process of developing this protocol, we have engaged physicians working in the proposed 52 County hospitals and they have identified the challenge of antibiotic resistance as real and 53 54 appreciate the opportunities that use of antibiotic guidelines and ASPs may provide in 55 combating resistance, improving clinical care and saving costs. We plan to continue involving 56 all the stakeholders in the process who include clinicians, laboratory personnel, pharmacists, 57 public health officials, patients and scientists in the arena. We anticipate that the study findings 58 will inform county and national policy on mitigating antibiotic resistance and raise public 59 awareness on the need for judicious use of antibiotics. The project will use social media 60

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 10 of 29

1 2 3

platforms, websites of the collaborating institutions, and publications in peer-reviewed BMJ Open: first published as 10.1136/bmjopen-2019-030823 on 31 March 2020. Downloaded from 4 journals, local dailies and presentations in scientific meetings to further engage stakeholders 5 6 and the public on this important issue and enhance the learning approach inherent in the 7 strategy of implementation science for improved performance of the ASPs and the study in 8 general. 9 10 Dissemination 11 12 Obtained results will be disseminated at different platforms which include research conferences 13 and in peer-reviewed journals. 14 15 Discussion 16 17 18 The emergence of antimicrobial-resistantFor peer reviewpathogens and a lackonly of new drugs to effectively treat 19 these pathogens are the two main challenges in human health. There is thus the need to advocate 20 for proper use of the currently available antimicrobial agents by safeguarding their 21 effectiveness. Antibiotic stewardship has been shown to contribute to reducing antibiotic 22 resistance 23,24 but this strategy has not been rolled out in most sub-Saharan countries in level 23 4 and 5 hospitals. The proposed work will employ an implementation research approach to 24 25 evaluate the best strategies and derive lessons on mainstreaming antibiotic stewardship in these 26 facilities. By leveraging on the health system approach, the implementation research will 27 unmask real life impediments and opportunities and working with hospital teams co-design 28 execution plans, monitoring and evaluation and sustainability of the stewardship programs. 29 30 Competing Interests 31 32 The authors declare no conflict of interest. 33 34 35 Grant Information 36 37 This protocol is supported by the Government of Kenya through the National Research Fund http://bmjopen.bmj.com/ 38 (NRF) grant to JG. The funder did not contribute to the design and decision to publish this 39 protocol. 40 41 Availability of data and material 42 43 All data sets will be available to the public upon request. 44 45 Author Contributions on September 28, 2021 by guest. Protected copyright. 46 47 JG conceived the project, JG , DM, DN, MM, GO, MN, FM, PM, RM, FM, and MM 48 developed the protocol, MK prepared the protocol for publication, all authors reviewed and 49 approved the protocol. 50 51 Acknowledgment 52 53 Not applicable 54 55 Additional files 56 57 Additional file 1: Laboratory preparedness assessment tool 58 59 Additional file 2: Laboratory knowledge, attitude and practice (KAP) tool 60

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 11 of 29 BMJ Open

1 2 3

Additional file 3: Health system tool BMJ Open: first published as 10.1136/bmjopen-2019-030823 on 31 March 2020. Downloaded from 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 For peer review only 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 http://bmjopen.bmj.com/ 38 39 40 41 42 43 44 45 on September 28, 2021 by guest. Protected copyright. 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 12 of 29

1 2 3

Figure Legend BMJ Open: first published as 10.1136/bmjopen-2019-030823 on 31 March 2020. Downloaded from 4 5 6 Figure 1: Statistical power analysis for sample size determination for the prevalence of 7 8 inappropriate use of broad-spectrum antibiotics 9 10 11 Figure 2: Statistical power analysis for sample size determination for the prevalence of 12 antibiotics resistance 13 14 15 16 17 18 For peer review only 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 http://bmjopen.bmj.com/ 38 39 40 41 42 43 44 45 on September 28, 2021 by guest. Protected copyright. 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 13 of 29 BMJ Open

1 2 3

References BMJ Open: first published as 10.1136/bmjopen-2019-030823 on 31 March 2020. Downloaded from 4 5 1. Bejon, P. et al. Invasive Gram-negative bacilli are frequently resistant to standard 6 7 antibiotics for children admitted to hospital in Kilifi, Kenya. J. Antimicrob. 8 Chemother. 56, 232–235 (2005). 9 10 2. Livermore, D. M. Has the era of untreatable infections arrived? J. Antimicrob. 11 Chemother. 64, 29–36 (2009). 12 13 3. Dougan, G. et al. Typhoid in Kenya Is Associated with a Dominant Multidrug- 14 Resistant Salmonella enterica Serovar Typhi Haplotype That Is Also Widespread in 15 Southeast Asia. J. Clin. Microbiol. 48, 2171–2176 (2010). 16 17 18 4. WHO. UnitedFor Nations peer meeting on reviewantimicrobial resistance only. (2016). 19 20 5. Baquero, F. et al. The global threat of antimicrobial resistance: science for 21 intervention. New Microbes New Infect. 6, 22–29 (2015). 22 23 6. Sanchez-Ortega, I. et al. Bacteraemia due to multidrug-resistant Gram-negative bacilli 24 in cancer patients: risk factors, antibiotic therapy and outcomes. J. Antimicrob. 25 Chemother. 66, 657–663 (2010). 26 27 28 7. Gyssens, I. C., Kern, W. V. & Livermore, D. M. The role of antibiotic stewardship in 29 limiting antibacterial resistance among hematology patients. Haematologica 98, 1821– 30 1825 (2013). 31 32 8. Vialle-Valentin, C. E., Lecates, R. F., Zhang, F., Desta, A. T. & Ross-Degnan, D. 33 Predictors of antibiotic use in African communities: Evidence from medicines 34 household surveys in five countries. Trop. Med. Int. Heal. 17, 211–222 (2012). 35 36

9. Paruk, F. et al. Antibiotic prescription practices and their relationship to outcome in http://bmjopen.bmj.com/ 37 38 South Africa: findings of the prevalence of infection in South African intensive care 39 units (PISA) study. S. Afr. Med. J. 102, 613–6 (2012). 40 41 10. Ingram, P. R., Seet, J. M., Budgeon, C. A. & Murray, R. Point-prevalence study of 42 inappropriate antibiotic use at a tertiary Australian hospital. Intern. Med. J. 42, 719– 43 721 (2012). 44 on September 28, 2021 by guest. Protected copyright. 45 11. Fridkin, S. K. et al. Vital Signs: Improving Antibiotic Use Among Hospitalized 46 Patients. MMWR. Morb. Mortal. Wkly. Rep. (2014). 47 48 49 12. Kariuki, S. et al. Increasing prevalence of multidrug-resistant non-typhoidal 50 salmonellae, Kenya, 1994–2003. Int. J. Antimicrob. Agents 25, 38–43 (2005). 51 52 13. Kariuki, S. Situational analysis and recommendations on antimicrobial use and 53 resistance in Kenya. Heal. (San Fr. (2011). 54 55 14. Gerding, D. N. The search for good antimicrobial stewardship. Jt. Comm. J. Qual. 56 Improv. 27, 403–404 (2001). 57 58 59 15. Drummond, M. & Jönsson, B. Moving beyond the drug budget silo mentality in 60 Europe. Value Heal. 6, 74–77 (2003).

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 14 of 29

1 2 3

16. Richards, P. G. et al. Clinical and economic outcomes from a community hospital’s BMJ Open: first published as 10.1136/bmjopen-2019-030823 on 31 March 2020. Downloaded from 4 antimicrobial stewardship program. Am. J. Infect. Control 41, 145–148 (2012). 5 6 7 17. Dellit, T. H. et al. Infectious Diseases Society of America and the Society for 8 Healthcare Epidemiology of America Guidelines for Developing an Institutional 9 Program to Enhance Antimicrobial Stewardship. Clin. Infect. Dis. 44, 159–177 (2007). 10 11 18. CLSI. Performance Standards for Antimicrobial Susceptibility Testing. (2019). 12 13 19. Karanika, S., Paudel, S., Grigoras, C., Kalbasi, A. & Mylonakis, E. Systematic review 14 and meta-analysis of clinical and economic outcomes from the implementation of 15 hospital-based antimicrobial stewardship programs. Antimicrob. Agents Chemother. 16 17 60, 4840–4852 (2016). 18 For peer review only 19 20. Borde, J. P. et al. Implementing an intensified antibiotic stewardship programme 20 targeting daptomycin use in orthopaedic surgery: a cost–benefit analysis from the 21 hospital perspective. Infection 44, 301–307 (2016). 22 23 21. WHO. WHO | Metrics: Disability-Adjusted Life Year (DALY). WHO (2014). 24 Available at: https://www.who.int/healthinfo/global_burden_disease/metrics_daly/en/. 25 26 (Accessed: 25th February 2019) 27 28 22. Duane, S. et al. Using qualitative insights to change practice: Exploring the culture of 29 antibiotic prescribing and consumption for urinary tract infections. BMJ Open 6, 30 (2016). 31 32 23. Kauffmann, R. M. et al. Infection Reduction Strategies Including Antibiotic 33 Stewardship Protocols in Surgical and Trauma Intensive Care Units Are Associated 34 with Reduced Resistant Gram-Negative Healthcare-Associated Infections. Surg. Infect. 35 36 (Larchmt). 12, 15–25 (2010). 37 http://bmjopen.bmj.com/ 38 24. Yong, M. K., Buising, K. L., Cheng, A. C. & Thursky, K. A. Improved susceptibility 39 of Gram-negative bacteria in an intensive care unit following implementation of a 40 computerized antibiotic decision support system. J. Antimicrob. Chemother. 65, 1062– 41 1069 (2010). 42 43 44 45 on September 28, 2021 by guest. Protected copyright. 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 15 of 29 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2019-030823 on 31 March 2020. Downloaded from

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 For peer review only 17 18 19 20 21 22 23 24 25 26 27 28 29 30 Figure 1: Statistical power analysis for sample size determination for the prevalence of inappropriate use of 31 broad-spectrum antibiotics 32 http://bmjopen.bmj.com/ 33 226x164mm (96 x 96 DPI) 34 35 36 37 38 39 40 on September 28, 2021 by guest. Protected copyright. 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 16 of 29 BMJ Open: first published as 10.1136/bmjopen-2019-030823 on 31 March 2020. Downloaded from

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 For peer review only 17 18 19 20 21 22 23 24 25 26 27 28 29 30 Figure 2: Statistical power analysis for sample size determination for the prevalence of antibiotics resistance 31 32 225x163mm (96 x 96 DPI) http://bmjopen.bmj.com/ 33 34 35 36 37 38 39 40 on September 28, 2021 by guest. Protected copyright. 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 17 of 29 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2019-030823 on 31 March 2020. Downloaded from

1 2 3 4 A survey on Knowledge, Attitudes and Practice about antibiotic prescribing and resistance among 5 6 medical practitioners in Kenyan local hospitals. 7 8 Thank you very much for accepting to participate in this study. 9 10 *You are kindly requested to answer the questionnaire honestly and completely independent of 11 cross-consultations and/or verifications. 12 13 14 Survey quality control 15 16 Date of interview: ……………………………………………….For peer review Start time...... only End time...... 17 18 19 Interviewed by: ...... Approved...... 20 21 Name of the Hospital…………………………………………… Respondent’s code……………………………………………………. 22 23 24 QUESTIONS ANSWERS 25 PART 1: GENERAL QUESTIONS 26 27 1. For how many years, since you graduated from medical school ❖ I am on attachment 28 ❖ I am a trainee in medicine (internship) /medical training College, have you been working in a hospital 29 ❖ Less than one year 30 (indicate cumulative years if worked in different hospitals) ❖ 1-3 years 31 ❖ 32 4 – 6 years http://bmjopen.bmj.com/ 33 ❖ 7 years and more 34 2. In which department do you work? o Medicine /Emergency 35 o Surgery 36 37 o Paediatrics 38 39 o Obstetrics and Gynaecology 40 o on September 28, 2021 by guest. Protected copyright. 41 Outpatient/A/E 42 o Pharmacy 43 44 o Other: ………………………………………………… 45 46 3. Designation (e.g. Consultant, Pharmacist, Nurse, etc.) 47 …………………………………………………………………… 48 PART 2: PRESCRIPTION PATTERN (PRACTICE) 49 50 4. How frequently do you prescribe antibiotics? ❖ More than once daily 51 ❖ Once daily 52 53 ❖ 3 – 5 times a week 54 55 ❖ 1 – 2 times a week 56 57 58 Page 1 of 7 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 18 of 29 BMJ Open: first published as 10.1136/bmjopen-2019-030823 on 31 March 2020. Downloaded from

1 2 3 ❖ less than once a week) 4 5 5. To whom do you prescribe? ❖ Patients at outpatient department 6 7 ❖ Hospitalized patients 8 ❖ Patients in out-patient department and 9 10 hospitalised patients 11 12 6. Do you follow any antibiotic prescription guidelines? ❖ Yes 13 ❖ No 14 15 PART 3: AWARENESS AND ATTITUDE ON THE CURRENT SCOPE OF ANTIBIOTIC RESISTANCE 16 For peer review only 17 7. Antibiotic resistance is a world-wide problem ❖ I strongly agree 18 19 ❖ I agree 20 ❖ Neutral 21 22 ❖ I disagree 23 24 ❖ I strongly disagree 25 8. Antibiotic resistance is a problem in my country ❖ I strongly agree 26 27 ❖ I agree 28 29 ❖ Neutral 30 ❖ I disagree 31 32 ❖ I strongly disagree http://bmjopen.bmj.com/ 33 34 9. Antibiotic resistance is a problem in my hospital ❖ I strongly agree 35 ❖ I agree 36 37 ❖ Neutral 38 39 ❖ I disagree 40 ❖ I strongly disagree on September 28, 2021 by guest. Protected copyright. 41 42 10. Antibiotics are overused in my hospital and in other hospitals ❖ I strongly agree 43 44 of my country Kenya ❖ I agree 45 46 ❖ Neutral 47 ❖ I disagree 48 49 ❖ I strongly disagree 50 51 11. Patients’ demands for antibiotics contribute to the overuse of ❖ I strongly agree 52 antibiotics in the hospital ❖ I agree 53 54 ❖ Neutral 55 56 57 58 Page 2 of 7 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 19 of 29 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2019-030823 on 31 March 2020. Downloaded from

1 2 3 ❖ I disagree 4 5 ❖ I strongly disagree 6 7 12. I think over-the-counter (OTC) medicines contribute to ❖ I strongly agree 8 antibiotic misuse and subsequent antibiotic resistance ❖ I agree 9 10 ❖ Neutral 11 12 ❖ I disagree 13 ❖ I strongly disagree 14 15 13. My awareness on local antibiotic resistance pattern is? ❖ Excellent 16 For peer review only 17 ❖ Good 18 ❖ Average 19 20 ❖ Very little 21 22 ❖ None 23 24 PART 4: CHOICE OF ANTIBIOTIC 25 14. How confident are you about your knowledge of antibiotics? ❖ Very confident 26 27 ❖ Confident 28 29 ❖ A bit confident 30 ❖ Neutral/ I have no idea 31 32 ❖ Not confident at all http://bmjopen.bmj.com/ 33 34 15. What is your confidence level in prescribing antibiotics ❖ Very confident 35 ❖ Confident 36 37 ❖ A bit confident 38 39 ❖ Neutral/ I have no idea 40 ❖ Not confident at all on September 28, 2021 by guest. Protected copyright. 41 42 16. How often do you check your decisions on antibiotic ❖ Never 43 44 prescribing with a colleague? ❖ Sometimes 45 ❖ 46 Half of the times 47 ❖ Mostly 48 49 ❖ Always 50 51 17. If you do consult a senior colleague, how frequent does he/she ❖ Never 52 recommend prescription of a different antibiotic? ❖ Sometimes 53 54 ❖ Half of the times 55 56 57 58 Page 3 of 7 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 20 of 29 BMJ Open: first published as 10.1136/bmjopen-2019-030823 on 31 March 2020. Downloaded from

1 2 3 ❖ Mostly 4 5 ❖ Always 6 7 18. How often do you depend on antibiotic sensitivity data from ❖ Never 8 the laboratory to vary your prescription ❖ Sometimes 9 10 ❖ Half of the times 11 12 ❖ Mostly 13 ❖ Always 14 15 PART 5: SOURCE OF INFORMATION ON ANTIBIOTICS PRESCRIBING AND RESISTANCE 16 For peer review only 17 19. During the past years, how many courses or trainings did you ❖ 0 18 ❖ 19 receive relating to antibiotics? 1-3 20 ❖ 4-6 21 22 ❖ 6-10 23 24 ❖ >10 25 20. Among the sources of information about antibiotics listed below, which one did you consult in the last month? 26 27 ▪ Information supplied by pharmaceutical companies ❖ Yes 28 29 ❖ No 30 31 ▪ Knowledge from training institutions ❖ Yes 32 http://bmjopen.bmj.com/ 33 34 ❖ No 35 ▪ Internet ❖ Yes 36 37 38 ❖ No 39 40 ▪ National guideline for empiric antimicrobial therapy ❖ Yes on September 28, 2021 by guest. Protected copyright. 41 42 ❖ No 43 44 ▪ The World Health Organization’s (WHO) guidelines for ❖ Yes 45 46 treatment of bacterial diseases ❖ No 47 48 21. How do you appreciate the sources of information about antibiotics listed below? 49 50 ▪ Information supplied by pharmaceutical companies ❖ Very useful 51 52 ❖ Useful 53 ❖ Not at all useful 54 55 ❖ I do not know 56 57 58 Page 4 of 7 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 21 of 29 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2019-030823 on 31 March 2020. Downloaded from

1 2 3 ▪ Information from University courses ❖ Very useful 4 5 ❖ Useful 6 7 ❖ Not at all useful 8 ❖ I do not know 9 10 ▪ Internet ❖ Very useful 11 12 ❖ Useful 13 ❖ Not at all useful 14 15 ❖ I do not know 16 For peer review only 17 ▪ National guideline for empiric antimicrobial therapy ❖ Very useful 18 ❖ Useful 19 20 ❖ Not at all useful 21 22 ❖ I do not know 23 24 ▪ The World Health Organization’s (WHO) guidelines for ❖ Very useful 25 treatment of bacterial diseases ❖ Useful 26 27 ❖ Not at all useful 28 29 ❖ I do not know 30 ▪ Does your facility have a frequently released antibiogram? ❖ Yes 31 32 ❖ No http://bmjopen.bmj.com/ 33 34 ▪ If yes, how useful is the antibiogram to you ❖ Very useful 35 ❖ Useful 36 37 ❖ Not at all useful 38 39 ❖ I do not know 40 PART 6: DECISION ABOUT ANTIBIOTIC PRESCRIBING on September 28, 2021 by guest. Protected copyright. 41 42 22. When one prescribes an antibiotic, it is important to know the ❖ I strongly agree 43 44 resistance pattern of the bacteria in the local setting ❖ I agree 45 46 ❖ Neutral 47 ❖ I disagree 48 49 ❖ I strongly disagree 50 51 23. My choice of prescribing antibiotic is more influenced by the ❖ I strongly agree 52 availability of antibiotics than by the cause of the infection ❖ I agree 53 54 ❖ Neutral 55 56 57 58 Page 5 of 7 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 22 of 29 BMJ Open: first published as 10.1136/bmjopen-2019-030823 on 31 March 2020. Downloaded from

1 2 3 ❖ I disagree 4 5 ❖ I strongly disagree 6 7 24. My choice of prescribing antibiotic is more influenced by the ❖ I strongly agree 8 cost of the drug to the patient ❖ I agree 9 10 ❖ Neutral 11 12 ❖ I disagree 13 ❖ I strongly disagree 14 15 25. I’m always concerned about effectiveness and quality of an ❖ I strongly agree 16 For peer review only 17 antibiotic when making my prescribing decisions ❖ I agree 18 ❖ Neutral 19 20 ❖ I disagree 21 22 ❖ I strongly disagree 23 24 26. In regard to antibiotic guidelines, local guidelines are more ❖ I strongly agree 25 useful than international guidelines ❖ I agree 26 27 ❖ Neutral 28 29 ❖ I disagree 30 ❖ I strongly disagree 31 32 27. Antibiotic guidelines and antibiotic committees are rather ❖ I strongly agree http://bmjopen.bmj.com/ 33 34 obstacles than a help ❖ I agree 35 ❖ Neutral 36 37 ❖ I disagree 38 39 ❖ I strongly disagree 40 28. I welcome the implementation of a training program about ❖ I strongly agree on September 28, 2021 by guest. Protected copyright. 41 42 antibiotics ❖ I agree 43 44 ❖ Neutral 45 ❖ I disagree 46 47 ❖ I strongly disagree 48 49 PART 7: KNOWLEDGE ON USE OF ANTIBIOTICS 50 51 29. A 4-year-old child had diarrhoea in the last 4 days (3 stools ❖ Amoxicillin orally 52 daily). She had no fever during the past days nor at ❖ Trimethoprim/sulphamethoxazole orally 53 54 consultation. What is your treatment choice? ❖ Amoxicillin/clavulanic acid orally 55 56 57 58 Page 6 of 7 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 23 of 29 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2019-030823 on 31 March 2020. Downloaded from

1 2 3 ❖ Oral rehydration salts with no antibiotic 4 5 30. A 6-year-old child has fever (38°C), nasal discharge and a ❖ Trimethoprim/sulphamethoxazole orally 6 7 painful throat for two days. At visual inspection, the throat is ❖ Amoxicillin orally 8 reddish. What is your treatment choice? ❖ Amoxicillin/clavulanic acid orally 9 10 ❖ No antibiotic 11 12 31. During ward round, you have seen two patients with impaired ❖ Patient A 13 renal function. ❖ Patient B 14 15 - Patient A is a 68-year-old male with cellulitis in the lower limb. ❖ Patient A & B 16 For peer review only 17 He is administered clindamycin. ❖ Neither patient A nor patient B 18 - Patient B is a 64-year-old woman with diabetes who received 19 20 treatment for sepsis with ceftriaxone empirically. 21 22 In which case will you need to adjust the antibiotic dose? 23 24 32. Which one of the following antibiotics may be safely given ❖ Amoxicillin 25 during the first trimester of pregnancy? ❖ Ciprofloxacin 26 27 ❖ Gentamicin 28 29 33. Which of the following antibiotics has the best activity against ❖ Ciprofloxacin 30 anaerobes? ❖ Metronidazole 31 32 ❖ Trimethoprim/sulphamethoxazole http://bmjopen.bmj.com/ 33 34 34. Methicillin resistant - Staphylococcus aureus is susceptible to: ❖ Amoxicillin clavulanic acid 35 ❖ Cefotaxime 36 37 ❖ Ceftriaxone 38 39 ❖ None of these antibiotics 40 35. Which of the following antibiotics most effectively crosses the ❖ Clindamycin on September 28, 2021 by guest. Protected copyright. 41 42 blood-brain barrier? ❖ Ceftriaxone 43 44 ❖ Vancomycin 45 46 36. Aminoglycoside antibiotics such as gentamicin are most active ❖ Orally, three times daily 47 when they are administered as follows: ❖ Parenterally, once daily 48 49 ❖ Parenterally, three times daily 50 51 52 53 Thank you very much for your kind and honest participation 54 55 56 57 58 Page 7 of 7 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open: first published as 10.1136/bmjopen-2019-030823 on 31 March 2020. Downloaded from

BMJ Open Page 24 of 29

1 2 3 Questionnaire to assess laboratory capacity for AMR testing. 4 5 Objective: To assess capacity for Study sites laboratories to perform antimicrobial susceptibility testing and contribute to AMR surveillance 6 7 Laboratory infrastructure and equipment Additional Information 8 Q1 Does your lab have capacity to carry out basic bacteriology (process stool, urine, Yes 9 10 urethral/ cervical swabs, and blood)? (Perform bacterial culture, identification, and Partial susceptibility testing)? 11 No 12 For peer review only 13 Q2 Does your lab have resources for basic aerobic bacterial culture? Yes 14 Partial 15 No 16 Q3 Does your lab possess CO2 incubators and CO2 tank? Yes 17 Partial http://bmjopen.bmj.com/ 18 No 19 Q4 Does your lab perform susceptibility testing by disc diffusion? Yes 20 Partial 21 No 22 Q5 Please indicate the presence and status of the following in your lab Present Functional 23 (Tick) (Tick) 24 ❖ Petri dishes 25 ❖ Swabs for surface application of cultures on September 28, 2021 by guest. Protected copyright. 26 27 ❖ Standardized susceptibility testing discs 28 ❖ control strains of known susceptibility patterns 29 ❖ Incubators 30 ❖ Refrigerator 31 ❖ Autostart backup for refrigerator/incubator 32 ❖ Media preparation room 33 ❖ Autoclave 34 ❖ Compound microscope 35 ❖ Weighing scale 36 ❖ Biosafety cabinet Class 2 37 ❖ Candle jar 38 ❖ pH meter 39 ❖ water distiller 40 41 Q6 Does your lab have automated system (Vitek) to conduct Antimicrobial Susceptibility Yes 42 Testing? Partial 43 No 44 45 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 BMJ Open: first published as 10.1136/bmjopen-2019-030823 on 31 March 2020. Downloaded from

Page 25 of 29 BMJ Open

1 2 3 USE OF STANDARDIZED METHODS 4 Additional Information 5 Q7 Does your laboratory use Clinical Laboratory Standards Institute (CLSI) Yes 6 guidelines? Partial 7 No 8 Q8 Does your laboratory use CLSI interpretation breakpoints? Yes 9 Partial 10 No 11 12 Q9 Does your laboratory select individual antibiotics following CLSI guidelines? Yes For peer reviewPartial only 13 14 No 15 Q10 Are single isolates or pure cultures only used for final performance of antimicrobial Yes 16 susceptibility testing? Partial 17 No http://bmjopen.bmj.com/ 18 Q11 Is the inoculum size standardized using a turbidity standard (0.5 McFarland) or Yes 19 other acceptable method? Partial 20 No 21 Q12 Does your lab have provision of standard microorganisms (ATCC) for internal Yes 22 quality control (useful in determining the potency of drugs or checking the quality Partial 23 24 of media)? No 25 on September 28, 2021 by guest. Protected copyright. 26 Q13 For disk susceptibility tests, are zone sizes of controls measured and recorded? Yes 27 Partial 28 No 29 Q14 Are zone sizes of tests measured and used for recording sensitivity resistance? Yes 30 Partial 31 No 32 Q15 Does your lab use commercially prepared dehydrated AST media? Yes 33 Partial 34 No 35 Q16 Does your lab perform Susceptibility Testing directly from specimen based on Yes 36 clinical information? 37 Partial 38 No 39 Q17 If direct susceptibility testing from specimen show mixed cultures, does your lab Yes 40 repeat susceptibility testing with isolated organisms? Partial 41 No 42 USE OF STANDARDIZED OPERATING PROCEDURES (SOPs) 43 Q18 For antimicrobial susceptibility testing systems, are there do cumented criteria in Yes 44 45 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 BMJ Open: first published as 10.1136/bmjopen-2019-030823 on 31 March 2020. Downloaded from

BMJ Open Page 26 of 29

1 2 3 your institutions’ SOPs for interpretation of the endpoint or zone size? Partial 4 No 5 Q19 Are guidelines established for the number and type of antibiotics reported for Yes 6 organisms isolated from different sites of infection? Partial 7 No 8 Q20 Do you report Antimicrobial Susceptibility Testing results based on Hospital policy Yes 9 (in consultation with Pharmacy, Infection control and Infectious diseases Partial 10 physicians. 11 No 12 QUALITY ASSURANCE For peer review only 13 Q21 Is each new lot of susceptibility disks checked for activity before use? Yes 14 Partial 15 No 16 Q22 Does your lab use QC (quality control) strains to assess new lot of susceptibility Yes 17 discs?

Partial http://bmjopen.bmj.com/ 18 No 19 Q23 Are tolerance limits for potency of antimicrobials established (criteria for "out of Yes 20 control")? Partial 21 No 22 Q24 Does your laboratory procedure manual address unusual or inconsistent Yes 23 antimicrobial testing results? Partial 24 25 No on September 28, 2021 by guest. Protected copyright. 26 Q25 Does your lab participate in any Antimicrobial Susceptibility Testing related Yes 27 internal quality assurance program? Partial 28 No 29 Q26 Does your lab participate in any Antimicrobial Susceptibility Testing Yes 30 related external quality assurance program? Partial 31 No 32 Q27 Are out of control results reported to supervisory personnel? Yes 33 Partial 34 35 No 36 37 38 39 40 41 42 43 44 45 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 BMJ Open: first published as 10.1136/bmjopen-2019-030823 on 31 March 2020. Downloaded from

Page 27 of 29 BMJ Open

1 2 3 READINESS FOR AMR SURVEILLANCE 4 Q28 Does your lab participate in antimicrobial resistance surveillance? Yes 5 Partial 6 7 No 8 Q29 Does your lab generate on routine basis antibiogram for purpose of monitoring the Yes 9 resistant and sensitivity patterns in your institution? Partial 10 No 11 Q30 Does your lab conduct all Antimicrobial Susceptibility Testing or forwards it to Yes 12 other labs? For peer reviewPartial only 13 No 14 15 Q31 Does your lab receive samples for Antimicrobial Susceptibility Testing from other Yes 16 labs? Partial 17 No http://bmjopen.bmj.com/ 18 Q32 Is Antimicrobial Susceptibility Testing cumulative data collected manually? Yes 19 Partial 20 No 21 22 Q33 Is Antimicrobial Susceptibility Testing cumulative data collected automatically Yes 23 using lab information system (LIS)? Partial 24 No 25 DETECTION OF SPECIFIC ORGANISMS on September 28, 2021 by guest. Protected copyright. 26 Q34 Does your laboratory have the capacity of identifying resistance genotypes or Yes 27 resistant bacterial clones? Partial 28 No 29 30 EQUIPMENT MAINTENANCE 31 Q35 Are Antimicrobial Susceptibility Testing equipment maintained appropriately and Yes 32 calibrated? Partial 33 No 34 Q36 Does your lab monitor incubator temperatures on a daily basis? Yes 35 Partial 36 No 37 38 CONTINUING MEDICAL EDUCATION 39 Q37 How often do you receive training in Bacteriology? Yes 40 Partial 41 No 42 Q38 How often are you trained in conducting Antimicrobial Susceptibility Testing? 43 44 45 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 BMJ Open: first published as 10.1136/bmjopen-2019-030823 on 31 March 2020. Downloaded from

BMJ Open Page 28 of 29

1 2 3 STAFFING 4 Q39 How many laboratory technologists are in the station? 5 Q40 How many microbiologists are in the station? 6 Q41 Does your laboratory have staff with Bachelor’s degree qualification or higher? Yes 7 8 Partial 9 No 10 Q42 Do you engage a Consultant clinical microbiologist(s)? Yes 11 Partial 12 For peer reviewNo only 13 CONSUMABLES 14 Q43 How often do you experience unavailability of consumables in Microbiology Yes 15 section? Eg Lack of biochemical reagents and media Partial 16 No 17 Q44 Does your lab experience delays in Antimicrobial Susceptibility Testing Yes http://bmjopen.bmj.com/ 18 due to lack of reagents? Partial 19 No 20 Q45 Do frequent stock outs lead to low demand of cultures by clinicians? Yes 21 Partial 22 No 23 BIOSAFETY 24 25 Q46 Does your lab autoclave/incinerate cultures prior to discard? Yes on September 28, 2021 by guest. Protected copyright. 26 Partial 27 No 28 Q47 Do you have handwashing facility in the laboratory? Yes 29 Partial 30 No 31 Q48 Does your lab get continuous supply of running water? Yes 32 Partial 33 No 34 Q49 Does your lab have soap supply in the handwash facility? Yes 35 Partial 36 No 37 38 39 40 41 42 43 44 45 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Page 29 of 29 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2019-030823 on 31 March 2020. Downloaded from

1 2 3 Core element 1: Senior hospital management leadership towards Yes No 4 antimicrobial stewardship 5 6 1 Has your hospital management formally identified antimicrobial stewardship 7 as a priority objective for the institution and included it in its key performance 8 indicators? 9 2 Is there dedicated, sustainable and sufficient budgeted financial support for 10 antimicrobial stewardship activities (e.g., support for salary, training, or IT 11 (information technology) support)? 12 3 Does your hospital follow any (national or international) staffing standards for 13 antimicrobial stewardship activities (e.g. number of full-time equivalent (FTE) 14 15 per 100 beds for the different members of the antimicrobial stewardship 16 team)? For peer review only 17 Core element 2: Accountability and responsibilities 18 4 Does your hospital have a formal/written antimicrobial stewardship 19 programme/strategy accountable for ensuring appropriate antimicrobial use? 20 5 Does your hospital have a formal organizational multidisciplinary structure 21 responsible for antimicrobial stewardship (e.g., a committee focused on 22 23 appropriate antimicrobial use, pharmacy committee, patient safety 24 committee or other relevant structure)? 25 6 Is there a healthcare professional identified as a leader for antimicrobial 26 stewardship activities at your hospital and responsible for implementing the 27 programme? 28 7 Is there a document clearly defining roles, procedures of collaboration and 29 responsibilities of the antimicrobial stewardship team members? 30 31 8 Are clinicians, other than those part of the antimicrobial stewardship team 32 (e.g. from the ICU, Internal Medicine and Surgery) involved in the http://bmjopen.bmj.com/ 33 antimicrobial stewardship committee? 34 9 Does the antimicrobial stewardship committee produce regularly (indicate 35 minimum time) a dedicated report which includes e.g. antimicrobial use data 36 and/or prescription improvement initiatives, with time-committed short term 37 and long term measurable goals/ targets for optimizing antimicrobial use? 38 10 Is there a document clearly defining the procedures of collaboration of the 39 40 antimicrobial stewardship team/committee with the infection prevention and on September 28, 2021 by guest. Protected copyright. 41 control team/committee? 42 Core element 3: Available expertise on infection management 43 11 Do you have access to laboratory/imaging services and to timely results to be 44 able to support the diagnosis of the most common infections at your hospital? 45 12 In your hospital are there, or do you have access to, trained and experienced 46 healthcare professionals (medical doctor, pharmacist, nurse …) in infection 47 48 management (diagnosis, prevention and treatment) and stewardship willing 49 to constitute an antimicrobial stewardship team? 50 Core element 4: Education and practical training 51 13 Does your hospital offer a range of educational resources to support staff 52 training on how to optimize antimicrobial prescribing? 53 14 Do the antimicrobial stewardship team members receive regular training in 54 antimicrobial prescribing and stewardship? 55 56 Core element 5: Other actions aiming at responsible antimicrobial use 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 30 of 29 BMJ Open: first published as 10.1136/bmjopen-2019-030823 on 31 March 2020. Downloaded from

1 2 3 15 Is a multidisciplinary antimicrobial stewardship team available at your hospital 4 (e.g., greater than one trained staff member supporting clinical decisions to 5 6 ensure appropriate antimicrobial use)? 7 16 Does your hospital support the antimicrobial stewardship activities/ strategy 8 with adequate information technology services? 9 17 Does your hospital have an antimicrobial formulary (i.e. a list of antimicrobials 10 that have been approved for use in a hospital, specifying whether the drugs 11 are unrestricted, restricted (approval of an antimicrobial stewardship team 12 member is required) or permitted for specific conditions)? 13 18 Does your hospital have available and up-to-date recommendations for 14 15 infection management (diagnosis, prevention and treatment), based on 16 international/nationalFor peer evidence-based review guidelines and only local susceptibility 17 (when possible), to assist with antimicrobial selection (indication, agent, dose, 18 route, duration) for common clinical conditions? 19 19 Does your hospital have a written policy that requires prescribers to 20 document an antimicrobial plan (includes indication, name, dosage, duration, 21 route and interval of administration) in the medical record or during order 22 entry for all antimicrobial prescriptions? 23 24 20 Does the antimicrobial stewardship team review/audit courses of therapy for 25 specified antimicrobial agents or clinical conditions at your hospital? 26 21 Is advice from antimicrobial stewardship team members easily available to 27 prescribers? 28 22 Is advice from antimicrobial stewardship team members easily available to 29 prescribers? 30 Core element 6: Monitoring and surveillance (on a continuous basis) 31 32 23 Does your hospital monitor the quality of antimicrobial use at the unit and/or http://bmjopen.bmj.com/ 33 hospital wide level? 34 24 Does your stewardship programme monitor compliance with one or more of 35 the specific interventions put in place by the stewardship team (e.g. indication 36 captured in the medical record for all antimicrobial prescriptions)? 37 25 Does your hospital monitor antibiotic susceptibility rates for a range of key 38 bacteria? 39 40 26 Does your hospital monitor the quantity of antimicrobials prescribed/ on September 28, 2021 by guest. Protected copyright. 41 dispensed/purchased at the unit and/or hospital wide level? 42 Core element 7: Reporting and feedback (on a continuous basis) 43 27 Does your stewardship programme share hospital-specific reports on the 44 quantity of antimicrobials prescribed/dispensed/purchased with prescribers? 45 28 Does your stewardship programme share facility-specific reports on antibiotic 46 susceptibility rates with prescribers? 47 48 29 Are results of audits/reviews of the quality/appropriateness of antimicrobial 49 use communicated directly with prescribers? 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open BMJ Open: first published as 10.1136/bmjopen-2019-030823 on 31 March 2020. Downloaded from

Combating Antibiotic Resistance using Guidelines and Enhanced Stewardship in Kenya: A protocol for an Implementation Science Approach ForJournal: peerBMJ Open review only Manuscript ID bmjopen-2019-030823.R1

Article Type: Protocol

Date Submitted by the 15-Aug-2019 Author:

Primary Subject Health economics Heading:

Secondary Subject Heading: Evidence based practice, Medical management

Antimicrobial stewardship, Antimicrobial resistance, Implementation Keywords:

science on September 28, 2021 by guest. Protected copyright.

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 1 of 30 BMJ Open

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 2 of 30

1 2 3

Combating Antibiotic Resistance using Guidelines and Enhanced Stewardship in BMJ Open: first published as 10.1136/bmjopen-2019-030823 on 31 March 2020. Downloaded from 4 Kenya: a protocol for an Implementation Science Approach 5 6 1 2 3 4 5 7 Jesse Gitaka , Dominic Mureithi , Davies Ndegwa , Moses Masika , Geoffrey Omuse , 6 1 1 1 1 8 Moses Ngari , Francis Makokha , Peter Mwaura , Moses Kamita , Ronald Mathai , Francis 9 Muregi1, Matilu Mwau7 10 11 Affiliations 12 1. Directorate of Research and Innovation, Mount Kenya University, P.O. Box 342-01000, 13 Thika, Kenya 14 2. Department of Animal Health and Production, Maasai Mara University, P.O. Box 861- 15 16 20500, Narok, Kenya. 17 3. Department of Medical Laboratory Sciences, Kenya Medical Training College P.O. Box 18 30195-00100, Nairobi,For Kenya peer review only 19 4. Department of Medical Microbiology, College of Health Sciences, University of Nairobi, 20 P.O. Box 30197-00100, Nairobi, Kenya. 21 5. Department of Pathology, Aga Khan University Hospital, P.O. Box 30270-00100, 22 23 Nairobi, Kenya 24 6. KEMRI/Wellcome Trust Research Programme, Centre for Geographic Medicine 25 Research - Coast, P.O. Box 230-80108, Kilifi, Kenya 26 7. Centre for Infectious and Parasitic Diseases Control Research, Kenya Medical Research 27 Institute, P.O. Box 3-50400, Busia, Kenya 28 29 Correspondence: Jesse Gitaka 30 31 [email protected] 32 Directorate of Research and Innovation, Mount Kenya University, P.O. Box 342-01000, 33 Thika, Kenya 34 35 Word count: 3333 words 36 37 http://bmjopen.bmj.com/ 38 39 40 41 42 43 44 45 on September 28, 2021 by guest. Protected copyright. 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 3 of 30 BMJ Open

1 2 3

Abstract BMJ Open: first published as 10.1136/bmjopen-2019-030823 on 31 March 2020. Downloaded from 4 5 Introduction 6 7 8 Antibiotic resistance (AMR) is a growing problem globally especially in Sub-Saharan Africa 9 including Kenya. Without any intervention, developing countries will be most affected due to 10 the high burden of diseases. Studies have consistently shown that inappropriate use of 11 antimicrobials is the major drivers of AMR. To address this challenge hospital are now 12 implementing antibiotic stewardship programs (ASPs), which have been observed to reduced 13 antibiotic usage, decrease the prevalence of resistance and lead to significant economic 14 benefits. However, the implementation of the guideline is highly dependent on settings in 15 16 which they are rolled out. This study, employing an implementation science approach aims to 17 address the knowledge and research gap in this area and provide critical data and experiences 18 in using antibiotic guidelinesFor peer and stewardship review programs in onlythe public health sector. This will 19 provide evidence of ASP performance and potentially contribute to the county, national and 20 regional policies on antibiotics use. 21 22 Methods 23 24 25 The study will be conducted in three geographically diverse regions each represented by two 26 hospitals. A baseline study on antibiotic usage, resistance and de-escalation, duration of 27 hospital stay, rates of readmission and costs will be carried out in the pre-implementation 28 phase. The intervention, that is, the use of antibiotic guidelines and antibiotic stewardship 29 programs will be instituted for 18 months using a stepwise implementation strategy that will 30 facilitate learning and continuous improvement of stewardship activities and updating of 31 guidelines to reflect the evolving antibiotic needs. 32 33 34 Ethics and dissemination 35 36 The proposal has been approved by the University of Nairobi-Kenyatta National Hospital 37 Ethical and Research Committee (ERC), the Mount Kenya University Ethics Review http://bmjopen.bmj.com/ 38 Committee and Hospital-based Review Committees. Study findings will be presented to 39 policy stakeholders and published in peer-reviewed scientific journals. It is anticipated the 40 findings will inform local-based appropriate antibiotic use guidelines. 41 42 Key Words 43 44 45 Antimicrobial stewardship; Implementation science; Antimicrobial resistance on September 28, 2021 by guest. Protected copyright. 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 4 of 30

1 2 3

Strengths and limitations of this study BMJ Open: first published as 10.1136/bmjopen-2019-030823 on 31 March 2020. Downloaded from 4 5  First study aimed at rolling out antimicrobial stewardship committees in multiple 6 hospitals in Kenya concurrently. 7 8  Use of implementation approach to support implement suggested guidelines for 9 10 antimicrobial resistance surveillance. 11  First hand evidence on the antimicrobial resistance in three diverse counties in Kenya. 12 13  The study is limited to only three counties of the 47 counties in Kenya 14 15 16 17 18 For peer review only 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 http://bmjopen.bmj.com/ 38 39 40 41 42 43 44 45 on September 28, 2021 by guest. Protected copyright. 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 5 of 30 BMJ Open

1 2 3

Introduction BMJ Open: first published as 10.1136/bmjopen-2019-030823 on 31 March 2020. Downloaded from 4 5 Antibiotic resistance is a serious public health concern globally1,2 and locally3,4 and fears of 6 7 running out of antibiotics options in the near future have been expressed. In 2016, the World 8 Health Organization (WHO), called for immediate and concerted efforts to mitigate this 9 threat to global health that was estimated to contribute to 700,000 deaths in 2014 and 10 projected to cause 10 million deaths in 2050 if inadequately mitigated.5 The driving force 11 escalating rates of resistance is the injudicious use of antibiotics in patients and in livestock 12 or release into the environment. These forces exert selective pressure which rise and increase 13 spread of resistant pathogens that may emerge by mutations or acquisition of plasmids 14 6 15 carrying resistance genes. In one study, prior antibiotic exposure was the key independent 16 risk factor for the acquisition of antibiotic multi-resistant.7 Broad-spectrum antibiotics have 17 the unintended consequence of selecting multidrug-resistant pathogens and increasing the 18 likelihood of infectionFor by fungi peer and Clostridium review difficile.8 only 19 20 Nonetheless, the injudicious use of antibiotics is not unusual. In Africa, many patients do not 21 receive treatment from the conventional health care system. Of those who receive antibiotics, 22 23 31.7 % do not consult a doctor for a prescription and a further 26.4% obtain the antibiotics 9 24 over-counter. A study in South Africa found that 54.9% of antibiotics were inappropriately 25 prescribed in intensive care unit settings while in the US, 20-50% of prescribed antibiotics 26 were unnecessary or unwarranted.10–12 27 28 The rate of antimicrobial resistance in Kenya is worrying and rising. In one study, the 29 prevalence of Salmonella typhi resistant to two or more antimicrobials was observed to have 30 increased from 50% in 1998 to 78% in 2004 at Kenyatta National Hospital.13 The Global 31 32 Antibiotic Resistance Partnership – Kenya Working Group Report of 2011 identified 33 antibiotic resistance as a key issue in Kenya and made recommendations to curtail the spread. 34 These recommendations included the use of antibiotic guidelines that took into consideration 35 local resistance surveillance data and enhanced antibiotic stewardship programs (ASPs).14 36 Even so, these ASPs have not been instituted at county hospitals, and key implementation http://bmjopen.bmj.com/ 37 data and experiences are lacking in their roll out. 38 39 Antibiotic stewardship is defined as the optimum selection, dosage, and duration of 40 41 antimicrobial treatment that yields the best clinical outcomes for the treatment or prevention 42 of infection with the least toxicity to the patient and minimal impact on subsequent 43 resistance.15 It has the potential to lower treatment costs and realize economic benefits to the 44 patient, health care system and the country at large.16,17 Moreover, optimizing antibiotic use on September 28, 2021 by guest. Protected copyright. 45 by minimizing exposure, fine-tuning dosage and reducing superfluous therapy and focusing 46 treatment to the likely culprit pathogens is a strategy that boosts patient safety18 and 47 48 ultimately safeguards against antibiotic resistance. 49 50 Justification 51 52 Antibiotic resistance is a major health challenge globally. Studies in Kenya have shown rising 53 antibiotic resistance over the last 3 decades. Nonetheless, antibiotic guidelines and antibiotic 54 stewardship programs which have been observed to lead to significant economic benefits, 55 reduce antibiotic usage and lower the prevalence of resistance especially in Europe, North 56 America, Japan, and South Africa have not been employed to tackle the challenge in the 57 58 public health sector in Kenya. The GARP report of 2011, recognized use of guidelines and 59 stewardship programs as a potential strategy in ‘saving antibiotics’ but noted the need for 60 local studies to understand implementation challenges, derive lessons and embed the strategy

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 6 of 30

1 2 3

within the Kenyan health care system. This study, employing an implementation science BMJ Open: first published as 10.1136/bmjopen-2019-030823 on 31 March 2020. Downloaded from 4 approach aims to address the knowledge and research gap in this area and provide critical 5 6 data and experiences in using antibiotic guidelines and stewardship programs in the public 7 health sector. 8 9 Objectives of the Study 10 11 General Objectives 12 13 The overall aim is to evaluate the impact of antibiotic guidelines for empirical treatment of 14 urinary tract infections, community-acquired pneumonia, bacteremia and meningitis and 15 antibiotic stewardship program on reducing usage of broad-spectrum antibiotics, antibiotic 16 17 de-escalation, duration of hospital stays, rates of readmission and prevalence of antibiotic 18 resistance in 6 countyFor hospitals peer in Kenya. review only 19 20 Specific Objectives 21 22 1. To develop guidelines for antibiotic use for common infections. 23 2. To set up antibiotic steward committees (ASCs) in 6 county hospitals. 24 3. To train the ASCs to be able to perform their mandate. 25 26 4. To measure the usage of broad-spectrum antibiotics, antibiotic de-escalation, duration 27 of hospital stays and rates of readmission in hospitals using the antibiotic stewardship 28 strategy. 29 5. To ascertain antibiotic resistance patterns using culture, sensitivity, and genetic 30 markers. 31 6. To establish the health care workers knowledge, attitudes and prescription practices 32 regarding antibiotics resistance, use of guidelines and ASPs. 33 34 7. To evaluate the economic benefits of using guidelines for antibiotics use and ASPs. 35 36 Expected Outputs of the research 37 http://bmjopen.bmj.com/ 38 1. Antibiotic guidelines informed by local bacteria susceptibility patterns. 39 2. Antibiotic Stewardship Committees (ASCs) in 6 County hospitals in Kenya. 40 3. Trained ASCs 41 4. Data on antibiotics usage, antibiotic de-escalation, duration of hospital stay and rates 42 of readmission in hospitals using the antibiotic stewardship strategy 43 44 5. Map and data of antibiotic resistance pattern before and after implementation of the 45 strategy. on September 28, 2021 by guest. Protected copyright. 46 6. Qualitative data on health care workers’ knowledge, attitudes, and practices on the 47 antibiotic usage and stewardship programs 48 7. Publication on cost-benefit analysis for using antibiotic guidelines and ASPs 49 50 Methodology 51 52 53 Setting 54 55 To be able to evaluate the antibiotic stewardship strategy across Kenya using an 56 implementation science approach, the study will be conducted in 6 county hospitals (Kiambu 57 County Referral Hospital, Bungoma County Hospital, Webuye Sub-County Hospital, Nakuru 58 County Teaching and Referral Hospital, Thika Level 5 Hospital and Naivasha Sub-County 59 Hospital). 60

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 7 of 30 BMJ Open

1 2 3

Design BMJ Open: first published as 10.1136/bmjopen-2019-030823 on 31 March 2020. Downloaded from 4 5 The study will utilize the reach, effectiveness, adoption, implementation and maintenance (RE-AIM) 6 conceptual framework. RE-AIM involves the interaction of all the five factors and fits well with 7 approaches that are system based19. The study design will encompass a pre-implementation phase, 8 9 a stepwise implementation phase, and an endline study to measure the changes in outcomes 10 between the phases. A baseline study on antibiotic usage (Defined Daily Doses per 1000), 11 antibiotic resistance (culture and genetic markers), antibiotic de-escalation, duration of 12 hospital stays, rates of readmission, prescription patterns, and costs analysis will be carried 13 out in the pre-implementation phase. The intervention, which is the use of antibiotic 14 guidelines and antibiotic stewardship programs, will be instituted in the 6 county hospitals for 15 16 18 months. The stepwise implementation phase will involve introduction of interventions as 17 well as monitoring and evaluation of their effectiveness and thereafter improving the 18 interventions whereFor necessary. peer The strategy review will facilitate only learning cycles every 4-6 weeks 19 and continuous improvement of stewardship activities and updating of guidelines to reflect 20 evolving antibiotic needs in diverse settings. The end line study will be conducted and 21 differences between the 2 phases evaluated as below (Statistical analysis). 22 23 In the baseline and endline studies, multidisciplinary strategies will be employed as 24 25 follows 26 i. Health care workers knowledge, attitudes, and practices on antibiotic resistance, 27 guidelines, and ASP will be studied qualitatively and quantitatively. 28 ii. Basic science approaches encompassing antibiotics culture sensitivity and molecular 29 biology-genetic markers of resistance will be analyzed as detailed below. 30 iii. Clinical- patient outcomes will be studied to evaluate the guidelines. 31 iv. Health economics-cost savings on using guidelines and ASP will be evaluated as 32 33 below. 34 35 Antibiotic guidelines and ASPs: Development 36 37 Antibiotic guidelines will be formulated in consultation with senior clinicians in the study http://bmjopen.bmj.com/ 38 hospitals taking into account each hospital’s antimicrobial resistance patterns. ASPs 39 committee will comprise of the hospital physician, microbiologist, and pharmacist. Broad 40 spectrum antibiotic prescriptions will be brought to the attention of the ASP committees who 41 42 will also perform regular ward rounds three times a week in the initial stages and later once a 43 week to optimize adherence to antibiotic guidelines. Furthermore, the guidelines will be 44 promoted through teaching sessions, provision of pocket-size guideline cards to clinicians 45 and pharmacists, large poster displays in the wards and through hospital and project websites. on September 28, 2021 by guest. Protected copyright. 46 47 Data collection 48 49 Laboratory assessment tools will be used to determine the preparedness of the laboratories to 50 perform antibiotic sensitivity tests (see Additional file 1). The tool will assess whether the 51 52 laboratories have the equipment that are necessary to perform bacterial culture, identification, 53 and 54 susceptibility testing such as the carbon dioxide incubators, safety cabinets, and refrigerators 55 among others. The tool will also determine whether the laboratories have the internationally 56 recommended guidelines to perform susceptibility and quality assurance tests. In addition, 57 knowledge, attitude and practices (KAP) about antibiotic prescribing and resistance among 58 medical practitioners will be assessed using a KAP tool (See Additional file 2). The tool will 59 60 target those medical practitioners who usually prescribe antimicrobial drugs in the hospitals

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 8 of 30

1 2 3

and will include the consultants, medical officers and the interns, clinical officers and the BMJ Open: first published as 10.1136/bmjopen-2019-030823 on 31 March 2020. Downloaded from 4 interns, as well as the pharmacists. The KAP tool will also determine the guidelines that the 5 6 prescribers use to decide on the appropriate antimicrobial drug. A health system assessment 7 tool will be used to gather baseline information on the antimicrobial stewardship activities in 8 each hospital (see Additional file 3). Information on each hospital bed occupancy, antibiotic 9 usage, and antibiotic resistance data will be collected from hospital information systems, 10 pharmacy management systems and laboratory reporting systems respectively. Antibiotic 11 usage will be analyzed in order to determine the impact of the intervention on antibiotic 12 usage comparing the intervention and the control arms. Data on antibiotics issued to adult and 13 14 children inpatients only will be factored excluding discharge and outpatient supplies. 15 16 To consistently compare antibiotic usage, the defined daily doses (DDD) will be used and 17 expressed per 1000 occupied bed days (OBDs) to account for fluctuations in activity. The 18 OBDs will be obtainedFor from peerthe hospital reviewinformation systems. only 19 20 Culture sensitivity and genetic analyses 21 22 Nasal swabs or tracheal aspirate, urine, wound swabs and blood samples will be taken from 23 24 patients who have consented to the study for bacteriological analysis. Samples will be 25 subjected to standard bacteriological analysis to isolate the culprit bacteria. Bacterial species 26 will be confirmed by use of biochemical test and analytical profiles index API strips (Bio 27 Merieux France). Antimicrobial susceptibility test will be performed on isolated bacteria as 28 per the Kirby-Bauer Method following manufacturer’s instruction. Results will be interpreted 29 using the Clinical and Laboratory Standards Institute (CLSI) tables.20 30 31 Any bacteria isolate found to be resistant to third generation cephalosporin’s will be tested for 32 33 production of extended spectrum Beta-lactamase (ESBLs) using the synergy disk diffusion 34 test. Vancomycin Resistance Enterococci (VRE) will be identified using disc diffusion tests. 35 Methicillin resistance in S. aureus (MRSA) will be detected by testing isolates resistant to 36 cefoxitin by E test (AB Biodisk, Solna, Sweden) on Mueller–Hinton agar supplemented with 37 2% NaCl and incubated at 37°C for 24 h. The identified ESBLs, VRE and MRSA will be http://bmjopen.bmj.com/ 38 analyzed by PCR and sequencing to identify the resistance genotype. In vitro conjugation 39 tests will be performed to determine if resistance in bacteria is transferable. 40 41 42 Cost-benefit analysis for the use of antibiotic guidelines and Antibiotic Stewardship 43 Program 44 45 A cost-benefit analysis (CBA) from a health facility and a national perspective will be on September 28, 2021 by guest. Protected copyright. 46 performed. For health facilities, three cost drivers will be considered: pharmacy spending, 47 length of stay, and antimicrobial stewardship interventions (training, infection control 48 measures, etc.). For the country, we will consider, Disability-Adjusted Life years (DALYS), 49 21–23 50 work-days lost, and cost of treatment. This will be done by collecting and analyzing data 51 on patient income, length of hospital stay, death or disability occasioned by drug-resistant 52 pathogens, hospital pharmacy expenditure and cost of training/rolling out antimicrobial 53 stewardship guidelines. This data will be collected before and after antimicrobial stewardship 54 interventions. 55 56 Study size 57 58 Prevalence of inappropriate use of broad-spectrum antibiotics 59 60

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 9 of 30 BMJ Open

1 2 3

The sample size is based on the prevalence of inappropriate use of broad-spectrum BMJ Open: first published as 10.1136/bmjopen-2019-030823 on 31 March 2020. Downloaded from 4 antibiotics of ~50%10 in South Africa. To detect a reduction of the 50% inappropriate use of 5 6 broad-spectrum antibiotics by 20% to 30% with a power of 90% and α of 0.05 in a two-sided 7 test, a sample size of 410 in each county hospital would be adequate. In order to detect a 8 reduction (20%) but with a power 80% the sample size would be 320 as shown in Figure 1 9 below. With a 15% allowance for loss to follow up, a total of about 500 participants would be 10 enough in each of the selected county hospitals. Therefore, we shall evaluate the prevalence 11 of inappropriate use of broad-spectrum antibiotics in 500 x 6= 3,000 patients from the 6 12 facilities. 13 14 15 Prevalence of antibiotic resistance 16 17 The sample size is based on the prevalence of antibiotic resistance of 78%13 in 2004 at 18 Kenyatta National Hospital,For peerKenya. To detectreview a reduction onlyof the 78% antibiotic resistance by 19 10% to 68% with a power of 90% and α of 0.05 two-sided test, a sample size of 500 in each 20 county hospital would be adequate. While to achieve a reduction (10%) but with power, 80% 21 the sample size would be 220 as shown in Figure 2 below. With 15% allowance for loss to 22 23 follow up, a total of 600 participants would be enough in each of the selected county 24 hospitals. Therefore, we shall evaluate the prevalence of resistance in 600 x 6 = 3,600 25 patients for the 6 facilities. 26 27 28 29 Data Management 30 31 Data will be recorded on a standardized Case Report Form (CRF) at every hospital by the 32 trained study staff. The data will be held locally and uploaded to the secure central study 33 34 server hosted at the Mount Kenya University main campus in Thika. This will be overseen by 35 the Data Manager/statistician who will run regular reconciliation to derive the final study 36 database. Access to the study database will be restricted and password protected. 37 http://bmjopen.bmj.com/ 38 Statistical analyses 39 40 The data will be analyzed using qualitative and quantitative methods. Qualitative data will be 41 analyzed by subjecting the information to content analysis and presenting it in different 42 emerging themes. The summaries of the data emanating from these themes will then be 43 24 44 arranged on a case by case basis through the use of an Excel spreadsheet and the analyses 45 done by using NVIVO software. The quantitative data analyses will be done using Stata on September 28, 2021 by guest. Protected copyright. 46 version 14 (Stata, Inc.). The differences between baseline and endline on all the study 47 outcomes will be compared using appropriate statistical methods like McNemar’s test, paired 48 t-test and the Wilcoxon signed-rank test non-parametric test accounting for pairing and 49 clusters (hospitals). Multivariable log-binomial regression analysis will be used to get risk 50 51 factors for antibiotic resistance. 52 53 Ethical considerations 54 55 Approvals to carry out the study will be sought from the University of Nairobi-Kenyatta 56 National Hospital Ethical and Research Committee (ERC), the Mount Kenya University 57 Ethics Review Committee and each Hospital-based Review Committees. The study will 58 follow all provisions of the Declaration of Helsinki. Participants in the study will not incur 59 any cost in the transport nor processing of the samples, neither will they receive any 60

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 10 of 30

1 2 3

monetary inducements to participate in the study. Material transfer to laboratories outside of BMJ Open: first published as 10.1136/bmjopen-2019-030823 on 31 March 2020. Downloaded from 4 Kenya shall not be undertaken in this study. Informed written consent will be sought from the 5 6 participants enrolled in the study. 7 8 Exit strategy and stakeholder involvement 9 10 In the process of developing this protocol, we have engaged physicians working in the 11 proposed County hospitals and they have identified the challenge of antibiotic resistance as 12 real and appreciate the opportunities that use of antibiotic guidelines and ASPs may provide 13 in combating resistance, improving clinical care and saving costs. We plan to continue 14 involving all the stakeholders in the process who include clinicians, laboratory personnel, 15 16 pharmacists, public health officials, patients and scientists in the arena. We anticipate that the 17 study findings will inform county and national policy on mitigating antibiotic resistance and 18 raise public awarenessFor on the peer need for judicious review use of antibiotics. only The project will use social 19 media platforms, websites of the collaborating institutions, and publications in peer-reviewed 20 journals, local dailies and presentations in scientific meetings to further engage stakeholders 21 and the public on this important issue and enhance the learning approach inherent in the 22 23 strategy of implementation science for improved performance of the ASPs and the study in 24 general. 25 26 Dissemination 27 28 Obtained results will be disseminated at different platforms which include research 29 conferences and in peer-reviewed journals. In addition, the findings will be shared in a 30 dissemination forum bring together members of the health management teams at both the 31 country and county levels, clinicians who do prescription of antimicrobial drugs, researchers 32 33 and other key stakeholders. 34 35 Discussion 36 37 The emergence of antimicrobial-resistant pathogens and a lack of new drugs to effectively http://bmjopen.bmj.com/ 38 treat these pathogens are the two main challenges in human health. There is thus the need to 39 advocate for proper use of the currently available antimicrobial agents by safeguarding their 40 effectiveness. Antibiotic stewardship has been shown to contribute to reducing antibiotic 41 resistance25,26 but this strategy has not been rolled out in most sub-Saharan countries in level 42 43 4 and 5 hospitals. The proposed work will employ an implementation research approach to 44 evaluate the best strategies and derive lessons on mainstreaming antibiotic stewardship in 45 these facilities. By leveraging on the health system approach, the implementation research on September 28, 2021 by guest. Protected copyright. 46 will unmask real life impediments and opportunities and working with hospital teams co- 47 design execution plans, monitoring and evaluation and sustainability of the stewardship 48 programs. 49 50 51 Most hospitals in Kenya lack the capacity to do antimicrobial sensitivity tests due to the lack 52 of enough resources and technical knowhow among other challenges. This study will first do 53 an assessment to determine the challenges that the hospitals are facing through interviews 54 with clinicians and assessment of the capacity of the laboratories to perform the sensitivity 55 tests. These steps will make up the initial phase and will guide the nature of implementations 56 to be used during the implementation phase. It is hoped that the project will capacitate the 57 laboratories in the six hospitals to do the tests and sensitize the clinicians on the need to 58 59 prescribe antimicrobial drugs based on results obtained from the laboratory tests. Once the 60 implementation phase is done, the final phase will be the endline phase where surveys similar

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 11 of 30 BMJ Open

1 2 3

to the ones that were conducted during the baseline phase will be done in order to determine BMJ Open: first published as 10.1136/bmjopen-2019-030823 on 31 March 2020. Downloaded from 4 the impact of the implementation strategies taken in reducing the cost of treatment, days of 5 6 stay in the hospital and burden on the health system among others. 7 8 The hospital management plays a key role in determining the allocation of resources to the 9 hospitals. In this project we have proposed to involve the hospital management by including 10 its members in the stewardship committees that will be established. This will ensure that the 11 management is well informed about the challenges that are in the sections that are involved in 12 surveillance and the progress that is being made. Involvement of the hospital management 13 will also ensure that there is a buy-in of the recommendations made by the stewardship 14 15 committee. 16 17 Although there is the need to establish antimicrobial stewardship committees in all the 18 hospitals, the projectFor proposes peer to start withreview the selected sixonly hospitals with the hope that the 19 same can be reproduced by other hospitals to establish similar committees in their hospitals. 20 21 Study timeline 22 23 The study is designed to take three years to complete. The study started enrolling on 2018 24 25 and participant recruitment will continue up to 2020 (Table 1). Currently, data collection on 26 antibiotic usage in the hospitals is going on. 27 28 Table 1: Proposed study timelines. 29 2018 2019 2020 30 31 32 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 33 Activity 34 35 Ethics clearance and 36 37 authorization from study http://bmjopen.bmj.com/ 38 sites 39 40 Baseline study on 41 42 prevalence of antibiotic 43 44 resistance and on September 28, 2021 by guest. Protected copyright. 45 stewardship activities 46 47 48 Identification of members 49 for the antibiotic 50 stewardship programs 51 52 Development of antibiotic 53 guidelines 54 55 Adopting and preparing 56 57 materials for ASP training 58 Development of antibiotic 59 60 guidelines and

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 12 of 30

1 2 3

stewardship mobile BMJ Open: first published as 10.1136/bmjopen-2019-030823 on 31 March 2020. Downloaded from 4 5 application 6 7 Training the ASP 8 members 9 10 Collecting data on 11 antibiotic usage in the 12 hospitals 13 14 Conducting KAPs study 15 on use of guidelines and 16 ASPs 17 18 Analyze and presentFor peer review only 19 antibiotic resistance data 20 using maps and charts 21 Presenting findings to 22 stakeholders and 23 preparing the final report 24 25 26 Competing Interests 27 28 The authors declare no conflict of interest. 29 30 Grant Information 31 32 This work was supported by The Kenya National Research Fund (grant number 33 NRF/MKU/2017/007 to JG). The funders had no role in study design, data collection and 34 analysis, decision to publish, or preparation of the manuscript. 35 36

Patient and Public Involvement http://bmjopen.bmj.com/ 37 38 No patient involved. 39 40 Availability of data and material 41 42 All data sets will be available to the public upon request. 43 44 Author Contributions 45 on September 28, 2021 by guest. Protected copyright. 46 JG conceived the project, JG, DM, DN, MM, GO, MN, FM, PM, RM, FM, and MM 47 developed the protocol, MK prepared the protocol for publication, all authors reviewed and 48 approved the protocol. 49 50 Acknowledgment 51 52 Not applicable 53 54 Additional files 55 56 Additional file 1: Laboratory preparedness assessment tool 57 58 Additional file 2: Laboratory knowledge, attitude and practice (KAP) tool 59 60 Additional file 3: Health system tool

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 13 of 30 BMJ Open

1 2 3

Figure Legends BMJ Open: first published as 10.1136/bmjopen-2019-030823 on 31 March 2020. Downloaded from 4 5 Figure 1: Statistical power analysis for sample size determination for the prevalence of 6 inappropriate use of broad-spectrum antibiotics 7 8 Figure 2: Statistical power analysis for sample size determination for the prevalence of 9 antibiotics resistance 10 11 12 13 14 15 16 17 18 For peer review only 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 http://bmjopen.bmj.com/ 38 39 40 41 42 43 44 45 on September 28, 2021 by guest. Protected copyright. 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 14 of 30

1 2 3

References BMJ Open: first published as 10.1136/bmjopen-2019-030823 on 31 March 2020. Downloaded from 4 5 1. Aslam, B. et al. Antibiotic resistance: a rundown of a global crisis. Infect. Drug Resist. 6 7 11, 1645–1658 (2018). 8 9 2. Livermore, D. M. Has the era of untreatable infections arrived? J. Antimicrob. 10 Chemother. 64, 29–36 (2009). 11 12 3. Bejon, P. et al. Invasive Gram-negative bacilli are frequently resistant to standard 13 antibiotics for children admitted to hospital in Kilifi, Kenya. J. Antimicrob. 14 Chemother. 56, 232–235 (2005). 15 16 4. Dougan, G. et al. Typhoid in Kenya Is Associated with a Dominant Multidrug- 17 18 Resistant SalmonellaFor entericapeer Serovar review Typhi Haplotype only That Is Also Widespread in 19 Southeast Asia. J. Clin. Microbiol. 48, 2171–2176 (2010). 20 21 5. WHO. United Nations meeting on antimicrobial resistance. (2016). 22 23 6. Baquero, F. et al. The global threat of antimicrobial resistance: science for 24 intervention. New Microbes New Infect. 6, 22–29 (2015). 25 26 7. Sanchez-Ortega, I. et al. Bacteraemia due to multidrug-resistant Gram-negative bacilli 27 28 in cancer patients: risk factors, antibiotic therapy and outcomes. J. Antimicrob. 29 Chemother. 66, 657–663 (2010). 30 31 8. Gyssens, I. C., Kern, W. V. & Livermore, D. M. The role of antibiotic stewardship in 32 limiting antibacterial resistance among hematology patients. Haematologica 98, 1821– 33 1825 (2013). 34 35 9. Vialle-Valentin, C. E., Lecates, R. F., Zhang, F., Desta, A. T. & Ross-Degnan, D. 36

Predictors of antibiotic use in African communities: Evidence from medicines http://bmjopen.bmj.com/ 37 38 household surveys in five countries. Trop. Med. Int. Heal. 17, 211–222 (2012). 39 40 10. Paruk, F. et al. Antibiotic prescription practices and their relationship to outcome in 41 South Africa: findings of the prevalence of infection in South African intensive care 42 units (PISA) study. S. Afr. Med. J. 102, 613–6 (2012). 43 44 11. Ingram, P. R., Seet, J. M., Budgeon, C. A. & Murray, R. Point-prevalence study of on September 28, 2021 by guest. Protected copyright. 45 inappropriate antibiotic use at a tertiary Australian hospital. Intern. Med. J. 42, 719– 46 721 (2012). 47 48 49 12. Fridkin, S. K. et al. Vital Signs: Improving Antibiotic Use Among Hospitalized 50 Patients. MMWR. Morb. Mortal. Wkly. Rep. (2014). 51 52 13. Kariuki, S. et al. Increasing prevalence of multidrug-resistant non-typhoidal 53 salmonellae, Kenya, 1994–2003. Int. J. Antimicrob. Agents 25, 38–43 (2005). 54 55 14. Kariuki, S. Situational analysis and recommendations on antimicrobial use and 56 resistance in Kenya. Heal. (San Fr. (2011). 57 58 59 15. Gerding, D. N. The search for good antimicrobial stewardship. Jt. Comm. J. Qual. 60 Improv. 27, 403–404 (2001).

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 15 of 30 BMJ Open

1 2 3

16. Drummond, M. & Jönsson, B. Moving beyond the drug budget silo mentality in BMJ Open: first published as 10.1136/bmjopen-2019-030823 on 31 March 2020. Downloaded from 4 Europe. Value Heal. 6, 74–77 (2003). 5 6 7 17. Richards, P. G. et al. Clinical and economic outcomes from a community hospital’s 8 antimicrobial stewardship program. Am. J. Infect. Control 41, 145–148 (2012). 9 10 18. Dellit, T. H. et al. Infectious Diseases Society of America and the Society for 11 Healthcare Epidemiology of America Guidelines for Developing an Institutional 12 Program to Enhance Antimicrobial Stewardship. Clin. Infect. Dis. 44, 159–177 (2007). 13 14 19. Glasgow, R. E., Vogt, T. M. & Boles, S. M. Evaluating the public health impact of 15 health promotion interventions: The RE-AIM framework. American Journal of Public 16 17 Health 89, 1322–1327 (1999). 18 For peer review only 19 20. CLSI. Performance Standards for Antimicrobial Susceptibility Testing. (2019). 20 21 21. WHO. WHO | Metrics: Disability-Adjusted Life Year (DALY). WHO (2014). 22 Available at: https://www.who.int/healthinfo/global_burden_disease/metrics_daly/en/. 23 (Accessed: 25th February 2019) 24 25 22. Karanika, S., Paudel, S., Grigoras, C., Kalbasi, A. & Mylonakis, E. Systematic review 26 27 and meta-analysis of clinical and economic outcomes from the implementation of 28 hospital-based antimicrobial stewardship programs. Antimicrob. Agents Chemother. 29 60, 4840–4852 (2016). 30 31 23. Borde, J. P. et al. Implementing an intensified antibiotic stewardship programme 32 targeting daptomycin use in orthopaedic surgery: a cost–benefit analysis from the 33 hospital perspective. Infection 44, 301–307 (2016). 34 35 24. Duane, S. et al. Using qualitative insights to change practice: Exploring the culture of 36 37 antibiotic prescribing and consumption for urinary tract infections. BMJ Open 6, http://bmjopen.bmj.com/ 38 (2016). 39 40 25. Kauffmann, R. M. et al. Infection Reduction Strategies Including Antibiotic 41 Stewardship Protocols in Surgical and Trauma Intensive Care Units Are Associated 42 with Reduced Resistant Gram-Negative Healthcare-Associated Infections. Surg. Infect. 43 (Larchmt). 12, 15–25 (2010). 44 on September 28, 2021 by guest. Protected copyright. 45 26. Yong, M. K., Buising, K. L., Cheng, A. C. & Thursky, K. A. Improved susceptibility 46 47 of Gram-negative bacteria in an intensive care unit following implementation of a 48 computerized antibiotic decision support system. J. Antimicrob. Chemother. 65, 1062– 49 1069 (2010). 50 51 52 53 54 55 56 57 58 59 60

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 16 of 30 BMJ Open: first published as 10.1136/bmjopen-2019-030823 on 31 March 2020. Downloaded from

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 For peer review only 17 18 19 20 21 22 23 24 25 26 27 28 29 30 Figure 1: Statistical power analysis for sample size determination for the prevalence of inappropriate use of 31 broad-spectrum antibiotics 32 http://bmjopen.bmj.com/ 33 226x164mm (96 x 96 DPI) 34 35 36 37 38 39 40 on September 28, 2021 by guest. Protected copyright. 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 17 of 30 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2019-030823 on 31 March 2020. Downloaded from

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 For peer review only 17 18 19 20 21 22 23 24 25 26 27 28 29 30 Figure 2: Statistical power analysis for sample size determination for the prevalence of antibiotics resistance 31 32 225x163mm (96 x 96 DPI) http://bmjopen.bmj.com/ 33 34 35 36 37 38 39 40 on September 28, 2021 by guest. Protected copyright. 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 18 of 30 BMJ Open: first published as 10.1136/bmjopen-2019-030823 on 31 March 2020. Downloaded from

1 2 3 4 A survey on Knowledge, Attitudes and Practice about antibiotic prescribing and resistance among 5 6 medical practitioners in Kenyan local hospitals. 7 8 Thank you very much for accepting to participate in this study. 9 10 *You are kindly requested to answer the questionnaire honestly and completely independent of 11 cross-consultations and/or verifications. 12 13 14 Survey quality control 15 16 Date of interview: ……………………………………………….For peer review Start time...... only End time...... 17 18 19 Interviewed by: ...... Approved...... 20 21 Name of the Hospital…………………………………………… Respondent’s code……………………………………………………. 22 23 24 QUESTIONS ANSWERS 25 PART 1: GENERAL QUESTIONS 26 27 1. For how many years, since you graduated from medical school ❖ I am on attachment 28 ❖ I am a trainee in medicine (internship) /medical training College, have you been working in a hospital 29 ❖ Less than one year 30 (indicate cumulative years if worked in different hospitals) ❖ 1-3 years 31 ❖ 32 4 – 6 years http://bmjopen.bmj.com/ 33 ❖ 7 years and more 34 2. In which department do you work? o Medicine /Emergency 35 o Surgery 36 37 o Paediatrics 38 39 o Obstetrics and Gynaecology 40 o on September 28, 2021 by guest. Protected copyright. 41 Outpatient/A/E 42 o Pharmacy 43 44 o Other: ………………………………………………… 45 46 3. Designation (e.g. Consultant, Pharmacist, Nurse, etc.) 47 …………………………………………………………………… 48 PART 2: PRESCRIPTION PATTERN (PRACTICE) 49 50 4. How frequently do you prescribe antibiotics? ❖ More than once daily 51 ❖ Once daily 52 53 ❖ 3 – 5 times a week 54 55 ❖ 1 – 2 times a week 56 57 58 Page 1 of 7 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 19 of 30 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2019-030823 on 31 March 2020. Downloaded from

1 2 3 ❖ less than once a week) 4 5 5. To whom do you prescribe? ❖ Patients at outpatient department 6 7 ❖ Hospitalized patients 8 ❖ Patients in out-patient department and 9 10 hospitalised patients 11 12 6. Do you follow any antibiotic prescription guidelines? ❖ Yes 13 ❖ No 14 15 PART 3: AWARENESS AND ATTITUDE ON THE CURRENT SCOPE OF ANTIBIOTIC RESISTANCE 16 For peer review only 17 7. Antibiotic resistance is a world-wide problem ❖ I strongly agree 18 19 ❖ I agree 20 ❖ Neutral 21 22 ❖ I disagree 23 24 ❖ I strongly disagree 25 8. Antibiotic resistance is a problem in my country ❖ I strongly agree 26 27 ❖ I agree 28 29 ❖ Neutral 30 ❖ I disagree 31 32 ❖ I strongly disagree http://bmjopen.bmj.com/ 33 34 9. Antibiotic resistance is a problem in my hospital ❖ I strongly agree 35 ❖ I agree 36 37 ❖ Neutral 38 39 ❖ I disagree 40 ❖ I strongly disagree on September 28, 2021 by guest. Protected copyright. 41 42 10. Antibiotics are overused in my hospital and in other hospitals ❖ I strongly agree 43 44 of my country Kenya ❖ I agree 45 46 ❖ Neutral 47 ❖ I disagree 48 49 ❖ I strongly disagree 50 51 11. Patients’ demands for antibiotics contribute to the overuse of ❖ I strongly agree 52 antibiotics in the hospital ❖ I agree 53 54 ❖ Neutral 55 56 57 58 Page 2 of 7 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 20 of 30 BMJ Open: first published as 10.1136/bmjopen-2019-030823 on 31 March 2020. Downloaded from

1 2 3 ❖ I disagree 4 5 ❖ I strongly disagree 6 7 12. I think over-the-counter (OTC) medicines contribute to ❖ I strongly agree 8 antibiotic misuse and subsequent antibiotic resistance ❖ I agree 9 10 ❖ Neutral 11 12 ❖ I disagree 13 ❖ I strongly disagree 14 15 13. My awareness on local antibiotic resistance pattern is? ❖ Excellent 16 For peer review only 17 ❖ Good 18 ❖ Average 19 20 ❖ Very little 21 22 ❖ None 23 24 PART 4: CHOICE OF ANTIBIOTIC 25 14. How confident are you about your knowledge of antibiotics? ❖ Very confident 26 27 ❖ Confident 28 29 ❖ A bit confident 30 ❖ Neutral/ I have no idea 31 32 ❖ Not confident at all http://bmjopen.bmj.com/ 33 34 15. What is your confidence level in prescribing antibiotics ❖ Very confident 35 ❖ Confident 36 37 ❖ A bit confident 38 39 ❖ Neutral/ I have no idea 40 ❖ Not confident at all on September 28, 2021 by guest. Protected copyright. 41 42 16. How often do you check your decisions on antibiotic ❖ Never 43 44 prescribing with a colleague? ❖ Sometimes 45 ❖ 46 Half of the times 47 ❖ Mostly 48 49 ❖ Always 50 51 17. If you do consult a senior colleague, how frequent does he/she ❖ Never 52 recommend prescription of a different antibiotic? ❖ Sometimes 53 54 ❖ Half of the times 55 56 57 58 Page 3 of 7 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 21 of 30 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2019-030823 on 31 March 2020. Downloaded from

1 2 3 ❖ Mostly 4 5 ❖ Always 6 7 18. How often do you depend on antibiotic sensitivity data from ❖ Never 8 the laboratory to vary your prescription ❖ Sometimes 9 10 ❖ Half of the times 11 12 ❖ Mostly 13 ❖ Always 14 15 PART 5: SOURCE OF INFORMATION ON ANTIBIOTICS PRESCRIBING AND RESISTANCE 16 For peer review only 17 19. During the past years, how many courses or trainings did you ❖ 0 18 ❖ 19 receive relating to antibiotics? 1-3 20 ❖ 4-6 21 22 ❖ 6-10 23 24 ❖ >10 25 20. Among the sources of information about antibiotics listed below, which one did you consult in the last month? 26 27 ▪ Information supplied by pharmaceutical companies ❖ Yes 28 29 ❖ No 30 31 ▪ Knowledge from training institutions ❖ Yes 32 http://bmjopen.bmj.com/ 33 34 ❖ No 35 ▪ Internet ❖ Yes 36 37 38 ❖ No 39 40 ▪ National guideline for empiric antimicrobial therapy ❖ Yes on September 28, 2021 by guest. Protected copyright. 41 42 ❖ No 43 44 ▪ The World Health Organization’s (WHO) guidelines for ❖ Yes 45 46 treatment of bacterial diseases ❖ No 47 48 21. How do you appreciate the sources of information about antibiotics listed below? 49 50 ▪ Information supplied by pharmaceutical companies ❖ Very useful 51 52 ❖ Useful 53 ❖ Not at all useful 54 55 ❖ I do not know 56 57 58 Page 4 of 7 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 22 of 30 BMJ Open: first published as 10.1136/bmjopen-2019-030823 on 31 March 2020. Downloaded from

1 2 3 ▪ Information from University courses ❖ Very useful 4 5 ❖ Useful 6 7 ❖ Not at all useful 8 ❖ I do not know 9 10 ▪ Internet ❖ Very useful 11 12 ❖ Useful 13 ❖ Not at all useful 14 15 ❖ I do not know 16 For peer review only 17 ▪ National guideline for empiric antimicrobial therapy ❖ Very useful 18 ❖ Useful 19 20 ❖ Not at all useful 21 22 ❖ I do not know 23 24 ▪ The World Health Organization’s (WHO) guidelines for ❖ Very useful 25 treatment of bacterial diseases ❖ Useful 26 27 ❖ Not at all useful 28 29 ❖ I do not know 30 ▪ Does your facility have a frequently released antibiogram? ❖ Yes 31 32 ❖ No http://bmjopen.bmj.com/ 33 34 ▪ If yes, how useful is the antibiogram to you ❖ Very useful 35 ❖ Useful 36 37 ❖ Not at all useful 38 39 ❖ I do not know 40 PART 6: DECISION ABOUT ANTIBIOTIC PRESCRIBING on September 28, 2021 by guest. Protected copyright. 41 42 22. When one prescribes an antibiotic, it is important to know the ❖ I strongly agree 43 44 resistance pattern of the bacteria in the local setting ❖ I agree 45 46 ❖ Neutral 47 ❖ I disagree 48 49 ❖ I strongly disagree 50 51 23. My choice of prescribing antibiotic is more influenced by the ❖ I strongly agree 52 availability of antibiotics than by the cause of the infection ❖ I agree 53 54 ❖ Neutral 55 56 57 58 Page 5 of 7 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 23 of 30 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2019-030823 on 31 March 2020. Downloaded from

1 2 3 ❖ I disagree 4 5 ❖ I strongly disagree 6 7 24. My choice of prescribing antibiotic is more influenced by the ❖ I strongly agree 8 cost of the drug to the patient ❖ I agree 9 10 ❖ Neutral 11 12 ❖ I disagree 13 ❖ I strongly disagree 14 15 25. I’m always concerned about effectiveness and quality of an ❖ I strongly agree 16 For peer review only 17 antibiotic when making my prescribing decisions ❖ I agree 18 ❖ Neutral 19 20 ❖ I disagree 21 22 ❖ I strongly disagree 23 24 26. In regard to antibiotic guidelines, local guidelines are more ❖ I strongly agree 25 useful than international guidelines ❖ I agree 26 27 ❖ Neutral 28 29 ❖ I disagree 30 ❖ I strongly disagree 31 32 27. Antibiotic guidelines and antibiotic committees are rather ❖ I strongly agree http://bmjopen.bmj.com/ 33 34 obstacles than a help ❖ I agree 35 ❖ Neutral 36 37 ❖ I disagree 38 39 ❖ I strongly disagree 40 28. I welcome the implementation of a training program about ❖ I strongly agree on September 28, 2021 by guest. Protected copyright. 41 42 antibiotics ❖ I agree 43 44 ❖ Neutral 45 ❖ I disagree 46 47 ❖ I strongly disagree 48 49 PART 7: KNOWLEDGE ON USE OF ANTIBIOTICS 50 51 29. A 4-year-old child had diarrhoea in the last 4 days (3 stools ❖ Amoxicillin orally 52 daily). She had no fever during the past days nor at ❖ Trimethoprim/sulphamethoxazole orally 53 54 consultation. What is your treatment choice? ❖ Amoxicillin/clavulanic acid orally 55 56 57 58 Page 6 of 7 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 24 of 30 BMJ Open: first published as 10.1136/bmjopen-2019-030823 on 31 March 2020. Downloaded from

Page 25 of 30 BMJ Open

BMJ Open Page 26 of 30

Page 27 of 30 BMJ Open

BMJ Open Page 28 of 30

Page 29 of 30 BMJ Open

1 2 3 STAFFING 4 Q39 How many laboratory technologists are in the station? 5 Q40 How many microbiologists are in the station? 6 Q41 Does your laboratory have staff with Bachelor’s degree qualification or higher? Yes 7 8 Partial 9 No 10 Q42 Do you engage a Consultant clinical microbiologist(s)? Yes 11 Partial 12 For peer reviewNo only 13 CONSUMABLES 14 Q43 How often do you experience unavailability of consumables in Microbiology Yes 15 section? Eg Lack of biochemical reagents and media Partial 16 No 17 Q44 Does your lab experience delays in Antimicrobial Susceptibility Testing Yes http://bmjopen.bmj.com/ 18 due to lack of reagents? Partial 19 No 20 Q45 Do frequent stock outs lead to low demand of cultures by clinicians? Yes 21 Partial 22 No 23 BIOSAFETY 24 25 Q46 Does your lab autoclave/incinerate cultures prior to discard? Yes on September 28, 2021 by guest. Protected copyright. 26 Partial 27 No 28 Q47 Do you have handwashing facility in the laboratory? Yes 29 Partial 30 No 31 Q48 Does your lab get continuous supply of running water? Yes 32 Partial 33 No 34 Q49 Does your lab have soap supply in the handwash facility? Yes 35 Partial 36 No 37 38 39 40 41 42 43 44 45 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 BMJ Open Page 30 of 30 BMJ Open: first published as 10.1136/bmjopen-2019-030823 on 31 March 2020. Downloaded from

1 2 3 Core element 1: Senior hospital management leadership towards Yes No 4 antimicrobial stewardship 5 6 1 Has your hospital management formally identified antimicrobial stewardship 7 as a priority objective for the institution and included it in its key performance 8 indicators? 9 2 Is there dedicated, sustainable and sufficient budgeted financial support for 10 antimicrobial stewardship activities (e.g., support for salary, training, or IT 11 (information technology) support)? 12 3 Does your hospital follow any (national or international) staffing standards for 13 antimicrobial stewardship activities (e.g. number of full-time equivalent (FTE) 14 15 per 100 beds for the different members of the antimicrobial stewardship 16 team)? For peer review only 17 Core element 2: Accountability and responsibilities 18 4 Does your hospital have a formal/written antimicrobial stewardship 19 programme/strategy accountable for ensuring appropriate antimicrobial use? 20 5 Does your hospital have a formal organizational multidisciplinary structure 21 responsible for antimicrobial stewardship (e.g., a committee focused on 22 23 appropriate antimicrobial use, pharmacy committee, patient safety 24 committee or other relevant structure)? 25 6 Is there a healthcare professional identified as a leader for antimicrobial 26 stewardship activities at your hospital and responsible for implementing the 27 programme? 28 7 Is there a document clearly defining roles, procedures of collaboration and 29 responsibilities of the antimicrobial stewardship team members? 30 31 8 Are clinicians, other than those part of the antimicrobial stewardship team 32 (e.g. from the ICU, Internal Medicine and Surgery) involved in the http://bmjopen.bmj.com/ 33 antimicrobial stewardship committee? 34 9 Does the antimicrobial stewardship committee produce regularly (indicate 35 minimum time) a dedicated report which includes e.g. antimicrobial use data 36 and/or prescription improvement initiatives, with time-committed short term 37 and long term measurable goals/ targets for optimizing antimicrobial use? 38 10 Is there a document clearly defining the procedures of collaboration of the 39 40 antimicrobial stewardship team/committee with the infection prevention and on September 28, 2021 by guest. Protected copyright. 41 control team/committee? 42 Core element 3: Available expertise on infection management 43 11 Do you have access to laboratory/imaging services and to timely results to be 44 able to support the diagnosis of the most common infections at your hospital? 45 12 In your hospital are there, or do you have access to, trained and experienced 46 healthcare professionals (medical doctor, pharmacist, nurse …) in infection 47 48 management (diagnosis, prevention and treatment) and stewardship willing 49 to constitute an antimicrobial stewardship team? 50 Core element 4: Education and practical training 51 13 Does your hospital offer a range of educational resources to support staff 52 training on how to optimize antimicrobial prescribing? 53 14 Do the antimicrobial stewardship team members receive regular training in 54 antimicrobial prescribing and stewardship? 55 56 Core element 5: Other actions aiming at responsible antimicrobial use 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 31 of 30 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2019-030823 on 31 March 2020. Downloaded from

Article Type: Protocol

Date Submitted by the 26-Sep-2019 Author:

Complete List of Authors: Gitaka, Jesse; Mount Kenya University, Research and Innovation Kamita, Moses; Mount Kenya University, Research and Innovation Mureithi, Dominic; Maasai Mara University, Department of Animal Health and Production Ndegwa, Davies; Kenya Medical Training College Masika, Moses; University of Nairobi College of Health Sciences, Department of Medical Microbiology, Omuse, Geoffrey; Aga Khan University - Kenya Ngari, Moses; KEMRI/Wellcome Trust, Clinical Trial Facility Makokha, Francis; Mount Kenya University, Research and Innovation Mwaura, Peter; Mount Kenya University, Research and Innovation Mathai, Ronald; Mount Kenya University, Research and Innovation Muregi, Francis; Mount Kenya University, Research and Innovation http://bmjopen.bmj.com/ Mwau, Matilu; Kenya Medical Research Institute

Primary Subject Health economics Heading:

Secondary Subject Heading: Evidence based practice, Medical management

Antimicrobial stewardship, Antimicrobial resistance, Implementation Keywords:

science on September 28, 2021 by guest. Protected copyright.

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 1 of 32 BMJ Open

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 2 of 32

1 2 3

Combating Antibiotic Resistance using Guidelines and Enhanced Stewardship in BMJ Open: first published as 10.1136/bmjopen-2019-030823 on 31 March 2020. Downloaded from 4 Kenya: a protocol for an Implementation Science Approach 5 6 1 1 2 3 4 7 Jesse Gitaka , Moses Kamita , Dominic Mureithi , Davies Ndegwa , Moses Masika , 5 6 1 1 1 8 Geoffrey Omuse , Moses Ngari , Francis Makokha , Peter Mwaura , Ronald Mathai , Francis 9 Muregi1, Matilu Mwau7 10 11 Affiliations 12 1. Directorate of Research and Innovation, Mount Kenya University, P.O. Box 342-01000, 13 Thika, Kenya 14 2. Department of Animal Health and Production, Maasai Mara University, P.O. Box 861- 15 16 20500, Narok, Kenya. 17 3. Department of Medical Laboratory Sciences, Kenya Medical Training College P.O. Box 18 30195-00100, Nairobi,For Kenya peer review only 19 4. Department of Medical Microbiology, College of Health Sciences, University of Nairobi, 20 P.O. Box 30197-00100, Nairobi, Kenya. 21 5. Department of Pathology, Aga Khan University Hospital, P.O. Box 30270-00100, 22 23 Nairobi, Kenya 24 6. KEMRI/Wellcome Trust Research Programme, Centre for Geographic Medicine 25 Research - Coast, P.O. Box 230-80108, Kilifi, Kenya 26 7. Centre for Infectious and Parasitic Diseases Control Research, Kenya Medical Research 27 Institute, P.O. Box 3-50400, Busia, Kenya 28 29 Correspondence: Jesse Gitaka 30 31 [email protected] 32 Directorate of Research and Innovation, Mount Kenya University, P.O. Box 342-01000, 33 Thika, Kenya 34 35 36 37 http://bmjopen.bmj.com/ 38 39 40 41 42 43 44 45 on September 28, 2021 by guest. Protected copyright. 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 3 of 32 BMJ Open

1 2 3

Abstract BMJ Open: first published as 10.1136/bmjopen-2019-030823 on 31 March 2020. Downloaded from 4 5 Introduction 6 7 8 Antibiotic resistance (AMR) is a growing problem globally especially in Sub-Saharan Africa 9 including Kenya. Without any intervention, developing countries will be most affected due to 10 the high burden of diseases. Studies have consistently shown that inappropriate use of 11 antimicrobials is the major drivers of AMR. To address this challenge hospital are now 12 implementing antibiotic stewardship programs (ASPs), which have been observed to reduced 13 antibiotic usage, decrease the prevalence of resistance and lead to significant economic 14 benefits. However, the implementation of the guideline is highly dependent on settings in 15 16 which they are rolled out. This study, employing an implementation science approach aims to 17 address the knowledge and research gap in this area and provide critical data and experiences 18 in using antibiotic guidelinesFor peer and stewardship review programs in onlythe public health sector. This will 19 provide evidence of ASP performance and potentially contribute to the county, national and 20 regional policies on antibiotics use. 21 22 Methods and analysis 23 24 25 The study will be conducted in three geographically diverse regions each represented by two 26 hospitals. A baseline study on antibiotic usage, resistance and de-escalation, duration of 27 hospital stay, rates of readmission and costs will be carried out in the pre-implementation 28 phase. The intervention, that is, the use of antibiotic guidelines and antibiotic stewardship 29 programs will be instituted for 18 months using a stepwise implementation strategy that will 30 facilitate learning and continuous improvement of stewardship activities and updating of 31 guidelines to reflect the evolving antibiotic needs. 32 33 34 Ethics and dissemination 35 36 Approvals to carry out the study have been sought from the National Commission for 37 Science, Technology and Innovation and the Mount Kenya University Ethics Review http://bmjopen.bmj.com/ 38 Committee. In addition, approvals will be sort from each Hospital-based Review Committees 39 where such committees are in place. Study findings will be presented to policy stakeholders 40 and published in peer-reviewed scientific journals. It is anticipated the findings will inform 41 local-based appropriate antibiotic use guidelines. 42 43 44 Key Words 45 on September 28, 2021 by guest. Protected copyright. 46 Antimicrobial stewardship; Implementation science; Antimicrobial resistance 47 48 49 50 51 52 53 54 55 56 57 58 59 60

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 4 of 32

1 2 3

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 5 of 32 BMJ Open

1 2 3

Introduction BMJ Open: first published as 10.1136/bmjopen-2019-030823 on 31 March 2020. Downloaded from 4 5 Antibiotic resistance is a serious public health concern globally1,2 and locally3,4 and fears of 6 7 running out of antibiotics options in the near future have been expressed. In 2016, the World 8 Health Organization (WHO), called for immediate and concerted efforts to mitigate this 9 threat to global health that was estimated to contribute to 700,000 deaths in 2014 and 10 projected to cause 10 million deaths in 2050 if inadequately mitigated.5 The driving force 11 escalating rates of resistance is the injudicious use of antibiotics in patients and in livestock 12 or release into the environment.6 These forces exert selective pressure which rise and increase 13 the spread of resistant pathogens that may emerge by mutations or acquisition of plasmids 14 7 15 carrying resistance genes. In one study, prior antibiotic exposure was the key independent 16 risk factor for the acquisition of antibiotic multi-resistant.8 Broad-spectrum antibiotics have 17 the unintended consequence of selecting multidrug-resistant pathogens and increasing the 18 likelihood of infectionFor by fungi peer and Clostridium review difficile.9 only 19 20 Nonetheless, the injudicious use of antibiotics is not unusual. In Africa, many patients do not 21 receive treatment from the conventional health care system. Of those who receive antibiotics, 22 23 31.7 % do not consult a doctor for a prescription and a further 26.4% obtain the antibiotics 10 24 over-counter. A study in South Africa found that 54.9% of antibiotics were inappropriately 25 prescribed in intensive care unit settings while in the US, 20-50% of prescribed antibiotics 26 were unnecessary or unwarranted.11–13 27 28 The rate of antimicrobial resistance in developing countries Kenya included is worrying and 29 rising.14 In one study, the prevalence of Salmonella typhi resistant to two or more 30 antimicrobials was observed to have increased from 50% in 1998 to 78% in 2004 at Kenyatta 31 15 32 National Hospital. The Global Antibiotic Resistance Partnership – Kenya Working Group 33 Report of 2011 identified antibiotic resistance as a key issue in Kenya and made 34 recommendations to curtail the spread.16 These recommendations included the use of 35 antibiotic guidelines that took into consideration local resistance surveillance data and 36 enhanced antibiotic stewardship programs (ASPs).16 Even so, these ASPs have not been http://bmjopen.bmj.com/ 37 instituted at county hospitals, and key implementation data and experiences are lacking in 38 their roll out. 39 40 41 Antibiotic stewardship is defined as the optimum selection, dosage, and duration of 42 antimicrobial treatment that yields the best clinical outcomes for the treatment or prevention 43 of infection with the least toxicity to the patient and minimal impact on subsequent 44 resistance.17 It has the potential to lower treatment costs and realize economic benefits to the on September 28, 2021 by guest. Protected copyright. 45 patient, health care system and the country at large.18,19 Moreover, optimizing antibiotic use 46 by minimizing exposure, fine-tuning dosage and reducing superfluous therapy and focusing 47 20 48 treatment to the likely culprit pathogens is a strategy that boosts patient safety and 21 49 ultimately safeguards against antibiotic resistance. 50 51 Justification 52 53 Antibiotic resistance is a major health challenge globally. Studies in Kenya have shown rising 54 antibiotic resistance over the last 3 decades. Nonetheless, antibiotic guidelines and antibiotic 55 stewardship programs which have been observed to lead to significant economic benefits, 56 reduce antibiotic usage and lower the prevalence of resistance especially in Europe, North 57 58 America, Japan, and South Africa have not been employed to tackle the challenge in the 59 public health sector in Kenya. The GARP report of 2011, recognized use of guidelines and 60 stewardship programs as a potential strategy in ‘saving antibiotics’ but noted the need for

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 6 of 32

1 2 3

local studies to understand implementation challenges, derive lessons and embed the strategy BMJ Open: first published as 10.1136/bmjopen-2019-030823 on 31 March 2020. Downloaded from 4 within the Kenyan health care system. This study, employing an implementation science 5 6 approach aims to address the knowledge and research gap in this area and provide critical 7 data and experiences in using antibiotic guidelines and stewardship programs in the public 8 health sector. 9 10 Objectives of the Study 11 12 General Objectives 13 14 The overall aim is to evaluate the impact of antibiotic guidelines for empirical treatment of 15 urinary tract infections, community-acquired pneumonia, bacteremia and meningitis and 16 17 antibiotic stewardship program on reducing usage of broad-spectrum antibiotics, antibiotic 18 de-escalation, durationFor of hospitalpeer stays, review rates of readmission only and prevalence of antibiotic 19 resistance in 6 county hospitals in Kenya. 20 21 Specific Objectives 22 23 1. To develop guidelines for antibiotic use for common infections. 24 2. To set up antibiotic steward committees (ASCs) in 6 county hospitals. 25 26 3. To train the ASCs to be able to perform their mandate. 27 4. To measure the usage of broad-spectrum antibiotics, antibiotic de-escalation, duration 28 of hospital stays and rates of readmission in hospitals using the antibiotic stewardship 29 strategy. 30 5. To ascertain antibiotic resistance patterns using culture, sensitivity, and genetic 31 markers. 32 6. To establish the health care workers knowledge, attitudes and prescription practices 33 34 regarding antibiotics resistance, use of guidelines and ASPs. 35 7. To evaluate the economic benefits of using guidelines for antibiotics use and ASPs. 36 37 Expected Outputs of the research http://bmjopen.bmj.com/ 38 39 1. Antibiotic guidelines informed by local bacteria susceptibility patterns. 40 2. Antibiotic Stewardship Committees (ASCs) in 6 County hospitals in Kenya. 41 3. Trained ASCs 42 4. Data on antibiotics usage, antibiotic de-escalation, duration of hospital stay and rates 43 44 of readmission in hospitals using the antibiotic stewardship strategy 45 5. Map and data of antibiotic resistance pattern before and after implementation of the on September 28, 2021 by guest. Protected copyright. 46 strategy. 47 6. Qualitative data on health care workers’ knowledge, attitudes, and practices on the 48 antibiotic usage and stewardship programs 49 7. Publication on cost-benefit analysis for using antibiotic guidelines and ASPs 50 51 52 Methods and analysis 53 54 Setting 55 56 To be able to evaluate the antibiotic stewardship strategy across Kenya using an 57 implementation science approach, the study will be conducted in 6 county hospitals (Kiambu 58 County Referral Hospital, Bungoma County Hospital, Webuye Sub-County Hospital, Nakuru 59 60

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 7 of 32 BMJ Open

1 2 3

County Teaching and Referral Hospital, Thika Level 5 Hospital and Naivasha Sub-County BMJ Open: first published as 10.1136/bmjopen-2019-030823 on 31 March 2020. Downloaded from 4 Hospital). 5 6 7 Design 8 9 The study will utilize the reach, effectiveness, adoption, implementation and maintenance (RE-AIM) 10 conceptual framework. RE-AIM involves the interaction of all the five factors and fits well with 11 approaches that are system based22. The study design will encompass a pre-implementation phase, 12 a stepwise implementation phase, and an endline study to measure the changes in outcomes 13 between the phases. A baseline study on antibiotic usage (Defined Daily Doses per 1000), 14 antibiotic resistance (culture and genetic markers), antibiotic de-escalation, duration of 15 hospital stays, rates of readmission, prescription patterns, and costs analysis will be carried 16 17 out in the pre-implementation phase. The intervention, which is the use of antibiotic 18 guidelines and antibioticFor stewardship peer programs, review will be instituted only in the 6 county hospitals for 19 18 months. The stepwise implementation phase will involve introduction of interventions as 20 well as monitoring and evaluation of their effectiveness and thereafter improving the 21 interventions where necessary. The strategy will facilitate learning cycles every 4-6 weeks 22 and continuous improvement of stewardship activities and updating of guidelines to reflect 23 evolving antibiotic needs in diverse settings. During the intervention phase, challenges faced 24 25 during the implementation of the guidelines will be documented and used to advice on the 26 changes that need to be made during the review of the guidelines. The end line study will be 27 conducted and differences between the 2 phases evaluated as below (Statistical analysis). 28 29 In the baseline and endline studies, multidisciplinary strategies will be employed as 30 follows 31 i. Health care workers knowledge, attitudes, and practices on antibiotic resistance, 32 guidelines, and ASP will be studied qualitatively and quantitatively. 33 34 ii. Basic science approaches encompassing antibiotics culture sensitivity and molecular 35 biology-genetic markers of resistance will be analyzed as detailed below. 36 iii. Clinical- patient outcomes will be studied to evaluate the guidelines. 37 iv. Health economics-cost savings on using guidelines and ASP will be evaluated as http://bmjopen.bmj.com/ 38 below. 39 40 Antibiotic guidelines and ASPs: Development 41 42 43 Antibiotic guidelines will be formulated in consultation with senior clinicians in the study 44 hospitals taking into account each hospital’s antimicrobial resistance patterns. ASPs 45 committee will comprise of the hospital physician, microbiologist, and pharmacist. Broad on September 28, 2021 by guest. Protected copyright. 46 spectrum antibiotic prescriptions will be brought to the attention of the ASP committees who 47 will also perform regular ward rounds three times a week in the initial stages and later once a 48 week to optimize adherence to antibiotic guidelines. Furthermore, the guidelines will be 49 promoted through teaching sessions, provision of pocket-size guideline cards to clinicians 50 51 and pharmacists, large poster displays in the wards and through hospital and project websites. 52 53 Data collection 54 55 Laboratory assessment tools will be used to determine the preparedness of the laboratories to 56 perform antibiotic sensitivity tests (see Additional file 1). The tool will assess whether the 57 laboratories have the equipment that are necessary to perform bacterial culture, identification, 58 and 59 susceptibility testing such as the carbon dioxide incubators, safety cabinets, and refrigerators 60

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 8 of 32

1 2 3

among others. The tool will also determine whether the laboratories have the internationally BMJ Open: first published as 10.1136/bmjopen-2019-030823 on 31 March 2020. Downloaded from 4 recommended guidelines to perform susceptibility and quality assurance tests. In addition, 5 6 knowledge, attitude and practices (KAP) about antibiotic prescribing and resistance among 7 medical practitioners will be assessed using a KAP tool (See Additional file 2). The tool will 8 target those medical practitioners who usually prescribe antimicrobial drugs in the hospitals 9 and will include the consultants, medical officers and the interns, clinical officers and the 10 interns, as well as the pharmacists. The KAP tool will also determine the guidelines that the 11 prescribers use to decide on the appropriate antimicrobial drug. A health system assessment 12 tool will be used to gather baseline information on the antimicrobial stewardship activities in 13 14 each hospital (see Additional file 3). Information on each hospital bed occupancy, antibiotic 15 usage, and antibiotic resistance data will be collected from hospital information systems, 16 pharmacy management systems and laboratory reporting systems respectively. Antibiotic 17 usage will be analyzed in order to determine the impact of the intervention on antibiotic 18 usage comparing theFor intervention peer and the review control arms. Data only on antibiotics issued to adult and 19 children inpatients only will be factored excluding discharge and outpatient supplies. 20 21 22 To consistently compare antibiotic usage, the defined daily doses (DDD) will be used and 23 expressed per 1000 occupied bed days (OBDs) to account for fluctuations in activity 24 following WHO guidelines.23 The OBDs will be obtained from the hospital information 25 systems. 26 27 Culture sensitivity and genetic analyses 28 29 Nasal swabs or tracheal aspirate, urine, wound swabs and blood samples will be taken from 30 patients who have consented to the study for bacteriological analysis. Samples will be 31 32 subjected to standard bacteriological analysis to isolate the culprit bacteria. Bacterial species 33 will be confirmed by use of biochemical test and analytical profiles index API strips (Bio 34 Merieux France). Antimicrobial susceptibility test will be performed on isolated bacteria as 35 per the Kirby-Bauer Method following manufacturer’s instruction. Results will be interpreted 36 using the Clinical and Laboratory Standards Institute (CLSI) tables.24 37 http://bmjopen.bmj.com/ 38 Any bacteria isolate found to be resistant to third generation cephalosporin’s will be tested for 39 production of extended spectrum Beta-lactamase (ESBLs) using the synergy disk diffusion 40 41 test. Vancomycin Resistance Enterococci (VRE) will be identified using disc diffusion tests. 42 Methicillin resistance in S. aureus (MRSA) will be detected by testing isolates resistant to 43 cefoxitin by E test (AB Biodisk, Solna, Sweden) on Mueller–Hinton agar supplemented with 44 2% NaCl and incubated at 37°C for 24 h. The identified ESBLs, VRE and MRSA will be on September 28, 2021 by guest. Protected copyright. 45 analyzed by PCR and sequencing to identify the resistance genotype. In vitro conjugation 46 tests will be performed to determine if resistance in bacteria is transferable. 47 48 49 Cost-benefit analysis for the use of antibiotic guidelines and Antibiotic Stewardship 50 Program 51 52 A cost-benefit analysis (CBA) from a health facility and a national perspective will be 53 performed. For health facilities, three cost drivers will be considered: pharmacy spending, 54 length of stay, and antimicrobial stewardship interventions (training, infection control 55 measures, etc.). For the country, we will consider, Disability-Adjusted Life years (DALYS), 56 work-days lost, and cost of treatment.25–27 This will be done by collecting and analyzing data 57 58 on patient income, length of hospital stay, death or disability occasioned by drug-resistant 59 pathogens, hospital pharmacy expenditure and cost of training/rolling out antimicrobial 60

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 9 of 32 BMJ Open

1 2 3

stewardship guidelines. This data will be collected before and after antimicrobial stewardship BMJ Open: first published as 10.1136/bmjopen-2019-030823 on 31 March 2020. Downloaded from 4 interventions. 5 6 7 Study size 8 9 Prevalence of inappropriate use of broad-spectrum antibiotics 10 11 The sample size is based on the prevalence of inappropriate use of broad-spectrum 12 antibiotics of ~50%11 in South Africa. To detect a reduction of the 50% inappropriate use of 13 broad-spectrum antibiotics by 20% to 30% with a power of 90% and α of 0.05 in a two-sided 14 test, a sample size of 410 in each county hospital would be adequate. In order to detect a 15 reduction (20%) but with a power 80% the sample size would be 320 as shown in Figure 1 16 17 below. With a 15% allowance for loss to follow up, a total of about 500 participants would be 18 enough in each of theFor selected peer county hospitals. review Therefore, only we shall evaluate the prevalence 19 of inappropriate use of broad-spectrum antibiotics in 500 x 6= 3,000 patients from the 6 20 facilities. 21 22 Prevalence of antibiotic resistance 23 24 The sample size is based on the prevalence of antibiotic resistance of 78%15 in 2004 at 25 26 Kenyatta National Hospital, Kenya. To detect a reduction of the 78% antibiotic resistance by 27 10% to 68% with a power of 90% and α of 0.05 two-sided test, a sample size of 500 in each 28 county hospital would be adequate. While to achieve a reduction (10%) but with power, 80% 29 the sample size would be 220 as shown in Figure 2 below. With 15% allowance for loss to 30 follow up, a total of 600 participants would be enough in each of the selected county 31 hospitals. Therefore, we shall evaluate the prevalence of resistance in 600 x 6 = 3,600 32 patients for the 6 facilities. 33 34 35 Data Management 36 37 Data will be recorded on a standardized Case Report Form (CRF) at every hospital by the http://bmjopen.bmj.com/ 38 trained study staff. The data will be held locally and uploaded to the secure central study 39 server hosted at the Mount Kenya University main campus in Thika. This will be overseen by 40 the Data Manager/statistician who will run regular reconciliation to derive the final study 41 database. Access to the study database will be restricted and password protected. 42 43 Statistical analyses 44 45 on September 28, 2021 by guest. Protected copyright. 46 The data will be analyzed using qualitative and quantitative methods. Qualitative data will be 47 analyzed by subjecting the information to content analysis and presenting it in different 48 emerging themes. The summaries of the data emanating from these themes will then be 49 arranged on a case by case basis through the use of an Excel spreadsheet28 and the analyses 50 done by using NVIVO software. The quantitative data analyses will be done using Stata 51 version 14 (Stata, Inc.). The differences between baseline and endline on all the study 52 53 outcomes will be compared using appropriate statistical methods like McNemar’s test, paired 54 t-test and the Wilcoxon signed-rank test non-parametric test accounting for pairing and 55 clusters (hospitals). Multivariable log-binomial regression analysis will be used to get risk 56 factors for antibiotic resistance. 57 58 Ethical considerations 59 60

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 10 of 32

1 2 3

Approvals to carry out the study have been sought from the National Commission for BMJ Open: first published as 10.1136/bmjopen-2019-030823 on 31 March 2020. Downloaded from 4 Science, Technology and Innovation (NACOSTI) (NACOSTI/P/18/33304/25986) and the 5 6 Mount Kenya University Ethics Review Committee (MKU/ERC/0764). In addition, 7 approvals will be sort from each Hospital-based Review Committees where such committees 8 are in place. The study will follow all provisions of the Declaration of Helsinki. Participants 9 in the study will not incur any cost in the transport nor processing of the samples, neither will 10 they receive any monetary inducements to participate in the study. Material transfer to 11 laboratories outside of Kenya shall not be undertaken in this study. Informed written consent 12 will be sought from the participants enrolled in the study. 13 14 15 Exit strategy and stakeholder involvement 16 17 In the process of developing this protocol, we have engaged physicians working in the 18 proposed County hospitalsFor andpeer they have review identified the challengeonly of antibiotic resistance as 19 real and appreciate the opportunities that use of antibiotic guidelines and ASPs may provide 20 in combating resistance, improving clinical care and saving costs. We plan to continue 21 involving all the stakeholders in the process who include clinicians, laboratory personnel, 22 23 pharmacists, public health officials, patients and scientists in the arena. We anticipate that the 24 study findings will inform county and national policy on mitigating antibiotic resistance and 25 raise public awareness on the need for judicious use of antibiotics. The project will use social 26 media platforms, websites of the collaborating institutions, and publications in peer-reviewed 27 journals, local dailies and presentations in scientific meetings to further engage stakeholders 28 and the public on this important issue and enhance the learning approach inherent in the 29 strategy of implementation science for improved performance of the ASPs and the study in 30 31 general. 32 33 Dissemination 34 35 Obtained results will be disseminated at different platforms which include research 36 conferences and in peer-reviewed journals. It is hoped that five publications will be generated 37 by the end of the study. The publications will cover areas ranging from the study protocol http://bmjopen.bmj.com/ 38 itself to the different aspects that will be focused on in the course of the project. 39 Contextualized guidelines on judicious use of antibiotics in the six hospitals in Kenya will 40 41 also be published for easy access by other health facility in and outside Kenya. In addition, 42 the findings will be shared in a dissemination forum bring together members of the health 43 management teams at both the country and county levels, clinicians who do prescription of 44 antimicrobial drugs, researchers and other key stakeholders. 45 on September 28, 2021 by guest. Protected copyright. 46 Discussion 47 48 The emergence of antimicrobial-resistant pathogens and a lack of new drugs to effectively 49 50 treat these pathogens are the two main challenges in human health. There is thus the need to 51 advocate for proper use of the currently available antimicrobial agents by safeguarding their 52 effectiveness. Antibiotic stewardship has been shown to contribute to reducing antibiotic 53 resistance29,30 but this strategy has not been rolled out in most sub-Saharan countries in level 54 4 and 5 hospitals. The proposed work will employ an implementation research approach to 55 evaluate the best strategies and derive lessons on mainstreaming antibiotic stewardship in 56 these facilities. By leveraging on the health system approach, the implementation research 57 58 will unmask real life impediments and opportunities and working with hospital teams co- 59 design execution plans, monitoring and evaluation and sustainability of the stewardship 60 programs.

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 11 of 32 BMJ Open

1 2 3

Most hospitals in Kenya lack the capacity to do antimicrobial sensitivity tests due to the lack BMJ Open: first published as 10.1136/bmjopen-2019-030823 on 31 March 2020. Downloaded from 4 of enough resources and technical knowhow among other challenges.31 This study will first 5 6 do an assessment to determine the challenges that the hospitals are facing through interviews 7 with clinicians and assessment of the capacity of the laboratories to perform the sensitivity 8 tests. These steps will make up the initial phase and will guide the nature of implementations 9 to be used during the implementation phase. It is hoped that the project will capacitate the 10 laboratories in the six hospitals to do the tests and sensitize the clinicians on the need to 11 prescribe antimicrobial drugs based on results obtained from the laboratory tests. Once the 12 implementation phase is done, the final phase will be the endline phase where surveys similar 13 14 to the ones that were conducted during the baseline phase will be done in order to determine 15 the impact of the implementation strategies taken in reducing the cost of treatment, days of 16 stay in the hospital and burden on the health system among others. 17 18 The hospital managementFor plays peer a key rolereview in determining only the allocation of resources to the 19 hospitals.32 In this project we have proposed to involve the hospital management by including 20 its members in the stewardship committees that will be established. This will ensure that the 21 22 management is well informed about the challenges that are in the sections that are involved in 23 surveillance and the progress that is being made. Involvement of the hospital management 24 will also ensure that there is a buy-in of the recommendations made by the stewardship 25 committee. 26 27 Although there is the need to establish antimicrobial stewardship committees in all the 28 hospitals, the project proposes to start with the selected six hospitals with the hope that the 29 same can be reproduced by other hospitals to establish similar committees in their hospitals. 30 31 32 Study timeline 33 34 The study is designed to take three years to complete. The study started enrolling on 2018 35 and participant recruitment will continue up to 2020 (Table 1). So far, stewardship 36

committees have been established and workshops to sensitize the members conducted. http://bmjopen.bmj.com/ 37 Currently, data collection on knowledge attitude and practice by antibiotic prescribers in the 38 39 hospitals is going on. 40 41 Table 1: Proposed study timelines. 42 2019 2020 2021 43 44 45 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 on September 28, 2021 by guest. Protected copyright. 46 Activity 47 48 Ethics clearance and 49 50 authorization from study 51 sites 52 53 Baseline study on 54 55 prevalence of antibiotic 56 resistance and 57 58 stewardship activities 59 60

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 12 of 32

1 2 3

Identification of members BMJ Open: first published as 10.1136/bmjopen-2019-030823 on 31 March 2020. Downloaded from 4 5 for the antibiotic 6 stewardship programs 7 8 Development of antibiotic 9 guidelines 10 11 Adopting and preparing 12 materials for ASP training 13 14 Development of antibiotic 15 guidelines and 16 17 stewardship mobile 18 application For peer review only 19 20 Training the ASP 21 members 22 23 Collecting data on 24 antibiotic usage in the 25 hospitals 26 27 Conducting KAPs study 28 29 on use of guidelines and 30 ASPs 31 Analyze and present 32 antibiotic resistance data 33 using maps and charts 34 Presenting findings to 35 36 stakeholders and 37 preparing the final report http://bmjopen.bmj.com/ 38 Patient and Public Involvement 39 40 No patient involved. 41 42 Availability of data and material 43 44 All data sets will be available to the public upon request. 45 on September 28, 2021 by guest. Protected copyright. 46 Acknowledgment 47 Not applicable 48 49 Additional files 50 51 Additional file 1: Laboratory preparedness assessment tool 52 53 Additional file 2: Laboratory knowledge, attitude and practice (KAP) tool 54 55 Additional file 3: Health system tool 56 57 Figure Legends 58 59 60

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 13 of 32 BMJ Open

1 2 3

Figure 1: Statistical power analysis for sample size determination for the prevalence of BMJ Open: first published as 10.1136/bmjopen-2019-030823 on 31 March 2020. Downloaded from 4 inappropriate use of broad-spectrum antibiotics 5 6 Figure 2: Statistical power analysis for sample size determination for the prevalence of 7 8 antibiotics resistance 9 10 11 12 13 14 15 16 17 18 For peer review only 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 http://bmjopen.bmj.com/ 38 39 40 41 42 43 44 45 on September 28, 2021 by guest. Protected copyright. 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 14 of 32

1 2 3

References BMJ Open: first published as 10.1136/bmjopen-2019-030823 on 31 March 2020. Downloaded from 4 5 1. Aslam, B. et al. Antibiotic resistance: a rundown of a global crisis. Infect. Drug Resist. 6 7 11, 1645–1658 (2018). 8 9 2. Livermore, D. M. Has the era of untreatable infections arrived? J. Antimicrob. 10 Chemother. 64, 29–36 (2009). 11 12 3. Bejon, P. et al. Invasive Gram-negative bacilli are frequently resistant to standard 13 antibiotics for children admitted to hospital in Kilifi, Kenya. J. Antimicrob. 14 Chemother. 56, 232–235 (2005). 15 16 4. Dougan, G. et al. Typhoid in Kenya Is Associated with a Dominant Multidrug- 17 18 Resistant SalmonellaFor entericapeer Serovar review Typhi Haplotype only That Is Also Widespread in 19 Southeast Asia. J. Clin. Microbiol. 48, 2171–2176 (2010). 20 21 5. WHO. United Nations meeting on antimicrobial resistance. (2016). 22 23 6. Boxall, A. et al. The role of the natural environment in the emergence of antibiotic 24 resistance in Gram- negative bacteria The role of the natural environment in the 25 emergence of antibiotic resistance in Gram-negative bacteria. Lancet Infect. Dis. 13, 26 27 155–165 (2017). 28 29 7. Baquero, F. et al. The global threat of antimicrobial resistance: science for 30 intervention. New Microbes New Infect. 6, 22–29 (2015). 31 32 8. Sanchez-Ortega, I. et al. Bacteraemia due to multidrug-resistant Gram-negative bacilli 33 in cancer patients: risk factors, antibiotic therapy and outcomes. J. Antimicrob. 34 Chemother. 66, 657–663 (2010). 35 36

9. Gyssens, I. C., Kern, W. V. & Livermore, D. M. The role of antibiotic stewardship in http://bmjopen.bmj.com/ 37 38 limiting antibacterial resistance among hematology patients. Haematologica 98, 1821– 39 1825 (2013). 40 41 10. Vialle-Valentin, C. E., Lecates, R. F., Zhang, F., Desta, A. T. & Ross-Degnan, D. 42 Predictors of antibiotic use in African communities: Evidence from medicines 43 household surveys in five countries. Trop. Med. Int. Heal. 17, 211–222 (2012). 44 on September 28, 2021 by guest. Protected copyright. 45 11. Paruk, F. et al. Antibiotic prescription practices and their relationship to outcome in 46 South Africa: findings of the prevalence of infection in South African intensive care 47 48 units (PISA) study. S. Afr. Med. J. 102, 613–6 (2012). 49 50 12. Ingram, P. R., Seet, J. M., Budgeon, C. A. & Murray, R. Point-prevalence study of 51 inappropriate antibiotic use at a tertiary Australian hospital. Intern. Med. J. 42, 719– 52 721 (2012). 53 54 13. Fridkin, S. K. et al. Vital Signs: Improving Antibiotic Use Among Hospitalized 55 Patients. MMWR. Morb. Mortal. Wkly. Rep. (2014). 56 57 58 14. Ayukekbong, J. A., Ntemgwa, M. & Atabe, A. N. The threat of antimicrobial 59 resistance in developing countries: Causes and control strategies. Antimicrob. Resist. 60 Infect. Control 6, 1–8 (2017).

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 15 of 32 BMJ Open

1 2 3

15. Kariuki, S. et al. Increasing prevalence of multidrug-resistant non-typhoidal BMJ Open: first published as 10.1136/bmjopen-2019-030823 on 31 March 2020. Downloaded from 4 salmonellae, Kenya, 1994–2003. Int. J. Antimicrob. Agents 25, 38–43 (2005). 5 6 7 16. Global Antibiotic Resistance Partnership - Kenya Working. Situational analysis and 8 recommendations on antimicrobial use and resistance in Kenya. Center for Disease 9 Dynamics, Economics & Policy (2011). 10 11 17. Gerding, D. N. The search for good antimicrobial stewardship. Jt. Comm. J. Qual. 12 Improv. 27, 403–404 (2001). 13 14 18. Drummond, M. & Jönsson, B. Moving beyond the drug budget silo mentality in 15 Europe. Value Heal. 6, 74–77 (2003). 16 17 18 19. Richards, P. ForG. et al. Clinicalpeer and economicreview outcomes only from a community hospital’s 19 antimicrobial stewardship program. Am. J. Infect. Control 41, 145–148 (2012). 20 21 20. Dellit, T. H. et al. Infectious Diseases Society of America and the Society for 22 Healthcare Epidemiology of America Guidelines for Developing an Institutional 23 Program to Enhance Antimicrobial Stewardship. Clin. Infect. Dis. 44, 159–177 (2007). 24 25 21. Bogaert, D. & van Belkum, A. Antibiotic treatment and stewardship in the era of 26 27 microbiota-oriented diagnostics. European Journal of Clinical Microbiology and 28 Infectious Diseases 37, 795–798 (2018). 29 30 22. Glasgow, R. E., Vogt, T. M. & Boles, S. M. Evaluating the public health impact of 31 health promotion interventions: The RE-AIM framework. American Journal of Public 32 Health 89, 1322–1327 (1999). 33 34 23. World Health Organization. Essential medicines and health products. (2019). 35 Available at: https://www.who.int/medicines/regulation/medicines- 36 37 safety/toolkit_indicators/en/index2.html. (Accessed: 21st September 2019) http://bmjopen.bmj.com/ 38 39 24. CLSI. Performance Standards for Antimicrobial Susceptibility Testing. (2019). 40 41 25. Karanika, S., Paudel, S., Grigoras, C., Kalbasi, A. & Mylonakis, E. Systematic review 42 and meta-analysis of clinical and economic outcomes from the implementation of 43 hospital-based antimicrobial stewardship programs. Antimicrob. Agents Chemother. 44

60, 4840–4852 (2016). on September 28, 2021 by guest. Protected copyright. 45 46 26. Borde, J. P. et al. Implementing an intensified antibiotic stewardship programme 47 48 targeting daptomycin use in orthopaedic surgery: a cost–benefit analysis from the 49 hospital perspective. Infection 44, 301–307 (2016). 50 51 27. WHO. WHO | Metrics: Disability-Adjusted Life Year (DALY). WHO (2014). 52 Available at: https://www.who.int/healthinfo/global_burden_disease/metrics_daly/en/. 53 (Accessed: 25th February 2019) 54 55 28. Duane, S. et al. Using qualitative insights to change practice: Exploring the culture of 56 57 antibiotic prescribing and consumption for urinary tract infections. BMJ Open 6, 58 (2016). 59 60 29. Kauffmann, R. M. et al. Infection Reduction Strategies Including Antibiotic

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 16 of 32

1 2 3

Stewardship Protocols in Surgical and Trauma Intensive Care Units Are Associated BMJ Open: first published as 10.1136/bmjopen-2019-030823 on 31 March 2020. Downloaded from 4 with Reduced Resistant Gram-Negative Healthcare-Associated Infections. Surg. Infect. 5 6 (Larchmt). 12, 15–25 (2010). 7 8 30. Yong, M. K., Buising, K. L., Cheng, A. C. & Thursky, K. A. Improved susceptibility 9 of Gram-negative bacteria in an intensive care unit following implementation of a 10 computerized antibiotic decision support system. J. Antimicrob. Chemother. 65, 1062– 11 1069 (2010). 12 13 31. Odhiambo, F. et al. Antimicrobial resistance: capacity and practices among clinical 14 laboratories in Kenya, 2013. Pan Afr. Med. J. 19, 332 (2014). 15 16 17 32. Barasa, E. W., Molyneux, S., English, M. & Cleary, S. Setting healthcare priorities in 18 hospitals: a reviewFor of peerempirical studies. review Health Policy only Plan. 30, 386–96 (2015). 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 http://bmjopen.bmj.com/ 38 39 40 41 42 43 44 45 on September 28, 2021 by guest. Protected copyright. 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 17 of 32 BMJ Open

1 2 3

Authors’ Contributions BMJ Open: first published as 10.1136/bmjopen-2019-030823 on 31 March 2020. Downloaded from 4 5 JG initiated the concepts of the study. JG, DM, DN, MMasika, GO, MN, FMakokha, PM, 6 RM, FMuregi, and MMwau developed the protocol. JG, DM, DN, and MMasika will be 7 8 involved in training the ASCs. DM, GO, MN, FMakokha, PM, RM, DN, and MK will help in 9 data collection. MN and MK will be involved in data analysis. MK wrote the first draft of the 10 manuscript. All authors reviewed and approved the protocol. 11 12 Funding Statement 13 14 This work was supported by The Kenya National Research Fund (grant number 15 16 NRF/MKU/2017/007 to JG). The funders had no role in study design, decision to publish, or 17 preparation of the manuscript. 18 For peer review only 19 Competing Interests 20 21 The authors declare no conflict of interest. 22 23 24 Word count: 3466 words 25 26 27 28 29 30 31 32 33 34 35 36 37 http://bmjopen.bmj.com/ 38 39 40 41 42 43 44 45 on September 28, 2021 by guest. Protected copyright. 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 18 of 32 BMJ Open: first published as 10.1136/bmjopen-2019-030823 on 31 March 2020. Downloaded from

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 For peer review only 17 18 19 20 21 22 23 24 25 26 27 28 29 30 Figure 1: Statistical power analysis for sample size determination for the prevalence of inappropriate use of 31 broad-spectrum antibiotics 32 http://bmjopen.bmj.com/ 33 226x164mm (96 x 96 DPI) 34 35 36 37 38 39 40 on September 28, 2021 by guest. Protected copyright. 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 19 of 32 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2019-030823 on 31 March 2020. Downloaded from

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 For peer review only 17 18 19 20 21 22 23 24 25 26 27 28 29 30 Figure 2: Statistical power analysis for sample size determination for the prevalence of antibiotics resistance 31 32 225x163mm (96 x 96 DPI) http://bmjopen.bmj.com/ 33 34 35 36 37 38 39 40 on September 28, 2021 by guest. Protected copyright. 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 20 of 32 BMJ Open: first published as 10.1136/bmjopen-2019-030823 on 31 March 2020. Downloaded from

1 2 3 4 A survey on Knowledge, Attitudes and Practice about antibiotic prescribing and resistance among 5 6 medical practitioners in Kenyan local hospitals. 7 8 Thank you very much for accepting to participate in this study. 9 10 *You are kindly requested to answer the questionnaire honestly and completely independent of 11 cross-consultations and/or verifications. 12 13 14 Survey quality control 15 16 Date of interview: ……………………………………………….For peer review Start time...... only End time...... 17 18 19 Interviewed by: ...... Approved...... 20 21 Name of the Hospital…………………………………………… Respondent’s code……………………………………………………. 22 23 24 QUESTIONS ANSWERS 25 PART 1: GENERAL QUESTIONS 26 27 1. For how many years, since you graduated from medical school ❖ I am on attachment 28 ❖ I am a trainee in medicine (internship) /medical training College, have you been working in a hospital 29 ❖ Less than one year 30 (indicate cumulative years if worked in different hospitals) ❖ 1-3 years 31 ❖ 32 4 – 6 years http://bmjopen.bmj.com/ 33 ❖ 7 years and more 34 2. In which department do you work? o Medicine /Emergency 35 o Surgery 36 37 o Paediatrics 38 39 o Obstetrics and Gynaecology 40 o on September 28, 2021 by guest. Protected copyright. 41 Outpatient/A/E 42 o Pharmacy 43 44 o Other: ………………………………………………… 45 46 3. Designation (e.g. Consultant, Pharmacist, Nurse, etc.) 47 …………………………………………………………………… 48 PART 2: PRESCRIPTION PATTERN (PRACTICE) 49 50 4. How frequently do you prescribe antibiotics? ❖ More than once daily 51 ❖ Once daily 52 53 ❖ 3 – 5 times a week 54 55 ❖ 1 – 2 times a week 56 57 58 Page 1 of 7 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 21 of 32 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2019-030823 on 31 March 2020. Downloaded from

1 2 3 ❖ less than once a week) 4 5 5. To whom do you prescribe? ❖ Patients at outpatient department 6 7 ❖ Hospitalized patients 8 ❖ Patients in out-patient department and 9 10 hospitalised patients 11 12 6. Do you follow any antibiotic prescription guidelines? ❖ Yes 13 ❖ No 14 15 PART 3: AWARENESS AND ATTITUDE ON THE CURRENT SCOPE OF ANTIBIOTIC RESISTANCE 16 For peer review only 17 7. Antibiotic resistance is a world-wide problem ❖ I strongly agree 18 19 ❖ I agree 20 ❖ Neutral 21 22 ❖ I disagree 23 24 ❖ I strongly disagree 25 8. Antibiotic resistance is a problem in my country ❖ I strongly agree 26 27 ❖ I agree 28 29 ❖ Neutral 30 ❖ I disagree 31 32 ❖ I strongly disagree http://bmjopen.bmj.com/ 33 34 9. Antibiotic resistance is a problem in my hospital ❖ I strongly agree 35 ❖ I agree 36 37 ❖ Neutral 38 39 ❖ I disagree 40 ❖ I strongly disagree on September 28, 2021 by guest. Protected copyright. 41 42 10. Antibiotics are overused in my hospital and in other hospitals ❖ I strongly agree 43 44 of my country Kenya ❖ I agree 45 46 ❖ Neutral 47 ❖ I disagree 48 49 ❖ I strongly disagree 50 51 11. Patients’ demands for antibiotics contribute to the overuse of ❖ I strongly agree 52 antibiotics in the hospital ❖ I agree 53 54 ❖ Neutral 55 56 57 58 Page 2 of 7 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 22 of 32 BMJ Open: first published as 10.1136/bmjopen-2019-030823 on 31 March 2020. Downloaded from

1 2 3 ❖ I disagree 4 5 ❖ I strongly disagree 6 7 12. I think over-the-counter (OTC) medicines contribute to ❖ I strongly agree 8 antibiotic misuse and subsequent antibiotic resistance ❖ I agree 9 10 ❖ Neutral 11 12 ❖ I disagree 13 ❖ I strongly disagree 14 15 13. My awareness on local antibiotic resistance pattern is? ❖ Excellent 16 For peer review only 17 ❖ Good 18 ❖ Average 19 20 ❖ Very little 21 22 ❖ None 23 24 PART 4: CHOICE OF ANTIBIOTIC 25 14. How confident are you about your knowledge of antibiotics? ❖ Very confident 26 27 ❖ Confident 28 29 ❖ A bit confident 30 ❖ Neutral/ I have no idea 31 32 ❖ Not confident at all http://bmjopen.bmj.com/ 33 34 15. What is your confidence level in prescribing antibiotics ❖ Very confident 35 ❖ Confident 36 37 ❖ A bit confident 38 39 ❖ Neutral/ I have no idea 40 ❖ Not confident at all on September 28, 2021 by guest. Protected copyright. 41 42 16. How often do you check your decisions on antibiotic ❖ Never 43 44 prescribing with a colleague? ❖ Sometimes 45 ❖ 46 Half of the times 47 ❖ Mostly 48 49 ❖ Always 50 51 17. If you do consult a senior colleague, how frequent does he/she ❖ Never 52 recommend prescription of a different antibiotic? ❖ Sometimes 53 54 ❖ Half of the times 55 56 57 58 Page 3 of 7 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 23 of 32 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2019-030823 on 31 March 2020. Downloaded from

1 2 3 ❖ Mostly 4 5 ❖ Always 6 7 18. How often do you depend on antibiotic sensitivity data from ❖ Never 8 the laboratory to vary your prescription ❖ Sometimes 9 10 ❖ Half of the times 11 12 ❖ Mostly 13 ❖ Always 14 15 PART 5: SOURCE OF INFORMATION ON ANTIBIOTICS PRESCRIBING AND RESISTANCE 16 For peer review only 17 19. During the past years, how many courses or trainings did you ❖ 0 18 ❖ 19 receive relating to antibiotics? 1-3 20 ❖ 4-6 21 22 ❖ 6-10 23 24 ❖ >10 25 20. Among the sources of information about antibiotics listed below, which one did you consult in the last month? 26 27 ▪ Information supplied by pharmaceutical companies ❖ Yes 28 29 ❖ No 30 31 ▪ Knowledge from training institutions ❖ Yes 32 http://bmjopen.bmj.com/ 33 34 ❖ No 35 ▪ Internet ❖ Yes 36 37 38 ❖ No 39 40 ▪ National guideline for empiric antimicrobial therapy ❖ Yes on September 28, 2021 by guest. Protected copyright. 41 42 ❖ No 43 44 ▪ The World Health Organization’s (WHO) guidelines for ❖ Yes 45 46 treatment of bacterial diseases ❖ No 47 48 21. How do you appreciate the sources of information about antibiotics listed below? 49 50 ▪ Information supplied by pharmaceutical companies ❖ Very useful 51 52 ❖ Useful 53 ❖ Not at all useful 54 55 ❖ I do not know 56 57 58 Page 4 of 7 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 24 of 32 BMJ Open: first published as 10.1136/bmjopen-2019-030823 on 31 March 2020. Downloaded from

1 2 3 ▪ Information from University courses ❖ Very useful 4 5 ❖ Useful 6 7 ❖ Not at all useful 8 ❖ I do not know 9 10 ▪ Internet ❖ Very useful 11 12 ❖ Useful 13 ❖ Not at all useful 14 15 ❖ I do not know 16 For peer review only 17 ▪ National guideline for empiric antimicrobial therapy ❖ Very useful 18 ❖ Useful 19 20 ❖ Not at all useful 21 22 ❖ I do not know 23 24 ▪ The World Health Organization’s (WHO) guidelines for ❖ Very useful 25 treatment of bacterial diseases ❖ Useful 26 27 ❖ Not at all useful 28 29 ❖ I do not know 30 ▪ Does your facility have a frequently released antibiogram? ❖ Yes 31 32 ❖ No http://bmjopen.bmj.com/ 33 34 ▪ If yes, how useful is the antibiogram to you ❖ Very useful 35 ❖ Useful 36 37 ❖ Not at all useful 38 39 ❖ I do not know 40 PART 6: DECISION ABOUT ANTIBIOTIC PRESCRIBING on September 28, 2021 by guest. Protected copyright. 41 42 22. When one prescribes an antibiotic, it is important to know the ❖ I strongly agree 43 44 resistance pattern of the bacteria in the local setting ❖ I agree 45 46 ❖ Neutral 47 ❖ I disagree 48 49 ❖ I strongly disagree 50 51 23. My choice of prescribing antibiotic is more influenced by the ❖ I strongly agree 52 availability of antibiotics than by the cause of the infection ❖ I agree 53 54 ❖ Neutral 55 56 57 58 Page 5 of 7 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 25 of 32 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2019-030823 on 31 March 2020. Downloaded from

1 2 3 ❖ I disagree 4 5 ❖ I strongly disagree 6 7 24. My choice of prescribing antibiotic is more influenced by the ❖ I strongly agree 8 cost of the drug to the patient ❖ I agree 9 10 ❖ Neutral 11 12 ❖ I disagree 13 ❖ I strongly disagree 14 15 25. I’m always concerned about effectiveness and quality of an ❖ I strongly agree 16 For peer review only 17 antibiotic when making my prescribing decisions ❖ I agree 18 ❖ Neutral 19 20 ❖ I disagree 21 22 ❖ I strongly disagree 23 24 26. In regard to antibiotic guidelines, local guidelines are more ❖ I strongly agree 25 useful than international guidelines ❖ I agree 26 27 ❖ Neutral 28 29 ❖ I disagree 30 ❖ I strongly disagree 31 32 27. Antibiotic guidelines and antibiotic committees are rather ❖ I strongly agree http://bmjopen.bmj.com/ 33 34 obstacles than a help ❖ I agree 35 ❖ Neutral 36 37 ❖ I disagree 38 39 ❖ I strongly disagree 40 28. I welcome the implementation of a training program about ❖ I strongly agree on September 28, 2021 by guest. Protected copyright. 41 42 antibiotics ❖ I agree 43 44 ❖ Neutral 45 ❖ I disagree 46 47 ❖ I strongly disagree 48 49 PART 7: KNOWLEDGE ON USE OF ANTIBIOTICS 50 51 29. A 4-year-old child had diarrhoea in the last 4 days (3 stools ❖ Amoxicillin orally 52 daily). She had no fever during the past days nor at ❖ Trimethoprim/sulphamethoxazole orally 53 54 consultation. What is your treatment choice? ❖ Amoxicillin/clavulanic acid orally 55 56 57 58 Page 6 of 7 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 26 of 32 BMJ Open: first published as 10.1136/bmjopen-2019-030823 on 31 March 2020. Downloaded from

Page 27 of 32 BMJ Open

BMJ Open Page 28 of 32

Page 29 of 32 BMJ Open

BMJ Open Page 30 of 32

Page 31 of 32 BMJ Open

1 2 3 STAFFING 4 Q39 How many laboratory technologists are in the station? 5 Q40 How many microbiologists are in the station? 6 Q41 Does your laboratory have staff with Bachelor’s degree qualification or higher? Yes 7 8 Partial 9 No 10 Q42 Do you engage a Consultant clinical microbiologist(s)? Yes 11 Partial 12 For peer reviewNo only 13 CONSUMABLES 14 Q43 How often do you experience unavailability of consumables in Microbiology Yes 15 section? Eg Lack of biochemical reagents and media Partial 16 No 17 Q44 Does your lab experience delays in Antimicrobial Susceptibility Testing Yes http://bmjopen.bmj.com/ 18 due to lack of reagents? Partial 19 No 20 Q45 Do frequent stock outs lead to low demand of cultures by clinicians? Yes 21 Partial 22 No 23 BIOSAFETY 24 25 Q46 Does your lab autoclave/incinerate cultures prior to discard? Yes on September 28, 2021 by guest. Protected copyright. 26 Partial 27 No 28 Q47 Do you have handwashing facility in the laboratory? Yes 29 Partial 30 No 31 Q48 Does your lab get continuous supply of running water? Yes 32 Partial 33 No 34 Q49 Does your lab have soap supply in the handwash facility? Yes 35 Partial 36 No 37 38 39 40 41 42 43 44 45 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 BMJ Open Page 32 of 32 BMJ Open: first published as 10.1136/bmjopen-2019-030823 on 31 March 2020. Downloaded from

1 2 3 Core element 1: Senior hospital management leadership towards Yes No 4 antimicrobial stewardship 5 6 1 Has your hospital management formally identified antimicrobial stewardship 7 as a priority objective for the institution and included it in its key performance 8 indicators? 9 2 Is there dedicated, sustainable and sufficient budgeted financial support for 10 antimicrobial stewardship activities (e.g., support for salary, training, or IT 11 (information technology) support)? 12 3 Does your hospital follow any (national or international) staffing standards for 13 antimicrobial stewardship activities (e.g. number of full-time equivalent (FTE) 14 15 per 100 beds for the different members of the antimicrobial stewardship 16 team)? For peer review only 17 Core element 2: Accountability and responsibilities 18 4 Does your hospital have a formal/written antimicrobial stewardship 19 programme/strategy accountable for ensuring appropriate antimicrobial use? 20 5 Does your hospital have a formal organizational multidisciplinary structure 21 responsible for antimicrobial stewardship (e.g., a committee focused on 22 23 appropriate antimicrobial use, pharmacy committee, patient safety 24 committee or other relevant structure)? 25 6 Is there a healthcare professional identified as a leader for antimicrobial 26 stewardship activities at your hospital and responsible for implementing the 27 programme? 28 7 Is there a document clearly defining roles, procedures of collaboration and 29 responsibilities of the antimicrobial stewardship team members? 30 31 8 Are clinicians, other than those part of the antimicrobial stewardship team 32 (e.g. from the ICU, Internal Medicine and Surgery) involved in the http://bmjopen.bmj.com/ 33 antimicrobial stewardship committee? 34 9 Does the antimicrobial stewardship committee produce regularly (indicate 35 minimum time) a dedicated report which includes e.g. antimicrobial use data 36 and/or prescription improvement initiatives, with time-committed short term 37 and long term measurable goals/ targets for optimizing antimicrobial use? 38 10 Is there a document clearly defining the procedures of collaboration of the 39 40 antimicrobial stewardship team/committee with the infection prevention and on September 28, 2021 by guest. Protected copyright. 41 control team/committee? 42 Core element 3: Available expertise on infection management 43 11 Do you have access to laboratory/imaging services and to timely results to be 44 able to support the diagnosis of the most common infections at your hospital? 45 12 In your hospital are there, or do you have access to, trained and experienced 46 healthcare professionals (medical doctor, pharmacist, nurse …) in infection 47 48 management (diagnosis, prevention and treatment) and stewardship willing 49 to constitute an antimicrobial stewardship team? 50 Core element 4: Education and practical training 51 13 Does your hospital offer a range of educational resources to support staff 52 training on how to optimize antimicrobial prescribing? 53 14 Do the antimicrobial stewardship team members receive regular training in 54 antimicrobial prescribing and stewardship? 55 56 Core element 5: Other actions aiming at responsible antimicrobial use 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 33 of 32 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2019-030823 on 31 March 2020. Downloaded from

Article Type: Protocol

Date Submitted by the 17-Dec-2019 Author:

Primary Subject Health economics Heading:

Secondary Subject Heading: Evidence based practice, Medical management

Antimicrobial stewardship, Antimicrobial resistance, Implementation Keywords:

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 1 of 32 BMJ Open

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 2 of 32

1 2 3

Combating Antibiotic Resistance using Guidelines and Enhanced Stewardship in BMJ Open: first published as 10.1136/bmjopen-2019-030823 on 31 March 2020. Downloaded from 4 Kenya: a protocol for an Implementation Science Approach 5 6 1 1 2 3 4 7 Jesse Gitaka , Moses Kamita , Dominic Mureithi , Davies Ndegwa , Moses Masika , 5 6 1 1 1 8 Geoffrey Omuse , Moses Ngari , Francis Makokha , Peter Mwaura , Ronald Mathai , Francis 9 Muregi1, Matilu Mwau7 10 11 Affiliations 12 1. Directorate of Research and Innovation, Mount Kenya University, P.O. Box 342-01000, 13 Thika, Kenya 14 2. Department of Animal Health and Production, Maasai Mara University, P.O. Box 861- 15 16 20500, Narok, Kenya. 17 3. Department of Medical Laboratory Sciences, Kenya Medical Training College P.O. Box 18 30195-00100, Nairobi,For Kenya peer review only 19 4. Department of Medical Microbiology, College of Health Sciences, University of Nairobi, 20 P.O. Box 30197-00100, Nairobi, Kenya. 21 5. Department of Pathology, Aga Khan University Hospital, P.O. Box 30270-00100, 22 23 Nairobi, Kenya 24 6. KEMRI/Wellcome Trust Research Programme, Centre for Geographic Medicine 25 Research - Coast, P.O. Box 230-80108, Kilifi, Kenya 26 7. Centre for Infectious and Parasitic Diseases Control Research, Kenya Medical Research 27 Institute, P.O. Box 3-50400, Busia, Kenya 28 29 Correspondence: Jesse Gitaka 30 31 [email protected] 32 Directorate of Research and Innovation, Mount Kenya University, P.O. Box 342-01000, 33 Thika, Kenya 34 35 36 37 http://bmjopen.bmj.com/ 38 39 40 41 42 43 44 45 on September 28, 2021 by guest. Protected copyright. 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 3 of 32 BMJ Open

1 2 3

Abstract BMJ Open: first published as 10.1136/bmjopen-2019-030823 on 31 March 2020. Downloaded from 4 5 Introduction 6 7 8 Antibiotic resistance (AMR) is a growing problem globally especially in Sub-Saharan Africa 9 including Kenya. Without any intervention, lower- and middle-income countries (LMIC) will 10 be most affected due to already higher AMR levels compared to higher income countries and 11 due to the far higher burden of diseases in the LMICs. Studies have consistently shown that 12 inappropriate use of antimicrobials is the major driver of AMR. To address this challenge, 13 hospital are now implementing antibiotic stewardship programs (ASPs), which have been 14 shown to achieve reduced antibiotic usage, to decrease the prevalence of resistance and lead 15 16 to significant economic benefits. However, the implementation of the guideline is highly 17 dependent on the settings in which they are rolled out. This study, employing an 18 implementation scienceFor approach, peer aims toreview address the knowledge only gap in this area and provide 19 critical data as well as practical experiences when using antibiotic guidelines and stewardship 20 programs in the public health sector. This will provide evidence of ASP performance and 21 potentially contribute to the county, national and regional policies on antibiotics use. 22 23 24 Methods and analysis 25 26 The study will be conducted in three geographically diverse regions, each represented by two 27 hospitals. A baseline study on antibiotic usage, resistance and de-escalation, duration of 28 hospital stay, rates of readmission and costs will be carried out in the pre-implementation 29 phase. The intervention, that is, the use of antibiotic guidelines and antibiotic stewardship 30 programs will be instituted for 18 months using a stepwise implementation strategy that will 31 facilitate learning and continuous improvement of stewardship activities and updating of 32 33 guidelines to reflect the evolving antibiotic needs. 34 35 Ethics and dissemination 36 37 Approvals to carry out the study have been obtained from the National Commission for http://bmjopen.bmj.com/ 38 Science, Technology and Innovation and the Mount Kenya University Ethics Review 39 Committee. The approvals from the two institutions were used to obtain permission to 40 conduct the study at each of the participating hospitals. Study findings will be presented to 41 policy stakeholders and published in peer-reviewed scientific journals. It is anticipated the 42 43 findings will inform local-based appropriate antibiotic use guidelines. 44 45 Key Words on September 28, 2021 by guest. Protected copyright. 46 47 Antimicrobial stewardship; Implementation science; Antimicrobial resistance 48 49 50 51 52 53 54 55 56 57 58 59 60

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 4 of 32

1 2 3

Strengths and limitations of this study BMJ Open: first published as 10.1136/bmjopen-2019-030823 on 31 March 2020. Downloaded from 4 5  First study aimed at concurrently rolling out antimicrobial stewardship committees in 6 multiple hospitals in Kenya. 7 8  Use of implementation approach to support implementation of suggested guidelines for 9 10 antimicrobial resistance surveillance. 11  Firsthand evidence on the antimicrobial resistance in three diverse counties in Kenya. 12 13  The study is limited to only three of the 47 counties in Kenya. 14 15 16 17 18 For peer review only 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 http://bmjopen.bmj.com/ 38 39 40 41 42 43 44 45 on September 28, 2021 by guest. Protected copyright. 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 5 of 32 BMJ Open

1 2 3

Introduction BMJ Open: first published as 10.1136/bmjopen-2019-030823 on 31 March 2020. Downloaded from 4 5 Antibiotic resistance is a serious public health concern globally1,2 and locally3,4 and fears of 6 7 running out of antibiotic options in the near future have been expressed. In 2016, the World 8 Health Organization (WHO) called for immediate and concerted efforts to mitigate this threat 9 to global health, which was estimated to contribute to 700,000 deaths in 2014 and projected to 10 cause 10 million deaths in 2050, if inadequately mitigated.5 The driving force escalating rates 11 of resistance is the injudicious use of antibiotics in patients, in livestock and in agriculture, and 12 its unchecked release into the environment.6 These forces raise and increase the spread of 13 resistant pathogens that may emerge by mutations or acquisition of plasmids carrying 14 7 15 resistance genes. In one study, prior antibiotic exposure was the key independent risk factor 16 for the acquisition of antibiotic multi-resistance.8 Broad-spectrum antibiotics have the 17 unintended consequence of selecting multidrug-resistant pathogens and increasing the 18 likelihood of infectionFor by fungi peer and Clostridium review difficile.9 only 19 20 The injudicious use of antibiotics is widespread. In Africa, many patients do not receive 21 treatment from the conventional health care system. Of those who receive antibiotics, 31.7 % 22 23 do not consult a doctor for a prescription and a further 26.4% obtain the antibiotics over the 10 24 counter. A study in South Africa found that 54.9% of antibiotics were inappropriately 25 prescribed in intensive care unit settings while in the US, 20-50% of prescribed antibiotics were 26 unnecessary or unwarranted.11–13 27 28 The rate of antimicrobial resistance in developing countries like Kenya is worrying and rising.14 29 For example, the prevalence of Salmonella typhi resistant to two or more antimicrobials was 30 found to have increased from 50% in 1998 to 78% in 2004 at Kenyatta National Hospital.15 31 32 The Global Antibiotic Resistance Partnership – Kenya Working Group Report of 2011 33 identified antibiotic resistance as a key issue in Kenya and made recommendations to curtail 34 the spread.16 These recommendations included the use of antibiotic guidelines that took into 35 consideration local resistance surveillance data and enhanced antibiotic stewardship programs 36 (ASPs).16 However, these ASPs have not been instituted at county hospitals, and key http://bmjopen.bmj.com/ 37 implementation data and experiences are lacking in their roll out. 38 39 Antibiotic stewardship is defined as ‘the optimum selection, dosage, and duration of 40 41 antimicrobial treatment that yields the best clinical outcomes for the treatment or prevention 42 of infection with the least toxicity to the patient and minimal impact on subsequent 43 resistance’.17 It has the potential to lower treatment cost and realize economic benefits to the 44 patient, health care system and the country at large.18,19 Moreover, optimizing antibiotic use by on September 28, 2021 by guest. Protected copyright. 45 minimizing exposure, fine-tuning dosage and reducing superfluous therapy and focusing 46 treatment to the likely culprit pathogens is a strategy that boosts patient safety20 and ultimately 47 21 48 safeguards against antibiotic resistance. 49 50 Justification 51 52 Antibiotic resistance is a major health challenge globally. Studies in Kenya have shown rising 53 antibiotic resistance over the last 3 decades. Whilst antibiotic guidelines and antibiotic 54 stewardship programs have been observed to lead to significant economic benefits, reduce 55 antibiotic usage and lower the prevalence of resistance especially in Europe, North America, 56 Japan, and South Africa, they have not been employed to tackle the challenge in the public 57 58 health sector in Kenya. The GARP report of 2011, recognized the use of guidelines and 59 stewardship programs as a potential strategy in ‘saving antibiotics’ but noted the need for local 60 studies to understand implementation challenges, derive lessons and embed the strategy within

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 6 of 32

1 2 3

the Kenyan health care system. The here proposed study, employing an implementation science BMJ Open: first published as 10.1136/bmjopen-2019-030823 on 31 March 2020. Downloaded from 4 approach, aims to address the knowledge and research gap in this area and provide critical data 5 6 and experiences in using antibiotic guidelines and stewardship programs in the public health 7 sector. 8 9 Objectives of the Study 10 11 General Objectives 12 13 The overall aim is to evaluate the impact of antibiotic guidelines (antibiotic stewardship 14 program) for empirical treatment of (i) urinary tract infections, (ii) community-acquired 15 pneumonia, (iii) bacteremia and (iv) meningitis in terms of reducing usage of broad-spectrum 16 17 antibiotics, antibiotic de-escalation, duration of hospital stays, rates of readmission and 18 prevalence of antibioticFor resistance peer in six countyreview hospitals inonly Kenya. 19 20 Specific Objectives 21 22 1. To develop guidelines for antibiotic use for common infections. 23 2. To set up antibiotic steward committees (ASCs) in six county hospitals. 24 3. To train the ASC members such that they can fully perform their mandate. 25 26 4. To measure the usage of broad-spectrum antibiotics, antibiotic de-escalation, duration 27 of hospital stays and rates of readmission in hospitals using the antibiotic stewardship 28 strategy. 29 5. To ascertain antibiotic resistance patterns using culture, sensitivity, and genetic 30 markers. 31 6. To establish the knowledge of health care workers, their attitudes and prescription 32 practices regarding antibiotics resistance, use of guidelines and ASPs. 33 34 7. To evaluate the economic benefits of using guidelines for antibiotics use and ASPs. 35 36 Expected Outputs of the research 37 http://bmjopen.bmj.com/ 38 1. Antibiotic guidelines informed by local bacteria susceptibility patterns. 39 2. Antibiotic Stewardship Committees (ASCs) in six county hospitals in Kenya. 40 3. Trained ASC members. 41 4. Data on antibiotics usage, antibiotic de-escalation, duration of hospital stay and rates 42 of readmission in hospitals using the antibiotic stewardship strategy. 43 44 5. Map and data of antibiotic resistance patterns before and after implementation of the 45 strategy. on September 28, 2021 by guest. Protected copyright. 46 6. Qualitative data on the knowledge of health care workers, their attitudes, and practices 47 regarding antibiotic usage and stewardship programs. 48 7. Publication on cost-benefit analysis for using antibiotic guidelines and ASPs. 49 50 Methods and analysis 51 52 53 Setting 54 55 To be able to evaluate the antibiotic stewardship strategy across Kenya using an 56 implementation science approach, the study will be conducted in six county hospitals, namely 57 (1) Kiambu County Referral Hospital, (2) Bungoma County Hospital, (3) Webuye Sub-County 58 Hospital, (4) Nakuru County Teaching and Referral Hospital, (5) Thika Level 5 Hospital and 59 (6) Naivasha Sub-County Hospital). 60

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 7 of 32 BMJ Open

1 2 3

Design BMJ Open: first published as 10.1136/bmjopen-2019-030823 on 31 March 2020. Downloaded from 4 5 The study will utilize the Reach, Effectiveness, Adoption, Implementation and Maintenance 6 7 (RE-AIM) conceptual framework. RE-AIM involves the interaction of all the five factors and 22 8 fits well with approaches that are system based . The study design will encompass a pre- 9 implementation phase, a stepwise implementation phase, and an endline study to measure the 10 changes in outcomes between the phases. A baseline study on antibiotic usage (Defined Daily 11 Doses per 1,000), antibiotic resistance (culture and genetic markers), antibiotic de-escalation, 12 duration of hospital stays, rates of readmission, prescription patterns, and cost analysis will be 13 carried out in the pre-implementation phase. The intervention, which is the use of antibiotic 14 15 guidelines and antibiotic stewardship programs, will be instituted in the six county hospitals 16 for 18 months. The stepwise implementation phase will involve introduction of interventions 17 as well as monitoring and evaluation of their effectiveness, and thereafter improving the 18 interventions whereFor necessary. peer The strategy review will facilitate learningonly cycles every 4-6 weeks and 19 continuous improvement of stewardship activities and updating of guidelines to reflect 20 evolving antibiotic needs in diverse settings. During the intervention phase, challenges faced 21 22 during the implementation of the guidelines will be documented and used to advise on the 23 changes required during the review of the guidelines. The end line study will be conducted and 24 differences between the two phases evaluated as per statistical analysis laid out below. 25 26 In the baseline and endline studies, multidisciplinary strategies will be employed as 27 follows 28 i. Knowledge of health care workers, attitudes, and practices on antibiotic resistance, 29 guidelines, and ASP will be studied qualitatively and quantitatively. 30 31 ii. Basic science approaches, encompassing antibiotics culture sensitivity and molecular 32 biology-genetic markers of resistance, will be analyzed as detailed below. 33 iii. Clinical patient outcomes will be studied to evaluate the guidelines. 34 iv. Health economics cost savings on using guidelines and ASP will be evaluated as below. 35 36 Antibiotic guidelines and ASPs: Development 37 http://bmjopen.bmj.com/ 38 Antibiotic guidelines will be formulated in consultation with senior clinicians in the study 39 hospitals taking into account each hospital’s antimicrobial resistance patterns. The ASP 40 41 committee will comprise of a hospital physician, a microbiologist, and a pharmacist. Broad 42 spectrum antibiotic prescriptions will be brought to the attention of the ASP committee 43 members. The committee members will perform regular ward rounds, three times a week, in 44 the initial stages. This will be reduced to once a week later in the project, to optimize adherence on September 28, 2021 by guest. Protected copyright. 45 to antibiotic guidelines. Furthermore, the guidelines will be promoted through training 46 sessions, provision of pocket-size guideline cards to clinicians and pharmacists, large poster 47 48 displays in the wards and through hospital and project websites and social media (?). 49 50 Data collection 51 52 Laboratory assessment tools will be used to determine the preparedness of the laboratories to 53 perform antibiotic sensitivity tests (see Additional file 1). The tool will assess whether the 54 laboratories have the equipment necessary to perform bacterial culture, identification, and 55 susceptibility testing, such as carbon dioxide incubators, safety cabinets, and refrigerators. The 56 tool will also determine whether the laboratories have the internationally recommended 57 58 guidelines to perform susceptibility and quality assurance tests. In addition, knowledge, 59 attitude and practices (KAP) about antibiotic prescription and resistance among medical 60 practitioners will be assessed using a KAP tool (See Additional file 2). The tool will target

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 8 of 32

1 2 3

those medical practitioners, who usually prescribe antimicrobial drugs in the hospitals and will BMJ Open: first published as 10.1136/bmjopen-2019-030823 on 31 March 2020. Downloaded from 4 include the consultants, medical officers as well as interns, clinical officers and pharmacists. 5 6 The KAP tool will also determine the guidelines that the prescribers use to decide on the 7 appropriate antimicrobial drug. A health system assessment tool will be employed to gather 8 baseline information on the antimicrobial stewardship activities in each hospital (see 9 Additional file 3). Information on bed occupancy, antibiotic usage, and antibiotic resistance 10 data will be collected from each of the hospital information systems, pharmacy management 11 systems and laboratory reporting systems, respectively. Antibiotic usage will be analyzed in 12 order to determine the impact of the intervention on antibiotic usage comparing the intervention 13 14 and the control arms. Data on antibiotics issued to adult and children inpatients only will be 15 factored, excluding discharge and outpatient supplies. 16 17 To consistently compare antibiotic usage, the defined daily doses (DDD) will be used and 18 expressed per 1,000For occupied peer bed days (OBDs) review to account foronly fluctuations in activity following 19 WHO guidelines.23 The OBDs will be obtained from the hospital information systems. 20 21 Culture sensitivity and genetic analyses 22 23 24 Nasal swabs or tracheal aspirate, urine, wound swabs and blood samples will be taken from 25 patients who have consented to the study for bacteriological analysis. Samples will be subjected 26 to standard bacteriological analysis to isolate the culprit bacteria. Bacterial species will be 27 confirmed by use of biochemical test and analytical profiles index API strips (bioMérieux 28 France). Antimicrobial susceptibility tests will be performed on isolated bacteria as per the 29 Kirby-Bauer Method following manufacturer’s instruction. Results will be interpreted using 30 the Clinical and Laboratory Standards Institute (CLSI) tables.24 31 32 33 Any bacteria isolate found to be resistant to third generation cephalosporins will be tested for 34 production of extended spectrum beta-lactamase (ESBLs) using the synergy disk diffusion 35 test. Vancomycin Resistant Enterococci (VRE) will be identified using disc diffusion tests. 36 Methicillin resistant S. aureus (MRSA) will be detected by testing isolates resistant to cefoxitin 37 using the E test (AB Biodisk, Solna, Sweden) on Mueller–Hinton agar supplemented with 2% http://bmjopen.bmj.com/ 38 NaCl and incubated at 37 °C for 24 h. The identified ESBLs, VRE and MRSA samples will be 39 analyzed by PCR and sequencing to identify the resistance genotype. In vitro conjugation tests 40 41 will be performed to determine if resistance in bacteria is transferable. 42 43 Cost-benefit analysis for the use of antibiotic guidelines and Antibiotic Stewardship 44 Program 45 on September 28, 2021 by guest. Protected copyright. 46 A cost-benefit analysis (CBA) from a health facility and a national perspective will be 47 performed. For health facilities, three cost drivers will be considered: (a) pharmacy spending, 48 (b) length of stay, and (c) antimicrobial stewardship interventions (training, infection control 49 50 measures, etc.). At country level, we will consider Disability-Adjusted Life years (DALYS), 25–27 51 work-days lost, and cost of treatment. This will be done by collecting and analyzing data 52 on patient income, length of hospital stay, death or disability occasioned by drug-resistant 53 pathogens, hospital pharmacy expenditure and cost of training/rolling out antimicrobial 54 stewardship guidelines. This data will be collected before and after antimicrobial stewardship 55 interventions. 56 57 Study size 58 59 60 Prevalence of inappropriate use of broad-spectrum antibiotics

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 9 of 32 BMJ Open

1 2 3

The sample size is based on the prevalence of inappropriate use of broad-spectrum antibiotics, BMJ Open: first published as 10.1136/bmjopen-2019-030823 on 31 March 2020. Downloaded from 4 documented as being about 50% in South Africa.11 To detect a reduction of the 5 6 50% inappropriate use of broad-spectrum antibiotics by 20% to 30% with a power of 90% and 7 α of 0.05 in a two-sided test, a sample size of 410 in each county hospital would be adequate. 8 In order to detect a reduction (20%) but with a power 80%, the sample size would be 320 as 9 shown in Figure 1. With a 15% allowance for loss to follow up, a total of about 500 participants 10 would be sufficient in each of the selected county hospitals. Therefore, we shall evaluate the 11 prevalence of inappropriate use of broad-spectrum antibiotics in 500 x 6= 3,000 patients from 12 the six facilities. 13 14 15 Prevalence of antibiotic resistance 16 17 The sample size is based on the prevalence of antibiotic resistance of 78%15 in 2004 at Kenyatta 18 National Hospital, ForKenya. Topeer detect a reductionreview of the 78% only antibiotic resistance by 10% to 19 68% with a power of 90% and α of 0.05 two-sided test, a sample size of 500 in each county 20 hospital would be adequate; while to achieve a reduction of 10% with power 80%, the sample 21 size would need to be 220 (Figure 2). With 15% allowance for loss to follow up, a total of 600 22 23 participants would be sufficient in each of the selected county hospitals. Therefore, we shall 24 evaluate the prevalence of resistance in 600 x 6 = 3,600 patients for the six facilities. 25 26 Data Management 27 28 Data will be recorded on a standardized Case Report Form (CRF) at every hospital by the 29 trained study staff. The data will be held locally and uploaded to the secure central study server 30 hosted at the Mount Kenya University main campus in Thika. This will be overseen by the 31 Data Manager/statistician, who will run regular reconciliation to derive the final study 32 33 database. Access to the study database will be restricted and password protected. The data will 34 be stored in excel, jpeg and word files for the electronic cases. For archiving purposes, the 35 Mount Kenya University electronic repository will be used to ensure the data is publicly 36 available. Collected data will be retained accord to the policy governing the Mount Kenya 37 University repository. The principal investigators involved in this study do not have any http://bmjopen.bmj.com/ 38 reasons that might prohibit the sharing of the data emanating from the study. 39 40 Statistical analysis 41 42 43 The data will be analyzed using qualitative and quantitative methods. Qualitative data will be 44 analyzed by subjecting the information to content analysis and presenting it in different 45 emerging themes. The summaries of the data emanating from these themes will then be on September 28, 2021 by guest. Protected copyright. 46 arranged on a case by case basis through the use of an Excel spreadsheet28 and the analyses 47 performed using NVIVO software. Quantitative data analyses will be carried out using Stata 48 version 14 (Stata, Inc.). The differences between baseline and endline on all the study outcomes 49 50 will be compared using appropriate statistical methods, such as the McNemar’s test, paired t- 51 test and the Wilcoxon signed-rank test non-parametric test, accounting for pairing and clusters 52 (hospitals). Multivariable log-binomial regression analysis will be used to elucidate risk factors 53 for antibiotic resistance. 54 55 Ethics and Dissemination 56 57 Approvals to carry out the study have been obtained from the National Commission for 58 Science, Technology and Innovation (NACOSTI) (NACOSTI/P/18/33304/25986) and the 59 60 Mount Kenya University Ethics Review Committee (MKU/ERC/0764). In addition,

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 10 of 32

1 2 3

permission was obtained from each of the participating hospitals. This entailed submission of BMJ Open: first published as 10.1136/bmjopen-2019-030823 on 31 March 2020. Downloaded from 4 an introduction letter attaching the certificate of clearance from NACOSTI and the ethical 5 6 review certificate from MKU. In Thika, Nakuru and Naivasha a research fee was paid as 7 required by the hospital board. The study will follow all provisions of the Declaration of 8 Helsinki. Participants in the study will not incur any cost associated with the transport or 9 processing of the samples; neither will they receive any monetary inducements to participate 10 in the study. Material transfer to laboratories outside of Kenya shall not be undertaken in this 11 study. Informed written consent will be sought from the participants enrolled in the study. 12 13 Obtained results will be disseminated via different streams, including research conferences and 14 15 in peer-reviewed journals. We are aiming for five publications to be generated by the end of 16 the study. These will include the study protocol itself and the different aspects emerging during 17 in the course of the project. Contextualized guidelines on judicious use of antibiotics in the six 18 hospitals in Kenya Forwill also peerbe published review in the Ministry onlyof Health’s website and that of the 19 collaborating hospitals for easy access by other health facility in and outside Kenya. In 20 addition, the findings will be shared in a dissemination forum bring together members of the 21 22 health management teams at both the country and county levels, clinicians who do prescription 23 of antimicrobial drugs, researchers, members of the public, and other key stakeholders. The 24 dissemination forum will be held once at the end of the study. 25 26 Patient and Public Involvement 27 28 The study is mainly targeting clinicians who do antimicrobial proscription and, therefore, 29 patients will not be directly involved in the study. Reports obtained from this study will, 30 however, be made available to the patients in the participating hospitals through a 31 32 dissemination forum and to the general public through publications. 33 34 Exit strategy and stakeholder involvement 35 36 In the process of developing this protocol, we have engaged physicians working in the proposed 37 county hospitals, and they have identified the challenge of antibiotic resistance as real and http://bmjopen.bmj.com/ 38 appreciate the opportunities that use of antibiotic guidelines and ASPs may provide in 39 combating resistance, improving clinical care and saving costs. We plan to continue involving 40 all the stakeholders in the process, including clinicians, laboratory personnel, pharmacists, 41 42 public health officials, patients and scientists. We anticipate that the study findings will inform 43 county and national policy on mitigating antibiotic resistance and raise public awareness on 44 the need for judicious use of antibiotics. The project will use social media platforms, websites 45 of the collaborating institutions, and publications in peer-reviewed journals, local dailies and on September 28, 2021 by guest. Protected copyright. 46 presentations in scientific meetings to further engage stakeholders and the public on this 47 important issue and enhance the learning approach inherent in the strategy of implementation 48 49 science for improved performance of the ASPs and the study in general. In case any of the 50 hospitals drop out either during the baseline phase or early stages of the implementation phase, 51 another hospital will be brought onboard. Replacements will not be considered during the late 52 stages of the project. 53 54 Discussion 55 56 The emergence of antimicrobial-resistant pathogens and a lack of new drugs to effectively treat 57 these pathogens are the two main challenges in human health. There is thus the need to advocate 58 59 the proper use of the currently available antimicrobial agents by safeguarding their 60 effectiveness. Antibiotic stewardship has been shown to contribute to reducing antibiotic

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 11 of 32 BMJ Open

1 2 3 29,30

resistance, but this strategy has not been rolled out in most sub-Saharan countries in level BMJ Open: first published as 10.1136/bmjopen-2019-030823 on 31 March 2020. Downloaded from 4 4 and 5 hospitals. The proposed work will employ an implementation research approach to 5 6 evaluate the best strategies and derive lessons on mainstreaming antibiotic stewardship in these 7 facilities. By leveraging on a health system approach, the implementation research will unravel 8 real life impediments and opportunities by working with hospital teams to co-design execution 9 plans, monitor and evaluate and the sustainability of the stewardship programs. 10 11 Most hospitals in Kenya lack the capacity to carry out antimicrobial sensitivity tests, due to the 12 lack of resources and technical knowhow, among other challenges.31 This study will first 13 provide an assessment to determine the challenges hospitals are facing; this will be done 14 15 through interviews with clinicians and assessment of the capacity of the laboratories to perform 16 the sensitivity tests. These steps will make up the initial phase and will guide the nature of 17 implementations to be used during the implementation phase. It is hoped that the project will 18 capacitate the laboratoriesFor in peer the six hospitals review to do the tests only and sensitize the clinicians on the 19 need to prescribe antimicrobial drugs based on results obtained from the laboratory tests. After 20 the implementation phase, the endline phase will involve surveys similar to those conducted 21 22 during the baseline phase. This will allow determination of the impact of the implementation 23 strategies taken in reducing the cost of treatment, length of hospital stay and burden on the 24 health system, among others. 25 26 The hospital management teams play a key role in determining the allocation of resources to 27 and within the hospitals.32 In this project, we have proposed to involve hospital management 28 by including its members in the stewardship committees that will be established. This will 29 ensure that the management is well informed about the challenges in the sections involved in 30 31 surveillance and the progress that is being made. Involvement of the hospital management will 32 also ensure that there is a buy-in of the recommendations made by the stewardship committee. 33 34 Although there is the need to establish antimicrobial stewardship committees in all the 35 hospitals, the project proposes to start with the selected six hospitals with the hope that the 36 same can be reproduced by other hospitals to establish similar committees in their hospitals. 37 http://bmjopen.bmj.com/ 38 Study timeline 39 40 The study is designed for a three-year period. Enrolment already started in 2018 and 41 42 participant recruitment will continue up to 2020 (Table 1). So far, stewardship committees 43 have been established and workshops to sensitize the members conducted. Data collection on 44 knowledge, attitude and practice of antibiotic prescribers in the hospitals is already in on September 28, 2021 by guest. Protected copyright. 45 progress. 46 47 48 Table 1: Proposed study timelines. 49 2019 2020 2021 50 51 52 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 53 Activity 54 55 Ethics clearance and 56 authorization from 57 58 study sites 59 60

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 12 of 32

1 2 3

Baseline study on BMJ Open: first published as 10.1136/bmjopen-2019-030823 on 31 March 2020. Downloaded from 4 5 prevalence of antibiotic 6 7 resistance and 8 9 stewardship activities 10 11 Identification of 12 members for the 13 antibiotic stewardship 14 15 programs 16 17 Development of 18 antibiotic guidelinesFor peer review only 19 20 Adopting and preparing 21 materials for ASP 22 training 23 24 Development of 25 antibiotic guidelines and 26 27 stewardship mobile 28 application 29 30 Training the ASP 31 members 32 33 Collecting data on 34 antibiotic usage in the 35 36 hospitals 37 http://bmjopen.bmj.com/ 38 Conducting KAPs study 39 on use of guidelines and 40 ASPs 41 Analyze and present 42 antibiotic resistance 43 data using maps and 44

charts on September 28, 2021 by guest. Protected copyright. 45 46 Presenting findings to 47 stakeholders and 48 preparing the final 49 report 50 51 52 Patient and Public Involvement 53 54 No patient involved. 55 56 Availability of data and material 57 58 All data sets will be available to the public upon request. 59 60 Acknowledgment

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 13 of 32 BMJ Open

1 2 3

Authors acknowledge Professor Nicole Pamme, University of Hull, for proofreading the BMJ Open: first published as 10.1136/bmjopen-2019-030823 on 31 March 2020. Downloaded from 4 5 manuscript. 6 Additional files 7 8 Additional file 1: Laboratory preparedness assessment tool 9 10 Additional file 2: Laboratory knowledge, attitude and practice (KAP) tool 11 12 Additional file 3: Health system tool 13 14 Figure Legends 15 16 Figure 1: Statistical power analysis for sample size determination for the prevalence of 17 inappropriate use of broad-spectrum antibiotics 18 For peer review only 19 Figure 2: Statistical power analysis for sample size determination for the prevalence of 20 antibiotics resistance 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 http://bmjopen.bmj.com/ 38 39 40 41 42 43 44 45 on September 28, 2021 by guest. Protected copyright. 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 14 of 32

1 2 3

References BMJ Open: first published as 10.1136/bmjopen-2019-030823 on 31 March 2020. Downloaded from 4 5 1. Aslam, B. et al. Antibiotic resistance: a rundown of a global crisis. Infect. Drug Resist. 6 7 11, 1645–1658 (2018). 8 9 2. Livermore, D. M. Has the era of untreatable infections arrived? J. Antimicrob. 10 Chemother. 64, 29–36 (2009). 11 12 3. Bejon, P. et al. Invasive Gram-negative bacilli are frequently resistant to standard 13 antibiotics for children admitted to hospital in Kilifi, Kenya. J. Antimicrob. 14 Chemother. 56, 232–235 (2005). 15 16 4. Dougan, G. et al. Typhoid in Kenya Is Associated with a Dominant Multidrug- 17 18 Resistant SalmonellaFor entericapeer Serovar review Typhi Haplotype only That Is Also Widespread in 19 Southeast Asia. J. Clin. Microbiol. 48, 2171–2176 (2010). 20 21 5. WHO. United Nations meeting on antimicrobial resistance. (2016). 22 23 6. Boxall, A. et al. The role of the natural environment in the emergence of antibiotic 24 resistance in Gram- negative bacteria The role of the natural environment in the 25 emergence of antibiotic resistance in Gram-negative bacteria. Lancet Infect. Dis. 13, 26 27 155–165 (2017). 28 29 7. Baquero, F. et al. The global threat of antimicrobial resistance: science for 30 intervention. New Microbes New Infect. 6, 22–29 (2015). 31 32 8. Sanchez-Ortega, I. et al. Bacteraemia due to multidrug-resistant Gram-negative bacilli 33 in cancer patients: risk factors, antibiotic therapy and outcomes. J. Antimicrob. 34 Chemother. 66, 657–663 (2010). 35 36

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 15 of 32 BMJ Open

1 2 3

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 16 of 32

1 2 3

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 17 of 32 BMJ Open

1 2 3

Authors’ Contributions BMJ Open: first published as 10.1136/bmjopen-2019-030823 on 31 March 2020. Downloaded from 4 5 JG initiated the concepts of the study. JG, DM, DN, MMasika, GO, MN, FMakokha, PM, 6 RM, FMuregi, and MMwau developed the protocol. JG, DM, DN, and MMasika will be 7 8 involved in training the ASCs. DM, GO, MN, FMakokha, PM, RM, DN, and MK will help in 9 data collection. MN and MK will be involved in data analysis. MK wrote the first draft of the 10 manuscript. All authors reviewed and approved the protocol. 11 12 Funding Statement 13 14 This work was supported by The Kenya National Research Fund (grant number 15 16 NRF/MKU/2017/007 to JG). The funders had no role in study design, decision to publish, or 17 preparation of the manuscript. 18 For peer review only 19 Competing Interests 20 21 The authors declare no conflict of interest. 22 23 24 Word count: 3526 words 25 26 27 28 29 30 31 32 33 34 35 36 37 http://bmjopen.bmj.com/ 38 39 40 41 42 43 44 45 on September 28, 2021 by guest. Protected copyright. 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 18 of 32 BMJ Open: first published as 10.1136/bmjopen-2019-030823 on 31 March 2020. Downloaded from

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 For peer review only 17 18 19 20 21 22 23 24 25 26 27 28 29 30 Figure 1: Statistical power analysis for sample size determination for the prevalence of inappropriate use of 31 broad-spectrum antibiotics 32 http://bmjopen.bmj.com/ 33 557x405mm (72 x 72 DPI) 34 35 36 37 38 39 40 on September 28, 2021 by guest. Protected copyright. 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 19 of 32 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2019-030823 on 31 March 2020. Downloaded from

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 For peer review only 17 18 19 20 21 22 23 24 25 26 27 28 29 30 Figure 2: Statistical power analysis for sample size determination for the prevalence of antibiotics resistance 31 32 557x405mm (72 x 72 DPI) http://bmjopen.bmj.com/ 33 34 35 36 37 38 39 40 on September 28, 2021 by guest. Protected copyright. 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 20 of 32 BMJ Open: first published as 10.1136/bmjopen-2019-030823 on 31 March 2020. Downloaded from

1 2 3 4 A survey on Knowledge, Attitudes and Practice about antibiotic prescribing and resistance among 5 6 medical practitioners in Kenyan local hospitals. 7 8 Thank you very much for accepting to participate in this study. 9 10 *You are kindly requested to answer the questionnaire honestly and completely independent of 11 cross-consultations and/or verifications. 12 13 14 Survey quality control 15 16 Date of interview: ……………………………………………….For peer review Start time...... only End time...... 17 18 19 Interviewed by: ...... Approved...... 20 21 Name of the Hospital…………………………………………… Respondent’s code……………………………………………………. 22 23 24 QUESTIONS ANSWERS 25 PART 1: GENERAL QUESTIONS 26 27 1. For how many years, since you graduated from medical school ❖ I am on attachment 28 ❖ I am a trainee in medicine (internship) /medical training College, have you been working in a hospital 29 ❖ Less than one year 30 (indicate cumulative years if worked in different hospitals) ❖ 1-3 years 31 ❖ 32 4 – 6 years http://bmjopen.bmj.com/ 33 ❖ 7 years and more 34 2. In which department do you work? o Medicine /Emergency 35 o Surgery 36 37 o Paediatrics 38 39 o Obstetrics and Gynaecology 40 o on September 28, 2021 by guest. Protected copyright. 41 Outpatient/A/E 42 o Pharmacy 43 44 o Other: ………………………………………………… 45 46 3. Designation (e.g. Consultant, Pharmacist, Nurse, etc.) 47 …………………………………………………………………… 48 PART 2: PRESCRIPTION PATTERN (PRACTICE) 49 50 4. How frequently do you prescribe antibiotics? ❖ More than once daily 51 ❖ Once daily 52 53 ❖ 3 – 5 times a week 54 55 ❖ 1 – 2 times a week 56 57 58 Page 1 of 7 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 21 of 32 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2019-030823 on 31 March 2020. Downloaded from

1 2 3 ❖ less than once a week) 4 5 5. To whom do you prescribe? ❖ Patients at outpatient department 6 7 ❖ Hospitalized patients 8 ❖ Patients in out-patient department and 9 10 hospitalised patients 11 12 6. Do you follow any antibiotic prescription guidelines? ❖ Yes 13 ❖ No 14 15 PART 3: AWARENESS AND ATTITUDE ON THE CURRENT SCOPE OF ANTIBIOTIC RESISTANCE 16 For peer review only 17 7. Antibiotic resistance is a world-wide problem ❖ I strongly agree 18 19 ❖ I agree 20 ❖ Neutral 21 22 ❖ I disagree 23 24 ❖ I strongly disagree 25 8. Antibiotic resistance is a problem in my country ❖ I strongly agree 26 27 ❖ I agree 28 29 ❖ Neutral 30 ❖ I disagree 31 32 ❖ I strongly disagree http://bmjopen.bmj.com/ 33 34 9. Antibiotic resistance is a problem in my hospital ❖ I strongly agree 35 ❖ I agree 36 37 ❖ Neutral 38 39 ❖ I disagree 40 ❖ I strongly disagree on September 28, 2021 by guest. Protected copyright. 41 42 10. Antibiotics are overused in my hospital and in other hospitals ❖ I strongly agree 43 44 of my country Kenya ❖ I agree 45 46 ❖ Neutral 47 ❖ I disagree 48 49 ❖ I strongly disagree 50 51 11. Patients’ demands for antibiotics contribute to the overuse of ❖ I strongly agree 52 antibiotics in the hospital ❖ I agree 53 54 ❖ Neutral 55 56 57 58 Page 2 of 7 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 22 of 32 BMJ Open: first published as 10.1136/bmjopen-2019-030823 on 31 March 2020. Downloaded from

1 2 3 ❖ I disagree 4 5 ❖ I strongly disagree 6 7 12. I think over-the-counter (OTC) medicines contribute to ❖ I strongly agree 8 antibiotic misuse and subsequent antibiotic resistance ❖ I agree 9 10 ❖ Neutral 11 12 ❖ I disagree 13 ❖ I strongly disagree 14 15 13. My awareness on local antibiotic resistance pattern is? ❖ Excellent 16 For peer review only 17 ❖ Good 18 ❖ Average 19 20 ❖ Very little 21 22 ❖ None 23 24 PART 4: CHOICE OF ANTIBIOTIC 25 14. How confident are you about your knowledge of antibiotics? ❖ Very confident 26 27 ❖ Confident 28 29 ❖ A bit confident 30 ❖ Neutral/ I have no idea 31 32 ❖ Not confident at all http://bmjopen.bmj.com/ 33 34 15. What is your confidence level in prescribing antibiotics ❖ Very confident 35 ❖ Confident 36 37 ❖ A bit confident 38 39 ❖ Neutral/ I have no idea 40 ❖ Not confident at all on September 28, 2021 by guest. Protected copyright. 41 42 16. How often do you check your decisions on antibiotic ❖ Never 43 44 prescribing with a colleague? ❖ Sometimes 45 ❖ 46 Half of the times 47 ❖ Mostly 48 49 ❖ Always 50 51 17. If you do consult a senior colleague, how frequent does he/she ❖ Never 52 recommend prescription of a different antibiotic? ❖ Sometimes 53 54 ❖ Half of the times 55 56 57 58 Page 3 of 7 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 23 of 32 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2019-030823 on 31 March 2020. Downloaded from

1 2 3 ❖ Mostly 4 5 ❖ Always 6 7 18. How often do you depend on antibiotic sensitivity data from ❖ Never 8 the laboratory to vary your prescription ❖ Sometimes 9 10 ❖ Half of the times 11 12 ❖ Mostly 13 ❖ Always 14 15 PART 5: SOURCE OF INFORMATION ON ANTIBIOTICS PRESCRIBING AND RESISTANCE 16 For peer review only 17 19. During the past years, how many courses or trainings did you ❖ 0 18 ❖ 19 receive relating to antibiotics? 1-3 20 ❖ 4-6 21 22 ❖ 6-10 23 24 ❖ >10 25 20. Among the sources of information about antibiotics listed below, which one did you consult in the last month? 26 27 ▪ Information supplied by pharmaceutical companies ❖ Yes 28 29 ❖ No 30 31 ▪ Knowledge from training institutions ❖ Yes 32 http://bmjopen.bmj.com/ 33 34 ❖ No 35 ▪ Internet ❖ Yes 36 37 38 ❖ No 39 40 ▪ National guideline for empiric antimicrobial therapy ❖ Yes on September 28, 2021 by guest. Protected copyright. 41 42 ❖ No 43 44 ▪ The World Health Organization’s (WHO) guidelines for ❖ Yes 45 46 treatment of bacterial diseases ❖ No 47 48 21. How do you appreciate the sources of information about antibiotics listed below? 49 50 ▪ Information supplied by pharmaceutical companies ❖ Very useful 51 52 ❖ Useful 53 ❖ Not at all useful 54 55 ❖ I do not know 56 57 58 Page 4 of 7 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 24 of 32 BMJ Open: first published as 10.1136/bmjopen-2019-030823 on 31 March 2020. Downloaded from

1 2 3 ▪ Information from University courses ❖ Very useful 4 5 ❖ Useful 6 7 ❖ Not at all useful 8 ❖ I do not know 9 10 ▪ Internet ❖ Very useful 11 12 ❖ Useful 13 ❖ Not at all useful 14 15 ❖ I do not know 16 For peer review only 17 ▪ National guideline for empiric antimicrobial therapy ❖ Very useful 18 ❖ Useful 19 20 ❖ Not at all useful 21 22 ❖ I do not know 23 24 ▪ The World Health Organization’s (WHO) guidelines for ❖ Very useful 25 treatment of bacterial diseases ❖ Useful 26 27 ❖ Not at all useful 28 29 ❖ I do not know 30 ▪ Does your facility have a frequently released antibiogram? ❖ Yes 31 32 ❖ No http://bmjopen.bmj.com/ 33 34 ▪ If yes, how useful is the antibiogram to you ❖ Very useful 35 ❖ Useful 36 37 ❖ Not at all useful 38 39 ❖ I do not know 40 PART 6: DECISION ABOUT ANTIBIOTIC PRESCRIBING on September 28, 2021 by guest. Protected copyright. 41 42 22. When one prescribes an antibiotic, it is important to know the ❖ I strongly agree 43 44 resistance pattern of the bacteria in the local setting ❖ I agree 45 46 ❖ Neutral 47 ❖ I disagree 48 49 ❖ I strongly disagree 50 51 23. My choice of prescribing antibiotic is more influenced by the ❖ I strongly agree 52 availability of antibiotics than by the cause of the infection ❖ I agree 53 54 ❖ Neutral 55 56 57 58 Page 5 of 7 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 25 of 32 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2019-030823 on 31 March 2020. Downloaded from

1 2 3 ❖ I disagree 4 5 ❖ I strongly disagree 6 7 24. My choice of prescribing antibiotic is more influenced by the ❖ I strongly agree 8 cost of the drug to the patient ❖ I agree 9 10 ❖ Neutral 11 12 ❖ I disagree 13 ❖ I strongly disagree 14 15 25. I’m always concerned about effectiveness and quality of an ❖ I strongly agree 16 For peer review only 17 antibiotic when making my prescribing decisions ❖ I agree 18 ❖ Neutral 19 20 ❖ I disagree 21 22 ❖ I strongly disagree 23 24 26. In regard to antibiotic guidelines, local guidelines are more ❖ I strongly agree 25 useful than international guidelines ❖ I agree 26 27 ❖ Neutral 28 29 ❖ I disagree 30 ❖ I strongly disagree 31 32 27. Antibiotic guidelines and antibiotic committees are rather ❖ I strongly agree http://bmjopen.bmj.com/ 33 34 obstacles than a help ❖ I agree 35 ❖ Neutral 36 37 ❖ I disagree 38 39 ❖ I strongly disagree 40 28. I welcome the implementation of a training program about ❖ I strongly agree on September 28, 2021 by guest. Protected copyright. 41 42 antibiotics ❖ I agree 43 44 ❖ Neutral 45 ❖ I disagree 46 47 ❖ I strongly disagree 48 49 PART 7: KNOWLEDGE ON USE OF ANTIBIOTICS 50 51 29. A 4-year-old child had diarrhoea in the last 4 days (3 stools ❖ Amoxicillin orally 52 daily). She had no fever during the past days nor at ❖ Trimethoprim/sulphamethoxazole orally 53 54 consultation. What is your treatment choice? ❖ Amoxicillin/clavulanic acid orally 55 56 57 58 Page 6 of 7 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 26 of 32 BMJ Open: first published as 10.1136/bmjopen-2019-030823 on 31 March 2020. Downloaded from

Page 27 of 32 BMJ Open

BMJ Open Page 28 of 32

Page 29 of 32 BMJ Open

BMJ Open Page 30 of 32

Page 31 of 32 BMJ Open

1 2 3 STAFFING 4 Q39 How many laboratory technologists are in the station? 5 Q40 How many microbiologists are in the station? 6 Q41 Does your laboratory have staff with Bachelor’s degree qualification or higher? Yes 7 8 Partial 9 No 10 Q42 Do you engage a Consultant clinical microbiologist(s)? Yes 11 Partial 12 For peer reviewNo only 13 CONSUMABLES 14 Q43 How often do you experience unavailability of consumables in Microbiology Yes 15 section? Eg Lack of biochemical reagents and media Partial 16 No 17 Q44 Does your lab experience delays in Antimicrobial Susceptibility Testing Yes http://bmjopen.bmj.com/ 18 due to lack of reagents? Partial 19 No 20 Q45 Do frequent stock outs lead to low demand of cultures by clinicians? Yes 21 Partial 22 No 23 BIOSAFETY 24 25 Q46 Does your lab autoclave/incinerate cultures prior to discard? Yes on September 28, 2021 by guest. Protected copyright. 26 Partial 27 No 28 Q47 Do you have handwashing facility in the laboratory? Yes 29 Partial 30 No 31 Q48 Does your lab get continuous supply of running water? Yes 32 Partial 33 No 34 Q49 Does your lab have soap supply in the handwash facility? Yes 35 Partial 36 No 37 38 39 40 41 42 43 44 45 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 BMJ Open Page 32 of 32 BMJ Open: first published as 10.1136/bmjopen-2019-030823 on 31 March 2020. Downloaded from

1 2 3 Core element 1: Senior hospital management leadership towards Yes No 4 antimicrobial stewardship 5 6 1 Has your hospital management formally identified antimicrobial stewardship 7 as a priority objective for the institution and included it in its key performance 8 indicators? 9 2 Is there dedicated, sustainable and sufficient budgeted financial support for 10 antimicrobial stewardship activities (e.g., support for salary, training, or IT 11 (information technology) support)? 12 3 Does your hospital follow any (national or international) staffing standards for 13 antimicrobial stewardship activities (e.g. number of full-time equivalent (FTE) 14 15 per 100 beds for the different members of the antimicrobial stewardship 16 team)? For peer review only 17 Core element 2: Accountability and responsibilities 18 4 Does your hospital have a formal/written antimicrobial stewardship 19 programme/strategy accountable for ensuring appropriate antimicrobial use? 20 5 Does your hospital have a formal organizational multidisciplinary structure 21 responsible for antimicrobial stewardship (e.g., a committee focused on 22 23 appropriate antimicrobial use, pharmacy committee, patient safety 24 committee or other relevant structure)? 25 6 Is there a healthcare professional identified as a leader for antimicrobial 26 stewardship activities at your hospital and responsible for implementing the 27 programme? 28 7 Is there a document clearly defining roles, procedures of collaboration and 29 responsibilities of the antimicrobial stewardship team members? 30 31 8 Are clinicians, other than those part of the antimicrobial stewardship team 32 (e.g. from the ICU, Internal Medicine and Surgery) involved in the http://bmjopen.bmj.com/ 33 antimicrobial stewardship committee? 34 9 Does the antimicrobial stewardship committee produce regularly (indicate 35 minimum time) a dedicated report which includes e.g. antimicrobial use data 36 and/or prescription improvement initiatives, with time-committed short term 37 and long term measurable goals/ targets for optimizing antimicrobial use? 38 10 Is there a document clearly defining the procedures of collaboration of the 39 40 antimicrobial stewardship team/committee with the infection prevention and on September 28, 2021 by guest. Protected copyright. 41 control team/committee? 42 Core element 3: Available expertise on infection management 43 11 Do you have access to laboratory/imaging services and to timely results to be 44 able to support the diagnosis of the most common infections at your hospital? 45 12 In your hospital are there, or do you have access to, trained and experienced 46 healthcare professionals (medical doctor, pharmacist, nurse …) in infection 47 48 management (diagnosis, prevention and treatment) and stewardship willing 49 to constitute an antimicrobial stewardship team? 50 Core element 4: Education and practical training 51 13 Does your hospital offer a range of educational resources to support staff 52 training on how to optimize antimicrobial prescribing? 53 14 Do the antimicrobial stewardship team members receive regular training in 54 antimicrobial prescribing and stewardship? 55 56 Core element 5: Other actions aiming at responsible antimicrobial use 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 33 of 32 BMJ Open BMJ Open: first published as 10.1136/bmjopen-2019-030823 on 31 March 2020. Downloaded from