IN OUR 2016 ARTICLE, examining the responsiveness of the Disorders of Consciousness Scale (DOCS-25), we reported the minimal detectable change (MDC) along with several other indices of responsiveness including anchor and distribution-based minimally clinically important differences (MCIDs).1 Similar to others, we used a formula in which the SEM was included within the square root. Bland2 points out the correct formula for the MDC is when the standard error of measurement (SEM) is external to the square root (MDC₉₅ = 1.96 x SEM x [square root]2).2 We have recalculated the MDC for Rasch-transformed DOCS-25 person measures using this formula: 9.98, 11.22, and 11.47 for nonimprovers, improvers, and all participants, respectively. These MDC indices apply to Rasch-transformed person measures and not to total raw scores. The revised MDC is somewhat larger than our previously reported anchor-based MCID (8.6).1

As noted in our earlier article, MDCs can be clinically useful, particularly in early phases of recovery when patients may demonstrate fluctuating levels of neurobehavioral function on a day-to-day basis. Knowing when such change is beyond measurement error better enables clinicians to identify when variation is consequential enough to warrant attention. Clinicians may find MCIDs useful for informing treatment decisions such as when a change in intervention strategy may be warranted. In addition, anchor-based MCIDs may support clinicians to engage families in discussions about treatment goals.

We encourage readers to use discretion when applying this type of MDC since the calculation assumes that measurement error is consistent across all total raw scores, which it is not.3 For Rasch-based measures, the standard error is larger at the ends of the range and smaller in the middle of the range; for raw score scales the standard error is larger at the middle of the range and smaller in the ends.3 To provide rehabilitation clinicians with person-centered indices of change, future studies could examine alternative MDC approaches conditioned on patient admission and discharge measures.4

REFERENCES