Low Serum Total Protein at Admission Predicts in-Hospital Mortality Among General Inpatients: Historical Cohort Study

PURPOSE: Low total protein (TP) is associated with mortality among patients with specific diseases, but its association was uncertain among general patients. We evaluated the effects of low TP on in-hospital mortality among general inpatients. PATIENTS AND METHODS: We used data from the Japan Adverse Drug Events study series. We enrolled adult inpatients (≥ 16 years) admitted to a tertiary care hospital between September 1 and November 30, 2013. We excluded patients with multiple myeloma, pregnant women, and bone marrow graft donors. Patient data were extracted from electronic medical records. All patients were stratified into those with and without malignancy and divided into the low and normal TP groups. Low TP was defined as < 6.5 g/dL. We compared the in-hospital mortality of the low and normal TP groups stratified by the presence of malignancy. RESULTS: Among the 2235 enrolled patients (mean age, 67.8 years), the TP value was lower in 516 patients with malignancy than in 1719 patients without malignancy (6.6 g/dL vs 6.8 g/dL, P = 0.0002). Among patients without or with malignancy, 27% (462/1719) and 35% (183/516) were in the low TP group, respectively. Mortality was higher in the low TP group among patients without malignancy (23.2% vs 10.2%, P < 0.0001). Likewise, among patients with malignancy, mortality was higher in the low TP group (34.7% vs 11.3%, P = 0.0029). The adjusted hazard ratio of the low TP group was 1.75 (95% confidence interval, 1.04–2.96) in patients without malignancy and 2.45 (95% confidence interval, 1.12–5.37) in patients with malignancy, but the interaction was not significant (P = 0.23). CONCLUSION: Low TP values were associated with higher in-hospital mortality among general inpatients, and this association was observed among patients with and those without malignancy. Routinely measured TP should be utilized to risk stratification on admission.