Analysing 3429 digital supervisory interactions between Community Health Workers in Uganda and Kenya: the development, testing and validation of an open access predictive machine learning web app.

BACKGROUND: Despite the growth in mobile technologies (mHealth) to support Community Health Worker (CHW) supervision, the nature of mHealth-facilitated supervision remains underexplored. One strategy to support supervision at scale could be artificial intelligence (AI) modalities, including machine learning. We developed an open access, machine learning web application (CHWsupervisor) to predictively code instant messages exchanged between CHWs based on supervisory interaction codes. We document the development and validation of the web app and report its predictive accuracy. METHODS: CHWsupervisor was developed using 2187 instant messages exchanged between CHWs and their supervisors in Uganda. The app was then validated on 1242 instant messages from a separate digital CHW supervisory network in Kenya. All messages from the training and validation data sets were manually coded by two independent human coders. The predictive performance of CHWsupervisor was determined by comparing the primary supervisory codes assigned by the web app, against those assigned by the human coders and calculating observed percentage agreement and Cohen's kappa coefficients. RESULTS: Human inter-coder reliability for the primary supervisory category of messages across the training and validation datasets was 'substantial' to 'almost perfect', as suggested by observed percentage agreements of 88-95% and Cohen's kappa values of 0.7-0.91. In comparison to the human coders, the predictive accuracy of the CHWsupervisor web app was 'moderate', suggested by observed percentage agreements of 73-78% and Cohen's kappa values of 0.51-0.56. CONCLUSIONS: Augmenting human coding is challenging because of the complexity of supervisory exchanges, which often require nuanced interpretation. A realistic understanding of the potential of machine learning approaches should be kept in mind by practitioners, as although they hold promise, supportive supervision still requires a level of human expertise. Scaling-up digital CHW supervision may therefore prove challenging. TRIAL REGISTRATION: This was not a clinical trial and was therefore not registered as such.

Reanalysis of the Pfizer mRNA BNT162b2 SARS-CoV-2 vaccine data fails to find any increased efficacy following the boost: Implications for vaccination policy and our understanding of the mode of action

BackgroundIn clinical trials two vaccinations with mRNA vaccines have shown high efficacy in preventing COVID-19. However, in the context of a pandemic, the time to generation of protective immunity, the need for and timing of a second vaccination are matters of legitimate debate. This manuscript explores the efficacy and timing of the second dose COVID-19 vaccines, including a reanalysis of data from the Pfizer mRNA BNT162b2 mRNA SARS-CoV-2 vaccine phase 3 study. Methods and findingsA non-weighted three-segment, two knot linear regression was fitted to the published cumulative infection incidence from the Pfizer BNT162b2 vaccine Phase III trial using the lspine routine in R. The optimal knot days were estimated through sensitivity analysis and the confidence limits for efficacy estimates were determined by Monte Carlo Simulations. This analysis showed the vaccine was effective from day 11 post first vaccination. The estimated efficacy over the period 11 to 28 days post first vaccination was 0.94 and there was no detectable increase in efficacy following the second vaccination. The efficacy post first vaccination substantially preceded the development of detectable serum neutralizing antibody. ConclusionsStrongly protective immunity develops rapidly following a single vaccination and at least in the short period covered by the timetable of the Phase III trial, there was no additional benefit from a second vaccination. This increases options for use of this vaccine, e.g., for ring fence vaccination, for use in travelers and for mass vaccination rollout. It highlights the need for further research into duration of immunity following a single vaccination and for understanding mechanisms of protection.

'We are the people whose opinions don't matter'. A photovoice study exploring challenges faced by community health workers in Uganda.

Understanding the experiences of community health workers (CHWs) through the use of participatory visual methods (PVMs) has been relatively underexplored. One such PVM is photovoice, which involves the capture of photographic images related to issues of social importance. In this study, we explore challenges faced by eight CHWs in Mukono District, Uganda through the use of photovoice. Over a six-week period, CHWs captured 62 relevant photographs. Subsequent individual interviews and group discussions were held with the CHWs regarding the content of the photographs. Using traditional content analysis, a range of themes related to perceived challenges faced by the CHWs were highlighted, including poor infrastructure, insufficient on-going training and supervision, relationships with other health professionals and equipment supplies. Suggestions were raised as to why such challenges existed and how they could be addressed; mainly through increased roles of the government and supporting NGOs. Overall, photovoice was generally a feasible method to highlight the challenges faced by CHWs, however community acceptability regarding image capture and consent taking may prove challenging, given past historical experiences. The use of photovoice in this study highlighted the need to address the multiple and complex challenges faced by CHWs in order to help them fulfil their roles.