Evaluation of malaria outbreak detection methods, Uganda, 2022.

BACKGROUND: Malaria outbreaks are detected by applying the World Health Organization (WHO)-recommended thresholds (the less sensitive 75th percentile or mean + 2 standard deviations [2SD] for medium-to high-transmission areas, and the more sensitive cumulative sum [C-SUM] method for low and very low-transmission areas). During 2022, > 50% of districts in Uganda were in an epidemic mode according to the 75th percentile method used, resulting in a need to restrict national response to districts with the highest rates of complicated malaria. The three threshold approaches were evaluated to compare their outbreak-signaling outputs and help identify prioritization approaches and method appropriateness across Uganda. METHODS: The three methods were applied as well as adjusted approaches (85th percentile and C-SUM + 2SD) for all weeks in 2022 for 16 districts with good reporting rates ( ≥ 80%). Districts were selected from regions originally categorized as very low, low, medium, and high transmission; district thresholds were calculated based on 2017-2021 data and re-categorized them for this analysis. RESULTS: Using district-level data to categorize transmission levels resulted in re-categorization of 8/16 districts from their original transmission level categories. In all districts, more outbreak weeks were detected by the 75th percentile than the mean + 2SD method (p < 0.001). For all 9 very low or low-transmission districts, the number of outbreak weeks detected by C-SUM were similar to those detected by the 75th percentile. On adjustment of the 75th percentile method to the 85th percentile, there was no significant difference in the number of outbreak weeks detected for medium and low transmission districts. The number of outbreak weeks detected by C-SUM + 2SD was similar to those detected by the mean + 2SD method for all districts across all transmission intensities. CONCLUSION: District data may be more appropriate than regional data to categorize malaria transmission and choose epidemic threshold approaches. The 75th percentile method, meant for medium- to high-transmission areas, was as sensitive as C-SUM for low- and very low-transmission areas. For medium and high-transmission areas, more outbreak weeks were detected with the 75th percentile than the mean + 2SD method. Using the 75th percentile method for outbreak detection in all areas and the mean + 2SD for prioritization of medium- and high-transmission areas in response may be helpful.

Evaluation of the sentinel yellow fever surveillance system in Uganda, 2017-2022: strengths and weaknesses.

BACKGROUND: Uganda has a sentinel surveillance system in seven high-risk sites to monitor yellow fever (YF) patterns and detect outbreaks. We evaluated the performance of this system from 2017 to 2022. METHODS: We evaluated selected attributes, including timeliness (lags between different critical time points), external completeness (proportion of expected sentinel sites reporting ≥ 1 suspect case in the system annually), and internal completeness (proportion of reports with the minimum required data elements filled), using secondary data in the YF surveillance database from January 2017-July 2022. We conducted key informant interviews with stakeholders at health facility and national level to assess usefulness, flexibility, simplicity, and acceptability of the surveillance system. RESULTS: In total, 3,073 suspected and 15 confirmed YF cases were reported. The median time lag from sample collection to laboratory shipment was 37 days (IQR:21-54). External completeness was 76%; internal completeness was 65%. Stakeholders felt that the surveillance system was simple and acceptable, but were uncertain about flexibility. Most (71%) YF cases in previous outbreaks were detected through the sentinel surveillance system; data were used to inform interventions such as intensified YF vaccination. CONCLUSION: The YF sentinel surveillance system was useful in detecting outbreaks and informing public health action. Delays in case confirmation and incomplete data compromised its overall effectiveness and efficiency.

Individual and household risk factors for Ebola disease among household contacts in Mubende and Kassanda districts, Uganda, 2022.

BACKGROUND: In 2022, an Ebola disease outbreak caused by Sudan virus (SUDV) occurred in Uganda, primarily affecting Mubende and Kassanda districts. We determined risk factors for SUDV infection among household members (HHM) of cases. METHODS: We conducted a case-control and retrospective cohort study in January 2023. Cases were RT-PCR-confirmed SUDV infection in residents of Mubende or Kassanda districts during the outbreak. Case-households housed a symptomatic, primary case-patient for ≥ 24 h and had ≥ 1 secondary case-patient with onset < 2 weeks after their last exposure to the primary case-patient. Control households housed a case-patient and other HHM but no secondary cases. A risk factor questionnaire was administered to the primary case-patient or another adult who lived at home while the primary case-patient was ill. We conducted a retrospective cohort study among case-household members and categorized their interactions with primary case-patients during their illnesses as none, minimal, indirect, and direct contact. We conducted logistic regression to explore associations between exposures and case-household status, and Poisson regression to identify risk factors for SUDV infection among HHM. RESULTS: Case- and control-households had similar median sizes. Among 19 case-households and 51 control households, primary case-patient death (adjusted odds ratio [ORadj] = 7.6, 95% CI 1.4-41) and ≥ 2 household bedrooms (ORadj=0.19, 95% CI 0.056-0.71) were associated with case-household status. In the cohort of 76 case-HHM, 44 (58%) were tested for SUDV < 2 weeks from their last contact with the primary case-patient; 29 (38%) were positive. Being aged ≥ 18 years (adjusted risk ratio [aRRadj] = 1.9, 95%CI: 1.01-3.7) and having direct or indirect contact with the primary case-patient (aRRadj=3.2, 95%CI: 1.1-9.7) compared to minimal or no contact increased risk of Sudan virus disease (SVD). Access to a handwashing facility decreased risk (aRRadj=0.52, 95%CI: 0.31-0.88). CONCLUSION: Direct contact, particularly providing nursing care for and sharing sleeping space with SVD patients, increased infection risk among HHM. Risk assessments during contact tracing may provide evidence to justify closer monitoring of some HHM. Health messaging should highlight the risk of sharing sleeping spaces and providing nursing care for persons with Ebola disease symptoms and emphasize hand hygiene to aid early case identification and reduce transmission.

Understanding the delay in identifying Sudan Virus Disease: gaps in integrated disease surveillance and response and community-based surveillance to detect viral hemorrhagic fever outbreaks in Uganda, September 2022.

BACKGROUND: Early detection of outbreaks requires robust surveillance and reporting at both community and health facility levels. Uganda implements Integrated Disease Surveillance and Response (IDSR) for priority diseases and uses the national District Health Information System (DHIS2) for reporting. However, investigations after the first case in the 2022 Uganda Sudan virus outbreak was confirmed on September 20, 2022 revealed many community deaths among persons with Ebola-like symptoms as far back as August. Most had sought care at private facilities. We explored possible gaps in surveillance that may have resulted in late detection of the Sudan virus disease (SVD) outbreak in Uganda. METHODS: Using a standardized tool, we evaluated core surveillance capacities at public and private health facilities at the hospital level and below in three sub-counties reporting the earliest SVD cases in the outbreak. Key informant interviews (KIIs) were conducted with 12 purposively-selected participants from the district local government. Focus group discussions (FGDs) were conducted with community members from six villages where early probable SVD cases were identified. KIIs and FGDs focused on experiences with SVD and Viral Hemorrhagic Fever (VHF) surveillance in the district. Thematic data analysis was used for qualitative data. RESULTS: Forty-six (85%) of 54 health facilities surveyed were privately-owned, among which 42 (91%) did not report to DHIS2 and 39 (85%) had no health worker trained on IDSR; both metrics were 100% in the eight public facilities. Weak community-based surveillance, poor private facility engagement, low suspicion index for VHF among health workers, inability of facilities to analyze and utilize surveillance data, lack of knowledge about to whom to report, funding constraints for surveillance activities, lack of IDSR training, and lack of all-cause mortality surveillance were identified as gaps potentially contributing to delayed outbreak detection. CONCLUSION: Both systemic and knowledge-related gaps in IDSR surveillance in SVD-affected districts contributed to the delayed detection of the 2022 Uganda SVD outbreak. Targeted interventions to address these gaps in both public and private facilities across Uganda could help avert similar situations in the future.

Rapid antiretroviral therapy initiation following rollout of point-of-care early infant diagnosis testing, Uganda, 2018-2021.

BACKGROUND: Uganda Ministry of Health (MOH) recommends a first HIV DNA-PCR test at 4-6 weeks for early infant diagnosis (EID) of HIV-exposed infants (HEI) and immediate return of results. WHO recommends initiating antiretroviral therapy (ART) ≤ 7 days from HIV diagnosis. In 2019, MOH introduced point-of-care (POC) whole-blood EID testing in 33 health facilities and scaled up to 130 facilities in 2020. We assessed results turnaround time and ART linkage pre-POC and during POC testing. METHODS: We evaluated EID register data for HEI at 10 health facilities with POC and EID testing volume of ≥ 12 infants/month from 2018 to 2021. We abstracted data for 12 months before and after POC testing rollout and compared time to sample collection, results receipt, and ART initiation between periods using medians, Wilcoxon, and log-rank tests. RESULTS: Data for 4.004 HEI were abstracted, of which 1.685 (42%) were from the pre-POC period and 2.319 (58%) were from the period during POC; 3.773 (94%) had a first EID test (pre-POC: 1.649 [44%]; during POC: 2.124 [56%]). Median age at sample collection was 44 (IQR 38-51) days pre-POC and 42 (IQR 33-50) days during POC (p < 0.001). Among 3.773 HEI tested, 3.678 (97%) had test results. HIV-positive infants' (n = 69) median age at sample collection was 94 (IQR 43-124) days pre-POC and 125 (IQR 74-206) days during POC (p = 0.04). HIV positivity rate was 1.6% (27/1.617) pre-POC and 2.0% (42/2.061) during POC (p = 0.43). For all infants, median days from sample collection to results receipt by infants' caregivers was 28 (IQR 14-52) pre-POC and 1 (IQR 0-25) during POC (p < 0.001); among HIV-positive infants, median days were 23 (IQR 7-30) pre-POC and 0 (0-3) during POC (p < 0.001). Pre-POC, 4% (1/23) HIV-positive infants started ART on the sample collection day compared to 33% (12/37) during POC (p < 0.001); ART linkage ≤ 7 days from HIV diagnosis was 74% (17/23) pre-POC and 95% (35/37) during POC (p < 0.001). CONCLUSION: POC testing improved EID results turnaround time and ART initiation for HIV-positive infants. While POC testing expansion could further improve ART linkage and loss to follow-up, there is need to explore barriers around same-day ART initiation for infants receiving POC testing.

Group-based Education and monitoring program delivered by community health workers to improve control of high blood pressure in island districts of lake victoria, Uganda.

BACKGROUND: Individuals living in communities with poor access to healthcare may be unaware of their high blood pressure (BP). While the use of community health workers (CHWs) can address gaps in human resources for health, CHWs in Uganda have not been used previously for BP screening and management. We report the results of an initiative to train CHWs to evaluate BP and to administer group-based education in Kalangala and Buvuma Island Districts of Lake Victoria, Uganda. METHODS: We randomly selected 42 of 212 villages. We trained CHWs based in island districts on measuring BP. CHWs visited all households in the selected villages and invited all adults ≥ 18 years to be screened for high BP. We used the World Health Organization's STEPwise tool to collect data on demographic and behavioral characteristics and BP measurements. High blood pressure was defined as systolic BP (SBP) ≥ 140 mm Hg and/or diastolic BP (DBP) ≥ 90 mm Hg over three readings. CHWs created and led fortnight support groups for individuals identified with high blood pressure at baseline. At each group meeting, CHWs re-measured BP and administered an intervention package, which included self-management and lifestyle education to participants. The paired t-test was used to compare mean values of systolic blood pressure (SBP) and diastolic blood pressure (DBP) before and after the intervention. Generalized estimating equations (GEE) were used to model longitudinal changes in BP. RESULTS: We trained 84 CHWs to measure BP and deliver the intervention package. Among 2,016 community members, 570 (28.3%) had high blood pressure; of these, 63 (11.1%) had a previous diagnosis of hypertension. The comparison of SBP and DBP before and after the intervention revealed significant reductions in mean SBP from 158mmHg (SD = 29.8) to 149 mmHg (SD = 29.8) (p < 0.001) and mean DBP from 97mmHg (SD = 14.3) to 92mmHg (p < 0.001). GEE showed decreases of -1.133 (SBP) and - 0.543 mmHg (DBP)/fortnight. CONCLUSION: High BP was common but previously undiagnosed. The CHW-led group-based self-management and education for controlling high BP was effective in the island districts in Uganda. Scaling up the intervention in other hard-to-reach districts could improve control of high BP on a large scale.

Improving maternal and neonatal outcomes among pregnant women who are HIV-positive or HIV-negative through the Saving Mothers Giving Life initiative in Uganda: An analysis of population-based mortality surveillance data.

HIV infection is associated with poor maternal health outcomes. In 2016, the maternal mortality ratio (MMR) in Uganda was 336/100,000, and the neonatal mortality rate (NMR) was 19/1,000. Saving Mothers, Giving Life (SMGL) was a five-year maternal and neonatal health strengthening initiative launched in 2012 in Uganda. We extracted maternal and neonatal data for 2015-2016 from the initiative's population-based mortality surveillance system in 123 health facilities in Western Uganda. We collected data on the facilities, HIV status, antiretroviral drug (ARV) use, death, birth weight, delivery type, parity, Apgar scores, and complications. We compared mother and baby outcomes between HIV-positive or HIV-negative, computed risk ratios (RR) for adverse outcomes, and used the chi-square to test for significance in differences observed. Among 116,066 pregnant women who attended and gave birth at SMGL-implementing facilities during 2015-2016, 8,307 (7.7%) were HIV-positive, of whom 7,809 (94%) used antiretroviral drugs (ARVs) at the time of delivery. During birth, 23,993 (21%) women experienced ≥1 complications. Neonate Apgar scores <7 (8.8%) and maternal haemorrhage during birth (1.6%) were the most common outcomes. Overall facility MMR was 258/100,000 and NMR was 7.6/1,000. HIV infection increased risk of maternal death (RR = 3.6, 95% Confidence Interval (CI) = 2.4-5.5), maternal sepsis (RR = 2.1, 95% CI = 1.3-3.3), and infant birth weight <2,500g (RR = 1.2, 95% CI = 1.1-1.3), but was protective against maternal complications (RR = 0.92, 95% CI = 0.87-0.97) and perinatal death (RR = 0.78, 95% CI = 0.68-0.89). Among the HIV-positive, ARV non-use increased risk of maternal death (RR = 15, 95% CI = 7.1-31) and perinatal death (RR = 2.3, 95% CI = 1.6-3.4). SMGL reduced facility MMR and NMR below national rates. HIV-infection was associated with maternal sepsis and death. Failure to use ARVs among women living with HIV increased the risk of maternal and perinatal death. Use of the SMGL approach and complementary interventions that further strengthen HIV care, may continue to reduce MMR and NMR.

Performance and impact of contact tracing in the Sudan virus outbreak in Uganda, September 2022-January 2023.

BACKGROUND: Contact tracing (CT) is critical for ebolavirus outbreak response. Ideally, all new cases after the index case should be previously-known contacts (PKC) before their onset, and spend minimal time ill in the community. We assessed the impact of CT during the 2022 Sudan Virus Disease (SVD) outbreak in Uganda. METHODS: We collated anonymized data from the SVD case and contacts database to obtain and analyze data on CT performance indicators, comparing confirmed cases that were PKC and were not PKC (NPKC) before onset. We assessed the effect of being PKC on the number of people infected using Poisson regression. RESULTS: There were 3844 contacts of 142 confirmed cases (mean: 22 contacts/case). Forty-seven (33%) confirmed cases were PKC. PKCs had fewer median days from onset to isolation (4 vs 6; P<0.007) and laboratory confirmation (4 vs 7; P<0.001) than NPKC. Being a PKC vs NPKC reduced risk of transmitting infection by 84% (IRR=0.16, 95% CI 0.08-0.32). CONCLUSION: Contact identification was sub-optimal during the outbreak. However, CT reduced the time SVD cases spent in the community before isolation and the number of persons infected in Uganda. Approaches to improve contact tracing, especially contact listing, may improve control in future outbreaks.

Time to care and factors influencing appropriate Sudan virus disease care among case patients in Uganda, September to November 2022.

OBJECTIVES: Early isolation and care for Ebola disease patients at Ebola Treatment Units (ETU) curb outbreak spread. We evaluated time to ETU entry and associated factors during the 2022 Sudan virus disease (SVD) outbreak in Uganda. METHODS: We included persons with RT-PCR-confirmed SVD with onset September 20-November 30, 2022. We categorized days from symptom onset to ETU entry ("delays") as short (≤2), moderate (3-5), and long (≥6); the latter two were "delayed isolation." We categorized symptom onset timing as "earlier" or "later," using October 15 as a cut-off. We assessed demographics, symptom onset timing, and awareness of contact status as predictors for delayed isolation. We explored reasons for early vs late isolation using key informant interviews. RESULTS: Among 118 case-patients, 25 (21%) had short, 43 (36%) moderate, and 50 (43%) long delays. Seventy-five (64%) had symptom onset later in the outbreak. Earlier symptom onset increased risk of delayed isolation (crude risk ratio = 1.8, 95% confidence interval (1.2-2.8]). Awareness of contact status and SVD symptoms, and belief that early treatment-seeking was lifesaving facilitated early care-seeking. Patients with long delays reported fear of ETUs and lack of transport as contributors. CONCLUSION: Delayed isolation was common early in the outbreak. Strong contact tracing and community engagement could expedite presentation to ETUs.