Nursing documentation of symptoms is associated with higher risk of emergency department visits and hospitalizations in homecare patients.

Abstract Background Nurses often document patient symptoms in narrative notes. Purpose This study used a technique called natural language processing (NLP) to: (1) Automatically identify documentation of seven common symptoms (anxiety, cognitive disturbance, depressed mood, fatigue, sleep disturbance, pain, and well-being) in homecare narrative nursing notes, and (2) examine the association between symptoms and emergency department visits or hospital admissions from homecare. Method NLP was applied on a large subset of narrative notes (2.5 million notes) documented for 89,825 patients admitted to one large homecare agency in the Northeast United States. Findings NLP accurately identified symptoms in narrative notes. Patients with more documented symptom categories had higher risk of emergency department visit or hospital admission. Discussion Further research is needed to explore additional symptoms and implement NLP systems in the homecare setting to enable early identification of concerning patient trends leading to emergency department visit or hospital admission.