Rob J.P.M. Scholten, MD, PhD Wim Opstelten, MD Cees G. van der Plas, MD Dick Bijl, MD Walter L.J.M. Devillè, MD, PhD Lex M. Bouter, PhD Utrecht, Amsterdam, and Monnickendam, The Netherlands
From the Dutch Cochrane Centre and Department of Clinical Epidemiology and Biostatistics, Academic Medical Center, Amsterdam (R.J.P.M.S.); Julius Center for Health Sciences and Primary Care, University Medical Center, Utrecht (W.O.); private practice, Monnickendam (C.G.v.d.P.); Institute for Research in Extramural Medicine, VU University Medical Center, Amsterdam (D.B., L.M.B.); Netherlands Institute for Health Services Research (Nivel), Utrecht (W.L.J.M.D.).
We reviewed 17 studies that examined the accuracy of physical diagnostic tests for assessing ACL ruptures of the knee. Of those tests, the pivot shift test seems to have favorable positive predictive value, and the Lachman test good negative predictive value. The anterior drawer test is of unproven diagnostic value in this setting. In view of the potential biases in the original studies, however, the accuracy of the various ACL tests might be overestimated and the poor quality of the studies impede sound conclusions about the usefulness of the tests for daily practice. In addition, no study has been performed in primary care.
Because test characteristics may be influenced substantially by referral filters leading to spectrum bias,31 and because primary care physicians will be less experienced in performing these tests, the tests will presumably be less accurate in a primary care setting. Furthermore, the pivot shift test is very difficult to perform, making it less attractive for the average primary care physician.
Future research
Useful answers would be derived from sound research on the diagnostic accuracy of the various tests (determined for each test separately and for all tests jointly) combined with patient characteristics (eg, age, physical fitness, and functional demands) and elements of the medical history (eg, type of trauma and nature of the complaints). The emergence of MRI will facilitate this research. Relevance to clinical practice would be enhanced by an assessment of the effect of a correct diagnosis on the functional outcome of patients.
Acknowledgments
The authors thank Afina Glas, MD, and Professor Koos Zwinderman, PhD, for their statistical advice. We are much obliged to 1 of the referees for his/her useful suggestions.
Corresponding author Rob J.P.M. Scholten, MD, PhD, Dutch Cochrane Centre, Department of Clinical Epidemiology and Biostatistics, J2-273, Academic Medical Center, P.O. Box 22700, 1100 DE Amsterdam, The Netherlands. E-mail: r.j.scholten@amc.uva.nl.
