Interval increase in right-left ventricular diameter ratios at CT as a predictor of 30-day mortality after acute pulmonary embolism: Initial experience

Purpose: To retrospectively determine if the interval increase of right ventricular-left ventricular (RV/LV) diameter ratio from negative prior to positive current computed tomographic (CT) examination findings for pulmonary embolism (PE) is more accurate for predicting 30-day mortality than positive CT ratio alone, by using patient 30-day mortality as reference standard. Materials and Methods: This IRB-approved, HIPAA-compliant study had waiver of informed consent and retrospectively reviewed 50 patients (19 men, 31 women; mean age, 60 years) with negative prior and positive current CT findings for acute PE (median interval, 63 days). Interval increase was defined as percentage change in RV/LV diameter ratio by using reformatted four-chamber views. Receiver operating characteristic (ROC) analysis compared the interval increase with the RV/LV diameter ratio from the positive findings alone for PE-related and all-cause mortality. Results: Twelve (24%) patients died in 30 days; nine were PE-related. The interval increase was significantly more accurate overall than the ratio from the positive study alone for PE-related (area under the ROC curve [AUC] = 0.95 vs 0.73, P = .003) and all-cause (AUC = 0.81 vs 0.66, P = .05) mortality. The respective sensitivity, specificity, positive predictive value, and negative predictive value were 0.78 (seven of nine; 95% confidence interval [CI]: 0.43, 1.00), 0.93 (38 of 41; 95% CI: 0.83, 1.00), 0.70 (seven of 10; 95% CI: 0.38, 1.00), and 0.95 (38 of 40; 95% CI: 0.87, 1.00) for PE-related mortality (interval increase, > 18%) and 0.75 (nine of 12; 95% CI: 0.49, 1.00), 0.89 (34 of 38; 95% CI: 0.80, 0.99), 0.69 (nine of 13; 95% CI: 0.44, 0.95), and 0.92 (34 of 37; 95% CI: 0.83, 1.00) for all-cause mortality (interval increase, > 15%). At target sensitivity (0.75), specificity of interval increase was significantly higher than from positive scans alone for both PE-related (0.93 vs 0.59, P = .001) and all-cause (0.89 vs 0.58, P = .05) mortality. Conclusion: The interval increase in four-chamber RV/LV diameter ratio is more accurate than the diameter ratio of the CT examination with with positive findings for PE alone for mortality prediction after acute PE. (c) RSNA, 2008.