Improvement of screening for pulmonary embolism with a standardized questionnaire

The anamnesis is believed to be poor in identifying patients with pulmonary embolism (PE), but the method of data collection may be critical for inference on this issue. We compared the prevalences of history findings recorded after a free verbal interview (VI) by the referring physicians with those recorded after completion of a standardized questionnaire (SQ) by the admitting physicians in a group of 177 consecutive patients referred to our Emergency Unit with the suspicion of PE (subsequently confirmed in 97). VI data were incomplete in 18 patients. In the remaining 159 patients, prevalences of symptoms and predisposing factors were higher after SQ than after VI. Accordingly, 8 items (obesity, prolonged immobilization, surgery, varicose leg veins, deep venous thrombosis, pleuritic chest pain, and sudden-onset dyspnea) were significantly more prevalent in patients with confirmed PE after SQ, compared to only 2 items (prolonged immobilization and pleuritic chest pain) after VI. When we tested for the agreement between the two methods of data collection, kappa values ranged from high values (for surgery and hemoptysis) to very low values (for prolonged immobilization and recurrent phlebitis). These results show that the use of an SQ could improve the accuracy of collecting clinical data in patients with suspected PE, as they are also consistent in separating patients with PE from those with unconfirmed suspicion of PE. Moreover, it allows the clinician to be alert towards findings which could be missed when not carefully searched for and which may be useful to raise or strengthen the suspicion of this disease.