Estimating the household secondary attack rate and serial interval of COVID-19 using social media

Abstract

We propose a method to estimate the household secondary attack rate (hSAR) of COVID-19 in the United Kingdom based on activity on the social media platform X, formerly known as Twitter. Conventional methods of hSAR estimation are resource intensive, requiring regular contact tracing of COVID-19 cases. Our proposed framework provides a complementary method that does not rely on conventional contact tracing or laboratory involvement, including the collection, processing, and analysis of biological samples. We use a text classifier to identify reports of people tweeting about themselves and/or members of their household having COVID-19 infections. A probabilistic analysis is then performed to estimate the hSAR based on the number of self or household, and self and household tweets of COVID-19 infection. The analysis includes adjustments for a reluctance of Twitter users to tweet about household members, and the possibility that the secondary infection was not acquired within the household. Experimental results for the UK, both monthly and weekly, are reported for the period from January 2020 to February 2022. Our results agree with previously reported hSAR estimates, varying with the primary variants of concern, e.g. delta and omicron. The serial interval (SI) is based on the time between the two tweets that indicate a primary and secondary infection. Experimental results, though larger than the consensus, are qualitatively similar. The estimation of hSAR and SI using social media data constitutes a new tool that may help in characterizing, forecasting and managing outbreaks and pandemics in a faster, affordable, and more efficient manner.

Fig. 1. Monthly/weekly household secondary attack rates with key events for the period from January 2020 to February 2022.

Monthly and weekly cohort sizes and household SAR (hSAR_br) marked with key events for the period from January 2020 to February 2022, inclusive, for users with p_h > 0 and an assumed maximum serial interval of 14 days. The UK Health Security Agency (UKHSA) (formerly Public Health England) SAR scores are the weighted average of SAR scores for different variants published by UKHSA.

Fig. 2. Comparison of hSAR_br across for various cohort sizes.

The solid line depicts hSAR_br values calculated using all the users in our Twitter cohort. The dotted values show hSAR_br for fixed cohort sizes of A. 100, B. 1000, C. 2000 D.4000 users. For a given cohort size, we create 50 cohorts, uniformly sampled from the available data, total cohort size permitting.

Fig. 3. Adjustments to the household secondary attack rate estimates.

A hSAR depicts the monthly household SAR assuming no reluctance to tweet about a household member. hSAR_b depicts the household SAR adjusted for reluctance. rSAR is an estimate of second infections from outside of the household. Our final estimate of household SAR is hSAR_br = hSAR_b − rSAR. B The monthly values for α₁ (the hSAR estimate calculated from the subset of users who tweeted about being infected and subsequently tweeted that a household member was infected) and α₂ (the estimate calculated from the subset of users who tweeted about a household member being infected and subsequently tweeted that they were infected) for users with p_h > 0 assuming a maximum serial interval of 14 days.

Fig. 4. Serial interval for Alpha, Delta, and Omicron dominant periods.

Histogram of serial interval for daily bins with an assumed maximum serial interval of 14 days for the periods A All the months from Jan 2020 to Feb 2022. B Alpha dominant period (18 December 2020 to 15 May 2021). C Delta dominant period (22 May 2021 to 19 December 2021). D Omicron dominant period (23 December 2021 to 28 February 2022).