Direct health care costs associated with COVID-19 in the United States
Publication: Journal of Managed Care & Specialty Pharmacy
Volume 28, Number 9
Abstract
BACKGROUND: Data on the real-world health care burden of COVID-19 in the United States are limited.
OBJECTIVE: To compare health care resource use (HRU), direct health care costs, and long-term COVID-19–related complications between patients with vs patients without COVID-19 diagnoses.
METHODS: Using IBM MarketScan Commercial Claims and Encounters and Medicare Supplemental and Coordination of Benefits administrative claims databases (January 1, 2018, to March 1, 2021), this retrospective, matched cohort study compared patients with a recorded COVID-19 diagnosis to control subjects with no recorded diagnosis for COVID-19, personal history of COVID-19, or pneumonia due to COVID-19. To capture typical health care utilization, the control group was analyzed in 2019 (prepandemic); their index date was assigned as 1 year before the index date (first observed COVID-19 diagnosis) of their matched COVID-19 patient. All patients had continuous health plan coverage for at least 6 months pre-index (baseline) and at least 6 months post-index (allowing censoring during month 6). Separately for commercial and Medicare cohorts, COVID-19 and control patients were matched 1:1 using propensity scores, number of followup months, and indicator of age 18 years or older. During each month of the 6-month follow-up, all-cause HRU, health care costs, and COVID-19–related complications were compared between patients with COVID-19 and controls.
RESULTS: After matching COVID-19 and control patients 1:1, a total of 150,731 commercial matched pairs and 1,862 Medicare matched pairs were retained; baseline characteristics were similar between patients with COVID-19 and controls. Patients with COVID-19 and controls had mean ages of 38.9 and 39.7 years in the commercial cohort and 74.3 and 75.3 years in the Medicare cohort, respectively.
In month 1 of follow-up, patients with COVID-19 relative to controls were significantly more likely to have at least 1 inpatient admission (commercial: 6.9% vs 0.5%; Medicare: 29.1% vs 1.3%; both P < 0.001) and at least 1 emergency department visit (commercial: 37.3% vs 3.4%; Medicare: 26.2% vs 4.1%; both P < 0.001). Total health care costs in month 1 were significantly higher among patients with COVID-19 than controls (mean differences: $3,706 for commercial; $10,595 for Medicare; both P < 0.001), driven by inpatient costs. Though the incremental HRU and cost burden of COVID-19 decreased over time, patients with COVID-19 continued to have significantly higher total costs through month 5 (all P < 0.001 for both commercial and Medicare). During follow-up, patients with COVID-19 had significantly higher rates of complications than controls (commercial: 52.8% vs 29.0% with any; Medicare: 74.5% vs 47.9% with any; both P < 0.001), most commonly cough, dyspnea, and fatigue.
CONCLUSIONS: COVID-19 was associated with significant economic and clinical burden, both in the short-term and over 6 months following diagnosis.
DISCLOSURES: Jessica K DeMartino is an employee of Janssen Scientific Affairs, LLC. Elyse Swallow, Debbie Goldschmidt, Karen Yang, Marta Viola, Tyler Radtke, and Noam Kirson are employees of Analysis Group, Inc., which has received consulting fees from Janssen Scientific Affairs, LLC.
This study was funded by Janssen Scientific Affairs, LLC. The sponsor was involved in the study design, interpretation of the results, manuscript review, and the decision to publish the article.
Plain language summary
This study assessed the burden of COVID-19 in the United States using health insurance claims data and compared patients with vs patients without a COVID-19 diagnosis. The findings showed that patients with COVID-19 had more medical visits, incurred higher health care costs, and experienced more complications than those without COVID-19. The burden of COVID-19 was the highest in the month after diagnosis but persisted for several months. These results can be used to inform policy decisions.
Implications for managed care pharmacy
This study quantified the excess burden of COVID-19 from the payer’s perspective and found that patients with COVID-19 had more health care resource utilization, health care costs, and complications than those without COVID-19. The incremental burden persisted for several months after the acute disease phase, a finding that can inform the design of clinical programs addressing long COVID-19. These results can be used in economic models that evaluate COVID-19 treatments or vaccines and public health policies.
Since the first case of novel COVID-19 was reported in the United States in January 2020,1 as of May 2022, more than 81 million people have been infected, with tens of thousands of new cases reported daily.2 The human toll of COVID-19 has been enormous, with approximately 990,000 deaths in the United States, the most of any country,2,3 and a decrease in life expectancy of 1.13 years estimated for 2020.4
COVID-19 has a broad clinical spectrum, and it is estimated that around 35% of cases were asymptomatic5 but still infectious,6 based on a variety of studies (not specific to a variant) before widespread vaccination. In studies characterizing early COVID-19 cases in January to February 2020, most patients (~80%) presented with mild symptoms,7 typically fever, cough, fatigue, anorexia, and dyspnea.8 Severe illness involved multiple organ dysfunction, respiratory impairment, sepsis, and death.9 Even after recovery from the acute phase of illness, a sizeable proportion of patients continued to experience sequelae, including fatigue, headache, attention disorder, hair loss, shortness of breath, and other respiratory, cardiovascular, and neurologic symptoms that may have lasted weeks or months10-12 and may have required intensive treatment.13,14
The medical costs of COVID-19 have been investigated in real-world data from a large, multihospital US database (ie, Premier Healthcare Database),15-17 commercial claims,18 and Medicare fee-for-service claims;19 however, these studies were often limited in scope (eg, to hospitalized patients or older adults) and did not examine long-term costs and complications associated with COVID-19. This study aimed to assess the incremental health care burden of COVID-19 in the United States by comparing health care resource use (HRU), direct health care costs, and long-term complications of COVID-19 between patients with vs patients without COVID-19 diagnoses.
Methods
DATA SOURCE
This study used real-world administrative claims data from the following IBM MarketScan Research databases:
(1)
Commercial Claims and Encounters (commercial) and
(2)
Medicare Supplemental and Coordination of Benefits (Medicare) (January 1, 2018, to March 1, 2021). The commercial database contains employer-sponsored private health insurance data of more than 200 million employees, spouses, and dependents, and the Medicare database includes beneficiaries with Medicare supplemental insurance paid by employers in all US census regions. All data were deidentified and complied with Health Insurance Portability and Accountability Act.
STUDY DESIGN
A retrospective, observational, matched cohort study design was used to assess the differences between 2 mutually exclusive groups: patients diagnosed with COVID-19 and subjects without a COVID-19 diagnosis (controls). Patients with COVID-19 diagnoses were identified using the International Classification of Diseases, Tenth Revision, Clinical Modification (ICD-10-CM) code for COVID-19 (U07.1), introduced on April 1, 2020. For claims between January 1, 2020, and March 31, 2020, the ICD-10-CM code for “other coronavirus as the cause of diseases classified elsewhere” (B97.29) was used to identify patients with COVID-19, per Centers for Disease Control and Prevention recommendations.20,21 IBM MarketScan conducted a preliminary match of patients with COVID-19 to control subjects based on region of residence, sex, and birth year.
Index Date, Baseline, and Follow-up Period. The index date for patients with COVID-19 was the date of the first observed claim with a COVID-19 diagnosis (or diagnosis for personal history of COVID-19 or pneumonia due to COVID-19; codes in Supplementary Table 1, available in online article) from January 1, 2020, onwards. For patients diagnosed with COVID-19 during a hospitalization, the date of admission was defined as the index date. The study period for the control group was limited to the prepandemic period (ie, 2019), given that HRU and costs for patients without COVID-19 in the pandemic period (2020) may have underestimated typical health care utilization, which would lead to overestimation of the excess burden of COVID-19. Therefore, the index date for controls was assigned as exactly 1 year prior to the index date of their preliminary COVID-19 matches.
For both groups, the baseline period was defined as the 6 months prior to the index date (exclusive), and the follow-up period was defined as the 6 months after the index date (inclusive).
Study Population. Patients were required to be continuously enrolled in a health plan for the 6-month baseline and 6-month follow-up periods. However, patients who were censored during the last month of follow-up were still included in the sample. Additional inclusion criteria were as follows: for patients with COVID-19, at least 1 claim with a COVID-19 diagnosis code between January 1, 2020, and 1 month before the end of data availability; and for controls, no diagnosis codes for COVID-19, personal history of COVID-19, or pneumonia due to COVID-19 (codes in Supplementary Table 1) observed at any time. COVID-19 vaccination status was not considered.
Propensity Score Matching of Study Subjects. Separately for commercial and Medicare cohorts, COVID-19 and control patients were matched 1:1 based on propensity scores. The propensity score model included an indicator for COVID-19 vs control as the dependent variable and index month (to control for seasonality), age on the index date, sex, region of residence, insurance plan type, Elixhauser Comorbidity Index (ECI),22 and total baseline health care costs as independent variables. Additionally, COVID-19 and control patients were matched exactly on number of follow-up months and a binary indicator of age 18 years or older.
ANALYZED VARIABLES
Demographic characteristics were evaluated on the index date; ECI, comorbidities (codes in Supplementary Table 1), and health care costs were evaluated during the 6-month baseline period. Monthly all-cause HRU (codes for select outcomes in Supplementary Table 2) and health care cost (pharmacy costs and inpatient [IP], outpatient [OP], emergency department [ED], and other medical costs) outcomes were evaluated in each month of follow-up. Health care costs from the payer’s perspective were inflated to 2020 US dollars (USD) using the medical care component of the US Consumer Price Index. For each month of follow-up, the presence of a diagnosis code for any of the following long-term COVID-19–related complications was reported: fatigue, dyspnea, chest pain or discomfort, joint pain, cough, anxiety, depression, posttraumatic stress disorder, cognitive symptoms, and cardiac injury (codes in Supplementary Table 1).
DATA ANALYSIS
Analyses were performed separately for commercial and Medicare cohorts using SAS Enterprise Guide 7.1 (SAS Institute Inc.) and R software (R Foundation for Statistical Computing).
Description of Baseline Characteristics. Demographic and baseline characteristics were compared between patients with COVID-19 and controls. Continuous variables were described using mean and SD; categorical variables were described using frequency count and percentage. Absolute standardized mean differences (SMDs) were calculated as the absolute difference in the mean (for continuous variables) or percentage (for binary variables) between patients with COVID-19 and controls divided by the pooled SD. The balance of baseline characteristics between COVID-19 and control groups was assessed using the SMD before and after matching, with SMD less than or equal to 0.1 considered well balanced.23
Comparisons of HRU, Health Care Costs, and Long-Term Complications of COVID-19 During Follow-up. The proportions of patients with each HRU category and long-term complications were described for each month of follow-up. Health care costs were described using the mean. Statistical comparisons between patients with COVID-19 and controls were performed with the Wilcoxon rank-sum test for continuous variables and chi-square test for categorical variables. For categorical variables with counts less than 5, Fisher exact test was used.
Stratification Analysis. For the commercial cohort, analyses were repeated in patient subgroups stratified by age at the index date (0-17, 18-49, and 50-64 years).
Results
PATIENT CHARACTERISTICS
After matching COVID-19 and control patients 1:1, there were 150,731 pairs in the commercial cohort and 1,862 pairs in the Medicare cohort (Supplementary Figure 1). Nearly all baseline characteristics were balanced between COVID-19 and control subjects (SMD ≤ 0.1) (Table 1). Patients with COVID-19 and controls had mean ages of 38.9 and 39.7 years in the commercial cohort and 74.3 and 75.3 years in the Medicare cohort, respectively. Commercial patients had a mean ECI of 0.3 for both COVID-19 and control groups; the most common conditions that increased risk of severe illness from COVID-19 included asthma (COVID-19: 5.1%, control: 3.8%), heart conditions (COVID-19: 2.3%, control: 2.1%), and smoking (COVID-19: 2.2%, control: 2.7%). In the Medicare cohort, the mean ECI was 3.8 for patients with COVID-19 and 4.3 for controls, and common risk factors included heart conditions (COVID-19: 21.5%, control: 22.6%), chronic kidney disease (COVID-19: 11.3%, control: 10.4%), cerebrovascular disease (COVID-19: 9.0%, control: 7.4%), and chronic obstructive pulmonary disease (COVID-19: 8.7%, control: 7.7%). Mean baseline health care costs were $3,664 for COVID-19 and $3,879 for controls in the commercial cohort and $9,502 for COVID-19 and $7,899 for controls in the Medicare cohort. Notably, prior to matching in the Medicare cohort, patients who were later diagnosed with COVID-19 tended to have more comorbidities and higher health care costs at baseline than control subjects (mean ECI: 4.2 vs 3.0; mean total costs: $11,704 vs $5,946).
| Commercial | Medicare | |||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Before matching | After matching | Before matching | After matching | |||||||||
| COVID-19 N = 150,834 | Control N = 384,273 | SMD | COVID-19 N = 150,731 | Control N = 150,731 | SMD | COVID-19 N = 1,947 | Control N = 4,568 | SMD | COVID-19 N = 1,862 | Control N = 1,862 | SMD | |
| Demographic information at index date | ||||||||||||
| Age, year | ||||||||||||
| Mean ± SD (median) | 38.9 ± 15.2 (40) | 39.0 ± 15.5 (41) | 0.004 | 38.9 ± 15.2 (40) | 39.7 ± 15.4 (42) | 0.049 | 74.3 ± 8.7 (71) | 75.8 ± 8.5 (74) | 0.179 | 74.3 ± 8.8 (71) | 75.3 ± 8.5 (73) | 0.118 |
| Pediatric patients, n (%) | 11,358 (7.53) | 35,659 (9.28) | 0.063 | 11,321 (7.51) | 11,321 (7.51) | 0 | — | — | — | — | — | — |
| Sex, n (%) | ||||||||||||
| Male | 67,383 (44.67) | 174,742 (45.47) | 0.016 | 67,337 (44.67) | 63,914 (42.40) | 0.046 | 879 (45.15) | 1,977 (43.28) | 0.038 | 837 (44.95) | 858 (46.08) | 0.023 |
| Female | 83,451 (55.33) | 209,531 (54.53) | 83,394 (55.33) | 86,817 (57.60) | 1,068 (54.85) | 2,591 (56.72) | 1,025 (55.05) | 1,004 (53.92) | ||||
| Region, n (%) | ||||||||||||
| North Central | 25,092 (16.64) | 71,786 (18.68) | 0.1 | 25,083 (16.64) | 27,433 (18.20) | 0.045 | 653 (33.54) | 2,008 (43.96) | 0.271 | 642 (34.48) | 612 (32.87) | 0.054 |
| Northeast | 34,230 (22.69) | 73,187 (19.05) | 34,183 (22.68) | 33,029 (21.91) | 558 (28.66) | 1,141 (24.98) | 542 (29.11) | 528 (28.36) | ||||
| South | 73,319 (48.61) | 195,041 (50.76) | 73,284 (48.62) | 71,603 (47.50) | 584 (29.99) | 1,266 (27.71) | 564 (30.29) | 591 (31.74) | ||||
| West | 18,186 (12.06) | 44,203 (11.50) | 18,175 (12.06) | 18,661 (12.38) | 152 (7.81) | 151 (3.31) | 114 (6.12) | 131 (7.04) | ||||
| Unknown | 7 (0.00) | 56 (0.01) | 6 (0.00) | 5 (0.00) | 0 (0.00) | 2 (0.04) | 0 (0.00) | 0 (0.00) | ||||
| ECIa | ||||||||||||
| Mean ± SD (median) | 0.3 ± 3.2 (0) | 0.2 ± 2.7 (0) | 0.048 | 0.3 ± 3.2 (0) | 0.3 ± 3.1 (0) | 0.004 | 4.2 ± 7.5 (0) | 3.0 ± 6.1 (0) | 0.187 | 3.8 ± 6.9 (0) | 4.3 ± 7.0 (0) | 0.06 |
| Conditions with increased risk of severe COVID-19, n (%) | ||||||||||||
| Asthma | 7,760 (5.14) | 11,654 (3.03) | 0.107 | 7,746 (5.14) | 5,691 (3.78) | 0.066 | 108 (5.55) | 175 (3.83) | 0.081 | 100 (5.37) | 86 (4.62) | 0.035 |
| Cerebrovascular disease | 1,200 (0.80) | 2,184 (0.57) | 0.028 | 1,184 (0.79) | 1,092 (0.72) | 0.007 | 180 (9.24) | 286 (6.26) | 0.112 | 167 (8.97) | 138 (7.41) | 0.057 |
| Chronic kidney disease | 1,614 (1.07) | 2,359 (0.61) | 0.05 | 1,591 (1.06) | 1,222 (0.81) | 0.025 | 238 (12.22) | 392 (8.58) | 0.12 | 210 (11.28) | 193 (10.37) | 0.029 |
| Chronic obstructive pulmonary disease | 1,174 (0.78) | 2,145 (0.56) | 0.027 | 1,161 (0.77) | 1,041 (0.69) | 0.009 | 180 (9.24) | 302 (6.61) | 0.098 | 161 (8.65) | 143 (7.68) | 0.035 |
| Down syndrome | 30 (0.02) | 61 (0.02) | 0.003 | 30 (0.02) | 29 (0.02) | 0 | 0 (0.00) | 0 (0.00) | 0 | 0 (0.00) | 0 (0.00) | 0 |
| Heart conditions | 3,542 (2.35) | 6,273 (1.63) | 0.051 | 3,507 (2.33) | 3,110 (2.06) | 0.018 | 440 (22.60) | 842 (18.43) | 0.103 | 401 (21.54) | 420 (22.56) | 0.025 |
| Immunocompromised state from solid organ transplantation | 244 (0.16) | 366 (0.10) | 0.019 | 236 (0.16) | 207 (0.14) | 0.005 | 4 (0.21) | 9 (0.20) | 0.002 | 2 (0.11) | 5 (0.27) | 0.037 |
| Pregnancy | 2,954 (1.96) | 6,468 (1.68) | 0.021 | 2,949 (1.96) | 3,233 (2.14) | 0.013 | 0 (0.00) | 2 (0.04) | 0.03 | 0 (0.00) | 1 (0.05) | 0.033 |
| Sickle cell disease | 183 (0.12) | 207 (0.05) | 0.023 | 180 (0.12) | 93 (0.06) | 0.019 | 3 (0.15) | 1 (0.02) | 0.045 | 2 (0.11) | 1 (0.05) | 0.019 |
| Smoking | 3,342 (2.22) | 8,682 (2.26) | 0.003 | 3,331 (2.21) | 4,069 (2.70) | 0.032 | 86 (4.42) | 222 (4.86) | 0.021 | 78 (4.19) | 112 (6.02) | 0.083 |
| All-cause health care costs (2020 USD) | ||||||||||||
| Mean ± SD (median) | 3,910 ± 20,973 (519) | 2,817 ± 15,312 (266) | 0.06 | 3,664 ± 16,961 (518) | 3,879 ± 18,173 (419) | 0.012 | 11,704 ± 33,707 (2,525) | 5,946 ± 16,657 (1,498) | 0.217 | 9,502 ± 21,176 (2,441) | 7,899 ± 18,050 (2,043) | 0.081 |
Characteristics that were not summarized for a particular cohort were marked with an em dash, “—”.
avan Walraven modification of the Elixhauser comorbidity score was reported.38
ECI = Elixhauser Comorbidity Index; SMD = standardized mean difference; USD = US dollar.
HRU AND COSTS
Across payer cohorts, the incremental burden of COVID-19 was highest during the index month and decreased over the follow-up period. COVID-19–related costs were driven by IP costs in the first month, which decreased substantially thereafter (Figures 1-2 and Supplementary Figures 2A-2F).
HRU During the 6-Month Follow-Up. In the matched commercial cohort, HRU was significantly higher for patients with COVID-19 than for control patients in month 1, with 6.9% and 37.3% of patients with COVID-19 experiencing an IP admission and ED visit, respectively, compared with 0.5% and 3.4% of control patients (P < 0.001 for both comparisons) (Supplementary Figures 2A-2B and Supplementary Table 3A). Severe events, such as intubation or ventilator use, were rare but highest among patients with COVID-19 in month 1 (COVID-19: 0.5%, control: 0.01%, P < 0.001) (Supplementary Figure 2F and Supplementary Table 3A). The proportion of patients with COVID-19 with IP admissions or ED visits decreased more than 80% from month 1 to month 2 (month 2 IP: 0.6%, ED: 6.2%), whereas the proportion of controls with these types of visits stayed relatively steady (month 2 IP: 0.4%, ED: 3.3%) (Supplementary Figures 2A-2B and Supplementary Table 3A). Compared with controls, patients with COVID-19 were significantly more likely to have at least 1 IP admission through month 4 (all P < 0.05) and at least 1 ED visit through month 6 (all P < 0.001), though the magnitude of the difference between groups after month 1 was small (< 0.2% for IP, < 3% for ED). Of patients with COVID-19, 82.6% had an OP visit in month 1; this dropped to 44.7% in month 2 and decreased steadily to 35.4% in month 6 (Supplementary Figure 2C and Supplementary Table 3A). Meanwhile, the proportion of controls with an OP visit remained steady at 34.2-36.1% across months 1-6. Mental health and rehabilitation visits were relatively rare (< 7%), and the magnitude of the difference between COVID-19 and control groups was minimal (Supplementary Figures 2D-2E). By month 6, HRU rates were numerically similar between COVID-19 and control groups.
These trends largely applied to the Medicare cohort as well. Among patients in the COVID-19 group, 29.1% had an IP admission in month 1; this dropped to 1.8% in month 2 and fluctuated between 1.4% and 1.9% through month 6 (Supplementary Figure 2A and Supplementary Table 3B). Meanwhile, between 1.3% and 1.9% of controls had an IP admission per month. ED visits followed a similar pattern, with 26.2% of patients with COVID-19 and 4.1% of controls having an ED visit in month 1 (Supplementary Figure 2B and Supplementary Table 3B). The rate of ED visits for patients with COVID-19 dropped to 6.3% in month 2 and steadily decreased to 4.1% in month 6. For the control group, ED visits remained relatively steady at 3.5%-4.1% per month. Though the proportion of patients with mental health and rehabilitation visits was small (patients with COVID-19 in month 1: 3.5% with mental health visits and 9.3% with rehabilitation visits), the COVID-19 group was significantly more likely to have mental health visits through month 5 (all P<0.01) and rehabilitation visits through month 4 (all P < 0.01) compared with controls (Supplementary Figures 2D-2E and Supplementary Table 3B).
Costs During the 6-Month Follow-Up. In the matched commercial cohort, total direct health care costs averaged over the 6 months of follow-up were significantly higher for patients with COVID-19 than for controls (mean monthly cost: $1,410 vs $647; P < 0.001) (Supplementary Table 4A). In month 1, mean total health care costs for patients with COVID-19 were $4,366 compared with $660 for controls, a difference of $3,706 (Figure 1). Total health care costs for the patients with COVID-19 then decreased by more than 75% to $987 in month 2, after which there was an average monthly decline of about 10% to $598 in month 6. Mean total health care costs for the controls remained relatively steady, at $602-$667 across all 6 months. Overall, mean total health care costs for COVID-19 patients remained significantly higher than that for controls until month 5 (all P < 0.001). By month 6, the magnitude of the difference between patients with COVID-19 and controls was minimal ($598 vs $602). For patients with COVID-19, total health care costs were mainly driven by mean IP costs, which were $3,170 in month 1 and decreased to $356 in month 2. Mean ED costs showed a similar trend, albeit with a lower magnitude ($476 in month 1 to $50-$74 in subsequent months). OP costs started at $540 in month 1 and declined over time to $271 in month 6. In contrast, pharmacy costs remained relatively constant. For control subjects, health care costs across all cost categories remained relatively steady over the 6-month follow-up.
Health care costs followed a similar trend within each age group, with higher relative costs of COVID-19 in month 1 that decreased over the 6-month follow-up period (Supplementary Figure 3). Incremental burden increased with age, with mean differences of total health care costs ranging from $1,421 in month 1 for patients younger than 18 years (COVID-19: $1,748 vs control: $327) to $2,309 for patients aged 18-49 years (COVID-19: $2,766 vs control: $457) and $7,117 for patients aged 50-64 years (COVID-19: $8,018 vs control: $901).
In the matched Medicare cohort, total direct health care costs averaged over the first 6 months of follow-up were significantly higher for patients with COVID-19 than for controls (mean monthly cost: $3,597 vs $1,260; P < 0.001) (Supplementary Table 4B). In month 1, mean total health care costs for patients with COVID-19 were significantly higher than those for the controls ($11,828 vs $1,233; P < 0.001), a difference of $10,595 (Figure 2). Mean total health care costs for patients with COVID-19 then decreased by approximately 75% in the following month (to $2,935) and steadily declined by an average of 17% per month thereafter (from $2,311 in month 3 to $1,349 in month 6), whereas mean costs for the controls remained consistent at $1,139-$1,366 per month. Mean total health care costs remained significantly higher for patients with COVID-19 than for controls through month 5 (all P < 0.001). Mean IP and ED costs for patients with COVID-19 rapidly declined after the first month (IP: from $9,258 to $1,203; ER: from $369 to $74) but remained significantly higher than that for controls through month 2 for IP (all P < 0.001) and month 3 for ED (all P < 0.05). Mean OP costs of patients with COVID-19 remained significantly higher than that of controls through month 5 (all P < 0.001). Mean pharmacy costs of patients with COVID-19 remained relatively constant ($294 in month 1 and $223 in month 6) and similar to those of controls ($268 in month 1 and $278 in month 6).
LONG-TERM COMPLICATIONS
In the matched commercial cohort, 52.8% of patients with COVID-19 and 29.0% of controls had at least 1 of the following complications in the first 6 months of follow-up: fatigue, dyspnea, chest pain or discomfort, joint pain, cough, anxiety, depression, posttraumatic stress disorder, cognitive symptoms, and cardiac injury (Figure 3 and Supplementary Table 5A). Among patients with COVID-19, 37.9% had complications in month 1, which decreased to 10.6%-14.7% in subsequent months and remained significantly higher than the share of controls with any complications (consistently ~10%) across all 6 months (all P < 0.001). In the first month, the most frequently observed complications in patients with COVID-19 and controls were cough (19.3% and 0.7%, respectively), dyspnea (13.0% and 0.5%, respectively), and fatigue (7.5% and 1.3%, respectively).
In the matched Medicare cohort, 74.5% of patients with COVID-19 and 47.9% of controls had at least 1 complication in the first 6 months after the index date (Figure 4 and Supplementary Table 5B). In month 1, 58.2% of patients with COVID-19 had a complication, which decreased to 29.4% in month 2 and 20.0% in month 6 but remained significantly higher than the rate among controls through month 5 (all P < 0.001). After the first month, all complications were relatively infrequent (observed in < 10% of patients with COVID-19), except for fatigue, anxiety, and depression. These 3 complications occurred at significantly higher rates in patients with COVID-19 than in controls in month 6 (all P < 0.01), though the numeric difference between groups was small.
Discussion
The COVID-19 pandemic is an ongoing public health emergency that has placed substantial burdens on the US health care system. Previous studies quantifying the burden of COVID-19 have focused on specific patient populations (eg, hospitalized or Medicare patients) and have not assessed the incremental burden relative to controls without COVID-19 diagnoses.15-19 The present study addressed this gap in knowledge and focused on the direct HRU, health care costs, and long-term complications associated with COVID-19 diagnosis using both commercial and Medicare supplemental insurance claims data from the IBM MarketScan Research database. The results are largely generalizable to commercial and Medicare beneficiaries in the United States, as all census regions and approximately 350 private and public sector payers are represented in the claims data. The use of propensity score matching minimized differences in key baseline characteristics between patients with and without COVID-19 diagnoses, allowing for comparisons of the 2 groups. Our analyses show that the economic burden and risk of COVID-19–related complications are highest in the first month following diagnosis but persist for several months.
This study found a mean incremental burden of COVID-19 of $3,706 for patients with commercial insurance and $10,595 for patients with Medicare insurance in the first month following COVID-19 diagnosis, with the incremental burden decreasing over time but remaining significant through month 5. Other studies focused on patients hospitalized for COVID-19 and found median IP costs of $11,267-$12,046, with higher costs for patients requiring mechanical ventilation or an intensive care unit (ICU) stay.15,17 Another study described patients with COVID-19 covered by Medicare fee-for-service insurance and found that 92.6% of total COVID-19–related medical care costs were for hospitalizations, with an average cost per hospitalization of $21,752.19 Interestingly, that study found that the cost per COVID-19–related hospitalization decreased with patient age; the authors speculated that this could be due to higher mortality rates among older patients, leading to shorter hospital stays and lower costs. Alternatively, our study found that costs increased with age, consistent with other reports that advanced age (> 50 years) is the strongest risk factor for severe COVID-19 requiring ICU admission in the United States.24 Weiner et al recently used commercial claims data to describe the use of telehealth vs in-person OP visits during the pandemic but also reported monthly costs of patients with and without COVID-19; however, they did not perform statistical comparisons or adjust for differences in patient characteristics.18 Patients with suspected or diagnosed COVID-19 had mean total monthly medical costs of $1,338-$2,214, whereas patients without COVID-19 had $456-$736.18 In our study, however, commercially insured patients with COVID-19 incurred a mean of $1,242 in total monthly medical costs over the first 6 months of follow-up compared with $486 for controls (Supplementary Table 4A). Differences in medical costs compared with Weiner et al may be due to their study using allowed charges (not paid amounts), using different codes to identify COVID-19 cases (including COVID-19 screening, exposure, and testing), and relying only on data at the start of the pandemic (March to June 2020), when COVID-19 treatment protocols were still developing. Unlike previous studies, which were purely descriptive of COVID-19 costs, ours is the first article that we are aware of that quantifies the incremental burden of COVID-19 by conducting rigorous statistical comparisons between matched patients with COVID-19 in 2020 and those without COVID-19 in 2019.
There is increasing evidence that COVID-19 is associated with long-term sequelae requiring ongoing care13,14 and use of health care resources.25 Although the condition of “long COVID” is not fully understood,26 one prospective study of 177 patients diagnosed with COVID-19, most of whom had mild illness and did not require hospitalization, found that 30% had symptoms, including fatigue and loss of sense of taste or smell, after a median follow-up time of 5.6 months.27 Another study found reports of symptoms, such as fatigue, dyspnea, cough, lack of taste, and diarrhea, 35 days after discharge among patients hospitalized with COVID-19 in the United States.28 An analysis of electronic health records of 236,379 patients diagnosed with COVID-19 found increased risk of neurologic and psychiatric disorders in the 6 months after diagnosis compared with matched patients with influenza or other respiratory tract infections.29 The current study adds to the evidence on long COVID, showing that patients with COVID-19 had higher rates of complications, most commonly cough, dyspnea, and fatigue, over the 5-6 months after diagnosis than matched controls. In addition, the persistence of significantly higher mean costs among patients with COVID-19 5 months after their diagnosis indicate that at least some of these patients require additional care following the acute phase of COVID-19 illness.
LIMITATIONS
There are limitations to this study. This sample of patients with COVID-19 derived from claims data may not cover certain subpopulations experiencing COVID-19: those without a formal COVID-19 diagnosis, those unlikely to seek medical attention (eg, for long COVID-19 symptoms), and those lacking coverage from these specific payers or health insurance altogether. COVID-19 cases prior to April 1, 2020, (when the COVID-19 ICD-10-CM diagnosis code was released) may be underreported or miscoded in claims data. Given that this study described COVID-19 burden and complications over time, patients who died prior to the 6th month of follow-up were not included.
Second, the true burden of COVID-19 may be misestimated, though the directionality is unclear. The economic burden may be underestimated for various reasons. Claims occurring after June 30, 2020, were considered “Early View” before all claim types have been paid. In a sensitivity analysis, Early View data appeared similar to the completely adjudicated data; however, some claims that take longer to process (ie, IP stays) may be missing from the final month of data. The incremental burden of COVID-19 relative to controls may be underestimated, given the study’s conservative approach of shifting the controls’ follow-up period to before the pandemic. Moreover, since the matching process reduced differences between COVID-19 and control groups, the final matched sample for the Medicare cohort may exclude the sickest and most costly patients from the COVID-19 group and the healthiest subjects from the control group. In addition, this study only considered the direct health care costs available in administrative medical claims data. The total cost of COVID-19 includes indirect costs, such as those from delayed diagnosis or treatment of other conditions,30,31 as well as lost income, child care costs, and other unmeasured costs. Nevertheless, the economic burden may be overestimated, as this study was not able to identify patients with asymptomatic or mild COVID-19 who did not seek medical care and who would have had lower costs.
Third, because data beyond March 1, 2021, were not available at the time the analyses were conducted, the current study does not take into account the changing course of the disease and treatments since then. This includes the emergence of the delta and omicron variants of concern, development of better treatments, and approval and distribution of COVID-19 vaccines. These changes may impact the disease burden as our knowledge continues to evolve.
Finally, as with all observational retrospective studies using administrative claims data, the results may be confounded by characteristics that were not measured or adjusted for. Administrative claims data have inherent limitations, such as miscoding and omissions of diagnoses, prescriptions, and procedures. Though patients diagnosed with COVID-19 during hospitalization were assumed to have COVID-19 at admission, claims data cannot reliably distinguish between whether COVID-19 was the reason for admission or an incidental finding.
Conclusions
This study provided real-world evidence on the HRU, cost burden, and complications associated with COVID-19, both in the short-term and over the 6 months following diagnosis. Although vaccination has reduced transmission and mortality from COVID-19 in the United States,32-34 the COVID-19 pandemic continues to place a significant burden on the health care system, as new variants emerge35,36 and about 30% of the US population remains unvaccinated (as of November 2021).37 Our findings can inform policy decisions aimed at mitigating the economic burden of COVID-19 and aid planning for future public health emergencies.
ACKNOWLEDGMENTS
Medical writing assistance was provided by Janice Imai, an employee of Analysis Group, Inc.
REFERENCES
1.
Holshue ML, DeBolt C, Lindquist S, et al. First case of 2019 novel coronavirus in the United States. N Engl J Med. 2020;382(10):929-36.
2.
US Centers for Disease Control and Prevention. COVID data tracker. 2022. Accessed May 4, 2022. https://covid.cdc.gov/covid-data-tracker/#trends_dailycases
3.
World Health Organization. WHO coronavirus (COVID-19) dashboard. 2022. Accessed May 4, 2022. https://covid19.who.int/
4.
Andrasfay T, Goldman, N. Reductions in 2020 US life expectancy due to COVID-19 and the disproportionate impact on the Black and Latino populations. Proc Natl Acad Sci USA. 2021;118(5):e2014746118.
5.
Sah P, Fitzpatrick MC, Zimmer CF, et al. Asymptomatic SARS-CoV-2 infection: A systematic review and meta-analysis. Proc Natl Acad Sci USA. 2021;118(34):e2109229118.
6.
Oran DP, Topol, EJ. Prevalence of asymptomatic SARS-CoV-2 infection: A narrative review. Ann Intern Med. 2020;173(5):362-7.
7.
Epidemiology Working Group for NCIP Epidemic Response: Chinese Center for Disease Control and Prevention. The epidemiological characteristics of an outbreak of 2019 novel coronavirus diseases (COVID-19) in China. Zhonghua Liu Xing Bing Xue Za Zhi. 2020;41(2):145-51. ssn.0254-6450.2020.02.003
8.
World Health Organization. Living guidance for clinical management of COVID-19. 2021. Accessed November 23, 2021. https://www.who.int/publications/i/item/WHO-2019-nCoV-clinical-2021-2
9.
Mehta OP, Bhandari P, Raut A, et al. Coronavirus disease (COVID-19): Comprehensive review of clinical presentation. Front Public Health. 2021;8:582932.
10.
Ghosn J, Piroth L, Epaulard O, et al. Persistent COVID-19 symptoms are highly prevalent 6 months after hospitalization: Results from a large prospective cohort. Clin Microbiol Infect. 2021;27(7):1041.e1-4.
11.
Lopez-Leon S, Wegman-Ostrosky T, Perelman C, et al. More than 50 Long-term effects of COVID-19: A systematic review and meta-analysis. Sci Rep. 2021;11:16144.
12.
Nasserie T, Hittle M, Goodman SN. Assessment of the frequency and variety of persistent symptoms among patients with COVID-19: A systematic review. JAMA Netw Open. 2021;4(5):e2111417.
13.
Nalbandian A, Sehgal K, Gupta A, et al. Post-acute COVID-19 syndrome. Nat Med. 2021;27(4):601-15.
14.
Romero-Duarte Á, Rivera-Izquierdo M, Guerrero-Fernández de Alba I, et al. Sequelae, persistent symptomatology and outcomes after COVID-19 hospitalization: The ANCOHVID multicentre 6-month follow-up study. BMC Med. 2021;19(1):129.
15.
Di Fusco M, Shea KM, Lin J, et al. Health outcomes and economic burden of hospitalized COVID-19 patients in the United States. J Med Econ. 2021;24(1): 308-17.
16.
Icten Z, Livingstone I, Menzin J. PIN178 impact of selected comorbidities on healthcare resource utilization among hospitalized patients with COVID-19 in a US population. Value Health. 2020;23:S573.
17.
Ohsfeldt RL, Choong CK, Mc Collam PL, et al. Inpatient hospital costs for COVID-19 patients in the United States. Adv Ther. 2021;38(11):5557-95.
18.
Weiner JP, Bandeian S, Hatef E, Lans D, Liu A, Lemke KW. In-person and telehealth ambulatory contacts and costs in a large US insured cohort before and during the COVID-19 pandemic. JAMA Netw Open. 2021;4(3):e212618.
19.
Tsai Y, Vogt TM, Zhou F. Patient characteristics and costs associated with COVID-19-related medical care among Medicare Fee-for-Service beneficiaries. Ann Intern Med. 2021;174(8):1101-9.
20.
US Centers for Disease Control and Prevention. ICD-10-CM Official coding guidelines - supplement coding encounters related to COVID-19 coronavirus outbreak (effective: February 20, 2020). 2020. Accessed October 9, 2020. https://www.cdc.gov/nchs/data/icd/ICD-10-CM-Official-Coding-Gudance-Interim-Advice-coronavirus-feb-20-2020.pdf
21.
Kadri SS, Gundrum J, Warner S, et al. Uptake and accuracy of the diagnosis code for COVID-19 among US hospitalizations. JAMA. 2020;324(24):2553-4.
22.
Elixhauser A, Steiner C, Harris DR, et al. Comorbidity measures for use with administrative data. Med Care. 1998;36(1):8-27.
23.
Austin PC. Using the standardized difference to compare the prevalence of a binary variable between two groups in observational research. Commun Stat Simul Comput. 2009;38(6):1228-34. https://doi.org/10.1080/03610910902859574
24.
Kim L, Garg S, O’Halloran A, et al. Risk factors for intensive care unit admission and in-hospital mortality among hospitalized adults identified through the US coronavirus disease 2019 (COVID-19)-associated hospitalization surveillance network (COVID-NET). Clin Infect Dis. 2021;72(9):e206-14.
25.
Al-Aly Z, Xie Y, Bowe B. Highdimensional characterization of post-acute sequelae of COVID-19. Nature. 2021;594(7862):259-64.
26.
US Centers for Disease Control and Prevention. Post-COVID conditions. 2021. Accessed November 30, 2021. https://www.cdc.gov/coronavirus/2019-ncov/long-term-effects/index.html?CDC_AA_refVal=https%3A%2F%2Fwww.cdc.gov%2Fcoronavirus%2F2019-ncov%2Flong-term-effects.html
27.
Logue JK, Franko NM, McCulloch DJ, et al. Sequelae in adults at 6 months after COVID-19 infection. JAMA Netw Open. 2021;4(2):e210830.
28.
Jacobs LG, Gourna Paleoudis E, Lesky-Di Bari D, et al. Persistence of symptoms and quality of life at 35 days after hospitalization for COVID-19 infection. PLoS One. 2020;15(12):e0243882.
29.
Taquet M, Geddes, JR, Husain, M, et al. 6-Month neurological and psychiatric outcomes in 236 379 survivors of COVID-19: A retrospective cohort study using electronic health records. Lancet Psychiatry. 2021;8(5):416-27.
30.
Kaczorowski J, Del Grande C. Beyond the tip of the iceberg: Direct and indirect effects of COVID-19. Lancet Digit Health. 2021;3(4):e205-6.
31.
Sabetkish N, Alireza Rahmani A. The overall impact of COVID-19 on healthcare during the pandemic: A multidisciplinary point of view. Health Sci Rep. 2021; 4(4):e386.
32.
Christie A, Henley, SJ, Mattocks L, et al. Decreases in COVID-19 cases, emergency department visits, hospital admissions, and deaths among older adults following the introduction of COVID-19 vaccine – United States, September 6, 2020–May 1, 2021. MMWR Morb Mortal Wkly Rep. 2021;70(23):858-64.
33.
Griffin JB, Haddix M, Danza P, et al. SARS-CoV-2 infections and hospitalizations among persons aged ≥ 16 years, by vaccination status – Los Angeles County, California, May 1–July 25, 2021. MMWR Morb Mortal Wkly Rep. 2021;70(34):1170-6.
34.
Tenforde MW, Self, WH, Naioti, EA, et al. Sustained effectiveness of Pfizer-BioNTech and Moderna vaccines against COVID-19 associated hospitalizations among adults – United States, March-July 2021. MMWR Morb Mortal Wkly Rep. 2021;70:1156-62.
35.
Murray CJL, Piot P. The potential future of the COVID-19 Pandemic: Will SARS-CoV-2 become a recurrent seasonal infection? JAMA. 2021;325(13):1249-50.
36.
World Health Organization. Tracking SARS-CoV-2 variants. 2021. Accessed November 30, 2021. https://www.who.int/en/activities/tracking-SARS-CoV-2-variants/
37.
Mayo Clinic. US COVID-19 vaccine tracker: See your state’s progress. 2021. Accessed November 29, 2021. https://www.mayoclinic.org/coronavirus-covid-19/vaccine-tracker
38.
van Walraven C, Austin PC, Jennings A, Quan H, Forster AJ. A modification of the Elixhauser comorbidity measures into a point system for hospital death using administrative data. Med Care. 2009;47(6):626-33.
Information & Authors
Information
Published In
Journal of Managed Care & Specialty Pharmacy
Volume 28 • Number 9 • September 2022
Pages: 936 - 947
PubMed: 35722829
Copyright
Copyright © 2022, Academy of Managed Care Pharmacy. All rights reserved.
History
Published online: 18 June 2022
Published in print: September 2022
Authors
Metrics & Citations
Metrics
Citations
Export citation
Select the citation format you wish to export for this article or chapter.