Paul M. Emerson

Population-based prevalence surveys are essential for decision-making on interventions to achieve trachoma elimination as a public health problem. This paper outlines the methodologies of Tropical Data, which supports work to undertake those surveys. Tropical Data is a consortium of partners that supports health ministries worldwide to conduct globally standardised prevalence surveys that conform to World Health Organization recommendations. Founding principles are health ministry ownership, partnership and collaboration, and quality assurance and quality control at every step of the survey process. Support covers survey planning, survey design, training, electronic data collection and fieldwork, and data management, analysis and dissemination. Methods are adapted to meet local context and needs. Customisations, operational research and integration of other diseases into routine trachoma surveys have also been supported. Between 29^th February 2016 and 24^th April 2023, 3373 trachoma surveys across 50 countries have been supported, resulting in 10,818,502 people being examined for trachoma. This health ministry-led, standardised approach, with support from the start to the end of the survey process, has helped all trachoma elimination stakeholders to know where interventions are needed, where interventions can be stopped, and when elimination as a public health problem has been achieved. Flexibility to meet specific country contexts, adaptation to changes in global guidance and adjustments in response to user feedback have facilitated innovation in evidence-based methodologies, and supported health ministries to strive for global disease control targets.

Background: Great progress has been made toward the elimination of trachoma as a public-health problem. Mathematical and statistical models have been used to forecast when the program will attain the goal of the elimination of active trachoma, defined as prevalence of trachomatous inflammation—follicular in 1–9 year olds (TF_1–9) 1–9.Methodology/Principal findings: We calculated the mean number of years (95% CI) observed for an implementation unit (IU) to move from a baseline TF_1–9prevalence ≥5% to the elimination threshold, based on the region (Ethiopia vs. non-Ethiopia) and baseline prevalence category. Ethiopia IUs had significantly different rates of reaching the TF_1–9elimination threshold after a trachoma impact survey (TIS) compared to non-Ethiopia IUs across all baseline categories. We used those estimates to predict when remaining active trachoma-endemic IUs (TF_1–9≥5%) would have their last round of mass drug administration (MDA) based on the mean number of years required and number of MDA rounds already completed. Our model predicts that elimination of TF_1–9will be achieved in 2028 in Ethiopia (95% CI: 2026–2033) and 2029 outside of Ethiopia (95% CI: 2023–2034), with some IUs in East Africa predicted to be the last requiring MDA globally.Conclusions/Significance: Our empirical estimate is similar to those resulting from previous susceptible-infectious-susceptible (SIS) and mathematical models, suggesting that the forecast achievement of TF_1–9elimination is realistic with the caveat that although disease elimination progress can be predicted for most IUs, there is an important minority of IUs that is not declining or has not yet started trachoma elimination activities. These IUs represent an important barrier to the timely global elimination of active trachoma.

Models predict that the negative effects of delayed implementation in trachoma elimination programmes caused by the COVID-19 pandemic will be minimal, except in high prevalence districts where progress may be reversed. During times of change we must stand by our principles of evidence-based decision-making, but also be willing to show flexibility. Slow progress to elimination in high prevalence districts was already a significant challenge to the global programme and mitigation of COVID-related delays with enhanced implementation provides an opportunity to simultaneously address an unprecedented challenge and a pre-existing one.

Conjunctival examination for trachomatous inflammation—follicular (TF) guides public health decisions for trachoma. Smartphone cameras may allow remote conjunctival grading, but previous studies have found low sensitivity. A random sample of 412 children aged 1–9 years received an in-person conjunctival examination and then had conjunctival photographs taken with 1) a single-lens reflex (SLR) camera and 2) a smartphone coupled to a 3D-printed magnifying attachment. Three masked graders assessed the conjunctival photographs for TF. Latent class analysis was used to determine the sensitivity and specificity of each grading method for TF. Single-lens reflex photo-grading was 95.0% sensitive and 93.6% specific, and smartphone photo-grading was 84.1% sensitive and 97.6% specific. The sensitivity of the smartphone-CellScope device was considerably higher than that of a previous study using the native smartphone camera, without attachment. Magnification of smartphone images with a simple attachment improved the grading sensitivity while maintaining high specificity in a region with hyperendemic trachoma.

Trachoma prevalence surveys, including impact surveys (TIS) and surveillance surveys (TSS), provide information to program managers on the impact of the SAFE (surgery, antibiotics, facial cleanliness, and environmental improvement) strategy and current burden of disease, and they provide a crucial component of the evidence base necessary for the validation of the elimination of trachoma as a public health problem. The prevalence surveys included in this analysis are multi-level cluster random surveys that provide population-based estimates for program planning. This study conducted an analysis of the cost of 8 rounds of TIS/TSS executed in Amhara, Ethiopia, 2012–2016, comprising 232,357 people examined over 1,828 clusters in 187 districts. Cost data were collected retrospectively from accounting and procurement records from the implementing partner, The Carter Center, and coded by survey activity (i.e. training and field work) and input category (i.e. personnel, transportation, supplies, venue rental, and other). Estimates of staff time were obtained from The Carter Center Ethiopia. Data were analyzed by activity and input category. The mean total cost per cluster surveyed was $752 (standard deviation $101). Primary cost drivers were personnel (39.6%) and transportation (49.2%), with costs increasing in the last 3 rounds of TIS/TSS. Despite the considerable cost of conducting TIS and TSS, these surveys provide necessary information for program managers. Limited options are available to reduce the costs of TIS/TSS and gain economies of scale, as the surveys must be designed to achieve their designated sample size. However, surveys must also be designed in a way that is possible to be executed given the financial resources, personnel, and time required. Program managers can use these findings to improve estimates of the total cost of a survey and its components to ensure that sufficient resources are budgeted accordingly. Population-based trachoma surveys are necessary to determine the impact of interventions and to build the case for the validation of elimination of trachoma as a public health problem. Many trachoma-endemic areas are currently receiving mass drug administration, which will require trachoma impact surveys (TIS) and trachoma surveillance surveys (TSS). We analyzed the costs accrued by the implementing non-governmental organization during 8 rounds of trachoma surveys in Amhara, Ethiopia, that included 187 districts surveyed over four years. The costs were sorted by activity (i.e. training and field work) and input category (i.e. transportation, personnel, venue rental, supplies, and other). Field work was the most expensive activity for trachoma surveys, with transportation and personnel as the most significant drivers of cost. As the global trachoma program matures, these surveys must be executed given the financial resources, personnel, and time available. Program managers can use these findings to improve total cost estimates of trachoma surveys and survey components to ensure that sufficient resources are being budgeted accordingly. Additionally, managers in hyperendemic settings can use these findings to better understand the drivers in cost of repeated surveys in a single geography.

Nucleic acid amplification tests are increasingly used to detect ocular chlamydia infection in trachoma research and programs. To evaluate the reliability of Chlamydia trachomatis detection by the Abbott RealTime CT/NG assay on the m2000 platform, three conjunctival samples were collected from each of 200 children aged 0–9 years in Ethiopia: two from the right eye and one from the left eye. Four aliquots were processed for each child: two from the first right eye sample, one from the second right eye sample, and one from the left eye sample. Sixty-nine swabs were processed in a U.S. laboratory and 131 in an Ethiopian laboratory. Intra-class correlation coefficients (ICCs) were high when comparing two aliquots from the same swab (ICC ranged from 0.96 to 0.99), two separate swabs from the right eye (0.89–0.91), and one right and one left eye swab (0.87–0.89), indicating reliable chlamydial load assessment across different samples and laboratory settings.

At baseline in 2006, Amhara National Regional State, Ethiopia, was the most trachoma-endemic region in the country. Trachoma impact surveys (TIS) were conducted in all districts between 2010 and 2015, following 3–5 years of intervention with the WHO-recommended SAFE (surgery, antibiotics, facial cleanliness, and environmental improvement) strategy. A multistage cluster random sampling design was used to estimate the district-level prevalence of trachoma. In total, 1,887 clusters in 152 districts were surveyed, from which 208,265 individuals from 66,089 households were examined for clinical signs of trachoma. The regional prevalence of trachomatous inflammation-follicular (TF) and trachomatous inflammation-intense among children aged 1–9 years was 25.9% (95% CI: 24.9–26.9) and 5.5% (95% CI: 5.2–6.0), respectively. The prevalence of trachomatous scarring and trachomatous trichiasis among adults aged ≥ 15 years was 12.9% (95% CI: 12.2–13.6) and 3.9% (95% CI: 3.7–4.1), respectively. Among children aged 1–9 years, 76.5% (95% CI: 75.3–77.7) presented with a clean face; 66.2% (95% CI: 64.1–68.2) of households had access to water within 30 minutes round-trip, 48.1% (95% CI: 45.5–50.6) used an improved water source, and 46.2% (95% CI: 44.8–47.5) had evidence of a used latrine. Nine districts had a prevalence of TF below the elimination threshold of 5%. In hyperendemic areas, 3–5 years of implementation of SAFE is insufficient to achieve trachoma elimination as a public health problem; additional years of SAFE and several rounds of TIS will be required before trachoma is eliminated.

Mass drug administration (MDA) with azithromycin may reduce under-5 child mortality (U5M) in sub-Saharan Africa. Here, we conducted a pooled analysis of all published cluster-randomized trials evaluating the effect of azithromycin MDA on child mortality. We pooled data from cluster-randomized trials randomizing communities to azithromycin MDA versus control. We calculated mortality rates in the azithromycin and control arms in each study, and by country for multisite studies including multiple countries. We conducted a two-stage individual community data meta-analysis to estimate the effect of azithromycin for prevention of child mortality. Three randomized controlled trials in four countries (Ethiopia, Malawi, Niger, and Tanzania) were identified. The overall pooled mortality rate was 15.9 per 1,000 person-years (95% confidence interval [CI]: 15.5–16.3). The pooled mortality rate was lower in azithromycin-treated communities than in placebo-treated communities (14.7 deaths per 1,000 person-years, 95% CI: 14.2–15.3 versus 17.2 deaths per 1,000 person-years, 95% CI: 16.5–17.8). There was a 14.4% reduction in all-cause child mortality in communities receiving azithromycin MDA (95% CI: 6.3–21.7% reduction, P = 0.0007). All-cause U5M was lower in communities receiving azithromycin MDA than in control communities, suggesting that azithromycin MDA could be a new tool to reduce child mortality in sub-Saharan Africa. However, heterogeneity in effect estimates suggests that the magnitude of the effect may vary in time and space and is currently not predictable.

The World Health Organization recommends annual mass azithromycin administration in communities with at least 10% prevalence of trachomatous inflammation–follicular (TF) in children, with further treatment depending on reassessment after 3–5 years. However, the effect of stopping mass azithromycin distribution after multiple rounds of treatment is not well understood. Here, we report the results of a cluster-randomized trial where communities that had received 4 years of treatments were then randomized to continuation or discontinuation of treatment. In all, 48 communities with 3,938 children aged 0–9 years at baseline in northern Ethiopia had received 4 years of annual or twice yearly mass azithromycin distribution as part of the TANA I trial. We randomized these communities to either continuation or discontinuation of treatment. Individuals in the communities in the continuation arm were offered either annual or twice yearly distribution of a single directly observed dose of oral azithromycin. The primary outcome was community prevalence of ocular chlamydial infection in a random sample of children aged 0–9 years, 36 months after baseline. We also assessed the change from baseline to 36 months in ocular chlamydia prevalence within each arm. We compared 36-month ocular chlamydia prevalence in communities randomized to continuation versus discontinuation in a model adjusting for baseline ocular chlamydia prevalence. A secondary prespecified analysis assessed the rate of change over time in ocular chlamydia prevalence between arms. In the continuation arm, mean antibiotic coverage was greater than 90% at all time points. In the discontinuation arm, the mean prevalence of infection in children aged 0–9 years increased from 8.3% (95% CI 4.2% to 12.4%) at 0 months to 14.7% (95% CI 8.7% to 20.8%, P = 0.04) at 36 months. Ocular chlamydia prevalence in communities where mass azithromycin distribution was continued was 7.2% (95% CI 3.3% to 11.0%) at baseline and 6.6% (95% CI 1.1% to 12.0%, P = 0.64) at 36 months. The 36-month prevalence of ocular chlamydia was significantly lower in communities continuing treatment compared with those discontinuing treatment (P = 0.03). Limitations of the study include uncertain generalizability outside of trachoma hyperendemic regions. In this study, ocular chlamydia infection rebounded after 4 years of periodic mass azithromycin distribution. Continued distributions did not completely eliminate infection in all communities or meet WHO control goals, although they did prevent resurgence. This study was prospectively registered at clinicaltrials.gov (clinicaltrials.gov NCT01202331).