Covid-19 patients were at increased risk of psychiatric sequelae, and a psychiatric diagnosis appeared to be an independent risk factor for Covid-19, a retrospective analysis of more than 62,000 U.S. electronic health records suggested.
From 14 to 90 days after a Covid-19 diagnosis, the incidence of any psychiatric diagnosis was 18.1% (95% CI 17.6-18.6), including 5.8% (95% CI 5.2-6.4) that were a first diagnosis, reported Paul Harrison, FRCPsych, of the University of Oxford in England, and coauthors.
A psychiatric diagnosis in the previous year was associated with a higher incidence of Covid-19 diagnosis (RR 1.65, 95% CI 1.59-1.71; P<0.0001).
“Although preliminary, our findings have implications for clinical services, and prospective cohort studies are warranted,” Harrison and colleagues wrote in Lancet Psychiatry.
In patients with no previous psychiatric history, a Covid-19 diagnosis was associated with a higher incidence of a first psychiatric diagnosis compared with six other health events:
- HR 2.1 (95% CI 1.8-2.5) versus influenza.
- HR 1.7 (95% CI 1.5-1.9) versus other respiratory tract infection.
- HR 1.6 (95% CI 1.4-1.9) versus skin infection.
- HR 1.6 (95% CI 1.3-1.9) versus cholelithiasis.
- HR 2.2 (95% CI 1.5-1.9) versus urolithiasis.
- HR 2.1 (95% CI 1.5-1.9) versus large bone fracture.
“To our knowledge, this is the first dataset allowing the psychiatric sequelae and antecedents of Covid-19 to be measured reliably in terms of clinical diagnoses,” Harrison and colleagues noted. “The study uses propensity score matching to control for many variables, including established physical risk factors for Covid-19 and for more severe Covid-19 illness, and uses large-scale real-world data, thus providing more clinically relevant findings.”
Distinct from datasets based on national healthcare systems, data from available electronic medical records in the U.S. capture events that occur only in participating health-care organizations, observed Robert Yolken, MD, of Johns Hopkins University in Baltimore, in an accompanying editorial.
“Since the identity of participating healthcare organizations and their relative contributions to the dataset are not disclosed, the generalizability of data derived from this population is difficult to assess,” he wrote. “In this regard, although the 62,354 Covid-19 cases presented in this report is a large study population, they represent only a fraction of the number of cases reported in the USA during the same time period.”
Psychiatric symptoms may complicate medical care for Covid-19, and many people with significant mental illness have a high comorbidity burden that may further complicate Covid-19, including obesity, hypertension, smoking, and diabetes, and high exposure to respiratory viruses. A prior study using electronic medical records reported an increase in Covid-19 for people with a pre-existing psychiatric condition, with an overall risk of 1.65, Yolken noted.
In this analysis, Harrison and colleagues analyzed 62,354 Covid-19 cases from Jan. 20 and Aug. 1, 2020 in the TriNetX health network database, which contains anonymized data from hospitals, primary care, and specialist providers in the U.S. The average patient age was 49, and 55% were female.
The researchers created cohorts of patients who had been diagnosed with Covid-19 or a range of other health events and used propensity score matching to control for confounding by risk factors for Covid-19 and for severity of illness.
The most frequent new psychiatric diagnoses were forms of anxiety disorder (HRs 1.59-2.62, all P<0.0001) with 90-day probability of 4.7% (95% CI 4.2–5.3). Adjustment disorder and generalized anxiety disorder were the most frequent anxiety disorders.
Mood disorders (90-day probability 2%, 95% CI 1.7–2.4) were also seen, with depressive episode being most common.
A new diagnosis of psychotic disorder was uncommon (0.1%), similar to the risk associated with non-Covid-19 diagnoses. Insomnia (90-day probability 1.9%) and, in people older than 65, dementia (1.6%), were also observed.
The elevated risk of psychiatric sequelae after Covid-19 diagnosis versus control health events could not be readily explained by differences in illness severity, the researchers said. “Risks persisted in all sensitivity analyses,” they wrote.
“The changing nature of the Covid-19 pandemic presents a moving target for clinicians, investigators, readers of medical literature, and the general public,” Yolken pointed out. “Infection rates in different areas are frequently changing. Additionally, new cases, clinical data, and analytic functionalities are being added to available databases. Conclusions based on any one dataset thus require frequent re-examination and re-interpretation. The recent retraction of articles on Covid-19 based on another database highlights the necessity of data sharing and transparency.”
In addition, electronic medical record data often lack information relevant to Covid-19, Yolken noted. “These data include detailed information relating to housing density, family size, current employment and immigration status, specific geographic location, and contact with others with Covid-19. Therefore, it is imperative that data derived from electronic medical records be supported by cohort studies that prospectively collect relevant information and biological samples.”
With psychiatric disorders, “it is possible that the first entry of a diagnosis into the database might not represent the first occurrence of the condition, but rather the first time it is recognized by a health-care provider at a participating health-care organization, making the timing of symptom onset relating to Covid-19 difficult to evaluate,” he added.
- Covid-19 patients appeared to be at increased risk of psychiatric sequelae, and a psychiatric diagnosis appeared to be an independent risk factor for Covid-19, a retrospective analysis of more than 62,000 U.S. electronic health records suggested.
- The identity of participating healthcare organizations and their relative contributions to the dataset were not disclosed, and the generalizability of data derived from this population is difficult to assess, the editorialist said.
Paul Smyth, MD, Contributing Writer, BreakingMED™
This study was funded by the National Institute for Health Research.
Harrison reported no competing interests.
Yolken reported no competing interests.
Cat ID: 146
Topic ID: 87,146,730,933,190,926,192,927,146