|
D-dimer in ruling out acute aortic dissection: a systematic review and prospective cohort study |
Sodeck G, Domanovits H, Schillinger M, Ehrlich M P, Endler G, Herkner H, Laggner A |
|
|
CRD summary This review assessed the sensitivity and optimal cut-off level of D-dimer test levels to exclude the presence of acute aortic dissection (AAD). It concluded that this test is likely to be useful in accurately excluding the presence of ADD in patients with low or intermediate suspicion of the disease. However, these conclusions were based on a small poor-quality data set and may not be reliable. Authors' objectives To determine the sensitivity and optimal cut-off level of D-dimer levels to exclude the presence of acute aortic dissection (AAD). This article also reported on a prospective cohort study which is not covered in this abstract. Searching MEDLINE, EMBASE, CINAHL and BIOSIS Previews were searched from inception to January 2007; the search terms were reported. Articles in English, German, French, Dutch, Danish, Spanish, Italian and Portuguese were considered eligible. Study selection Eligible studies included clinical trials and case reports. The included study designs were cohort studies with or without control groups and single patient case reports. No criteria for eligible patients were reported. The index test was the measurement of D-dimer levels. The included studies used one or more of the following assay types: enzyme-linked immunosorbent assay, immunoturbidimetry, latex agglutination. The reference standard test was the diagnosis of AAD; no further details of the diagnostic criteria were reported. The principal outcome measures used in this review were the sensitivity, specificity and cut-off levels of D-dimer measurement.
The authors did not state how the papers were selected for the review, or how many reviewers performed the selection. Assessment of study quality The clinical validity of the primary studies was assessed using the following criteria: study design, number of patients, classification of AAD, types of assay used and predefined thresholds. Each included study was also assessed according to the Quality Assessment of Diagnostic Accuracy Studies (QUADAS) statement.
Two independent reviewers extracted these data, with any disagreements resolved by discussion amongst the reviewers. Data extraction Data were extracted to calculate the sensitivity and, where possible, the specificity. Continuous data were also extracted as a mean (with standard deviation, SD). Where a median plus interquartile range had been reported, the median was taken to be representative of the mean and the interquartile range was converted into an SD by dividing by 1.35.
Two independent reviewers extracted the data, with any disagreements resolved by discussion amongst the reviewers. Methods of synthesis A a random-effects model was used to produce pooled estimates of the true positive rate. Statistical heterogeneity was assessed using the I2 statistic. Meta-regression was used to extend the meta-analysis and to explore the impact of variations in cut-off values on the heterogeneity of the true positive rate. Where diseased and non-diseased populations were available, likelihood ratios (LRs), diagnostic odds ratios and summary receiver operating characteristic (sROC) estimates were calculated. Results of the review A total of 16 studies (n=437) were included in this review: 8 cohort studies with a control group, 3 uncontrolled cohort studies and 5 single case reports.
The results from the QUADAS quality assessment suggested that a large proportion of the included studies did not satisfy these criteria. The reported diagnostic cut-off values ranged from 0.1 to 0.9 μg/mL.
D-dimer testing to determine the presence of AAD was shown to have high sensitivity (0.97, 95% confidence interval, CI: 0.94, 0.98), based on the results from 9 studies, and negative LR (0.06, 95% CI: 0.0, 0.13). Five studies contributed to the estimate of pooled specificity (0.59, 95% CI: 0.53, 0.64) and positive LR (2.58, 95% CI: 1.76, 3.78) which, overall, indicates that a positive result on this test does not increase the certainty of diagnosis.
The sROC curve results suggested that negative D-dimer test results could be used to exclude AAD with high certainty levels (area under the curve 0.94).
The pooled diagnostic odds ratio was 21.27 (95% CI: 11.64, 38.88), indicating that the test provided fairly strong diagnostic evidence.
Heterogeneity tests found no significant variability between the studies, while a meta-regression found that changing the cut-off values had no effect on the sensitivity of the D-dimer test. Authors' conclusions The measurement of D-dimer levels is likely to be a useful addition when performing triage on patients with suspected AAD, allowing individuals to be prioritised for further diagnostic tests. While imaging will still be required to definitively diagnose AAD, a negative D-dimer test may be used to exclude the presence of AAD in up to 100% of cases and will be particularly useful in patients judged to be at low or intermediate risk of AAD. CRD commentary This review addressed a clear question with detailed inclusion criteria, although details of the reference test could have been clearer. A number of databases were searched for studies reported in several languages, but the failure to explore the unpublished literature may have introduced some publication bias. The selection process was poorly reported and cannot be assumed to have excluded bias and/or error at this stage. Quality assessments of the primary studies were carried out, although the results were unclear. The summary table highlights that none of the included studies used an appropriate patient spectrum or a reference standard. The meta-analysis combined controlled and uncontrolled cohort studies without taking into account the potentially variable quality of the evidence. It appears that the authors might have double-counted some of the primary data (where two difference assays were carried out on the same patient group, the data have been entered as though independent from separate studies). The synthesis may not, therefore, be reliable. Overall, the review methodology was inadequately reported, the available data were of a relatively poor quality, and the authors' conclusions appear overoptimistic. Implications of the review for practice and research Practice: The authors stated that D-dimer testing should be used as part of a diagnostic pathway to exclude ADD in patients with low or intermediate suspicion of the disease, Where patients have a high clinical likelihood of ADD, the appropriate imaging tests should still be carried out even if a negative D-dimer test result was returned.
Research: The authors did not state any implications for further research. Bibliographic details Sodeck G, Domanovits H, Schillinger M, Ehrlich M P, Endler G, Herkner H, Laggner A. D-dimer in ruling out acute aortic dissection: a systematic review and prospective cohort study. European Heart Journal 2007; 28(24): 3067-3075 Indexing Status Subject indexing assigned by NLM MeSH Aged; Aneurysm, Dissecting /diagnosis; Aortic Aneurysm /diagnosis; Biomarkers /blood; Epidemiologic Methods; Female; Fibrin Fibrinogen Degradation Products /analysis; Humans; Male; Middle Aged AccessionNumber 12008000025 Date bibliographic record published 01/09/2008 Date abstract record published 23/12/2008 Record Status 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. |
|
|
|