INTRODUCTION
Selecting patients with lumbar degenerative spondylolisthesis (LDS) for surgery is not always straightforward. Appropriate use criteria (AUC) have been developed to clarify the indications for LDS surgery, but their validity has not been assessed in controlled prospective studies.
METHODS
This was a prospective, controlled, multicentre (5 in Switzerland; 1 in USA), study of 908 patients (561 surgical and 347 non-surgical controls; 69.5 ± 9.7 y old; 623 (69%) female). Based on their baseline characteristics, their appropriateness for surgery was determined using the AUC. They completed the Core Outcome Measures Index (COMI) at baseline and 12 months' follow-up (FU). The treatment plan was decided by the treating physician, blind to the AUC, together with the patient. The influence of appropriateness designation and treatment received (controlling for confounders) on the 12-month COMI and the achievement of a ≥2.2-point reduction in COMI score (i.e. the Minimum Clinically Important Change score; MCIC) were investigated using multiple regression.
RESULTS
As per convention, appropriate (A) and uncertain (U) groups were combined for comparison with the inappropriate (I) group. Surgery of some type was considered A/U in 406/561 (72.4%) and inappropriate (I) in 114/561 (20.3%) of the surgical patients (7.3% unclassifiable); it was A/U in 182/347 (52.4%), and I in 133/347 (38.3%) of the non-surgical patients (9.3% unclassifiable). For the adjusted 12-month COMI, the benefit of surgery relative to non-surgical care was not significantly greater for the A/U than the I group (p=0.189). There was, however, a greater treatment effect of surgery over non-surgical care for those with higher baseline COMI (p=0.035), independent of appropriateness status. The groups’ adjusted probabilities of achieving MCIC were: 83% (A/U, receiving surgery), 71% (I, receiving surgery), 50% (A/U, receiving non-surgical care), and 32% (I, receiving non-surgical care).
DISCUSSION
The patients for whom surgery was deemed appropriate (A/U) and who actually received surgery achieved MCIC more frequently than did surgical patients in the I group or patients in either group (A/U or I) treated non-surgically. However, the AUC were not able to identify which patients benefited most from surgery relative to non-surgical care. The identification of other characteristics that predict a greater treatment effect of surgery, in addition to baseline COMI, is required to improve decision-making.