Disparity in Diagnosis of Pulmonary Embolism in Hospitalized Patients with Pneumonia.

Pulmonary embolism (PE) is a major cause of preventable hospital deaths. Little is known about PE in the setting of pneumonia, as symptoms and signs of PE and pneumonia overlap, confounding diagnosis. With infrequent thromboprophylaxis, PE continues to be a major public health problem. In order to determine the incidence and characteristics of PE in patients with pneumonia, we analyzed statewide discharge data, 1997–2001, collected by the Pennsylvania Health Care Cost Containment Council (PHC4). The analysis included all PE discharges with ICD-9 codes 415.11 and 415.19, excluding pregnant patients and all second or later PE admissions. Infectious disease diagnoses were enumerated by ICD-9 codes, including 030–041, 042, 790.7, 996–998; pneumonia by ICD-9 codes 480–483. 485–487, and 996–998; and diagnostic procedures by revenue code 352 for computerized tomographic (CT), including spiral CT scans. Severity of illness was determined by MediQual Altas Admission Severity Categories. Of 33,036 total PE discharges (all-PE) from Pennsylvania hospital discharges 1997–2001, 5,757 (16.8%) had infection, more than half of which, 3,543 (52.3%), were due to pneumonia (pneumonia-PE). There was a marginal increase in the incidence of pneumonia-PE during this period, 615 (10.3%) in 1997 to 839 (10.5%), p>0.05. Those with pneumonia-PE were typically Caucasian, 77.3%, p<0.01; female, 51.5%, p<0.01; and elderly, 53.1% ≥ 70 years of age, p<0.01, mean age, 66.4±17.3 yr. The proportion of patients undergoing CT scans increased by 15.3%, from 18.9% in 1997 to 21.8% in 2001, p<0.01, an increase 20% lower than than the increase in the all-PE group, 88.9%, from 10.8% in 1997 to 20.4% in 2001, p<0.01. There was a significantly smaller improvement in the severity of illness scores in the pneumonia-PE group, increasing (46.0% to 53.0%) for the three lowest Altas categories and decreasing (46.8% to 42.0%) for the two highest categories, 1997–2001, respectively, compared with the all-PE group, (56.9% to 68.7%) in the three lowest and (35.3% to 26.5%) for the two highest categories, both p<0.01. During this period, mortality rates decreased significantly less in the pneumonia-PE group, by 18.5%, from (21.1% to 17.2%) as compared to a 20% decrease in all-PE, from (11.5% to 9.2%), 1997–2001, p<0.01. The apparent disparity in CT scan frequency and associated lesser improvements in severity of illness and mortality in the pneumonia-PE group, as compared with the all-PE group, suggest early PE diagnosis may have been missed. Modeling the rate of increase in PE diagnosis per increase in CT scan in the all-PE group, we estimate 454 missed cases of PE, which is 12.8% more PE cases among those with pneumonia. We conclude that the lack of increase in CT scans among hospitalized patients with pneumonia is associated with lower rates of PE diagnoses and poorer outcomes. Thromboprophylaxis should be considered in hospitalized pneumonia patients, particularly those at risk for PE.

View largeDownload slide

Figure

View largeDownload slide

Figure

Close modal