CORD-19:029c1c588047f1d612a219ee15494d2d19ff7439 JSONTXT 8 Projects

Protective Population Behavior Change in Outbreaks of Emerging Infectious Disease 1 2 Abstract estimation is complicated, however, by efforts to isolate infected individuals in hospitals or 45 other settings to decrease contact with the susceptible population. While the isolation of 46 infected individuals is beneficial and should be encouraged, it challenges data analysts because 47 it is time-varying and reflects dynamic and often unpredictable human behavior. Moreover, the 48 rate at which infected individuals are removed from the population typically accelerates 49 throughout an epidemic as awareness of the infectious threat increases, 8 a process Drake et al 50 referred to as "societal learning." 9 Obtaining accurate estimates of this time-varying removal of 51 infected persons, while difficult, improves the quality of compartmental models for epidemics 52 of infectious disease. 9,10 To our knowledge, however, no work has directly compared the rate of 53 behavioral adaption across multiple epidemics, societies, and geographic settings. 54 55 Many factors can affect how quickly effective isolation practices are implemented, such as 56 access to health care, local public health funding, international aid, and the efficacy of 57 information campaigns. 11 Local health care practices and non-formal healthcare systems also 58 provide care to patients during epidemics and can play a part in quarantining infected 59 individuals. 12 Previous work in Liberia has shown that a combination of these approaches 60 through simultaneous community engagement and clinical intervention is more effective than 61 any single intervention, with both health care access and utilization increasing hand-in-hand to 62 decrease EVD transmission during the 2013-2016 Ebola epidemic. 13 While infection prevention 63 often includes vaccination, progress to develop effective vaccines for emerging infections is 64 slow and not necessarily more effective than isolation of infected individuals. 14 Ring vaccination 65 with the rVSV-ZEBOV-GP Ebola vaccine 15 in the Democratic Republic of the Congo is 66 promising, 16 but previous work has suggested that ring vaccination may only provide a marginal 67 benefit to rigorous contact tracing and patient isolation. 17 68 69 The focus of this paper is the identification of key similarities and differences in the behavioral 70 response to outbreaks of three emerging zoonotic infections. We sought to determine how the 71 mean removal rate of infected individuals changed over the course of each outbreak as 72 measured by epidemic week and viral serial interval. Individuals often experience zoonotic and 73 emerging infections as innately more frightening than "familiar" diseases, leading to rapid 74 behavioral adaptations due to high perceived risk. 18 We compiled data for each outbreak location binned by epidemic week, to produce comparable 97 data for regression analysis. Epidemic weeks came from weekly onset dates described above. 98 We also binned the same data by serial interval, using 12 days as the estimated serial interval 99 for Ebola, 23 8 days for SARS, 24 and 7 days for MERS; 27 this was calculated as epidemic 100 week/(serial interval/7). Each dataset included, per week, the number of new cases, the 101 cumulative number of cases, mean DSOH and associated standard deviation, and MRR and 102 associated standard deviation. We removed epidemic weeks from the beginning of each 103 outbreak so that the first three epidemic weeks had greater than 0 cases of disease each in 104 order to focus on population-level behavioral adaptation to large-scale disease outbreaks 105 instead of adaptations to individual disease events early in an epidemic. We performed all 106 regression analyses using this binned data. 107 108 Initial regression analyses fit linear models to predict DSOH and MRR ( week) more than the 165 mean change in the 166 MRR of the Liberian 167 Ebola epidemic (Fig. 168 2). The mean change 169 of the MRR in the 170 Ebola epidemic in Lofa 171 County, Liberia, was 172 significantly higher than the mean change of the MRR for the overall Liberian epidemic and the 173 . CC-BY-NC-ND 4.0 International license author/funder. It is made available under a The copyright holder for this preprint (which was not peer-reviewed) is the . https://doi.org/10.1101/2020.01.27.921536 doi: bioRxiv preprint outbreak in Montserrado County, Liberia, regardless of predictor (epidemic week or serial 174 interval) (Fig. 2) . The three MERS outbreaks (Riyadh, Jeddah, and South Korea) did not differ 175 significantly from one another and had limited precision (Fig. 2) . 176 177 We narrowed the 195 associated confidence intervals (Fig. 2) . We identified little difference in the mean change of 196 the MRR for the Ebola outbreak depending on predictor (Fig. 2) While our findings demonstrate large and statistically significant differences in MRR, it is 231 notable that the calculated rates of change in the MRRs are within a factor of ten (when 232 calculated using epidemic week) to seven (when calculated using serial interval) of each other 233 (Fig. 2) , with the mean change being the lowest in the EVD outbreak in Liberia and the highest 234 in the MERS outbreak in South Korea. For modelers seeking to understand the epidemiology of 235 emerging infectious diseases with limited or no data from previous outbreaks, this study 236 provides a range of acceptable values for the MRR based on seven geographically distinct 237 outbreaks of three emerging diseases. Similarly, while large disparities in DSOH are obvious (Fig. 238 1), these data highlight that all societies quickly adapt to outbreaks of emerging infections. We have shown that public health practices for isolating infected individuals from the 249 susceptible population vary significantly by pathogen and location, but can in some cases be 250 predicted by the timing and serial interval of the epidemic. This study detected variation in 251 DSOH and MRR based on epidemic location and outbreak type, indicating that it may be 252 possible to estimate a general range of the rate of change in these variables over time. Due to 253 location-specific differences in DSOH and MRR, modelers who seek to develop forecasts early in 254 an outbreak would benefit from estimating an expected range for removal of infected 255 individuals using data from past outbreaks of the same pathogen in a similar setting. 256 Furthermore, the quality of these estimates will be impacted by the metric chosen, as seen by 257 the notable, but distinct, trends detected in DSOH and MRR. As seen in this study, utilizing a 258 well-chosen response variable with a relatively small amount of data can provide material for 259 making effective forecasts about public health behavior.