Assessing the sustainability of the Systems Analysis and Improvement Approach to increase HIV testing in family planning clinics in Mombasa, Kenya: results of a cluster randomized trial

Study design and randomization

This study was a two-stage cluster-randomized trial to evaluate the use of SAIA to improve HTC at family planning clinics in Mombasa, Kenya. The trial design and Stage 1 results have been previously reported [16]. In brief, SAIA is a blended implementation strategy that iteratively uses a 5-step cycle to improve performance across care cascades [11]. Twenty-four family planning clinics in Mombasa were selected for study inclusion and randomized to either receive the SAIA implementation strategy (n=12) or to be included as controls receiving standard care (n=12). Restricted randomization of clinics (1:1) was conducted based on clinic size and delivery of HTC services prior to the study start. Due to the nature of the implementation strategy delivered to clinics in the intervention arm, participating clinics were not blinded. Randomization was conducted by an independent statistician at the Center for AIDS Research Biometrics Core at the University of Washington who did not serve in any other role in the study.

County DOHS collaborators

Kenya has a decentralized system of government, in which the national Ministry of Health (MOH) provides policy, but each county independently operates a County Department of Health Services. In Mombasa County, an Executive of Health oversees two branches, Public Health and Medical Services, which are each led by a Chief Officer and Director. Under this leadership, the County Health Management Team (CHMT) operates across both branches, divided into departments to address important health topics. For this research, we worked with the Reproductive Health (RH) Officer and HIV/Sexually Transmitted Infections (STI) Officer. These officers oversee sub-county RH and STI coordinators, who have direct oversight over family planning clinics within their respective sub-counties. The sub-county STI and RH coordinators are primarily nurses and clinical officers by training who have risen to a supervisory role through years of service and professional development. Their primary role is to supervise the delivery of STI and RH services, respectively, in their sub-county jurisdictions. As part of their standard duties, RH and STI coordinators visit family planning clinics monthly to address any problems and track the progress of programmatic activities. All study activities were conducted in coordination with collaborators within the Mombasa County DOHS.

Stage 1: Study setting

Trial Stage 1 was conducted from December 2018 to November 2019. During this time, Mombasa County did not experience any systematic disruptions to the healthcare system. At the study start, Mombasa had approximately 170 family planning clinics, including public and private facilities. All facilities receive HIV-testing supplies at no cost from the Mombasa County DOHS. HIV-testing commodities are tracked on MOH-provided registers. Specific training and certification are required to perform HTC. In the context of this study, the counseling aspect of HTC refers to pre-test counseling, in which care providers recommend opt-out HIV testing and ask family planning clients if they are willing to be tested. Anyone who reports a previous HIV-positive diagnosis is not eligible for HIV testing. For each family planning client, MOH-provided registers record if they received this counseling, HIV serostatus at the time of counseling, and if they received HIV testing. These data were used to calculate HTC rates in clinics in the intervention arm versus control clinics, with all new family planning clients considered eligible for counseling, and all new clients who did not have a previous HIV-positive diagnosis eligible for testing.

Stage 1: Procedures

During Stage 1 of the trial, study staff implemented SAIA at each clinic in the intervention arm. The SAIA steps and roles played by the clinic staff and facilitators are explained in Table 1. As previously described [16], this included the creation of a “cascade analysis tool,” an Excel-based system for quantifying and displaying the number of individuals who complete each step of a process to identify where improvement may be needed [9, 17]. The tool also shows the expected impact on HIV testing when each step of the cascade is optimized to full performance. Cascade analysis was followed by sequential process flow mapping, in which study staff helped clinic staff to map clinic processes to identify modifiable bottlenecks in their workflow for HTC. Study staff then worked with clinic staff to conduct plan-do-study-act (PDSA) cycles, in which they identified workflow modifications that clinic staff would implement during the following month (termed “micro-interventions”) chosen to address barriers to implementing HTC specific to each clinic, then evaluate during the following cycle. Study staff conducted monthly SAIA visits with clinic staff at facilities in the intervention arm to assess the implementation and impact of the micro-interventions with real-time data input into the cascade analysis tool and plan a micro-intervention for the next month. Micro-intervention activities were enacted by the clinic staff at each family planning clinic over the following month. Examples of micro-interventions implemented in Stage 1 have been previously published [16]. Research staff conducted monthly SAIA visits for 12 months at each participating clinic in the intervention arm, during which time study data on HTC outcomes were also collected.

Table 1 Description of the 5-step SAIA cycle conducted with intervention clinics

Control clinics were aware of the study but did not receive any of the SAIA implementation strategy components described above and instead continued with standard care and delivery of HTC. Kenya MOH National Guidelines recommend integration of HTC at family planning clinics, and this is overseen by DOHS RH and STI coordinators [18]. However, no specific strategies are in place to promote HTC uptake at these clinics. Study staff visited control clinics every 3 months to collect data, but otherwise had no interaction with control clinic staff. During this stage, the County DOHS leadership were updated regularly regarding study activities and served in an advisory role.

At the end of study Stage 1, there was a brief gap in the trial before Stage 2 was launched. During this time, study staff continued to actively deliver SAIA at clinics in the intervention arm.

Power and sample size

Sample size estimates are based on Stage 1 of the trial and have been described previously [16]. Briefly, sample size determination was made based on an average of 15 new family planning clients per clinic per 3-month period, 20% HIV testing among new clients in the control clinics, and a 50% increase in HTC with the SAIA implementation strategy. At an alpha level of 0.05 and a two-sided test, the inclusion of 11 clinics per study arm would provide 80% power to detect this effect in clinics in the intervention arm compared to control clinics. To allow for potential loss to follow up of one clinic per arm, 24 clinics were randomized. Twenty-three clinics remained in follow-up throughout both stages of the study.

Stage 2: Study setting

Trial Stage 2 was conducted from February 2020 to January 2021. In March 2020, the first cases of COVID-19 were detected in Kenya. This led to government-mandated restrictions beginning March 18, 2020, including curfews, travel restrictions, school closures, bans on gatherings of >15 people, and drastic restrictions on public transportation. In Mombasa County, a full lockdown was issued for some areas between May 6 to July 7. In October and November 2020, healthcare workers in Mombasa began a “go-slow” period, in which healthcare services were restricted to emergencies only, followed by a full strike from December 28, 2020, to February 19, 2021. These events resulted in temporary closures of some family planning clinics, reduced staff and capacity at public clinics that remained open, and reduced capacity at the county level to oversee family planning clinics. While these events were disruptive to care delivery, the purpose of Stage 2 of the trial was to assess if both SAIA delivery and higher HTC performance in family planning clinics would be sustained when implemented by the County in real-world circumstances, so it did not impact the timeline of data collection for this study.

Stage 2: Procedures

In study Stage 2, clinics maintained the study arm that they were randomized to in the first year of the study. To support the transition of SAIA implementation to the Mombasa County DOHS, the research team trained 16 sub-county RH and STI Coordinators who were appointed by the RH and HIV/STI Officers as “implementers” to conduct SAIA visits at each clinic in the intervention arm as part of their normally scheduled monthly family planning clinic supervision visits. Four teams of STI and RH Coordinators conducted SAIA visits at all clinics in the intervention arm within their coverage area (between 1 and 6 clinics per team, depending on coverage area).

Trainings led by study staff provided the sub-county RH and STI Coordinators with an overview of SAIA, practice in collecting and recording data, and mock SAIA visits. The implementers were then responsible for traveling to assigned clinics in the intervention arm for SAIA visits to conduct or update flow mapping (as necessary), cascade analysis, and development and assessment of micro-interventions. The targeted SAIA visit schedule was one visit per month to each clinic for 12 months. The initial Stage 2 SAIA cycle at each clinic was conducted by the County implementers with oversight and mentorship from the study staff. After this initial mentored hand-off, study staff were available to answer questions and conducted periodic check-ins to monitor progress. Aside from the initial mentored cycle, the study staff did not participate in the implementation of SAIA during Stage 2 of the trial. County implementers were provided tablets preloaded with training materials that they used for data collection at sites. No other funding support or incentives were given by the study to complete these visits. County implementers reported to study staff when each SAIA visit was completed to allow study staff to track when DOHS-led SAIA cycles were conducted at each site. To avoid behavior changes induced by observation, study staff did not attend any additional SAIA cycle meetings after the mentored cycle, and therefore, we were unable to collect data on potential adaptations made at SAIA visits led by DOHS staff.

Three types of data were collected during this stage of the study. First, DOHS-appointed implementers recorded information about each SAIA visit in a REDCap questionnaire, which provided fidelity data on completion of SAIA cycles. Second, the clinical outcomes of interest (number of new family planning clients, number counseled, number tested) were independently collected quarterly by study staff directly from register data at each family planning clinic. Third, interviews were conducted with clinic staff and managers to assess barriers and facilitators to uptake and sustainment of HTC and SAIA, and reflections on why improvements were or were not sustained at their clinic.

Participants

Data for this study were collected at the clinic level. Family planning clinic staff and managers worked with the STI and RH coordinators to implement SAIA at each clinic and participated in exit interviews at the end of the study. This study did not involve direct contact with family planning clients, and all client data were de-identified and aggregated.

Outcomes

Outcomes of interest included sustained delivery of SAIA and maintenance of improvements in HTC observed in Stage 1 of the trial. Sustained delivery of SAIA was measured as the proportion of the planned monthly SAIA visits that were completed. Interviews with clinic staff were used to provide context for gaps in SAIA delivery. Interviews also assessed institutionalization of HTC, as well as facilitators and barriers that impacted sustainment of SAIA implementation and HTC at clinics in the intervention arm.

HTC delivery was measured as the proportion of new family planning clients tested for HIV in the final 3 months of the study, comparing clinics in the intervention arm to control clinics. A secondary measure of continued HTC was the proportion of new family planning clients who received pre-test counseling in the final 3 months in clinics in the intervention arm compared to controls. These outcome variables were collected quarterly even if no SAIA visits had been conducted and were recorded as zero new family planning clients during temporary clinic closures.

Statistical analysis

The primary analysis followed an intent-to-treat design based on the arm of the trial each clinic was randomized to, regardless of participation in SAIA implementation procedures. Study data were collected each quarter (Q) of the 2-year study, with Stage 2 data collected in Q5–Q8. We calculated prevalence rate ratios (PRR) using Poisson regression with a log link, comparing the rates of HIV testing in the final 3 months of Stage 2 (Q8) in clinics in the intervention arm versus control clinics. A secondary analysis used the same method to compare rates of pre-test counseling in the intervention arm compared to the control. For both outcomes, we also examined if performance differed between public and private family planning clinics. Results were stratified by public versus private family planning clinics if an interaction term p value was <0.05. As an exploratory analysis, we further examined HTC rates over the course of Stage 2 using a difference-in-differences analysis in which we compared the change in HIV testing and counseling rates from Q5 to Q8 at clinics in the intervention arm versus control clinics. All analyses used Stata version 15.1 (College Station, TX, USA, 2017).

All clinics in the intervention arm (n=12) were invited to participate in exit interviews after study Stage 2 was complete. An interview guide was developed using adapted measures from the Consolidated Framework for Implementation Research (CFIR) interview guide tool (https://cfirguide.org/) and a psychometrically validated tool developed by Weiner et al. [19]. Interviews were conducted among family planning clinic staff members who were involved in SAIA implementation, and were analyzed using a rapid assessment approach guided by the CFIR [20] and the Implementation Outcomes Framework [21]. This analysis includes interview responses focused on barriers and facilitators of sustaining the implementation strategy (SAIA) and the evidence-based intervention (HTC), with a specific focus on CFIR domains of intervention characteristics, inner setting, and process, as well as concepts specific to sustainability, such as institutionalization. Interviews were recorded through field notes and audio recordings (GW). A structured codebook was created to allow for categorization of elicited constructs and emergent themes and was populated by two coders using field notes (GW, JL). Interpretation of coding was discussed iteratively between the two coders until consensus was reached.

Ethical considerations

This research was approved by the Kenyatta National Hospital-University of Nairobi Ethics and Research Committee, the Human Subjects Research Institutional Review Board at the University of Washington, and the Mombasa County DOHS. This trial is registered at ClinicalTrials.gov (NCT02994355). All Mombasa County DOHS implementers verbally agreed to participate in the trial, and clinic staff and managers who participated in interviews provided written informed consent prior to COVID-19 and verbal assent for remote interviews during the pandemic.

留言 (0)

沒有登入
gif