|Meta analysis on mortality of ruptured abdominal aortic aneurysms
|Hoornweg L L, Storm-Versloot M N, Ubbink D T, Koelemay M J, Legemate D A, Balm R
This review concluded that mortality rates of patients with ruptured abdominal aortic aneurysms treated by open surgery have not changed over the past 15 years. These conclusions are supported by the data presented, but their reliability is limited by the inherent methodological flaws of the study designs on which they are based.
To assess the mortality rates of open surgery for patients with ruptured abdominal aortic aneurysms (AAAs) and to examine changes in mortality over time.
MEDLINE, PubMed, EMBASE, SUMSearch, CINAHL, the Cochrane CENTRAL Register and Excerpta Medica were searched for articles published after 1991; the search terms were reported. The reference lists of relevant articles were handsearched for additional articles, but there were no manual searches of journals or conference proceedings and no contact with leading topic experts. Two independent researchers conducted the searches, which were restricted to articles written in English, German, French, Spanish and Dutch.
Prospective and retrospective studies assessing the mortality rates (intra-operative or overall) of any conventional open surgery (trans- or retroperitoneal) for patients with ruptured AAA were eligible for inclusion. Eligible studies had to include at least 40 patients and describe an original consecutive patient series. Ruptured AAA was defined as ultrasound or computed tomography (CT) diagnosis of blood outside the aortic wall, or as confirmed by surgery. Studies had to clearly distinguish ruptured AAAs from symptomatic AAAs. Studies of thoracic, inflammatory or mycotic aneurysms were excluded from the review, as were studies focusing on specific subgroups of patients such as octogenarians. Where described, the included studies were mainly retrospective, although some prospective studies were also included. Patients undergoing surgery were examined over a period of 33 years (mid time of studies was between 1970 and 2003). Where reported, the mean age of the included participants ranged from 67 to 82 years and the proportion of women from 0 to 36%, and the majority of surgeries were performed by vascular surgeons. The studies were mainly carried out in North America, Europe and the UK.
Two independent reviewers assessed the eligibility of articles and any disagreements were resolved by discussion or by a third reviewer, after reading the full article.
Assessment of study quality
The authors did not state that they assessed validity.
Percentage mortality rates were extracted from the mid point of the study (i.e. half way through the inclusion period). The overall and intra-operative mortality rates, study design, participant age, source of data, type of surgeon, country of origin, hospital volume and number of patients dying before surgery were also recorded.
Two independent reviewers extracted the study data.
Methods of synthesis
Meta-regression was used to plot intra-operative and overall mortality rates against time and weighted by study size. Weighted mean overall and intra-operative mortality rates were calculated. Publication bias was assessed using funnel plots.
Results of the review
One hundred and sixteen studies (n=60,822) were included in the review. Seventeen studies were prospective, 52 were retrospective and 47 studies failed to describe their study methods.
The pooled weighted mean was 48.5% (95% confidence interval, CI: 48.1, 48.9) for overall mortality and 13.3% (95% CI: 12.3, 14.3) for the 37 studies that reported intra-operative mortality. Meta-regression showed that the reported 1.6% reduction in overall mortality over 33 years and the 1.2% increase in intra-operative mortality over 29 years were not statistically significant (p=0.84 and p=0.69, respectively). Similarly, there was no significant difference in overall mortality in articles using national registry data as compared with data from hospital records. Significant differences were, however, reported for change in age over time (p=0.03), and for prospective versus retrospective study designs (46.7%, 95% CI: 36.7, 56.7 versus 41.1%, 95% CI: 32.5, 49.7), although the latter may not be reliable as only a small number of prospective studies were identified. A positive association between hospital volume and overall mortality rates was also reported (p=0.04), but again this may not be reliable as the CIs were wide and there were very few centres which performed over 30 ruptured AAA repairs a year.
Overall, there was no evidence of publication bias for studies reporting overall mortality. However, there was evidence of publication bias in the assessment of intra-operative mortality (37 studies).
Mortality rates of patients with ruptured AAA treated by open surgery have not changed over the past 15 years. This may be due to the increased age of patients undergoing repair of ruptured AAAs.
This clearly defined review used a broad range of study designs identified from a number of different electronic databases. However, as acknowledged by the authors and subsequent statistical tests, there may be some risk of publication bias as no specific attempts were made to locate unpublished studies. Some restrictions were also placed on the language of publication, but it is difficult to assess whether this would have had a detrimental affect on the review. Precautions were taken to reduce the risk of reviewer bias and error when selecting studies and extracting their data, but it appears that study quality was not formally assessed, although the reviewers stated that they adhered to guidelines for the assessment of observational studies. This, along with the limited details of included study designs, makes it difficult to assess the reliability of the data. However, it is likely to be poor given that the majority of studies were reported as being retrospective, with some also relying on data from national registries. Given the variability in design, patients, surgeon and setting, it is difficult to assess whether the pooled estimates are reliable and the meta-regression suggests that the data may be confounded by age, study design and hospital. Overall, the authors' conclusions appear to be supported by the data presented, but their reliability is unclear given the lack of any assessment of validity and the likely inherent methodological flaws of the included study designs.
Implications of the review for practice and research
The authors did not state any implications for practice or further research.
Dutch Heart Foundation, grant number 2002B197.
Hoornweg L L, Storm-Versloot M N, Ubbink D T, Koelemay M J, Legemate D A, Balm R. Meta analysis on mortality of ruptured abdominal aortic aneurysms. European Journal of Vascular and Endovascular Surgery 2008; 35(5): 558-570
Subject indexing assigned by NLM
Aged; Aneurysm, Ruptured /mortality /surgery; Aortic Aneurysm, Abdominal /mortality /surgery; Hospital Mortality /trends; Humans; Registries
Date bibliographic record published
Date abstract record published
This is a critical abstract of a systematic review that meets the criteria for inclusion on DARE. Each critical abstract contains a brief summary of the review methods, results and conclusions followed by a detailed critical assessment on the reliability of the review and the conclusions drawn.