The implementation of a clinical decision rule for pulmonary CT angiography (CTA) may improve diagnosis of pulmonary emboli (PE), with one hospital significantly increasing the likelihood that a patient referred for CTA would have an embolism, according to a study published in the May issue of the American Journal of Roentgenology.
Pulmonary CTA has become the procedure of choice for the diagnosis of PE, although as utilization of the exam grows, several studies have investigated guidelines for more appropriate use of the exam, according to Guy W. Soo Hoo, MD, MPH, of the pulmonary and critical care section at West Los Angeles Medical Center in Los Angeles, and colleagues.
“At our institution, requests for pulmonary CTA examinations increased 56 percent between 2000 and 2005 without a significant change in the positivity rate for PE, which was 3.1 percent (6/196) in the two years preceding our study. This increased utilization suggested an inappropriate use of pulmonary CTA, possibly as a screening tool,” stated Soo Hoo and co-authors.
In December 2006 a rule was introduced into the computerized provider order entry (CPOE) system, requiring that for patients with suspected PE, a Wells score higher than 4 (PE likely) generated an automatic approval for pulmonary CTA. For Wells scores of 4 or lower, PE was categorized as unlikely, so that patients were required to obtain enzyme-linked immunosorbent assay (ELISA) D-dimer levels.
D-dimer levels above 500 ng/mL resulted in automatic approval for CTA, while Wells scores of 4 or lower, along with D-Dimer levels at 500 ng/mL or below, required consultation with a chest radiologist prior to approval.
Two hundred sixty-one pulmonary CTA exams were performed between 2006 and 2008 (252 patients) to evaluate suspected PE. The average patient age was 65, while 14 of the patients were women.
A total of 17 percent of patients who underwent CTA had PE, compared with just over 3 percent in the two years prior to the implementation of the clinical decision rule. Comparing the same two-year periods with and without the rule, the number of exams performed increased from 196 to 261.
No PE were found within three months of CTA in 82 percent of negative exams. The remaining patients died from cancers or other causes, minus five patients lost to follow-up.
The 43 patients with positive CTA exams had significantly higher Wells scores (mean 5.5) than patients with negative CTA findings (mean Wells score 4.5).
Sixty-five percent of PE-positive patients had D-dimer tests, compared with 68 percent of PE-negative patients. The mean level for the PE-positive group was 4,956 ng/mL, which was significantly higher than the mean 2,398-ng/mL level among PE-negative patients.
The combination of a Wells score of less than 4 and a D-dimer level of less than 1,000 ng/mL yielded a negative predictive value of 1.0, the authors reported. Among patients with Wells scores greater than 4 and D-dimer levels greater than 3,000 ng/mL, 72 percent were positive for PE on CTA.
“The main finding of our study is that pulmonary CTA utilization can be significantly improved by the successful incorporation of a clinical decision rule and ELISA d-dimer test into a computerized order entry menu,” explained Soo Hoo and co-authors.
“When coupled with ongoing educational efforts, these modifications effectively changed ordering patterns to better focus pulmonary CTA studies on the patient population at risk for PE and represented a significant change in institutional approach to the evaluation of patients with suspected PE.”
Despite the capacity to override the rule via D-dimer score or radiologist consultation, only two CTA studies were performed in patients with Wells scores below 4 and D-dimer levels below 500 ng/mL.
“The increase in the positive pulmonary CTA rate despite an increase in the actual number of pulmonary CTA examinations performed suggests that the implemented guidelines improved the process of identifying patients with PE for pulmonary CTA,” Soo Hoo and colleagues argued.
The authors acknowledged that the relatively small patient sample, particularly the scarcity of female patients, as well as a lack of information on the exams not ordered as a result of the guidelines, limited the generalizability of their findings.
Nevertheless, Soo Hoo and colleagues contended, “Our experience indicates that incorporation of a clinical decision rule and D-dimer test