Abstract
Background:
With the coronavirus outbreak of 2019 (COVID-19) came many changes in how health care is accessed and delivered. Perhaps most notable is the massive expansion of telemedicine, especially in the developed world. With pandemic-induced economic and health care system disruptions, it is reasonable to expect changes in how health care services are utilized by different patients. We examined how health care service usage trends changed for various patient demographics from the pre-COVID-19 era to the COVID-19 era.
Design and methods:
De-identified patient demographics and telemedicine, in-patient, in-person out-patient, radiology/procedures, and emergency department visit data (N = 1,164,719) between January 1st, 2019 and May 31st, 2021 were obtained from UHealth in Miami, Florida, USA. This cross-sectional study employed descriptive statistics and other tools to determine relationships between patient demographics and health system usage.
Results:
There were significant changes in health care usage and demographics for UHealth services from the pre-COVID-19 era to the COVID-19 era. There was an increase in telehealth visits and a corollary decrease of in-person out-patient visits (p < 0.001) along with increased health care utilization by those with commercial insurance (p < 0.001) during COVID-19. Lower-income patients had increased use of in-person out-patient services (p < 0.001). Non-Hispanic, English-speaking patients and those with higher median incomes had higher telemedicine usage.
Conclusions:
COVID-19 revealed differences in health care access, particularly telemedicine access, and highlighted differences in vulnerability among patient demographics. These trends are likely multifactorial and reflect changes in patients’ preferences and disparities in care access.
Keywords: COVID-19, telemedicine, telehealth, pandemic, health care
Introduction
The COVID-19 pandemic had major health, economic, and social impacts on the world. During the height of this outbreak in 2020, many local to international institutions struggled to contain the spread of the virus and provide adequate treatments to not only patients infected with COVID-19, but to patients in general. This struggle may be partially attributed to a large quantity of resources being devoted to fighting the pandemic and significant numbers of patients requiring hospitalization secondary to COVID-19 infection. These factors put notable strains on various health care systems and institutions.1
The pandemic brought about many changes on how health care resources are utilized. Just as many businesses and institutions reduced their in-person services to limit the spread of disease during the pandemic, so too did many doctor’s offices and clinics limit in-person patient visits, particularly in the out-patient setting. In addition, many patients, particularly those who are especially vulnerable to COVID-19 complications like the elderly and those with certain preexisting medical conditions, preferred to meet with medical providers virtually, delay visits, or even forgo receiving medical evaluations and treatments all together for fear of contracting the virus. Avoiding coming to the hospital or clinic for medical care may exacerbate preexisting and ongoing medical conditions, resulting in increased morbidity and mortality and putting further strain on health care systems.2
With the COVID-19 pandemic, we have seen marked changes with how health care is delivered. Perhaps most notable is the significant expansion of virtual health care across many parts of the world, often referred to as telemedicine or telehealth.3 With the increase in telehealth usage, we would predict to see a corresponding increase in discrepancies in health care utilization amongst different demographic groups. For instance, many elderly patients may struggle with technology required for virtual doctor visits and consequently have reduced access to care. In another aspect, patients who come from low-income backgrounds may not have easy access to reliable internet connections or technology to complete telehealth visits.4 The pandemic caused economic instability for many people, which may have led to decreased health care utilization in those more affected such as the uninsured and those with fixed income.
Other studies have found that patient demographics differed significantly regarding telemedicine usage during the COVID-19 pandemic, potentially exacerbating preexisting health care disparities.4 We focused on exploring the effects of the COVID-19 pandemic on different patient populations’ access to all health care modalities, including telemedicine, in-person out-patient, in-patient, radiology/procedures, and emergency department services. Specifically, we investigated health care services from January 1st, 2019 to May 31st, 2021 using patient demographics, such as race, gender, ethnicity, age, household income, and health insurance status. By examining health care usage over this period, we can compare usage trends from before the COVID-19 era to the times throughout the pandemic.
Design and methods
Data collection
This cross-sectional study analyzed patient information from the University of Miami Health System (UHealth) database.De-identified patient data, including demographics and medical visit information, were gathered on all patient visits to the University of Miami Hospital in Miami, Florida, USA from January 1st, 2019 to May 31st, 2021 (N = 1,164,719). The data and analysis were carried out in the latter half of 2021 and during 2022. For each encounter, we gathered patient data such as gender, primary language, ethnicity, race, religion, age, zip code, and health insurance type. We also specified encounter type including in-patient, in-person out-patient, telemedicine, radiology/procedures, and emergency room visits. Data were then classified as pre-COVID-19 versus COVID-19 with March 2020 as the dividing date, since it marks the first case of COVID-19 in Florida.5 Data that was missing greater than 50% of demographic information were excluded from the analysis (N = 7942). Additionally, encounters that had not actually been billed for (i.e. missing ICD-10 diagnosis code) were also excluded (N = 171,709). Excluded visits were checked to ensure that there was not a statistical association between these encounters and the pre-COVID-19 versus COVID-19 groups. Many of the visits (N = 412,621) that had occurred during the COVID-19 period were specific COVID-19 testing lab visits, given that University of Miami students were required to undergo weekly COVID-19 testing; these data were excluded from the analysis as they were not within the scope of our study. The data set was joined with zip code-level income data from the Internal Revenue Service for 2018.6 These data came pre-stratified by income level classification (i.e. $1 under $25,000, $25,000 under $50,000, $50,000 under $75,000, $75,000 under $100,000, $100,000 under $200,000, $200,000 or more), so a weighted-average income for each zip code was calculated.6
Statistical analysis
Analysis was completed using RStudio 1.2.1335.7 Categorical variables were analyzed using chi-squared tests.Continuous variables were analyzed using t-tests or ANOVA depending on the number of groups being compared. In order to correct for multiple comparisons, a Bonferroni correction was applied with the new ɑ value = 0.05/23 = 0.0022. Data visualizations were created with the ggplot2 package within RStudio.8 To show differences in our chosen variables between the pre-COVID-19 and COVID-19 time frames, we compared percent change trend. Percent change for each item is relative to the total volume for that time period (either pre-COVID-19 or COVID-19). The trend is the change in percent for each item from pre-COVID-19 to COVID-19.
Results
Health care service usage
There was a large shift in visit type with an increase of 15.3% for telemedicine visits accompanied by a decrease of 13.4% for (in-person) out-patient visits from the pre-COVID-19 to COVID-19 eras. During these time periods, there were no major changes in emergency room (−0.8%), in-patient (−0.1%), or radiology/procedure (−1.1%) visits (Table 1). Prior to the coronavirus pandemic of 2019, telemedicine comprised a trace percentage of total visits (0.0001% shown as 0% in Figure 1). Telemedicine visits increased to 3% of visits in March 2020 when telemedicine started to be implemented for COVID-19 virtual visits and then increased sharply in April 2020, making up 56% of all patient visits. At the same time, out-patient visits decreased significantly to 6% of all visits in April 2020, from a previous baseline of about 62% of all visits. Over the next year, the proportion of telehealth visits steadily declined with a corresponding steady increase in out-patient visits (Figure 1).
Table 1.
Visit type for pre-COVID-19 versus COVID-19 era. Compared to the pre-COVID-19 era, the COVID-19 era showed a significant increase in telemedicine visits and a decrease in (in-person) out-patient visits, while there were minimal changes in emergency room, in-patient, and radiology/procedure visits.
| Overall (N = 572,303) | Pre-COVID-19 (N = 300,889) | COVID-19 (N = 271,414) | Trend | p-value | |
|---|---|---|---|---|---|
| Visit type | |||||
| Emergency room | 85,532 (14.9%) | 46,106 (15.3%) | 39,426 (14.5%) | (−0.8%) | <0.001 |
| In-patient | 39,157 (6.8%) | 20,647 (6.9%) | 18,510 (6.8%) | (−0.1%) | <0.001 |
| Out-patient | 317,336 (55.4%) | 185,843 (61.8%) | 131,493 (48.4%) | (−13.4%) | <0.001 |
| Radiology/Procedures | 88,781 (15.5%) | 48,258 (16.0%) | 40,523 (14.9%) | (−1.1%) | <0.001 |
| Telemedicine | 41,497 (7.3%) | 35 (0.0%) | 41,462 (15.3%) | (+15.3%) | <0.001 |
Figure 1.
Composition of visit type from the pre-COVID-19 era to the COVID-19 era.
Demographics
There were minor yet statistically significant percentage changes in overall demographics for UHealth services in regards to sex, age, race, ethnicity, language, and religion. There was, however, a notable shift in health care services utilized based on insurance coverage. There was a 3.6% increase in all visits covered by commercial health insurance and a 2.4% decrease in all visits covered by Medicare (Table 2). However, these changes were not equally distributed among visit type. Commercial insurance had the greatest increases in telemedicine, radiology/procedures (+4.8%), out-patient visits (+3.6%) and emergency room visits (+3.3%) (Figure 2). In contrast, Medicare coverage had the greatest decrease in out-patient visits (−4.0%) and radiology/procedure (−3.3%) visits (Figure 2).
Table 2.
Demographic characteristics for pre-COVID-19 versus COVID-19 era. In the COVID-19 era, there was an increase in visits covered by commercial health insurance and a decrease in Medicare covered visits. Health care usage trends among uninsured patients, Medicaid patients, and patients with another insurance type showed smaller changes.
| Overall (N = 572,303) | Pre-COVID-19 (N = 300,889) | COVID-19 (N = 271,414) | Trend | p-value | |
|---|---|---|---|---|---|
| Sex | |||||
| Male | 239,176 (41.8%) | 125,663 (41.8%) | 113,513 (41.8%) | (0%) | 0.65 |
| Female | 333,127 (58.2%) | 175,226 (58.2%) | 157,901 (58.2%) | (0%) | |
| Age (years) | |||||
| Mean (SD) | 54.6 (18.0) | 55.1 (18.0) | 54.2 (17.9) | −0.9 | <0.001 |
| Median [Min, Max] | 57.0 [0, 109] | 57.0 [0, 109] | 56.0 [0, 108] | −1.0 | |
| Race | |||||
| White | 431,941 (75.5%) | 226,157 (75.2%) | 205,784 (75.8%) | (+0.6%) | <0.001 |
| Black | 102,075 (17.8%) | 54,301 (18.0%) | 47,774 (17.6%) | (−0.4%) | |
| Asian | 9614 (1.7%) | 5223 (1.7%) | 4391 (1.6%) | (−0.1%) | |
| Other | 6028 (1.1%) | 3463 (1.2%) | 2565 (0.9%) | (−0.3%) | |
| Unknown | 22,645 (4.0%) | 11,745 (3.9%) | 10,900 (4.0%) | (+0.1%) | |
| Ethnicity | |||||
| Hispanic | 309,575 (54.1%) | 162,469 (54.0%) | 147,106 (54.2%) | (+0.2%) | <0.001 |
| Non-Hispanic | 245,617 (42.9%) | 131,323 (43.6%) | 114,294 (42.1%) | (−1.5%) | |
| Unknown | 17,111 (3.0%) | 7097 (2.4%) | 10,014 (3.7%) | (+1.3%) | |
| Language | |||||
| English | 384,401 (67.2%) | 201,181 (66.9%) | 183,220 (67.5%) | (+0.6%) | <0.001 |
| Spanish | 179,699 (31.4%) | 95,136 (31.6%) | 84,563 (31.2%) | (−0.4%) | |
| Other | 6489 (1.1%) | 3639 (1.2%) | 2850 (1.1%) | (−0.1%) | |
| Unknown | 1714 (0.3%) | 933 (0.3%) | 781 (0.3%) | (0%) | |
| Religion | |||||
| Catholic | 188,433 (32.9%) | 100,864 (33.5%) | 87,569 (32.3%) | (−1.2%) | <0.001 |
| Non-Catholic Christian | 125,890 (22.0%) | 66,424 (22.1%) | 59,466 (21.9%) | (−0.2%) | |
| Non-Christian Faiths | 33,798 (5.9%) | 18,155 (6.0%) | 15,643 (5.8%) | (−0.2%) | |
| Other | 23,250 (4.1%) | 12,164 (4.0%) | 11,086 (4.1%) | (+0.1%) | |
| Unaffiliated | 116,860 (20.4%) | 60,256 (20.0%) | 56,604 (20.9%) | (+0.9%) | |
| Unknown | 84,072 (14.7%) | 43,026 (14.3%) | 41,046 (15.1%) | (+0.8%) | |
| Insurance | |||||
| Commercial | 289,515 (50.6%) | 147,148 (48.9%) | 142,367 (52.5%) | (+3.6%) | <0.001 |
| Medicare | 175,142 (30.6%) | 95,521 (31.7%) | 79,621 (29.3%) | (−2.4%) | |
| Medicaid | 55,721 (9.7%) | 30,055 (10.0%) | 25,666 (9.5%) | (−0.5%) | |
| Other insurance | 13,272 (2.3%) | 6659 (2.2%) | 6613 (2.4%) | (+0.2%) | |
| Uninsured | 38,653 (6.8%) | 21,506 (7.1%) | 17,147 (6.3%) | (−0.8%) | |
| Weighted average income (thousands) | |||||
| Mean (SD) | 88.8 (152) | 89.8 (167) | 87.7 (134) | −2.1 | <0.001 |
| Median [Min, Max] | 52.2 [23.3, 3890] | 52.2 [23.5, 3890] | 52.2 [23.3, 3890] | 0 | |
Figure 2.
Composition of visit type by insurance type between the pre-COVID-19 and COVID-19 eras.
Income
There was also a significant shift in health care services utilized based on median income. But much like health insurance coverage, median income changes were not equally distributed among visit type. In-person out-patient visits were being accessed by patients from lower-income zip codes more during the COVID-19 era and the median income of COVID-19 era out-patient patients was $2400 lower compared to pre-pandemic times (Table 3). However, the median income of emergency room patients during the COVID-19 era did not change from that of the pre-COVID-19 era; these data were not statistically significant at our corrected ɑ value.
Table 3.
Weighted median income by visit type for pre-COVID-19 versus COVID-19 era. Income median in thousands of dollars with ranges in brackets. In the COVID-19 era, patients with lower-income had an increased proportion of healthcare service utilization, most notably in out-patient in-person clinic visits. There was no change in median income of patients visiting the ER during the COVID-19 era.
| Overall | Pre-COVID-19 | COVID-19 | Trend | p-value | |
|---|---|---|---|---|---|
| Visit type | |||||
| Emergency room | 35.2 [23.2, 3890] N = 85,532 | 35.2 [23.5, 3890] N = 46,106 | 35.2 [23.3, 3890] N = 39,426 | 0 | 0.003 |
| In-patient | 46.4 [23.5, 3890] N = 39,157 | 47.7 [25.1, 3890] N = 20,647 | 46.1 [23.5, 3890] N = 18,510 | −1.6 | 0.83 |
| Out-patient | 56.6 [23.5, 3890] N = 317,336 | 56.9 [23.5, 3890] N = 185,843 | 54.5 [23.5, 3890] N = 131,493 | −2.4 | <0.001 |
| Radiology/Procedures | 53.5 [25.1, 3890] N = 88,781 | 52.8 [25.1, 3890] N = 48,258 | 55.2 [25.1, 3890] N = 40,523 | +2.4 | 0.24 |
| Telemedicine | 63.6 [23.5, 3890] N = 41,497 | 56.9 [31.5, 213] N = 35** | 63.6 [23.5, 3890] N = 41,462 | +6.7** | <0.001 |
| All visit types | 52.2 [23.3, 3890] N = 572,303 | 52.2 [23.5, 3890] N = 300,889 | 52.2 [23.3, 3890] N = 271,414 | 0 | <0.001 |
The sample for the pre-COVID-19 telemedicine category was only 35 patients.
Telemedicine outcomes
Telemedicine had too few pre-COVID-19 visits (N = 35) to reliably compare to telemedicine visits during the COVID-19 era. Therefore, we compared pre-COVID-19 in-person out-patient visits (N = 185,843) to COVID-19 era telemedicine visits (N = 41,462). There were major shifts in demographics with much less Hispanic ethnicity (−7.9%), black race (−2.8%) and Spanish language (−9.5%) patients accessing COVID-19 era telemedicine compared to pre-COVID-19 in-person visits. There was a corollary increase in non-Hispanic (+5.6%), unknown ethnicity (+2.4%), and English language (+9.8%) patients accessing telemedicine. Also, telemedicine tended to be used more by higher-income patients, which is reflected by a $6700 increase in median income compared to patients seeing in-person out-patient providers in the pre-COVID-19 era (Table 4).
Table 4.
Pre-COVID-19 (in-person) out-patient versus COVID-19 era telemedicine demographics. There was a shift in who used in-person services pre-COVID-19 compared to telemedicine services in the COVID-19 era. Telemedicine garnered increasing proportions of English-speaking patients with higher-income levels who were also not Hispanic or Black.
| Overall (N = 227,305) | Pre-COVID-19 out-patient (N = 185,843) | COVID-19 telemedicine (N = 41,462) | Trend | p-value | |
|---|---|---|---|---|---|
| Sex | |||||
| Male | 91,188 (40.1%) | 74,706 (40.2%) | 16,482 (39.8%) | (−0.4%) | 0.095 |
| Female | 136,117 (59.9%) | 111,137 (59.8%) | 24,980 (60.2%) | (+0.4%) | |
| Age (years) | |||||
| Mean (SD) | 54.4 (18.1) | 54.7 (18.0) | 53.3 (18.8) | −1.4 | <0.001 |
| Median [Min, Max] | 57.0 [0, 103] | 57.0 [0, 103] | 56.0 [0, 103] | -1.0 | |
| Race | |||||
| White | 176,707 (77.7%) | 144,129 (77.6%) | 32,578 (78.6%) | (+1.0%) | <0.001 |
| Black | 33,665 (14.8%) | 28,477 (15.3%) | 5188 (12.5%) | (−2.8%) | |
| Asian | 4407 (1.9%) | 3583 (1.9%) | 824 (2.0%) | (+0.1%) | |
| Other | 1783 (0.8%) | 1447 (0.8%) | 336 (0.8%) | (0%) | |
| Unknown | 10,743 (4.7%) | 8207 (4.4%) | 2536 (6.1%) | (+1.7%) | |
| Ethnicity | |||||
| Hispanic | 119,240 (52.5%) | 100,175 (53.9%) | 19,065 (46.0%) | (−7.9%) | <0.001 |
| Non-Hispanic | 101,135 (44.5%) | 80,795 (43.5%) | 20,340 (49.1%) | (+5.6%) | |
| Unknown | 6930 (3.0%) | 4873 (2.6%) | 2057 (5.0%) | (+2.4%) | |
| Language | |||||
| English | 159,537 (70.2%) | 127,125 (68.4%) | 32,412 (78.2%) | (+9.8%) | <0.001 |
| Spanish | 64,513 (28.4%) | 55,962 (30.1%) | 8551 (20.6%) | (−9.5%) | |
| Other | 2585 (1.1%) | 2198 (1.2%) | 387 (0.9%) | (−0.3%) | |
| Unknown | 670 (0.3%) | 558 (0.3%) | 112 (0.3%) | (0%) | |
| Religion | |||||
| Catholic | 73,515 (32.3%) | 61,516 (33.1%) | 11,999 (28.9%) | (−4.2%) | <0.001 |
| Non-Catholic Christian | 46,982 (20.7%) | 38,726 (20.8%) | 8256 (19.9%) | (−0.9%) | |
| Non-Christian Faiths | 16,100 (7.1%) | 12,518 (6.7%) | 3582 (8.6%) | (+1.9%) | |
| Other | 9296 (4.1%) | 7778 (4.2%) | 1518 (3.7%) | (−0.5%) | |
| Unaffiliated | 46,772 (20.6%) | 37,768 (20.3%) | 9004 (21.7%) | (+1.4%) | |
| Unknown | 34,640 (15.2%) | 27,537 (14.8%) | 7103 (17.1%) | (+2.3%) | |
| Insurance | |||||
| Commercial | 125,186 (55.1%) | 101,838 (54.8%) | 23,348 (56.3%) | (+1.5%) | <0.001 |
| Medicare | 68,829 (30.3%) | 56,018 (30.1%) | 12,811 (30.9%) | (+0.8%) | |
| Medicaid | 19,599 (8.6%) | 16,522 (8.9%) | 3077 (7.4%) | (−1.5%) | |
| Other Insurance | 4998 (2.2%) | 4329 (2.3%) | 669 (1.6%) | (−0.7%) | |
| Uninsured | 8693 (3.8%) | 7136 (3.8%) | 1557 (3.8%) | (0%) | |
| Weighted average income (thousands) | |||||
| Mean (SD) | 98.2 (173) | 96.8 (180) | 104 (137) | +7.2 | <0.001 |
| Median [Min, Max] | 56.9 [23.5, 3890] | 56.9 [23.5, 3890] | 63.6 [23.5, 3890] | +6.7 | |
Discussion
Principal results
During the COVID-19 pandemic, we saw a significant change in how health care systems delivered care to patients. Early in the pandemic, the majority of out-patient office visits were conducted virtually via telemedicine in order to contain the spread of the coronavirus.3 Some regions had in-patient hospital floors with COVID-19 surges affecting primarily elderly or those with chronic medical conditions. The number of elective procedures performed was reduced to divert resources to treating COVID-19 patients [3]. Many patients avoided seeking medical care out of fear of contracting the virus. While many of these changes would be expected to have significant consequences in the demographics of patients who accessed health care, notable trends were only identified in a few categories; in essence, the majority of health care system utilization and demographics showed no major changes, with notable exceptions that will be discussed in further detail.
Our data show that at the start of the pandemic-related lockdowns in Florida, after March 2020, there was a significant increase in telemedicine visits relative to other encounter types, expanding from 0.0001% of visits before March 2020 to 3% in March 2020, and peaking at 56% of all visits in April 2020. Consequently, out-patient visits during April 2020 consisted of only 6% of all visits, compared to about 62% of visits prior to March 2020. Using March 1st, 2020 as the separating date for pre-COVID-19 versus COVID-19, there was an overall 13.4% decrease in in-person out-patient visits and a 15.3% increase in telemedicine visits in the COVID-19 era. The abrupt onset of the lockdowns we witnessed in April 2020 seemed to correspond to the rapid transition of out-patient physician office visits to virtual visits. The percent change in other encounter types, including emergency room visits, in-patient visits, and radiology/procedure services, was minimal. The minimal change of only 0.1% decrease in in-patient visits from the pre-COVID-19 era to the COVID-19 era is unexpected, given repeated COVID-19 surges during several “waves” of the pandemic and presumed delay of care by many patients with chronic conditions. Other than telemedicine replacing a major portion of in-person out-patient visits, the frequencies of other visit types remained relatively stable.
Insurance coverage of patients accessing health care services had significant shifts during the COVID-19 crisis. There was a 3.6% proportional increase in UHealth visits covered by commercial health insurance (p < 0.001). There was a simultaneous 2.4% decrease in Medicare recipients and 0.8% decrease in uninsured patients from the pre-COVID-19 to COVID-19 era services (p < 0.001). This change may reflect health care utilization shifts due to economic, employment, or cultural influences. This phenomenon was not evenly distributed as evidenced by out-patient visits having a 3.6% increase in commercial insurance patients while also having a 4.0% decrease in Medicare patients. Almost 96% of the nation’s elderly (persons age 65 years and older) have Medicare coverage.9 This reduced number of out-patient visits in Medicare likely reflect older patients, who disproportionally represent Medicare patients, avoiding in-person clinic visits due to their increased vulnerability to suffering from COVID-19 complications. These older Medicare patients however did have a 1.3% increase in their use of in-patient services which could reflect higher acuity services required for these more vulnerable patients during the pandemic.
Given the changes seen in income across multiple visit types, it’s likely that health care services were utilized more by lower-income patients compared to before the pandemic. This significant shift in health care utilization based on median income was not equally distributed among visit type. In-person out-patient visits included more patients with lower median income (−$2400). Concurrently, telemedicine visits comprised a larger proportion of higher-income patients. Again, this could represent choices by lower-income patients to be seen in person, or due to reduced access to telemedicine.
Limitations
A limitation of this study is that the statistical methods are only descriptive and therefore cannot control for confounding variables. For example, while we found significant results in terms of health insurance composition in the pre-COVID-19 versus COVID-19 era, we cannot conclude that this is not the byproduct of additional factors such as income, race, or ethnicity that could be drivers of these significant results. Also, given the limitations in the structure of our external data set, we could only utilize weighted average mean income per each zip code instead of median which would have likely been more representative of our data. An additional limitation of this study is the low volume of telehealth visits (N = 35) between January 2019 and February 2020, corresponding to the pre-COVID-19 era in this study. This low number of virtual visits makes it difficult to compare demographic parameters before COVID-19 versus during the coronavirus pandemic. Given this limitation, we compared pre-COVID-19 out-patient visits (N = 185,843) with COVID-19 era telemedicine visits (N = 41,462). This provides a better comparison of which patient populations shifted from in-person visits to telemedicine visits during the COVID-19 crisis. We found that telemedicine patients were more likely to be non-Hispanic, English speaking, and/or come from higher-income backgrounds. This shift may reflect inadequate telemedicine access for disadvantaged and minority patients, but could also represent a cultural or preference to avoid telemedicine, and instead be seen in person. An additional limitation of this study is that it only examined patients from a single hospital, albeit a large academic hospital. The state of Florida also had COVID-19 policies that tended to favor less restrictions in contrast to many other states, which could potentially influence patients’ health and health care usage and services offered by health care systems.
Conclusions
The COVID-19 pandemic caused significant disruptions in economies and health care systems internationally and locally. As economies shut down and health care systems shifted strategies and resources during the coronavirus pandemic of 2019, we saw significant changes in health care delivery systems and how care is accessed. Our data from the UHealth services showed that telehealth usage increased substantially with a corresponding decrease for in-person out-patient visits. The coronavirus pandemic also led to an increase in health care utilization by those with commercial insurance and those with lower median incomes compared to the pre-pandemic era. The pandemic-induced transition to telemedicine showed a shift to more non-Hispanic and English-speaking patients with higher-incomes accessing these telehealth services. Meanwhile, in-person out-patient visits had a shift toward lower-income patient populations. These differences in health care access may reflect preferences to be seen in person or reduced access to telemedicine for Hispanic, Spanish-speaking, and/or lower-income patients. More research that focuses on these disparities is needed to further investigate these findings and may reveal similar trends across different cities, states, and countries.
Acknowledgments
The authors would like to thank Regina Shulman who carried out the clinical data request.
Footnotes
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding: The author(s) received no financial support for the research, authorship, and/or publication of this article.
Significance for public health: With the onset of the COVID-19 pandemic came many changes in how health care is delivered and accessed. Although telemedicine usage increased significantly at the height of the outbreak with a corresponding decline in face-to-face doctor’s visits, it is less apparent how health care service usage changed amongst different patient populations. This information is important from a public health perspective because it highlights how the COVID-19 pandemic affected existing inequalities and inequities in health care access. Changes in health care delivery systems ultimately affect the health outcomes of patients. This study closely examines an array of demographic characteristics of many patients in an ethnically, racially, economically, and socially diverse region. As health care delivery and technology continue to evolve, particularly telehealth advancement, it will become increasingly relevant to consider the different impacts these changes have on different patient populations and what we can do to address these disparities.
References
- 1. Moynihan R, Sanders S, Michaleff ZA, et al. Impact of COVID-19 pandemic on utilisation of healthcare services: a systematic review. BMJ Open 2021; 11: e045343. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. Jiang L, Ma H. Surgical protocol in a west China day surgery center during the COVID-19 pandemic: practice and experience. Surg Innov 2021; 28(1): 53–57. [DOI] [PubMed] [Google Scholar]
- 3. Contreras CM, Metzger GA, Beane JD, et al. Telemedicine: patient-provider clinical engagement during the COVID-19 pandemic and beyond. J Gastrointest Surg 2020; 24: 1692–1697. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Gmunder KN, Ruiz JW, Franceschi D, et al. Demographics associated with US healthcare disparities are exacerbated by the telemedicine surge during the COVID-19 pandemic. J Telemed Telecare. Epub ahead of print 23 June 2021. DOI: 10.1177/1357633X211025939. [DOI] [PubMed] [Google Scholar]
- 5. Rivkees SA, Roberson S. The Florida Department of Health STEPS public health approach: the COVID-19 response plan and outcomes through May 31, 2020. Public Health Rep 2020; 135(5): 560–564. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. IRS. SOI tax stats - individual income tax statistics - ZIP code data (SOI), https://www.irs.gov/statistics/soi-tax-stats-individual-income-tax-statistics-zip-code-data-soi (2018, accessed 14 February 2022).
- 7. R Core Team. R: A language and environment for statistical computing. R Foundation for Statistical Computing. https://www.R-project.org/ (2019, accessed 14 February 2022).
- 8. Wickham H. Ggplot2: elegant graphics for data analysis. New York, NY: Springer-Verlag, 2016. [Google Scholar]
- 9. Lohr KN; Committee to Design a Strategy for Quality Review and Assurance in Medicare, Institute of Medicine (US). The elderly population. In: Lohr KN; Committee to Design a Strategy for Quality Review and Assurance in Medicare, Institute of Medicine (US) (ed.) Medicare: a strategy for quality assurance. Washington, DC: National Academies Press, 1990, pp.69–91. [Google Scholar]