Does COVID-19 raise healthcare use and costs for months afterwards?
After a COVID-19 diagnosis, adults kept using more healthcare and costing more each month for at least six months than they had before infection, while children returned close to their previous level.
Source
Long-haul COVID: healthcare utilization and medical expenditures 6 months post-diagnosis
Study at a glance
- Design
- Cohort — Retrospective within-patient before-after cohort using US insurance claims: monthly visits and costs in the 6 months before versus 6 months after a lab-confirmed COVID-19 diagnosis (February-October 2020), modelled with linear regression and monthly lead/lag terms, excluding pulmonary encounters.
- N
- N=250514 · 250,514 patients (from 3,006,168 medical claims) with six months of records before and after diagnosis; commercially insured, Medicare, Medicaid and uninsured patients.
- Population
- US patients of all ages with a lab-confirmed COVID-19 diagnosis between February and October 1, 2020, in a national claims database covering commercial, Medicare, Medicaid and HRSA-uninsured payers.
- Outcome
- Monthly healthcare visits and medical expenditures across ten service categories (inpatient, emergency, telemedicine, surgery, cardiology and others), overall and by age group.
Structured fields used in claim comparison tables when every cited study has a complete layer.
What they did
The researchers used de-identified US insurance claims to follow 250,514 people with lab-confirmed COVID-19 diagnosed between February and October 2020, comparing each person's monthly healthcare use and spending in the six months before diagnosis with the six months after. They grouped services into ten categories such as inpatient, emergency, telemedicine and cardiology, deliberately leaving out lung-related visits, and adjusted for age, sex, insurance type and local socio-economic factors. Month-by-month models then estimated 'excess' use and cost by pairing each post-diagnosis month with its matching pre-diagnosis month.
What they found
On average a COVID-19 diagnosis was associated with 0.7269 extra visits (roughly three-quarters of a visit) and $223.59 extra spending per month over the following six months, driven mainly by inpatient, emergency and surgical care, with smaller rises in telemedicine and cardiology. Use peaked in the diagnosis month and declined but did not return to pre-diagnosis levels; excess visits fell from 0.499 in the second month to 0.158 in the sixth. Children's excess use and costs essentially disappeared by month six, whereas all adult groups stayed above baseline, with people aged 45-64 having the largest excess costs.
The limits
What it doesn't show
There was no comparison group of people without COVID-19, so the authors stress the results are descriptive and not causal; rising use could reflect other health problems or pandemic-era changes in care. The claims come from providers, so some care is missing and estimates are lower bounds, and mortality was not recorded, so likely deaths were excluded, possibly biasing older groups. Lung-related visits were deliberately excluded, the sample was diagnosed early in the pandemic before vaccines, and the study cannot tell whether extra care was for long-COVID symptoms or pre-existing conditions.
Key terms
- Long COVID (post-acute sequelae)
- Symptoms and health problems that persist or appear weeks to months after the acute phase of a SARS-CoV-2 infection.
- Claims data
- Records of billed medical services submitted to insurers, showing what care was delivered and what it cost, but not detailed clinical findings.
- Excess utilisation
- Here, the difference in healthcare use between a given month after diagnosis and the matching month before diagnosis for the same patients.
- Before-after (pre-post) design
- Comparing the same people before and after an event; it controls for stable personal traits but not for other changes happening over the same time.
- Fixed effects
- Indicator variables (here for state and diagnosis month) that absorb differences between places or periods so they do not distort the main estimate.
Flashcards
0 of 11 answers reviewed
Research intelligence for this paper
See its role on concept claims, tensions it is part of, placement history, and related discoveries.
Quiz yourself
What was the comparison used to estimate COVID-19's association with healthcare use?
Common questions
Why compare people with themselves before diagnosis rather than with uninfected people?
Using each patient's own earlier months controls for stable differences such as chronic illness or habits; the trade-off is that without an uninfected comparison group, the study cannot separate COVID-19's effect from general pandemic changes in care.
Why did the researchers leave out lung-related visits?
So that obvious respiratory care would not dominate the totals, letting them see smaller, less expected increases in other kinds of care; it also means the estimates understate total post-COVID use.
Who had the biggest long-term cost increases?
Adults aged 45-64 had the largest excess costs in every month after diagnosis, while people over 65 had the largest short-term jump in visits.
More on Infectious disease