Automated Stratification of Medication Prescriptions by Risk and Complexity as a Safety Strategy

演講摘要

More and better stratification is needed to understand medication use more precisely.

Our ultimate goal may be to personalize medication therapy with great precision, thereby consistently and effectively individualizing medication selection, dosing, and duration of use. But devising methods to achieve this ultimate goal raises profound questions. Through scientific discovery, what we reveal about human use of and responses to medications comes mainly from collectively analyzing individuals at the group level. In practice, we then typically apply group-level conclusions to individual cases. When we do this, the relevant question to ask is to what degree is any individual in our care like the members of the group studied to inform their care?

We cannot stop applying medication-related conclusions about groups to individual cases. One thing we can do is further stratify study cohorts, thereby making group conclusions incrementally more precise. For example, rather than study a medication's effects on "healthy adults," we could study its effects on "adults living at home who have complex medication regimens." Similarly, rather than study prescribing errors overall, we could study prescribing errors made when prescribing "high-risk medications." Of course, stratifying in cases like this requires reliable and valid methods of scoring regimen complexity and medication risk.

In this talk, I will present published work from others and myself towards automatically scoring medication regimen complexity and towards the development of a new type of medication risk index. Next, I will discuss how regimen complexity and medication risk measures and scores might be used to improve pharmacy practice and move us towards greater precision in medication use and better patient support with home medication use.