Controlling the first wave of the COVID-19 pandemic in Malawi: Results from a multi-round study

Abstract

We investigated behavioral responses to COVID-19 in Malawi, where a first wave of the pandemic occurred between June and August 2020. Contrary to many countries on the African continent, the Government of Malawi did not impose a lockdown or a stay-at-home order in response to the initial spread of SARS-CoV-2. We hypothesized that, in the absence of such requirements to restrict social interactions, individuals would primarily seek to reduce the risk of SARS-CoV-2 transmission during contacts, rather than reduce the extent of their social contacts. We analyzed 4 rounds of a panel survey spanning time periods before, during and after the first wave of the COVID-19 pandemic in Malawi. Five hundred and forty-three participants completed 4 survey interviews between April and November 2020. We found that the likelihood of attending various places and events where individuals work and/or socialize remained largely unchanged during that time. Over the same time frame, however, participants reported adopting on a large scale several behaviors that reduce the transmissibility of SARS-CoV-2 during contacts. The percentage of panel participants who reported practicing physical distancing thus increased from 9.8% to 47.0% in rural areas between April-May 2020 and June-July 2020, and from 11.4% to 59.4% in urban areas. The percentage of respondents who reported wearing a facial mask to prevent the spread of SARS-CoV-2 also increased, reaching 67.7% among rural residents in August-September 2020, and 89.6% among urban residents. The pace at which these behaviors were adopted varied between population groups, with early adopters of mask use more commonly found among more educated office workers, residing in urban areas. The adoption of mask use was also initially slower among women, but later caught up with mask use among men. These findings stress the importance of behavioral changes in containing future SARS-CoV-2 outbreaks in settings where access to vaccination remains low. They also highlight the need for targeted outreach to members of socioeconomic groups in which the adoption of protective behaviors, such as mask use, might be delayed.

Fig 1. Adoption of restrictions on social interactions in African countries, from the beginning of the COVID-19 pandemic to October, 1^st 2020.

Source: Data on COVID-related policies were extracted from the Oxford COVID-19 Government Response Tracker. Administrative boundaries (Admin 0) were obtained from a World Bank repository of publicly available geographic information (available at: https://datacatalog.worldbank.org/search/dataset/0038272/World-Bank-Official-Boundaries).

Fig 2. Recorded COVID-19 cases in Malawi and study timelines.

Notes: Data plotted in this graph were drawn from daily reports from the Public Health Institute of Malawi, and obtained via the data portal of the European Center for Disease Control (ECDC). The “o” symbols represent the counts of confirmed COVID-19 cases reported on a given day. The shaded area noted “R1” refers to round 1 of data collection, which occurred between April 22^nd and May 14^th, 2020. Areas noted “R2”, “R3” and “R4” refer to subsequent rounds of data collection. Vertical dashed lines indicate significant events in the course of the pandemic in Malawi.

Fig 3. Attendance of places and events in Malawi (April to November 2020).

Notes: the data plotted in this graph were collected from a series of questions asking respondents whether they had attended the places listed on the y-axis in the past 7 days prior to the survey. The places that appear on the y-axis are ordered according to their attendance in urban areas in round 1. Error bars represent 95% confidence intervals. In calculating confidence intervals, standard errors were adjusted for the clustering of observations within respondents.

Fig 4. Adoption of behaviors to reduce transmissibility of SARS-CoV2 in Malawi (April to November 2020).

Notes: the data plotted in this graph were collected from a series of questions asking respondents what they had done to prevent the spread of SARS-CoV-2 in the past month. The behaviors that appear on the y-axis are ordered according to their prevalence in urban areas in round 1. Error bars represent 95% confidence intervals. In calculating confidence intervals, standard errors were adjusted for the clustering of observations within respondents.

Fig 5. Frequency of mask use in Malawi (August to November 2020).

Notes: in rounds 3 & 4, we asked respondents who reported that they had used a mask to prevent the spread of SARS-CoV-2, how often they had done so. These data were not collected in rounds 1 & 2.