Satisfaction Guaranteed — the Right Way

Richard M. Plotzker, MD


September 07, 2022

People working on Madison Avenue know that we are all consumers at heart, partly seeking gratification before purchase, assessing whether we got our moneys' worth afterwards. With a car or house — these one-time expensive transactions — new buyers are more valuable than repeaters. Satisfaction with the transaction doesn't matter much. But for repeat purchases, loyalty matters. That brand of disposable diapers better not leak.

We've also become more critical of our consumer experiences with more forums to convey admiration and scorn. Political advertising, at least in the United States, has pushed many of us to vilify as our default option.

Medical care has joined the rest of the global economy. Those of us who see patients in office accumulate consumer ratings that are readily accessible to the potential patient. Some elements we don't control at all, like our race.  And we have little control over how a caller rates the person they speak with in our medical networks.

But the one thing we do control is the caliber of our medical decisions and our ability to navigate the healing or misadventures that follow. Some of our institutions track those ratings mostly as surrogate dollar signs, which generates some trepidation among those being scored. Others track them as a composite of patient care and process. But the most valuable to physicians, and to our patients, are those that bridge the formalities of a research tool yet remain applicable to clinical care. Even better when they lack punitive potential.

Such an assessment of patient satisfaction may be the Diabetes Treatment Satisfaction Questionnaire (DTSQ). This survey was developed about 40 years ago. It has undergone one revision since then to overcome a limitation of the original where baseline satisfaction was too high to detect incremental improvement. As a commercial product, it requires purchase by a practice or research grant, though sample pages of the questions are available by asking Google to display images of the search.

Unlike internet surveys that direct patients to what they think of us or our institutions, this survey restricts responses to assess care in several spheres: Convenience, flexibility, understanding, recommend to others, likely to comply with treatment, perceived hyperglycemia, and perceived hypoglycemia. But from these targeted data, inferences can also be made about the structural elements of care.

Diabetes: Global Problem, Variations in Care

Using this resource, a group in Japan undertook an assessment of a very common diabetic practice that all clinicians who see diabetic patients encounter. When glycemic control deteriorates on monotherapy, which it inevitably does as type 2 diabetes is characteristically progressive, how do patients respond to choices made by their doctors to add to the complexity of their care?

In the process, they also exposed the reality that diabetes is a global disorder with regional variations in care. As with many reported findings, their study — the Japanese subset J-DISCOVER — is a subset of the much larger multinational DISCOVER study. The goal is to assess the changes in treated type 2 diabetes, assessing clinical outcomes of different treatment decisions among patients with different clinical and demographic characteristics.

Recognizing that diabetes is largely a patient-managed disorder with professional oversight, they focused their data on how acceptable the medication decisions were to the patients who needed to implement them, using the DTSQ as their validated measuring survey. Their primary study ran for 3 years at 142 sites across Japan. They enrolled about 1800 patients, all on monotherapy for glycemic control. About half took a dipeptidyl peptidase 4 inhibitor, about one third used a biguanide, with the rest divided among the other available oral hypoglycemics. This baseline medicine distribution of initial monotherapy may be different in other countries.

Because all of the participants needed better control, the vast majority had a second agent added, though 14% were switched to a different monotherapy. Roughly 1% were prescribed an injectable, mostly insulin with a few put on a glucagon-like peptide 1­ receptor agonist.

The mean A1c was 7.7%, with about 28% having an A1c over 8%. About half had a body mass index (BMI) > 25, though more extreme obesity, which is so common in the US population, was not specified. Mean age was about 60 years, about 60% were male, and nearly 70% of the patients had an endocrinologist responsible for diabetic decisions, unlike the United States and other countries where patients with not extreme hyperglycemia are typically managed in a primary care setting.

The baseline DTSQ score was rather high at 25.9, but over the first 6 months of tracking the mean score increased to 27.3, where it plateaued for the remainder of the study. This increase appeared irrespective of what changes were made in the treatments throughout the 3 years of observation. Self-perceived hyperglycemia improved as treatment was intensified. Self-perceived fear of hypoglycemia remained low, even among those introduced to insulin or sulfonylureas.

So it looks like those patients were content with what their doctors did on their behalf, even before addressing the baseline hyperglycemia. The best improvements in satisfaction came from those who received the greatest glycemic benefit from the change, which usually came early in the course of treatment. Yet there were subsets of the cohort that became less satisfied over the 3 years of the study. Satisfaction scores did not increase in older patients, with scores of those aged 65-75 years staying about the same and those older than 75 years generating lower scores with more intense treatment.

Patients treated as regional hospital outpatients, about 20% of the total, expressed a sense of improvement that was not experienced by those treated at free-standing clinics. Those whose medication decisions were made by specialists other than endocrinologists showed a decline in satisfaction over time, as did those who lived alone. Patients with an estimated glomerular filtration rate < 60 mL/min/1.73 m2 also had a suggestive though not quite statistically significant decline in satisfaction as their treatment programs were modified.

It appears that getting good medical results, which is where physicians focus attention, makes for happier patients, even when they are inconvenienced by new medications or more medications. But assessment of satisfaction in the face of better medical results reveals other elements of medical care that are less obvious in our office encounters. Advanced age generated some pushback as did living alone or with marginal connections to others — things that cannot be overcome by better glycemic results.

There may be an element of confidence in the setting of care. Hospitals offer more one-stop availability of testing, dieticians, and diabetes educators (understanding of disease was a component of DTSQ surveys) than do the more limited facilities of a free-standing clinic. There also may be an element of confidence in the decision-maker, with the diabetes specialists having an advantage in prestige, if not the actual glycemic results. Even with the efficacy of medical care paramount, there remain social determinants and means of offering care that enhance or detract from how patients perceive their treatment.

And though this remains very different from the gestalt of 0 to 5 stars when somebody Googles us, in their discussion of limitations to their study, the authors acknowledge that the DTSQ intentionally omits assessment of certain parts of the patient experience.

It is reassuring to note that sound medical decisions with good outcomes remain paramount in patient assessments. But satisfaction scores as assessed in this and several other studies of diabetic care remain different than the quality of life score, which has its own literature on diabetes, though with overlaps. Necessary treatments and progression of disease usually burden people. This remains distinct from how people assess the care they have received and their willingness to accept that. Both measures are needed to enhance the care we offer. What may not be needed, and may even be detrimental, are those publicly broadcast ratings that our search engines generate, whether to our delight or dismay.

For more diabetes and endocrinology news, follow us on Twitter and Facebook

Follow Medscape on Facebook, Twitter, Instagram, and YouTube


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.