Background: Modification of the Norwood procedure has been reported to improve immediate postoperative mortality compared with the classic Norwood. Interstage mortality has not been shown to be improved with the modified Norwood probably because of the small number of patients from each institution. The goal of this study was to determine if meta-analysis would provide sufficient data to prove statistical difference in interstage mortality for the modified Norwood procedure.
Methods: PubMed was searched using six different terms individually for articles from January 2003 to October 2004. Manuscripts that compared the classic to modified Norwood were reviewed. Mantel-Haenszel analysis was used to evaluate the relationship between treatment method and mortality stratified across hospitals. The Breslow-Day procedure tested homogeneity of odds ratio across hospitals. Separate analyses were performed for inpatient and interstage periods.
Results: A total of 4,545 citations was screened. Five manuscripts met the criteria. Seventy-two patients undergoing classic Norwood and 84 patients undergoing modified Norwood survived to initial hospital discharge. The Breslow-Day statistic supported homogeneity of odds ratios for survival across hospitals (chi2 = 2.09, df = 4, p = 0.72). Odds of interstage death was 11.6 times greater (2.2 to 62.1, 95% CI) for the classic Norwood compared with the modified Norwood procedure. This difference was statistically significant by the Mantel-Haenszel chi2 (11.0, p = 0.001). The Breslow-Day statistic supported homogeneity of the odds ratios across hospitals (chi2 = 3.1, df = 4, p = 0.53).
Conclusions: The modified Norwood procedure has a significantly lower interstage mortality compared with the classic Norwood procedure. A large randomized study is needed to determine whether these results remain consistent.