Evaluation of the Diagnostic Accuracy of the Wells Score in Pulmonary Embolism

Author(s): Taryètba André Arthur Seghda, Tall/ Thiam Anna, Arnaud Diendéré, Martin Wendlassida Nacanabo, Sandrine Damoué/Seghda, Léa Wendlassida Sawadogo, Djième Claudine Dah, Abouga/Loya Mireille, Georges RC Millog, Nobila Valentin Yameogo, André K Samadoulougou, Patrice Zabsonré

Introduction: Pulmonary embolism pulmonary embolism (PE) is a major medical emergency responsible for high morbidity and mortality in the absence of early and appropriate diagnosis and management. Its diagnostic approach is well codified by validated scores, including the Wells score.

Patients and Methods: This was a 3-year descriptive and analytical cross-sectional study carried out in a Burkina Fas university hospital. Patients admitted for suspected pulmonary embolism were included. Study variables were socio-demographic, clinico-biological and scanographic data, clinico-biological and CT scan data. Patients were classified according to the clinical probability level of the Wells score as follows: low (score < 1 points); intermediate (score 2-6 points) and high (score > 7 points). Angioscan was the gold standard, with the Wells score as the diagnostic test.

Results: The hospital prevalence of pulmonary embolism was 21.74%. The mean age of patients was 54.33 ± 16.82 years, and the modal class corresponding to the [60-70] age group was the most represented. The variables significantly associated with the discovery of pulmonary embolism on CT-scan after logistic regression were a history of thromboembolic venous disease (p=0.0171), recent surgery (p=0.0005), shortness of breath dyspnea (p=0.0166), chest pain (p=0.0053), syncope (p=0.0010). The sensitivity and specificity of the Wells score at 2 probability levels were 78.65% and 92.19% respectively, while the positive predictive value value (PPV) and negative predictive value were 92.92% and 76.82% respectively.

Conclusion: Pulmonary embolism, previously rare in sub-Saharan Africa, has become a real health problem. The Wells predictive score appears to be one of the tools of choice in the diagnostic process in our context. However, the score still needs to be adapted to local data from prospective multicenter studies.