Response by Mistry et al regarding letter to the editor, “Mechanical Thrombectomy in Patients with Ischemic Stroke with Prestroke Disability”

We thank Appleton et al. for their interest in our paper. ¹ Their questions are certainly worthy of a discussion. The primary outcome of our study, defined as no increase in disability or 90-day modified Rankin score 0–1, was selected a priori to understand if patients with pre-stroke disability return to their functional baseline after thrombectomy at a similar rate as those without a significant pre-stroke disability. We assumed that if a patient has a certain level of a permanent disability prior to a devastating stroke then retaining that functional status would be an acceptable quality of life for an individual patient. Like Appleton et al point out, since this is a relatively evidence free zone, our goal was to understand the rates of efficacy and procedural safety first. We agree that ordinal regression analysis is an attractive statistical analysis method for 90-day modified Rankin score (mRS).² It yields a common odds ratio of having a worse outcome on an ordinal scale. ³ Ideally, the two groups being compared with ordinal regression are alike at baseline and the only factor affecting the “shift” across the scale is the group assignment. The interpretation of the common odds ratio becomes difficult when the two groups start at a different score of the scale at baseline, or one group is already “shifted” on a scale at baseline, as is the case in our study. Thus, for the simplicity of understanding of this relatively unstudied group of patients, we chose a binary outcome for analysis.

We agree that mRS is not an ideal measure of pre-stroke disability. However, in our retrospective observational study, this was the only measure of disability readily available for analysis both pre- and post-stroke. Like Appleton et al. point out, there were patients who seemingly improved in mRS after their stroke. We attributed this to modest inter-rater reliability of mRS and accepted a difference of 1 point on mRS.⁴ We plan to expand the measure of disability to incorporate more sophisticated methods in our future prospective studies.