Can ChatGPT Provide Quality Information on Integrative Oncology? A Brief Report

Abstract

This short report evaluated the accuracy and quality of information provided by ChatGPT regarding the use of complementary and integrative medicine for cancer. Using the QUality Evaluation Scoring Tool, a panel of 12 reviewers assessed ChatGPT's responses to 8 questions. The study found that ChatGPT provided moderate-quality responses that were relatively unbiased and not misleading. However, the chatbot's inability to reference specific scientific studies was a significant limitation. Patients with cancer should not rely on ChatGPT for clinical advice until further systematic validation. Future studies should examine how patients perceive ChatGPT's information and its impact on communication with health care professionals.

Introduction

Patients with cancer often turn to complementary and integrative medicine (CIM) to manage their symptoms or treatment-related toxicities.¹ The internet remains one of the most common sources of information regarding CIM.² However, misinformation and falsified claims on the internet, such as inaccurate information on the use of CIM for cancer, remain a concerning problem.³

The development of ChatGPT, a language model that was released to the public by OpenAI in 2022, has generated considerable interest in various fields such as health care.⁴ Using artificial intelligence technology, this chatbot can interact in a conversational manner and is capable of performing various tasks, such as answering questions, and coding, through training on data from the internet.⁵ Studies have examined the use of ChatGPT to provide information on cancer, such as cancer symptoms and general misconceptions, but not on the use of complementary or integrative medicine.^6,7

In a recent editorial, Cramer posted an intriguing question “But the changes in society toward an information society are not leaving medicine unscathed. But do they also affect CIM, where direct human contact is still so much more important? I think so.”⁸ It is becoming inevitable that the rise of artificial intelligence will affect CIM in terms of research and delivery of care. In response to this editorial, this brief report aimed to evaluate the accuracy and quality of information provided by ChatGPT on CIM use for cancer.

Methods

On March 30, 2023, the authors presented ChatGPT (GPT-3, February 13 version) with eight questions covering various aspects of CIM, including the indications of CIM modalities (Q1: Can I use acupuncture for general cancer pain or pain caused by taking anastrozole? Q2: Is omega-3 supplement helpful for peripheral neuropathy caused by chemo?), drug interactions (Q3: Can I take St. John's wort supplements when taking irinotecan? Q4: Can I take green tea supplements when taking gefitinib?), potential adverse effects (Q5: Will ginseng affect my sleep? Q6: Will ginseng damage my liver if I am a cancer patient?), and special precautions/controversies (Q7: I have estrogen-receptor positive breast cancer. Can I take soy products? Q8: Can I receive acupuncture if I have lymphedema?) (Table 1). These types of questions are common concepts and frequently asked questions on popular CIM websites, such as the American Cancer Society and the Memorial Sloan Kettering Cancer Center websites.

Table 1.

ChatGPT Responses to the Questions on Integrative Oncology

Topics	Questions	ChatGPT answers	Average total score (total: 26 points), Median [IQR], (Fleiss K)	Median score for each QUEST criterion
Indication/use	Q1: Can I use acupuncture for general cancer pain or pain caused by taking anastrozole?	Acupuncture can be a useful complementary therapy for managing cancer-related pain, including general cancer pain and pain caused by taking anastrozole, which is a medication used to treat breast cancer. Several studies have suggested that acupuncture may be effective in reducing cancer-related pain, improving quality of life, and reducing the need for pain medication. Acupuncture involves the insertion of thin needles into specific points on the body to stimulate nerve endings and promote the body's natural healing response. In addition to reducing pain, acupuncture may also have other benefits for people with cancer, including reducing chemotherapy-induced nausea and vomiting, improving sleep quality, and reducing anxiety and depression. It is important to note that acupuncture should not be used as a replacement for conventional cancer treatments, such as chemotherapy or radiation therapy. Instead, it should be used as a complementary therapy to help manage cancer-related symptoms and improve overall well-being. If you are interested in trying acupuncture to manage cancer-related pain, it is important to talk to your health care provider first to ensure that it is safe for you to try and to find a licensed and experienced acupuncturist who is trained to work with people with cancer.	Score 14 [13–16] points (κ = 0.523, p < 0.001)	Attribution: 1 Type of study: NA Conflict of interest: 2 Currency: 0 Complementarity: 1 Tone: 1.5
Indication/use	Q2: Is omega-3 supplement helpful for peripheral neuropathy caused by chemo?	There is some evidence to suggest that omega-3 supplements may be helpful in managing peripheral neuropathy caused by chemotherapy. Omega-3 fatty acids are important nutrients that are found in fatty fish, such as salmon, and can also be taken as supplements. Some studies have suggested that omega-3 supplements may help reduce inflammation and improve nerve function, which could help alleviate peripheral neuropathy symptoms. However, the evidence is not yet conclusive, and more research is needed to fully understand the potential benefits of omega-3 supplements for peripheral neuropathy. If you are considering taking omega-3 supplements to help manage your peripheral neuropathy symptoms, it is important to talk to your health care provider first to ensure that it is safe for you to try and to determine the appropriate dosage. Omega-3 supplements can interact with certain medications and may not be safe for everyone to take.	Score 15 [13–16] points (κ = 0.616, p < 0.001)	Attribution: 1 Type of study: NA Conflict of interest: 2 Currency: 0 Complementarity: 1 Tone: 2
Interaction	Q3: Can I take St. John's wort supplements when taking irinotecan?	No, it is not recommended to take St. John's wort supplements while taking irinotecan or any chemotherapy medication without consulting with your health care provider. St. John's wort is a herbal supplement that is often used to treat depression and anxiety, but it can interact with many medications, including irinotecan. St. John's wort can increase the activity of enzymes in the liver that break down drugs, including irinotecan. This can cause irinotecan to be metabolized and eliminated from the body more quickly, potentially reducing its effectiveness. In addition, taking St. John's wort with irinotecan may increase the risk of side effects, such as diarrhea, nausea, and vomiting. Therefore, it is important to talk to your health care provider before taking any herbal supplements or other medications while receiving chemotherapy. Your health care provider can evaluate your individual situation and provide guidance on whether it is safe and appropriate to use St. John's wort or any other supplements. It is always better to be cautious and check with your health care provider before taking any new supplements or medications to avoid potential interactions or complications.	Score 13 [10–13.5] points (κ = 0.511, p < 0.001)	Attribution: 0 Type of study: NA Conflict of interest: 2 Currency: 0 Complementarity: 1 Tone: 2
Interaction	Q4: Can I take green tea supplements when taking gefitinib?	It is generally not recommended to take green tea supplements while taking gefitinib without consulting with your health care provider first. Gefitinib is a medication used to treat certain types of cancer, and it works by blocking certain enzymes involved in the growth and spread of cancer cells. Green tea supplements, however, contain high concentrations of compounds called catechins that have antioxidant properties and have been studied for their potential anticancer effects. Some studies suggest that the catechins in green tea may interfere with the effectiveness of gefitinib. This is because catechins can activate some of the same enzymes that gefitinib blocks, potentially reducing the drug's effectiveness. Therefore, it is important to talk to your health care provider before taking any supplements or making significant changes to your diet while taking gefitinib or any other cancer medication. Your health care provider can evaluate your individual situation and provide guidance on whether it is safe and appropriate to use green tea supplements. They may also suggest ways to optimize your diet and nutrition to support your cancer treatment.	Score 15.5 [12.5–16] points (κ = 0.557, p < 0.001)	Attribution: 1 Type of study: NA Conflict of interest: 2 Currency: 0 Complementarity: 1 Tone: 2
Adverse events	Q5: Will ginseng affect my sleep?	Ginseng is a herbal supplement that is commonly used to improve energy levels, reduce stress, and enhance cognitive function. While ginseng is generally considered safe for most people when taken in appropriate doses, it can cause some side effects, including changes in sleep patterns. Ginseng is known to have stimulating effects, which can increase alertness and energy levels. As a result, taking ginseng late in the day or in the evening may make it difficult to fall asleep or stay asleep at night. If you are sensitive to caffeine or other stimulants, you may be more likely to experience these side effects. If you are considering taking ginseng to help manage symptoms related to cancer treatment or any other health condition, it is important to talk to your health care provider first to ensure that it is safe for you to try and to determine the appropriate dosage and timing of the supplement. They may recommend taking ginseng earlier in the day or in smaller doses to help minimize the risk of sleep disturbances.	Score 13 [10–13] points (κ = 0.583, p < 0.001)	Attribution: 0 Type of study: NA Conflict of interest: 2 Currency: 0 Complementarity: 1 Tone: 2
Adverse events	Q6: Will ginseng damage my liver if I am a cancer patient?	There is limited research on the potential effects of ginseng on liver function in cancer patients. Some studies have suggested that high doses of ginseng may cause liver damage in certain individuals, particularly if they have pre-existing liver conditions or if they are taking other medications that can affect liver function. If you are a cancer patient considering taking ginseng, it is important to talk to your health care provider first to ensure that it is safe for you to try and to determine the appropriate dosage and timing of the supplement. They may recommend regular liver function tests to monitor any potential effects on liver health. In addition, it is important to note that ginseng may interact with certain cancer treatments, including chemotherapy and radiation therapy, so it is important to inform your health care provider of any supplements or herbs that you are taking.	Score 16 [13.75–16] points (κ = 0.621, p < 0.001)	Attribution: 1 Type of study: NA Conflict of interest: 2 Currency: 0 Complementarity: 1 Tone: 2
Special precautions/controversies	Q7: I have estrogen-receptor positive breast cancer. Can I take soy products?	Soy products contain compounds called phytoestrogens, which have a chemical structure similar to estrogen. As a result, there has been some concern that consuming soy products may interfere with the effectiveness of hormone therapy for estrogen-receptor positive breast cancer. However, research on this topic has been mixed, and the current evidence suggests that moderate consumption of soy products is safe for women with estrogen-receptor positive breast cancer. In fact, some studies have suggested that soy may have a protective effect against breast cancer recurrence. It is always a good idea to talk to your doctor or a registered dietitian before making any significant changes to your diet, especially if you are currently undergoing treatment for breast cancer. They can help you weigh the potential benefits and risks of consuming soy products and make personalized recommendations based on your individual situation.	Score 16 [13.75–16] points (κ = 0.627, p < 0.001)	Attribution: 1 Type of study: NA Conflict of interest: 2 Currency: 0 Complementarity: 1 Tone: 2
Special precautions/controversies	Q8: Can I receive acupuncture if I have lymphedema?	Yes, acupuncture can be used to manage symptoms of lymphedema, a condition that can occur after cancer treatment and results in swelling in the arms, legs, or other parts of the body. Acupuncture is a form of traditional Chinese medicine that involves the insertion of thin needles into specific points on the body to promote healing and alleviate pain and discomfort. There is some evidence to suggest that acupuncture may help reduce swelling and pain associated with lymphedema. It may also help improve circulation and lymphatic flow, which can help reduce swelling and improve immune function. However, it is important to note that acupuncture should be used as a complementary therapy in conjunction with other treatments for lymphedema, such as compression therapy, exercise, and massage. It should not be used as a replacement for conventional medical care. If you have lymphedema and are considering acupuncture, it is important to talk to your health care provider before starting treatment. They can help you determine if acupuncture is a safe and appropriate treatment option for you, and can provide guidance on finding a qualified practitioner. It is also important to ensure that your acupuncturist is aware of your lymphedema diagnosis and takes appropriate precautions during treatment to minimize the risk of infection or injury.	Score 13 [10–16] points (κ = 0.433, p < 0.001)	Attribution: 1 Type of study: NA Conflict of interest: 2 Currency: 0 Complementarity: 1 Tone: 1

NA when reviewers provided a score <2 in “attribution” (see Table 2 for scoring details).

IQR, interquartile range; NA, not applicable; QUEST, QUality Evaluation Scoring Tool.

A panel of 12 reviewers, which comprised pharmacists, integrative medicine practitioners, oncologists, public health professionals, and cancer survivors, evaluated the responses from ChatGPT using the QUality Evaluation Scoring Tool (QUEST).⁹ The QUEST is a seven-item questionnaire that has been shown to be a reliable tool to evaluate online information.⁹ The authors modified the QUEST criteria to remove “authorship” because chatbot responses are not authored. The remaining 6 items were “attribution,” “type of study,” “conflict of interest,” “currency,” “complementarity,” and “tone” (Table 2), with a total score of 26 points.

Table 2.

Modified QUality Evaluation Scoring Tool Criteria for Evaluating Online Information

Aspect	Ratings
Attribution	0—No sources 1—Mention of expert source, research findings (although with insufficient information to identify the specific studies), links to various sites, advocacy body, or other 2—Reference to at least one identifiable scientific study, regardless of format (e.g., information in text, reference list) 3—Reference to mainly identifiable scientific studies, regardless of format (in >50% of claims)
Type of study (for all articles scoring 2 or 3 on Attribution)	0—In vitro, animal models, or editorials 1—All observational work 2—Meta-analyses, randomized controlled trials, clinical studies
Conflict of interest	0—Endorsement or promotion of intervention designed to prevent or treat condition (e.g., supplements, brain training games, foods) within the article 1—Endorsement or promotion of educational products and services (e.g., books, care home services) 2—Unbiased information
Currency	0—No date present 1—Article is dated but 5 years or older 2—Article is dated within the last 5 years
Complementarity	0—No support of the patient–physician relationship 1—Support of the patient–physician relationship
Tone (includes title)	0—Fully supported (authors fully and unequivocally support the claims, strong vocabulary such as “cure,” “guarantee,” and “easy,” mostly use of nonconditional verb tenses [“can,” “will”], no discussion of limitations) 1—Mainly supported (authors mainly support their claims but with more cautious vocabulary such as “can reduce your risk” or “may help prevent,” no discussion of limitations) 2—Balanced/cautious support (authors' claims are balanced by caution, includes statements of limitations and/or contrasting findings

The inter-rater agreement was evaluated using Fleiss' kappa, which is a test used to measure the inter-rater agreement between two or more raters (κ > 0.75 indicates excellent agreement, κ of 0.40 to 0.75 indicates fair to good agreement, and κ < 0.40 indicates poor agreement).¹⁰ Statistical analyses were conducted using R 4.1.2 (R Foundation, Vienna, Austria), the “irr” package (v0.84.1).¹¹

To assess the accuracy of the content, the authors also manually consolidated corresponding information from clinical guidelines and well-established databases for each question and presented them alongside the chatbot responses. The reviewers were asked to highlight any potential misinformation or falsified claims in the responses by ChatGPT.

Results

Table 1 shows the responses of ChatGPT to the eight questions presented on March 30, 2023. The median total scores rated by the 12 reviewers were the lowest at 13 points (out of 26 points) for the questions on the interaction between St. John's wort and irinotecan (Q3), adverse reactions of ginseng (Q5), and special precautions on the use of acupuncture for lymphedema (Q8). The scores were the highest at 16 points for the questions on the indication of a CIM modality (Q2) and special precaution/controversies on the use of soy products for patients with breast cancer (Q7). The inter-rater agreement for most questions was fair to good and ranged from a κ of 0.43 (lowest) for Q8 to 0.62 (highest) for Q7.

The distribution of scores for each criterion of the QUEST was relatively consistent across the eight questions. The reviewers generally provided high ratings for ChatGPT responses regarding providing unbiased information (“conflict of interest,” median score: 2/2), encouraging the patient–physician relationship (“complementarity,” median score: 1/1), and having a balanced or neutral tone (“tone,” median score: 1–2/2). However, the scores were low when referencing or presenting scientific studies in the responses (“attribution,” median score: 0–1/3), specifying the type of study or evidence (generally not applicable when reviewers provided a score <2 in “attribution”), and dating the evidence or references (“currency,” median score: 0/2).

In terms of accuracy of information, none of the reviewers identified any explicit misinformation, although two reviewers commented that the team's consolidated responses from clinical guidelines/literature are simply incomparable with the chatbot responses as the latter was vague in nature.

The ChatGPT responses were generated anew each time. Therefore, the authors extracted the responses to the same set of questions 1 month after the initial extraction and compared them (Supplementary Table S1). Generally, there were no major changes to the responses. The authors also used Google to search for answers to the eight questions, and the corresponding results are provided in Supplementary Table S2. Generally, the results are shorter compared with those generated by ChatGPT. The Google results typically originate from a single source, unlike ChatGPT that summarizes information from multiple sources. It is also worth noting that the answers obtained from Google are less straightforward to comprehend as they simply quote the words directly from the websites.

Discussion

Overall, this study showed that ChatGPT provided responses of moderate quality to the questions regarding integrative oncology. The responses were relatively unbiased and did not provide falsified claims or misinformation. Particularly, ChatGPT acknowledged the importance of health care professionals in evaluating the patient's individual situation and providing general guidance on the safety and appropriateness of CIM modalities. Many responses included statements such as “If you are considering/interested in …, it is important to talk to your healthcare provider first to ensure that it is safe for you to try ….” However, in reality, patients often fail to disclose their CIM use to physicians,¹² and a tool such as ChatGPT may further discourage communication on CIM use between patients and health care professionals.

Notably, the modest agreement among reviewers on the quality of information for certain questions indicates that readers may have different interpretations of the response provided. An example of this possibility is that Fleiss' kappa was the lowest for the question on the use of acupuncture for lymphedema, which remains a controversial topic because of limited evidence in the literature. Future studies should examine how patients interpret and perceive the information provided by ChatGPT, particularly when the responses are not definitive. Hence, it is also important to understand patients' downstream actions and tendencies to seek professional advice after using this chatbot.

One of the disadvantages of ChatGPT is its inability to refer to specific scientific studies or literature. When mentioning scientific evidence, ChatGPT often relies on vague statements, such as “several/some studies have suggested that …” without providing details on the type of study or any specific study from which the information was derived. This finding is expected because ChatGPT is supposed to provide simplified, succinct, and aggregated responses that laypersons can understand. Even if ChatGPT provides reference sources or lists citations, it can generate fake citations or summarize incorrect sources.^13,14 This error presents a considerable limitation because verifying the information or evaluating the contemporaneity of the references or evidence without access to the underlying sources is not possible.

In some cases, ChatGPT appears to fabricate convincing responses that contain factual errors or incorrect data when summarizing information about the effectiveness of therapies.⁵ Integrative oncology emphasizes evidence-informed cancer care that uses multiple approaches. Decisions on which approach to use should incorporate patients' values, along with practitioners' clinical experience and current research evidence for the approaches.¹⁵ Therefore, ChatGPT cannot be substituted for an integrative oncology consultation to determine the optimal modalities for a particular patient. Until ChatGPT can provide reliable reference sources and undergo more systematic validation, patients with cancer should be cautioned against relying on it for clinical advice on integrative oncology.

It is recommended that patients should seek reliable information about CIM in cancer from reputable sources. General information about integrative medicine can be obtained from various sources, such as textbooks, journals, and most commonly, reputable websites. In contrast to ChatGPT, these sources undergo rigorous review processes and are typically authored by professionals with expertise in the field. While ChatGPT can offer quick access to information and engage in interactive conversations, it may not have the same level of expertise and depth of information from those trusted sources. When finding information on the internet, patients can look for sources from government agencies, universities, hospitals, or health organizations.

This can be verified by examining the letters at the end of the web address. For example, the URL of government websites end with “.gov” and organizational websites usually end with “.org.” Some examples of reputable websites that provide information on integrative oncology include the National Cancer Institute (https://www.cancer.gov/about-cancer/treatment/cam/patient), the American Cancer Society (https://www.cancer.org/cancer/managing-cancer/treatment-types/complementary-and-integrative-medicine.html), Memorial Sloan Kettering Cancer Center's About Herbs (https://www.mskcc.org/cancer-care/diagnosis-treatment/symptom-management/integrative-medicine/herbs), as well as the National Center for Complementary and Integrative Health (https://www.nccih.nih.gov/health/atoz).

This study has a few limitations. First, this study consisted of only eight questions, which may not be representative of how ChatGPT provides information on the topic of integrative oncology; however, the authors believe that they still serve the study purpose of investigating the quality of information provided by ChatGPT and identifying potential areas for future research. Second, the authors applied the QUEST, which may not be well suited to evaluate chatbot responses. However, to date, there has been no quality assessment tool that is specific to assessing chatbot responses. Information provided by ChatGPT originates from online health information found on the internet. Therefore, the authors consider that the QUEST was a reasonable option for this exploratory study. Third, the level of inter-rater agreement may be influenced, to some extent, by the heterogeneity of the panel, which included a diverse range of health care professionals and patients.

The authors included patients and different health care professionals as participants of this study, considering that ChatGPT is freely accessible and used by various stakeholders involved in cancer care. In addition, the overall fair-to-good level of agreement observed in most questions suggested that there was no significant divergence among participant regarding their assessment of the quality of information generated by ChatGPT. Besides, the version of ChatGPT was trained on data before 2021, and how this chatbot can keep up to date with new information and continue to provide accurate responses are unclear. Notably, a newer version of ChatGPT (GPT-4) has been released at the time of preparing this article, but it requires a subscription.¹⁶ Future research should systematically evaluate updated models of the chatbot and its performance with different questions related to integrative oncology and cancer in general.

Conclusions

In conclusion, this study showed that ChatGPT can provide responses of moderate quality to questions regarding the use of CIM for cancer. The information from ChatGPT is relatively unbiased and not falsified. However, this chatbot's inability to refer to specific scientific studies or literature presents a major limitation. Importantly, patients with cancer should be reminded not to rely on ChatGPT for clinical advice until it undergoes more systematic validation. Future research should examine how patients perceive the information provided by ChatGPT and its effect on communication between patients and health care professionals.

Footnotes

Authors' Contributions

C.S.L.: Conceptualization, methodology, formal analysis, investigation, data curation, writing—original draft, writing—review and editing. R.H., H.K.K., K.Z., T.N.L., C.P.L., W.L.L., C.L.W., Y.M.L., H.H-.F.L., and V.C-.H.C.: Conceptualization, investigation, writing—review and editing. Y.T.C.: Conceptualization, methodology, formal analysis, investigation, supervision, writing—review and editing.

Ethics Approval

This study was approved by the Survey and Behavioural Research Ethics Committee of the Chinese University of Hong Kong (reference no. SBRE-22-0723).

Author Disclosure Statement

No competing financial interests exist.

Funding Information

No funding was received for this article.

Supplementary Material

Supplementary Table S1

Supplementary Table S2

References

Keene

, Heslop

, Sabesan

, et al. Complementary and alternative medicine use in cancer: A systematic review. Complement Ther Clin Pract, 2019; 35:33–47; doi: 10.1016/j.ctcp.2019.01.004

Verhoef

, Trojan

, Armitage

, et al. Complementary therapies for cancer patients: Assessing information use and needs. Chronic Dis Can, 2009; 29(2):80–88; doi: 10.24095/HPCDP.29.2.06

Teplinsky

, Ponce

, Drake

, et al. Online medical misinformation in cancer: Distinguishing fact from fiction. JCO Oncol Pract, 2022; 18(8):584–589; doi: 10.1200/OP.21.00764

OpenAI. Introduction ChatGPT. 2023. Available from: https://openai.com/blog/chatgpt [Last accessed: May 10, 2023 ].

van Dis

, Bollen

, Zuidema

, et al. ChatGPT: Five priorities for research. Nature, 2023; 614(7947):224–226; doi: 10.1038/d41586-023-00288-7

Hopkins

, Logan

, Kichenadasse

, et al. Artificial intelligence chatbots will revolutionize how cancer patients access information: ChatGPT represents a paradigm-shift. JNCI Cancer Spectr, 2023; 7(2):pkad010; doi: 10.1093/jncics/pkad010

Johnson

, King

, Warner

, et al. Using ChatGPT to evaluate cancer myths and misconceptions: Artificial intelligence and cancer information. JNCI Cancer Spectr, 2023; 7(2):pkad015; doi: 10.1093/jncics/pkad015

Cramer

. Artificial intelligence, complementary and integrative medicine: A paradigm shift in health care delivery and research?. J Integr Complement Med, 2023; 29(3):131–133; doi: 10.1089/jicm.2023.0040

Robillard

, Jun

, Lai

J-A

, et al. The QUEST for quality online health information: Validation of a short quantitative tool. BMC Med Inform Decis Mak, 2018; 18(1):1–15; doi: 10.1186/s12911-018-0668-9

10.

Fleiss

, Levin

, Paik

. The Measurement of Interrater Agreement. In: Statistical Methods for Rates and Proportions. ( Shewart

, Wilks

SS.

eds.) John Wiley & Sons, Inc.: US; 1981; pp. 598–626.

11.

Gamer

, Lemon

, Singh

. Package ‘irr’: Various Coefficients of Interrater Reliability and Agreement. R package version 0.84.1. 2012. Available from: https://cran.r-project.org/web/packages/irr/index.html [Last accessed: September 1, 2023 ].

12.

Davis

, Oh

, Butow

, et al. Cancer patient disclosure and patient-doctor communication of complementary and alternative medicine use: A systematic review. Oncologist, 2012; 17(11):1475–1481; doi: 10.1634/theoncologist.2012-0223

13.

Sanchez-Ramos

, Lin

, Romero

. Beware of References when Using ChatGPT as a Source of Information to Write Scientific Articles. Am J Obstet Gynecol, 2023; 229(3):356–357; doi: 10.1016/j.ajog.2023.04.004

14.

Welborn

. ChatGPT and Fake Citations. Duke University Libraries. 2023. Available from: https://blogs.library.duke.edu/blog/2023/03/09/chatgpt-and-fake-citations [Last accessed: May 10, 2023 ].

15.

Witt

, Balneaves

, Cardoso

, et al. A comprehensive definition for integrative oncology. J Natl Cancer Inst Monogr, 2017; 2017(52); doi: 10.1093/jncimonographs/lgx012.

16.

OpenAI. GPT-4 is OpenAI's Most Advanced System, Producing Safer and More Useful Responses. 2023. Available from: https://openai.com/product/gpt-4 [Last accessed: May 10, 2023 ].

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.02 MB

0.01 MB