Our objective was to assess whether the serum albumin level rises in p
atients given total parenteral nutrition (TPN). All randomized control
led studies of TPN for at least 7 days versus oral therapy were review
ed. Data on serum albumin had to be available to be acceptable for ana
lysis. Only trials of patients with cancer fulfilled these selection c
riteria for our analysis. None of the reported studies showed a signif
icant rise in serum albumin with TPN when compared to controls. The me
an change in serum albumin levels for all of the studies was -0.3 g/dl
in the TPN group and -0.3 g/dl in the control group. In published ran
domized controlled studies of TPN versus oral diet, there is no signif
icant increase in serum albumin levels in those receiving TPN or decre
ase in serum albumin in controls. Our study does not support the serum
albumin level as a nutritional marker in patients with cancer.