Timing of aneurysm surgery in subarachnoid hemorrhage: A systematic review of the literature

OBJECTIVE: Many practitioners favor early operation after aneurysmal rupture, but sound data supporting this practice are lacking. A systematic review was conducted to compare early aneurysm surgery (Days 0-3), intermediate surgery (Days 4-7), and late surgery (more than 7 d after subarachnoid hemorrhage). METHODS: We performed a MEDLINE search of the literature published between January 1974 and December 1998 and an additional manual search of selected journal titles from January 1998 to December 1998. Main outcome measures were death and poor outcome (defined as death or dependency) at the end of the follow-up period. Risk ratios (RRs) and corresponding 95% confidence intervals (CIs) were calculated; patients planned for late surgery were used as the reference. RESULTS: Identified were 1 randomized clinical trial and 268 observational studies, of which only 10 studies (assessing a total of 1814 patients) fulfilled a set of minimum requirements for methodological quality. In the trial, the RR of poor outcome was 0.42 (95% Cl, 0.17-1.04) for patients planned for early surgery and 1.07 (95% CI, 0.56-2.05) for intermediate surgery. In analyses with data from the 11 included studies, the RR of poor outcome for patients in good clinical condition at admission was 0.41 (95 % CI, 0.34-0.51) for early surgery and 0.47 (95% CI, 0.32-0.69) for intermediate surgery. For patients in poor clinical condition at admission, the RR of poor outcome was 0.84 (95% CI, 0.67-1.05) for early surgery and 0.54 (95% CI, 0.24-1.22) for intermediate surgery. Adjustment of the RRs for year of publication, study design, and aneurysm location yielded essentially the same results, as did a sensitivity analysis after exclusion of the data from the randomized trial. CONCLUSION: This meta-analysis suggests that both early and intermediate surgical treatment improve outcome after aneurysmal subarachnoid hemorrhage-in particular for patients in good clinical condition at admission. However, this impression is derived only from an indirect comparison between different cohorts of patients. Sound evidence on the best timing of surgery is still lacking. Observational studies with better methods-and ideally a new randomized trial-are needed.