Rehabilitation following first-time lumbar disc surgery - A systematic review within the framework of the Cochrane Collaboration: A systematic review within the framework of the Cochrane collaboration

RWJG Ostelo, HCW de Vet, G. Waddell, MR Kerckhoffs, P. Leffers, M van Tulder, Henrica C.W. De Vet, Maria R. Kerckhoffs

Research output: Contribution to journalArticleAcademicpeer-review

117 Citations (Scopus)


Study Design. A systematic review of randomized controlled trials. Background. Although several rehabilitation programs, physical fitness programs, or protocols regarding instruction for patients to return to work after lumbar disc surgery have been suggested, little is known about the efficacy of these treatments, and there are still persistent fears of causing reinjury, reherniation, or instability. Objectives. The objective of this systematic review was to evaluate the effectiveness of active treatments that are used in the rehabilitation after first-time lumbar disc surgery. Methods. The authors searched the MEDLINE, Embase, and Psyclit databases up to April 2000 and the Cochrane Controlled Trials Register 2001, issue 3. Both randomized and nonrandomized controlled trials on any type of active rehabilitation program after first-time disc surgery were included. Two independent reviewers performed the inclusion of studies, and two other reviewers independently performed the methodologic quality assessment. A rating system that consists of four levels of scientific evidence summarizes the results. Results. Thirteen studies were included, six of which were of high quality. There is no strong evidence for the effectiveness for any treatment starting immediately post-surgery, mainly because of the lack of good quality studies. For treatments that start 4 to 6 weeks postsurgery, there is strong evidence (level 1) that intensive exercise programs are more effective on functional status and faster return to work (short-term follow-up) as compared to mild exercise programs, and there is strong evidence (level 1) that on long-term follow-up there is no difference between intensive exercise programs and mild exercise programs with regard to overall improvement. For all other primary outcome measures for the comparison between intensive and mild exercise programs, there is conflicting evidence (level 3) with regard to long-term follow-up. Furthermore, there is no strong evidence for the effectiveness of supervised training as compared to home exercises. There is also no strong evidence for the effectiveness of multidisciplinary rehabilitation as compared to usual care. There is limited evidence (level 3) that treatments in working populations that aim at return to work are more effective than usual care with regard to return to work. Also, there is limited evidence (level 3) that low-tech and high-tech exercises, started more than 12 months postsurgery, are more effective in improving low-back functional status as compared to physical agents, joint manipulations, or no treatment. Finally, there is no strong evidence for the effectiveness of any specific intervention when added to an exercise program, regardless of whether exercise programs start immediately postsurgery or later. None of the investigated treatments seem harmful with regard to reherniation or reoperation. Conclusions. There is no evidence that patients need to have their activities restricted after first-time lumbar disc surgery. There is strong evidence for intensive exercise programs (at least if started about 4-6 weeks post-operative) and no evidence they increase the reoperation rate. It is unclear what the exact content of postsurgery rehabilitation should be. Moreover, there are no studies that investigated whether active rehabilitation programs should start immediately postsurgery or possibly 4 to 6 weeks later.

Original languageEnglish
Pages (from-to)209-218
Number of pages10
Issue number3
Publication statusPublished - 1 Feb 2003


  • exercises
  • low back pain
  • lumbar disc surgery
  • postsurgery
  • rehabilitation
  • systematic review

Cite this