[Utility of the ENEAS study screening guide in a long-stay chronic care hospital]

Introduction: Because of the characteristics of patients with multimorbidities, the methodology used in acute care hospitals is unsuitable to obtain valid results in long-stay chronic care hospitals. The present study aimed to determine the utility of the screening guide used in acute care hospitals to study adverse events (AE) in long-stay, chronic care hospitals and to identify the factors that could improve the adjustment of this screening guide to these hospitals.

Method: We performed a retrospective cohort study from January 1 to December 31, 2004 and cross-sectional studies in 2005, 2006, 2007 and 2008 in a long-stay chronic care hospital. The variables studied were AE and incidents. We used the AE screening guide adapted to the Harvard study and the Spanish version of the Modular Review Form (MRF2) for case record review. Data analysis consisted of calculation of the global predictive value of the screening guide, the item-by-item predictive value, and the percentage of AE detected by each alert.

Results: Of the 19 criteria in the screening guide, the most frequently marked criterion in both the cohort and prevalence studies was criterion number 1 (prior hospitalization during the last year in patients less than 65 years old or prior hospitalization in the last 6 months in patients 65 years old or older). Of all the AE and complications detected, 57.2% were screened with warning criteria 17 and 19, corresponding to "another AE" and "any kind of nosocomial infection". Overall, the screening guide had positive predictive value (PPV) of 24.3% in the cohort study, and a PPV of 37.4% in the cross-sectional study when all types of AE were included (i.e. AE associated with healthcare, disease and incidents). Screening criteria 10 (new neurological deficit at hospital discharge), 11 (acute myocardial infarction, stroke or pulmonary thromboembolism during or after an invasive procedure), 12 (cardiorespiratory arrest or low Apgar score) and 13 (injury or complication related to abortion, amniocentesis, delivery or pre-delivery) were not marked in any type of study.

Conclusions: The utility of the screening guide is low when used in a long-stay chronic care hospital. A new screening guide is required that excludes inefficient criteria and includes new, more specific criteria that take into account the characteristics of patients with multimorbidity.