Wu and McLaughlin (2012) have cast a thoughtful eye on the topic of the physician gender gap in the arena of communication skills, which has been supported by our recent findings (Swygert et al. 2011), along with many others. They note that the findings are very consistent in the literature, but we think it’s worthwhile to point out an additional benefit—and some caveats—to including the results of our study, which used USMLE® Step 2 Clinical Skills (CS) data, in the discussion of whether the communication gender gap can, or should, be addressed within undergraduate medical education.

We hope that the findings from our paper add some unique information to this well-documented topic, due to the sample size of our study and the psychometric soundness of information from an exam that is standardized and controls for the gender of the (standardized) patient. That much said, some cautions about the generalizability of standardized test findings in general, and Step 2 CS in particular, should be mentioned. The skills observed on Step 2 CS are a limited subset of observable, measurable skills that represent a sampling from the larger domain of appropriate physician–patient communication. The sampling is intended to enable the test outcome to be generalizable to the entire communication skills domain, but because the set of skills measured by the exam is limited, any planned improvement or intervention in the medical school curriculum should aim to do more than “teach to the test.” On a related note, any planned curriculum intervention should be clear about the ultimate goals of that intervention. Instruction that targets students in need (perhaps more males than females, or perhaps modifying instruction by gender) could result in improvement as measured by the school while still not changing greatly the outcomes on Step 2 CS. Any potential curriculum change should include an explicit statement of measurable goals, of which improvement in Step 2 CS pass rates would be only one of those, and not necessarily the most important one.

Moving on to the general comments in the Wu and McLaughlin commentary, one could consider the argument that they introduce; that one potential response to the general findings is to selectively admit women to medical school over men with the same MCAT scores. However, the validity evidence for this as a selection process would require two demonstrated outcomes: until we can show that women (or women who score higher on a communication skills assessment) actually provide better care than those who score lower, and we can show that there is an improvement in patient outcomes correspondent with that care, it would be unwise to selectively admit women because of the anticipated performance after undergraduate medical training.

However, if a demonstrated improvement in patient outcomes could be linked to the level of communication skills in medical school, either via standardized tests or other measures, one could argue that, in addition to (or in place of) improving the medical education communication skills curriculum, one could include a standardized assessment at an earlier stage, so that women and men with better communication skills are selected for medical school at the start. In part, this may be possible with existing measures of cognitive skills that happen to be predictive of non-cognitive skills as well; the Verbal Reasoning portion of the MCAT, for one, has been shown to have modest success in predicting communication scores on physician licensing examinations (Kulatunga-Morvzi and Norman 2002). Another potential path might be to improve the selection of medical students via a more interactive, yet still standardized, interview process that focuses on a prospective student’s strength of non-cognitive skills; screening tools such as the CMSENS developed by Dore et al. (2006, 2009) and the MMI developed by Eva et al. (2009) have promising potential. While the caveats about the generalizability of standardized assessments outcomes would still apply, these types of measures have the potential to add additional value to the efforts of medical schools to improve the communication skills of their students and, in turn, improve patient care and satisfaction.