Application of pulmonary embolism rule-out criteria to the UK Manchester Investigation of Pulmonary Embolism Diagnosis (MIOPED) study cohort

Abstract

The effect of introducing a structured approach to the diagnosis of pulmonary embolism in UK emergency departments has been complex. Unlike our US and Australasian counterparts, it can take several days to complete the pulmonary embolism investigative pathway, as delays for ventilation-perfusion and computed tomography (CT) scans are common. As in the US [1Kline J.A. Wells P.S. Methodology for a rapid protocol to rule out pulmonary embolism in the emergency department.Ann Emerg Med. 2003; 42: 266-75Abstract Full Text Full Text PDF PubMed Scopus (53) Google Scholar], a larger proportion of emergency department patients now undergo pulmonary embolism rule-out strategies. During pulmonary embolism exclusion patients can be admitted to hospital unnecessarily. Furthermore, patient care is passed from the emergency department to the admitting medical physician. On occasion, the admitting specialties disagree that exclusion of pulmonary embolism is necessary: for example, in the young well patient with isolated pleuritic chest pain. The PERC rule [2Kline J.A. Mitchell A.M. Kabrhel C. Richman P.B. Courtney D.M. Clinical criteria to prevent diagnostic testing in emergency department patients with suspected pulmonary embolism.J Thromb Haemost. 2004; 2: 1247-55Abstract Full Text Full Text PDF PubMed Scopus (0) Google Scholar] provides a standardized approach to assessing such patients prior to commencing a rule-out strategy. The Manchester Investigation of Pulmonary Embolism Diagnosis study (MIOPED study) is a prospective cohort study which recruited 425 patients with pleuritic chest pain. Between February 2002 and May 2003 patients presenting to Manchester Royal Infirmary's emergency department with pleuritic chest pain were consented and recruited. Exclusion criteria included pneumothorax, electrocardiogram (ECG) changes of myocardial infarction, ischemia or pericarditis, pregnancy, trauma within 4 weeks, age under 18 and patients previously recruited to the study. Pulmonary embolism was excluded with combined normal IL d-dimer test and low clinical probability, a normal ventilation-perfusion scan or a low probability ventilation-perfusion scan with low clinical probability, normal CT pulmonary angiography with low clinical probability or normal digital subtraction pulmonary angiography. All patients were followed-up clinically for 3 months. Pulmonary embolism was confirmed by a high probability ventilation-perfusion scan with high clinical probability, positive CT pulmonary angiography or a positive digital subtraction angiography. The mean age was 38.3 years (SD 15.0) with 51.1% of the cohort female. Eighty-eight per cent of patients scored a low Wells' clinical probability of pulmonary embolism, 8.7% moderate and 3.3% high clinical probability. Mean PaO2 was 11.8 kPa (SD 2.2) and PaCO2 5.1 kPa (SD 0.7). Five per cent were tachycardic and 25% tachypnoeic, 20% had an elevated white cell count and 13% an abnormal chest X-ray. The prevalence of pulmonary embolism was 5.3%. All patients had complete data for the PERC decision rule. Two hundred and sixteen patients fulfillled all eight factors (age < 50, pulse < 100, pulse oximetry > 94%, no unilateral leg swelling, no hemoptysis, no recent surgery, nor prior thromboembolism and no oral hormone use). Within this subgroup, three patients had pulmonary embolism (1.39%, 95% CI 0.5–4.0%). The sensitivity of the PERC rule was 86.4% (95% CI 65.1–97.1%), specificity 53.9% (95% CI 48.9–58.9%) and negative likelihood ratio 0.25 (95% CI 0.09–0.62). In comparison, 3.2% (95% CI 1.7–5.6%) of the Wells' low clinical probability group had pulmonary embolism. During the MIOPED study Manchester Royal Infirmary saw 100 000 patients per year. Four per cent of presentations involved chest pain and a quarter of these were cases of pleuritic chest pain. Thus, on average, three patients per day presented with pleuritic chest pain. As demonstrated by our study, the prevalence of pulmonary embolism among this population is low. The PERC rule offers a method for screening these patients without taking a blood test for d-dimer level and subjecting patients with a false positive d-dimer to unnecessary diagnostic imaging. Of the remaining 201 patients who could not have pulmonary embolism ruled out using the PERC decision rule, 79% scored low Wells' clinical probability, 15% moderate and 6% high clinical probability. Applying the PERC rule in retrospect to our cohort, only 159 patients (with low clinical probability) would require a d-dimer analysis. Of these, 69 had a positive IL test d-dimer and would require diagnostic imaging, in addition to 42 patients who scored a moderate or high clinical probability. The total number of patients requiring further imaging would have been reduced from 186 (44.6%) to 111 (26.6%). The obvious disadvantage of the PERC rule is that it would have missed three cases of pulmonary embolism. In comparison, using the Wells' clinical probability score in combination with d-dimer missed two patients with pulmonary embolism. Despite the limitations, the PERC rule remains an objective and useful decision rule, and represents one of the first attempts to address the potential for over investigation of pulmonary embolism in low-risk patient groups.