The overall aim of this study is to compare the proportion of antimicrobial prescriptions adherent with prescribing guidelines delivered by mobile phone application (app) versus paper-based prescribing guidelines in both in- and outpatients in six hospitals in Laos.

This is an open cohort stepped-wedge cluster randomized controlled trial 3-step trial (4-month intervals at each step) with a 4-month pre-intervention period). The study conducts in six hospitals across Laos. This is 16 months-study. The intervention is a MicroGuide antimicrobial prescribing guideline mobile phone application and antimicrobial stewardship (AMS) training; whilst the reference is paper-based antimicrobial prescribing guidelines. The main outcome of interest is the proportion of antimicrobial prescriptions adherent with prescribing guidelines delivered by app-based versus paper-based prescribing guidelines at month 16 (12 months' exposure in each group). Outcome measurements are Point prevalence surveys (PPS) of hospital antimicrobial use, prescriber questionnaires and AMS knowledge survey of the prescribers. Simple random sampling will be used to select the order (step) of interventions implementation for a group of hospitals (two hospitals per group and one group per step). Neither hospitals nor investigators are blinded to the timing of intervention. Open Data Kit (ODK) system will be used as data collection instrument. All data will be kept securely, protected by password access with automatic daily backup. A logistic mixed-effects regression model will be used to compare the proportion of antimicrobial prescription adherence measured cross-sectionally at month 0, 4, 8, 12 and month 16 after introduction of prescribing guidelines delivered by mobile phone application versus paper-based. This study was approved by University of Health Sciences, Vientiane, Lao PDR and Oxford Tropical Ethics Committee (OxTREC), University of Oxford, United Kingdom.

Background Antimicrobial prescribing applications are a promising tool to improve prescribing. However, there has been no evidence from trials to support whether the antimicrobial prescribing guidelines in the format of a mobile phone application are superior to conventional means (i.e. paper-based) in low- and middle-income countries (LMIC). This trial could have a significant impact in Laos and in other LMICs, and inform policy globally, by improving prescribing and hence reducing AMR.

Objectives Primary objective To compare the proportion of antimicrobial prescriptions adherent with prescribing guidelines delivered by mobile phone application (app-based) versus paper-based prescribing guidelines in both in- and outpatients in six general hospitals in Laos

Secondary objectives are to compare:

1. change in antimicrobial defined daily doses (DDDs) after introduction of paper-based versus app-based guidelines + antimicrobial stewardship (AMS) training in both in- and out patients
2. change in proportion of patients receiving intravenous antimicrobials route between paper-based and app-based guidelines + AMS training in both in- and out patients
3. AMS knowledge level of the prescribers before and after app-based guidelines and AMS training
4. prescriber satisfaction with/acceptability of paper-based versus app-based guidelines
5. antimicrobial costs before and after app-based guidelines and AMS training
6. proportion of patients admitted to hospital receiving an antimicrobial prescription before and after paper-based; and the app-based guidelines + AMS training Methods This is an open cohort stepped wedge cluster randomized controlled trial in six hospitals across Laos. The six hospitals have been purposively selected based on their capacity, location, previous participation in PPS program with available hospital antimicrobial use and antimicrobial susceptibility data. The sample size of 6 hospitals was estimated assuming 30% of antimicrobial prescription adherence with the paper-based prescribing guideline, and assuming 0.05 intra-cluster correlation coefficient. This will have 80% of power of the test to be able to detect 15% minimum difference after introduction of the mobile phone app prescribing guidelines at 0.05 significant level. This is a 3-step trial with one baseline measurement after introduction of the paper-based prescribing guidelines. Six hospitals have been divided into 3 pairs based on the geographical location and convenience for conducting the surveys.

Study intervention and reference

1. Reference phase

Study duration is divided into two parts (reference and intervention phases). The order of activities at the start of the project are listed below:

1. PPS on hospital antimicrobial use for all 6 hospitals
2. AMS knowledge survey
3. Pre-prescribing guidelines survey
4. Access to paper prescribing guidelines for all 6 hospitals with introductory session
5. Receive PPS reports (within a month after PPS) After month 0, for all hospitals which remain in the reference phase, PPS of hospital antimicrobial use is conducted every 4 months without other activities from the project.

2. Intervention phase

1. Antimicrobial Prescribing guidelines mobile phone application (MicroGuide) Prescribing guidelines were transferred to MicroGuide; which is a medical guideline application. The app is installed to the personal smart phones of local physicians by the PI. The project do not provide a smart phone to any physician without one. The contents of the mobile app guidelines are the same as the paper version but using hyperlinks to information, rather than displaying it as in paper guidelines. Both mobile app and paper guidelines are available in Lao language.
2. Antimicrobial stewardship (AMS) training AMS training sessions are given to all prescribers in participating hospitals when firstly enter to the intervention phase. This session is given alongside an introduction of mobile phone app prescribing guideline. The training details are based on the existing information in the paper-based therapeutic and prophylactic antimicrobial prescribing guidelines.

Intervention delivery

The intervention starts at month 4 after introducing the paper-based guidelines. This intervention is introduced in one pair of hospitals at a time and repeated every 4 months. The order of the activities during intervention phase of each step are listed below:

1. PPS on hospital antimicrobial use
2. AMS knowledge survey
3. Prescriber feedback survey on paper-based guidelines
4. Access to MicroGuide app-based guidelines with introductory session
5. AMS training session
6. Receive PPS reports

Note:

1. During the intervention period, prescribers in participating hospitals still have access to paper-based prescribing guidelines distributed at the start of the study.

   Contamination prevention

   There is a possibility of contamination of the app to the reference group by doctors from the intervention group:

   • move to the reference group (training or changing workplace)

   • give smart phones to other doctors in the reference group (unlikely)

   • lose smart phones that are subsequently found and used by doctors in a reference group (unlikely) To avoid or reduce the problem of contamination, all participating hospitals are requested not to share the mobile app guidelines to others while hospitals are still under evaluation. The app guidelines will be installed by the PI. This app cannot be accessed without the approval from the PI (entering passcode). The security system is controlled by the MicroGuide team and PI of the project.

   Endpoints Primary endpoint: The proportion of antimicrobial prescriptions adherent with prescribing guidelines delivered by app-based versus paper-based guidelines of 12 months' exposure in each group Secondary endpoints

   1. Total antimicrobial DDDs prescribed in app-based versus paper-based groups of 12 months' exposure in each group 2. Proportion of patients prescribed antimicrobials by the intravenous route in app-based versus paper-based groups of 12 months' exposure in each group 3. AMS knowledge level of prescribers prescribed in app-based versus paper-based groups of 12 months' exposure in each group 4. Prescriber satisfaction/acceptability of the guidelines in app-based versus paper-based groups of 4 months' exposure in each group 5. Total costs of antimicrobial prescribed in app-based versus paper-based groups of 12 months' exposure in each group 6. Proportion of patients admitted to hospital receiving an antimicrobial prescription before and after introduction of the paper-based; and the app-based guidelines + AMS training of 12 months' exposure in each group Definitions of adherence to the guidelines

   1. Full adherence with the guidelines=using correct antimicrobial agent(s) and correct dose based on the provided antimicrobial prescribing guidelines
   2. Partial adherence with the guidelines=using correct antimicrobial agent(s) with incorrect dose based on provided antimicrobial prescribing guidelines
   3. Non-adherence with the guidelines=using a different antimicrobial agent(s) compared to that suggested in provided antimicrobial prescribing guidelines

   Outcome measurement

   • Point prevalence surveys of hospital antimicrobial prescribing and antimicrobial costing in both in- and outpatients (every 4 months until month 16)
   • Prescriber surveys (month 0, paper-based antimicrobial prescribing guidelines survey before the implementation of the mobile phone app guidelines, mobile phone app-based antimicrobial prescribing guidelines survey 4 month after the implementation of the mobile phone app guidelines
   • Antimicrobial stewardship knowledge survey (month 0, before the implementation of the mobile phone app guidelines and AMS training and at month 16) Data collection method and monitoring

   IPD and OPD PPS data will be collected in Open Data Kit (ODK) system on patients' characteristics and treatment status. Data on the frequency of use of the mobile app can also be downloaded from the MicroGuide App. Prescribing behaviors are collected via survey forms. All data will be entered to MACRO system. Only antibiotic cost data is collected. All complete survey forms are kept securely in a locked filling cabinet. Only PI of the project and project data management team have access to the raw data. Data checking and cleaning are done by PI. Data inconsistency will be discussed with the research team immediately.

   Statistical analysis A logistic mixed effects regression model will be used to compare the proportion of antimicrobial prescriptions adherence measured cross-sectionally at month 0, 4, 8, 12 until month 16 after introduction of prescribing guidelines by mobile phone application versus paper-based. The model will compare all wedges in the intervention from month 4 to month 16 with all wedges in the reference arm from month 0 to month 12. Likert scale analysis will be used to evaluate prescriber satisfaction/acceptability of paper-based compared to mobile phone app delivered antimicrobial prescribing guidelines. Mann-Whitney test will be used to assess the differences between paper-based population and mobile phone app population. The overall proportion of correct answers from AMS survey in each time point will be compared. Cost of antibiotic therapy will be calculated from the start date of treatment to the survey day. DDD will be calculated using data from PPS in both IPD and OPD.

   Safety considerations/Adverse Events No adverse event monitoring is planned Ethics Ethical approvals were received from the University of Health Sciences Ethics Committee, Ministry of Health, Lao PDR and the Oxford Tropical Research Ethics Committee (OxTREC), University of Oxford, United Kingdom to conduct this project.

   Benefits This first therapeutic and prophylactic antimicrobial prescribing guidelines will encourage the rational use of antimicrobials. The mobile phone application version will allow clinicians to access guidelines easier. This will help to improve clinical decision of doctors and consistency of care. This will give the potential to improve health outcomes of patients, reduce morbidity and mortality and give value for money.

   Risks There might be some risk that MoH would like to introduce the guidelines and the app before the study finishes if the results are very positive. In addition, it could be challenging to manage and control the travel of Lao doctors from one hospital to another as trainings, workshops or conferences (not related to this project) are conducted across the country. Moreover, another AMR program is also conducted in three participating hospitals (Salavan, Xiengkhang and Luang Namtha) to improve the quality of infectious diagnosis in these three hospitals. This program plans to start at the same time as this project.

   Patient and doctor confidentiality There will not be any record of patients' personal details to be able to identify any individual. There will not be any direct interaction between the research team and patients.

   Trial insurance The University of Oxford would operate in the event of any participant suffering harm as a result of their involvement in the research.

   Data sharing and dissemination Survey data from this study will be shares using AMU-Laos app (computer dashboard). Preliminary results will be shared with the Lao Ministry of Health and the Wellcome Trust. Paper reports will be shared with participating hospitals, local public health offices, and the MoH. Oral presentation and posters will be used as tools to deliver progress updates and the results of the study in national, regional and international conferences. Results will also be published in open access peer reviewed medical journals.

   Finance This research is supported by the Wellcome Trust of Great Britain, UK. No funding sources have any role in study design, data collection, or preparation of the protocol.

Inclusion Criteria:

• Six hospitals have been purposively selected based on their capacity, location, previous participation in our PPS program with available hospital antimicrobial use and antimicrobial susceptibility data

Exclusion Criteria:

• Other hospitals in Laos which are in indicated in the inclusion criteria

we are not recruiting individual participants. Therefor, this is not applicable and data will be shared according to Mahidol Oxford Tropical Medicine Research Unit (MORU) data sharing policy