Quantifying Contact Patterns: Development and Characteristics of the British Columbia COVID-19 Population Mixing Patterns Survey

Abstract

Purpose: Several non-pharmaceutical interventions such as physical distancing, self-isolation, a stay-at-home order, hand washing, and schools and businesses closures were implemented in British Columbia (BC) following the first lab-tested case of COVID-19 on January 26, 2020. These interventions were aimed at minimizing in-person contacts that could potentially lead to new COVID-19 infections. The BC COVID-19 Population Mixing Patterns survey (BC-Mix) was established as a surveillance system to measure behaviour and contact patterns in BC over time to inform the timing of the easing/re-imposition of control measures. We describe the BC-Mix survey design and the demographic characteristics of respondents. Methods & Materials: The ongoing repeated online survey was launched in September 2020. Participants are recruited through a variety of social media platforms including Instagram, Facebook, YouTube, and community group mailing lists. A follow up survey is sent to participants two to four weeks after completing the first iteration. Survey responses are weighted to BC's population by age, sex, geography, and ethnicity to obtain generalizable estimates. A survey completion rate of at least 33% AND a valid response for the sex questionnaire item AND a valid response for age questionnaire item were required for inclusion in weighting and further analysis. Additional indices such as material and social deprivation index, and residential instability are generated using census and location data. Results: As of June 14, 2021, over 58,000 residents of BC had participated in the survey of which 31,007 survey responses were eligible for analysis. Of the eligible participants, about 60% provided consent for monthly follow up and about 26% provided their personal health numbers for linkage with other healthcare utilization databases. Approximately 51% were females 39% were 55 years or older, 63% identified as white or not a visible minority and 48% had at least a university degree. Conclusion: The pandemic response is best informed by surveillance systems capable of timely assessment of behaviour patterns. BC-Mix survey respondents represented a large cohort of British Columbians providing near real-time information on behavioural and contact patterns in BC. Data from the BC-Mix survey continues to inform provincial COVID-19-related control measures.