| INTRODUC TI ONAlbumin, which is the most important serum protein, has multiple physiological functions including maintenance of osmotic pressure, binding a variety of compounds and provision of the bulk of plasma antioxidant activity. 1 To date, several studies have shown the strong correlation between serum albumin and health outcomes. 2,3 Hypoalbuminemia has been shown to be a good predictive tool for several clinical outcomes such as hospital readmissions, functional recovery and mortality in patients with stroke, heart failure and cancer. [4][5][6][7] In addition, hypoalbuminemia has been associated with increased hospital stay and 30-day all-cause mortality in acutely admitted medical patients. 3,6,[8][9][10][11][12] Hospital readmission of medical patients comprises a significant proportion of all admissions, 13 ranging from 10% to 16% after 1 month and from 30% to 49% within 1 year. Slightly higher figures were reported for geriatric patients. 13 Disease-related characteristics also affect readmission rates, being higher for patients with heart failure, 14 chronic obstructive pulmonary disease 15 and cerebrovascular disease. 16 A better understanding of an individual's risk of death within 6-12 months after hospital stay would allow health care providers to properly allocate resources and possibly withhold unnecessary tests or procedures. This is perhaps particularly important in patients with hospital readmission.Previous studies have emphasised the prognostic significance of persistent hypoalbuminemia during hospitalisation 10 and after hospital discharge. 17 However, it is unclear what is the prognostic significance of persistent hypoalbuminemia for long-term mortality and whether normalisation of albumin levels after hospital discharge is associated with better prognosis. This study aimed Summary Aims: To determine the relationship between trends in admission serum albumin and long-term mortality in medical patients with hospital readmission.
Materials and methods:We used a cohort of patients admitted to five departments of internal medicine during 3 years. Survival analysis was performed based on mean admission serum albumin levels and trends in albumin values from recurrent admissions.
Results:A total of 5396 patients had 16 640 admissions (readmission cohort), another 9422 patients were admitted only once (single admission cohort). Readmitted patients with low mean albumin were older, predominantly females and had higher comorbidity index than patients with normal mean albumin. The 6-month all-cause mortality rate of the normal and low mean albumin groups was 5.2% and 24.2%, respectively (P < 0.001). Survival analysis showed that patients with persistently normal albumin levels had the highest survival rates at 6 months (97.7%), compared with patients who had hypoalbuminemia at index admission but normalised their albumin levels in subsequent admissions (92%), patients with declining albumin trends (85.6%) and patients with persistently low albumin levels (68.9%) (P < 0.0001).
Conclusions:Serum al...