Comments on: Validation Study of an Electronic Method of Condensed Outcomes Tools Reporting in Orthopaedics (J Knee Surg 2013;26:445–452)

Farr and co-authors are interested in providing clinicians and researchers with the ability to administer multiple knee scores and the SF-36 by one electronic “superquestionnaire.” Items and answer options for the different patient-reported outcomes (PROs) that were included in the “superquestionnaire” were reworded when “similar” and “redundant” items were thereby removed to reduce the response burden for the patient. The authors calculated correlation coefficients between paper and electronic versions of the respective questionnaires and the values of these correlation coefficients were compared with “historical” test-retest reliability correlation coefficients. Because these correlations did not appear to be different, the authors interpreted this to mean that the electronic “superquestionnaire” provided valid versions of the originally formatted IKDC Subjective Knee Form, KOOS, Lysholm, and SF-36.

We are concerned that the methods applied do not verify or allow data obtained by this “superquestionnaire” to be presented or interpreted as similar to data obtained by the original PROs. Changes in the wording of the items may influence the interpretation and meaning of the item by the respondents, thus changing the interpretation and meaning of the resulting score. Additionally, it appears that the response scales for many of the items were modified, making comparison to the original PRO scores invalid. It appears that the authors determined the relationship between the “superquestionnaire” version of the outcome measures and the scores from the original versions of the PROs using either Pearson or Spearman correlations, which only describe the general relationship between the scores (high scores on the “superquestionnaire” version are associated with high scores on the original version of the PRO). This is not the same as saying that the two scores are exactly equal in value. Thus, these results do not support replacing the original PRO scores with the scores obtained from the “superquestionnaire” version of the outcome measure.

There is no valid reason to obtain data by multiple knee-specific questionnaires from a patient. Instead, this approach allows for introduction of bias when the score showing the best result can be arbitrarily chosen for presentation. This approach is not compatible with high-quality design of outcomes studies.

Finally, no conflicts of interest are stated by any of the authors. As readers of the Journal of Knee Surgery, we would appreciate verification that this study was undertaken by authors not otherwise involved in the marketing and sales of this “superquestionnaire” or any other related products.