Documentation-derived nursing process indicators and in-hospital outcomes in patients with acute myocardial infarction undergoing PCI: A cohort study

In-hospital outcomes after acute myocardial infarction (AMI) treated with percutaneous coronary intervention remain variable. Documentation-derived nursing process indicators may act as proxy markers of clinical acuity, surveillance intensity, and care complexity, but their association with short-term outcomes remains unclear. This single-center retrospective cohort study included consecutive adults with AMI. Nursing process indicators were extracted from routine nursing records, including assessment frequency, monitoring intensity, documentation density, and selected observational items. The primary endpoint was a composite of all-cause in-hospital death, recurrent myocardial infarction, malignant arrhythmia, cardiogenic shock, major bleeding, or prolonged hospitalization. Multivariable logistic regression was used to estimate adjusted odds ratios (ORs). Intensive care unit admission was examined as a contextual care-setting variable but was not included in the final extended discrimination model. Internal model discrimination was assessed using receiver operating characteristic analysis and area under the curve comparison by DeLong test; no external validation was performed. Of 512 screened patients, 438 were included in the final analysis; the mean age was 61.4 years, and 73.5% were male. During hospitalization, 168 patients (38.4%) experienced the composite primary endpoint; all-cause in-hospital mortality occurred in 4.1%, and major adverse cardiovascular events occurred in 14.2%. After multivariable adjustment, higher nursing documentation density was associated with the composite endpoint (per additional record per day: adjusted OR: 1.29, 95% confidence interval [CI]: 1.08–1.55; P  = .005), as was documented cardiac rhythm monitoring (adjusted OR: 1.88, 95% CI: 1.25–2.84; P  = .002). In an internal discrimination analysis without external validation, the baseline clinical model showed an AUC of 0.71 (95% CI: 0.66–0.76), whereas the extended model including selected documentation-derived nursing process indicators showed an AUC of 0.76 (95% CI: 0.71–0.81; ΔAUC = 0.05; DeLong P  = .012). In this real-world cohort of patients with AMI undergoing percutaneous coronary intervention, documentation-derived nursing process indicators were associated with in-hospital adverse outcomes and provided modest incremental internal discriminative information beyond traditional clinical predictors.