Racial and Ethnic Disparities in Preventive Service Use Among Adults Before and During the COVID-19 Pandemic

Abstract

Our objective was to assess changes in preventive services use before and during the COVID-19 pandemic. We obtained secondary survey data from the National Health Interview Survey for 2019 and 2021. We examined, six preventive services among all adults. Descriptive and multivariate analyses assessed changes in preventive service use among adults and by race/ethnicity for 2019 and 2021 (drawing from an unweighted sample of 60 843 weighted to be 386.2 million across both years). We used Ordinary least squares estimation to conduct a difference-in-differences analysis that assessed changes in service use for non-white racial/ethnic groups relative to changes for white non-Hispanic adults between 2019 and 2021. We found preventive services use declined overall for each screening service assessed. Asian adults experienced the largest declines relative to white adults for “well visit within the last year” (−7.45 percentage points (pp) relative to white adults), “blood pressure screening within the last year” (−7.85 pp), and “mammograms within the last year” (−12.3 pp). While adults in other racial/ethnic groups did not experience significant declines in preventive services use relative to white adults between 2019 and 2021, pre-existing disparities remained for Hispanic and American Indian/Alaska Native (AIAN) adults compared to white adults. In conclusion, preventive service use declined in the first years of the COVID-19 public health emergency, and existing disparities in access for Hispanic and AIAN adults continued. Future research should investigate barriers Asian adults may face in obtaining access to preventive services after the conclusion of the public health emergency and federal pandemic-related protections.

Introduction

The COVID-19 pandemic significantly disrupted health care utilization patterns in the United States. Following the public health emergency declaration on January 31, 2020 and the National Emergency declaration on March 13, 2020, the US Centers for Disease Control and Prevention (CDC) recommended health care providers defer non-emergency in-person care in order to manage health system capacity and minimize risks to patients and their communities. ¹ While disruptions to in-person care were partially offset by a rapid increase in telehealth services,^{2 -5} some health care procedures, including many types of preventive screening services, could not be provided via telehealth. Even as providers reopened their doors to non-emergency care later in 2020, rates of service use did not immediately return to pre-pandemic levels for some types of health care. ⁶

Systemic racism and other structural barriers to health produce racial and ethnic, as well as economic, inequities within the health care system and across all aspects of people’s lives–inequities that were exacerbated by the pandemic.^7,8 As a result, people of color may have experienced more disruption to their health care than non-Hispanic white people during the pandemic. For example, because of greater exposure to COVID-19 at work^{9 -11} and differential or discriminatory treatment by the medical system, ¹² about 30% of Black and Hispanic adults had a relative or someone close to them who died from COVID-19, ¹³ —more than double the share of white adults—which may have led to stronger concerns among people of color about contracting COVID-19 during health care visits. ¹⁴ Moreover, non-Hispanic Asian and Hispanic adults were more likely than non-Hispanic white adults to lose their employer health insurance coverage during the first years of the pandemic. ¹⁵ Communities of color were also disproportionately impacted by public transportation disruptions, creating potential access problems. ¹⁶ For people who are not fluent in English, the inability to bring another person to health care visits during the pandemic may have been an additional barrier. ¹⁷ And finally, racist anti-Asian rhetoric may have contributed to fears of seeking health care among Asian Americans; by early 2021, 80% of Asian Americans said violence against them in the US is increasing, 45% experienced an incident tied to their racial and ethnic background since the pandemic began, and 32% feared someone might physically attack them. ¹⁸

Prior to the pandemic, use of preventive screening services varied considerably across adults of different racial and ethnic groups, patterns influenced by a wide range of structural factors. Non-Hispanic Black adults generally had the highest rates of preventive screenings, or rates comparable to non-Hispanic White adults, for well visits, chronic disease, and cancer.^19,20 Hispanic and American Indian/Alaska Native adults had much lower rates of preventive screenings than non-Hispanic White adults, in part due to lower rates of insurance coverage, but buffered, in part, for American Indian/Alaska Natives who have access to the Indian Health Services.^19,21 Asian adults had higher rates of screening for some types of services and lower rates for others compared to non-Hispanic white adults.^16,22

At the same time, prior to the pandemic, people of color were more likely than non-Hispanic white adults to receive advanced-stage breast cancer diagnoses, contributing to higher rates of breast cancer morbidity and mortality.^{23 -26} There were also significant racial and ethnic disparities in colorectal cancer mortality rates, especially for Black individuals.^27,28 These disparities are likely driven in part by systemic barriers such as discrimination and inequitable access to resources such as insurance coverage and high quality medical care.^{29 -33}

Evidence from early in the pandemic indicates that Black adults under age 65 were more likely to forgo or delay needed care compared to white and Hispanic adults under age 65 due to both fear of contracting COVID and their health care providers remaining closed. ³⁴ Another study suggested that Asian adults were less likely than non-Hispanic white adults to use telehealth during the first year of the pandemic, while Black and Hispanic adults were more likely to use such care. ³⁵ If similar patterns occurred for preventive screenings, the pandemic may have increased pre-existing disparities in health and health care treatment.^36,37

In this study, we use data from the National Health Interview Survey (NHIS) to examine the extent to which well visits, preventive screenings for hypertension and diabetes, and preventive cancer screenings recommended by the U.S. Preventive Services Task Force (USPSTF) changed between 2019 and 2021. We estimate the changes between 2019 and 2021, overall and separately for people of different racial and ethnic groups, in (1) having a screening in the past year and (2) being up-to-date on each screening. The implications of our study are important because delayed or missed screenings can lead to more advanced stage chronic disease and cancer diagnoses, with potentially worse health outcomes as a result.^{38 -40}

While prior research has documented the decline in preventive services since the onset of COVID-19, to our knowledge, none of these studies provide a comprehensive assessment of the impact of the pandemic on use of preventive screenings services for people of different racial and ethnic groups and whether individuals stayed up-to-date on these screenings in general.^41,42 Some prior studies suggest differences across racial and ethnic groups, but the literature is generally limited by region of study, sample size, or lack of available information on the racial and ethnic composition of those receiving care.^{43 -49} Given this gap in the literature, it is not known to what extent preventive care use changed differentially for patients of different races and ethnicities over the course of the pandemic.

Methods

Data

We use the 2019 and 2021 National Health Interview Survey (NHIS) to examine the extent to which preventive service use changed during the early years of the pandemic. We focus on adult well visits and general preventive screenings for hypertension (blood pressure screenings) and diabetes typically provided during a well visit, as well as cervical, breast, and colon cancer screenings recommended by USPSTF. The NHIS asks respondents whether they have ever had certain preventive screenings and, if they did, when the most recent screening occurred. This allows us to estimate (1) whether individuals had each screening in the past year; and (2) whether they are up-to-date on screenings based on the USPSTF recommendations. Well visits and blood pressure screenings are calculated among all adults (18 and older), whereas all other measures for screenings in the past year are calculated among the populations recommended for each screening. Table 1 describes the recommendations, the populations for which the recommendations apply, and how we define “up-to-date” for those populations using sex, age, and other risk factors identifiable on the NHIS.

Table 1.

USPSTF Recommendations and Operationalization with 2019 and 2021 NHIS.

Screening	Recommendation	Operationalization of up-to-date on screening using NHIS	Notes
General preventive screening
Well Visit	No current recommendation.	N/A
Chronic condition screening
Hypertension	Adults aged 40 years or older and those who are at increased risk for high blood pressure should be screened annually. Persons at increased risk include those who have high-normal blood pressure (130-139/85-89 mm Hg), those who are overweight or obese, and Black adults.	Adults age 40 and older, all adults who are overweight or obese, Black adults: “up-to-date” if screening for high blood pressure in the past year.	Notes: We operationalized the recommendations for hypertension that include all Black adults in the high risk category for screening. We note, however, that the risk factor is racism and not race.
	Adults aged 18 to 39 years with normal blood pressure (<130/85 mm Hg) who do not have other risk factors should be rescreened every 3 to 5 years.	Adults age 18 to 39 who are not overweight or obese and are not black: “up-to-date” if screening for high blood pressure in the past 3 years
Pre-diabetes and diabetes screening	Adults age 35 to 70: Screen every 3 years for prediabetes and type 2 diabetes in adults aged 35 to 70 years who have overweight or obesity.	Adults age 35 to 70 who are overweight or obese and not diabetic: “up-to-date” if diabetes screening in the past 3 years
Cancer screening
Breast cancer screening	Women age 50 to 74: biennial mammograms. Women age 40 to 49: individual choice of biennial mammograms.	Women age 50 to 74: “up-to-date” if mammogram in the past 2 years
Cervical cancer screening	Women age 21 to 29: screening every 3 years with cervical cytology alone Women age 30 to 65: screening every 3 years with cervical cytology alone, OR every 5 years with human papillomavirus (HPV) testing alone, OR every 5 years with HPV testing in combination with cytology.	Women age 21 to 29: “up-to-date” if screened in the past 3 years Women 30 to 65: “up-to-date” if screened in the past 5 years with HPV; if no HPV testing in last 5 years, “up-to-date” if screened in the past 3 years.	Notes: Women who have had a hysterectomy are removed from the sample. While cervical cytology (pap smears) are exclusively recommended for 21 to 29, we include as up-to-date the small share of women 21 to 29 who had another kind of cervical cancer screening in the last 3 years.
Colon cancer screening	Adults age 40 to 85: High-sensitivity gFOBT or FIT every year; or sDNA-FIT every 1 to 3 years; or CT colonography every 5 years; or flexible sigmoidoscopy every 5 years; or flexible sigmoidoscopy every 10 years + FIT every year; or Colonoscopy screening every 10 years	Adults age 40 to 85: “up-to-date” if screened with: (1) colonoscopy in the last 10 years, or (2) sigmoidoscopy in the past 5 years.	Notes: Changes in the NHIS coding prevented us from incorporating CT colonography, DNA fit, and at-home FIT tests into our measure of up-to-date.

Note. Recommendations derived from the published recommendations of the U.S. Preventive Services Task Force: https://www.uspreventiveservicestaskforce.org/uspstf/topic_search_results?topic_status=P.

Using the 7 race and ethnicity groups available on the publicly use NHIS files, we create the following mutually exclusive race and ethnicity categories: Hispanic, non-Hispanic white only (white), non-Hispanic Black only (Black), Non-Hispanic Asian only (Asian), American Indian/Alaska Native (AIAN), and non-Hispanic “Other and Multiple” race (other and multiple race). We combine individuals who report AIAN alone or in combination with any other group into one group due small sample size concerns. We do not present certain estimates for AIAN and other and multiple race adults for screenings where there are fewer than 250 respondents. ⁱ Adults in these groups are included in the overall estimates to produce national estimates of changes in service use. Sample sizes for each preventive screening and race ethnicity group are presented in Supplemental Appendix Table 1.

Due to the sampling structure of the NHIS, for adults responding to the 2019 NHIS, the screenings they report on taking place “in the last year” could have occurred in first quarter of 2018 through the last quarter of 2019, depending on the quarter in which the survey was administered. Similarly, for adults responding to the 2021 NHIS, these screenings could have occurred in the first quarter of 2020 through the last quarter of 2021. Therefore, estimates from the 2021 survey year mostly, but not exclusively, represent the pandemic period because the public health emergency was declared on January 31, 2020 and the national emergency was declared on March 13, 2020. A sensitivity analysis excluding data collected during January to March 2019 and January to March 2021—therefore ensuring all data from the 2021 sample represent the pandemic period—showed results consistent with full-year data. The estimates we produce also represent a mix of impacts that occurred both before and after introduction of COVID-19 vaccines, which became widely available in the second quarter of 2021.

Analysis

We use descriptive and multivariate analyses to estimate the use of preventive services and being up-to-date on preventive services in 2019 (prior to the pandemic) and in 2021 overall and by race and ethnicity; changes experienced by adults in each racial and ethnic group from 2019 to 2021; and whether changes over this period for non-White adults differ from changes for adults of each of the other race and ethnicity groups (the difference-in-differences). We choose to examine differences between white adults and adults of other races and ethnicities not because white adults have the highest level of preventive care use, they do not, but rather because we believe that access to care for people of color would have been more impacted by the pandemic due to the many attributes of structural racism discussed in the introduction.

To examine changes over time within and across racial and ethnic groups, we use ordinary least squares (OLS) to estimate the following difference-in-differences model for each preventive service:

$$\begin{array}{c}Servic{e}_{it}={\beta }_0+{\beta }_{1}Year{21}_t+{\beta }_{2}Rac{e}_i\\ +{\beta }_3(Year{21}_t*Rac{e}_i)+{X^{\prime }}_i{\beta }_4+{\epsilon }_{it}\end{array}$$

In this equation, $Servic{e}_{it}$ indicates whether individual $i$ received the service (or was up-to-date on the service) in survey year $t$ . $Year{21}_t$ is an indicator for survey year 2021 and $Rac{e}_i$ denotes a set of dummy variables for each of the 6 racial and ethnic groups, each of which is interacted with $Year{21}_t$ . The key coefficients are the ${\beta }_3$ ’s which represent the differential change in the probability of service use between the 2019 and 2021 surveys for adults in each of the race/ethnicity categories relative to non-Hispanic white adults (the difference-in-differences estimate). We estimate this model with and without additional individual-level covariates denoted by $X$ , including age (13 5-year age categories from 19 to 85), and sex (male and female categories). We include these factors because at a population-level, they were more or less fixed over the course of the pandemic, unlike income and insurance coverage which were directly impacted by the pandemic. ⁱⁱ Consequently, our estimates reflect the overall change in preventive service use across racial and ethnic groups from 2019 to 2021, which could be partially attributable to changes in insurance coverage and income over this period. Estimates are produced using STATA 18 and final annual weights (WTFA-A) for sample adults are used to account for the complex survey design, and standard errors are estimated using Taylor linearization.

Results

Table 2 presents results for our analysis of preventive care screening in the past year and includes unadjusted rates of screenings for 2019 and 2021, adjusted differences between 2019 and 2021 for all race and ethnicity groups combined and for each race and ethnicity group separately, and unadjusted and adjusted difference-in-differences estimates comparing changes for white adults relative to adults of each of the other race and ethnicity groups. Results for being up-to-date on preventive screening visits are presented in Table 3. While we focus on the adjusted results in the narrative, Supplemental Appendix Table 2 presents adjusted and unadjusted changes for all outcomes, and the patterns are similar.

Table 2.

Use of Preventive Services Within the Last Year by Race and Ethnicity, NHIS 2019 and 2021.

	2019 Survey year	2021 Survey year	Adjusted percentage point difference between 2019 and 2021	Difference-in-differences (relative to change for white adults)
					Unadjusted		Adjusted
Wellness visit in last year
All	78.9%	76.2%	–2.93	**
Hispanic	72.2%	69.5%	–2.89	**	–0.44	**	–0.47
White	79.6%	77.4%	–2.42	**
Black	84.4%	82.5%	–2.38	**	0.34	**	0.04
Asian	81.3%	71.6%	–9.88	**	–7.43	**	–7.45	**
AIAN only and oth/mult	77.3%	76.6%	–0.98		1.49	**	1.44
Other and multiple	68.9%	66.6%	–0.79		–0.04		1.64
Diabetes screening within last year
All	73.6%	69.5%	–4.26	**
Hispanic	71.2%	65.3%	–5.75	**	–2.00	**	–1.72
White	72.8%	68.9%	–4.03	**
Black	81.6%	78.8%	–3.21		1.12	**	0.81
Asian	75.8%	68.5%	–7.28	*	–3.44	**	–3.25
Blood pressure screening within last year
All	88.1%	84.0%	–4.27	**
Hispanic	81.9%	76.5%	–5.56	**	–2.02	**	–2.06	*
White	89.4%	86.1%	–3.50	**	0.00
Black	90.4%	88.1%	–2.67	**	1.09	**	0.84
Asian	87.6%	76.4%	–11.35	**	–7.83	**	–7.85	**
AIAN only and oth/mult	89.9%	83.0%	–7.05	*	–3.48	**	–3.55
Other and multiple	81.1%	79.4%	–0.64		1.66	**	2.86
Pap test within last year, women 21 to 65
All	46.5%	40.1%	–6.36	**
Hispanic	43.3%	37.7%	–5.79	**	0.75	**	0.57
White	46.7%	40.3%	–6.36	**	0.00
Black	52.4%	47.4%	–4.85	*	1.35	**	1.51
Asian	43.0%	32.3%	–10.35	**	–4.31	**	–3.99
Mammogram within last year, women 50 to 74
All	59.9%	57.1%	–2.85	**
Hispanic	61.0%	55.0%	–5.96	**	–4.46	**	–4.36
White	59.3%	57.8%	–1.60		0.00
Black	63.2%	61.6%	–1.55		–0.06	*	0.05
Asian	62.2%	48.3%	–13.91	**	–12.32	**	–12.32	**
Colon cancer screening within last year, 40 to 85
All	12.1%	10.9%	–1.21	**
Hispanic	9.5%	8.7%	–0.61		0.22	**	0.55
White	12.2%	11.1%	–1.16	**	0.00
Black	16.4%	14.1%	–2.67	**	–1.17	**	–1.50
Asian	9.0%	8.3%	–0.89		0.44	**	0.27
AIAN only and oth/mult	14.5%	13.7%	–0.82		0.36	**	0.34

Source. National Health Interview Survey 2019 and 2021.

Note. **/*indicates difference is significant at the .05/.10 level.

Table 3.

Up-to-Date Use of Preventive Services by Race and Ethnicity, NHIS 2019 and 2021.

	2019 Survey year	2021 Survey year	Adjusted percentage point difference between 2019 and 2021	Difference-in-differences (relative to change for white adults)
					Unadjusted		Adjusted
Diabetes screening up-to-date
All	88.3%	88.0%	−0.36
Hispanic	87.7%	85.8%	−1.81		−1.79	**	−1.56
White	88.0%	87.8%	−0.24
Black	91.4%	92.2%	0.44		0.85	**	0.68
Asian	88.0%	90.1%	2.05		2.19	**	2.29
Blood pressure screening up-to-date
All	89.8%	86.5%	−3.42	**
Hispanic	84.6%	80.0%	−4.67	**	−1.90	**	−1.85	*
White	91.1%	88.3%	−2.82	**
Black	90.4%	88.1%	−2.54	**	0.44	**	0.28
Asian	90.6%	82.8%	−7.96	**	−5.12	**	−5.14	**
AIAN only and oth/mult	91.2%	85.2%	−6.19		−3.27	**	−3.37
Other and multiple	84.7%	86.1%	2.24		4.14	**	5.06
Pap test up-to-date, women 21 to 65
All	76.3%	74.6%	−1.71	**
Hispanic	70.2%	68.5%	−2.15		−1.34	**	−1.71
White	79.0%	78.6%	−0.44		0.00
Black	78.3%	73.4%	−4.65	**	−4.56	**	−4.20	*
Asian	68.2%	63.5%	−4.42		−4.35	**	−3.97
Mammogram up-to-date, women 50-74
All	76.5%	76.0%	−0.47
Hispanic	78.4%	74.0%	−4.37	*	−4.56	**	−4.49
White	76.2%	76.3%	0.12		0.00
Black	79.1%	82.1%	3.18		2.91	**	3.06
Asian	73.9%	67.3%	−6.48		−6.70	**	−6.60
Colon cancer screening up-to-date, 40 to 85
All	51.5%	52.2%	0.35
Hispanic	34.7%	36.5%	2.61	*	1.45	**	2.67	*
White	56.3%	56.6%	−0.06		0.00
Black	51.2%	53.3%	0.75		1.72	**	0.81
Asian	37.3%	40.1%	1.82		2.43	**	1.89
AIAN only and oth/mult	45.1%	46.1%	1.56		0.67	**	1.62

Source. National Health Interview Survey 2019 and 2021.

Note. **/* indicates difference is significant at the .05/.10 level.

Well Visits

In 2019, 78.9% of all adults reported having had a well visit in the past year, with rates varying by race and ethnicity. Black adults had the highest rate of well visit receipt (84.4%), while adults of other and multiple races and Hispanic adults had the lowest (68.9% and 72.2%, respectively). From 2019 to 2021 the overall rate of having a well visit in the past year declined to 76.2%, a significant reduction of 2.9 percentage points (P < .05). While all race and ethnicity groups experienced a reduction between 2019 and 2021, the magnitude and significance of the reduction varied by race and ethnicity. The decline for Asian adults (9.9 percentage points) was >3 times larger than the declines among Hispanic adults (2.9 percentage points), white adults (2.4 percentage points), and Black adults (2.4 percentage points). The declines among AIAN adults, and adults reporting other or multiple races were smaller and not statistically significant.

The changes between the 2019 and 2021 survey years for race/ethnicity groups relative to the change for white adults (ie, the difference-in-differences estimates) were not statistically significant for Black, Hispanic, AIAN, or “other and multiple race” adults. However, Asian adults experienced a significant decline in in screening relative to white adults, an adjusted difference-in-differences estimate of −7.5 percentage points (P < .05).

Chronic Disease Screening

Diabetes and blood pressure screenings also declined during the pandemic across all racial and ethnic groups. In 2019, 73.6% of adults who were recommended for screening by the USPSTF (ie, ages 35-70, overweight or obese, and not previously diagnosed with diabetes) reported having been screened for diabetes within the last year. Eighty-eight percent of adults recommended for blood pressure screening (ie, age 40 or older or under age 40 but at high risk of high blood pressure) had their blood pressure screened within the last year. Black adults recommended for screening had the highest rates of receipt for both diabetes and blood pressure screenings in the last year in 2019 (81.6% and 90.4% respectively), while Hispanic adults recommended for screening had the lowest for both (71.2% and 81.9% respectively).

Between 2019 and 2021, the rates of diabetes screening declined from 73.6% to 69.5%, a significant reduction of 4.3 percentage points. Compared to white adults (4.0 percentage points), Hispanic adults saw a sharper decline of 5.8 percentage points and Asian adults experienced the largest decrease in diabetes screenings (7.3 percentage points). No significant changes for diabetes screening in the last year were observed between 2019 and 2021 for Black adults. Blood pressure screenings declined from 88.1% to 84.0% among all adults during the same period, a significant reduction of 4.3 percentage points. Similar to declines in well visits and diabetes screenings, Asian adults experienced the largest decline (11.4 percentage points), while white and Hispanic adults experienced declines similar to the overall reduction (3.5 percentage points and 5.6 percentage points, respectively), and Black adults experienced the smallest decline (2.7 percentage points).

The difference-in differences estimate between the 2019 and 2021 surveys for diabetes screenings were not significantly different from white adults for Black, Hispanic, or Asian adults after adjusting for age and sex. For blood pressure screening, the difference-in-differences estimates were not significant between white adults and any other race or ethnicity category except for Asian adults, who experienced a 7.9 percentage point decline in blood pressure screening relative to white adults between survey years 2019 and 2021.

Cancer Screening

Declines in rates of cancer screening varied by the type of screening and by race and ethnicity. In 2019, 46.5% of women 21 to 65 reported receiving a cervical cancer screening within the last year, 59.9% of women 50 to 74 reported receiving a breast cancer screening through mammography within the last year, and 12.1% of adults 40 to 85 reported receiving a colon cancer screening within the last year. Black adults had the highest rates of cancer screenings within the past year for all 3 measures in survey year 2019; 52.4%, 63.2%, and 16.4% for cervical cancer screening, breast cancer screening, and colon cancer screening, respectively. Hispanic women 21 to 65 had the lowest rate for cervical cancer screening within the last year (43.3%), white women 50 to 74 had the lowest rate for mammogram within the last year (59.3%), and Asian adults 40 to 85 had the lowest rate for colon cancer screening within the last year (9.0%).

Between 2019 and 2021, the rate of women reporting they had received a cervical cancer screening within the last year declined from 46.5% to 40.1%, a 6.4 percentage point decline. White, Black, and Hispanic women experienced similar declines in receipt of a cervical cancer screening in the last year between 2019 and 2021 (6.4 percentage points, 4.9 percentage points, and 5.8 percentage points, respectively). Asian women experienced the largest decline in receipt of a cervical cancer screening in the last year: a 10.4 percentage point drop, representing a 25% decline from survey year 2019.

For mammograms, the rate of women who had a mammogram in the last year declined slightly from 59.9% to 57.1%, a 2.9 percentage point decline. Among race and ethnicity subgroups, Asian women experienced a significant decline between 2019 and 2021, a 13.9 percentage point drop, while Hispanic women experienced a 6.0 percentage point decline in the same period. Mammogram receipt for white women and Black women did not change significantly between 2019 and 2021 when adjusting for age and sex.

The rate of adults who had a colon cancer screening within the last year dropped from 12.1% to 10.9%, a 1.2 percentage point decline. For colon cancer screenings, only white adults and Black adults experienced a significant decline of 1.2 percentage points and 2.7 percentage points respectively between 2019 and 2021.

The difference-in-differences estimates show that the rates of cervical cancer screening and colon cancer screening for nonwhite race/ethnicity groups declined at similar rates relative to white adults between 2019 and 2021. For mammograms, the only nonwhite race/ethnicity group that did not experience declines at similar rates to white women were Asian women, who experienced a 12.3 percentage point decline in receipt of a mammogram within the last year relative to white women between 2019 and 2021, significant at the P < .05 level.

Up-to-Date Use of Preventive Services

Compared to rates of preventive screenings reported within the last year, rates of being up-to-date on preventive screenings showed little variation between the pre-pandemic and mid-pandemic period (Table 3). Notable differences include blood pressure screenings and cervical cancer screenings. Being up-to-date on screenings for blood pressure dropped significantly between 2019 and 2021, with an overall decrease of 3.4 percentage points. All racial and ethnic groups, except for AIAN adults and adults of other races, experienced significant declines in up-to-date blood pressure screenings between 2019 and 2021. The highest decline was among Asian adults, with an 8.0 percentage point decline, representing a 5.1 percentage point drop relative to white adults. Being up-to-date on cervical cancer screenings declined slightly overall, from 76.3% in 2019 to 74.6% in 2021, an adjusted decrease of 1.7 percentage points. This appears to be driven largely by the decline experienced by Black women in being up-to-date on cervical cancer screenings: they experienced a 4.7 percentage point decline between 2019 and 2021, when no other racial and ethnic groups experienced significant changes in cervical cancer screenings. Relative to white women in the same period, Black women experienced a 4.2 percentage point decline (significant at the P < .10 level).

Finally, to better understand the impact of the NHIS interviewing strategy on our estimates, we compare changes in outcomes between the 2 years by interview quarter (Supplemental Appendix Table 3). These estimates indicate that the largest declines in preventive service use occurred for individuals surveyed in the first quarter of 2021 compared to those surveyed in later quarters. These individuals surveyed in the first quarter were reporting on their use of services in a period prior to widespread availability of the COVID-19 vaccine.

Discussion

We provide new evidence of considerable declines in preventive care use among adults from 2019 to 2021, that is, before and after the start of the COVID-19 pandemic, using nationally representative survey data. The declines in preventive care use occurred for all adults, but Asian adults experienced significantly larger declines compared to white adults in several of the services we examined, including well visits, blood pressure screening, and mammograms. For example, Asian adults experienced a decline in well visits rates that was 7.5 percentage points greater than the decline for white adults (9.2% relative to baseline), and a decline in mammogram rates that was 12.3 percentage point greater (19.8% relative to baseline).

These findings are consistent with prior studies documenting declines in outpatient service use during the pandemic and evidence of racial disparities in these declines.^41,42,50 We add to this literature by focusing specifically on preventive care use and by examining changes in preventive service use by race and ethnicity in a nationally representative sample of US adults. These broad declines in preventive care use were likely driven by several factors. First, many health care providers suspended all non-emergency in-person services during the early months of the pandemic to reduce the spread of the virus, and many adults may have missed a scheduled visit during this time. Second, even after in-person care became fully available again, people may have chosen to forgo routine care to minimize potential exposure to the virus, especially prior to widespread distribution of the vaccine. While telehealth was rapidly adopted as a substitute for some in-person care and especially mental health and primary care services, ⁵¹ other preventive services such as mammograms and colonoscopies could not have been provided via telehealth. Third, pandemic-related income loss or health coverage changes may have reduced people’s ability to pay for routine health services or disrupted access to their preferred provider. Finally, people may have experienced ongoing or new access barriers, such as increased barriers to transportation or childcare, during the pandemic.

Except for Asian adults compared to white adults, we did not find significant differences in changes in access to preventive care by race and ethnicity. While this suggests the pandemic did not significantly worsen racial disparities in preventive care access for Black, AIAN, Hispanic, and multiracial adults relative to white adults, addressing pre-existing disparities remains an important priority for health equity. Understanding the reasons these disparities, which may derive from a host of factors including access to health insurance and health care and educational and cultural beliefs, is key to devising effective strategies for their reduction.

It will also be important to understand why Asian adults experienced significantly larger declines in preventive care during the pandemic (for instance, by investigating impacts of attitudes toward COVID-19 and fear of violence on access to care and health inequities ¹⁸ ) and to ensure Asian adults are able to receive missed care in a timely manner to avoid delayed diagnoses of health conditions that could lead to worse long-term health outcomes. People who identify as Asian represent an extremely diverse group of individuals, and Asian immigrants represent a large share of the nation’s refuges. Asian people are over-represented in both the top and bottom 10% of the income distribution, illustrating the diversity of Asian people’s lived experiences and socioeconomic circumstances in the US. ⁵² Understanding which Asian adults were most impacted by the pandemic, and why, is critical to identifying and targeting strategies to ensure all adults get “caught up” on preventive screenings.

This study has several limitations. First, we assess preventive care use during time periods before and after the start of the COVID-19 pandemic but cannot interpret our findings as causally related to the pandemic; rather, we identify associations between the start of the pandemic and use of preventive care. Second, the race/ethnicity groups we study are limited by the categories available on the NHIS and do not capture significant heterogeneity within race/ethnicity groups. For example, individuals belonging to the Asian race/ethnicity group may have very different lived experiences that may differ based on their heritage or socio-economic status in terms of access to health care and effects of the pandemic. Third, the NHIS data reflect self-reported measures of preventive care use and may be subject to recall bias. A small number of 2021 survey respondents may have been reporting on a 1-year lookback period that included time prior to the declaration of the Public Health emergency and National Emergency; measurement error introduced by this would serve to bias our findings toward understating the decline in preventive care use after the pandemic.

This study documents considerable declines in preventive service use among US adults of all races and ethnicities during the initial months of the COVID-19 pandemic and suggests Asian adults may have experienced greater declines in care relative to non-Hispanic white adults. It will be important for future research to examine longer-term trends in preventive service use to document whether missed care during this period is made up for in a timely manner or whether it had lasting impacts on health outcomes. Further, as the accessibility of preventive services comes into question due to potential changes to zero cost sharing requirements on insurers,^53,54 it will be important to closely monitor access to these important services over time even as the COVID-19 public health emergency has ended.