Open-Source Geriatric Beers Criteria–Part D Crosswalk, for Analyses of Drugs Adverse to Elders (DAE)

[Under development]

A crucial resource for describing the quality of prescribing practices for elder patients (those age 65+) is essentially unusable for research and data analysis. This blog post documents the challenges and decision-making for creating a crosswalk for these drugs that can be used in Part D data.

The Beers Criteria published in 2019 by the American Geriatric Society are an important resource for all who interact with geriatric patients’ medications. For clinicians trusted to manage or add to the prescriptions issued to these patients, the Beers Criteria provide a list of drugs adverse to the elderly, succinct recommendation (e.g., “avoid” or “avoid as first-line therapy…”, a rationale for each recommendation, and the strength and evidence-quality of each.

Major challenges arise from the structure of Part D data, which contain the generic name (GNN) and National Drug Code (NDC) to identify the drug sensitively (GNN) and specifically (NDC).


The generic name seems promising: it’s an abstraction of the drug itself — agnostic of its dose, its form, and the brand. For example, Xanax, Niravam, and generic alprazolam would all have the GNN of “alprazolam.” If we had all generic names for drugs falling under the Beers Criteria recommendation, we could simplify our search to just a couple hundred terms, matched against the entire GNN field in Part D Event data. This approach may prove most efficacious and efficient.

My fear for the GNN is that we under-represent “all generic names falling under the recommendation.” When working in an FDA metathesaurus software, I find complex relationships among drugs that indicate complex relationships persist in the world of GNNs: drugs may be used in combination

The goal of this project is to create an algorithm for

My initial list of 15 benzodiazepine names (suggested by peers) is exceeded by the 2015 Beers Criteria, which includes “amitriptyline” and “clidinium.” On the other hand, my list exceeds the 2019 Beers Criteria, which omits Halazepam, Midazolam, and Prazepam. Relying theon the
At least in the 2019 Beers Criteria, “Chlordiazepoxide (alone or in combination with amitriptyline or clidinium)”

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s