Total parenteral nutrition during acute pancreatitis: clinical experience with 156 patients.

Over a 3-year period, 156 of 815 patients admitted to a single institution with acute pancreatitis received total parenteral nutrition (TPN) for 2,572 patient days. Seventy had "simple" acute pancreatitis (group I) and 86 (group II) developed local complex disease (pseudocyst, abscess, or necrotic gland). In groups I and II, respectively, days without oral intake (NPO) were 13.6 +/- 1.5 (SEM) and 24.0 +/- 2.1 (p less than 0.005), hospital days were 19.8 +/- 1.7 and 35.8 +/- 3.2 (p less than 0.005), and duration of TPN was 10.9 +/- 1.0 and 21.0 +/- 2.3 days (p less than 0.005). Thirty-three patients in group I and 53 in group II required exogenous insulin. Alteration of standard formulas was necessary in 87 patients, but cessation of therapy was necessary in only one instance. Twenty catheters were removed for suspected sepsis with only 3 confirmed cases. Fat-based formulas were well tolerated in 15% of patients. During TPN, body weight rose from 95.0 +/- 2.4% to 97.4 +/- 4.3% of ideal in group I and remained at 90.5 +/- 1.8% in group II. Albumin rose from 3.36 +/- 0.10 to 3.50 +/- 0.08 g/dl in group I and from 3.01 +/- 0.07 to 3.35 +/- 0.07 g/dl in group II. The entire cohort differed from 10 randomly chosen patients who did not receive TPN in terms of days NPO (2.8 +/- 0.3) and hospital days (5.5 +/- 0.6). Variables associated with prolongation of hospital stay and time NPO were number of prognostic criteria, local complex disease, and underlying chronic pancreatitis only in select groups. We conclude that during acute pancreatitis, TPN can be administered safely but with careful monitoring and we recommend early aggressive therapy in the subgroups noted above and when underlying malnutrition exists. In the borderline patient, TPN may be administered by peripheral vein until the severity of disease is manifest.