Establishing The Minimal Clinically Important Difference And Detectable Change Of The Stroke Rehabilitation Assessment Of Movement

ObjectiveThis article was written to establish the minimal clinically importantdifferece and detectable change of the Stroke rehabilitation assessment ofmovement scale(STREAM) and its3subscales (including upper extremity,lower extremity, and mobility subscales). At the same time, we detected thereliability of the Chinese version of this scale.MethodsBoth anchor-based and distribution-based methods were used to determinethe MCID. In the anchor-based method,60patients with stroke wasevaluated before and after the rehabilitation treatment (rehabilitationprogram for2to4weeks) by using the STREAM and the Likert-type scaleof seven grades. The mean change scores on STREAM and its3subscales ofthe MCID group, based on the patients’ ratings on the Likert-typescale(scored on+2to+3), served as the estimates of the MCID. In thedistribution-based method, we first detected the intra-rater and inter-raterreliability of STREAM scale and3subscales, then calculated the Standarderror of measurement (SEM) according to the formula and denoted it as theMCID. Eventually, we used the SEM to calculate the MDC (90%CI).ResultsIn the anchor-based method, There were23,18,15and13patients included in the MCID group of the four scales (upper extremity, lower extremity,mobility subscales and total STREAM scale), calculated MCID values were1.78,1.55,2.87and4.85, respectively; In the study of the distribution method,There were60and39patients finished the intra-rater test and the inter-ratertest respectively, the four scales`coefficient correlation of the intra-ratertest were0.935,0.982,0.961and0.973, according to the formula, the MCID(SEM values) were1.46,0.74,1.54and2.90, the MDC (90%CI) were3.4,1.7,3.6, and6.8, They accounted for the percentage of the total score of17%,8.7%,11.9%and9.7%; the coefficient correlation of the inter-rater testwere0.910,0.960,0.984, and0.976, MCID values were1.64,1.09,0.95and2.63, the MDC (90%CI) were3.8,2.5,2.2and6.14, They accounted for thepercentage of the total score of19%,12.7%,7.4%and8.8%. Conservativeprinciples, We finally selected the most greater one as the MCID values, sothe final MCID value of the four scales were1.78,1.55,2.87and4.85,respectively; The MDC values (percentage of the total score) were3.4~3.8(17%~19%),1.7~2.5(8.7%~12.7%),2.2~3.6(7.4%~11.9%) and6.1~6.8(8.8%~9.7%).Conclusionwe established the minimal clinically important differece and detectablechange of this scale. In clinical or research work, if the mean change scoreson the four scales (upper extremity, lower extremity, mobility subscales andtotal STREAM scale) within a stroke group have reached1.78,1.55,2.87and4.85points, the change scores can be perceived by patients as clinicallyimportant, we can assume that there is a clinically meaningful improvementin motor function and life activities function and beyond measurement error.The minimal detectable changes of the Stroke Rehabilitation Assessment of Movement measure are useful for both clinicians and researchers todetermine whether the change score of an individual patient is real andstatistical significantly. By the same time, Our research show that thisChinese version of Stroke Rehabilitation Assessment of Movement alsohave an excellent test-retest reliability.