COVID-19: Effectiveness of Non-Pharmaceutical Interventions in the United States before Phased Removal of Social Distancing Protections Varies by Region
This article has been Reviewed by the following groups
Listed in
- Evaluated articles (ScreenIT)
Abstract
Although coronavirus disease 2019 (COVID-19) emerged in January 2020, there is no quantified effect size for non-pharmaceutical interventions (NPI) to control the outbreak in the continental US. Objective. To quantify national and sub-national effect sizes of NPIs in the US. Design. This is an observational study for which we obtained daily county level COVID-19 cases and deaths from January 22, 2020 through the phased removal of social distancing protections. A stepped-wedge cluster-randomized trial (SW-CRT) analytical approach is used, leveraging the phased implementation of policies. Data include 3142 counties from all 50 US states and the District of Columbia. Exposures. County-level NPIs were obtained from online county and state policy databases, then classified into four intervention levels: Level 1 (low) – declaration of a State of Emergency; Level 2 (moderate) – school closures, restricting nursing home access, or closing restaurants and bars; Level 3 (high) – non-essential business closures, suspending non-violent arrests, suspending elective medical procedures, suspending evictions, or restricting mass gatherings of at least 10 people; and Level 4 (aggressive) – sheltering in place / stay-at-home, public mask requirements, or travel restrictions. Additional county-level data were obtained to record racial (Black, Hispanic), economic (educational level, poverty), demographic (rural/urban) and climate factors (temperature, specific humidity, solar radiation). Main Outcomes. The primary outcomes are rates of COVID-19 cases, deaths and case doubling times. NPI effects are measured separately for nine US Census Region (Pacific, Mountain, West North Central, East North Central, West South Central, East South Central, South Atlantic, Middle Atlantic, New England). Results. Aggressive NPIs (level 4) significantly reduced COVID-19 case and death rates in all US Census Regions, with effect sizes ranging from 4.1% to 25.7% and 5.5% to 25.5%, respectively, for each day they were active. No other intervention level achieved significance across all US Regions. Intervention levels 3 and 4 both increased COVID-19 doubling times, with effects peaking at 25 and 40 days after initiation of each policy, respectively. The effectiveness of level 3 NPIs varied, reducing case rates in all regions except North Central states, but associated with significantly higher death rates in all regions except Pacific states. Intervention levels 1 and 2 did not indicate any effect on COVID-19 propagation and, in some regions, these interventions were associated with increased COVID-19 cases and deaths. Heterogeneity of NPI effects are associated with racial composition, poverty, urban-rural environment, and climate factors. Conclusion. Aggressive NPIs are effective tools to reduce COVID-19 propagation and mortality. Reducing social and environmental disparities may improve NPI effects in regions where less strict policies are in place.
Article activity feed
-
SciScore for 10.1101/2020.08.18.20177600: (What is this?)
Please note, not all rigor criteria are appropriate for all manuscripts.
Table 1: Rigor
Institutional Review Board Statement not detected. Randomization not detected. Blinding not detected. Power Analysis not detected. Sex as a biological variable not detected. Table 2: Resources
No key resources detected.
Results from OddPub: We did not detect open data. We also did not detect open code. Researchers are encouraged to share open data when possible (see Nature blog).
Results from LimitationRecognizer: We detected the following sentences addressing limitations in the study:There are several limitations to note in our study. First, the analysis does not have an accurate representation of the availability of testing (or the number of tests administered) at the county level for the time series. As …
SciScore for 10.1101/2020.08.18.20177600: (What is this?)
Please note, not all rigor criteria are appropriate for all manuscripts.
Table 1: Rigor
Institutional Review Board Statement not detected. Randomization not detected. Blinding not detected. Power Analysis not detected. Sex as a biological variable not detected. Table 2: Resources
No key resources detected.
Results from OddPub: We did not detect open data. We also did not detect open code. Researchers are encouraged to share open data when possible (see Nature blog).
Results from LimitationRecognizer: We detected the following sentences addressing limitations in the study:There are several limitations to note in our study. First, the analysis does not have an accurate representation of the availability of testing (or the number of tests administered) at the county level for the time series. As this availability changed over time for all counties in the US, we cannot accurately characterize the population at risk for COVID-19 case detection. However, we are confident in COVID-related deaths reported. Second, we do not have measures of public acceptance and adherence to NPIs. Anecdotal data suggest significant variation of NPI adherence and that with phased reopening of some states, NPIs were viewed as an affront to civil liberties. Data on NPI adherence are critical to accurately measuring “dose” of the NPI to accurately characterize the population-level effect. Third, we do not look at individual interventions. However, this is a choice rather than limitation as sets of NPIs are considered a more appropriate response than single interventions, which have never been employed historically without others. Finally, we would ideally be applying these methods to a randomized design, which is impossible for COVID-19. Thus, our inferences are drawn from observational data.
Results from TrialIdentifier: No clinical trial numbers were referenced.
Results from Barzooka: We did not find any issues relating to the usage of bar graphs.
Results from JetFighter: We did not find any issues relating to colormaps.
Results from rtransparent:- Thank you for including a conflict of interest statement. Authors are encouraged to include this statement when submitting to a journal.
- Thank you for including a funding statement. Authors are encouraged to include this statement when submitting to a journal.
- No protocol registration statement was detected.
-
