Post-Ebola Community Health Worker programme performance

The devastating 2014-2015 Ebola outbreak in Sierra Leone Background: could erode the gains of the health system including the Community Health Worker (CHW) programme. We conducted a study to ascertain if the positive trend in reporting cases of malaria, pneumonia and diarrhoea treated by CHWs in the post-Ebola period has been sustained 18 months post-Ebola. We conducted a retrospective cross-sectional study using Methods: aggregated CHW programme data (2013-2017) from all Primary Health Units in Kenema district. Data was extracted from the District Health Information System and analysed using STATA. Data in the pre(June 2013-April 2014), during(June 2014-April 2015) and post-Ebola recovery (June 2016-April 2017) periods was compared and analysed for reporting completeness; Rapid Diagnostic Tests (RDTs) performed and cases of malaria, diarrhoea and pneumonia treated per month. Differences across periods were tested using two-sample t-test with significance set at 0.05. CHW reporting increased from pre-Ebola by 8% ( -value=0.29) Results: p intra-Ebola and 19% ( -value=0.012) post-Ebola. Compared to the p pre-Ebola period, in the post-Ebola recovery period, there was a significant increase in the mean monthly reported RDTs of 35% ( -value=0.020); p malaria treatments 66% ( -value<0.001); and pneumonia treatments p increased by 80% ( -value=0.004). Conversely, the mean monthly p diarrhoea cases treated decreased by 20% ( -value=0.16) in the p post-Ebola period. The resiliency demonstrated by the CHW programme during Conclusion: and immediately after the Ebola outbreak has been sustained in the post-Ebola recovery period. Continued programme investments in supportive supervision and financial incentives for CHWs will be critical to ensure uninterrupted contribution towards Sustainable Development Goal 3.

Creative Commons Attribution IGO , which permits copying, adaptation and distribution in any medium or format for any purpose, provided the original work is properly cited, a Licence link is provided to the license, and any changes made are indicated. Any such copying, adaptation and distribution must not in any way suggest that World Health Organisation endorses you or your use. Thomas

Introduction
The effect of the 2014-2015 West African Ebola outbreak on Community Health Worker (CHW) services in Kenema district, Sierra Leone, was assessed through a retrospective crosssectional study comparing CHW reporting and services before (June 2013-April 2014), during (June 2014-April 2015) and post-outbreak (November 2015-April 2016) 1 . The study found CHW reporting completeness and reported treatments for malaria increased post-Ebola, while those for pneumonia and diarrhoea returned to pre-outbreak levels. Results showed CHWs stopped performing Rapid Diagnostic Tests (RDTs) during the Ebola period and did not resume until after the outbreak. The study recommended additional investigations into preliminary positive trends in the early post-Ebola period to determine if these post-outbreak tendencies represent a resumption of pre-Ebola CHW programme performance or a continued progression. Therefore, we aimed to understand whether health system performance was sustained 18 months after the outbreak by reporting the trend and comparing CHW system performance (reporting completeness, reported diagnosis and treatment services among reports received) in Kenema district in the pre-, intra-and post-Ebola recovery periods for children under five years.

Methods
This was a retrospective cross-sectional study using aggregate CHW programme data from Kenema district, Sierra Leone. Details of Kenema and the CHW programme were previously described 1  CHW programme data were extracted from the electronic Ministry of Health and Sanitation (MoHS) District Health Information Systems (DHIS2), which is aggregated from monthly summary sheets submitted by PHU supervisors. Data on the following variables were extracted per month: facility reporting completeness; reported malaria RDTs; treatment for malaria, diarrhoea and pneumonia. In this study we defined CHW reporting completeness as number of facilities with fully (100%) complete reporting for CHW monthly summary sheets over the total number of expected facility reports. We imported data into STATA v14.2 2 . We produced descriptive summary statistics and examined statistical differences between periods using two-sample t-test, with significance set at 0.05.
The Sierra Leone Ethics and Scientific Review Committee (dated 18 December 2018) and the Ethics Advisory Group of the International Union Against Tuberculosis Lung Disease (UAG number 65/18) provided ethics and institutional approval. Since aggregate programme data were used, participant consent was not sought.

Results
Proportion of PHUs reporting complete reports Figure 1 shows the proportion of facilities with CHW complete reports monthly in the pre-, intra-and post-Ebola recovery periods. The number of PHUs submitting complete CHW reports ranged between 39-129 (27-89%) and 98-134 (68-92%) in the pre-and intra-outbreak periods respectively. Conversely, in the recovery period, 136 (94%) facilities consistently completed reports each month.
Mean monthly RDTs, malaria, diarrhoea and pneumonia cases reported Figure 2 shows trends across pre-, intra-and post-Ebola recovery periods for CHW programme performance indicators. Table 1 shows that the mean monthly reported RDTs had an increasing trend in the pre-, intra-and post-Ebola recovery periods, with increases of 11% intra-Ebola and 35% (p-value=0.012) post-Ebola compared to the pre-Ebola period.
The mean monthly diarrhoea treatments reported exhibited a downward trend, in the pre-and intra-Ebola periods, followed by a small uptick in the post-Ebola recovery period. However, compared to the pre-Ebola period, overall monthly mean reported treatments significantly decreased by 42% (p-value=0.013) in the intra-Ebola, while the reduction in the post-Ebola recovery period was not statistically significant (p-value=0.16).
The mean reported monthly pneumonia treatments declined by 11% intra-Ebola but grossly increased by 80% (p-value=0.004) during the recovery period compared to the pre-Ebola period ( Table 1).

Discussion
Our study results indicate a general trend of the CHW programme in Kenema district sustaining comparable performance levels into the post-Ebola recovery period as compared to pre-Ebola for reporting completeness and reported RDTs, malaria and pneumonia treatments. Comparable to results from the previous study 1 and the region 3 , this may be due to sustained investment in key areas affecting supply and services. The Integrated Disease Surveillance and Response approach, which emphasises active community-based surveillance and completeness of reporting, was strengthened in the intra-Ebola period 3 . In addition, supportive supervision, a critical element to the success of CHW programmes 4 , was revived after the outbreak 3 . Finally, financial incentives for CHWs initiated during the outbreak were subsequently incorporated into the national policy in 2016 5 . Furthermore, the CHWs have reported increased community awareness of signs and symptoms of common childhood illnesses, translating into increased demand for services 3 .
We found fewer missing reports in the pre-and intra-Ebola periods compared to the previous study 2 , most likely due to retrospective data entry into the DHIS2. While we found higher mean reported RDTs performed in the post-Ebola period, we still observed fewer in the intra-Ebola period corresponding with the enactment of the "no touch policy," similar to the previous report 1,3,6 .
Our results reveal an absolute decline in the reported diarrhoea treatments during the recovery period. The promotion of hygiene practices through community sensitization 7 and institution of bye-laws by community stakeholders in the intra-Ebola period may have had lasting effects on behaviour and thus contributing to the reduction in reported diarrhoea treatments during the post-Ebola period 3 . However, the difference in reported treatments between pre-and post-Ebola periods was not found to be significant, so this trend may warrant future investigation.
Utilising pre-Ebola service levels for comparisons, while useful, still reflect benchmarks of weak systems functioning, a factor which contributed to the impact of the Ebola outbreak itself 8 . Therefore, it is imperative that such comparisons be evaluated in this light. Therefore, in order to achieve the vision of the Sustainable Development Goal (SDG) for good health and wellbeing beyond pre-Ebola-level benchmarks, sustained investments in supportive supervision and financial incentives for CHWs are essential for the programme.
A strength of the study is the use of complete district PHU data for the study period. Furthermore, we adhered to the Strengthening the Reporting of Observational Studies in Epidemiology guidelines for the reporting of observational data and sound ethical principles 9 . Primary weaknesses of our study were the use of routinely collected data, which was influenced by reporting completeness, thus restricting reporting of morbidity; lack of data triangulation; inability to validate the electronic data base against the raw data and to generalise to the national CHW programme.  In conclusion, although our study established a sustained trend towards the pre-Ebola CHW service levels 18 months after the outbreak, there is need for continued investment in the CHW programme to continue gains in programme performance in order to contribute towards SDG 3.

Grant information
The programme was funded by the Special Programme for Research and Training in Tropical Diseases hosted at the World Health Organization (TDR).
The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.