Occult pulmonary embolism in intensive care unit patients undergoing chest computed tomography scan: incidence and effect on outcomes.

To determine the incidence of occult pulmonary embolism (PE) and the associated morbidity in an intensive care unit (ICU).Retrospective study.Fifteen-bed ICU of a university hospital.Two hundred patients who underwent chest computed tomography (CT) scans with administration of contrast.The patients were classified into 3 groups: (1) Occult PE if the chest CT scan was not taken, specifically for elucidating the diagnosis of PE, but it confirmed this diagnosis; (2) non-occult PE if the chest CT scan was taken to elucidate a suspected diagnosis of PE and confirmed this diagnosis; and (3) the chest CT scan did not confirm this diagnosis. The analysis was conducted to identify the effect of a diagnosis of occult PE on the outcomes of patients.Among the 200 patients who underwent chest CT scan, 27 (13%) patients had PE, in whom 18 (9%) were classified as occult PE and 9 (4.5%) as non-occult PE. The duration of ICU stay was increased in patients with PE, as compared with the controls (23 [18-48] days v 17 [10-20] days v 14 [7-29] days; p = 0.02 for occult PE, non-occult PE, and controls, respectively). No difference was observed in mortality rate among the 3 groups.Occult PE was found in 9% of the cohort. This emphasized the need for developing diagnostic strategies in high-risk patients. Future studies should aim at assessing interventions for preventing this event.