Multidetector Computed Tomography for Acute Pulmonary Embolism

Stein PD. NEJM 1 June 2006; 354(22): 2317-27.

RESULTS

Randomization: 7284 screened, 3262 eligible, 1090 enrolled. 238 did not receive reference diagnosis, 824 underwent subsequent analysis (11% of those screened, 25% of those eligible).

Baseline: mean age 51.7y, majority had low or moderate probability of pulmonary embolism (Wells score).

Reference Diagnosis: 192/824 diagnosed with PE (23%), of the 632 ruled out, 592 had interpretable CTA, 590 did not receive anticoagulatants, 2/590 had initially unrecognized PE.

Wells Score | Table 4 | Table 5

Results of CTA:

Sensitivity 83% (76-92%)
Specificity 96% (93-97%)
Likelihood ratio for a positive test was 19.6 (13.3 - 29.0)
Likelihood ratio for a negative test was 0.18 (0.13 - 0.24)
PPV 86% (79-90%)
NPV 95% (92-96%)

PPV for PE in main or lobar artery 97%
PPV for PE in segmental vessel 68%
PPV for PE in subsegmental vessel 25%

Results of CTA-CTV:

Sensitivity 90% (84-93%)
Specificity 95% (92-96%)
Likelihood ratio for a positive test was 16.5 (11.6 - 23.5)
Likelihood ratio for a negative test was 0.11 (0.07 - 0.16)
PPV 85% (78-94%)
NPV 97% (94-97%)

Results using the Wells score:

High Clinical Probabiltiy: PPV 96%, NPV 60%
with CTV, PPV 96%, NPV 82%

Discussion: In high clinical probability patients, a positive CTA provides sufficient evidence to start treatment. Venous phase imaging increased the sensitivity by 7%. In high clinical probability patients who have a negative CT, the false negative rate was 40% which poses a dilemna regarding treatment.

METHODS

Study: Prospective, multicenter (8) study sponsered by NHLB Institute

Inclusion: (a) 18y +, (b) clinically suspected of acute PE.

These patients were seen in either outpatient or inpatient centers, and mostly recruited during the daytime and weekdays.

Exclusion: (a) inability to complete testing in 36h (19%), (b) abnormal creatinine levels or on long-term HD (19%), (c) long-term anticoagulant use (13%), (d) critically ill, (e) on mechanical ventilation, (f) allergic to contrast, (g) had an MI in last 30d, (h) possible pregnancy, (i) s/p IVC filter, (j) no suspected PE, (k) had UE DVT, (l) previously enrolled in the study.

Time Frame: September 2001 to July 2003

Composite Reference Standard: (1) V/Q scan showing high probability of PE, (2) Abnormal findings on DSA (digital subtraction angiography), (3) Abnormal findings on venous ultrasonography and nondiagnostic V/Q scanning.

Exclusion of PE Required: (1) Normal findings on DSA or (2) Normal V/Q scan or (3) V/Q scan with low or very low probability, Wells score < 2, and normal venous LE U/S

Follow-up: 3 months & 6 months

Machines: 4-slice, 8-slice, and 16-slice MDCT, mostly 4-slice

ADDITIONAL READING

van Belle A. Effectiveness of managing suspected pulmonary embolism using an algorithm combining clinical probability, D-dimer testing, and computed tomography. JAMA 11 Jan 2006; 295(2): 172-9.

Kline JA. Prospective multi-center evaluation of the pulmonary embolism rule-out criteria. J Thromb Haemost 2008; 6(5): 772.

Goodman LR. CT venography: continuous helical images versus reformatted discontinuous images using PIOPED II data. Am J Roentgenol 2007; 189: 409-12.

Goodman LR. CT venography and compression ultrasound are diagnostically equivalent: data from PIOPED II. Am J Roentgenol 2007; 189: 1071-76.