Dosage adjustments in renal impairment among medical ward patients - ChatGPT® and DeepSeek® models’ effectiveness in assessing those adjustments

Q: Why are renal dose adjustments necessary for medical ward patients?

Renal impairment reduces drug clearance, so incorrect dosing can cause drug accumulation, toxicity, adverse drug reactions, and worsening renal function. Many recommendations rely on creatinine clearance (CrCl) and related measures.

Q: What was the objective of comparing ChatGPT® and DeepSeek® models?

The study aimed to evaluate renal dose adjustment practices in hospitalized patients and compare AI model performance (ChatGPT®, DeepSeek®) against UpToDate® guidelines.

Q: What were the key findings about dosing errors and AI performance?

A substantial share of drug orders required adjustment, and many were inappropriately dosed across patients. DeepThink® showed the highest sensitivity but low specificity, while improved ChatGPT® versions demonstrated modest gains in prediction performance.

Sep 30, 2026

Uploaded by River Wang

Renal impairment alters drug clearance, so medication dose adjustments are required to avoid toxicity or therapeutic failure, yet inappropriate dosing remains common. This prospective observational study (Jan–Apr 2024) evaluated hospitalized adults in a general medicine ward with creatinine clearance below 60 mL/min, comparing medication regimens to UpToDate® guidance and testing five AI models using 348 tailored prompts. Among 1461 orders, 23.8% needed adjustment, but 63.5% were inappropriately dosed in 76.4% of patients. Errors included unadjusted, inappropriately adjusted, and contraindicated regimens.

LanguageEnglish

CategoryHealthcare

FormatPDF

Pages9

Reading Time23 min

AccessFree to download

Contents lists available at ScienceDirect

Exploratory Research in Clinical and Social Pharmacy

A R T I C L E I N F O

Introduction

Exploratory Research in Clinical and Social Pharmacy 21 (2026) 100698 Contents lists available at ScienceDirect Exploratory Research in Clinical and Social Pharmacy journal [homepage:](homepage: www.elsevier.com/locate/rcsop)[ www.elsevier.com/locate/rcsop](homepage: www.elsevier.com/locate/rcsop) | Dosage adjustments in renal impairment among medical ward patients: ChatGPT® and DeepSeek® models’ effectiveness in assessing
those adjustments
Rahi Bikram Thapaa,c,*, Ravindra Khadkab, Prasanna Dahalc,d, Subash Karkic, Sabin Shresthae
a Hospital Development and Medical Services Division, Ministry of Health, Koshi Province, Biratnagar, Nepal b Department of Pharmacy, School of Medicine and Allied Health Sciences, Manmohan Technical University, Morang, Nepal c Department of Pharmacy, Purbanchal University School of Health Science, Gothgaun, Morang, Nepal
d Chettinad School of Pharmaceutical Sciences, Chettinad Hospital and Research Institute, Chettinad Academy of Research and Education, Kellambakkam, Tamilnadu, India
e Department of Pharmacy, Manmohan Memorial Institute of Health Science, Soalteemode, Kathmandu, Nepal | | | | --- | --- | --- | | A R T I C L E I N F O | | A B S T R A C T | | Keywords:
Artificial intelligence ChatGPT® DeepSeek®
Dose adjustment Renal impairment CKD
Nepal | | Background: Due to altered drug clearance, renal impairment necessitates drug dose adjustments to prevent toxicity or therapeutic failure, yet inappropriate dosing persists. The utility of AI tools (e.g., ChatGPT®, DeepSeek®) in supporting renal dose adjustments remains understudied.
Objective: Evaluate renal dose adjustment practices in hospitalized patients and compare AI models (ChatGPT®, DeepSeek®) against UpToDate® guidelines.
Method: A prospective observational study (January–April 2024) included hospitalized patients with creatinine clearance <60 mL/min in a general medicine ward. Medication regimens were assessed against UpToDate® guidelines. Five AI models (ChatGPT® 3.5, ChatGPT® 4.0, ChatGPT® 5.0, DeepSeek®, DeepThink®) were tested using 348 tailored prompts; sensitivity, specificity, and accuracy were calculated.
Results: Renal impairment prevalence was 30.9 %. Of 1461 drug orders, 23.8 %(348) required adjustment, with 63.5 %(221/348) inappropriately dosed in 76.4 %(113/148) of patients. Errors included 134 (38.5 %) unadjusted, 75 (21.6 %) inappropriately adjusted, and 12 (3.4 %) contraindicated regimens. Piperacillin/tazobactam (66.7 %), levofloxacin (83.33 %), and ranitidine (89.1 %) were most frequently inappropriately dosed. Severe renal impairment (CrCl ≤30 mL/min) increased improper dosing risk (AOR: 3.34; p = 0.004). DeepThink® showed the highest sensitivity (81.6 %) but low specificity (29 %) and 67.5 % accuracy. With advances in the ChatGPT® model, there was a modest improvement in prediction capacity, with the latest ChatGPT® 5.0 achieving balanced performance (64.3 % sensitivity, 54.8 % specificity, 61.7 % accuracy).
Conclusion: Inappropriate renal dosing is prevalent among hospitalized patients with renal impairment. While AI models show promise as clinical decision support tools, their accuracy limitations require further optimization via evidence-based database training, prompt refinement, and integration of clinical expertise for reliable implementation. | 1. Introduction Medications play a crucial role in managing various diseases, with many drugs and their metabolites primarily eliminated through the kidneys.1,2 Optimizing drug therapy in patients with renal impairment is essential, as impaired renal function significantly affects drug elimination.3 Failure to adjust dosages appropriately can lead to drug accumulation, toxicity, adverse drug reactions (ADRs), and even worsening renal function.1,2 Renal impairment, characterized by a reduced estimated glomerular filtration rate (eGFR <60 mL/min/1.73m2), includes both acute kidney injury (AKI) and chronic kidney disease

FAQs

Why are renal dose adjustments necessary for medical ward patients?

Renal impairment reduces drug clearance, so incorrect dosing can cause drug accumulation, toxicity, adverse drug reactions, and worsening renal function. Many recommendations rely on creatinine clearance (CrCl) and related measures.

What was the objective of comparing ChatGPT® and DeepSeek® models?

The study aimed to evaluate renal dose adjustment practices in hospitalized patients and compare AI model performance (ChatGPT®, DeepSeek®) against UpToDate® guidelines.

What were the key findings about dosing errors and AI performance?

A substantial share of drug orders required adjustment, and many were inappropriately dosed across patients. DeepThink® showed the highest sensitivity but low specificity, while improved ChatGPT® versions demonstrated modest gains in prediction performance.

Sep 30, 2026

Uploaded by River Wang