CoChat

The Capability of ChatGPT in Predicting and Explaining Common Drug-Drug Interactions

Open access
Juhi, Ayesha, Pipil, Neha, Santra, Soumya, Mondal, Shaikat, Behera, Joshil Kumar, Mondal, Himel · 2023 · Cureus
Cited by 128

The DOI matches its Crossref record; 7 of the 7 references it cites were found in a registry or an index.

Abstract

Background Drug-drug interactions (DDIs) can have serious consequences for patient health and well-being. Patients who are taking multiple medications may be at an increased risk of experiencing adverse events or drug toxicity if they are not aware of potential interactions between their medications. Many times, patients self-prescribe medications without knowing DDI. Objective The objective is to investigate the effectiveness of ChatGPT, a large language model, in predicting and explaining common DDIs. Methods A total of 40 DDIs lists were prepared from previously published literature. This list was used to converse with ChatGPT with a two-stage question. The first question was asked as "can I take X and Y together?" with two drug names. After storing the output, the next question was asked. The second question was asked as "why should I not take X and Y together?" The output was stored for further analysis. The responses were checked by two pharmacologists and the consensus output was categorized as "correct" and "incorrect." The "correct" ones were further classified as "conclusive" and "inconclusive." The text was checked for reading ease scores and grades of education required to understand the text. Data were tested by descriptive and inferential statistics. Results Among the 40 DDI pairs, one answer was incorrect in the first question. Among correct answers, 19 were conclusive and 20 were inconclusive. For the second question, one answer was wrong. Among correct answers, 17 were conclusive and 22 were inconclusive. The mean Flesch reading ease score was 27.64±10.85 in answers to the first question and 29.35±10.16 in answers to the second question, p = 0.47. The mean Flesh-Kincaid grade level was 15.06±2.79 in answers to the first question and 14.85±1.97 in answers to the second question, p = 0.69. When we compared the reading levels with hypothetical 6th grade, the grades were significantly higher than expected (t = 20.57, p < 0.0001 for first answers and t = 28.43, p < 0.0001 for second answers). Conclusion ChatGPT is a partially effective tool for predicting and explaining DDIs. Patients, who may not have immediate access to the healthcare facility for getting information about DDIs, may take help from ChatGPT. However, on several occasions, it may provide incomplete guidance. Further improvement is required for potential usage by patients for getting ideas about DDI.

Overview

The claims and statistics of this paper haven’t been checked yet.

Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.

A few free checks a day · papers already verified open instantly · results are public

Integrity signals

Registry

Crossref · 10.7759/cureus.36272

Cites

7 of 7 references verified0 citation errors · 0 not checkable

Statistics

Verification

Claims the paper makes

The claims and statistics of this paper haven’t been checked yet.

Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.

A few free checks a day · papers already verified open instantly · results are public

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

The self-check needs the paper’s full text.

Citations

Cites

Cited by

See all 128

Metadata

Authors
Juhi, A., Pipil, N., Santra, S., Mondal, S., Behera, J. K., & Mondal, H.
Venue
Cureus
Published
2023-03-17
Volume
15
Issue
3
Pages
e36272-e36272
ISSN
2168-8184 (ISSN-L)
OpenAlex
W4327715333
Cited by
128
References
7
FWCI
4.43
Type
journal article
Language
en
Open access
diamond

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review

Data sources

Bibliographic data: Semantic Scholar, OpenAlex, Crossref.

Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).

Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.