PURPOSE: The incidence of pulmonary embolism (PE) is estimated to be approximately one in one thousand patients with an annual mortality of 300,000. Many clinical prediction scoring systems have been developed that aid the clinician in determining the pretest probability for the diagnosis of PE. One of the best known prediction systems is the Well’s criteria. Our study’s purpose was to determine whether practitioners at an urban academic medical center consider established risk factors to assess pretest probability prior to ordering a spiral CTA (Computed Tomography Angiography).
METHODS: A retrospective chart review of fifty-three consecutive patients who had a CTA for possible PE was conducted. Only patients from emergency department or those who were admitted for < 7 days were selected. Primary end point was the use of Well’s criteria explicitly or not, and secondary end points were the costs and clinically relevant of incidental findings.
RESULTS: Only two (3.7%) out of fifty-three patients suspected for a PE had a documented Well’s score (WS) prior to ordering a CTA. Chest pain and dyspnea were the most common reasons for ordering CTA, followed by dyspnea and chest pain alone. High WS (>6) was present in 13%, intermediate WS (2-6) in 23%, and low WS (< 2) in 64% of patients. CTA for PE was high probability in 15%, intermediate probability in 2%, and low probability in 83% of patients. One of thirty-four patients with low WS had CTA with high probability for PE. The P value was 0.007. Evaluation of patients who have a low WS by CTA has a cost of $93,000 per diagnosis.
CONCLUSION: The WS is a valuable and underutilized asset in the evaluation of a suspected PE. Utilization of WS is effective and when combined with D-dimer, it has been shown to be valuable in an evaluation of a suspected PE.
CLINICAL IMPLICATIONS: Utilization of WS is a valuable tool in an evaluation of a PE and can have profound cost saving implication if employed more broadly.
DISCLOSURE: Mehul Patel, No Financial Disclosure Information; No Product/Research Disclosure Information