Operational feasibility of lot quality assurance sampling (LQAS) as a tool in routine process monitoring of filariasis control programmes

Estimation of bed net coverage indicators in Tanzania using mobile phone surveys: a comparison of sampling approaches

BACKGROUND

Threats to maintaining high population access with effective bed nets persist due to errors in quantification, bed net wear and tear, and inefficiencies in distribution activities. Monitoring bed net coverage is therefore critical, but usually occurs every 2-3 years through expensive, large-scale household surveys. Mobile phone-based survey methodologies are emerging as an alternative to household surveys and can provide rapid estimates of coverage, however, little research on varied sampling approaches has been conducted in sub-Saharan Africa.

METHODS

A nationally and regionally representative cross-sectional mobile phone survey was conducted in early 2021 in Tanzania with focus on bed net ownership and access. Half the target sample was contacted through a random digit dial methodology (n = 3500) and the remaining half was reached through a voluntary opt-in respondent pool (n = 3500). Both sampling approaches used an interactive voice response survey. Standard RBM-MERG bed net indicators and AAPOR call metrics were calculated. In addition, the results of the two sampling approaches were compared.

RESULTS

Population access (i.e., the percent of the population that could sleep under a bed net, assuming one bed net per two people) varied from a regionally adjusted low of 48.1% (Katavi) to a high of 65.5% (Dodoma). The adjusted percent of households that had a least one bed net ranged from 54.8% (Pemba) to 75.5% (Dodoma); the adjusted percent of households with at least one bed net per 2 de facto household population ranged from 35.9% (Manyara) to 55.7% (Dodoma). The estimates produced by both sampling approaches were generally similar, differing by only a few percentage points. An analysis of differences between estimates generated from the two sampling approaches showed minimal bias when considering variation across the indicator for households with at least one bed net per two de facto household population.

CONCLUSION

The results generated by this survey show that overall bed net access in the country appears to be lower than target thresholds. The results suggest that bed net distribution is needed in large sections of the country to ensure that coverage levels remain high enough to sustain protection against malaria for the population.

Lot quality assurance sampling to assess coverage and compliance following mass drug administration to eliminate lymphatic filariasis in Fiji: A methodological approach

Background

Assessing the quality of mass drug administration (MDA) rounds is a key component of lymphatic filariasis (LF) elimination programs. Routine collection of administrative coverage is unreliable, especially when pockets with low program coverage exist. To address this gap, we used lot quality assurance sampling (LQAS) following the 10th annual LF-MDA round in Fiji to explore whether there was any area in which target coverage was not reached. We also assessed the level of drug compliance and satisfaction with the LF-MDA implementation strategy.

Methodology/principal findings

We conducted a cross-sectional household survey in 3 divisions of Fiji. For LQAS, we defined 19 lots in 7 medical areas of the Suva sub-division and another 12 sub-divisions in the Central, Northern, and Eastern Divisions. A sample of 16 randomly selected household members was taken un each lot. We defined our decision rule as follows: if more than 1 person in a given lot did not swallow the medication, coverage was considered inadequate, i.e. less than 80%.

Of the 7 lots in Suva sub-division and 12 lots in the 3 divisions, five and two lots, respectively, were identified as having inadequate coverage. The overall program coverage estimated from 304 samples was 92%, which was higher than the reported administrative coverage of 82%. About 98% of interviewees were offered the medication and 96% swallowed it. Non-participation arose from insufficient information on how to obtain the drugs. At least 92% were satisfied with the LF-MDA implementation strategy.

Conclusions

Areas of low program coverage with results discordant with the reported administrative coverage existed in both urban and rural settings. Drug compliance and satisfaction were high, even after repeated rounds. We recommend increasing efforts to deliver the service in those areas with inadequate program coverage, as well as conducting timely coverage assessment through LQAS for corrective action.

Improved assessment of mass drug administration and health district management performance to eliminate lymphatic filariasis

Lymphatic filariasis (LF) elimination as a public health problem requires the interruption of transmission by administration of preventive mass drug administration (MDA) to the eligible population living in endemic districts. Suboptimal MDA coverage leads to persistent parasite transmission with consequential infection, disease and disability, and the need for continuing MDA rounds, requiring considerable investment. Routine coverage reports must be verified in each MDA implementation unit (IU) due to incorrect denominators and numerators used to calculate coverage estimates with administrative data. IU are usually the health districts. Coverage is verified so IU teams can evaluate their outreach and take appropriate action to improve performance. Mozambique and the Democratic Republic of Congo (DRC) have conducted MDA campaigns for LF since 2009 and 2014, respectively. To verify district reports and assess the declared achievement using administrative data of the minimum 80% coverage of eligible people (or 65% of the total population), both countries conducted rapid probability surveys using Lot Quality Assurance Sampling (LQAS)(n = 1102) in 2015 and 2016 in 58 IU in 49 districts. The surveys identified IU with suboptimal coverage, reasons residents did not take the medication, place where the medication was received, information sources, and knowledge about diseases prevented by the MDA. LQAS identified four inadequately covered IU triggering district team performance reviews with provincial and national teams and district retreatment. Provincial estimates using probability samples (weighted by populations sizes) were 10 and 17 percentage points lower than reported coverage in DRC and Mozambique. The surveys identified: absence from home during annual MDA rounds as the main reason for low performance and provided valuable information about pre-campaign and campaign activities resulting in improved strategies and continued progress towards elimination of LF and co-endemic Neglected Tropical Diseases.

Global elimination of lymphatic filariasis (LF) is achieved through treatment of at-risk populations with annual or bi-annual mass drug administration campaigns. In Africa campaigns need to be completed in 32 countries with 343 million people at risk. The World Health Organisation recommends verification of the campaign’s administrative records using household cluster surveys at least once every 5-years. However, cluster surveys are expensive and usually completed in a few districts only or at sub-national levels. Together with National Programmes in Mozambique and the Democratic Republic of the Congo (DRC) we adapted Lot Quality Assurance Sampling (LQAS) methods to verify campaign coverage because it is relatively inexpensive in comparison to other survey approaches, provides information and facilitates actions at the Implementation Unit (IU) level, which is usually the district. LQAS signals IU whose performance is likely to need improvement because MDA coverage is below the coverage target. Our results show consistently that administrative records over-estimated campaign coverage and did not detect implementation and coverage problems due to errors in numerators and denominators, incorrect reporting, and/or incorrect aggregation of tally sheets. The LQAS verification approach prompted immediate action to remedy coverage shortfalls averting persistent LF transmission and disease, and the costs associated with failed campaigns. Our study demonstrates that a rapid probability sample to verify coverage provides district teams with information after each campaign which can be used for action, and that one coverage survey every 5-years is insufficient for infectious disease elimination in highly endemic settings where achievement of optimal coverage is essential. It also demonstrates that LQAS provides a decentralised assessment, sensitive for detecting and ameliorating programme bottlenecks and can be used to verify MDA in other countries.

Optimal design of multiple-objective Lot Quality Assurance Sampling (LQAS) plans

Lot Quality Assurance Sampling (LQAS) plans are widely used for health monitoring purposes. We propose a systematic approach to design multiple-objective LQAS plans that meet user-specified type 1 and 2 error rates and targets for selected diagnostic accuracy metrics. These metrics may include sensitivity, specificity, positive predictive value, and negative predictive value in high or low anticipated prevalence rate populations. We use Mixed Integer Nonlinear Programming (MINLP) tools to implement our design methodology. Our approach is flexible in that it can directly generate classic LQAS plans that control error rates only and find optimal LQAS plans that meet multiple objectives in terms of diagnostic metrics. We give examples, compare results with the classic LQAS and provide an application using a malaria outcome indicator survey in Mozambique.

Monitoring maternal, newborn, and child health interventions using lot quality assurance sampling in Sokoto State of northern Nigeria

BACKGROUND

Maternal mortality ratio and infant mortality rate are as high as 1,576 per 100,000 live births and 78 per 1,000 live births, respectively, in Nigeria's northwestern region, where Sokoto State is located. Using applicable monitoring indicators for tracking progress in the UN/WHO framework on continuum of maternal, newborn, and child health care, this study evaluated the progress of Sokoto toward achieving the Millennium Development Goals (MDGs) 4 and 5 by December 2015. The changes in outcomes in 2012-2013 associated with maternal and child health interventions were assessed.

DESIGN

We used baseline and follow-up lot quality assurance sampling (LQAS) data obtained in 2012 and 2013, respectively. In each of the surveys, data were obtained from 437 households sampled from 19 LQAS locations in each of the 23 local government areas (LGAs). The composite state-level coverage estimates of the respective indicators were aggregated from estimated LGA coverage estimates.

RESULTS

None of the nine indicators associated with the continuum of maternal, neonatal, and child care satisfied the recommended 90% coverage target for achieving MDGs 4 and 5. Similarly, the average state coverage estimates were lower than national coverage estimates. Marginal improvements in coverage were obtained in the demand for family planning satisfied, antenatal care visits, postnatal care for mothers, and exclusive breast-feeding. Antibiotic treatment for acute pneumonia increased significantly by 12.8 percentage points. The majority of the LGAs were classifiable as low-performing, high-priority areas for intensified program intervention.

CONCLUSIONS

Despite the limited time left in the countdown to December 2015, Sokoto State, Nigeria, is not on track to achieving the MDG 90% coverage of indicators tied to the continuum of maternal and child care, to reduce maternal and childhood mortality by a third by 2015. Targeted health system investments at the primary care level remain a priority, for intensive program scale-up to accelerate impact.

Secondary mapping of lymphatic filariasis in Haiti-definition of transmission foci in low-prevalence settings

To eliminate Lymphatic filariasis (LF) as a public health problem, the World Health Organization (WHO) recommends that any area with infection prevalence greater than or equal to 1% (denoted by presence of microfilaremia or antigenemia) should receive mass drug administration (MDA) of antifilarial drugs for at least five consecutive rounds. Areas of low-antigen prevalence (< 1%) are thought to pose little risk for continued transmission of LF. Five low-antigen prevalence communes in Haiti, characterized as part of a national survey, were further assessed for transmission in this study. An initial evaluation of schoolchildren was performed in each commune to identify antigen-positive children who served as index cases for subsequent community surveys conducted among households neighboring the index cases. Global positioning system (GPS) coordinates and immunochromatographic tests (ICT) for filarial antigenemia were collected on approximately 1,600 persons of all ages in the five communes. The relationship between antigen-positive cases in the community and distance from index cases was evaluated using multivariate regression techniques and analyses of spatial clustering. Community surveys demonstrated higher antigen prevalence in three of the five communes than was observed in the original mapping survey; autochthonous cases were found in the same three communes. Regression techniques identified a significantly increased likelihood of being antigen-positive when living within 20 meters of index cases when controlling for age, gender, and commune. Spatial clustering of antigen-positive cases was observed in some, but not all communes. Our results suggest that localized transmission was present even in low-prevalence settings and suggest that better surveillance methods may be needed to detect microfoci of LF transmission.

Methods for estimating population coverage of mass distribution programmes: a review of practices in relation to trachoma control

In the context of trachoma control, population coverage with mass drug administration (MDA) using antibiotics is measured using routine data. Due to the limitations of administrative records as well as the potential for bias from incomplete or incorrect records, a literature review of coverage survey methods applied in neglected tropical disease control programmes and immunisation outreach was conducted to inform the design of coverage surveys for trachoma control. Several methods were identified, including the '30 × 7' survey method for the Expanded Programme on Immunization (EPI 30×7), other cluster random sampling (CRS) methods, lot quality assurance sampling (LQAS), purposive sampling and routine data. When compared against one another, the EPI and other CRS methods produced similar population coverage estimates, whilst LQAS, purposive sampling and use of administrative data did not generate estimates consistent with CRS. In conclusion, CRS methods present a consistent approach for MDA coverage surveys despite different methods of household selection. They merit use until standard guidelines are available. CRS methods should be used to verify population coverage derived from LQAS, purposive sampling methods and administrative reports.

Incremental cost of conducting population-based prevalence surveys for a neglected tropical disease: the example of trachoma in 8 national programs

Background

Trachoma prevalence surveys provide the evidence base for district and community-wide implementation of the SAFE strategy, and are used to evaluate the impact of trachoma control interventions. An economic analysis was performed to estimate the cost of trachoma prevalence surveys conducted between 2006 and 2010 from 8 national trachoma control programs in Africa.

Methodology and Findings

Data were collected retrospectively from reports for 165 districts surveyed for trachoma prevalence using a cluster random sampling methodology in Ethiopia, Ghana, Mali, Niger, Nigeria, Sudan, Southern Sudan and The Gambia.

The median cost per district survey was $4,784 (inter-quartile range [IQR] = $3,508–$6,650) while the median cost per cluster was $311 (IQR = $119–$393). Analysis by cost categories (personnel, transportation, supplies and other) and cost activity (training, field work, supervision and data entry) revealed that the main cost drivers were personnel and transportation during field work.

Conclusion

Population-based cluster random surveys are used to provide the evidence base to set objectives and determine when elimination targets have been reached for several neglected tropical diseases, including trachoma. The cost of conducting epidemiologically rigorous prevalence surveys should not be a barrier to program implementation or evaluation.

The costs of conducting population-based prevalence surveys for neglected tropical diseases such as trachoma are often cited as a reason that program managers do not conduct baseline or impact assessments when guidelines suggest they are warranted. The authors conducted a review of actual costs incurred during the implementation of 165 district level surveys in 8 national trachoma control programs to identify the median and mean costs per district and per cluster. In addition, the costs of the principal activities that are the most expensive were measured. The data show that field work is the most expensive activity for a prevalence survey, with personnel (per diems, allowances and accommodation) and transport costs driving the total cost of the survey. These findings can be used by program managers to budget for population-based prevalence surveys that are recommended for baseline and evaluation surveys, and periodic uptake surveys for neglected tropical diseases such as trachoma.

Rapid mapping of schistosomiasis and other neglected tropical diseases in the context of integrated control programmes in Africa

There is growing interest and commitment to the control of schistosomiasis and other so-called neglected tropical diseases (NTDs). Resources for control are inevitably limited, necessitating assessment methods that can rapidly and accurately identify and map high-risk communities so that interventions can be targeted in a spatially-explicit and cost-effective manner. Here, we review progress made with (1) mapping schistosomiasis across Africa using available epidemiological data and, more recently, climate-based risk prediction; (2) the development and use of morbidity questionnaires for rapid identification of high-risk communities of urinary schistosomiasis; and (3) innovative sampling-based approaches for intestinal schistosomiasis, using the lot quality assurance sampling technique. Experiences are also presented for the rapid mapping of other NTDs, including onchocerciasis, loiasis and lymphatic filariasis. Future directions for an integrated rapid mapping approach targeting multiple NTDs simultaneously are outlined, including potential challenges in developing an integrated survey tool. The lessons from the mapping of human helminth infections may also be relevant for the rapid mapping of malaria as its control efforts are intensified.