Coming full circle: an evidence-based definition of the timing and type of surgical management of very low-birth-weight (<1000 g) infants with signs of acute intestinal perforation.

OBJECTIVE

Gut disruption in very low birth weight follows 1 of 3 clinical pathways: isolated perforation with sudden free air, metabolic derangement (MD) complicated by appearance of free air, or progressive metabolic deterioration without evidence of free air. To refine evidence-based indications for peritoneal drainage (PD) vs laparotomy (LAP), we hypothesized that MD acuity is the determinant of outcome and should dictate choice of PD or LAP.

METHODS

Very low-birth-weight infants referred for surgical care because of free intraperitoneal air or MD associated with signs of enteritis were evaluated by univariate or multivariate logistic regression to investigate the effect on mortality of MD and initial surgical care (LAP vs PD). Metabolic derangement was scaled by assigning 1 point each for thrombocytopenia, metabolic acidosis, neutropenia, left shift of segmented neutrophils, hyponatremia, bacteremia, or hypotension. Laparotomy and PD were stratified by MD acuity, and odds of mortality were calculated for each surgical option.

RESULTS

From October 1991 to December 2003, 65 very low-birth-weight infants with suspected gut disruption were referred for surgical care. Peritoneal drainage and LAP infants had similar birth weight and gastrointestinal age, neither of which predicted mortality. Despite a higher incidence of isolated perforation with sudden free air in PD infants, the incidence of MD and overall mortality were similar for PD and LAP. Multivariate logistic regression demonstrated MD to be the best predictor of mortality (odds ratio [OR], 4.76; confidence interval [CI], 1.41-16.13, P = .012), which significantly increased with interval between diagnosis to surgical intervention (P < .05). Infants with MD receiving PD had a 4-fold increase in mortality (OR, 4.43; CI, 1.37-14.29; P = .0126). Conversely, those without MD and sudden free air who underwent LAP had a 3-fold increase in mortality (OR, 2.915; CI, 1.107-7.692; P = .03.) Of 5, 3 failed PD were "rescued" by LAP.

CONCLUSIONS

The dramatic difference in mortality odds based on surgical option in the presence of MD defines the critical importance of a thorough assessment of physiological status to exclude MD. Absence of MD warrants consideration for PD, especially for sudden intraperitoneal free air. Overwhelming MD may limit options to PD; however, salvage of 3 of 5 infants with failed PD demonstrates the value of LAP, whenever possible, for infants with MD.