Background: To determine whether oral protein energy supplements, prescribe
d during hospitalization to elderly medical admissions, affect nutritional
status and if baseline nutritional state influences this status. We also co
nsidered the effects on mortality, length of hospital stay, functional reco
very, and institutionalization. Methods: A prospective randomized controlle
d trial with no placebo. Consenting patients were stratified in 3 nutrition
al categories, and patients from each stratum were randomized into treatmen
t or control. Observers were blinded to randomization. The participants wer
e emergency admissions from home to a Medicine for the Elderly Unit in a Sc
ottish hospital. The inclusion criteria were as follows: no known malignanc
y, the ability to swallow, and nonobesity (BMI < 75th percentile). The inte
rvention was a prescription of 120 mL sip feed, 3 times daily (540 kcal, 22
.5 g protein per day) throughout hospitalization, using the medicine prescr
iption chart. The trial was powered to detect change in mean percentage wei
ght. The following outcomes were also considered: anthropometry; mortality,
length of hospital stay, functional recovery, and rates of institutionaliz
ation. Results: Included in the trial were 381 patients. Nutritional supple
mentation was associated with significantly better energy intake (p = .001)
and weight gain (p = .003) pooled across all nutritional categories. In th
e most poorly nourished patients, the intervention was associated with redu
ced mortality (5/34 versus 14/40, p < .05) and more patients improved funct
ionally (17/25 versus 11/28, p < .04). Overall mortality results were 21/18
6 versus 33/195, odds ratio (OR) 0.62, 95% confidence interval (CI) 0.35, 1
.13. Conclusions: Prescribing sip feed supplements in the medicine prescrip
tion chart during hospital stay reduces weight loss. Our data also support
other evidence for a reduction in mortality noted in elderly patients on nu
tritional supplementation. There were suggestions of other clinical benefit
s.