Difference in HIV prevalence by testing venue: results from population level survey in Uganda.

Growing demand for use of Health Facility (HF) HIV testing data, in addition to other testing data to obtain district level HIV prevalence requires understanding the comparability of these various sources. We analysed the 2011 Uganda AIDS indicator survey data to assess: the proportion of people tested for HIV across Uganda by venue of testing; HIV prevalence ratio for those tested in a HF compared to those tested in community setting; [Katz, D., Baptista, J., Azen, S. P., & Pike, M. C. (1978). Obtaining confidence intervals for the risk ratio in cohort studies. International Biometric Society, 34(3), 469-474. https://doi.org/10.2307/2530610] and factors associated with HIV positivity in each subgroup. Of the 11,685 individuals, 8978 (77.1%) had ever tested for HIV in a HF. Fifty nine per cent tested in a HF in the 12 months preceding the survey (female: 5507, 72.7% versus male: 1413, 34.9%). HIV prevalence ratio was 1.8 times among those tested in a HF compared to those tested at community setting (10.9% [95% CI: 10.0-11.7] versus 6.2% [95% CI: 5.4-7.0]). Among HF testers, older age group, previously married and having no sexual partner was associated with significantly higher HIV prevalence. Using facility testing data for planning and decisions should take into consideration the elevated and varying HIV prevalence among individuals accessing HIV testing services at HFs as well as differences in their social-demographic characteristics.

Hybrid prevalence estimation: Method to improve intervention coverage estimations.

Delivering excellent health services requires accurate health information systems (HIS) data. Poor-quality data can lead to poor judgments and outcomes. Unlike probability surveys, which are representative of the population and carry accuracy estimates, HIS do not, but in many countries the HIS is the primary source of data used for administrative estimates. However, HIS are not structured to detect gaps in service coverage and leave communities exposed to unnecessary health risks. Here we propose a method to improve informatics by combining HIS and probability survey data to construct a hybrid estimator. This technique provides a more accurate estimator than either data source alone and facilitates informed decision-making. We use data from vitamin A and polio vaccination campaigns in children from Madagascar and Benin to demonstrate the effect. The hybrid estimator is a weighted average of two measurements and produces SEs and 95% confidence intervals (CIs) for the hybrid and HIS estimators. The estimates of coverage proportions using the combined data and the survey estimates differ by no more than 3%, while decreasing the SE by 1-6%; the administrative estimates from the HIS and combined data estimates are very different, with 3-25 times larger CI, questioning the value of administrative estimates. Estimators of unknown accuracy may lead to poorly formulated policies and wasted resources. The hybrid estimator technique can be applied to disease prevention services for which population coverages are measured. This methodology creates more accurate estimators, alongside measured HIS errors, to improve tracking the public's health.

Population-level effective coverage of adolescent weekly iron and folic acid supplementation is low in rural West Bengal, India.

OBJECTIVE: To assess the coverage of the adolescent weekly iron and folic acid supplementation (WIFS) programme in rural West Bengal, India. DESIGN: We conducted a population-based cross-sectional survey of intended WIFS programme beneficiaries (in-school adolescent girls and boys and out-of-school adolescent girls). SETTING: Birbhum Health and Demographic Surveillance System. PARTICIPANTS: A total of 4448 adolescents 10-19 years of age participated in the study. RESULTS: The percentage of adolescents who reported taking four WIFS tablets during the last month as intended by the national programme was 9·4 % among in-school girls, 7·1 % for in-school boys and 2·3 % for out-of-school girls. The low effective coverage was due to the combination of large deficits in WIFS provision and poor adherence. A large proportion of adolescents reported they were not provided any WIFS tablets in the last month: 61·7 % of in-school girls, 73·3 % of in-school boys and 97·1 % of out-of-school girls. In terms of adherence, only 41·6 % of in-school girls, 38·1 % of in-school boys and 47·4 % of out-of-school girls reported that they consumed all WIFS tablets they received. Counselling from teachers, administrators and school staff was the primary reason adolescents reported taking WIFS tablets, whereas the major reasons for non-adherence were lack of perceived benefit, peer suggestion not to take WIFS and a reported history of side effects. CONCLUSIONS: The effective coverage of the WIFS programme for in-school adolescents and out-of-school adolescent girls is low in rural Birbhum. Integrated supply- and demand-side strategies appear to be necessary to increase the effective coverage and potential benefits of the WIFS programme.

Combining national survey with facility-based HIV testing data to obtain more accurate estimate of HIV prevalence in districts in Uganda.

BACKGROUND: National or regional population-based HIV prevalence surveys have small sample sizes at district or sub-district levels; this leads to wide confidence intervals when estimating HIV prevalence at district level for programme monitoring and decision making. Health facility programme data, collected during service delivery is widely available, but since people self-select for HIV testing, HIV prevalence estimates based on it, is subject to selection bias. We present a statistical annealing technique, Hybrid Prevalence Estimation (HPE), that combines a small population-based survey sample with a facility-based sample to generate district level HIV prevalence estimates with associated confidence intervals. METHODS: We apply the HPE methodology to combine the 2011 Uganda AIDS indicator survey with the 2011 health facility HIV testing data to obtain HIV prevalence estimates for districts in Uganda. Multilevel logistic regression was used to obtain the propensity of testing for HIV in a health facility, and the propensity to test was used to combine the population survey and health facility HIV testing data to obtain the HPEs. We assessed comparability of the HPEs and survey-based estimates using Bland Altman analysis. RESULTS: The estimates ranged from 0.012 to 0.178 and had narrower confidence intervals compared to survey-based estimates. The average difference between HPEs and population survey estimates was 0.00 (95% CI: - 0.04, 0.04). The HPE standard errors were 28.9% (95% CI: 23.4-34.4) reduced, compared to survey-based standard errors. Overall reduction in HPE standard errors compared survey-based standard errors ranged from 5.4 to 95%. CONCLUSIONS: Facility data can be combined with population survey data to obtain more accurate HIV prevalence estimates for geographical areas with small population survey sample sizes. We recommend use of the methodology by district level managers to obtain more accurate HIV prevalence estimates to guide decision making without incurring additional data collection costs.

Using lot quality assurance sampling to assess access to water, sanitation and hygiene services in a refugee camp setting in South Sudan: a feasibility study.

BACKGROUND: Humanitarian agencies working in refugee camp settings require rapid assessment methods to measure the needs of the populations they serve. Due to the high level of dependency of refugees, agencies need to carry out these assessments. Lot Quality Assurance Sampling (LQAS) is a method commonly used in development settings to assess populations living in a project catchment area to identify their greatest needs. LQAS could be well suited to serve the needs of refugee populations, but it has rarely been used in humanitarian settings. We adapted and implemented an LQAS survey design in Batil refugee camp, South Sudan in May 2013 to measure the added value of using it for sub-camp level assessment. METHODS: Using pre-existing divisions within the camp, we divided the Batil catchment area into six contiguous segments, called 'supervision areas' (SA). Six teams of two data collectors randomly selected 19 respondents in each SA, who they interviewed to collect information on water, sanitation, hygiene, and diarrhoea prevalence. These findings were aggregated into a stratified random sample of 114 respondents, and the results were analysed to produce a coverage estimate with 95% confidence interval for the camp and to prioritize SAs within the camp. RESULTS: The survey provided coverage estimates on WASH indicators as well as evidence that areas of the camp closer to the main road, to clinics and to the market were better served than areas at the periphery of the camp. This assumption did not hold for all services, however, as sanitation services were uniformly high regardless of location. While it was necessary to adapt the standard LQAS protocol used in low-resource communities, the LQAS model proved to be feasible in a refugee camp setting, and program managers found the results useful at both the catchment area and SA level. CONCLUSIONS: This study, one of the few adaptations of LQAS for a camp setting, shows that it is a feasible method for regular monitoring, with the added value of enabling camp managers to identify and advocate for the least served areas within the camp. Feedback on the results from stakeholders was overwhelmingly positive.

Assessing delivery practices of mothers over time and over space in Uganda, 2003-2012.

BACKGROUND: It is well known that safe delivery in a health facility reduces the risks of maternal and infant mortality resulting from perinatal complications. What is less understood are the factors associated with safe delivery practices. We investigate factors influencing health facility delivery practices while adjusting for multiple other factors simultaneously, spatial heterogeneity, and trends over time. METHODS: We fitted a logistic regression model to Lot Quality Assurance Sampling (LQAS) data from Uganda in a framework that considered individual-level covariates, geographical features, and variations over five time points. We accounted for all two-covariate interactions and all three-covariate interactions for which two of the covariates already had a significant interaction, were able to quantify uncertainty in outputs using computationally intensive cluster bootstrap methods, and displayed outputs using a geographical information system. Finally, we investigated what information could be predicted about districts at future time-points, before the next LQAS survey is carried out. To do this, we applied the model to project a confidence interval for the district level coverage of health facility delivery at future time points, by using the lower and upper end values of known demographics to construct a confidence range for the prediction and define priority groups. RESULTS: We show that ease of access, maternal age and education are strongly associated with delivery in a health facility; after accounting for this, there remains a significant trend towards greater uptake over time. We use this model together with known demographics to formulate a nascent early warning system that identifies candidate districts expected to have low prevalence of facility-based delivery in the immediate future. CONCLUSIONS: Our results support the hypothesis that increased development, particularly related to education and access to health facilities, will act to increase facility-based deliveries, a factor associated with reducing perinatal associated mortality. We provide a statistical method for using inexpensive and routinely collected monitoring and evaluation data to answer complex epidemiology and public health questions in a resource-poor setting. We produced a model based on this data that explained the spatial distribution of facility-based delivery in Uganda. Finally, we used this model to make a prediction about the future priority of districts that was validated by monitoring and evaluation data collected in the next year.

Associations with HIV testing in Uganda: an analysis of the Lot Quality Assurance Sampling database 2003-2012.

Beginning in 2003, Uganda used Lot Quality Assurance Sampling (LQAS) to assist district managers collect and use data to improve their human immunodeficiency virus (HIV)/AIDS program. Uganda's LQAS-database (2003-2012) covers up to 73 of 112 districts. Our multidistrict analysis of the LQAS data-set at 2003-2004 and 2012 examined gender variation among adults who ever tested for HIV over time, and attributes associated with testing. Conditional logistic regression matched men and women by community with seven model effect variables. HIV testing prevalence rose from 14% (men) and 12% (women) in 2003-2004 to 62% (men) and 80% (women) in 2012. In 2003-2004, knowing the benefits of testing (Odds Ratio [OR] = 6.09, 95% CI = 3.01-12.35), knowing where to get tested (OR = 2.83, 95% CI = 1.44-5.56), and secondary education (OR = 3.04, 95% CI = 1.19-7.77) were significantly associated with HIV testing. By 2012, knowing the benefits of testing (OR = 3.63, 95% CI = 2.25-5.83), where to get tested (OR = 5.15, 95% CI = 3.26-8.14), primary education (OR = 2.01, 95% CI = 1.39-2.91), being female (OR = 3.03, 95% CI = 2.53-3.62), and being married (OR = 1.81, 95% CI = 1.17-2.8) were significantly associated with HIV testing. HIV testing prevalence in Uganda has increased dramatically, more for women than men. Our results concurred with other authors that education, knowledge of HIV, and marriage (women only) are associated with testing for HIV and suggest that couples testing is more prevalent than other authors.

Can local staff reliably assess their own programs? A confirmatory test-retest study of Lot Quality Assurance Sampling data collectors in Uganda.

BACKGROUND: Data collection techniques that routinely provide health system information at the local level are in demand and needed. LQAS is intended for use by local health teams to collect data at the district and sub-district levels. Our question is whether local health staff produce biased results as they are responsible for implementing the programs they also assess. METHODS: This test-retest study replicates on a larger scale an earlier LQAS reliability assessment in Uganda. We conducted in two districts an LQAS survey using 15 local health staff as data collectors. A week later, the data collectors swapped districts, where they acted as disinterested non-local data collectors, repeating the LQAS survey with the same respondents. We analysed the resulting two data sets for agreement using Cohens' Kappa. RESULTS: The average Kappa score for the knowledge indicators was k = 0.43 (SD = 0.16) and for practice indicators k = 0.63 (SD = 0.17). These scores show moderate agreement for knowledge indicators and substantial agreement for practice indicators. Analyses confirm that respondents were more knowledgeable on retest; no evidence of bias was found for practice indicators. CONCLUSION: The findings of this study are remarkably similar to those produced in the first reliability study. There is no evidence that using local healthcare staff to collect LQAS data biases data collection in an LQAS study. The bias observed in the knowledge indicators was most likely due to a 'practice effect', whereby respondents increased their knowledge as a result of completing the first survey; no corresponding effect was seen in the practice indicators.

Comparing two survey methods of measuring health-related indicators: Lot Quality Assurance Sampling and Demographic Health Surveys.

OBJECTIVES: Two common methods used to measure indicators for health programme monitoring and evaluation are the demographic and health surveys (DHS) and lot quality assurance sampling (LQAS); each one has different strengths. We report on both methods when utilised in comparable situations. METHODS: We compared 24 indicators in south-west Uganda, where data for prevalence estimations were collected independently for the two methods in 2011 (LQAS: n = 8876; DHS: n = 1200). Data were stratified (e.g. gender and age) resulting in 37 comparisons. We used a two-sample two-sided Z-test of proportions to compare both methods. RESULTS: The average difference between LQAS and DHS for 37 estimates was 0.062 (SD = 0.093; median = 0.039). The average difference among the 21 failures to reject equality of proportions was 0.010 (SD = 0.041; median = 0.009); among the 16 rejections, it was 0.130 (SD = 0.010, median = 0.118). Seven of the 16 rejections exhibited absolute differences of <0.10, which are clinically (or managerially) not significant; 5 had differences >0.10 and <0.20 (mean = 0.137, SD = 0.031) and four differences were >0.20 (mean = 0.261, SD = 0.083). CONCLUSION: There is 75.7% agreement across the two surveys. Both methods yield regional results, but only LQAS provides information at less granular levels (e.g. the district level) where managerial action is taken. The cost advantage and localisation make LQAS feasible to conduct more frequently, and provides the possibility for real-time health outcomes monitoring.

Finding the gap: revealing local disparities in coverage of maternal, newborn and child health services in South Sudan using lot quality assurance sampling.

OBJECTIVES: We adapted a rapid monitoring method to South Sudan, a new nation with one of the world's highest maternal and child mortality rates, aiming to assess coverage of maternal, neonatal and child health (MNCH) services at the time of independence, and introducing a monitoring and evaluation system (M&E) for equity-sensitive tracking of progress related to Millennium Development Goals (MDG) 4 and 5 at national, state and county levels to detect local variability. METHODS: We conducted a national cross-sectional household survey among women from six client populations in all, but six of South Sudan's 79 counties. We used lot quality assurance sampling (LQAS) to measure coverage with diverse MNCH indicators to obtain information for national-, state- and county-level health system management decision-making. RESULTS: National coverage of MNCH services was low for all maternal and neonatal care, child immunisation, and child care indicators. However, results varied across states and counties. Central Equatoria State (CES), where the capital is located, showed the highest coverage for most indicators (e.g. ≥4 antenatal care visits range: 4.5% in Jonglei to 40.1% in CES). Urban counties often outperformed rural ones. CONCLUSIONS: This adaptation of LQAS to South Sudan demonstrates how it can be used in the future as an M&E system to track progress of MDGs at national, state and county levels to detect local disparities. Overall, our data reveal a desperate need for improving MNCH service coverage in all states.

Estimation after classification using lot quality assurance sampling: corrections for curtailed sampling with application to evaluating polio vaccination campaigns.

OBJECTIVES: To assess the bias incurred when curtailment of Lot Quality Assurance Sampling (LQAS) is ignored, to present unbiased estimators, to consider the impact of cluster sampling by simulation and to apply our method to published polio immunization data from Nigeria. METHODS: We present estimators of coverage when using two kinds of curtailed LQAS strategies: semicurtailed and curtailed. We study the proposed estimators with independent and clustered data using three field-tested LQAS designs for assessing polio vaccination coverage, with samples of size 60 and decision rules of 9, 21 and 33, and compare them to biased maximum likelihood estimators. Lastly, we present estimates of polio vaccination coverage from previously published data in 20 local government authorities (LGAs) from five Nigerian states. RESULTS: Simulations illustrate substantial bias if one ignores the curtailed sampling design. Proposed estimators show no bias. Clustering does not affect the bias of these estimators. Across simulations, standard errors show signs of inflation as clustering increases. Neither sampling strategy nor LQAS design influences estimates of polio vaccination coverage in 20 Nigerian LGAs. When coverage is low, semicurtailed LQAS strategies considerably reduces the sample size required to make a decision. Curtailed LQAS designs further reduce the sample size when coverage is high. CONCLUSIONS: Results presented dispel the misconception that curtailed LQAS data are unsuitable for estimation. These findings augment the utility of LQAS as a tool for monitoring vaccination efforts by demonstrating that unbiased estimation using curtailed designs is not only possible but these designs also reduce the sample size.

Assessing the quality of care in a new nation: South Sudan's first national health facility assessment.

OBJECTIVES: We adapted a rapid quality of care monitoring method to a fragile state with two aims: to assess the delivery of child health services in South Sudan at the time of independence and to strengthen local capacity to perform regular rapid health facility assessments. METHODS: Using a two-stage lot quality assurance sampling (LQAS) design, we conducted a national cross-sectional survey among 156 randomly selected health facilities in 10 states. In each of these facilities, we obtained information on a range of access, input, process and performance indicators during structured interviews and observations. RESULTS: Quality of care was poor with all states failing to achieve the 80% target for 14 of 19 indicators. For example, only 12% of facilities were classified as acceptable for their adequate utilisation by the population for sick-child consultations, 16% for staffing, 3% for having infection control supplies available and 0% for having all child care guidelines. Health worker performance was categorised as acceptable in only 6% of cases related to sick-child assessments, 38% related to medical treatment for the given diagnosis and 33% related to patient counselling on how to administer the prescribed drugs. Best performance was recorded for availability of in-service training and supervision, for seven and ten states, respectively. CONCLUSIONS: Despite ongoing instability, the Ministry of Health developed capacity to use LQAS for measuring quality of care nationally and state-by-state, which will support efficient and equitable resource allocation. Overall, our data revealed a desperate need for improving the quality of care in all states.

Improving malaria treatment and prevention in India by aiding district managers to manage their programmes with local information: a trial assessing the impact of Lot Quality Assurance Sampling on programme outcomes.

OBJECTIVES: This paper reports the first trial of Lot Quality Assurance Sampling (LQAS) assessing associations between access to LQAS data and subsequent improvements in district programming. This trial concerns India's approach to addressing an increase in malaria-attributable deaths by training community health workers to diagnose, treat and prevent malaria, while using LQAS to monitor sub-district performance and make programme improvements. METHODS: The Ministry of Health introduced LQAS into four matched high malaria burden districts (Annual Parasite Incidence >5) (N > 5 million). In each sub-district, we sampled four populations in three 6-monthly surveys: households, children <5 years, people with fever in the last 2 weeks and community health workers. In three districts, trained local staff collected, analysed and used data for programme management; in one control district, non-local staff collected data and did not disseminate results. For eight indicators, we calculated the change in proportion from survey one to three and used a Difference-in-Differences test to compare the relative change between intervention and control districts. RESULTS: Coverage increased from survey one to three for 24 of 32 comparisons. Difference-in-Differences tests revealed that intervention districts exhibited significantly greater change in four of six vertical strategies (insecticide treated bed-nets and indoor residual spraying), one of six treatment-seeking behaviours and four of 12 health worker capacity indicators. The control district displayed greater improvement than two intervention districts for one health worker capacity indicator. One district with poor management did not improve. CONCLUSIONS: In this study, LQAS results appeared to support district managers to increase coverage in underperforming areas, especially for vertical strategies in the presence of diligent managers.

Assessing attitudes towards violence against women and girls, their determinants and health-seeking behaviour among women and men in South Sudan: a cross-sectional national survey.

OBJECTIVE: Investigating attitudes accepting two categories of violence against women and girls (VAWG) (intimate partner violence-IPV-and other expressions of VAWG) and their association with seven demographic/social determinants and health-seeking behaviours in South Sudan. DESIGN: Cross-sectional study using data from the South Sudan National Household Survey 2020. SETTING: South Sudan. PARTICIPANTS AND METHODS: 1741 South Sudanese women and 1739 men aged 15-49 years; data captured between November 2020 and February 2021 and analysed using binary logistic regression. RESULTS: People with secondary or higher education displayed attitudes rejecting acceptance of IPV (OR 0.631, 95% CI 0.508 to 0.783). Women and men living in states with more numerous internally displaced people (IDP) or political/military violence had attitudes accepting IPV more than residents of less violence-affected regions (OR 1.853, 95% CI 1.587 to 2.164). Women had a higher odd of having attitudes accepting IPV than men (OR 1.195, 95% CI 1.014 to 1.409). People knowing where to receive gender-based violence healthcare and psychological support (OR 0.703, 95% CI 0.596 to 0.830) and with primary (OR 0.613, 95% CI 0.515 to 0.729), secondary or higher education (OR 0.596, 95% CI 0.481 to 0.740) displayed attitudes rejecting acceptance of other expressions of VAWG. People residing in states with proportionately more IDP and who accepted IPV were more likely to have attitudes accepting other expressions of VAWG (OR 1.699, 95% CI 1.459 to 1.978; OR 3.195, 95% CI 2.703 to 3.775, respectively). CONCLUSION: Attitudes towards accepting VAWG in South Sudan are associated with women's and men's education, gender, residence and knowledge about health-seeking behaviour. Prioritising women's empowerment and gender transformative programming in the most conflict-affected areas where rates of VAWG are higher should be prioritised along with increasing girls' access to education. A less feasible strategy to decrease gender inequalities is reducing insecurity, military conflict, and displacement, and increasing economic stability.

Innovative approach to improve information accuracy in a two-district cross-sectional study in Bihar, India.

OBJECTIVE: Combine Health Management Information Systems (HMIS) and probability survey data using the statistical annealing technique (AT) to produce more accurate health coverage estimates than either source of data and a measure of HMIS data error. SETTING: This study is set in Bihar, the fifth poorest state in India, where half the population lives below the poverty line. An important source of data, used by health professionals for programme decision making, is routine health facility or HMIS data. Its quality is sometimes poor or unknown, and has no measure of its uncertainty. Using AT, we combine district-level HMIS and probability survey data (n=475) for the first time for 10 indicators assessing antenatal care, institutional delivery and neonatal care from 11 blocks of Aurangabad and 14 blocks of Gopalganj districts (N=6 253 965) in Bihar state, India. PARTICIPANTS: Both districts are rural. Bihar is 82.7% Hindu and 16.9% Islamic. PRIMARY OUTCOME MEASURES: Survey prevalence measures for 10 indicators, corresponding prevalences using HMIS data, combined prevalences calculated with AT and SEs for each type of data. RESULTS: The combined and survey estimates differ by <0.10. The combined and HMIS estimates differ by up to 84.2%, with the HMIS having 1.4-32.3 times larger error. Of 20 HMIS versus survey coverage estimate comparisons across the two districts only five differed by <0.10. Of 250 subdistrict-level comparisons of HMIS versus combined estimates, only 36.4% of the HMIS estimates are within the 95% CI of the combined estimate. CONCLUSIONS: Our statistical innovation increases the accuracy of information available for local health system decision making, allows evaluation of indicator accuracy and increases the accuracy of HMIS estimates. The combined estimates with a measure of error better informs health system professionals about their risks when using HMIS estimates, so they can reduce waste by making better decisions. Our results show that AT is an effective method ready for additional international assessment while also being used to provide affordable information to improve health services.

How to estimate health service coverage in 58 districts of Benin with no survey data: Using hybrid estimation to fill the gaps.

The global movement to use routine information for managing health systems to achieve the Sustainable Development Goals, relies on administrative data which have inherent biases when used to estimate coverage with health services. Health policies and interventions planned with incorrect information can have detrimental impacts on communities. Statistical inferences using administrative data can be improved when they are combined with random probability survey data. Sometimes, survey data are only available for some districts. We present new methods for extending combined estimation techniques to all districts by combining additional data sources. Our study uses data from a probability survey (n = 1786) conducted during 2015 in 19 of Benin's 77 communes and administrative count data from all of them for a national immunization day (n = 2,792,803). Communes are equivalent to districts. We extend combined-data estimation from 19 to 77 communes by estimating denominators using the survey data and then building a statistical model using population estimates from different sources to estimate denominators in adjacent districts. By dividing administrative numerators by the model-estimated denominators we obtain extrapolated hybrid prevalence estimates. Framing the problem in the Bayesian paradigm guarantees estimated prevalence rates fall within the appropriate ranges and conveniently incorporates a sensitivity analysis. Our new methodology, estimated Benin's polio vaccination rates for 77 communes. We leveraged probability survey data from 19 communes to formulate estimates for the 58 communes with administrative data alone; polio vaccination coverage estimates in the 58 communes decreased to ranges consistent with those from the probability surveys (87%, standard deviation = 0.09) and more credible than the administrative estimates. Combining probability survey and administrative data can be extended beyond the districts in which both are collected to estimate coverage in an entire catchment area. These more accurate results will better inform health policy-making and intervention planning to reduce waste and improve health in communities.

Does interviewer gender influence a mother's response to household surveys about maternal and child health in traditional settings? A qualitative study in Bihar, India.

BACKGROUND: Two probability surveys, conducted in the same districts of Bihar, India (Aurangabad and Gopalganj) at approximately the same time in 2016 using identical questionnaires and similar survey methods, produced significantly different responses for 37.2% (58/156) of the indicator comparisons. Interviewers for one survey were men while for the other they were women. Respondents were mothers of children aged 0-59 months living in a traditional rural setting. We examined the influence of interviewer gender on mothers' survey responses and their implications for interpreting survey results. METHODS: We used qualitative methods including 10 focus group discussions (FGDs) and 33 in-depth interviews (IDIs) in the same locations as the 2016 surveys. FGD participants were purposefully selected mothers with children 0-59 months, husbands and other in-law family members. IDIs were carried out with frontline health-workers, enumerators and supervisors from the two previous household surveys. RESULTS: Findings revealed a preference for female interviewers for household surveys in study districts as they facilitated access to mothers and reduced their discomfort as survey participants. However, this gender preference was related to the survey question. Regardless of age, caste and educational level, most mothers were not permitted to communicate with men (aside from husbands) about female-specific health topics, including birth preparedness, delivery, menstrual cycles, contraception, breastfeeding, sexual behaviour, sexually transmitted disease, and domestic violence. Mothers in higher castes perceived these social restrictions more acutely than mothers in lower castes. There was no systematic direction of the resulting error. Mothers were willing to discuss child health issues with interviewers of either gender. CONCLUSIONS: Interviewer gender is an important consideration when designing survey protocols for maternal and reproductive health studies and when selecting and training enumerators. Female interviewers are optimal for traditional settings in Bihar as they are more likely to obtain accurate data on sensitive topics and reduce the potential for non-sampling error due to their reduced social distance with maternal respondents.

Model-based small area estimation methods and precise district-level HIV prevalence estimates in Uganda.

BACKGROUND: Model-based small area estimation methods can help generate parameter estimates at the district level, where planned population survey sample sizes are not large enough to support direct estimates of HIV prevalence with adequate precision. We computed district-level HIV prevalence estimates and their 95% confidence intervals for districts in Uganda. METHODS: Our analysis used direct survey and model-based estimation methods, including Fay-Herriot (area-level) and Battese-Harter-Fuller (unit-level) small area models. We used regression analysis to assess for consistency in estimating HIV prevalence. We use a ratio analysis of the mean square error and the coefficient of variation of the estimates to evaluate precision. The models were applied to Uganda Population-Based HIV Impact Assessment 2016/2017 data with auxiliary information from the 2016 Lot Quality Assurance Sampling survey and antenatal care data from district health information system datasets for unit-level and area-level models, respectively. RESULTS: Estimates from the model-based and the direct survey methods were similar. However, direct survey estimates were unstable compared with the model-based estimates. Area-level model estimates were more stable than unit-level model estimates. The correlation between unit-level and direct survey estimates was (ß1 = 0.66, r2 = 0.862), and correlation between area-level model and direct survey estimates was (ß1 = 0.44, r2 = 0.698). The error associated with the estimates decreased by 37.5% and 33.1% for the unit-level and area-level models, respectively, compared to the direct survey estimates. CONCLUSIONS: Although the unit-level model estimates were less precise than the area-level model estimates, they were highly correlated with the direct survey estimates and had less standard error associated with estimates than the area-level model. Unit-level models provide more accurate and reliable data to support local decision-making when unit-level auxiliary information is available.

Measuring health system resilience in a highly fragile nation during protracted conflict: South Sudan 2011-15.

Health systems resilience (HSR) is defined as the ability of a health system to continue providing normal services in response to a crisis, making it a critical concept for analysis of health systems in fragile and conflict-affected settings (FCAS). However, no consensus for this definition exists and even less about how to measure HSR. We examine three current HSR definitions (maintaining function, improving function and achieving health system targets) using real-time data from South Sudan to develop a data-driven understanding of resilience. We used 14 maternal, newborn and child health (MNCH) coverage indicators from household surveys in South Sudan collected at independence (2011) and following 2 years of protracted conflict (2015), to construct a resilience index (RI) for 9 of the former 10 states and nationally. We also assessed health system stress using conflict-related indicators and developed a stress index. We cross tabulated the two indices to assess the relationship of resilience and stress. For maintaining function for 80% of MNCH indicators, seven state health systems were resilient, compared with improving function for 50% of the indicators (two states were resilient). Achieving the health system national target of 50% coverage in half of the MNCH indicators displayed no resilience. MNCH coverage levels were low, with state averages ranging between 15% and 44%. Central Equatoria State displayed high resilience and high system stress. Lakes and Northern Bahr el Ghazal displayed high resilience and low stress. Jonglei and Upper Nile States had low resilience and high stress. This study is the first to investigate HSR definitions using a resilience metric and to simultaneously measure health system stress in FCAS. Improving function is the HSR definition detecting the greatest variation in the RI. HSR and health system stress are not consistently negatively associated. HSR is highly complex warranting more in-depth analyses in FCAS.

Are health workers reduced to being drug dispensers of antiretroviral treatment? A randomized cross-sectional assessment of the quality of health care for HIV patients in northern Uganda.

High quality of care (QoC) for antiretroviral treatment (ART) is essential to prevent treatment failure. Uganda, as many sub-Saharan African countries, increased access to ART by decentralizing provision to districts. However, little is known whether this rapid scale-up maintained high-quality clinical services. We assess the quality of ART in the Acholi and Lango sub-regions of northern Uganda to identify whether the technical quality of critical ART sub-system needs improvement. We conducted a randomized cross-sectional survey among health facilities (HF) in Acholi (n = 11) and Lango (n = 10). Applying lot quality assurance sampling principles with a rapid health facility assessment tool, we assessed ART services vis-à-vis national treatment guidelines using 37 indicators. We interviewed health workers (n = 21) using structured questionnaires, directly observed clinical consultations (n = 126) and assessed HF infrastructure, human resources, medical supplies and patient records in each health facility (n = 21). The district QoC performance standard was 80% of HF had to comply with each guideline. Neither sub-region complied with treatment guidelines. No HF displayed adequate: patient monitoring, physical examination, training, supervision and regular monitoring of patients' immunology. The full range of first and second line antiretroviral (ARV) medication was not available in Acholi while Lango had sufficient stocks. Clinicians dispensed available ARVs without benefit of physical examination or immunological monitoring. Patients reported compliance with drug use (>80%). Patients' knowledge of preventing HIV/AIDS transmission concentrated on condom use; otherwise it was poor. The poor ART QoC in northern Uganda raises major questions about ART quality although ARVs were dispensed. Poor clinical care renders patients' reports of treatment compliance as insufficient evidence that it takes place. Further studies need to test patients' immunological status and QoC in more regions of Uganda and elsewhere in sub-Saharan Africa to identify topical and geographical areas which are priorities for improving HIV care.