Article: Sensitivity and Specificity of Helical Computed Tomography in the Diagnosis of Pulmonary Embolism: A Systematic Review
Authors: Rathbun SW, Raskob GE, Whitsett TL
Publication: Ann Intern Med. 2000;132:227-232
Study: systematic review of medline search for all reports from 1986 to October 1999 that evaluated the use of helical CT for the diagnosis of PE
Inclusion Criteria: all prospective English-language studies
Exclusion Criteria: retrospective studies, review articles, and case reports
Step 1: Were the results validity?
Step 1a: Did the the study ask a focused question?
yes
Step 1b: Were appropriate inclusion criteria used to select the primary studies?
Yes, and a specific search strategy is outlined in the methods section
Step 1c: Were important relevant studies missed?
Possibly, since the authors used only medline and the references from the primary studies. They did not make personal contact with experts in order to find studies not yet referenced or unpublished. However, there is less likelihood of publication bias given the negative results of the review
Step 1d: Was the validity of primary studies examined?
Yes, but no study met all of the predefined criteria for assessing validity - consecutive series of all patients with suspected PE, patients with and those without PE, a broad spectrum of patient characteristics, performance of helical CT and pulmonary angiography (or appropriate reference test) in all patients, independent interpretation of the CT scan and pulmonary angiogram (or reference test)
Step 1e: Was there reproducible assessment of the primary studies?
Two authors reviewed the primary studies independently and a third reviewer resolved disagreements although we are not told how often the authors agreed
Step 1f: Were the results of the primary studies similar?
No; only 2 studies stated that a consecutive series of all patients with suspected PE were evaluated, 8 studies interpreted helical CT and pulmonary angiography independently, 1 study gave enough data to conclude that a broad spectrum of patients was included, and there were large differences in outcomes among studies (test of homogeneity not reported)
Step 2: What are the results:
the results were not combined but instead reported separately in a table format:
| Study | Number of Patients | Sensitivity | Specificity |
| 1 | 42 | 100 | 95 |
| 2 | 10 | 100 | 100 |
| 3 | 25 | NA | NA |
| 4 | 20 | 64 | 89 |
| 5 | 249 | 90 | 97 |
| 6 | 75 | 91 | 100 |
| 7* | 149 | 94 | 96 |
| 82 | 93 | ||
| 8* | 53 | 77 | 93 |
| 69 | 100 | ||
| 62 | 100 | ||
| 77 | 80 | ||
| 92 | 73 | ||
| 9 | 20 | NA | NA |
| 10 | 142 | 87 | 95 |
| 11 | 164 | NA | NA |
| 12 | 39 | NA | NA |
| 13 | 185 | 67 | 100 |
| 14* | 47 | 60 | 81 |
| 53 | 97 | ||
| 15 | 110 | 92 | 88 |
* these studies reported inter observer variation
Step 3: Will the results help me in caring for my patients?
Limitations
1. None of the primary studies passed the tests for validity.
2. Large variation in the outcomes of the primary studies could be due to important differences in patients characteristics, techniques or experience in interpreting tests, and study designs.
3. A search for prospective studies evaluating the safety of withholding anticoagulation without additional testing (in patients suspected of PE with negative CT scans) was performed but no study was identified.
Strength
1. Well done systematic review with a clearly defined clinical question and criteria for assessing validity.
Conclusion
The use of helical CT in the diagnosis of PE has not been adequately
evaluated since no primary study passed tests of validity