Implementation of a Clinical Decision Support System for Antimicrobial Prescribing in Sub-Saharan Africa: Multisectoral Qualitative Study.

BACKGROUND: Suboptimal use of antimicrobials is a driver of antimicrobial resistance in West Africa. Clinical decision support systems (CDSSs) can facilitate access to updated and reliable recommendations. OBJECTIVE: This study aimed to assess contextual factors that could facilitate the implementation of a CDSS for antimicrobial prescribing in West Africa and Central Africa and to identify tailored implementation strategies. METHODS: This qualitative study was conducted through 21 semistructured individual interviews via videoconference with health care professionals between September and December 2020. Participants were recruited using purposive sampling in a transnational capacity-building network for hospital preparedness in West Africa. The interview guide included multiple constructs derived from the Consolidated Framework for Implementation Research. Interviews were transcribed, and data were analyzed using thematic analysis. RESULTS: The panel of participants included health practitioners (12/21, 57%), health actors trained in engineering (2/21, 10%), project managers (3/21, 14%), antimicrobial resistance research experts (2/21, 10%), a clinical microbiologist (1/21, 5%), and an anthropologist (1/21, 5%). Contextual factors influencing the implementation of eHealth tools existed at the individual, health care system, and national levels. At the individual level, the main challenge was to design a user-centered CDSS adapted to the prescriber's clinical routine and structural constraints. Most of the participants stated that the CDSS should not only target physicians in academic hospitals who can use their network to disseminate the tool but also general practitioners, primary care nurses, midwives, and other health care workers who are the main prescribers of antimicrobials in rural areas of West Africa. The heterogeneity in antimicrobial prescribing training among prescribers was a significant challenge to the use of a common CDSS. At the country level, weak pharmaceutical regulations, the lack of official guidelines for antimicrobial prescribing, limited access to clinical microbiology laboratories, self-medication, and disparity in health care coverage lead to inappropriate antimicrobial use and could limit the implementation and diffusion of CDSS for antimicrobial prescribing. Participants emphasized the importance of building a solid eHealth ecosystem in their countries by establishing academic partnerships, developing physician networks, and involving diverse stakeholders to address challenges. Additional implementation strategies included conducting a local needs assessment, identifying early adopters, promoting network weaving, using implementation advisers, and creating a learning collaborative. Participants noted that a CDSS for antimicrobial prescribing could be a powerful tool for the development and dissemination of official guidelines for infectious diseases in West Africa. CONCLUSIONS: These results suggest that a CDSS for antimicrobial prescribing adapted for nonspecialized prescribers could have a role in improving clinical decisions. They also confirm the relevance of adopting a cross-disciplinary approach with participants from different backgrounds to assess contextual factors, including social, political, and economic determinants.

[Thromboembolic manifestations in 36 HIV-infected patients in West Africa]. / Manifestations thromboemboliques chez 36 patients Ouest Africains infectés par le VIH.

In HIV-infected patients thromboembolic disease is a complication linked to heightened risk. In Ivory Coast no study has been conducted on HIV-infected patients treated in HIV Services. The aim of our study is to describe HIV-associated thromboembolic manifestations in patients treated or untreated with antiretroviral drugs whose data were collected in the Infectious and Tropical Diseases Service (ITDS). We conducted a retrospective study by reviewing the medical records of HIV-infected patients hospitalized with deep vein thrombosis (DVT), arterial thrombosis and/or pulmonary embolism over the period January 2005-July 2015. Diagnosis was based on Doppler ultrasound of vessels and/or on thoracic angioscanner. Diagnostic, therapeutic and evolutionary features of thromboembolic manifestations in these patients were analyzed. The medical records of 36 patients, including 23 women (64%), with a sex-ratio M/F of 0.57 and an average age of 43±12 years were selected. Deep venous thrombosis (DVT) was found in 26 (72.2%) patients, pulmonary embolism (PE) in 9 (25%) patients and arterial thrombosis in 1 patient (2.8%). DVT was unilateral in 81% of cases and predominantly left-sided in 77% of cases. PE was unilateral and right-sided in 100% of cases while arterial thrombosis was bilateral in 2.7% of cases. In patients with DVT, the femoral vein (39%) and the popliteal vein (35%) were most commonly affected by thrombosis. PE involved the pulmonary arteries in 77.8% of cases while arterial thrombosis involved the left and right internal carotid. The majority of patients was under antiretroviral treatment (69%). The most commonly associated opportunistic infections included oral candidiasis (31%) and tuberculosis (33%). Nine patients died (25%). This study highlights high rates of DVT in HIV-infected patients. Other studies are necessary to better understand the role of HIV in the occurrence of thromboembolic disease.