Population management program improves CV risk factor control in diabetes

Darren McGuire, MD

While analyzing secular trends in health care performance in general and for type 2 diabetes targeted risk factors specifically, as done here, is not new, there are a few added bells and whistles here that are noteworthy. No. 1, using the national contemporary data for comparison and contrast is useful in supporting the likelihood that the measured changes in quality performance were driven at least in part by the PHASE program and not just a sign of the evolving care in American medicine.

No. 2, except for the nurse/pharmacist navigator effort, the remainder of the PHASE intervention can largely be automated within a common electronic medical record system, so there is little effort required beyond the programming — and likely substantial clinical benefits for patients and for populations.

I particularly like the quarterly reporting of metrics to each center. When reported in the context of the program average and the top performing centers, this can prove a potent instrument to raise the level of care at each center performing below the top tier. We have used inpatient and outpatient cardiology registries in this way for decades with consistently good results.

Single clinics or centers could relatively easily apply some facets of the PHASE program, but only with the assumption that the results are attributable in part to each of the four parts of the intervention. For example, if a clinic applies all but the nurse/pharmacist part, one cannot be certain from these data what proportion of the improvement might be affected, or is all of the change observed exclusively attributable to the added role/interventions of these dedicated providers? Providers, clinics and systems can be and should be tracking the key evidence-based metrics and feeding back results for continued improvements in such interventions.

Programs like Kaiser Permanente Northern California have tremendous potential to embed research like this into everyday practice. For future such programs, the system might consider “cluster randomization,” with clinics being the unit of randomization to be able to more clearly separate effects of the program from secular trends within the health care system, which may not necessarily be present in the nationwide experience. Then among those randomized to the program, a secondary randomization to apply one, two, three or all four of the components would make it possible to see which is/are most important and which did not contribute at all.

First, though hyperglycemia is a prognostic risk factor for atherosclerotic vascular disease, pharmacological interventions targeting improved blood glucose have not yielded atherosclerotic CVD benefits. While important for microvascular disease risk mitigation, and in that context, quality patient care, the reporting, as is done here, tends to perpetuate the misperception that glycemic control is a metric of quality CV care. This bias of the authors is reflected in the ordered presentation throughout the report: “glucose, cholesterol and blood pressure.”

Also, acknowledging the intervention began in the early 2000s, the thresholds used to identify targeted LDL and BP are debatable — which was the case even when the program started. Most would consider LDL < 70 mg/dL and BP below 130 mm Hg systolic/80 mm Hg diastolic more evidence-based and more potent atherosclerotic CVD prevention targets. The authors might have addressed this by meeting in the middle and, in addition to the data presented that they selected a priori, could have also reported the frequency of achieving these more aggressive targets. Capitalizing on what are missed opportunities for a more global assessment of atherosclerotic CVD prevention in type 2 diabetes would have meant analyzing the frequency of use of ACE inhibitors or angiotensin receptor blockers  independent of BP, the frequency of use of daily aspirin for the highest-risk patients at least for those with prevalent atherosclerotic CVD and the frequency of smoking abstinence or, for those actively smoking, the frequency of counseling or referral for cessation programs.