gms | German Medical Science

Objective: The benefits of spine interventions for degenerative pathologies are usually evaluated through assessment of patient reported outcomes. The minimum clinically important difference (MCID) as a threshold should represent the smallest improvement considered worthwhile by a patient. As some authors try to establish the MCID concept we re-calculated pre-existing data concerning the outcome of patients treated with dynamic stabilization due to lumbar degenerative disease to understand whether MCID is of any value.

Method: Different threshold values have been proposed as MCID for those instruments ODI (Oswestry Disability Index), PCS (Physical Component Score/SF36), MCS (Mental Component Score/SF 36), VAS (Visual Analog Scale) despite a lack of agreement on the best MCID calculation method. Anchor-based methods are “within-patient” score change, “between-patient” score change and “sensitivity and specifity based approach” i.e. Receiver Operating Characteristic (ROC) analysis. They are calculated in correlation with an graduated PSI. Distribution-based calculation methods compare the change in scores to some measures of variability such as standard error of measurement (SEM), half standard deviation (SD/2), Small Size Effect and Minimum Detectable Change (MDC). Data collection was completed in 95/103 treated patients (median FU: 24 months). Clinical assessment using standard scales (ODI, PCS, MCS, VAS, PSI) was acquired prospectively in pre-defined time intervals.

Results: Anchor based results for ROC analysis (Graphs in ppt) showed a cutoff of 4.5 for PCS and a cutoff of 3.5 for MCS which indicates that neither PCS nor MCS are good predictors for patient satisfaction. A cutoff of 58.5 for VAS and a cutoff of 27 for ODI showed that those parmeters are well correlated with PSI. Distribution based results for ODI are 4.6 for SD/2 calculation, 1.8 for Small size effect, 2.9 for SEM and 8.1 for MDC. For PCS and same for MCS 1.5 with SD/2 calculation and 0.6 with Small Size Effect were calculated.

Conclusions: As long as there is no consensus among statisticans which calculating method for calculating MCID should be used as standard no real conclusion can be drawn. No reference values have been established to classify and rate own results. Thereby VAS and ODI seem to be good predictors for patient satisfaction. SF 36 measures such as PSC and MCS are not that conclusive in predicting patient satisfaction as demonstrated in our ROC Analysis.