Evaluation of a countrywide implementation of the world health organisation surgical safety checklist in Madagascar.

Abstract

The 2009 World Health Organisation (WHO) surgical safety checklist significantly reduces surgical mortality and morbidity (up to 47%). Yet in 2016, only 25% of East African anesthetists regularly use the checklist. Nationwide implementation of the checklist is reported in high-income countries, but in low- and middle-income countries (LMICs) reports of successful implementations are sparse, limited to single institutions and require intensive support. Since checklist use leads to the biggest improvements in outcomes in LMICs, methods of wide-scale implementation are needed. We hypothesized that, using a three-day course, successful wide-scale implementation of the checklist could be achieved, as measured by at least 50% compliance with six basic safety processes at three to four months. We also aimed to determine predictors for checklist utilization. Using a blended educational implementation strategy based on prior pilot studies we designed a three-day dynamic educational course to facilitate widespread implementation of the WHO checklist. The course utilized lectures, film, small group breakouts, participant feedback and simulation to teach the knowledge, skills and behavior changes needed to implement the checklist. In collaboration with the Ministry of Health and local hospital leadership, the course was delivered to 427 multi-disciplinary staff at 21 hospitals located in 19 of 22 regions of Madagascar between September 2015 and March 2016. We evaluated implementation at three to four months using questionnaires (with a 5-point Likert scale) and focus groups. Multivariate linear regression was used to test predictors of checklist utilization. At three to four months, 65% of respondents reported always using the checklist, with another 13% using it in part. Participant's years in practice, hospital size, or surgical volume did not predict checklist use. Checklist use was associated with counting instruments (p< 0.05), but not with verifying: patient identity, difficult intubation risk, risk of blood loss, prophylactic antibiotic administration, or counting needles and sponges. Use of a multi-disciplinary three-day course for checklist implementation resulted in 78% of participants using the checklist, at three months; and an increase in counting surgical instruments. Successful checklist implementation was not predicted by participant length of medical service, hospital size or surgical volume. If reproducible in other countries, widespread implementation in LMICs becomes a realistic possibility.