The Effectiveness of Face Mask Mandates on COVID-19 Transmission in Utah

Abstract

Objective:

Throughout the COVID-19 pandemic, the effectiveness of face mask mandates was intensely debated. The objective of this study was to describe how face mask mandates at the state, county, and local levels differed in their effectiveness in reducing the number of COVID-19 cases in the jurisdiction where the mandate was implemented and throughout Utah.

Methods:

We used publicly available data from the Utah Department of Health and Human Services. We calculated the effectiveness of face mask mandates (E_Fm) in each local health district after 3 key mandates: the enactment of face mask mandates in Salt Lake and Summit counties (SLSC) on June 28, 2020, and June 27, 2020, respectively; the enactment of a statewide face mask mandate on November 9, 2020; and the lifting of the statewide face mask mandate on April 10, 2021.

Results:

Most counties in Utah had a reduction in the growth rate of COVID-19 cases after enactment of face mask mandates. We found an average 11.9% reduction in E_Fm after the introduction of the SLSC face mask mandates, with 8 of 13 local health districts experiencing a reduction, and an average 12.7% reduction after introduction of the statewide mandate, with 12 of 13 local health districts experiencing a reduction. After mandates were lifted, many counties had an increase in E_Fm.

Conclusion:

Face mask mandates were an effective way to reduce transmission of COVID-19 in local jurisdictions and in neighboring jurisdictions in Utah. Our evidence supports the use of face mask mandates as a way to prevent disease transmission and be better equipped to respond to future pandemics.

When vaccines and other therapeutics are not available, public health decision makers must rely on nonpharmaceutical interventions (NPIs), such as school closures, social distancing, and face mask mandates, as community mitigation strategies to reduce disease transmission and severe outcomes. ¹ During the COVID-19 pandemic, vaccines against SARS-CoV-2 did not become available until December 2020 ² ; consequently, for the first year of the pandemic, all personal and community mitigation strategies were in the form of NPIs.

Although numerous NPIs were considered, face mask mandates became the primary approach to slowing the spread of COVID-19. ³ Face masks had been used as a method to slow disease transmission before the start of the COVID-19 pandemic, especially in countries outside the United States during influenza season. ⁴ However, face masks were the catalyst for intense debates fueled by politics, with states choosing to implement and enforce mandates often based on their political leanings.^5,6

The deployment of face mask mandates in Utah illustrates the fragmentation in the way that face mask mandates were implemented throughout the state (Figure 1). The first mandates were issued in Salt Lake and Summit counties (SLSC), which received approval on June 28, 2020, and June 27, 2020, respectively. The next mandate was issued in the city of Springdale, followed by Grand County on July 7, 2020. As COVID-19 cases began to spike in fall 2020, scattered attempts among local health districts (LHDs) (eFigure1 in the Supplement) dealt with the rising number of cases, creating a mosaic of interventions throughout the state, ranging from recommendations to actual mandates. Throughout this period, special mandates for schools and public transportation were issued.^7,8 A statewide mandate was issued on November 9, 2020, for all counties until April 10, 2021. After the statewide mandate ended, Salt Lake City and Grand County opted to keep their mandates.

Figure 1.

Timeline for COVID-19 face mask mandates, by jurisdiction, Utah, June 2020–June 2021. City and county jurisdictions are indicated in italics. Mandates issued by LHDs, cities, or counties overlapped with the statewide mandate when the statewide mandate was in effect, from November 9, 2020, through April 10, 2021. Abbreviation: LHD, local health district.

Many studies measured the effectiveness of face mask mandates on the spread of airborne diseases and showed face masks to be effective at protecting people from infection.^{9 -20} One method, described by Britton, ⁹ estimated the effectiveness of face mask mandates at the population level by comparing the change in the growth rate of COVID-19 cases before and after a mandate was implemented or lifted.

One benefit of the fragmented system of mandates in Utah is that it created a natural experiment to examine data across LHDs and compare the effectiveness of various interventions in the same state. The objective of this study was to describe how face mask mandates at the state, county, and local levels differed in their effectiveness in reducing the number of COVID-19 cases in the jurisdiction where the mandate was implemented and throughout Utah.

Methods

The data for this study were from the publicly available Utah Department of Health and Human Services COVID-19 surveillance dashboard. ²¹ Our study did not require ethical approval because the data were publicly available and fully deidentified; the University of Utah Institutional Review Board considered the study to be non–human subjects research.

We used the method outlined in Britton ⁹ to calculate the effectiveness of face mask mandates (E_Fm). E_Fm is calculated by comparing the effective reproduction number (R_e) before and after the intervention and is given by the following equation:

$$E_{Fm}=1-\frac{R_{Fm}}{R_{no}}$$

where R_Fm and R_no are calculated by

$$R={(c^{2}r\mu +1)}^{c-2}.$$

In addition, $r$ is calculated before ( $r_{no}$ ) or after ( $r_{Fm}$ ) a face mask mandate is implemented and is given by

$$r_{Fm}=\frac{ln(\frac{i(t+d)}{i(t)})}{d}$$

$$r_{no}=\frac{ln(\frac{i(t)}{i(t-d)})}{d}.$$

In this study, we considered 28 days before and after the face mask mandate to account for the variability in the number of reported cases with respect to day of the week. Because many laboratories in Utah were closed on weekends, tests were not processed during that time, resulting in what looked like a dip in COVID-19 cases during the weekend and a spike on Monday and Tuesday. Our 28-day model accounted for these dips and spikes. We defined all variables and parameter values (Table 1).

Table 1.

Variables, their meaning, and, if applicable, the value used for calculating the effectiveness of face mask mandates in local health districts before and after 2 key face mask mandates were issued and 1 mandate was lifted during the COVID-19 pandemic, Utah, 2020-2021

Variable	Meaning	Value used
EFm	Effectiveness of face mask mandates	Estimated
RFm	Effective reproduction number after implementation of a face mask mandate	Estimated
Rno	Effective reproduction number before implementation of a face mask mandate	Estimated
µ	Mean generation time	6.5 days 22
σ	Standard deviation of the generation time distribution	4 days 22
$c=\frac{\sigma }{\mu }$	Coefficient of variation of the generation time distribution	1.62522
r	Growth rate of cases	Estimated
d	Number of days before or after a mandate that was considered	28 days
rno	Growth rate of cases $d$ days before a face mask mandate is implemented	Estimated
rFm	Growth rate of cases $d$ days after a face mask mandate is implemented	Estimated
i(t)	Incidence at the time of the mandate (t)	Estimated

We calculated the E_Fm for all LHDs in Utah after 3 dates: June 27-28, 2020 (SLSC face mask mandates issued^23,24); November 9, 2020 (statewide face mask mandate issued^25,26); and April 10, 2021 (all mandates lifted except in Salt Lake City and Grand County). ²⁷ Because E_Fm is the ratio of the average effective reproductive number during the 28-day period before and after the mandate, it can be interpreted as the percentage reduction (or increase) in transmission during that period. We conducted all analyses in R (The R Project for Statistical Computing); all code is available on GitHub. ²⁸

Results

Face mask mandates were implemented in Utah when COVID-19 cases were trending upward (Figure 2). After the SLSC face mask mandates were issued on June 27-28, 2020, the Salt Lake County health district had a 22.4% reduction in transmission in the 28 days after the mandate compared with the 28 days before the mandate (Table 2 and Figure 3). Conversely, the Summit County health district had a 6.7% increase in transmission after the SLSC mandates were issued. After the statewide mandate was issued on November 9, 2020, all LHDs in Utah, except the San Juan health district, had reductions in transmission after the mandate was issued. After the statewide mandate was lifted on April 10, 2021, 6 LHDs had a rise in COVID-19 cases; some increases were more gradual than others. After lifting of the statewide mandate, the Southeast Utah health district had the largest increase in transmission (26.2%), and the Salt Lake County health district had the smallest increase in transmission (1.9%).

Figure 2.

Incident cases of COVID-19 by local health district, March 2020–September 2021, Utah, in (A) SLSC and (B) all health districts in Utah. Abbreviation: SLSC, Salt Lake and Summit counties.

Table 2.

Effectiveness of face mask mandates (E_Fm) during the COVID-19 pandemic, by local health districts, Utah, 2020-2021 ^a

Local health district	EFm for SLSC mandate, June 27-28, 2020	EFm for statewide mandate, November 9, 2020	EFm for lifting of statewide mandate, April 10, 2021
Salt Lake County b	–22.4	–17.2	1.9
Summit County	6.7	–14.2	–2.9
Wasatch County	22.0	–7.0	–17.8
Weber–Morgan	–30.2	–16.2	–17.1
TriCounty	–9.9	–30.8	9.5
Southwest Utah	–24.3	–7.2	–1.4
Southeast Utah	6.9	–17.4	26.2
San Juan	104.5	11.2	–11.0
Central Utah	–42.4	–2.4	15.6
Tooele County	9.1	–14.4	–0.04
Davis County	–24.6	–12.5	–5.7
Utah County	–17.1	–12.7	11.0
Bear River	–11.9	–9.3	14.4

Abbreviation: SLSC, Salt Lake and Summit counties.

^aData source: Utah Department of Health and Human Services COVID-19 surveillance dashboard. ²¹ All values are percentages. To interpret percentages, for example, after the SLSC face mask mandate was issued on June 28, 2020, the Salt Lake County health district had a 22.4% reduction in R_e (the effective reproduction number) in the 28 days after the mandate compared with the 28 days preceding the mandate.

^bSalt Lake City still had a mandate, but the county as a whole did not.

Figure 3.

The effectiveness of face mask mandates (E_Fm) on the transmission of COVID-19 in Utah after implementation or lifting of face mask mandates, where a negative E_Fm indicates a reduction in transmission and a positive E_Fm indicates an increase in transmission. I-15 runs north–south through Utah for 400 miles. (A) After implementation of face mask mandates in Salt Lake and Summit counties on June 27-28, 2020. (B) After implementation of the statewide mandate on November 9, 2020. (C) After lifting the statewide mandate on April 10, 2021. (D) Identification of local health districts in Utah.

In addition to face mask mandates affecting the jurisdictions in which they were issued, the SLSC mandates affected other LHDs across the state (Table 2 and Figure 3). The largest effect was in the Central Utah health district, which had a 42.4% reduction in transmission, but all LHDs along the interstate highway I-15 corridor had a reduction in transmission (Figure 3). However, not all LHDs had a reduction in transmission after the SLSC face mask mandates were issued: 5 counties had an increase in transmission, with the San Juan health district having the greatest increase (104.5%) (Table 2).

Discussion

Overall, we found that most LHDs in Utah had a reduction in transmission of COVID-19 after face mask mandates were implemented. We found the most comprehensive reductions after the statewide mandate was implemented, with all health districts except San Juan having some decrease. After the statewide mandate was lifted, most LHDs had an increase in COVID-19 transmission. Although the statewide mandate was lifted in April 2020, Salt Lake City and Grand County maintained a citywide face mask mandate until 2022, which may partially explain why the Salt Lake County health district had only a modest increase in transmission.

Although it is not surprising that face mask mandates resulted in reductions in COVID-19 transmission in the jurisdiction where they were issued, it was surprising that many LHDs had a reduction in transmission after a face mask mandate was issued in a different jurisdiction. In particular, we found that all LHDs along the I-15 corridor had a reduction in transmission after the SLSC face mask mandates were issued. This finding suggests that face mask mandates in large population hubs such as Salt Lake County can have far-reaching effects and that people may change their behavioral patterns in response to face mask mandates in other jurisdictions and rising case counts.

Another interesting finding was that the largest reductions in COVID-19 transmission occurred after implementation of the SLSC mandates, with most LHDs having a larger decrease after these mandates were issued than after the statewide mandate was issued. The statewide mandate reduced transmission in all LHDs (except one), including counties that did not have a decrease after the SLSC mandates. One potential explanation for this reduced effect could be that counties had already begun to apply their own face mask restrictions in the weeks leading up to the statewide mandate or that people had already started to change their own behavior in the absence of mandates and as cases of COVID-19 were increasing.

Exploring patterns in counties that were not affected by face mask mandates can also be illuminating. We found an increase in COVID-19 transmission in Summit County after the SLSC mandates were issued, which could be attributed to low case counts when the mandate was issued: an average of 39.1 cases per week before the mandate and 112.7 cases per week after the mandate. In addition, the San Juan health district had a 104.5% increase in transmission after the SLSC mandate and an 11.2% increase in transmission after the statewide mandate. One potential explanation for these findings is that 21% of the San Juan health district’s population self-identifies as Navajo, ²⁹ and the Navajo nation was experiencing a major outbreak of COVID-19 during this time.^29,30

Our findings are particularly interesting when considered in the broad context of face mask mandates. Seegert et al ³¹ explored the economic effects of face mask mandates and found that statewide mandates stimulate economic activity while county-wide mandates depress activity, suggesting that this finding could be a result of risk perception. ³¹ Taken together with our findings that face mask mandates generally reduced transmission in the jurisdiction where it was issued as well as in jurisdictions connected by highways, we suggest that in future disease outbreaks, swift statewide action may be equally effective at reducing disease transmission and more effective at reducing economic harms than a piecemeal, local approach.

Several studies have attempted to estimate the effectiveness of face mask mandates during the COVID-19 pandemic.^{9 -20} One randomized controlled trial concluded that wearing face masks reduced the risk of SARS-CoV-2 infection for the wearer by 18%, while another study found that symptomatic incidence was reduced by 9.5%.^10,11 Observational studies that directly estimated the effect of face mask mandates on transmission estimated a 29% reduction in Kansas,^9,15 a 21% to 43% reduction in Jena, Germany,^9,17 and a 15.1% reduction when averaged across 190 countries and adjusted for confounders. ¹⁶ Furthermore, a simulation-based study found that the use of face masks equated to an average 25% to 35% reduction in infectious contacts. ¹³ Overall, our study showed a similar median effectiveness: a 12.7% reduction across all LHDs (a 13.5% reduction when excluding the San Juan health district) after implementation of the statewide face mask mandate and an 11.9% reduction across all LHDs (a 7.9% reduction when considering only LHDs with face mask mandates) after implementation of the SLSC mandates.

Limitations

Our study had several limitations. First, we used government-issued face mask mandates as a proxy for face mask wearing. Unfortunately, the relationship between face mask mandates and face mask use is weak, with 1 study showing no significant relationships and other studies showing a 13% to 23% change in self-reported face mask use following a face mask mandate in the United States.^18,19,32 Consequently, just because face mask mandates were in place does not mean that the population fully complied. Many face mask mandates were designed to be lightly enforced, and the tense political climate surrounding face mask use reduced the willingness to enforce face mask mandates, potentially decreasing the effectiveness of those mandates. ³³ Furthermore, some people may have worn a face mask regardless of mandate status. Our study does not attempt to link actual face mask use to the implementation of face mask mandates.

Second, our study did not attempt to provide evidence on the individual effectiveness of using a face mask. We examined the effectiveness of face mask mandates at the population level. Although face mask mandates dictate that individuals must wear a face mask, many people used ineffective face masks. ³⁴

The third limitation of our study was the timing of COVID-19 vaccine availability. Vaccines first became available to health care workers in Utah in December 2020 and became widely available to the general public by March 2021.^3,35,36 Given the timing of the statewide face mask mandate on November 9, 2020, we do think that the availability of vaccines affected the estimates of the effect of lifting the face mask mandates. Because widespread access to COVID-19 vaccination began in March 2021, vaccination availability may have affected estimates of the effectiveness of lifting the statewide face mask mandates (in April 2021). We believe that this effect was limited because only 32% of the Utah population was fully vaccinated by May 8, 2021, ³⁷ the end of our study time frame.

Conclusion

Our study shows that the implementation of face mask mandates in Utah played a role in reducing COVID-19 transmission throughout the state. We found that implementing and lifting face mask mandates influenced transmission in jurisdictions where the mandate was issued as well as in connected jurisdictions. The COVID-19 pandemic is unlikely to be the last time we are faced with the transmission of an airborne, viral pathogen; consequently, understanding the most effective way to reduce transmission is essential to inform future pandemic response.