Study population and setting
This study used daily confirmed cases of SARS-CoV-2 infection in mainland China and Hong Kong between January 16 and March 24, 2020 to estimate the reproductive number Rt. Confirmed cases were defined as those with evidence of SARS-CoV-2 infection by RT-PCR and suspected cases with epidemiologic links to laboratory-confirmed cases. Within-city mobility data (measured as number of trips per person per day) sourced from Baidu (the Chinese equivalent of Google) for major metropolitan areas within each province from January 1 to March 24 were used as a proxy for economic activity. Biweekly correlations between estimated Rt and within-city mobility were estimated over time, with a lag of four days to account for delays between symptom onset and case reporting.
Summary of Main Findings
In Beijing and the five provinces with the most confirmed cases, within-city mobility and the estimated SARS-CoV-2 reproductive number were highly positively correlated through early February 2020. Correlations subsequently declined and became strongly negative for a sustained period in each province. In Hong Kong, correlations were weak. Using province-specific lags did not appreciably alter results. The authors infer from these results that efforts to restart the Chinese economy without a resurgence of SARS-CoV-2 spread have been successful.
Data and code for replicating the analyses are publicly available.
The simple measure of within-city mobility used here is a proxy for only one aspect of economic activity, and its limitations are not explored. Using the estimated reproduction rate of SARS-CoV-2 introduces an arguably unnecessary level of complication and uncertainty into the analysis; using the incidence rate of domestically acquired cases might suit the aims equally well. Estimated correlations over time are only reported for six regions, via a difficult-to-interpret figure without any measures of uncertainty. The authors did not explicitly consider the timing of non-pharmaceutical interventions in their analysis other than qualitatively associating them with mobility changes. Correlations became positive for regions other than Hubei at the end of the study period; these patterns do not fit well with the conclusions drawn by the authors.
This study does not provide much additional insight into the conventional understanding that China achieved domestic control over SARS-CoV-2 transmission despite a partial normalization of economic activity.
This review was posted on: 20 November 2020