Impact of the COVID-19 Pandemic on Women's Health Care Access: A Cross-Sectional Study

Abstract

Background:

There has been limited study of how the COVID-19 pandemic has affected women's health care access. Our study aims to examine the prevalence and correlates of COVID-19-related disruptions to (1) primary care; (2) gynecologic care; and (3) preventive health care among women.

Materials and Methods:

We recruited 4,000 participants from a probability-based online panel. We conducted four multinomial logistic regression models, one for each of the study outcomes: (1) primary care access; (2) gynecologic care access; (3) patient-initiated disruptions to preventive visits; and (4) provider-initiated disruptions to preventive visits.

Results:

The sample included 1,285 women. One in four women (28.5%) reported that the pandemic affected their primary care access. Sexual minority women (SMW) (odds ratios [OR]: 1.67; 95% confidence intervals [CI]: 1.19–2.33) had higher odds of reporting pandemic-related effects on primary care access compared to women identifying as heterosexual. Cancer survivors (OR: 2.07; 95% CI: 1.25–3.42) had higher odds of reporting pandemic-related effects on primary care access compared to women without a cancer history. About 16% of women reported that the pandemic affected their gynecologic care access. Women with a cancer history (OR: 2.34; 95% CI: 1.35–4.08) had higher odds of reporting pandemic-related effects on gynecologic care compared to women without a cancer history. SMW were more likely to report patient- and provider-initiated delays in preventive health care. Other factors that affected health care access included income, insurance status, and having a usual source of care.

Conclusions:

The COVID-19 pandemic disrupted women's health care access and disproportionately affected access among SMW and women with a cancer history, suggesting that targeted interventions may be needed to ensure adequate health care access during the COVID-19 pandemic.

Background

Women's health care access affects many outcomes, such as sexual and reproductive health, prenatal and perinatal care, and cancer prevention and control.¹ Despite the important role health care access plays in women's health, it is often limited by numerous factors. In the United States, women are more likely than men to forego health care or underuse medication due to cost.^2,3 In addition to cost, women often report logistical challenges to accessing health care, such as lack of time, problems with obtaining childcare, and the inability to take time off from work.³ The COVID-19 pandemic has led to additional employment disruption for women, increased childcare needs due to childcare facility closures, and the pivot to virtual home learning for school-aged children.^4–7 As a result, the COVID-19 pandemic may have worsened women's health care access.

The COVID-19 pandemic disrupted health care delivery across many specialties. Studies show that primary care and preventive health care visits declined during the pandemic.^8–13 Utilization of cancer screening, such as breast and colon cancer screening, rapidly declined.^{10–12,14,15} There has been less research, however, on how the pandemic has affected women's health care access.

Studies before the pandemic demonstrate persistent disparities in women's health care access based on race/ethnicity, income, language, insurance status, and sexual orientation.^16–20 For example, sexual minority women (SMW)—women who identify as lesbian, bisexual, or other sexual orientation—experience reduced access to care compared with women who identify as heterosexual. SMW are less likely to have health insurance, a usual source of care, receive preventive health care services (e.g., breast and cervical cancer screening), and are more likely to delay necessary care.^21–26 During the COVID-19 pandemic, SMW were more likely to experience employment disruption compared with women who identify as heterosexual^27,28—which may have further affected health care access. COVID-19-related employment disruption also disproportionately affected Black/African American women and women with lower income.²⁹ Given that the economic consequences of the COVID-19 pandemic have been unequal, it is critical to evaluate how women's health care access was affected and whether certain groups of women were more likely to experience reduced health care access.

To address this gap, our study aims to examine the prevalence and correlates of COVID-19-related disruptions to (1) primary care; (2) gynecologic care; and (3) preventive health care among women. Study findings can inform targeted interventions to address potential disruptions to women's health care access.

Materials and Methods

Study sample

We utilized a survey panel management company to recruit 4,000 participants from a probability-based online panel of 60,000 individuals. We restricted the analyses to individuals who identified as female sex at birth and female gender (due to small sample size of individuals identifying as transgender) and individuals aged 21–45 to capture women eligible for both cervical cancer screening and human papillomavirus (HPV) vaccination. To be eligible, individuals needed internet access and English proficiency. Following a systematic and extensive data cleaning approach, we removed data from individuals with nonsensical survey responses (e.g., straight lining of survey responses) (Fig. 1).^30,31

FIG. 1.

Creation of analytic sample.

Survey

Eligible and interested participants completed a one-time survey through Qualtrics software (Provo, UT) that took ∼30 minutes to complete. The survey collected information about the participant, such as knowledge, attitudes, and beliefs about preventive health care, and perceived impact of the pandemic on health care access (e.g., primary care, gynecologic care, preventive health care). The survey also collected information on potential correlates of health care access, including demographics (e.g., race/ethnicity), social determinants of health (e.g., income), prior preventive health care utilization (e.g., cervical cancer screening), and clinical characteristics (e.g., cancer history). Survey administration occurred ∼1-year from the onset of the COVID-19 pandemic, from February 25, 2021 to March 24, 2021, giving participants sufficient time to reflect on how health care access was affected. The response rate was 25.5%. Participants who completed the survey were compensated with reward points, which could be redeemed for gift cards.

Outcome measures

Impact of pandemic on primary care access

The survey asked individuals to report whether the pandemic impacted their ability to access their primary care provider. Response options were: (1) no; (2) yes; and (3) I do not have a primary care provider.

Impact of pandemic on gynecologic care access

The survey asked individuals to report whether the pandemic impacted their ability to access their gynecologic care provider (e.g., obstetrician and gynecologist). Response options were: (1) no; (2) yes; and (3) I do not have a gynecologic care provider.

Provider-initiated disruption of preventive health care visits

The survey asked individuals to report whether their health care provider's office canceled or delayed one or more preventive health care visits due to the pandemic. Response options were: (1) no; (2) yes; and (3) I didn't have any preventive health care appointments scheduled in the past year.

Patient-initiated disruption of preventive health care visits

The survey asked individuals to report whether they canceled or delayed one or more preventive health care visits due to their own concerns related to the pandemic. Participants could provide three response options: (1) no; (2) yes; and (3) I didn't have any preventive health care appointments scheduled in the past year.

Potential correlates of health care access during the pandemic

The survey assessed multiple factors associated with health care access, including (1) demographics (e.g., race/ethnicity, age, sexual orientation, relationship status, whether an individual or their parents were born in the United States, preference for receiving health information in a language other than English, geographic region); (2) social determinants of health (e.g., education, income, employment status, health literacy); (3) health care access (e.g., insurance type, usual source of care, timing of last health care visit); (4) preventive health care use (e.g., vaccination history, being up to date on cervical cancer screening based on the United States Preventive Services Taskforce [USPSTF] guidelines³²); (5) knowledge, attitudes, and beliefs about HPV-related cancers (e.g., scale measuring perceived risk of developing HPV-related cancers)³³; and (6) clinical characteristics (e.g., cancer history, sexual history, clinical risk for cervical cancer).

Statistical analyses

We calculated descriptive statistics, including percentages for categorical variables and means and standard deviations (SD) for continuous variables. We conducted four multinomial logistic regression models, one for each of the study outcomes. In the first model, we estimated the probability of (1) yes, the pandemic affected primary care access and (2) I do not have a primary care provider compared against the reference category of “no, the pandemic did not affect primary care access.” In the second model, we estimated the probability of (1) yes, the pandemic affected gynecologic care access and (2) I do not have a gynecologic care provider against the reference category of “no, the pandemic did not affect gynecologic care access.” In the third model, we estimated the probability of (1) yes, my provider delayed or cancelled my preventive health care visit and (2) I did not have a preventive health care visit scheduled in the past year against the reference category of “no, my provider did not delay or cancel my preventive healthcare visit.” In the fourth model, we estimated the probability of (1) yes, I delayed or cancelled my preventive health care visit and (2) I did not have a preventive health care visit scheduled during the pandemic against the reference category of “no, I did not delay or cancel my preventive healthcare visit.” We chose a multinomial logistic regression rather than an ordinal logistic regression model because it does not assume any intrinsic ordering across the categories.

Given the many variables likely to affect health care access and the exploratory nature of study, we used backwards selection set at the 10% significance level for variable selection. Factors likely to influence health care access (e.g., employment and insurance) are often highly correlated. To address this issue, we opted to use backwards selection over other selection approaches given its improved performance for dealing with potential collinearity.^34,35 We also present univariate estimates for all potential correlates and outcomes in Supplementary Tables S1–S4. The data analyses were conducted using SAS Software version 9.4. Results are reported using adjusted odds ratios (OR), 95% confidence intervals (CI), and in adherence with the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines.³⁶ The study was approved by Moffitt Cancer Center's Institutional Review Board of record, Advarra, and the Scientific Review Committee and was determined to be exempt.

Results

The sample included 1,285 women (Table 1). The mean age was 31 (SD: 7.3). The racial breakdown of the sample included 8.9% Black/African American, 15.7% other racial groups (e.g., multiple racial categories), and 75.4% White. The ethnic breakdown of the sample included 13.9% Hispanic/Latinx and 86.1% non-Hispanic/Latinx. About one-third of participants had a bachelor's degree (34.5%) or an associate degree or some college (32.7%), while fewer had a high school diploma or less (17.8%) or a graduate degree (15.0%). Most women (60.1%) were up to date on their cervical cancer screening. About one in three participants (29.9%) identified as being at risk for cervical cancer (e.g., abnormal Pap test, HPV diagnosis). On a modified scale assessing perceived risk for HPV-related cancers (Supplementary File S1), participants reported a mean score of 2.8 (SD: 1.2), suggesting low overall perceived risk of HPV-related cancers.

Table 1.

Sample Characteristics, N = 1,285

Variable	Level	N	%
Sexual minority	Yes	212	16.8
	No	1,049	83.2
	Missing	24	—
Race	White	968	75.4
	Black/African American	114	8.9
	Under-represented groupa	201	15.7
	Missing	2	—
Ethnicity	Hispanic/Latinx	178	13.9
	Non-Hispanic/Latinx	1,104	86.1
	Missing	3	—
Education	High school diploma, GED,b or less	228	17.8
	Some college/Associate degree	420	32.7
	Bachelor's degree	442	34.5
	Graduate degree	193	15.0
	Missing	2	—
Employment status	Employed	951	74.1
	Unemployed	141	11.0
	Otherc	192	15.0
	Missing	1	—
Annual income	$0–$19,999	137	10.8
	$20,000–$49,999	335	26.3
	$50,000–$74,999	316	24.8
	$75,000–$99,999	230	18.1
	$100,000 or more	255	20.0
	Missing	12	—
Relationship status	Married/partnered	774	60.3
	All othersd	510	39.7
	Missing	1	—
Were you born in the United States?	No	114	8.9
	Yes	1,170	91.1
	Missing	1	—
Were either of your parents born outside the United States?	No	1,022	80.1
	Yes	254	19.9
	Missing	9	—
Do you prefer to receive health information in a language other than English?	No	1,220	95.0
	Yes	64	5.0
	Missing	1	—
Geographic region	Midwest	260	20.2
	Northeast	227	17.7
	South	572	44.5
	West	226	17.6
Type of insurance	Medicare/Medicaid/Tricare	375	29.2
	Private insurance	699	54.4
	Other	27	2.1
	No insurance	184	14.3
Have you ever had vaginal, anal, or oral sex?	I prefer not to answer	48	3.7
	No	140	10.9
	Yes	1,096	85.4
	Missing	1	—
Have you ever been diagnosed as having cancer?	No	1,204	93.8
	Yes	79	6.2
	Missing	2	—
Is there a particular doctor, nurse, or other health professional that you see most often?	No	554	43.1
	Yes	730	56.9
	Missing	1	—
About how long has it been since you last visited a doctor or nurse for a routine checkup? A routine checkup is a general physical examination, not an examination for a specific injury, illness, or condition.	Within past year (12 months ago or less)	747	58.2
	More than 1 year but <2 years ago	260	20.2
	More than 2 years but <5 years ago	137	10.7
	Five or more years ago/I don't know/never	140	10.9
	Missing	1	—
In general, how difficult is it for you to understand written health information?	Very easy	474	37.0
	Difficult	754	58.8
	Other	54	4.2
	Missing	3	—
Clinical risk for cervical cancer (e.g., HPV diagnosis, abnormal pap smear)	No	800	70.1
	Yes	342	29.9
	Missing	143	—
Up to date on cervical cancer screeninge	No	513	39.9
	Yes	772	60.1
Age	Mean	31.07
	Median	30
	Minimum	21
	Maximum	45
	Standard deviation	7.25
	Missing	0
Average score of perceived HPV-related cancer risk	Mean	2.81
	Median	2.67
	Minimum	1
	Maximum	7
	Standard deviation	1.15
	Missing	0

^a Under-represented racial categories include Asian, American Indian/Alaskan Native, Native Hawaiian, Pacific Islander, and multiple racial categories.

^b GED refers to a high school equivalency diploma.

^c Other employment categories include homemaker/full-time parent, student, military personnel, retired, and disabled/unable to work.

^d Other relationship categories include divorced, widowed, separated, dating, and not dating and never been married.

^e Up to date on cervical cancer screening was defined based on the USPSTF guidelines.

HPV, human papillomavirus; USPSTF, United States Preventive Services Taskforce.

Impact of pandemic on primary care access

Among women with complete data for all covariates of interest (N = 1,245) (Table 2), ∼1 in 4 women (N = 355, 28.5%) reported that the pandemic affected their primary care access. Controlling for other factors, SMW (OR: 1.67; 95% CI: 1.19–2.33) had higher odds of reporting pandemic-related effects on primary care access compared to women identifying as heterosexual. Women with a cancer history (OR: 2.07; 95% CI: 1.25–3.42) had higher odds of reporting pandemic-related effects on primary care access compared to women without a cancer history. Additional factors that increased the odds of reporting pandemic-related effects on primary care access included having a usual source of care and having a health care visit within the past year (Table 2).

Table 2.

Multinomial Regression of Self-Reported Pandemic Impact on Primary Care Access Among Women, N = 1,245

Covariate	Reported impact on primary care access,a N = 355			Does not have primary care provider,a N = 129			Type III p-values for backward selection
	OR	95% CI	p	OR	95% CI	p
Sexual minority							0.011
No (ref)
Yes	1.67	1.19–2.33	0.003	1.18	0.67–2.05	0.570
Income							0.008
$100,000+ (ref)
$0–$19,999	0.70	0.40–1.22	0.212	2.27	0.85–6.08	0.103
$20,000–$49,999	1.29	0.88–1.91	0.196	3.75	1.53–9.20	0.004
$50,000–$74,999	1.46	0.99–2.14	0.055	3.54	1.44–8.73	0.006
$75,000 to $99,999	1.04	0.68–1.57	0.868	1.68	0.60–4.70	0.323
Geographic region							0.032
West (ref)
Midwest	0.57	0.37–0.87	0.008	1.18	0.53–2.60	0.689
Northeast	0.86	0.57–1.32	0.495	2.22	1.01–4.85	0.047
South	0.68	0.47–0.97	0.031	1.27	0.64–2.54	0.493
Insurance							<0.001
No insurance (ref)
Public payer	0.95	0.59–1.52	0.833	0.43	0.24–0.76	0.004
Private payer	0.90	0.57–1.40	0.628	0.26	0.15–0.44	<0.001
Other	1.28	0.49–3.38	0.612	0.41	0.08–2.19	0.298
Usual source of care							<0.001
No (ref)
Yes	1.52	1.15–2.02	0.004	0.40	0.24–0.65	<0.001
Last health care visit							<0.001
More than 5 years (ref)
<1 year	1.21	0.69–2.14	0.501	0.09	0.05–0.18	<0.001
1–2 years	1.91	1.05–3.46	0.034	0.45	0.25–0.80	0.006
3–5 years	2.25	1.17–4.33	0.015	0.61	0.32–1.14	0.119
Cancer history							0.017
No (ref)
Yes	2.07	1.25–3.42	0.005	1.66	0.61–4.56	0.323

^a The base category for the multinomial regression was that the COVID-19 pandemic had no impact on primary care provider visits, N = 761.

CI, confidence intervals; OR, odds ratios.

Approximately 1 in 10 women (N = 129; 10.4%) reported not having a primary care provider. Women with an annual income of $20,000–$49,999 (OR: 3.75; 95% CI: 1.53–9.20) or an annual income of $50,000–$74,999 (OR: 3.54; 95% CI: 1.44–8.73) had higher odds of lacking a primary care provider compared to women with an annual income of $100,000 or more. Women residing in the northeast (OR: 2.22; 95% CI: 1.01–4.85) had higher odds of lacking a primary care provider compared to women residing in the west. Several factors were associated with lower odds of lacking a primary care provider, including having a usual source of care, private insurance, and having a health care visit in the past year (Table 2).

Impact of pandemic on gynecologic care access

Among women with complete data for all covariates of interest (N = 1,236) (Table 3), 16.7% (N = 207) reported that the pandemic affected their gynecologic care access. Women with a cancer history (OR: 2.34; 95% CI: 1.35–4.08) had higher odds of reporting pandemic-related effects on gynecologic care compared to women without a cancer history. Women with a last health care visit 1–2 years ago (OR: 2.83; 95% CI: 1.30–6.15) or 3–5 years ago (OR: 2.36; 95% CI: 1.01–5.56) had higher odds of reporting pandemic-related effects on gynecologic care compared to women with a health care visit of more than 5 years ago. Additional factors that were associated with higher odds of reporting pandemic effects on gynecologic care access included higher perceived risk of HPV-related cancers and being married (Table 3).

Table 3.

Multinomial Regression of Self-Reported Pandemic Impact on Gynecologic Care Access Among Women, N = 1,236

Covariate	Reported impact on gynecologic care access,a N = 207			Does not have gynecologic care provider,a N = 213			Type III p-values for backward selection
	OR	95% CI	p	OR	95% CI	p
Sexual minority
No (ref)
Yes	1.47	0.98–2.19	0.063	1.55	0.99–2.41	0.054	0.053
Income
$100,000+ (ref)							0.031
$0–$19,999	1.34	0.68–2.64	0.399	1.29	0.60–2.77	0.508
$20,000–$49,999	1.55	0.96–2.49	0.073	2.34	1.25–4.37	0.008
$50,000–$74,999	1.33	0.83–2.14	0.235	2.68	1.44–5.01	0.002
$75,000 to $99,999	1.06	0.63–1.77	0.832	1.65	0.83–3.28	0.154
Employment status							0.004
Employed (ref)
Unemployed	0.89	0.49–1.61	0.703	2.09	1.24–3.52	0.006
Other	1.03	0.65–1.64	0.888	2.09	1.32–3.33	0.002
Relationship status							<0.001
All others (ref)
Married/partnered	1.45	1.03–2.05	0.034	0.57	0.40–0.82	0.002
Insurance							<0.001
No insurance (ref)
Public payer	0.95	0.54–1.69	0.870	0.26	0.16–0.44	<0.001
Private payer	1.18	0.68–2.05	0.548	0.44	0.28–0.70	0.001
Other	1.67	0.56–5.00	0.358	0.26	0.06–1.08	0.064
Up to date on cervical cancer screening							<0.001
No (ref)
Yes	0.96	0.68–1.36	0.829	0.19	0.13–0.28	<0.001
Perceived HPV-related cancer risk, mean score	1.19	1.04–1.37	0.011	0.80	0.69–0.94	0.006	<0.001
Last health care visit
More than 5 years (ref)							<0.001
<1 year	1.60	0.76–3.40	0.220	0.27	0.16–0.45	<0.001
1–2 years	2.83	1.30–6.15	0.009	0.52	0.31–0.90	0.020
3–5 years	2.36	1.01–5.56	0.049	0.73	0.39–1.34	0.307
Cancer history							0.007
No (ref)
Yes	2.34	1.35–4.08	0.003	1.84	0.88–3.85	0.106

^a The base category for the multinomial regression was that the COVID-19 pandemic had no impact on gynecologic care provider visits, N = 816.

About 17% of women (N = 213) reported that they did not have a gynecologic care provider. Women with an annual income of $20,000–$49,999 (OR: 2.34; 95% CI: 1.25–4.37) and women with an annual income of $50,000–$74,999 (OR: 2.68; 95% CI: 1.44–5.01) had higher odds of reporting lack of a gynecologic care provider compared to women with an annual income of $100,000 or more. Women who are unemployed (OR: 2.09; 95% CI: 1.24–3.52) and women with other employment status (OR: 2.09; 95% CI: 1.32–3.33) had higher odds of reporting lack of a gynecologic care provider compared to women who are employed. Factors associated with lower odds of reporting a lack of gynecologic care provider included being married, having private insurance, being up-to-date on cervical cancer screening, having a higher perceived risk of HPV-related cancer, and having a health care visit in the past year (Table 3).

Provider-initiated disruption of preventive health care visits

Among women with complete data for all covariates of interest (N = 1,244) (Table 4), about a quarter of women (N = 301, 24.2%) reported that they experienced a provider-initiated disruption of preventive health care visits during the pandemic. SMW (OR: 1.58; 95% CI: 1.10–2.28) had higher odds of reporting provider-initiated disruption of preventive health care visits compared to women identifying as heterosexual. Women with an annual income of $20,000–$49,999 (OR: 0.63; 95% CI: 0.42–0.93) or an annual income of $50,000–$74,999 (OR: 0.52; 95% CI: 0.35–0.79) reported lower odds of provider-initiated disruption of preventive health care visits compared to women with an annual income of $100,000 or more. Women who were unemployed (OR: 0.50; 95% CI: 0.28–0.88) had lower odds of provider-initiated disruption of preventive health care visits compared to women who were employed.

Table 4.

Multinomial Regression of Provider-Initiated Disruption of Preventive Health Visits Among Women, N = 1,244

Covariate	Provider-initiated disruption of preventive health visits,a N = 301			No preventive health visits scheduled,a N = 190			Type III p-values for backward selection
	OR	95% CI	p	OR	95% CI	p
Sexual minority							0.017
No (ref)
Yes	1.58	1.10–2.28	0.014	1.58	0.99–2.51	0.056
Income							0.022
$100,000+ (ref)
$0–$19,999	0.58	0.32–1.03	0.063	1.59	0.73–3.46	0.244
$20,000–$49,999	0.63	0.42–0.93	0.021	1.55	0.80–2.99	0.195
$50,000–$74,999	0.52	0.35–0.78	0.002	1.77	0.92–3.39	0.088
$75,000 to $99,999	0.73	0.48–1.10	0.132	1.12	0.54–2.36	0.758
Employment status							0.083
Employed (ref)
Unemployed	0.50	0.28–0.88	0.016	1.14	0.66–1.96	0.637
Other	0.76	0.51–1.15	0.192	1.24	0.75–2.05	0.413
Prefer to receive health information in language other than English							0.024
No (ref)
Yes	2.05	1.15–3.65	0.015	0.72	0.29–1.80	0.485
Insurance							<0.001
No insurance (ref)
Public payer	1.24	0.73–2.11	0.431	0.32	0.19–0.54	<0.001
Private payer	0.82	0.49–1.38	0.455	0.35	0.22–0.55	<0.001
Other	1.50	0.55–4.12	0.428	0.23	0.05–1.18	0.078
Usual source of care							<0.001
No (ref)
Yes	1.47	1.09–1.99	0.012	0.48	0.32–0.72	<0.001
Up to date on cervical cancer screening							<0.001
No (ref)
Yes	0.93	0.69–1.26	0.640	0.45	0.31–0.66	<0.001	<0.001
Last health care visit
More than 5 years (ref)
<1 year	2.43	1.19–4.98	0.015	0.17	0.10–0.30	<0.001
1–2 years	2.15	1.02–4.57	0.046	0.59	0.35–1.00	0.048
3–5 years	1.18	0.48–2.91	0.713	0.95	0.54–1.67	0.858

^a The base category for the multinomial regression was report of no provider-initiated disruption of preventive health visits, N = 753.

Factors associated with higher odds of provider-initiated disruption of preventive health care visits included preferring to receive health information in a language other than English, having a usual source of care, and having a health care visit in the past year (Table 4).

About 15% of women (N = 190) reported that they did not have any preventive health care visits scheduled during the pandemic. Women with insurance provided by a public payer (OR: 0.32; 95% CI: 0.19–0.54) or women with private insurance (OR: 0.35; 95% CI: 0.22–0.55) had lower odds of reporting no preventive health care visits scheduled in the past year compared to women who were uninsured. Women with a usual source of care (OR: 0.48; 95% CI: 0.32–0.72) had lower odds of reporting no preventive health care visits scheduled in the past year compared to women without a usual source of care. Other factors associated with lower odds of having no preventive health care visits in the past year included being up-to-date on cervical cancer screening and having a health care visit in the past year (Table 4).

Patient-initiated disruption of preventive health care visits

Among women with complete data for all covariates of interest (N = 1,245) (Table 5), about a quarter (N = 329, 26.4%) reported that they canceled or delayed a preventive health care visit. Women with an annual income of less than $19,999 (OR: 0.48; 95% CI: 0.28–0.83), an annual income of $20,000–$49,999 (OR: 0.56; 95% CI: 0.38–0.82), and an annual income of $50,000–$74,999 (OR: 0.57; 95% CI: 0.39–0.84) had lower odds of delaying or canceling a preventive health care visit compared to women with an annual income of greater than $100,000. Factors associated with higher odds of reporting a patient-initiated disruption of preventive health care visits included sexual minority status, having a usual source of care, and having a health care visit in the past year (Table 5).

Table 5.

Multinomial Regression of Patient-Initiated Disruption of Preventive Health Visits Among Women, N = 1,245

Covariate	Patient-initiated disruption of preventive health visits,a N = 329			No preventive health visits scheduled,a N = 165			Type III p-values for backward selection
	OR	95% CI	p	OR	95% CI	p
Sexual minority							0.013
No (ref)
Yes	1.49	1.05–2.11	0.025	1.80	1.11–2.91	0.018
Income							0.023
$100,000+ (ref)
$0–$19,999	0.48	0.28–0.83	0.008	1.28	0.58–2.81	0.546
$20,000–$49,999	0.56	0.38–0.82	0.003	1.33	0.67–2.62	0.417
$50,000–$74,999	0.57	0.39–0.84	0.005	1.38	0.70–2.72	0.358
$75,000–$99,999	0.76	0.51–1.14	0.185	0.95	0.43–2.08	0.896
Insurance							<0.001
No insurance (ref)
Public payer	1.12	0.69–1.83	0.638	0.34	0.20–0.59	<0.001
Private payer	0.73	0.45–1.18	0.195	0.35	0.22–0.57	<0.001
Other	1.16	0.43–3.08	0.771	0.31	0.06–1.59	0.159
Usual source of care							<0.001
No (ref)
Yes	1.37	1.03–1.83	0.031	0.41	0.27–0.64	<0.001
Up to date on cervical cancer screening							0.002
No (ref)
Yes	0.99	0.74–1.31	0.931	0.48	0.32–0.72	<0.001
Last health care visit							<0.001
More than 5 years (ref)
<1 year	2.24	1.13–4.42	0.021	0.14	0.08–0.26	<0.001
1–2 years	2.63	1.29–5.36	0.008	0.56	0.33–0.96	0.036
3–5 years	2.63	1.20–5.74	0.015	0.94	0.53–1.69	0.843

^a The base category for the multinomial regression was report of no patient-initiated disruption of preventive health visits, N = 751.

About 13% of women (N = 165) reported that they did not have any preventive health care visits scheduled during the pandemic. SMW (OR: 1.80; 95% CI: 1.11–2.91) had higher odds of reporting no preventive health care visits scheduled in the past year compared to women who identified as heterosexual. Factors associated with lower odds of not having any preventive health care visits in the past year included private insurance, having a usual source of care, being up-to-date on cervical cancer screening, and having a health care visit in the past year (Table 5).

Discussion

Overall, our study found that the COVID-19 pandemic disrupted health care access among women during the first year of the pandemic. About a quarter of women reported a disruption to primary care access and about a quarter of women reported provider- and patient-initiated disruptions to preventive health care visits. Fewer women (∼17%) reported disruptions to gynecologic care access. Our study found that cancer survivors reported greater disruptions in health care access, including primary care and gynecologic care. In addition, SMW were more likely to report disruptions in primary care access and patient- and provider-initiated disruptions in preventive health care visits. Other factors that commonly affect health care access,^37,38 such as health insurance, income, and having a usual source of care, were associated with primary care, gynecologic care, and preventive health care access.

Our study found that SMW were more likely to report pandemic-related disruptions to primary care and preventive care access compared to non-SMW. Before the pandemic, SMW were less likely to receive recommended women's health care services, such as breast and cervical cancer screening.^{21,23,39–43} Research has found that SMW face similar health care access barriers to non-SMW (e.g., cost barriers) and unique barriers, such as concerns about discrimination and lack of sexual identity assessment and disclosure.^44–48 SMW are also less likely to have health insurance and a usual source of care compared to non-SMW.^49,50 Findings from our study suggest that the pandemic may have exacerbated health care access disparities between SMW and non-SMW.

Prior studies show that state-level protections for SMW, inclusive clinic environments (e.g., SMW-friendly materials), and provider recommendation are key facilitators of receiving recommended women's health services among SMW.^41,51,52 Researchers have also recommended strategies such as targeted health communication materials that discuss SMW health issues (e.g., increased risk for breast and cervical cancer).⁵³ Given the complex barriers that SMW experience accessing health care, multilevel interventions are needed that improve health care access and utilization of recommended women's health care services among SMW.

Similar to other studies, our findings suggest that cancer survivors were more likely to report pandemic-related disruptions to primary care and gynecologic care compared to women without a cancer history.⁵⁴ Before the pandemic, primary care access among cancer survivors was suboptimal. Primary care providers report insufficient knowledge of cancer survivorship and a need for additional training around the unique needs of cancer survivors (e.g., long-term effects of cancer).^55–59 Researchers have proposed models for better integrating primary care into survivorship care, such as risk stratification to determine which patients should be transitioned to oncology-led, primary care-led, or shared care models during survivorship.^60,61 There has been limited evaluation of the effectiveness of such models, however.^62–64

Fewer studies have explored care coordination across oncologists and gynecologic care providers.^65–68 Women with a history of certain cancers (e.g., breast) may experience unique long-term sexual and reproductive health concerns that should be jointly managed by an oncology and gynecologic care provider team.^69,70 Available studies suggest that women have varying preferences about which concerns (e.g., problems with sexual function, fertility concerns) should be managed by the oncology versus gynecology provider during cancer survivorship.^65–68 Further studies are needed to develop and test survivorship care models to ensure adequate access to gynecologic care among women cancer survivors.

Our study found that women who are not up-to-date on cervical cancer screening according to USPSTF guidelines were more likely to report not having had preventive health visits scheduled during the COVID-19 pandemic. Researchers have called for efforts to prioritize patients who are out-of-date with cancer screening during the pandemic, a time when there may be a backlog of overdue screenings.⁷¹ A recent qualitative study found that some primary care providers developed triage systems to prioritize cancer screening among overdue patients, while other providers described not having sufficient data on screening history to develop a triage system.⁷² Prior studies have documented challenges with having complete cancer screening history and other relevant factors (e.g., smoking history) in the electronic health record and have suggested strategies such as encouraging patients to review their patient health record and submit corrections and additions.^73–75 Additional studies are needed to test strategies for integrating patient-generated data into the electronic health record that can support cancer screening documentation.

Prior studies consistently demonstrate that health care access varies based on income, insurance status, and Black/African American race and Hispanic/Latinx ethnicity.^16–18 In our study, we found that income and insurance status were consistent predictors of primary care, gynecologic care, and preventive care access. In contrast, we did not find that Black/African American race or Hispanic/Latinx ethnicity were significant correlates of health care access in either the unadjusted or adjusted analyses. One potential reason for this may be that our study has under-representation from Black/African American (8.9% in sample vs. 13.4% in national estimates) and Hispanic/Latinx women (13.9% in sample vs. 18.7% in national estimates).⁷⁶ Therefore, it is possible that our study was underpowered to detect racial and ethnic differences in health care access. Future studies are needed that oversample participants based on race/ethnicity to better examine racial disparities in women's health care access during the COVID-19 pandemic.

Limitations

Our study has several limitations. First, data are self-reported and subject to bias. It is possible that participants may have over or underestimated the effects of the COVID-19 pandemic on health care access. We are also unable to evaluate the effects of nonresponse bias, such as comparing response rates by participant demographics (e.g., race/ethnicity) because only data on overall response rate (∼25%) were collected. Our study does not account for other patient characteristics that are likely to affect health care access (e.g., chronic conditions, disability, or rural residence). Our study excluded individuals with limited English proficiency. Further studies are needed to better characterize the health care experiences of women with limited English proficiency during the COVID-19 pandemic. Our study did not capture important elements of women's health care (e.g., contraception use, current pregnancy status) that should be examined in future studies. Furthermore, our survey was not designed to capture reasons for the primary care or gynecologic care visit (e.g., preventive care, reproductive health care, and so on).

Our study examined the first year of the pandemic; future studies are needed to examine longitudinal changes in women's health care access. Studies have shown that certain types of health care, such as primary care visits, have rebounded since the first year of the pandemic.⁷⁷ Our survey was not designed to capture differences in in-person versus telehealth visits (e.g., videoconferencing, phone visits). Further studies are needed to describe how telehealth may have expanded or hindered women's health care access during the COVID-19 pandemic. Finally, our survey did not define what was meant by preventive health care visits so it is possible that participants may have under- or overestimated preventive health care receipt.

Conclusions

Overall, the COVID-19 pandemic disrupted health care access among women during the first year of the pandemic. About a quarter of women reported a disruption to primary care access and about a quarter of women reported provider- and patient-initiated disruptions to preventive health care visits. Fewer women (∼17%) reported disruptions to gynecologic care access. Our study found that SMW and women with a cancer history were more likely to report disruptions in health care, suggesting that targeted interventions may be needed to ensure adequate health care access during the COVID-19 pandemic. Strategies may be needed that broaden health care access, such as home-based testing and mobile clinics, and public health campaigns that emphasize the importance of women's health care.

Ethics Approval

The study was approved by the Moffitt Cancer Center Scientific Review Board and the Institutional Review Board of record, Advarra.

Informed Consent

Informed consent was obtained for all individual participants in the study.

Authors' Contributions

K.T.: Conceptualization; Methodology; Writing—Original draft.

N.C.B.: Methodology; Formal analysis.

J.W.: Methodology; Formal analysis; Data curation.

M.A.: Writing—Review and editing.

J.Y.I.: Writing—Review and editing.

S.T.V.: Writing—Review and editing.

C.D.M.: Writing—Review and editing.

C.K.G.: Writing–Review and editing.

M.L.K.: Writing—Review and editing.

K.J.H.: Writing—Review and editing.

S.M.C.: Conceptualization; Methodology; Funding; Project administration; Supervision.

Author Disclosure Statement

Dr. Gwede currently serves in a leadership role for the American Association for Cancer Education.

Dr. Islam has received support to attend the following conferences: American Association for Cancer Research and the American Society of Preventive Oncology.

Dr. Kasting has received research grant funding from Merck unrelated to the current study.

Dr. Brownstein has received honoraria from the Statistical Consulting Section of the American Statistical Association (ASA) for Best Paper Award in 2019. Dr. Brownstein also received travel support to serve as an ad hoc grant reviewer for the American Cancer Society. Dr. Brownstein currently serves on a Data Safety Monitoring Board for Moffitt Cancer Center's Scientific Review Committee. Dr. Brownstein currently serves as Vice President of the Florida Chapter of the ASA and Section Representative for the ASA Statistical Consulting Section.

Dr. Christy serves as a Medical Advisory Board Member of the HPV Cancers Alliance.

Drs. Turner, Whiting, Arevalo, Vadaparampil, Meade, and Head do not have any conflicts of interest to report.

Funding Information

The study was supported with funding from a Moffitt Center for Immunization and Infection Research in Cancer Award (PI: Shannon M. Christy) and a Moffitt Merit Society Award (PI: Shannon M. Christy). This work has been supported, in part, by both the Participant Research, Interventions, and Measurement Core and the Biostatistics and Bioinformatics Shared Resource at the H. Lee Moffitt Cancer Center & Research Institute, a comprehensive cancer center designated by the National Cancer Institute and funded, in part, by Moffitt's Cancer Center Support Grant (P30-CA076292).

Additional support came from the South Carolina Clinical and Translational Science (SCTR) Institute at the Medical University of South Carolina. The SCTR Institute is funded by the National Center for Advancing Translational Sciences of the National Institutes of Health (Grant UL1TR001450). The contents of this article are solely the responsibility of the authors and do not necessarily represent the official views of the National Cancer Institute or National Institutes of Health. Monica L. Kasting's work on this project was made possible with support from grant numbers, KL2TR002530 (B. Tucker Edmonds, PI), and UL1TR002529 (S. Moe and S. Wiehe, co-PIs) from the National Institutes of Health, National Center for Advancing Translational Sciences, Clinical and Translational Sciences Award.

Supplementary Material

Supplementary File S1

Supplementary Table S1

Supplementary Table S2

Supplementary Table S3

Supplementary Table S4