Predictors of the efficacy of epidural steroid injections for structural lumbar degenerative pathology.
This study which has been accepted and is currently pending publication in The Spine Journal presents an interesting predictive model for lumbar epidural steroid injection efficacy for those who have specific defined degenerative pathology. Treatment success for this study was defined by achieving at least a 7.1% improvement on the Oswestry Disability Index (ODI).
Using logistic regression models, the authors of this study concluded that the presence of central stenosis, a transforaminal or interlaminar lumbar epidural steroid injection, and a diagnosis of disc herniation significantly increased the odds of achieving a “minimum clinically important difference” (MCID) on the ODI.
There are many concerns with the methodology of this study. First, the MCID of 7.1% was based off an anchor-based approach. This approach compares changes in treatment scores for a specific variable (“anchor”). The potential disadvantage of this approach is the variability in the outcome instrument (ODI in this case) and sample being studied. In our opinion, the diagnosis groups studied (spondylolisthesis, disc herniation, and spinal stenosis) would each have a different threshold to define treatment success. This study did not clearly define severity of the specific disease process.
Of the other concerns noted, over 63% of the study participants had symptoms exceeding 12 months and 57% had prior surgery. The paper does not clearly define if the injections given were at the same level(s) where prior surgery was performed. There is a strong potential that the presence of post-operative scarring would affect the outcomes of this study. It is also not clear from the methodology how the degenerative pathology was diagnosed and if the ESIs were performed at more than one level. It is also not clear how the type of ESI (transforaminal, interlaminar, caudal) was selected for each patient. A disproportionate percentage of patients (72.8%) had a transforaminal injection. The number
of injections given per patient also varied (“45% had additional injections following the baseline injection”).
In summary, there are many concerns / criticisms in regards to the overall methodology utilized in this study. Limited descriptive data and subgroup analysis was presented. The reviewers of this study question whether 7.1% MCID on ODI really has clinical significance? In our opinion, the results of this paper must be interpreted with extreme caution given the above referenced criticisms. Generalizablity of this data may also be limited given such a high percentage of patients who have had previous spine surgery. The results showing that central stenosis is a positive predictor for achieving clinical success (as defined by MCID) also contradicts findings from several other recent published studies. It is our opinion that a predictive model for determining success with epidural steroid injections would hold great value and is currently needed.