Serum albumin and prealbumin as predictors of clinical outcomes of hospitalized elderly nursing home residents.

OBJECTIVE

To investigate the prevalence of hypoalbuminemia and hypoprealbuminemia in hospitalized, elderly, skilled nursing facility residents and to correlate these findings with clinical outcomes.

DESIGN

Prospective cohort study.

SETTING

A 300-bed community hospital.

PARTICIPANTS

Eighty-one hospitalized, skilled nursing facility patients, average age 83.1 years.

INTERVENTIONS

None.

OUTCOME MEASURES

Serum albumin and prealbumin (transthyretin) were measured at admission, mid-week, 1 week, and 1 month. Patients were followed for 90 days for the outcomes of length of hospitalization and mortality.

RESULTS

The prevalence of hypoalbuminemia was 99% and of hypoprealbuminemia, 79%. Both means dropped significantly from admission to midweek nadirs of 25 g/L for albumin and 14 mg/L for prealbumin. Severe hypoalbuminemia at mid-week predicted mortality (RR = 4.1 95%, CI 2.0-8.5) and extended length of hospitalization (RR = 5.2 95%, CI 2.8-9.8). Severe hypoprealbuminemia predicted extended hospitalization (RR = 3.2, CI 1.5-6.7) but not mortality.

CONCLUSIONS

Hypoalbuminemia and hypoprealbuminemia are very common in this clinical setting and vary in parallel fashion over time. Severe hypoalbuminemia was a stronger predictor than hypoprealbuminemia of 90-day mortality and extended length of stay. Serum albumin on admission was not as strong a predictor of outcomes as serum albumin at mid-week.