Among hospital workers who caught COVID-19, the people most likely to still be sick months later were not the doctors. They were the nurse assistants.
That is the central finding of new research in BMC Public Health, which analyzed occupational health registry data from 8,441 healthcare workers diagnosed with COVID-19 at three public hospitals between 2020 and 2023. At all three sites, nurse assistants most consistently showed higher long COVID rates than physicians.
The pattern matters because it points to something modifiable. If long COVID risk inside a hospital tracks job role rather than personal biology, then how protection, tasks, and patient contact are distributed across a workforce becomes a health question and not just a scheduling one.
What the Study Measured, and Where
The work was led by researchers at the Center for Research in Occupational Health at Universitat Pompeu Fabra in Barcelona with the Center for Biomedical Research Network of Epidemiology and Public Health in Madrid.
Rather than surveying workers, the team drew on occupational health registries, meaning records already maintained by the hospitals' own health services. That design avoids some of the recall problems that affect questionnaire studies, but it also means a worker only appears if the system captured them.
Estimated long COVID prevalence came in at 4.9 percent and 5.0 percent at two hospitals and 3.0 percent at the third. Age was the strongest and cleanest signal: each additional year of age was associated with a 3 to 9 percent increase in prevalence, with the highest rates among workers aged 50 and older.
Why These Numbers Are So Much Lower Than Earlier Estimates
Anyone who has followed long COVID research will notice that 3 to 5 percent sits far below what previous studies of healthcare workers reported, which ranged from 15 percent to more than 70 percent. A Swiss study published earlier this year assessing data through 2024 put the figure as high as 60 percent.
That spread is not a contradiction so much as a measurement problem. Long COVID has no single agreed clinical definition, and studies differ enormously in how they identify it. A registry-based analysis that counts workers whose condition was formally recorded by occupational health will capture fewer people than a survey asking workers whether they still have symptoms.
The practical reading is that this study is probably better at describing who is affected than how many. The occupational comparison is internally consistent because every worker in it was assessed the same way. The headline prevalence figure is much harder to compare against anything else.
What Might Explain the Gradient
The researchers did not test a mechanism, and they were careful about that. They wrote that the differences may reflect differences in exposure intensity, task allocation, duration of patient contact, physical workload, and working conditions.
Each of those is plausible, and each is a workplace design choice. Nurse assistants typically spend more cumulative time in close physical contact with patients than physicians do, performing bathing, repositioning, toileting, and feeding. That work is hands-on, sustained, and physically demanding, and it is frequently done by the staff with the least control over scheduling and the least seniority.
The sex-specific findings were more striking and much less stable. Among women at one hospital, nurse assistants were nearly three times as likely as physicians to have long COVID. Among men, male nurses at one hospital were more than twice as likely, and male nurse assistants at another hospital were more than 20 times as likely.
That last number should not be repeated without its caveat, which belongs here rather than at the end. The researchers noted it rests on relatively few cases and carried wide confidence intervals, meaning the true value could be far smaller. It is a signal worth investigating, not a measurement to quote.
What the Evidence Does and Does Not Establish
This is an observational cohort analysis. It can show that certain job roles had higher recorded long COVID rates. It cannot show that the job caused the outcome.
Several alternatives remain live. Nurse assistants and physicians differ in age distribution, sex composition, underlying health, vaccination timing, and the likelihood of seeking occupational health assessment at all. A worker whose job is physically demanding may be more likely to report persistent symptoms because the work makes those symptoms impossible to work around, while a role with more schedule flexibility may absorb the same impairment quietly.
The setting is also specific. These were three public hospitals in Spain during a period with particular staffing patterns, protective equipment supply, and vaccination rollout. US hospitals differ on all of those. The authors call for future studies using standardized definitions and data collection.
What This Means for US Hospitals and the Patients in Them
Nothing here changes clinical guidance, and no worker should draw a personal conclusion from a prevalence estimate.
What it does support is treating long COVID as an occupational health issue with a distribution, not a random misfortune. If persistent illness concentrates among lower-paid, hands-on staff, the downstream effects land on patients through the staffing shortages that follow. MedicalDaily has reported on how staffing ratios have driven nearly every major nurse labor dispute of the past decade, and chronic illness among frontline staff feeds directly into that pressure.
Healthcare workers with symptoms that have persisted for months after a COVID infection, including fatigue that does not improve with rest, breathlessness, or difficulty concentrating, should raise it with their own clinician and with occupational health rather than absorbing it. Documentation matters for workplace accommodations and for any future claim.
Older workers and those in high contact roles are the groups this study suggests deserve the closest follow-up, and staying current on recommended COVID vaccination remains the intervention with the best evidence behind it.
As for what comes next, the authors are calling for standardized definitions across studies so that occupational and sex-specific patterns can be compared rather than guessed at. Until that exists, prevalence figures in this field will continue to swing by an order of magnitude depending on method.
The confirmed finding is that nurse assistants at three Spanish hospitals had consistently higher recorded long COVID rates than physicians. The people most affected are hands-on hospital staff, particularly those over 50. The most reasonable action is for workers with lingering symptoms to seek formal assessment. The central uncertainty is whether the gradient reflects exposure at work or differences between the people who do these jobs.
Frequently Asked Questions
What did the study find? That among 8,441 hospital workers diagnosed with COVID-19 at three Spanish hospitals, nurse assistants consistently had higher long COVID rates than physicians, and rates rose with age.
How common was long COVID in this study? Estimated at 4.9 percent, 5.0 percent, and 3.0 percent at the three hospitals, well below earlier estimates that ranged from 15 percent to more than 70 percent.
Why do estimates vary so much? Because long COVID has no single agreed definition and studies identify it differently. Registry-based counts capture fewer people than symptom surveys.
Does this prove the job caused long COVID? No. This is observational research showing association. Differences in age, sex, health status, and likelihood of seeking assessment could contribute.
What about the 20-fold figure for male nurse assistants? It was based on a few cases with wide confidence intervals, meaning the real difference could be much smaller. It should be treated as a signal for further study.
Does this apply to US hospitals? Not directly. The study covered three public hospitals in Spain with their own staffing and protective equipment conditions.
What should a healthcare worker with lingering symptoms do? Raise it with a clinician and with occupational health, and keep documentation. This article is general information and not a diagnosis.
