To the Editor:
I have read with interest the article by Dr Konstantinides and colleagues1 regarding clinical outcomes after treatment of pulmonary embolism (PE) with either heparin or thrombolysis. Although not a randomized trial, appropriate multivariate analysis was performed so that although not the “final word,” the results might have some validity.
Unfortunately, as published, this study has a tremendous design flaw that could potentially result in a selection bias that is not addressed by the multivariate analysis. Specifically, 27% of the patients did not have confirmation of PE by either ventilation/perfusion scanning or pulmonary angiography. The use of only clinical criteria to diagnose some of these patients (we don’t know how many of this group had other studies, such as Doppler ultrasound) is fraught with difficulty. In the Prospective Investigation of Pulmonary Embolism Diagnosis (PIOPED) study, only about a third of patients with suspected PE were found to have PE at angiogram.2
How, then, does this issue result in selection bias? It is probable that physicians would be less likely to give thrombolytic agents to patients without a confirmatory ventilation/perfusion scan or pulmonary angiogram. Because it is almost certain that some of these patients were inaccurately diagnosed with PE, they may not have been receiving appropriate care for their true problem. Therefore, it is possible that patients with cardiopulmonary disease other than PE, who may have been more likely to die secondary to inaccurate diagnosis, were more frequently treated with heparin as opposed to thrombolysis. Given the relatively low number of deaths, there need not have been many such patients to skew the results.
Fortunately, there is an easy way to address this concern. Reanalysis of the data excluding patients without definitive diagnosis of PE should be performed. If the results remain compelling, then the evidence in support of thrombolytic therapy for major PE is much stronger. If the results, however, fail to reveal a benefit of thrombolysis, the readers of Circulation deserve to know this, given the increased cost and bleeding risk associated with thrombolytic therapy.
- Copyright © 1998 by American Heart Association
Konstantinides S, Geibel A, Olschewski M, Heinrich F, Grosser K, Rauber K, Iversen S, Redecker M, Kienast J, Just H, Kasper W. Association between thrombolytic treatment and the prognosis of hemodynamically stable patients with major pulmonary embolism: results of a multicenter registry. Circulation. 1997;96:882–888.
Dr Metersky’s letter gives us the opportunity to stress some important issues regarding the management of acute major pulmonary embolism (PE):
It is not correct to question the diagnosis of acute PE in 27% of our patients just because they did not undergo pulmonary angiography or ventilation/perfusion lung scan. In fact, all but 6 of these 191 patients (97%) had echocardiographic evidence of acute right ventricular pressure overload and/or pulmonary hypertension according to the criteria listed in the Methods section of our article.R1 Bedside echocardiography is a valuable noninvasive diagnostic tool in patients with clinically suspected acute major PE, especially in the emergency room or intensive care unit. The diagnostic accuracy and reliability of echocardiographic findings has been demonstrated repeatedly by our work groupR2 R3 as well as by others.R4 Besides, echocardiography is particularly useful for risk stratification and prognostic assessment of patients in the setting of acute PE.R5 Therefore, confirmation of major PE by echocardiography “alone” is by no means a “tremendous design flaw” or a factor causing selection bias, as Dr Metersky suggests. In fact, rather the opposite is true: a study including only patients who are able to undergo angiographic or scintigraphic procedures would unavoidably select a clinically stable patient group with favorable prognosis. This point of view is strongly supported by the recently published results of the Management and Prognosis of Pulmonary Embolism Registry.R6 In our study,R1 statistical analysis focusing only on the subgroup of patients with pulmonary angiograms or lung scans (n=528) would yield a 30-day mortality rate of 4.3% in patients given thrombolytic treatment as opposed to 9.7% in heparin-treated patients (P=0.05). Of particular importance, however, is the fact that the independent effect of early thrombolysis on death risk as assessed by multiple logistic regression analysis would remain virtually unchanged (odds ratio of 0.44 as opposed to 0.46 in the entire patient populationR1 ). As might be expected, the 95% confidence interval would become slightly larger (0.18 to 1.11) and the P value slightly higher (0.08) due to the smaller number of patients considered in such a statistical model.
In conclusion, it would be, in our opinion, inappropriate and misleading to alter the inclusion criteria of the present registry retrospectively, because we would only consider patients with better in-hospital outcome without further improving the (already high) diagnostic accuracy of cardiac ultrasound. On the other hand, the association found between thrombolysis and the prognosis of acute major PE remains compelling irrespective of which diagnostic procedures were performed in each case. Thus, our results do suggest a clinical benefit of thrombolytic treatment for clinically stable patients with acute major PE. Of course, a definite statement on this issue must await the results of a prospective randomized trial, as already stressed in our article.R1
Konstantinides S, Geibel A, Olschewski M, Heinrich F, Grosser KD, Rauber K, Iversen S, Redecker M, Kienast J, Just H, Kasper W. Association between thrombolytic treatment and the prognosis of hemodynamically stable patients with major pulmonary embolism: results of a multicenter registry. Circulation. 1997;96:882–888.
Kasper W, Geibel A, Tiede N, Bassenge D, Kauder E, Konstantinides S, Meinertz T, Just H. Distinguishing between acute and subacute massive pulmonary embolism by conventional and Doppler echocardiography. Br Heart J. 1993;70:352–356.
Konstantinides S, Geibel A, Olschewski M, Kasper W, Just H. Acute pulmonary embolism: the value of echocardiography for identification of high-risk patients. Circulation. 1997;96:I-25. Abstract.