NEW YORK – Management of type 2 diabetes patients has shifted "from one-size-fits-all to recommendations that are more personalized," according to Dr. Patrick J. O’Connor, a family physician and senior clinical investigator at HealthPartners Research Foundation in Minneapolis.
From a risk-management perspective, that means assessing the role of each cardiovascular risk factor in a given patient and identifying which modifiable factor packs the biggest punch and is likely to yield the most benefit when changed, he said.
Dr. Patrick J. O'Connor
Cardiovascular events account for about two-thirds of all deaths in patients with type 2 diabetes, and most of a patient’s cardiovascular risk depends on age and sex, two risk factors that resist modification. The question then becomes which modifiable risk factor smoking, hypertension, hyperlipidemia, hemoglobin A1c level, and aspirin use – warrants initial attention.
The best way to find out is to run a risk-factor engine analysis on each patient. For the analysis, each of the patient’s actual values is run individually through a risk-assessment formula, each time substituting normal values for all the modifiable factors except one. This process helps to identify the modifiable factor that, if normalized, stands to most benefit the patient.
The best known risk-factor formulas are the Framingham Risk Score, and for patients with diabetes the score derived from data in the UK Prospective Diabetes Study (UKPDS). "No risk engines available today are perfect; the epidemiologic data [that the risk scoring formula derives from] may be out of date," he said. But among the risk formulas currently available, "the UKPDS is the most appropriate for patients with diabetes," and includes an entry for hemoglobin A1c.
Dr. O’Connor and his associates are in the process of producing a new risk-engine website designed for patients with diabetes.
Electronic medical records have made risk-engine assessments easy, because a physician can set up the electronic record to run the engine automatically using a patient’s stored risk-factor data.
For example, the UKPDS risk-scoring formula calculates that a 70-year-old man who doesn’t smoke and has had diabetes for 15 years with a hemoglobin A1c of 7.2%, a systolic blood pressure of 170 mm Hg, and moderate levels of total cholesterol and high density lipoprotein cholesterol has a 15% risk for a coronary heart disease event and a 21% risk for a stroke over the next 10 years. If his systolic pressure dropped to 140 mm Hg, his coronary event risk would fall to 13%, and his stroke risk would drop to 18% – modest declines.
An analysis like this might suggest that it is not worth the effort of undertaking the treatment required to cut the patient’s systolic pressure by 30 mm Hg to achieve such modest risk improvements.
On the other hand, if the same 70-year-old man also smoked at baseline, his 10-year coronary event risk would be 19% and his stroke risk 30%. It might be worth working with the patient to get him to stop smoking and to treat his systolic hypertension, as both actions would cut his 10-year stroke risk by an absolute value of 10%.
"Use the risk engine to get a sense of what the treatment should be. If the patient has little or no reversible risk, then stop. Not all interventions are of equal benefit to all patients at a given time," Dr. O’Connor noted.
Dr. O’Connor said he had no relevant financial disclosures.