The ProCCM Trial is a Randomized Control Trial designed to detect a relative difference in under-five mortality between the WHO-recommended integrated Community Case Management (iCCM) package of care for Community Health Workers (CHWs), and the intervention of CHWs providing that same care through Proactive Community Case Management (ProCCM). In both arms, participants access primary health care with no fees, and primary health centers (PHCs) received strengthening measures to deliver improved primary care. The trial protocol is published in BMJ Open.
Changes to the initial trial protocol
The following changes or deviations were made after the initial protocol was approved by the Institutional Review Board (IRB):
October 24, 2016: Amendment 1 to trial protocol and procedures, prior to baseline The ethics committee (FMPOS) in Mali approved version 2 of the ProCCM trial protocol, dated October 14, 2016. This amendment addressed several important lessons learned at the pilot site, where study procedures were tested. Changes to study procedures included stakeholder engagement, participant consent, and the survey tool. These changes were approved prior to the trial baseline and implemented at the baseline data collection.
February 24, 2017: Amendment 2 to trial protocol and procedures, after baseline The FMPOS ethics committee approved version 3 of the ProCCM trial protocol, dated February 21, 2017. This amendment addressed the realities of implementing the interventions and study procedures in the trial area at baseline. Changes included updates to the power calculations, stopping rules, and consent forms
February 26, 2017: Site-based services delivered by CHWs in intervention clusters According to the trial protocol, intervention CHWs would also provide care at a designated community health site for two hours a day in addition to their proactive home visits. However, this was renegotiated by village leaders and not implemented. Therefore, CHWs in the intervention clusters provided services at residents’ homes or at their own home when solicited by care-seeking patients. CHWs in the control clusters were instructed to only provide services at their designated community health site.
January 10, 2019: Protocol changes due to insecurity Beginning in 2018, the security situation in the study area deteriorated with the increase of violent conflicts across the Bankass district. This situation had a significant impact with respect to key activities in our study protocol, in particular the proactive search for patients (proactive case detection), access to health facilities, and service delivery by CHWs and at PHCs. Thus, to ensure the continuity of the study, we took certain measures such as the relocation or replacement of certain CHWs who were at risk in their cluster. We also introduced a mobile clinical team made up of nurses and midwives to serve affected populations in nine clusters who could not safely reach PHCs. We informed the Mali ethics committee (FMPOS) of these deviations due to insecurity on January 10, 2019, and the committee acknowledged receipt of the notice on February 11, 2019.
December 31, 2019: Added malaria sub-study to be conducted at study endline The FMPOS ethics committee in Mali approved the Addendum to the ProCCM trial protocol version 3. The Addendum described only the malaria sub-study, which was conducted during the end line ProCCM survey. This sub-study sought primarily to determine whether proactive case detection by CHWs reduced malaria prevalence in the population, including among vulnerable groups, compared to the passive case detection in the control arm.
January 14, 2020: Changes to consent forms We updated the women’s consent and assent forms, as well as the household consent form, in advance of the year 3 follow-up survey. The changes related to the list of external collaborators with whom data could be shared, and the names and telephone numbers for participants to contact in the event they had questions or concerns. No substantive changes were made to the consent forms at this time. A request to make these changes to the consent forms was approved by the Mali ethics committee in an amendment to the original study application.
July 2, 2020: Creation of an SOP for data collection procedures during the COVID-19 pandemic Faced with the COVID-19 pandemic, the Mali ethics committee (FMPOS) recommended that all human subject researchers develop protocols for the prevention of COVID-19 disease, in particular compliance with barrier measures to reduce the spread of disease. Although our planned data collections were complete at this point, we complied with the ethics committee’s request in the event that future follow-up surveys were completed.
March 12, 2021: Clarification of open cohort enrollment The original study protocol described following a cohort of study participants over the course of three years, with lifetime birth histories collected at baseline and updated in each following survey wave. At the beginning of the study, we revised the study protocol, allowing enrollment of newly eligible participants at each survey: baseline, year 1 follow-up, year 2 follow-up, and year 3 follow-up (endline). Commensurately, the survey protocol was changed to collect full birth histories for all women in each follow-up wave (year 1, year 2, and year 3), including newly enrolled and already enrolled women. However, this change in enrollment strategy was not recorded in the study protocol. We submitted this revised study protocol, along with the corresponding statistical analysis plan, to the trial’s Data Safety and Monitoring Board, which approved both the revised protocol and statistical analysis plan in 2018. This revised version of the study protocol was also submitted for peer review and published in BMJ Open in 2019. The independent trial compliance monitoring company hired by Muso, Pharmalys, pointed out to us that while we did receive a review and approval of the revised protocol from the DSMB and as part of protocol publication, we neglected to inform the study sponsor and the University of Bamako ethics committee of the revision at the time. We corrected this oversight with the ethics committee per their recommendation.
November 18, 2021: Reconciling of cluster randomization assignments The original study protocol described a stratified cluster randomization procedure that included 15 strata: one stratum less than or equal to five kilometres from the PHC and one stratum more than five kilometres from the PHC for each of the seven health catchment areas, and a fifteenth stratum that included the seven villages in which each PHC was located (one per catchment area). This randomization procedure was carried out as specified in the trial protocol on August 7, 2016. However, a prior version of the randomization procedure was conducted the prior day, August 6, 2016, and this version included 21 strata: two strata for each of the seven health catchment areas (<=5 kilometres and >5 kilometres from the PHC), and one stratum for each of the seven villages in which the PHC was located (one per catchment area). When the trial statisticians were unblinded, it was revealed while confirming the randomization assignments that in practice, the August 6, 2016 version of the randomization assignment was implemented in the field, rather than the August 7 version with the correct number of strata. We have thus followed the randomization assignments as implemented in the field in all trial analyses. In practice, this difference had no material impact on the randomisation because the difference in the two schemes only affected the seven villages <1.0 kilometre, which were evenly distributed across arms in both schemes.
The ProCCM Trial pilot site
Before launching the ProCCM Trial in the seven health catchment areas in Bankass, Muso tested the ProCCM Trial protocol in the Tori health catchment area, a pilot site that was in the same district as the Trial, but not contiguous to the study area. At the start of Muso’s activities in Tori, eight months prior to the full launch in Bankass, Tori had an estimated population of 21,347. The study design and intervention in Tori was exactly as described in the published trial protocol for the seven study sites.
Conducting the baseline survey first in Tori also allowed the team to identify technical difficulties with the tablet platform, assess the flow of the questionnaire, and to identify particular questions or sections that were challenging for respondents. It also allowed an opportunity to obtain feedback from interviewers and survey managers by discussing their experience. Muso and their research partners were able to identify and address problems related to survey implementation and revise specific questions prior to launching the baseline survey in Bankass. They were also able to refine data management systems and protocols related to the tablet-based platform.
The pilot also allowed the research team to identify and mitigate a range of operational challenges associated with the study and all aspects of the intervention prior to the full launch in Bankass.
Preparations for the intervention launch, particularly PHC reconstruction and training of CHWs, supervisors, and PHC personnel at the pilot site permitted the team to pilot these processes and identify operational challenges, such as the appropriateness of the training materials.
The intervention and study in Tori continued for the remaining three years, with follow-up surveys after 12 and 24 months. However, because of security concerns, Muso was unable to safely conduct a follow-up survey at 36 months. As in the Bankass study site, all households—regardless of whether they were assigned to the ProCCM intervention or a control cluster—now receive the full ProCCM intervention.
Comments