| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Ann Thorac Surg 1996;62:1570-1571
© 1996 The Society of Thoracic Surgeons
Columbia-Presbyterian Medical Center 177 Fort Washington Ave 7th Fl, Rm 435 New York, NY 10032
To the Editor:
The recent review of mechanical valves by Akins [1] provides an excellent overall summary of the literature regarding these prostheses; however, the conclusions regarding the relative merit of the prostheses must be recognized to be highly subjective. Doctor Akins evaluates the five mechanical heart valves approved for implantation in the United States based on functional characteristics, a review of the published long-term complications, and a proposed new "index of valve thromboembolism and bleeding." Unfortunately, the subjective nature of several data elements limits the scope of new information potentially derived from the study.
Doctor Akins reviews the technical characteristics of the valves, including aspects such as structural integrity and transvalvular gradient, using an otherwise unspecified scale of 0 for worst and 5 for best. Objective grading criteria are not defined. Later he refers to the loss of structural integrity by "15 to 20 St. Jude valves"; however, supportive references or objective data are remarkably absent. Next, a collection of published reports on valve function is reviewed and summarized, but the lack of cumulative statistical analysis limits the contribution of new information. Finally, Dr Akins describes a new "thromboembolism and bleeding index" in which unpublished data are used to infer that the Medtronic Hall valve enjoys a reduced thrombogenicity over other mechanical prostheses. Because the data remain unpublished and the proposed index does not conform to previously published guidelines for reporting mortality and morbidity after valve operations [2], we remain uncertain of the contribution this report makes to the understanding of long-term valve function. The conclusion that observed differences in the incidence of thromboembolism and hemorrhage in patients receiving mechanical valvular prostheses are due to the intrinsic nature of the prostheses rather than patient or pharmacologic variables is not supported by the data presented.
Evaluation of critical issues such as valve safety and efficacy must use standardized methods of reporting. Reviews of available information by experienced surgeons are potentially useful additions to the literature. Nevertheless, subjectively supported conclusions regarding the comparative merits of valve prostheses must be regarded as personal opinion.
Accepted for publication June 7, 1996.
References
Cardiac Surgical Unit Massachusetts General Hospital White 503 32 Fruit St Boston, MA 02114
To the Editor:
Doctors O'Hair and Rose's comments about the relatively subjective nature of my review of mechanical valves are generally appropriate and obviously reflect the intended tone of the message in my review. Indeed, I used the phrase "very subjective" in my review to describe my assessment of the functional mechanical characteristics of the available mechanical valves. The assessment of functional characteristics was drawn from a large volume of in vivo and in vitro literature, some of which presents conflicting information. The absence of a complete list of supporting references followed discussion with the staff of this journal, who hoped to limit the length of the article.
In answer to Drs O'Hair and Rose's specific observation about the number of St. Jude Medical valves that have suffered clinical loss of structural integrity, that number was derived from a personal communication from the St. Jude Medical company. All of us are aware that determining actually how many valves of a particular type have failed is very difficult, even for the manufacturer, and merely counting the MDRs from the Food and Drug Administration is not a valid approach. However, 15 to 20 failed valves out of more than 600,000 implanted is a trivial risk.
Doctors O'Hair and Rose suggest that the absence of a statistical analysis of the cumulative results of valve-related complications limits the value of the information. As noted in response to the letters above, it is my belief that these data are not appropriate for statistical analysis, whether it be metaanalysis or any other approach, because those methods impose a univariate approach to a multivariate problem.
Doctors O'Hair and Rose's comments about my composite thromboembolism and bleeding index deserve response. In the first place the only "new" aspect of this index is that it is done in a composite fashion. Previous authors, including Edmunds [1], have already used the approach of combining thromboembolic and bleeding events. Drs O'Hair and Rose's comment that this index does not conform to the previously published guidelines from 1988 is absolutely correct. In fact, a group of cardiac surgeons, including myself, have been reviewing those guidelines for reconsideration of their efficacy. I am certain Drs O'Hair and Rose noted that the last sentence in my review suggests that my presentation "may encourage us to reassess the methods we use to evaluate mechanical valves." Time will tell whether this approach has any merit.
Finally, one attribute of my review may be that it makes cardiac surgeons face the important limitations of our assessment of the functional characteristics of mechanical valves, as well as the long-term complications of their use, which are so heavily influenced by other factors.
Reference
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ANN THORAC SURG | ASIAN CARDIOVASC THORAC ANN | EUR J CARDIOTHORAC SURG |
| J THORAC CARDIOVASC SURG | ICVTS | ALL CTSNet JOURNALS |
