Expert Blog

Primary Care View on Case 2: Failing Metformin

Doron Schneider, MD, FACP
General Internist

Case 2: Metformin Fails in T2DM Patient represents a clinical dilemma faced by every primary care doctor. Patients, such as this one, who are initially diagnosed with diabetes, should be initiated on metformin and encouraged to engage in lifestyle modification. Multiple guidelines are in agreement about this approach.[1]  Unfortunately, diabetes is a progressive disease, and because this initial approach will usually fail to maintain patients, intensification of medication is often required. In this blog I focus on several concepts: medication adherence, intensification, a patient-centered approach, and the differences in the medications themselves.

Medication Adherence
The only medication that really works is the one that gets out of the bottle and into the patient! Prior to intensification it is critical that this patient get screened for medication non-adherence (also termed non-persistence or noncompliance). Diabetes is a chronic disease that early on may be relatively asymptomatic. These characteristics make it similar to osteoporosis and hypertension, conditions where non-adherence is unfortunately common. It is unclear from the case presentation if our patient is, in fact, actually taking all doses of her medication. What does she know of the importance of doing so? What are her goals of care? What does she believe the medication will do for her? Does she skip doses intentionally or does she forget to take them regularly? Screening questions may help find patients who are not adherent to their medications and can be incorporated into the intake process with the vital signs. If our patient really is adherent, then we can move into the next discussion: intensification.
Intensification in diabetes usually revolves around increasing dosages of current medication or adding additional medications in an effort to help patients achieve goals. There are many reasons why patients such as the one presented in our case do not get intensified. While the full scope of reasons is outside this blog, primary practices can increase the likelihood that intensification occurs if they redesign their offices with the Wagner Chronic Care Model[2] in mind. In this model:
  1. Patients are empowered to know their goals and partner with their physicians in developing management plans and strategies to achieve them.
  2. Practices are redesigned to allow team-based care, thus offloading the entire burden from the physician.  The physician emerges from the current chaos of the medical encounter to one where a highly orchestrated team exists (including physician, medical assistant/nurse, office staff, and, if available, nutritionist and CDE), each with a unique role.
  3. Evidence-based decision support becomes integrated into the workflow through algorithms or other agreed-to-order sets.  Such protocols ensure that this patient will be seen more frequently than the six-month interval in our case study. Infrequent contact is a barrier to goal attainment.
  4.  Patients not at goal are identified through the use of electronic registries or databases.   
I challenge you to begin the redesign of your offices. For this patient, she is still relatively early into her diabetes (eighteen months since diagnosis), and now is the time to aim for optimal control. Recent studies and discussions regarding metabolic memory indicate that tighter glycemic control early can have a long lasting positive impact on reducing diabetes-related endpoints.[3]
A Patient-Centered Approach
The redesign above and the approach to optimal care for our patient must include her voice in decision-making. By presenting the therapeutic options available for intensification (within the framework of agreed-to protocols at the practice level), the patient may be more likely to understand her medications and take them more faithfully. This approach, in a very small study, demonstrated increased knowledge and involvement in care but was underpowered to reach statistical significance in A1C level or medication adherence in comparison to usual care.[4] I believe that we must continue to build on that innovative study. We must help clinicians and patients by developing tools that will assist in their selection of the next medication. That selection should occur with a foundation of knowledge about the tradeoffs between clinical efficacy, safety, tolerability, durability, impact on weight, etc. 
The Medications Themselves
Having said all that, what about the choice of medication? Here I will let the experts responses stand on their own!  The reasons laid out in the three expert opinions for different clinical approaches to intensify treatment are all valid, and thus by definition none are really “wrong.” The only inappropriate position is to DO NOTHING! However, it is also fair to say that these choices cannot, and should not be viewed in a vacuum. These decisions can be made only with a focus on the biomedical/pathophysiologic approach. The decision-making lens also needs to capture the psychosocial dynamics and forces at play. Unfortunately physicians often do not consider this context.[5]
My own synthesis includes:
  •  Sulfonylurea—will not help her with weight loss, may lead to hypoglycemia and have poor durability. If this patient loses her insurance, however, I would certainly consider this agent.
  • TZD—not a bad option given that it can be given in combination with metformin and result in easier administration (and one copay) than two different pills. While it may help with hyperlipidemia, this patient is actually at goal with a statin (a class of medication with a proven track record of decreasing cardiac endpoints). I would discuss these added benefits as well as the possible weight gain with her.
  • GLP agonists—may help her with weight loss and attainment of goal A1C (affecting both fasting and post prandial blood sugars). Given how new they are on the market the durability of these (how well they continue to work over the years) is not as well studied as the other options. These medications are administered through injection devices so this needs to be discussed and reviewed with the patient. These agents are most likely to positively affect beta cell longevity and function, although continued study of these effects is warranted. 
We must stay tuned over the next several years to determine how the beta cell story plays out. While we wait for more data to determine if newer agents can really prolong beta cell viability in humans we cannot wait to grasp the bull by the horns and prioritize our approach to the care of the multitudes with diabetes as outlined above. Meanwhile, we must focus on medication adherence, intensification (through practice redesign), and a patient-centered approach.