Long COVID remains common across the healthcare workforce after SARS-CoV-2 infection
Long COVID persisted in a substantial share of healthcare workers at IRCCS Policlinico San Matteo Foundation, where an anonymous survey captured 771 employees with documented SARS-CoV-2 infection. The cumulative prevalence reached 27.9%, meaning more than one in four respondents experienced persistent symptoms at some point after acute infection. Even at 12 months, 16.0% still reported ongoing illness.
The pattern is clinically familiar and operationally disruptive. Fatigue, musculoskeletal pain, and cognitive complaints are not remote sequelae in an abstract cohort; they affect bedside endurance, concentration, and the ability to sustain complex clinical work across long shifts.
Symptom burden declined, but never disappeared
Follow-up at 3, 6, and 12 months showed a gradual fall in point prevalence: 26.0% at 3 months, 20.0% at 6 months, and 16.0% at 12 months. The median number of symptoms per person also eased from 3 at 3 months to 2 at both 6 and 12 months, suggesting partial recovery rather than uniform resolution.
The dominant symptom profile stayed remarkably stable. Fatigue led the list at every time point, followed by joint and muscle pain, concentration problems, and memory difficulties. Dyspnea faded over time, dropping from 4.5% at 3 months to 1.4% at 12 months, while dysgeusia, dysosmia, and sleep problems also became less frequent. The trajectory fits a syndrome in which inflammatory, autonomic, and neurocognitive disturbances improve unevenly, while energy failure and brain fog linger long after the respiratory phase has passed.
Nursing staff carried the heaviest burden
Job position mattered. Long COVID prevalence differed significantly across occupational groups, with nurses showing the highest prevalence at 35.0%. Other health professionals reached 38.0%, and administrative staff 32.0%. The association between occupation and persistent symptoms was statistically significant (_p_ = 0.004).
The bedside implication is straightforward. Healthcare roles combine repeated exposure to acute infection with physical strain, emotional load, and limited recovery windows. Even when the survey could not isolate one single occupational mechanism, the clustering of cases in nurses and other healthcare workers close to direct patient care suggests that post-viral disability may disproportionately erode the workforce segments most difficult to replace.
Multiple comorbidities doubled the prevalence signal
Comorbidity burden also shaped the outcome. Participants with two to five pre-existing conditions had a 49.0% prevalence of Long COVID compared with 24.0% among those without multiple comorbidities, a difference that was statistically significant (_p_ < 0.001). Hypertension, rheumatologic disorders, migraine, and obesity were the most frequent associated conditions.
This aligns with a broader biological rationale: baseline vascular, inflammatory, or neurologic vulnerability may lower the threshold for prolonged post-infectious dysfunction. In practical terms, healthcare workers who enter infection with a heavier chronic disease load may be less able to clear residual symptom burden, especially when fatigue and cognitive slowing are part of the dominant phenotype.
Cognitive and musculoskeletal symptoms defined the lingering phenotype
The symptom clusters that emerged at 3 months add another layer. Unsupervised learning methods identified three patient groups after dimensionality reduction with UMAP and clustering with DBSCAN. One cluster had minimal symptoms, another showed moderate respiratory and musculoskeletal burden, and a third carried the heaviest profile, dominated by concentration and memory difficulties, fatigue, joint pain, and dysthymia.
The clusters did not separate cleanly by age, gender, or medical history. That lack of neat demographic sorting matters because it suggests Long COVID cannot be reduced to a single occupational stereotype or one predictable clinical pathway. Instead, the syndrome appears to split into overlapping symptom burdens that may need different forms of workplace support, follow-up, and rehabilitation.
The survey design captured a wide window, but not every confounder
The survey drew on employees infected between February 2020 and December 2024 and was administered from late March to mid-May 2025, giving at least 3 months of follow-up to every participant and as much as 4 years for early pandemic infections. All analyses used standard statistical testing for prevalence comparisons and symptom evolution, and the study also performed a sensitivity analysis that treated all 1,091 non-responders as asymptomatic.
Under that conservative assumption, the prevalence fell to 11.5%. Even this lower-bound estimate leaves a sizable post-infectious burden among healthcare employees. The response rate was 41.41%, so non-response bias remains a meaningful consideration, especially in a workforce survey where those most affected may be more or less likely to answer.
Several potentially important variables were not captured in a way that allowed analysis: vaccination status, education level, income, acute COVID-19 severity, and specific workplace characteristics such as ventilation or ward versus office environment. Those gaps limit causal interpretation, but they do not erase the central finding that persistent symptoms were common, unevenly distributed, and slow to resolve.
Occupational health planning has to account for prolonged symptom tail
What this cohort makes clear is that Long COVID is not simply a recovery issue for individual workers. It is a staffing, scheduling, and occupational health issue. A workforce in which one-quarter of infected employees experience prolonged symptoms, and one in six still report illness at a year, has to absorb reduced stamina, concentration problems, and intermittent disability across clinical teams.
For hospitals, the relevant question is not whether the syndrome exists. The practical challenge is how to identify workers at highest risk early enough to provide support before persistent fatigue and neurocognitive symptoms translate into absences, reduced productivity, or unsafe workload compression. In this cohort, nurses, workers with multiple comorbidities, and those with heavy symptom burdens at 3 months stood out as groups warranting close attention, even as the syndrome gradually receded over time.