TO THE EDITOR:
I read with interest "Closed Fractures of the Tibial Shaft. A Meta-Analysis of Three Methods of Treatment" (80-A: 174—183, Feb. 1998), by Littenberg et al. I have several comments regarding the outcome measures that are used to assess fracture-healing.
The strength of a study relies on its initial design and the outcome measure that it uses. With regard to fracture-healing and time to union, the measures used to assess the end point of healing (when independent weight-bearing is possible) are clinical and radiographic assessment. The accuracy of a physician's ability to assess fracture stability is low; only a deflection on the order of 3 degrees can be reliably assessed1, and as healing progresses the deflection required for assessment approaches 1 degree. In a study in which seven fracture models were assessed by twenty surgeons and twenty controls, manual testing was shown to overestimate stiffness 83 percent of the time4. The assessment of fracture-healing with use of radiographic methods also is inaccurate; the probability of correct evaluation2 is on the order of 0.5. The inaccuracy of radiographic assessment, even in the evaluation of mechanically similar fractures, puts into question the errors in the use of this outcome measure. The comparison of different methods of treatment that are characterized by different degrees of callus formation (primary compared with secondary healing) increases these errors further.
The ideal measure of fracture-healing is strength (in other words, force to failure) in that it measures the end point directly. This is not clinically practical except by the measurement of failure after removal of a splint. (The authors made no mention of when internal fixation devices were removed.) At the present time, the only validated method for the assessment of fracture-healing is the measurement of stiffness, but this is only accurate for fractures treated with external fixation3.
Conclusions regarding healing times should be interpreted with care, particularly when different treatment methods are compared. The rate of nonunion, defined at a point when these inaccuracies are somewhat negated, was not found to differ significantly among the three types of treatment (p = 0.99). Much can be gained from meta-analysis but only when the outcome measures are accurate. Accurate and precise outcome measures allow investigators to design studies that require fewer patients and that lead to stronger conclusions. This must not be forgotten when studies are designed to assess treatment protocols for fractures.
Roger Wade, F.R.C.S. (Ed): 350 Telegraph Road, Heswell, Wirral, Merseyside, United Kingdom
Dr. Littenberg, Dr. Weinstein, Dr. McCarren, Mr. Mead, Dr. Swiontkowski, Dr. Rudicel, and Dr. Heck reply:
Mr. Wade is absolutely correct regarding the low quality of outcome measures available in the orthopaedic literature. We agree that the strength of any study relies on good design and accurate measurement. We were very disappointed with the quality of study designs and outcome measures available in the literature regarding the repair of tibial fractures. We join with him in requesting that investigators in this field employ accurate and precise outcome measures.
Benjamin Littenberg, M.D.: Division of General Medical Sciences, Washington University School of Medicine, 660 South Euclid Avenue, Campus Box 8005, St. Louis, Missouri 63110
Loryn P. Weinstein, M.D.: Department of Orthopedics, Mayo Clinic, 200 First Street S.W., Rochester, Minnesota 55905
Madeline McCarren, Ph.D.: Center for Health Services Research, M/C 922, University of Illinois, 2121 West Taylor Street, Chicago, Illinois 60612
Thomas Mead, M.L.S.: Dana Biomedical Library, 6168 Dartmouth College, Hanover, New Hampshire 03755
Marc F. Swiontkowski, M.D.: Department of Orthopedics, University of Minnesota, 420 Delaware Street, Box 492, Minneapolis, Minnesota 55455
Sally A. Rudicel, M. D.: Department of Orthopaedics, New England Medical Center, 750 Washington Street, P.O. Box 306, Boston, Massachusetts 02111
David Heck, M.D.: Indiana University, 541 Clinical Drive, Suite 600, Indianapolis, Indiana 46202