key: cord-0021081-yzmfciw7
authors: nan
title: Enhanced postoperative surveillance versus standard of care to reduce mortality among adult surgical patients in Africa (ASOS-2): a cluster-randomised controlled trial
date: 2021-08-19
journal: Lancet Glob Health
DOI: 10.1016/s2214-109x(21)00291-6
sha: ae023f83c693a913511f8cfecef960821d7674ca
doc_id: 21081
cord_uid: yzmfciw7

BACKGROUND: Risk of mortality following surgery in patients across Africa is twice as high as the global average. Most of these deaths occur on hospital wards after the surgery itself. We aimed to assess whether enhanced postoperative surveillance of adult surgical patients at high risk of postoperative morbidity or mortality in Africa could reduce 30-day in-hospital mortality. METHODS: We did a two-arm, open-label, cluster-randomised trial of hospitals (clusters) across Africa. Hospitals were eligible if they provided surgery with an overnight postoperative admission. Hospitals were randomly assigned through minimisation in recruitment blocks (1:1) to provide patients with either a package of enhanced postoperative surveillance interventions (admitting the patient to higher care ward, increasing the frequency of postoperative nursing observations, assigning the patient to a bed in view of the nursing station, allowing family members to stay in the ward, and placing a postoperative surveillance guide at the bedside) for those at high risk (ie, with African Surgical Outcomes Study Surgical Risk Calculator scores ≥10) and usual care for those at low risk (intervention group), or for all patients to receive usual postoperative care (control group). Health-care providers and participants were not masked, but data assessors were. The primary outcome was 30-day in-hospital mortality of patients at low and high risk, measured at the participant level. All analyses were done as allocated (by cluster) in all patients with available data. This trial is registered with ClinicalTrials.gov, NCT03853824. FINDINGS: Between May 3, 2019, and July 27, 2020, 594 eligible hospitals indicated a desire to participate across 33 African countries; 332 (56%) were able to recruit participants and were included in analyses. We allocated 160 hospitals (13 275 patients) to provide enhanced postoperative surveillance and 172 hospitals (15 617 patients) to provide standard care. The mean age of participants was 37·1 years (SD 15·5) and 20 039 (69·4%) of 28 892 patients were women. 30-day in-hospital mortality occurred in 169 (1·3%) of 12 970 patients with mortality data in the intervention group and in 193 (1·3%) of 15 242 patients with mortality data in the control group (relative risk 0·96, 95% CI 0·69–1·33; p=0·79). 45 (0·2%) of 22 031 patients at low risk and 309 (5·6%) of 5500 patients at high risk died. No harms associated with either intervention were reported. INTERPRETATION: This intervention package did not decrease 30-day in-hospital mortality among surgical patients in Africa at high risk of postoperative morbidity or mortality. Further research is needed to develop interventions that prevent death from surgical complications in resource-limited hospitals across Africa. FUNDING: Bill & Melinda Gates Foundation and the World Federation of Societies of Anaesthesiologists. TRANSLATIONS: For the Arabic, French and Portuguese translations of the abstract see Supplementary Materials section.

Protocol v2 was the first approved ASOS-2 protocol by the UCT Human Research Ethics Committee (HREC) v2 to v3 amendment 1. Page 2. Steering Committee: This has been changed to include essential local and international collaborators who have helped further refine the sample size and statistical analysis, and the trial processes. 2. Page 2. Funders of ASOS-2 have now been added to the protocol. 3. Page 6, 10, 14. Sample size amendment. Based on the coefficient of covariance of ASOS-1 we have increased the sample size to a more conservative number. This has required an increase in the recruitment period of potentially up to four weeks (in low recruiting sites). We have increased the recruitment period to four weeks at low surgical volume sites, as this decreases the coefficient of variance, which limits the sample size. 4. Page 18. Monitoring and auditing. Based on the increase in the sample size, and the potential increase in the duration of recruitment for some sites, we have added an independent international advisor (a world leading perioperative clinical trialist) who will advise on whether; i) hospital recruitment can continue after the planned recruitment window, and ii) whether an interim analysis would be required prior to continuing recruitment, if recruitment is slower than expected. 5. Page 11. Intervention arm. Based on the findings of the ASOS-2 Pilot Trial, we have made the following amendments; i) changed the definition of proximity to the nurses station, as a bed assigned to within view of the nurses' station, and ii) have added the 'Postoperative surveillance bedside guide' for all high-risk patients. 5. Page 18. Training of investigators. We have added 'remote electronic site initiation', as we believe it will be impossible for the national leaders to visit every site prior to the trial. 6. Page 6 and 20. We have increased the potential recruitment window to 4 months, as we believe that the sites will not all be ready to recruit within the same starting month. 7. The amended CRF's are attached. They also include capturing 'source data' of in-hospital mortality, which is the primary outcome. 8. There are minor grammatical changes in the protocol, to clarify (or correct the text) where questions have arisen from National Leaders, and other ASOS-2 investigators. v3 to v4 amendment Revision 1. Page 11. Section 6. 3 i. Old wording: Participating sites will be randomised to normal postoperative care or increased postoperative surveillance. Randomisation will be stratified according to the level of the surgical facility and the expected weekly surgical case-load.

ii.

New wording: Participating sites will be randomised to normal postoperative care ("SOC"), or increased postoperative surveillance ("Intervention") in four recruitment blocks (calendar time blocks). Randomisation in each block will be stratified by country and by level of the surgical facility with a fixed block size of 2. The randomisation algorithm will seek to balance randomisation arms by simulating randomisations with an automated filter to screen out poorly balanced randomisations. In the second and subsequent recruitment blocks, the randomisation algorithm will use knowledge of the previous randomisations and a similar simulation and filtering approach to ensure reasonable balance across arms and across level of surgical facility. This approach was developed on the basis of a simulation model for the randomisation that was used to evaluate the probability of overall balance of randomisation arms and balance within level of care strata if randomising recruitment phases in waves. Results of 10,000 simulation runs indicated that a balanced simulation within the bounds described below would be obtained >50% of the time without the filtering step.

The algorithm is described briefly here: For recruitment phase 1: Simulate stratified block randomisation within country and level of facility Check that simulated randomisation has percent SOC between [48% -52%] If NO, then repeat steps 1 and 2. If YES then use this randomisation. For all other recruitment phases: Simulate stratified block randomisation for this phase within country and level of facility Combine this simulation with prior actual randomisations Check that combined randomisation has percent SOC between [48% -52%] If NO, then repeat steps 1-3. If YES proceed to next check. Check that combined randomisation has balance within level of care between [45% -55%] If NO, then repeat steps 1-5. If YES then use this randomisation.

iii.

Detailed rationale/ justification/ explanation: The randomisation procedure was incompletely described in the previous protocol. The updated description of the randomisation procedure was tested through simulation and describes the procedure more comprehensively. 15-17. Section 7.5 i.

Old wording: Not applicable ii. New wording: 7.5.

Process evaluation

During the conduct of the ASOS-2 Trial we will perform a concurrent 'process evaluation' to gain insight into the implementation process. The process evaluation aims to: 1.

Measure protocol compliance (fidelity of implementation), 2. Identify factors that influence success of implementation, 3.

Verify specified process steps that form part of the implementation of the protocol and logical framework of the trial, and 4.

Generate contextual information about the trial setting.

The questions asked by the process evaluation are: 1.

Which factors determined fidelity of implementation at the local hospital level? 2.

How did implementation of the intervention affect patient care? 3.

Why does the intervention produce (or fail to produce) a change in patient morbidity and mortality?

The measurement tools used for the process evaluation are: 1.

The post-education, REDCap online test, 2.

The individual participant case record form (CRF), 3.

An online database, called the screening log, which is a REDCap application for participating hospitals to record the number of eligible patients daily, 4.

Research fellows will perform selected site visits in order to verify specified process steps and perform structured interviews with a sample of stakeholders, 5.

Telephonic structured interviews with a sample of stakeholders that cannot be reached by the two research fellows, and 6.

A post-trial, semi-quantitative, REDCap based, online questionnaire built around the key components identified in the structured interviews and the pilot trial.

Each member of the local hospital investigator team will be asked to complete the post-education online test prior to patient recruitment. The online test comprises eight short multiple-choice questions which test the comprehension of the key points from the education material. The individual participant CRF (Appendix 3) captures information that allows checking accuracy of risk stratification, fidelity of implementation, and differences in patient experience between non-high risk patients in the two arms. The research fellows will visit specific sites during the trial on appointed days agreed upon by the local hospital investigators. They will visit sites during the initiation phase, the recruitment phase, and the follow-up phase of the trial. They will use standard site visit checklists to identify non-compliance, barriers and facilitators of implementation. During the follow-up phase they will interview a sample of stakeholders who indicate willingness to participate in a qualitative interview about their personal experience of the trial. Verbal consent will be obtained and recorded for each interviewee who agrees to participate. No personal identifying information will be recorded for interviewees. Interviews will be recorded. A semistructured script will be used for the interviews. Recordings will be stored on a password protected cloud based drive. Interviews will be analysed deductively by independent investigators using the Consolidated Framework for Implementation Research (CFIR) as a guide. The recordings will be destroyed following analysis.

The post-trial questionnaire will be anonymous. At the start of the questionnaire respondents will be informed about its content and intent. Respondents will be given the opportunity to opt out, or to give consent prior to continuing to the questions. Questions will test the CFIR constructs that were highlighted in the structured stakeholder interviews.

ii.

Detailed rationale/ justification/ explanation: The previous protocol did not describe any process evaluation. Various experienced international trialists urged the steering committee to describe a formal process evaluation as part of the trial. The process evaluation is essential for translation of the findings of the trial to subsequent clinical practice. I will allow improved understanding of the trial context which influences implementation and effectiveness of the intervention.

i. Old wording: Not applicable ii.

New wording: He [referring to the independent international advisor] will also check the event rate in the control arm of the study when at least 80% of all recruited patients have been captured on REDCap. If the event rate of the control arm is lower than predicted, the independent international advisor will use the event rate in the control arm to decide whether recruitment of further hospitals is required beyond the original sample size of 664 hospitals, in order to maintain 80% power for the primary outcome. This specified interim analysis of data will not lead to adjustment of the prespecified alpha of 0.05.

iii.

Detailed rationale/ justification/ explanation: Event rate driven recruitment is an established method to ensure that a trial does not conclude with a sample that is underpowered for the outcome(s) of interest. Such an outcome would waste the considerable time, effort and resources that have been invested by thousands of researchers across Africa. Testing the event rate does not test any of the outcomes of the trial and does therefore not incur any penalty for repeated testing.

i. Old wording: Status at hospital discharge or 30 th postoperative in-hospital day:

ii. New wording: Status at hospital discharge or 30 th postoperative in-hospital day:

 Alive & still in hospital  Dead  Alive & discharged  if alive and discharged, was patient transferred to another facility for higher care?  Yes  No iii.

Detailed rationale/ justification/ explanation: The same amendment is made to both the control and intervention case report forms. This amendment to the case report forms allows us to distinguish between patients who have been discharged home or to a convalescent facility (i.e. experienced a favourable outcome) compared to patients who were transferred to another hospital for higher level of care (i.e. lost to follow-up). Detailed rationale/ justification/ explanation: Two steering committee members added. Patrice Forget is a crucial member of the trial team and will be the only Francophone steering committee member. Tim Stephens is a crucial member of the process evaluation team bringing expertise in this area to the steering committee.

iii.

Old wording: The inter-cluster correlation coefficient (ICC) in ASOS was 0.01. iv.

New wording: The intra-cluster correlation coefficient (ICC) in ASOS was 0.01.

Detailed rationale/ justification/ explanation: The correct term is intra-cluster correlation coefficient. This was a spelling mistake. Revision 3. Pages 15. Section 7. 1 i. Old wording: Based on a relative risk reduction of 25% in the intervention arm, the power for the secondary outcome, based on the sample sizes for the primary outcome are shown in Table 1. ii.

New wording: Based on a relative risk reduction of 25% in the intervention arm, the sample sizes for the primary outcome are shown in Table 1. iii.

Detailed rationale/ justification/ explanation: The sample size refers to the primary outcome, not the secondary outcome.

vii. Old wording: To determine whether increased postoperative surveillance reduces in-hospital mortality in high-risk adult surgical patients aged 18 years and over in Africa.

viii. New wording: To determine whether increased postoperative surveillance in high-risk adult surgical patients reduces overall in-hospital mortality in surgical patients aged 18 years and over in Africa.

ix. Detailed rationale/ justification/ explanation: The sentence structure of the outcome definition was corrected as it became apparent that the former version suggests we are only interested in the outcome of high-risk patients, whereas, we are principally interested in the outcome of all patients in the trial.

vi. Old wording: Recruitment end date August 2019 vii.

New wording: Recruitment end date March 2020 viii.

Detailed rationale/ justification/ explanation: At the end of August we had not reached our recruitment target. The progress of the trial was evaluated by our external auditor, Professor Paul Myles, Monash University, who advised that the trial continue. We plan to provide a next progress report to Professor Paul Myles during January 2020.

iv.

Old wording: To determine whether increased postoperative surveillance reduces in-hospital mortality in high-risk adult surgical patients aged 18 years and over in Africa.

v. New wording: To determine whether increased postoperative surveillance in high-risk adult surgical patients reduces overall in-hospital mortality in adult surgical patients aged 18 years and over in Africa.

Detailed rationale/ justification/ explanation: Duplication of revision 1.

i. Old wording: To determine whether increased postoperative surveillance reduces the incidence of the composite of severe in-hospital complications and mortality in high-risk adult surgical patients aged 18 years and over in Africa.

ii. New wording: To determine whether increased postoperative surveillance in high-risk adult surgical patients reduces the overall incidence of the composite of severe in-hospital complications and mortality in adult surgical patients aged 18 years and over in Africa.

iii. Detailed rationale/ justification/ explanation: Duplication of revision 1.

i. Old wording: Recruitment will run in 2019.

ii. New wording: Recruitment will run until March 2020.

iii. Detailed rationale/ justification/ explanation: Duplication of revision 2.

i. Old wording:

ii. New wording:

Adult ≥ 18 years, undergoing in-patient surgery 

In-hospital mortality

In-hospital composite of severe complications and mortality Usual postoperative care

All patients

In-hospital mortality

In-hospital composite of severe complications and mortality

iii. Detailed rationale/ justification/ explanation: The flow diagram was updated to make it clear that the hospital is the unit of randomisation in this trial and that within intervention hospitals it is only the subset of high-risk patients who should receive increased postoperative surveillance.

The statistical analysis plan was added to version 6. 

In-hospital mortality

In-hospital composite of severe complications and mortality

This hospital is participating in a surgical trial, known as the African Surgical OutcomeS-2 (ASOS-2) Trial. The ASOS-2 Trial will determine whether increased postoperative surveillance reduces in-hospital death in high-risk adult surgical patients across Africa. Hospitals are randomised to provide usual care, or increased surveillance of high-risk patients. All adult patients aged 18 years and over who have surgery in this hospital are included in this trial. All patients in this hospital will receive similar postoperative surgical management, based on which arm of the trial that this hospital has been randomised to. Potential harm to participants: We do not believe that this trial provides any potential harms to participants. You may not benefit directly from participation in this trial, however the results of this trial may help us to provide better care to future surgical patients in Africa. Confidentiality: The study doctor and his/her study team will look at your personal health information and collect only the information they need for the study. This information includes:

• Your name, gender and date of birth • Information about your surgery, and test results before your surgery and while you are in hospital after the surgery • Information about your health status and health outcomes The information that is collected for the study will be kept in a locked and secure area by the study doctor. Only the study team or the people or groups listed below will be allowed to look at your records. Your participation in this study also may be recorded in your medical record at this hospital. The following people may come to the hospital to look at the study records and at your personal health information to check that the information collected for the study is correct and to make sure the study followed proper laws and guidelines:

• The study lead investigator(s) or his/her representatives • Representatives of the University of Cape Town Research Ethics Board Some of your health information will be sent outside of the hospital to the trial lead investigators and stored at a secure research database hosted by Safe Surgery South Africa, of the South African Society of Anaesthesiologists. Any information about you that is sent out of the hospital will have a code and will not show your name or address, or any information that directly identifies you. All information collected during this study, including your personal health information, will be kept confidential and will not be shared with anyone outside the study unless required by law. You will not be named in any reports, publications, or presentations that may come from this study. Treatment Options: If you choose not to be in this study you will still get the best available treatment and be managed according to standard international guidelines. If you decide to leave the study, the information about you that was collected before you left the study will still be used. No new information will be collected without your permission.

The results of this study may be presented at conferences or seminars and may be published in a medical journal. All information during presentations will be anonymous. Study results will be available from the study investigators once the study is complete and the data is analysed.

There is no cost to participate in this study. You will not be reimbursed for your participation. Compensation for Injury: If you suffer a physical injury from participation in this study, medical care will be provided to you in the same manner as you would ordinarily obtain any other medical treatment. Participation in this trial does not waive your legal rights nor release the study doctors, the hospital from their legal and professional responsibilities. Reasonable medical expenses incurred through researchrelated injuries will be covered by the University of Cape Town's no-fault insurance cover. Participation and Withdrawal: Participation in any research study is voluntary. If you choose not to participate, you and your family will continue to have access to surgical care at this hospital. If you decide to participate in this study, you can change your mind without giving a reason, and you may withdraw from the study at any time without any effect on the care you and your family will receive any Hospital. New Information: If the trial investigators learn new information that affects your participation, you will be informed in a timely manner. The study protocol and consent form have been reviewed by a committee called the Human Research Ethics Committee at University of Cape Town. The Human Research Ethics Committee is a group of scientists, medical staff, and individuals from other backgrounds (including law and ethics) as well as members from the community. The committee is established by the hospital to review studies for their scientific and ethical merit. The Board pays special attention to the potential harms and benefits involved in participation to the research participant, as well as the potential benefit to society. This committee is also required to do periodic review of ongoing research studies. As part of this review, someone may contact you from the Human Research Ethics Committee to discuss your experience in the research study.

If you require any further information or have any concerns, please feel free to contact the investigators or study coordinator: _______________________________

 Site initiation educational material attached  Upload ethics approval  Upload Hospital Leader GCP certificate  Weeks three and two before recruitment begins Identify local ASOS-2 team of at least two people, ideally three or more  Hospital Leader to educate Local ASOS-2 team with site initiation educational material  Local ASOS-2 team to complete site initiation online questionnaire after completing educational material  Local ASOS-2 team decides who will perform each task (data collection, follow up)  Local ASOS-2 team members to educate all involved departments e.g. surgery department etc., about the trial 

 Site initiation educational material attached  Upload ethics approval  Upload Hospital Leader GCP certificate  Weeks three and two before recruitment beginsIdentify local ASOS-2 team of at least two people, ideally three or more  Hospital Leader to educate Local ASOS-2 team with site initiation educational material  Local ASOS-2 team to complete site initiation online questionnaire after completing educational material  Local ASOS-2 team decides who will perform each task (risk calculation, interventions, follow up)  Local team decides how interventions will be instituted in the hospital  Local ASOS-2 team members to educate all involved departments e.g. surgery department etc., about the trial and the intervention procedures The patient has at least one of the following: i. purulent drainage from the superficial incision ii. organisms isolated from an aseptically obtained culture of fluid or tissue from the superficial incision and at least one of the following signs or symptoms of infection: pain or tenderness, localized swelling, redness, or heat, or superficial incision is deliberately opened by surgeon and is culture positive or not cultured. A culture-negative finding does not meet this criterion. iii. diagnosis of an incisional surgical site infection by a surgeon or attending physician

An infection which involves both superficial and deep parts of surgical incision and meets the following criteria: a. Infection occurs within 30 days after surgery if no surgical implant is left in place or one year if an implant is in place and b. The infection appears to be related to the surgical procedure and involves deep soft tissues of the incision (e.g. fascial and muscle layers) and c.

The patient has at least one of the following: i. purulent drainage from the deep incision but not from the organ/space component of the surgical site ii. a deep incision spontaneously dehisces or is deliberately opened by a surgeon and is culture-positive or no cultures were taken whilst the patient has at least one of the following signs or symptoms of infection: fever (>38°C) or localized pain or tenderness. A culture-negative finding does not meet this criterion. iii. an abscess or other evidence of infection involving the deep incision is found on direct examination, during surgery, or by histopathologic or radiologic examination iv. diagnosis of a deep incisional surgical site infection by a surgeon or attending physician III.

Surgical site infection (organ/space) An infection which involves any part of the body excluding the fascia or muscle layers and meets the following criteria: a. Infection occurs within 30 days after surgery and b. The infection appears to be related to the surgical procedure and involves any part of the body, excluding the skin incision, fascia, or muscle layers, that is opened or manipulated during the operative procedure and c.

The patient has at least one of the following: i. purulent drainage from a drain that is placed through a stab wound into the organ/space ii. organisms isolated from an aseptically obtained culture of fluid or tissue in the organ/ space iii. an abscess or other evidence of infection involving the organ/space that is found on direct examination, during reoperation, or by histopathologic or radiologic examination iv. diagnosis of an organ/space surgical site infection by a surgeon or attending physician

An infection which is not related to infection at another site and which meets at least one of the following criteria: a. Patient has a recognised pathogen cultured from blood cultures which is not related to an infection at another site b. Patient has at least one of the following signs or symptoms: fever (>38°C), chills, or hypotension and at least one of the following:

i. common skin contaminant cultured from two or more blood cultures drawn on separate occasions ii. common skin contaminant cultured from at least one blood culture from a patient with an intravascular line, and a physician starts antimicrobial therapy iii. positive blood antigen test V.

Acute Respiratory Distress Syndrome (ARDS)

Respiratory failure, or new or worsening respiratory symptoms, commencing within one week of surgery; and a chest radiograph or computed tomography scan which demonstrates bilateral opacities not fully explained by effusions, lobar/lung collapse, or nodules; and respiratory failure not fully explained by cardiac failure or fluid overload. Need objective assessment (e.g. echocardiography) to exclude hydrostatic oedema if no risk factor is present. Severity grading: Severe: PaO2:FiO2 ≤100 mmHg with PEEP ≥5 cmH2O Guidance: If altitude is higher than 1000 m, a correction factor should be calculated as follows: (PaO2:FiO2 x [barometric pressure/760 mmHg]). PEEP, positive endexpiratory pressure; CPAP, non-invasive continuous positive airways pressure

Chest radiographs with new or progressive and persistent infiltrates, or consolidation, or cavitation, and at least one of the following: a. fever (>38°C) with no other recognized cause b. leucopaenia (<4,000 white blood cells/mm3) or leucocytosis (>12,000 white blood cells/mm3) c.

for adults >70 years old, altered mental status with no other recognised cause; d. and at least two of the following:

i. new onset of purulent sputum or change in character of sputum, or increased respiratory secretions, or increased suctioning requirements ii. new onset or worsening cough, or dyspnoea, or tachypnoea iii. rales or bronchial breath sounds iv. worsening gas exchange (hypoxaemia, increased oxygen requirement or increased ventilator demand) Guidance: Two radiographs are required for patients with underlying pulmonary or cardiac disease. The definition may be used to identify ventilator associated pneumonia.

An infection associated with at least one of the following signs or symptoms which should be identified within a 24 hour period; fever (>38 °C), urgency, frequency, dysuria, suprapubic tenderness, costovertebral angle pain or tenderness with no other recognised cause, and a positive urine culture of ≥105 colony forming units/mL with no more than two species of microorganisms. Guidance: Baseline serum creatinine must be measured before surgery but an estimated value can be used if the patient does not have chronic kidney disease.

Blood loss occurring within 72 hours after the end of surgery which would normally result in transfusion of blood.

The cessation of cardiac mechanical activity, as confirmed by the absence of signs of circulation. ECG changes may corroborate the incidence of cardiac arrest.

If any of the following complications result in a significant prolongation of hospital stay and/or permanent functional limitation or death, then mark 'Other severe complication' as 'Yes'. Note that they will almost always requires clinical treatment. a.

Leak of luminal contents from a surgical connection between two hollow viscera. The luminal contents may emerge either through the wound or at the drain site, or they may collect near the anastomosis, causing fever, abscess, septicaemia, metabolic disturbance and/or multiple-organ failure. 

Other prespecified sensitivity and subgroup analyses on the effectiveness outcomes were also conducted; i) the potential impact of unobserved outcomes on the primary and secondary effectiveness estimates, ii) subgroups treated as stratification variables; including hospital level, recruitment wave, and income category of country, and iii) analysis stratified by individual patient and surgical characteristics that form part of the ASOS Surgical Risk Calculator. The full prespecified additional secondary analyses are shown below.

The following prespecified sensitivity and subgroup analyses on the effectiveness outcomes were also conducted. 1. Sensitivity analyses investigated the potential impact of unobserved outcomes (transfer out or lost to follow up) on the primary and secondary effectiveness estimates, where individuals with unobserved outcomes were alternately assumed to be alive at discharge (censored at thirty days) or to have died in-hospital within 30 days. 2. Primary endpoints were also analysed under the same model as the primary analysis for the following subgroups treated as stratification variables: i) hospital level (primary, secondary, tertiary), iii) recruitment wave, and iii) income category of country (low or middle, according to the World Bank classification in 2020).

3. The relative risk of experiencing the primary and secondary outcome by exposure to the intervention (defined below) was estimated in stratified models for the following individual characteristics: i) age (years), ii) American Society of Anesthesiologists (ASA) status (1) (2) (3) (4) (5) , iii) surgical procedure category (gynaecologic, obstetric, neurosurgery, ear nose and throat (ENT), orthopaedic, plastics and breast, urology, gastrointestinal and hepatobiliary, cardio-thoracics, and other), iv) indication for surgery (non-communicable disease, caesarean section, trauma, infection), v) urgency of surgery (elective, urgent, emergent), and vi) severity of surgery (minor, intermediate, major).

(Secondary analyses included two analytic approaches to two per protocol populations, based on implementation fidelity. In the first per protocol analysis, we compared all patients from hospitals with data in the standard of care arm to all patients from hospitals with data in the intervention arm where the intervention hospital provided the intervention with fidelity to at least 80% of high-risk patients. Patients from hospitals in the intervention arm where the hospital delivered the intervention with fidelity to fewer than 80% of patients will be excluded from this analysis. In the second per protocol analysis, we compared all patients from hospitals with data in the standard of care arm to all patients from hospitals in the top two tertiles of implementation fidelity. Patients from intervention hospitals in the bottom tertile of implementation fidelity were excluded from this analysis. At the hospital level we report implementation fidelity as the proportion of high-risk patients who received the intervention with fidelity. We used two definitions for implementation fidelity; i) provision of at least the high-risk bedside guide plus one additional component of the intervention on days 0 and 1 postoperatively and ii) provision of at least 2 components of the intervention on days 0 and days 1 postoperatively, regardless whether the high-risk beside guide is one of the components.) Per protocol definition 1; where the intervention hospital provided the intervention with fidelity to at least 80% of high-risk patients Per protocol definition 2; hospitals in the top two tertiles of implementation fidelity Fidelity definition 1; provision of at least the high-risk bedside guide plus one additional component of the intervention on days 0 and 1 postoperatively Fidelity definition 2; provision of at least any two components of the intervention on days 0 and days 1 postoperatively, which did not necessarily have to include the high-risk beside guide as one of the components ASA=American Society of Anesthesiologists ASA=American Society of Anesthesiologists

Mpoki Ulisubisya MMed (Anaesthesia and Critical Care), Muhimbili University of Health and Allied Sciences

Mikiyas Teferi DM, National Coordinator for Surgical Improvement Case Team, Ministry of Health

Hôpital de la mère et de l'enfant, Lagune de Cotonou, Benin Supplementary Material S3. Randomisation algorithm The randomisation algorithm is described briefly here: For recruitment phase 1: 1. Simulate stratified block randomisation within country and level of facility. 2. Check that simulated randomisation has percent

If NO, then repeat steps 1 and 2. If YES then use this randomisation. For all other recruitment phases: 1. Simulate stratified block randomisation for this phase within country and level of facility. 2. Combine this simulation with prior actual randomisations

Check that combined randomisation has percent 'standard of care' between

If NO, then repeat steps 1-3. If YES proceed to next check

Check that combined randomisation has balance within level of care between

If NO, then repeat steps 1-5. If YES then use this randomisation

Standards for definitions and use of outcome measures for clinical effectiveness research in perioperative medicine: European Perioperative Clinical Outcome (EPCO) definitions: a statement from the ESA-ESICM joint taskforce on perioperative outcome measures