Estimating the Impact of Statewide Policies to Reduce Spread of Severe Acute Respiratory Syndrome Coronavirus 2 in Real Time, Colorado, USA

Andrea G. Buchwald; Jude Bayham; Jimi Adams; David Bortz; Kathryn Colborn; Olivia Zarella; Meghan Buran; Jonathan Samet; Debashis Ghosh; Rachel Herlihy; Elizabeth J. Carlton

Disclosures

Emerging Infectious Diseases. 2021;27(9):2312-2322.

Abstract and Introduction

Abstract

The severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) pandemic necessitated rapid local public health response, but studies examining the impact of social distancing policies on SARS-CoV-2 transmission have struggled to capture regional-level dynamics. We developed a susceptible-exposed-infected-recovered transmission model, parameterized to Colorado, USA–specific data, to estimate the impact of coronavirus disease–related policy measures on mobility and SARS-CoV-2 transmission in real time. During March–June 2020, we estimated unknown parameter values and generated scenario-based projections of future clinical care needs. Early coronavirus disease policy measures, including a stay-at-home order, were accompanied by substantial decreases in mobility and reduced the effective reproductive number well below 1. When some restrictions were eased in late April, mobility increased to near baseline levels, but transmission remained low (effective reproductive number <1) through early June. Over time, our model parameters were adjusted to more closely reflect reality in Colorado, leading to modest changes in estimates of intervention effects and more conservative long-term projections.

Introduction

Mathematical transmission models are useful tools for predicting the magnitude, duration, and severity of the severe acute respiratory coronavirus 2 (SARS-CoV-2) pandemic. However, widely used national-level models might not capture regional heterogeneity. The coronavirus disease (COVID-19) outbreak in Colorado, USA, has been the subject of numerous discrepant projections from the Institute for Health Metrics and Evaluation and other modeling groups,^[1] which might have structural and data source explanations, highlighting the need for ensuring that models are fit to local epidemiologic data.^[2–4]

We report on our experience using a locally tailored model to inform policy in Colorado. Social distancing policies, intended to decrease contact rates, have been cornerstone public health tools for pandemic control, and these strategies have been adopted to control SARS-CoV-2 globally.^[2,5] Until recently, evidence of the effects of social distancing has come primarily from studies of the consequences of school and transit closures on influenza transmission.^[3,4,6] Early evidence suggests that social distancing policies can suppress transmission of SARS-CoV-2,^[7,8] and recent evidence suggests a strong correlation between mobility and transmission reduction.^[9] However, these studies largely focused on periods when social distancing policies were in place, leaving critical questions unanswered regarding how long populations will comply with such measures and what happens when policies are relaxed.

One of the defining characteristics of the COVID-19 pandemic is the need for rapid response in the face of imperfect and incomplete information. Mathematical models of infectious disease transmission can be used in real-time to estimate parameters, such as the effective reproductive number (R_e) and the efficacy of current and future intervention measures, providing time-sensitive data to policy-makers.^[10] We describe development of such a model, in close collaboration with the Colorado Department of Health and Environment and the Governor's office, to gauge the current and future effects of early policies to decrease social contacts and, later, the gradual relaxing of stay-at-home orders.

We developed a compartmental susceptible-exposed-infected-recovered (SEIR) model calibrated to statewide COVID-19 case and hospitalization data to estimate changes in the contact rate and the R_e after emergence of SARS-CoV-2 and the implementation of statewide social distancing policies in Colorado. We supplemented model estimates with an analysis of mobility by using mobile-device location data. Estimates were generated in near real time, at multiple time-points, with a rapidly evolving understanding of SARS-CoV-2. At each time point, we generated projections of the possible course of the outbreak under future social distancing scenarios. Findings were regularly provided to key Colorado decision-makers. We present estimates generated at multiple time points to document how our model, estimates and projections evolved over time. Although our analysis is specific to Colorado, our experience highlights the need for locally calibrated transmission models to inform public health preparedness and policymaking, along with ongoing analyses of the impact of policies to slow the spread of SARS-CoV-2.

1 2 3 4

Next Section

Abstract and Introduction
Methods
Results
Discussion
References

Table 1. Key state-level events, executive orders, and policies directed at controlling transmission of SARS-CoV-2, Colorado, USA, 2020*
Policy/event	Description	Date announced	Policy effective date	Policy effective until
First case of COVID-19	First case of infection with SARS-CoV-2 reported	Mar 5	NA	NA
Executive Order D 2020 003	Disaster emergency	Mar 10	Mar 11	Apr 11
Executive Order D 2020 004	Ski resort closure	Mar 14	Mar 15	Mar 22
Executive Order D 2020 006	Extension of ski resort closure	Mar 18	Mar 18	Apr 17
CDPHE Order 20–22	Closure of bars, restaurants, theaters, gymnasiums, and casinos	Mar 16	Mar 17	Apr 16
CDPHE Order 20–23	Prohibition of >10 person gatherings	Mar 19	Mar 19	Apr 19
Executive Order D 2020 007	School closures during Mar 18–Apr 17	Mar 18	Mar 18	Apr 17
Executive Order D 2020 013	Reduction of in-person workforce by 50%	Mar 22	Mar 24	May10
Executive Order D 2020 017	Stay-at-home order: directive to require all residents of Colorado to stay home unless in pursuant of essential items (i.e., food) or working for critical businesses and ordering noncritical businesses to close temporarily.	Mar 25	Mar 26	Apr 11
Executive Order D 2020 021	Extension to school closures until Apr 30	Apr 1	Apr 1	May 1
Executive Order D 2020 024	Stay-at-Home extension	Apr 6	Apr 6	Apr 26
Executive Order D 2020 039	Ordering workers in critical businesses and government functions to wear nonmedical face coverings	Apr 17	Apr 17	May 17
Executive Order D 2020 041	Suspension of school closures until end of school year	Apr 22	Apr 22	May 20
Executive Order D 2020 044	Safer at home: All susceptible persons and those who have COVID-19 instructed to stay at home. State residents directed to limit interactions, only travel for essential needs, and limit gatherings to <10 people in public and private spaces. Nonmedical mask coverings recommended. Retail businesses can open for curbside delivery, elective medical, dental, and veterinary surgeries and procedures resume. Retail businesses and personal services (e.g., salons) can open. Offices can open at 50% capacity.†	Apr 26	Apr 27–May 4	May 27
Executive Order D 2020 058	Disaster emergency extension	May 7	May 7	Jun 6
Executive Order D 2020 067	Extending EO D 2020 039, ordering workers in critical businesses and government, to wear nonmedical face coverings	May 16	May 16	Jun 16
Executive Order D 2020 079	Extension to EO D 2020 044: Safer at Home to permit public places to offer outdoor dining, and limited indoor dining	May 25	May 25	Jun 1

*COVID-19–relevant executive orders are detailed in (11) and CDPHE policies are described in. (12) CDPHE, Colorado Department of Public Health and Environment; COVID-19, coronavirus disease; EO, executive order; NA, not applicable; SARS-CoV-2, severe acute respiratory syndrome coronavirus 2.
†Retail business and personal services were permitted to open on May 1, offices were permitted to open at 50% capacity on May 4. All other measures went into effect April 27.

Table 2. Model-estimated levels of social distancing, mask wearing and other parameter values at 4 time points over the course of the SARS-CoV-2 epidemic, Colorado, USA, 2020*
Characteristic	Range of possible values and sources (ref.)	Fitted value as of Apr 3† (95% CI)	Fitted value as of Apr 16† (95% CI)	Fitted value as of May 15† (95% CI)	Fitted value as of Jun 16† (95% CI)
Estimated effectiveness of social distancing
Phase 1: early closures, Mar 17–25, %‡	10–70	45 (42–53)	65 (63–72)	52 (49–66)	52 (52–53)
Phase 2: state-wide stay-at-home, Mar 26–Apr 26, %	50–99	NA	76 (72–77)	80 (80–83)	81 (80–82)
Phase 3: half of state under stay-at-home, half transitioned to safer at home, Apr 27–May 8, %	45–99	NA	NA	80 (78–84)	85 (83–90)
Phase 4: statewide safer at home, May 9–Jun 3, %§	NA	NA	NA	NA	90 (85–93)
Proportion of population wearing masks starting Apr 4	0.1–0.8	NA	NA	0.4 (0.11–0.64)	0.40 (0.15–0.76)
Other parameter values
Rate of infection	0.2–0.6 (24)	0.41 (0.39–0.42)	0.50 (0.49–0.51)	0.48 (0.46–0.49)	0.48 (0.48–0.48)
Reduction in infectious contacts due to symptomatic persons who self-isolate after March 5	0.3–0.8 (15)	0.38 (0.22–0.43)	0.47 (0.34–0.50)	0.30 (0.29–0.31)	0.32 (0.32–0.32)
Ratio of infectiousness for symptomatic vs. asymptomatic persons	1.0–4.0 (25,26)	2.27 (2.22–2.29)	1.50 (1.35–1.56)	1.65 (1.60–1.77)	1.68 (1.67–1.69)
Probability that symptomatic cases are identified by state surveillance	0.05–0.6 (24)	0.28 (0.16–0.45)	0.33 (0.27–0.44)	NA	NA

*COVID-19, coronavirus disease; NA, not available; ref, reference; SARS-CoV-2, severe acute respiratory syndrome coronavirus 2.
†Fit 1 uses all reported SARS-CoV-2 cases as of Apr 3 with an onset date of Mar 26 or earlier; fit 2 uses all reported cases as of Apr 16 with an onset date of Apr 8 or earlier; fit 3 uses COVID-19 hospital census through May 15; fit 4 uses hospital census data through Jun 16.
‡For the purpose of model fitting, phase 1 social distancing was modeled by using a start date of Mar 17, which corresponded with the closure of bars, restaurants, casinos, and many public schools in the state.
§The statewide Safer at Home policy remained in effect through the summer, but because of the ≈5.1-d mean incubation period and the 8-d lag between symptom onset and hospitalization, we were able to generate transmission estimates through Jun 3.