Reducing Variance in Diagnosis of Mental Disorders to Improve Patient Safety - Summary/Abstract: Psychiatric diagnosis can be regarded as an inexact science due to a comparative absence of biomarkers and lab tests for psychopathology as compared to physical health conditions. Nevertheless, accurate psychiatric diagnosis is critical for patient safety, because misdiagnosis causes psychoactive medications to be misprescribed, which results in substantial patient morbidity and mortality every year. More generally, psychiatric diagnoses have been critiqued as themselves being constructs that are influenced by social, political, and cultural histories. This has resulted in a general trend toward overdiagnosis of mental disorders in a manner harmful to patients but beneficial to the manufacturers of pharmaceuticals Reducing variance in psychiatric diagnostic practice is one step toward solving these problems. Without consistency of diagnoses (i.e. precision in statistical terms), diagnostic accuracy cannot be achieved. Consistency/precision also can aid the community to recognize inaccuracy in diagnosis better than it can be recognized in a diagnostic context rife with variance. This proposed project seeks to reduce diagnostic variance and thereby reduce diagnostic error and improve patient safety by achieving the following Specific Aims: Aim 1. Assess the level of agreement across medical providers in their diagnostic processes and outputs for mental disorders. Aim 2. Determine the agreement between diagnostic practice by providers with (i) major diagnostic systems published in psychiatry literature (e.g. DSM-5, ICD-10, HiTOP, dimensional systems) and (ii) non-diagnostic systems that might still influence diagnostic practice (e.g. screening tools like PHQ-9, folk taxonomies for mental disorders). Aim 3. Disseminate diagnostic agreement and disagreement information back to study participants, who will then self-evaluate patient cases from their own caseloads about whom they were more or less certain of their original diagnosis. Aim 4. Disseminate more generalizable information on diagnostic agreement based on a national sample of physicians to the general scientific and medical communities. The project will be conducted on site at two health system partners in Massachusetts and Washington state (50 physicians at each site), and in an additional nationally recruited sample of 150 physicians.
