Depressed or Delirious? Overdiagnosis of Depression in Medically Hospitalized Patients

Abstract: Background/Significance:

Misdiagnosis of depression and delirium by non-psychiatric clinicians is common.^1,2 Depression over-diagnosis can cause stigmatization of normal emotional reactions and delays in addressing delirium.

In our multisite retrospective chart review, we hypothesized that the consultation-liaison psychiatry (CLP) service would frequently diagnose delirium and adjustment disorder in patients referred for depression. We postulated that depression is more often overidentified than delirium by non-psychiatrists and examined factors that may lead primary services to view a patient as depressed who is not truly depressed per the CLP service’s evaluation.

Methods:

We reviewed inpatient depression and delirium referrals across two Cleveland Clinic sites; one site was analyzed at the abstract submission date. Agreement between referral reason (depression or delirium) and CLP service diagnosis (depression, delirium, mania, anxiety, adjustment disorder, cognitive disorder, personality disorder, psychotic disorder, no psychiatric disorder) was assessed using raw percent agreement and Cohen’s Kappa, including a bootstrapping analysis to determine whether the Cohen’s kappa values were different. Among depression referrals, frequency and percent of final psychiatric diagnostic categories were computed. We fit multivariable logistic regression models for the dependent variables of overdiagnosis of depression and delirium masquerading as depression, with independent variables that included the primary team discipline (medical versus surgical) and demographic and clinical variables. The study was approved by the Cleveland Clinic Institutional Review Board.

Results:

The CLP service agreed that depression referrals (N=201) were depressed in 66.4% of cases (fair agreement) and agreed that delirium referrals (N=278) were delirious in 86.6% of cases (substantial agreement), a statistically significant difference. Of depression referrals, the top alternative diagnoses were adjustment disorder (51%), delirium (19%), anxiety/OCD (15.1%), and no psychiatric diagnosis (4.8%). Female sex increased the primary service’s overdiagnosis of depression, while older age and prior psychiatric diagnosis decreased the chance of depression overdiagnosis. Older age, however, increased the odds of the primary service misidentifying CLP-diagnosed delirium as depression.

Discussion:

Primary services overidentify depression and more accurately identify delirium. Prior psychiatric diagnoses such as depression increase the likelihood of future depressive episodes, likely increasing accurate depression identification. Female sex may contribute to depression overidentification due to higher levels of emotional expressivity. The correlation between older age and under-detection of delirium has been previously demonstrated.² Conversely, older age was found to decrease depression overdiagnosis, which suggests that primary services are better able to appreciate severe depressive symptoms constituting a major depressive episode (versus less severe depressive symptoms of adjustment disorder) in older adults. This is a novel finding that may relate to increased stoicism in elderly adults, leading to demonstrable depressive symptoms only when the patient is suffering with major depressive disorder.