Linking Patient Satisfaction Scores to Physicians’ Compensation

Five Ways to Survey your Patients

When it’s so easy to get caught up in the daily challenges of running a practice, how do you make sure data from patient satisfaction surveys—which often provide details your practice couldn’t get else­where—result in action? One way is to link results with physi­cian compensation.

I think we are overdue for physicians having personal risk in patient satisfaction – I don’t understand why other industries (e.g., airlines) are so far ahead of healthcare in aligning employee incentives with improv­ing customer satisfaction. It is simply good business sense to improve patients’ perceptions of your medical practice. Remember that it is less expensive to retain the patients you al­ready have than to find new ones. Keeping your patients happy also increases the chances that they’ll recommend your practice to family and friends.

To tie satisfaction to compensation, you need to employ a fair, reliable survey method to overcome physician resistance and ensure that your pay-for-performance program achieves its intended purpose. Choose a methodology that generates results that are as accurate, unbiased, and statistically valid as possible. You want to be able to defend how the outcomes affect physicians’ pay.

Five basic methodologies for surveying your patients:

1. Written surveys via computer kiosks at the office

This format is convenient, but respondents decide whether to take it, so the sample is not random. Some doctors may achieve higher response rates than others, resulting in an unbalanced number of surveys per physician. In addition, installing kiosks/computer terminals will gen­erate moderate to high cost com­pared to other survey methods.

2. Written surveys mailed to pa­tients

This method may draw responses from only the extremely happy or unhappy patients, skewing results. Substantial variability will likely occur in the number of completed surveys per physician, giving physi­cians fodder to challenge the statisti­cal reliability of the results.

3. E-mail surveys

E-mail surveys tend to achieve high response rates, are easy to take and tabulate, and are inexpensive. However, there is a high poten­tial bias toward computer- and Internet-savvy patients. However, as you know, computers are now commonplace and used by patients of all ages.

4. Telephone surveys via automated/voice response

Automated telephone surveys can easily generate an appropriately sized random sample for each physi­cian However, patients may find this survey method impersonal, resulting in low response rates.

5. Telephone surveys in real time

Although this approach is the most expensive, it might be the most appropriate method to tie to pay plans because it generates the best coverage of all patient segments (i.e., representing different health plans) and pro­vides a random sample, with high statistical relia­bility and identical sample sizes for each physician.

Select an appropriate scoring system

Next, determine how you’ll score the survey results and how they’ll affect physicians’ pay. As with any compensa­tion plan, the devil is in the details. Several factors will influence how much or little financial risk your physicians will have in connection with patient-satisfaction scores.

When developing your survey instrument or question­naire, keep the following points in mind:

  • Make survey questions answerable on a five-point scale (1 = poor, 5 = excellent) so you can calculate average grades for each question.
  • Establish a committee, including physicians, to help finalize the wording of questions (i.e., wording should be understandable to patients based on spe­cialty and local vernacular).
  • Assign a physician influence value, ranging from zero to five or zero to 10 that is decided by a committee.
  • Multiply these values by the grade on each question to find the score.

The influence value represents how much responsibility your group assigns to physicians for their grade on any given question. For example, the question, “How would you rate the comfort of the waiting area?” would probably deserve a physician influence value of 0 because physi­cians have little or no control over this aspect of the prac­tice. In contrast, the question, “How would you rate the physician’s style?” is completely physician-driven, earning an influence value of 10.

Apply scores to pay

Finally, decide how you’ll incorporate patient satisfac­tion scores into physicians’ pay. One method is to calculate each physician’s score (i.e., average patient grade x influ­ence value) for each question and add all scores together. Determine each physician’s percentage of the total score. The funds with which to reward patient satisfaction come from a pool of money the group sets aside for this purpose (e.g., 5% of the group’s compensation pool). That pool is then divided in accordance with each doctor’s percentage of the total score.

Additional Resources


Have questions? I’m here to help.