Patient Satisfaction Verus Treatment Effect

One our of colleagues, Dr. Steve George, just published a nice paper characterizing the ability of a particular questionnaire item assessing patient satisfaction to distinguish between satisfaction with treatment effect versus satisfaction with treatment delivery after physical therapy treatment of low back pain. The findings suggest that patient satisfaction with symptoms is weakly associated with other satisfaction items and more related to self-report measures of treatment effect. The results remind us that patient satisfaction is a multi-dimensional construct and that a patient being "satisfied" with care may have little to do with whether the patient is actually better.

Steve, feel free to comment on any particular issues you think might be pertinent for the practicing clinician based on the findings from this study.

John

Download george_arch_phys_med_rehab_2005.pdf

Comments

[Steven George]

Hi John. Thanks for the chance to comment on this article...

I think the biggest take home message for clinicians is that they need to carefully select items if they are currently using or are planning to use patient satisfaction as an outcome measure.

Satisfaction questionnaires are often loaded up with many (or all) items that assess satisfaction with treatment delivery. As a result, only one component of a "multidimensional satisfaction construct" is assessed.

Also troubling is that satisfaction with the most common item type (treatment delivery) is often very high, and lacks variability needed to adequately discriminate outcomes. For example, in our study patients rated their satisfaction with the 2 treatment delivery items at close to 90%. The cause of "high" patient satisfaction when measured like this can be from several sources (outlined in the paper), some of which have nothing to do with patient being "satisfied" with the results of their treatment.

In contrast, patient satisfaction with symptoms was rated much lower, about 40%. And as you already mentioned, satisfaction with symptoms was significantly associated with changes in physical function, pain intensity, and pain bothersomeness. The other treatment delivery items in our study, were not consistently associated with factors indicative of an improvement in status. Again, this suggests that patients are likely to say there were satisfied with their treatment delivery, regardless of the outcome.

So, if you have an interest in measuring patient satisfaction, make sure you consider factors other than those related to treatment delivery. This will give you a better perspective on whether the patient was satisfied with the physical therapy process and also its outcome. I would think this would be especially important for clinics that look to use patient satisfaction ratings in performance evaluations.

Another interesting issue is the 2 factors (patient expectations met and symptom bothersomeness) that best predicted patient satisfaction at 6 months are not typically measured by PT's. I am curious about what other people think of that. At UF, a psychologist I work with is currently looking at ways to assess patient expectations of treatment for chronic pain, and how it affects satisfaction with outcome and future health care utilization. We have a preliminary paper in press and if there is any interest, I can post it when it is published as I think it is interesting and relevant to some issues in rehabilitation.

[Selena Horner]

I would assume that patient expectations generally are not assessed when analyzing satisfaction because those interested in capturing satisfaction have a better ability to control treatment delivery and the processes involved in treatment delivery.

Patient expectations may have a large amount of variability and maybe when analyzing satisfaction, those analyzing the satisfaction may be fearful of not having the ability to meet or surpass patient expectations as consistently which would actually lower the satisfaction scores.

There was some article that I read either here or on Medscape reporting that orthopaedic surgeons seem to be quite satisfied with total knee replacements, whereas the patients had substantially less satisfaction from the procedure. If I remember correctly, the patients reduced satisfaction had to do with their expectations not being met. It was quite interesting.

[Lisa M.]

My employer recently posed a question to the staff which I am having difficulty finding a good answer for. It is regarding patient satisfaction. It seems we are scoring low on the ability to reduce pain. The question was...how can we improve the patients satisfaction in decreasing their pain? Have you ever put together a survey that has been presented to the patient regarding pain reduction? I often find that the patients do not rate their pain level in accordance with the level of function they have achieved (i.e. increased range of motion, increased tolerance of ADL's, but no change in pain rating). Any ideas on how we can achieve better results?

[Laurie]

We have long suspected that true patient satisfaction is a complex issue that has surprisingly little correlation to the treatment process and more to do with the overall patient “experience”. With that in mind, we recently revamped our Patient Satisfaction Survey and even went as far to rename it a Patient Loyalty Survey. This came as we realized that having satisfied patients was not our goal. We wanted loyal patients. We wanted to develop patients that would sing our praises, refuse to seek treatment from another provider even when referred elsewhere, and would establish a singular relationship with their therapist much like they do their physician.

While we continue to be interested in treatment outcomes and do measure them separately, we realize the value in measuring experience-related items as well. The questions on our patient loyalty survey now also measure the “softer” elements of the physical therapy experience – convenience of appointment times, courtesy of the front desk, consistency in the treatment team, the patient's involvement in goal-setting and if they were given contact information for their therapist (our therapists often give patients their cell phone numbers and encourage them to call if a problem should occur outside of work hours) In addition, we ask if they have been a patient at one of out clinics previously, or if they are a friend/ family member of a former patient as we feel this may be our best singular indicator of patient “loyalty”.