A comparison of delirium diagnosis in elderly medical inpatients using the CAM, DRS-R98, DSM-IV and DSM-5 criteria

Background: The recently published DSM-5 criteria for delirium may lead to different case identification and rates of delirium than previous classifications. The aims of this study are to determine how the new DSM-5 criteria compare with DSM-IV in identification of delirium in elderly medical inpatients and to investigate the agreement between different methods, using CAM, DRS-R98, DSM-IV, and DSM-5 criteria. Methods: Prospective, observational study of elderly patients aged 70+ admitted under the acute medical teams in a regional general hospital. Each participant was assessed within 3 days of admission using the DSM-5, and DSM-IV criteria plus the DRS-R98, and CAM scales. Results: We assessed 200 patients [mean age 81.1±6.5; 50% female; pre-existing cognitive impairment in 63%]. The prevalence rates of delirium for each diagnostic method were: 13.0% (n = 26) for DSM-5; 19.5% (n = 39) for DSM-IV; 13.5% (n = 27) for DRS-R98 and 17.0%, (n = 34) for CAM. Using tetrachoric correlation coefficients the agreement between DSM-5 and DSM-IV was statistically significant (ρ_tetr = 0.64, SE = 0.1, p < 0.0001). Similar significant agreement was found between the four methods. Conclusions: DSM-IV is the most inclusive diagnostic method for delirium, while DSM-5 is the most restrictive. In addition, these classification systems identify different cases of delirium. This could have clinical, financial, and research implications. However, both classification systems have significant agreement in the identification of the same concept (delirium). Clarity of diagnosis is required for classification but also further research considering the relevance in predicting outcomes can allow for more detailed evaluation of the DSM-5 criteria.