Inferensys

Glossary

Beers Criteria

An explicit list of potentially inappropriate medications for older adults, maintained by the American Geriatrics Society, used as a rules-based filter in automated medication safety reviews.
Knowledge engineer constructing knowledge base on laptop, document hierarchy visible, casual office setup.
POTENTIALLY INAPPROPRIATE MEDICATION LIST

What is Beers Criteria?

The Beers Criteria is an explicit list of medications considered potentially inappropriate for older adults, maintained by the American Geriatrics Society, serving as a critical rules-based filter in automated medication safety reviews.

The Beers Criteria is a standardized compendium of medications for which the risks of adverse drug events generally outweigh the clinical benefits in patients aged 65 and older. Published and periodically updated by the American Geriatrics Society (AGS) since 1991, it categorizes drugs into three explicit domains: medications to avoid in most older adults, medications to avoid in patients with specific diseases or syndromes, and medications to use with caution. Each entry is supported by evidence-based rationales and graded by strength of recommendation.

In automated medication reconciliation workflows, the Beers Criteria functions as a deterministic rules engine that cross-references a patient's active medication list, diagnoses, and renal function against the published tables. When a match is detected—such as a first-generation antihistamine prescribed to a patient with dementia—the system generates a clinical decision support alert flagging the potentially inappropriate medication. This algorithmic integration reduces reliance on manual pharmacist review for high-volume screening while supporting polypharmacy risk reduction in geriatric populations.

GERIATRIC PHARMACOTHERAPY

Core Characteristics of the Beers Criteria

The Beers Criteria is an explicit list of potentially inappropriate medications (PIMs) for older adults, maintained by the American Geriatrics Society. It serves as a foundational rules-based filter in automated medication safety reviews, categorizing drugs to avoid, drugs to use with caution, and drug-drug interactions of concern in patients aged 65 and older.

01

Potentially Inappropriate Medication Classes

The criteria explicitly identify drug classes with an unfavorable risk-benefit ratio in geriatric populations. Key categories include:

  • First-generation antihistamines (e.g., diphenhydramine) due to anticholinergic effects causing confusion and urinary retention
  • Tricyclic antidepressants (e.g., amitriptyline) linked to orthostatic hypotension and cardiac conduction abnormalities
  • Long-acting benzodiazepines (e.g., diazepam) associated with prolonged sedation and increased fall risk
  • Non-COX-selective NSAIDs (e.g., indomethacin) in patients with Stage 4 or higher chronic kidney disease
02

Drug-Disease Interaction Avoidance

The Beers Criteria cross-references medications against specific geriatric syndromes to flag drug-disease interactions where a drug may exacerbate an existing condition:

  • Antipsychotics in patients with dementia due to increased cerebrovascular accident and mortality risk
  • Peripheral alpha-1 blockers (e.g., doxazosin) in women with stress incontinence due to exacerbation of symptoms
  • Opioids in patients with a history of falls or fractures, increasing the likelihood of recurrent injury
  • Acetylcholinesterase inhibitors in patients with syncope due to bradycardia risk
03

Drug-Drug Interaction Identification

The criteria catalog high-severity drug-drug interactions prevalent in polypharmacy regimens. Automated systems use these rules to flag combinations such as:

  • Three or more CNS-active drugs concurrently, dramatically increasing fall and cognitive impairment risk
  • Warfarin combined with NSAIDs, elevating the risk of life-threatening gastrointestinal hemorrhage
  • Lithium with ACE inhibitors, which can precipitate lithium toxicity through reduced renal clearance
  • Opioids with gabapentinoids, potentiating respiratory depression beyond either agent alone
04

Renal Function-Adjusted Avoidance

A dedicated subset of the criteria lists medications requiring dose reduction or complete avoidance based on creatinine clearance thresholds:

  • Dabigatran at CrCl < 30 mL/min due to accumulation and bleeding risk
  • Spironolactone at CrCl < 30 mL/min linked to dangerous hyperkalemia
  • Enoxaparin at CrCl < 30 mL/min requiring anti-Xa monitoring or alternative anticoagulation
  • Nitrofurantoin at CrCl < 30 mL/min due to inadequate urine concentration for efficacy and peripheral neuropathy risk
05

Anticholinergic Burden Stratification

The Beers Criteria explicitly ranks medications by their anticholinergic potency to help clinicians minimize cumulative burden. High-potency agents to avoid include:

  • Scopolamine and atropine for their potent central nervous system effects
  • Oxybutynin immediate-release for overactive bladder, with preferential use of non-anticholinergic alternatives like mirabegron
  • Paroxetine among SSRIs due to its uniquely strong muscarinic receptor affinity
  • The criteria emphasize that combining multiple agents with even moderate anticholinergic activity can produce a cumulative toxic effect manifesting as delirium, constipation, and functional decline
06

Scheduled Updates and Evidence Grading

The American Geriatrics Society maintains the criteria through a rigorous, transparent update cycle approximately every three years:

  • Systematic literature review conducted by an expert panel using the GRADE methodology
  • Each recommendation carries an explicit Quality of Evidence rating (High, Moderate, Low) and a Strength of Recommendation designation (Strong, Conditional)
  • The 2023 update introduced new PIMs including glimepiride and sliding-scale insulin due to hypoglycemia risk
  • This structured evidence framework allows automated clinical decision support systems to weight alerts by clinical significance, reducing alert fatigue from low-evidence flags
BEERS CRITERIA CLARIFIED

Frequently Asked Questions

Precise answers to common clinical and technical questions about the AGS Beers Criteria and its role in automated medication safety for older adults.

The Beers Criteria is an explicit list of potentially inappropriate medications (PIMs) for older adults, maintained and periodically updated by the American Geriatrics Society (AGS) . It is used as a rules-based clinical decision support filter to flag medications where the risk of adverse drug events typically outweighs the clinical benefit in patients aged 65 and older. The criteria categorize medications into three main buckets: medications to avoid in most older adults, medications to avoid in older adults with specific diseases or syndromes (drug-disease interactions), and medications to be used with caution. In automated medication reconciliation systems, the Beers Criteria serves as a deterministic safety net, cross-referencing active medication lists against these tables to generate actionable alerts for deprescribing or dose adjustment.

EXPLICIT TOOLS FOR GERIATRIC PRESCRIBING

Beers Criteria vs. STOPP/START Criteria

Comparison of two leading explicit criteria frameworks used to identify potentially inappropriate medications in older adults and optimize prescribing safety.

FeatureBeers CriteriaSTOPP CriteriaSTART Criteria

Primary Purpose

Identify potentially inappropriate medications to avoid

Detect potentially inappropriate prescriptions linked to adverse drug events

Detect potential prescribing omissions of indicated therapies

Maintaining Body

American Geriatrics Society

European expert consensus panel

European expert consensus panel

Year of Latest Update

2023

2015 (Version 2)

2015 (Version 2)

Number of Criteria

~100 explicit criteria

80 criteria

34 criteria

Organized By

Therapeutic category and disease state

Physiological system

Physiological system

Renal Dose Adjustment Guidance

Drug-Disease Interaction Flags

Prescribing Omission Detection

Prasad Kumkar

About the author

Prasad Kumkar

CEO & MD, Inference Systems

Prasad Kumkar is the CEO & MD of Inference Systems and writes about AI systems architecture, LLM infrastructure, model serving, evaluation, and production deployment. Over 5+ years, he has worked across computer vision models, L5 autonomous vehicle systems, and LLM research, with a focus on taking complex AI ideas into real-world engineering systems.

His work and writing cover AI systems, large language models, AI agents, multimodal systems, autonomous systems, inference optimization, RAG, evaluation, and production AI engineering.