Clinical cholera surveillance sensitivity in Bangladesh and implications for large-scale disease control

Introduction A surveillance system that is sensitive to detecting high burden areas is critical for achieving widespread disease control. In 2014, Bangladesh established a nationwide, facility-based cholera surveillance system for Vibrio cholerae infection. We sought to measure the sensitivity of this surveillance system to detect cases to assess whether cholera elimination targets outlined by the Bangladesh national control plan can be adequately measured. Methods We overlaid maps of nationally-representative annual V. cholerae seroincidence onto maps of the catchment areas of facilities where confirmatory laboratory testing for cholera was conducted, and identified its spatial complement as surveillance greyspots, areas where cases likely occur but go undetected. We assessed surveillance system sensitivity and changes to sensitivity given alternate surveillance site selection strategies. Results We estimated that 69% of Bangladeshis (111.7 million individuals) live in surveillance greyspots, and that 23% (25.5 million) of these individuals live in areas with the highest V. cholerae infection rates. Conclusions The cholera surveillance system in Bangladesh has the ability to monitor progress towards cholera elimination goals among 31% of the country's population, which may be insufficient for accurately measuring progress. Increasing surveillance coverage, particularly in the highest risk areas, should be considered.


Introduction 26
Bangladesh has among the highest national rates of Vibrio cholerae infection in the 27 world [1]; a nationally-representative serosurvey estimated that roughly 17% (95% CI: 11-24%) 28 Bangladesh, however, with which to compare and measure the reduction in morbidity and 40 mortality. While vaccination campaigns, water, sanitation, and hygiene interventions, and 41 improved case management are the primary tools to achieve these elimination targets, a 42 cholera surveillance system with widespread geographical coverage is necessary to target 43 interventions to the highest burden areas and monitor progress from endemic transmission to 44 elimination. 45 The US Centers for Disease Control and Prevention (CDC) has a standardized 46 framework for evaluating public health surveillance systems, which may be applied flexibly to 47 systems with varying goals [5,6]. The public health goal of widespread disease control and 48 elimination, like that for cholera in Bangladesh, requires the identification and monitoring of 49 areas with high case counts and high relative risk across the population in a timely manner. 50 However, quantitative evaluations of sensitivity are hard to obtain when the surveillance system 51 . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted June 5, 2021.  Figure 1A). 76 . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted June 5, 2021. ; https://doi.org/10.1101/2021.06.02.21258249 doi: medRxiv preprint 4 We defined the relative and absolute magnitudes of the infection risk and defined 77 thresholds for high, moderate, and low relative and absolute risk across 25 km 2 grid cells. We 78 used the 25th and 75th percentiles of the mean grid cell-level risk (relative or absolute) to define 79 cutoffs for moderate and high risk areas.

Results 126
Surveillance system sensitivity 127 In the year preceding the 2015/16 serosurvey, 16% (95% CI: 13-23%) or 8 million (95% 128 CI: 4.6-11.9) of the 51 million people living in the cholera surveillance zone had been infected 129 . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (16%, 95% CI: 9-23% and 16%, 95% CI: 9-24%, respectively) as the current surveillance 155 system and more directed strategies such as the Relative Risk Equity and Absolute Risk Equity 156 . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted June 5, 2021. ; https://doi.org/10.1101/2021.06.02.21258249 doi: medRxiv preprint 7 (18%, 95% CI: 10-25% and 17%, 95% CI: 10-24%, respectively). While percentage differences 157 were small, the mean number of infections captured varied by up to 1 million between some 158 pairs of strategies (Full results in Table S3). If all 491 public facilities were used in the 159 surveillance system, 97% (95% CI: 96.6-97.5%) of infections would be captured (27 million, 160 95% CI: 16.6 -37.9). 161 An examination of the ICC across multiple models of infections in the cholera 162 surveillance zone revealed that there was substantially greater uncertainty in the underlying 163 seroincidence risk estimates than in simulations for the same strategy. The ICC ranged from 164 0.84 to 0.97 for models with random effects on seroincidence posterior draws, while it ranged 165 only from 0.01 to 0.1 for models with random effects on simulations (full results in Table S4). 166 There were no major differences between strategies. 167

Discussion 168
The cholera sentinel surveillance system in Bangladesh is the only data source available 169 to monitor progress towards national disease control by 2030. Our study described the 170 characteristics of cholera surveillance greyspots, geographic areas where cases are unlikely to 171 be detected because they reside outside the catchment areas of sentinel surveillance sites. We 172 estimated that roughly 111.7 million individuals (69% of Bangladesh's population) live in 173 greyspots, and that 23% of these individuals (25.5 million people) live in areas with extremely 174 high risk of cholera infection (where the mean annual seroincidence rate is 22%) (Figure 3). The 175 alternative methods for selecting sentinel sites that we explored produced only minor 176 improvements in the capture of cholera infections, although more optimized strategies could be 177 devised. Without changes to the surveillance system, it will be impossible to monitor high 178 cholera burden areas in much of the country, which is a substantive impediment to measuring 179 progress on elimination. 180 The original stated goals for Bangladesh's national cholera sentinel surveillance system 181 were to monitor cholera seasonality and epidemiology while also tracking the burden of disease 182 . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted June 5, 2021. ; https://doi.org/10.1101/2021.06.02.21258249 doi: medRxiv preprint 8 in areas believed to have high prevalence; the objectives may not have had the disease 183 elimination goal in mind. Sentinel surveillance systems that are sensitive to capturing high risk 184 areas are critical to disease elimination efforts to measure disease burden, identify at-risk 185 populations, and monitor the health impacts of interventions in target populations. For cholera 186 control specifically, the drivers of disease transmission are highly local (i.e., fecal-contaminated 187 water and food), with great variation even between households, and campaigns against 188 waterborne diseases must be targeted effectively to high risk areas to achieve 189 success. Depending on whether risk is defined in relative or absolute terms, 58-74% of 190 individuals living in high risk areas were not captured within the cholera surveillance zone. While 191 it is often difficult to quantify the performance of a sentinel system to monitor high risk 192 populations, future work may use our framework to assess how multiple, simulated sentinel 193 selection site strategies may be better suited to achieving different system goals. Our approach has several limitations. We assumed that hospital catchment areas could 220 be defined with simple radial buffers, similar to previous work [14]. A more accurate approach to 221 estimating hospital catchment areas would use patient demographic, symptom, and home 222 address data, and account for barriers to healthcare seeking [15,16]; in reality, the cholera 223 surveillance zone is likely smaller than what we assumed resulting in overestimates of system 224 sensitivity. Conversely, the functional coverage of the cholera surveillance zone may be more 225 expansive than our stated assumptions if private clinics and facilities outside of the national 226 sentinel surveillance system use RDTs or culture to confirm suspected cholera cases, and 227 event-based surveillance systems like media surveillance and hotlines routinely detect disease 228 outbreaks, though samples still have to be processed and confirmed in a lab [17]. Discussions 229 with experts suggest that testing outside of the sentinel surveillance system is low, however, 230 and unlikely to change our results substantially. Finally, while clinical cholera incidence is almost 231 certainly lower than seroincidence, their geographic distributions of burden are likely to be 232 similar and our results should serve as a reasonable proxy for system sensitivity for clinical 233 cholera detection. 234 The surveillance evaluation framework proposed here, which aims to quantify 235 surveillance system sensitivity to monitoring large-scale reductions in cholera morbidity, may 236 . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted June 5, 2021. ; https://doi.org/10.1101/2021.06.02.21258249 doi: medRxiv preprint nonetheless prove useful in the context of nationwide control or elimination efforts for other 237 vaccine preventable diseases, like typhoid or Japanese encephalitis [18,19]. By comparing 238 surveillance data to an external validation instrument like a population-representative 239 serosurvey, it is possible to quantify surveillance system sensitivity and perform targeting of 240 interventions that can contribute to an effective elimination strategy. Beyond providing 241 surveillance metrics, an external validation instrument like cross-sectional serology can be used 242 to motivate specific system improvements such as the selection of alternate sentinel sites to 243 increase system sensitivity or even a more cost-effective surveillance system to capture the risk 244 of both asymptomatic and symptomatic infection. Further, by applying multiple definitions of 245 disease risk (e.g., relative versus absolute risk), we can identify surveillance greyspots that are 246 robust to multiple dimensions of information. For example, though the relative risk of V. cholerae 247 infection may be considered low in an urban area, the estimated absolute number of infections 248 could be high; we would not want sentinel surveillance sites to be concentrated only in high 249 relative risk areas. Monitoring changes in relative and absolute risk over time, and in rural 250 versus urban areas is important, especially as access to care changes. 251 Ultimately, the goal of public health surveillance systems are to generate data for action 252 towards improving public health, but if significant gaps in the surveillance system exist such 253 goals may never be met. In Bangladesh, the goal of cholera elimination will likely be hindered by 254 the lack of geographic or population coverage if changes to the system are not made; any 255 documented reductions in morbidity and mortality to quantitate progress will only be among 31% 256 of the country's population. For any disease, a strong elimination plan should demand high 257 quality surveillance data and using more rigorous and cost-effective methods to evaluate 258 surveillance data is an imperative first step. 259 260 . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted June 5, 2021. is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted June 5, 2021. CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted June 5, 2021. ; https://doi.org/10.1101/2021.06.02.21258249 doi: medRxiv preprint . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted June 5, 2021. ; https://doi.org/10.1101/2021.06.02.21258249 doi: medRxiv preprint