Skip to main content
Clinical SupportMixed Methods2025

AI in Medication Dosing Optimization and Therapeutic Individualization

Key Finding

Early trials and qualitative studies of AI-guided dosing tools (for example, CURATE.AI and related platforms) show promise in achieving therapeutic targets more quickly and at lower average doses, with one oncology example reporting a 20% reduction in capecitabine dosing compared with projected standard-of-care doses while maintaining response. Adoption barriers include trust, workflow integration, and regulatory oversight.

8 min read2 sources cited
oncologyinternal-medicineall

Executive Summary

Traditional dosing strategies often rely on population averages and fixed protocols that may not optimally balance efficacy and toxicity for individual patients. AI-enabled dosing platforms, such as CURATE.AI, use patient-specific response data to dynamically adjust dosing, exploring a personalized dose–response space with relatively few data points. Early clinical applications in oncology and transplant medicine suggest that AI-guided dosing can lead to faster attainment of therapeutic targets and reduced cumulative drug exposure compared with standard approaches.

A 2024 oncology report describes AI-discovered correlations between patient-specific capecitabine doses and tumor-marker trajectories, enabling an approximate 20% reduction in average prescribed dose versus projected standard-of-care while maintaining tumor control, with clinicians accepting AI recommendations in 26 of 27 dosing decisions. A 2025 qualitative study of clinicians using AI dosing tools highlights perceived benefits (time savings, decision support in high-volume settings) but also concerns about overreliance, explainability, and workflow fit.

Detailed Research

Methodology

Evidence includes pilot clinical trials, protocol papers, and qualitative evaluations of AI-enabled dosing systems. CURATE.AI, for example, maps individualized dose–response relationships based on serial biomarker and dose data, then recommends dose adjustments within prespecified safety bounds.

Qualitative work explores clinicians' expectations, trust, and integration experiences with AI dosing tools in practice.

Key Studies

PRECISE CURATE.AI Pilot Trial Protocol (2021)

  • Design: Pilot clinical trial protocol
  • Sample: Metastatic colorectal cancer patients
  • Findings: Small-data AI platform uses tumor marker levels and capecitabine doses to build personalized dose–response profiles and guide dynamic dosing. Goal is optimal efficacy with minimal toxicity through continual dose recalibration.
  • Clinical Relevance: Demonstrates feasibility of AI-guided dosing

Oncology Dosing Optimization Commentary (2024)

  • Design: Clinical commentary
  • Sample: Early trial results
  • Findings: AI-assisted capecitabine dosing enabled 20% reduction in average prescribed dose compared with standard-of-care with high clinician acceptance (26 of 27 recommendations followed).
  • Clinical Relevance: Promising real-world acceptance

Qualitative Study on AI-Guided Dosing (2025)

  • Design: Qualitative interviews
  • Sample: Clinicians using AI dosing tools
  • Findings: Providers see potential to save time and reduce uncertainty in high-volume settings, but raise concerns about performance in edge cases, transparency, and liability.
  • Clinical Relevance: Identifies adoption barriers

Clinical Implications

For osteopathic physicians managing complex medication regimens—such as oncology, transplant, or anticoagulation—AI-guided dosing tools may eventually offer more precise balancing of efficacy and toxicity, potentially reducing adverse events and improving quality of life.

In primary care, similar approaches could support personalized titration of antihypertensives, insulin, or disease-modifying agents, although robust evidence in these domains is still emerging.

Limitations & Research Gaps

Most AI dosing evidence remains in early trials or specific high-risk domains; broad, multi-condition clinical outcome data are lacking. Regulatory frameworks and integration standards for AI dosing tools are still being developed.

There are no osteopathy-specific studies examining how AI dosing intersects with OMT-driven changes in pain, function, or medication requirements.

Osteopathic Perspective

Osteopathic care often involves efforts to reduce medication burden through OMT and lifestyle interventions; AI-enabled dosing that safely minimizes drug exposure is conceptually aligned with this goal.

DOs should approach AI dosing tools as adjuncts that can inform, but not dictate, decisions—integrating algorithmic suggestions with careful assessment of structure–function changes, patient preferences, and the body's self-regulatory responses.

References (2)

  1. Ho D, Tan YJ, et al. “A Protocol for the PRECISE CURATE.AI Pilot Clinical Trial.” Frontiers in Digital Health, 2021;3:635524. DOI: 10.3389/fdgth.2021.635524
  2. Roy S “Optimizing Oncology Drug Dosing: Is Artificial Intelligence the Future?.” ASCO Daily News, 2024;44:e240220. DOI: 10.1200/EDBK-25-473590

Related Research

Accuracy of AI Systems in Generating Differential Diagnoses

Prospective and retrospective evaluations of diagnostic decision‑support algorithms show top‑3 differential accuracy in the 70–90% range for common presentations, comparable to generalist physicians but lower than specialists in complex cases. Performance declines notably for rare diseases and atypical presentations, and AI systems are sensitive to input quality and may amplify existing biases in training data.

Impact of AI on Diagnostic Errors in Clinical Practice

Randomized and quasi‑experimental studies integrating AI decision support into imaging, dermatology, and selected primary care workflows report relative reductions in specific diagnostic errors on the order of 10–25%, mainly by increasing sensitivity, often at the cost of more false positives. Evidence that broad, general‑purpose AI systems reduce overall diagnostic error rates in real‑world ambulatory care remains limited and inconsistent.

AI‑Enhanced Drug Interaction Checking and Medication Safety

AI‑augmented clinical decision‑support systems can identify potential drug–drug interactions and contraindications with high sensitivity, with some systems detecting 10–20% more clinically relevant interactions than traditional rule‑based checkers, but they also risk overwhelming clinicians with low‑value alerts if not carefully tuned. Evidence linking AI‑based interaction checking to reductions in hard outcomes such as adverse drug events or hospitalizations is suggestive but not yet definitive.