Are We Measuring Value in Primary Care?
Mrs. Susan Stevenson (identifying details altered) is a 63-year-old female who first presented to a physician assistant (PA) with a seven-year history of type 2 diabetes. Prior to seeing the PA, her care team included a PCP as well as an endocrinologist and her Hemoglobin A1C had been uncontrolled for several years. Her “A1C” was 10.5%, indicating that her current approach using food choices, physical activity, and two oral medications were not working and was putting her at risk for spending a lot of her retirement time and money in doctor’s offices.
Not wanting his patient to continue to be exposed to the harmful effects of high glucose levels, our PA suggested she consider starting insulin. But the patient did not want to take insulin. Given that this clinical scenario (and millions of others that share similar aspects) occurred at a time when many are working to improve healthcare by shifting from “fee for service” care to “fee for value”, this situation and the ensuing outcome raises the questions: What would high value care would look like for this patient? Do we have standardized measures that accurately assess the value of care she received?
Often our misuse or misunderstandings of words and phrases are inconsequential (with apologies to my relatives who were English majors). For example, I am nonplussed when someone uses “begs the question” to communicate that a question has been raised. But the lack of a shared understanding of the word value amongst healthcare payors and providers is impeding efforts to improve patient and provider experience, improve clinical outcomes, and decrease costs.
Economist Michael Porter put it this way: “In any field, improving performance and accountability depends on having a shared goal that unites the interests and activities of all stakeholders. In health care, however, stakeholders have myriad, often conflicting goals…. Lack of clarity about goals has led to divergent approaches, gaming of the system, and slow progress in performance improvement….[V]alue in healthcare remains largely unmeasured and misunderstood.”
At the recent Bridge to PopHealth conference, one speaker defined value as (Quality + Service) / Cost (where Service = Patient Experience + Access). Including patient experience when assessing value is important because (i) as Porter states, “value should always be defined around the customer”, and (ii) research suggests that patients often care more about the experience of care than clinical outcomes.
If Value = (Quality + Service) / Cost, how well are we measuring it? Clearly, mechanisms are being implemented to measure value when “episodes of care” can be defined, as in the case of joint replacements. But how well are we measuring value for the care of chronic conditions where discreet units are not so easily defined, such as care for patients with diabetes?
In my opinion, attempts to incorporate clinical value into reimbursement too often use only quality, which is related but not synonymous. For example, if assessing the performance of care to diabetic patients only uses a clinical outcome measure (i.e. the percentage of diabetic patients achieving an A1C < 7%), we are not evaluating the value of care for diabetes (e.g. (Quality + Service) / Cost). This is not just a semantic problem. Incentivizing based on quality, often measured by clinical outcomes alone, will not achieve the same results as incentivizing value. Not all patients who have attained a goal A1C have received the same value of care.
I agree wholeheartedly with the Utes who stated: “We can no longer afford time nor energy wasted in pursuit of isolated improvement. Improving quality without a careful understanding of cost, or reducing cost without measuring the impact on outcomes or experience decreases patient value. Once clinicians, teams, and the organization have the measures at hand, finding solutions that improve one or more outcomes without raising costs or lowering quality becomes the norm.”
So what happened with Mrs. Stevenson? Through a bit of negotiation and shared decision-making, both the PA and the patient landed on the following plan: she committed to making specific, measurable and attainable goals related to physical activity, food choices, and glucose monitoring, and they agreed to regularly interact with each other in between visits to assess how she was doing in pursuit of these. Within 6 months, Mrs. Stevenson’s achieved an A1C of 7% without insulin.
At least during the 12-month episode during which our clinic team cared for this patient—from when the patient presented in an uncontrolled state, to when her A1C was controlled—qualitatively it seems clear that she received high-value care:
- Quality was high, evidence by clinical outcomes (A1C 7) achieved
- Service was high: Patient experience was high (she did not have to go on insulin, as was her desire) and she had a high level of access to her PA
- The costs would be expected to be relatively low (no insulin, in-office visits minimized through the use of frequent interaction through patient portal, i.e. TwineHealth).
In my opinion, this represents a higher value than achievement of the same A1C had the PA merely insisted that the patient start insulin. Yet it seems current measurement schemes would not differentiate the two from a value perspective highlighting the shortcomings of our current approach to measuring value. For our healthcare system to make the needed transition to fee for value, payors and providers need a shared understanding of clinical value, as well as measurement tools that accurately discriminate between low value and high-value care. My hope is this story begs the question, “What is value in primary care?” amongst purchasers of and providers of healthcare alike so that if we are nonplussed in three years about the transition to “value-based care” it will be due to a profoundly positive impact.