Using the electronic health record as an innovative approach to measure delirium in older hospitalized patients

• Main Authors: Khan, Ariba, Simpson, Michelle R, Singh, Maharaj, Malone, Michael L, Hook, Mary L
• Format: Text
• Language: unknown
• Published: Advocate Aurora Health Institutional Repository 2014
• Subjects: delirium; diagnosis; electronic health record; EHR; Acute Care for Elders (ACE) Tracker; Advocate Aurora Research Institute; Morse; Braden’; Aurora Research Institute; morse
• Online Access: https://institutionalrepository.aah.org/geri/12

Background: “Acute Care for Elders (ACE) Tracker” is an automated checklist generated by the electronic health record (EHR) and used to identify geriatric conditions among older hospitalized patients. Purpose: The aim of this study was to examine the validity of the EHR in detecting delirium. Methods: Hospitalized adults 65 years and older were included in this cross-sectional study. The researchers utilized “confusion assessment method” as the gold standard. “Delirium marker” on the ACE Tracker was defined as presence of any of the following: delirium symptoms (by nurse) or use of physical restraints or delirium treatment (quetiapine, haloperidol, olanzapine and risperdone). The performance of the delirium tool was evaluated using sensitivity, specificity, positive and negative predictive values, and likelihood ratios from simple 2 × 2 tables. Results: Ninety-two participants in three hospitals were included. Of these, 54% were female; mean age was 77 ± 8.8 years, mean length of hospitalization was 5.9 ± 5.1 days, and mean number of scheduled medications was 11.6 ± 4.3. Overall, 70% of individuals had a Morse score > 45; the mean ADL score was 10, and the mean Braden score was 17.3. Dementia was present in 10% of participants. Delirium symptoms were present in 5.4%, delirium treatment 11% and restraints 4.3%. The prevalence of delirium marker was 16% by EHR and 17% by researchers. The delirium marker had a sensitivity of 44%, specificity of 89%, positive predictive value of 47% and negative predictive value of 88%. The likelihood ratios for positive and negative tests were 4 and 0.6, respectively. Limitations: The researchers examined the patient only once. To ensure accurate assessment of fluctuation in status, the researchers interviewed the nurses and, if needed, the physician and family. There may have been a change in mental status between the time participant was assessed by the researchers and documentation of delirium by the EHR. Conclusion: The EHR is a useful tool to alert health care professionals ...