How to validate the symptoms of pulmonary thromboembolism: how diagnostic scales help

• Published in: Terapevtic̆eskii arhiv Vol. 85; no. 4; p. 11
• Main Authors: Mironov, A V, Leont'ev, S G, Ustinov, F S, Efremova, O I, Lebedev, I S, Kirienko, A I
• Format: Journal Article
• Language: Russian
• Published: Russia (Federation) 2013
• Subjects: Adolescent; Adult; Aged; Aged, 80 and over; Area Under Curve; Female; Humans; Likelihood Functions; Male; Middle Aged; Predictive Value of Tests; Prospective Studies; Pulmonary Embolism - diagnosis; Pulmonary Embolism - epidemiology; Pulmonary Embolism - physiopathology; Reproducibility of Results; Severity of Illness Index; Young Adult
• ISSN: 0040-3660

To detect the most important clinical symptoms suggesting pulmonary thromboembolism (PTE) and to determine the diagnostic value of the scales used to estimate the likelihood of its occurrence. The prospective study included 130 patients admitted to hospital with a diagnosis of PTE and a referral for a surgery clinic. Scores of the likelihood of PTE were estimated using the Canada and Geneva scales in all the patients on admission. In all the patients with suspected PTE, the Canadian and revised Geneva scores averaged 4.2 +/- 0.48 and 6.21 +/- 0.5, respectively. These scores correspond to the intermediate clinical probability of PTE. In 96 patients whose diagnosis was verified by instrumental studies, the Canadian and Geneva scores were 4.41 +/- 0.57 and 6.17 +/- 0.63, respectively, which was also consistent with the intermediate clinical probability of PTE. In 34 patients, whose diagnosis of PTE was ruled out, the average scores did not virtually differ from those in the patients with the verified diagnosis and were 6.14 +/- 1.3 and 4.18 +/- 0.87, respectively. The area under characteristic curve for the Canadian scale was 0.428 and that for the Geneva scale was 0.512. With the use of a two-level interpretation system, a total of more than 6 Canadian scores and 10 Geneva scores suggested that there was a high probability of PTE. The investigation indicated the low value of integral systems for estimating the likelihood of PTE in the total population of patients with this disease. The authors recommend the two-level interpretation system, in which a total of more than 6 Canadian scores and 10 Geneva scores were identified with a high probability (up to 80%) of PTE.