Using Massachusetts Physician Payment Data Prescribing Patterns Under Attack (Again)

 

Prescribing practices of physicians are once again under attack. And, once again, the attacks leave consumers, patients, and their families feeling confused. This time, the confusion is a result of a study done by researchers from Harvard Medical School and published in JAMA Internal Medicine which found that “medical industry payments to physicians in Massachusetts are associated with higher rates of prescribing brand-name drugs that treat high cholesterol.”

Right off the bat, there are some interesting things to note relating to the study. First of all, the study only looked at physicians in Massachusetts, a relatively small state. Second, the study only focused on the prescribing of brand-name drugs for high cholesterol, a relatively small portion of total branded drugs prescribed since the largest players have since gone generic. These are not suggested to discredit the research, but instead, to help put it in perspective.

However, even acknowledging those realities, it is always beneficial to read through the entire study to ascertain the methods, the reasoning, and, if possible, the true results of the study, to best understand the study and the points behind it.

Design, Setting, and Participants

The study used cross-sectional linkage of Part D Medicare prescriptions claims data with the Massachusetts physician payment database. The database includes all licensed Massachusetts physicians who wrote prescriptions for statins that were paid for under the Medicare drug benefit in 2011.

Study Results

Among the 2444 Massachusetts physicians in the Medicare prescribing database in 2011, 899 (36.8%) received some form of industry payments. The most frequent payment was for company sponsored meals, accounting for 71.1% of payments.

Statins accounted for 1,559,003 prescription claims, 22.8% of which were for brand name drugs. According to the study, physicians with no industry payments listed had a median brand name prescribing rate of 17.8%. For every $1,000 received in payments, the brand name statin prescribing rate increased by a statistically insignificant .1%.

However, interestingly, payments for educational training were associated with a 4.8% increase in the rate of brand name prescribing. While that percentage is more significant than the overall percentage mentioned above, it is still not that significant. As we have mentioned before, and continue to opine on, we do not believe that physicians becoming more educated about certain drugs is typically a bad thing. The more they learn, they more they know and can understand how individual drugs can help their patients and what kind of side effects they need to be looking for.

Using the foregoing figures, the researchers concluded that industry payments to physicians are associated with higher rates of prescribing brand-name statins.

Analysis

As we have all know, and the Harvard researchers even stated, correlation does not equal causation. Simply because there is a correlation between receiving payments (such as a company sponsored meal) and prescribing rates does not mean that one is caused by the other. Even if it is a true causation, physicians who attend company sponsored meals learn more about the drug, and therefore feel more comfortable prescribing it, than a physician who does not have the benefit of that knowledge. Further, the correlation they are trying to bring to bear is a one in one thousand percent, a statistically insignificant amount.

Most patients likely prefer to have a physician who knows exactly what they are prescribing, and a physician who knows which individual patients may be someone who benefits from the name-brand drug, instead of the generic. In the case of statins not all patients can tolerate all statins some patients who have problems taking one statin often can use another statin and several of those statins that utilize different pathways are still branded products.

The study’s limitations included the possible inaccuracy of the reporting of payments and prescriptions covered outside of Medicare. Further, they could not determine “which physicians received payments from a specific company and analyze their prescribing of that company’s products.”