Delirium is under-reported in discharge summaries and in hospital administrative systems: a systematic review

Background Accurate recording of delirium in discharge summaries (DS) and hospital administrative systems (HAS) is critical for patient care. Objective To systematically review studies reporting the frequency of delirium documentation and coding in DS and HAS, respectively. Method We searched Medline, Embase, PsycINFO and Web of Science databases from inception to 23 June 2021. Eligibility criteria included requiring the term delirium in DS or HAS. Screening and full-text reviews were performed independently by two reviewers. Risk of bias (RoB) was assessed using the Effective Public Health Practice Project tool. Results The search yielded 7,910 results; 24 studies were included. The studies were heterogeneous in design and size (N=25 to 809,512). Mean age ranged from 57 to 84 years. Four studies reported only overall DS documentation and HAS coding in whole hospital or healthcare databases. Twenty studies used additional delirium ascertainment methods (e.g. chart review) in smaller patient subsets. Studies reported either DS figures only (N=8), HAS figures only (N=11), or both (N=5). Documentation rates in DS ranged from 0.1% to 64%. Coding rates in HAS ranged from 1.5% to 49%. Some studies explored the impact of race, and nurse versus physician practice. No significant differences were reported for race; one study reported that nurses showed higher documentation rates in DS relative to physicians. Most studies (N=22) had medium to high RoB. Conclusion Delirium is a common and serious medical emergency, yet studies show considerable under-documentation and under-coding in healthcare systems. This has important implications for patient care and service planning. Healthcare systems need to take action to reach satisfactory delirium documentation and coding rates.