Use of nutrition support in patients with AIDS: a four-year retrospective review.

A series of 55 patients with AIDS and opportunistic infections were admitted a total of 75 times to Robert Wood Johnson University Hospital over a 4-year period, and supplemental nutrition support--intravenous (IV), enteral, or both--was given during 32 of these admissions. Use of nutrition support was correlated retrospectively with pretreatment nutritional status, length of hospital stay (LOS), and survival and was found to be positively correlated with weight loss greater than or equal to 10% or weight less than or equal to 90% of ideal body weight (p less than 0.001), admission hemoglobin less than or equal to 10g (p less than 0.001), and LOS less than or equal to 21 days (p less than or equal to 0.003). Nutrition support intervention did not correlate with survival, admission total lymphocyte count (TLC), or serum albumin level. Survival was negatively correlated with LOS (p less than or equal to 0.04) and continuous daily fever for greater than or equal to 6 days (p less than 0.001). Survival was also significantly lower in patients who received IV rather than enteral nutrition support (p less than or equal to 0.03). Weight loss, admission TLC, albumin, and hemoglobin levels did not correlate with survival. These results suggest that nutrition support generally was given to the sickest patients with AIDS. There was no measurable benefit associated with use of supplemental nutritional support in this series. Properly designed trials will be necessary to define the optimum route, timing, and type of nutritional support for patients with AIDS.