The Opioid Epidemic: Patient Satisfaction and Physician Prescribing

The following is an excerpt from an article that originally appeared on the Credentialing Resource Center on September 11, 2017. 

by Robert J. Marder, president of Robert J. Marder Consulting. He can be reached at rmarder8@gmail.com.

Every day in the news, you can’t help but hear about the terrible impact of overuse and addiction to opioids. As our country desperately searches for solutions, one of the questions being asked is how much patient satisfaction survey questions on pain management have affected physician prescribing patterns.

In 2016, concerns were raised to CMS by 26 U.S. senators and multiple medical and hospital groups regarding whether linking pain management survey questions to economic rewards and penalties for hospitals had put pressure on physicians to overprescribe pain medication. In response, in November 2016, CMS temporarily removed the pain management questions from its HCAHPS survey so it could evaluate the issue. To refresh your memory, the three questions in the original survey were:

• During this hospital stay, did you need medicinefor pain?
• During this hospital stay, how often was your pain well controlled?
• During this hospital stay, how often did the hospital staff do everything they could to help you with your pain?

Although CMS is still committed to measuring patient satisfaction with pain management, its reevaluation of these questions determined that the focus should shift from patients’ perception of pain control to satisfaction with staff communication about their pain management. CMS has since indicated that the pain questions will be added back to the survey in January 2018, with the first and third questions intact and the second question changed to:

• During this hospital stay, how often did hospital staff talk with you about how much pain you had?

CMS also indicated that the pain questions would still be tied to reimbursement through its value-based purchasing initiative. My question is whether this will impact physician prescribing behavior. Let’s look at this issue from two perspectives: patient satisfaction data studies and physician perceptions.

While it is beyond the scope of this column to provide a literature review of the topic, most of the studies using patient satisfaction data look at whether satisfaction with pain management correlates with overall satisfaction with the episode of care. This is a common approach for analyzing patient satisfaction data. It does not attempt to validate the questions or imply causality, but merely seeks to identify whether one question or a set of questions correlates with another.

Interestingly, the results of these studies have been mixed. A 2012 study in The American Journal of Medical Quality conducted at a single organization showed a relationship between satisfaction with pain management and satisfaction with the episode of care. However, the study noted that patients were more satisfied when the staff did as much as possible to alleviate their pain compared to whether their pain was fully controlled. In 2016, several articles using larger-scale data indicated no relationship between the pain management questions and overall patient satisfaction, a position that has also been held by the largest provider of patient satisfaction surveys, Press Ganey, on its website.

Now let’s look at the information available on physician perception regarding patient satisfaction with pain management. Over the past few years, anecdotal reports have emerged regarding the potential impact of the use of pain management satisfaction data on physician prescribing. Emergency physicians have been particularly concerned because, unlike with hospital inpatient surveys, for patients who are just seen in the ED and not admitted, surveys can be more accurately attributed to the physician who cared for them. This direct attribution has led to direct economic consequences for the ED physicians at many hospitals.

Similar concerns have been expressed by primary care physicians in hospital-owned group practices, again because of the ability to attribute a patient survey to a specific physician. In 2014, a survey published in Patient Preference and Adherence showed that more than 48% of doctors say they have prescribed inappropriate narcotic pain medication because of patient satisfaction questions.

Why is there a difference between the two perspectives? While the correlation approach can provide valid information for a hospital to improve its approach to patient care, it does not take into account the potential impact of the use of this data when tied to economic consequences. I believe it is naïve to think that economics don’t affect perception.

Based on my experience as a VPMA and as a consultant for more than 15 years helping hospitals use patient satisfaction data to improve patient care, I am well aware that use of correlation data is often driven by the goal of raising a hospital’s overall patient satisfaction score in a comparative database. The impact is to improve the hospital’s standing in publicly available data for marketing purposes and its reimbursement from payers, including CMS, which use the data as part of value-based purchasing.

I have no doubt that measuring patient satisfaction with pain management has encouraged caregivers to be attentive to patient needs regarding pain. Unfortunately, once economic incentives or penalties are added to stimulate those improvements, we must ask whether the stimulus has an unintended trickle-down effect on physician prescribing behavior. While there might not be a national statistical correlation between pain management questions and overall patient satisfaction, the use of patient satisfaction data at the organizational level has resulted in the physician perception that these questions can affect their pay and ultimately their prescribing behavior.

Just as patient perceptions are valid, physician perceptions are equally valid and important in understanding this issue. Telling physicians that there is no correlation will not change their perception at the local level, especially when that perception is based on a reduced paycheck.