Time Trends in Pulmonary Embolism in the United States

As computed tomography (CT) scanners have become more sophisticated and highly sensitive, CT pulmonary angiography (CTPA) has largely become the test of choice for diagnosing pulmonary embolism (PE). In this retrospective time trend analysis, the authors question whether this increased sensitivity has led to over-diagnosis and treatment. The authors searched the Nationwide Inpatient Sample and Multiple Cause-of-Death databases and compared the age-adjusted incidence, mortality, and complications of treatment of PE among U.S. adults for the 5 years before (1993–1998) and after (1998–2006) the introduction of CTPA. For this study, treatment complications were described as in-hospital intracranial hemorrhage, gastrointestinal bleeding, or secondary thrombocytopenia. Incidence of PE increased by 81% during the time period after the advent of CTPA (from 62.1 to 112.3 per 100,000), but showed no significant change for the period before CTPA. Mortality from PE decreased during both time periods (down 8%, from 13.4 to 12.3 per 100,000 pre-CTPA and down 3%, from 12.3 to 11.9 post-CTPA). Case fatality improved pre-CTPA (8% decrease, from 13.2% to 12.1%) and more dramatically post-CTPA (36% decrease, from 12.1% to 7.8%). However, complication rate also substantially increased. Before CTPA, the complication rate was stable, but after, the rate increased by 71% (from 3.1 to 5.3 per 100,000). With these results, the authors caution that over-diagnosis may be leading to more complications without a change in mortality.