help rocmic and help fastmic and help rocmicd and help fastmicd 
                                                         Version 1.0 31-05-2009



rocmic -- Calculates the minimally important change (MIC) thresholds from continuous outcomes using ROC curves and an external reference criterion. fastmic -- As above, but suppresses the ROC curve for a faster process when using the bootstrap option. rocmicd -- Works in the same way as rocmic but is used to calculate minimally important deteriorations. fastmicd -- Works in the same way as fastmic but is used to calculate minimally important deteriorations.


rocmic refvar classvar , scale(minimum scale unit)

fastmic refvar classvar , scale(minimum scale unit)


rocmic estimates minimally important change (MIC) thresholds using two slightly > different methods. The first is the cut-point corresponding to a 45 degree tangent line intersection; this is mathematically > equivalent to the point at which the sensitivity and specificity are closest together (Farrar et al, 2001). The second is the cut-point corresponding to the smallest residual sum of sensi > tivity and specificity; this methodology has been proposed by researchers fro > m the EMGO Institute (de Vet et al, 2009).

The refvar should be the external criterion variable and must be either 0 or 1; > 1 representing an improvement in health status. The classvar should be the c > hange score variable (baseline minus follow-up). The minimal scale unit is the smallest increment me > asured by the instrument. In contrast to roctab, which when used with the opt > ion detail, presents sensitivity and specificity greater than or equal to each cut- > point, this program naturally calculates sensitivity and specificity for valu > es greater than a corresponding cut-point. Thus, to obtain the MIC, the scale's minimal in > crement must be added, in cases of improvement, and subtracted in cases of de > terioration. While in this prototype program, it is necessary to add this information, in fu > ture versions of the program I anticipate this will be unnecessary.

The program also calculated the ROC AUC with a 95% confidence interval in the s > ame way as the command roctab and produces a graph of sensitivity and sensiti > vity and plots a ROC curve (although the latter function is suppressed when using fa > stmic).


+------+ ----+ Main +-------------------------------------------------------------

There are currently no options available for rocmic. However, it will work with the bootstrap command. The program stores the MIC estimate for the 45 degree tangent line method in r(mic) and the EMGO MIC estimate is stored in r(emgo).


. rocmic ref change, scale(1)

. rocmicd ref change, scale(0.1)

. bootstrap MIC=(r(mic)): fastmic ref change, scale(1)

. bootstrap MIC=(r(mic)), reps(1000): fastmic ref change, scale(1)

. bootstrap MIC=(r(emgo)): fastmicd ref change, scale(1)

. bootstrap MIC=(r(mic)) MICemgo=(r(emgo): fastmic ref change, scale(1)


Farrar JT, Young JP, Jr., LaMoreaux L, Werth JL, Poole RM. Clinical importance of changes in chronic pain intensity measured on an 11-point numerical pain rating scale. Pain 2001;94(2):149-58.

de Vet H, Terluin B, Knol D, Roorda L, Mokkink B, Ostelo R, et al. There are three different ways to quantify the uncertainty when 'minimally important change' (MIC) values are applied to individual patients. J Clin Epidemiol 2009(IN PRESS).

R. Froud, S. Eldridge, R.Lall, M.Underwood, Estimating NNT from continuous outcomes in randomised controlled trials: Methodological challenges and worked example using data from the UK Back Pain Exercise and Manipulation (BEAM) trial ISRCTN32683578. BMC Health Services Research 2009 IN PRESS.


Thanks to my PhD supervisors S. Eldridge and M.Underwood.


Robert Froud

Also see

Manual: [ST] bootstrap [ST] roctab [ST] logistic [ST] lsens