An analysis of electronic health records from England suggests that diagnoses of several health conditions have not returned to expected levels more than four years after the start of the COVID-19 pandemic, according to a population-based cohort study published today in BMJ.
A team led by researchers from King’s College London used data from the National Health Service (NHS) OpenSAFELY network to track diagnosis rates for 19 long-term conditions from April 2016 through November 2024. They compared 2024 diagnosis levels with projected levels based on pre-pandemic trends and observed declines for several conditions.
The most pronounced decline was seen for depression, with diagnoses nearly one-third (27.7%) lower than expected. Other conditions with lower-than-expected diagnosis levels included asthma (16.4% below expected), chronic obstructive pulmonary disease (COPD; 15.8% below), psoriasis (17.1% below), and osteoporosis (11.5% below).
Multiple factors may be contributing to this trend, the authors speculate. Early in the pandemic, many patients avoided medical facilities owing to infection fears and delayed seeking care for less- urgent conditions. At the same time, many health care systems necessarily prioritized responding to COVID over routine care.
Low depression rates most evident in young adults
The decline in depression diagnoses was “particularly striking and somewhat puzzling,” said co-author Sam Norton, PhD, professor of medical statistics and applied health research at King’s College London, in a press release. “After an initial decrease during the early pandemic, diagnosis rates partially recovered by late 2021, but have declined markedly since 2022. This pattern was most evident among younger adults aged 20 to 39 years, and among individuals of white or mixed ethnicity.”
The authors speculate that the large cumulative decline in depression diagnoses may not indicate a decrease in the true incidence of depression but rather a delay in diagnosis due to lingering pandemic-related pressures on the health care system. “Although the total number of face-to-face [primary care] consultations has increased, so has the proportion of remote consultations, which together with time pressures in consultations, may make it more challenging to identify non-verbal cues associated with depression,” they write.
It’s also possible that more people with depressive symptoms are receiving care without a formal diagnosis. Referrals to NHS talk therapy services increased by nearly two-thirds from 2013 to 2024, with self-referrals (meaning no doctor referral or diagnosis was required) accounting for 69% of all referrals.
“The recent decrease in new diagnoses of depression warrants further investigation,” the authors write.
Large-scale analysis helps reveal gaps in care
The findings suggest that the pandemic’s effects on non–COVID-related health outcomes may be long-lasting, reflecting stresses on health care systems and changes in health-seeking behavior. The authors call for strategies to improve diagnostic capacity and to ensure that people with chronic conditions receive appropriate diagnosis and care.
The study also highlights the power of large-scale, anonymized data to reveal trends and gaps in health care. “Importantly, this study showed the potential for routinely collected health data to enable monitoring of disease epidemiology in near real time,” the authors write.
In an accompanying commentary, Mark Russell, MD, PhD, and James Galloway, PhD, of King’s College London argue that tracking care quality, time trends, and long-term conditions at scale in real time can transform care delivery.
“By enhancing our understanding of how common diseases are and how they vary by population and over time, we can better model workforce needs, target prevention efforts, and predict future demand. This enables policymakers and clinicians to act earlier, directing services and resources to where they are needed most,” they write.
