Judicious and Vigilant Use of Electronically Based Treatment Enhancements and Subjective Patient-Reported Outcomes: Commentary on an article by Richard L. Skolasky, ScD, et al.
The article by Skolasky et al. gives us some insight into the potential of a simple telephone intervention to help to engage patients in their treatment and recovery. The promise for improving patient outcomes is clear and persuasive. Nevertheless, there are several issues concerning the methodology in this article that are noteworthy for the reader.
Patient-reported outcomes are a relatively new approach to evaluating the results of treatment. The U.S. Federal Agency for Healthcare Research and Quality (AHRQ) has published a series of 6 monographs to help the reader to position patient-reported outcomes in the overall scheme of outcomes instruments1.
The other methodology issues with patient-reported outcomes center on the concern of eliminating bias in a patient-dependent, subjective evaluation of outcomes.
Any patient-reported outcomes questionnaire used in clinical research should be validated. The process of confirming a statistically valid instrument has been defined and refined over the years. Most readers will be familiar with the Oswestry Disability Index2. This is probably the most commonly used outcomes instrument in spinal research. It was developed by Jeremy Fairbank and his team at Oxford University in the United Kingdom. Fairbank and Pynsent published a subsequent article that outlined the process of validation of the Oswestry Disability Index3. Components of the validation process from face and content validity through the statistical analysis, including the establishment of clinically meaningful change, are described. It is an excellent resource document for the reader and researcher.
One should also employ a methodology that negates the placebo effect in patient-reported outcomes4. There are concerns that the simple act of paying more attention to the patient (e.g., with phone calls) can result in better compliance and improved outcomes compared with patients with no such interaction. As Kaptchuk and Miller emphasized in their article on “Placebo Effects in Medicine” published by the New England Journal of Medicine in 20155, “placebo effects are not just about dummy pills.” Readers should note that Skolasky et al. have addressed this issue in their methodology by having equal numbers of phone calls in both the treatment and control groups. Further, the time spent in conversation with the patients was the same. Only the content of the conversation differed between the treatment and control groups. This allowed the patient’s engagement in rehabilitation to be analyzed as an independent variable based on the specific interventions that occurred during the phone call in the treatment group.
With the advent of personal communications platforms such as the Internet, e-mail, social media, and texting, we are likely to see a proliferation of publications using patient-reported outcomes instruments as the primary treatment evaluation method. However, readers of these articles and the researchers need to keep the potential for the patients’ subjective biases in mind. This was brought home on a personal level several years ago when the results of a patient satisfaction survey done in our office were analyzed. The best satisfaction with treatment in the practice was not associated with how well the patient did after the surgical procedure, nor the time the physician spent with the patient, waiting time in the office, lag time to obtain an appointment, location of the office, or parking; the highest level of patient-reported subjective satisfaction was dependent on whether the patient left the office with a narcotic prescription. Pause for thought on potential implications.