Diagnostic accuracy of the Geneva clinical scale for diagnostic prediction of pulmonary embolism in adults aged 18 and older admitted between 2009 and 2020 with suspected pulmonary embolism at a Third-Level Institution in Colombia: A retrospective cohort study

To assess the overall applicability of the Geneva scale for diagnosing pulmonary embolism in adults aged 18 and older. A retrospective cohort study with diagnostic test analysis was conducted on patients in the emergency department or hospitalized between 2009 and 2020 with suspected pulmonary embolism at a Third-Level Institution in Colombia. Local study. The original and simplified Geneva scores were applied to 1237 subjects aged 18 and older with suspected pulmonary embolism and compared with confirmatory results from pulmonary angiography. All necessary variables for constructing the original and simplified Geneva rules were recorded, and calculations for sensitivity (S), specificity (E), likelihood ratios, and receiver operating characteristic curves were performed. The Geneva original score exhibited an S, E, positive likelihood ratio, negative likelihood ratio, and area under the curve of 60%, 54%, 1.3, 0.728, and 0.506, respectively. The simplified Geneva score showed 59%, 57%, 1.4, 0.7, and 0.546 for S, E, positive likelihood ratio, negative likelihood ratio, and area under the curve, respectively. The use of the original or simplified Geneva score in our population may not be useful for a diagnostic approach to pulmonary embolism. Both scales demonstrate almost negligible discriminatory capacity, necessitating the evaluation of other standardized clinical decision rules to assess the diagnosis and pretest probability of pulmonary thromboembolism.