This blog entry is a comparison of the content created by an AI tool as compared to content created by a licensed nurse writer. No medical advice is offered in this content. For questions about your health, please call or visit your provider
As a nurse writer, I’m faced with news about Artificial Intelligence (AI) “solutions” for content and copywriting. A popular AI platform called ChatGPT, a chatbot developed by OpenAI and launched in November 2022, is on the minds of healthcare, nurse, and medical content creators everywhere. ChatGPT has the ability to gather common knowledge information or create outlines or summaries of your original content. However, ChatGPT has limitations making the platform an ineffective tool for credibly written, evidence-based healthcare content.
ChatGPT pulls publicly available, dated information. According to an article by abcNEWS, information published after September 2021 is not available to the bot. Some articles require the use of common knowledge content like a blog about Americans’ favorite types of cars or general information about the food industry where scientific information is not needed. However, outdated information written as healthcare content is dangerous and affects consumer safety and outcomes.
To demonstrate, we decided to test ChatGPT against a nurse writer and compare the information. Using the question, “What is the newest technology in blood glucose testing?” we asked GPT-4 and a www.NurseWritersGroup.com team member to present a short outline. Here is what we discovered:
Chat GPT-4 content:
– Introductory sentence: “As of my knowledge cutoff date of September 2021, the newest technology in blood glucose monitoring includes continuous glucose monitoring (CGM) devices that are designed to be more accurate, user-friendly, and convenient than traditional glucose monitoring methods.”
– The first new device is the Evernsence CGM, with the ability to hold data for 90 days, has a less obtrusive monitoring device, and is more convenient than finger sticks.
– The second new device is the Abbott FreeStyle Libre 2, with the ability to pair with your SmartPhone, provide real-time blood glucose readings, track readings, and provide a warning alarm for blood sugars that are out of range.
– Closing sentence: ““It is important to note that as a language model, I do not have access to real-time updates or advancements in technology beyond my knowledge cutoff date. Therefore, there may be newer technologies available now or in the future.”
Nurse Writer Content:
– Introductory sentence: “The American Diabetes Association (2023) recommends several continuous blood glucose monitoring (CGMs) devices. We compare and contrast the top two recommended models from the Gold-Standard Accreditation body in diabetes care.
The Pros of the Dexcom G6 and the Eversense E3 CGM systems.
– The Dexcom G6 CGM system pros consist of the ability to measure interstitial glucose levels every 5 minutes. G6 pairs with several different types of devices, including SmartPhones, SmartWatches, and G6 receivers. The G6 does not require finger-stick blood calibration for verification of readings and stores glucose reading data in the cloud, allowing your provider and care team to access your information at any time of the day or night. This open access puts the power of control in the hands of the consumer and the provider.
– The Eversense E3 CGM system pros consist of continuous, not timed interval, monitoring of interstitial glucose levels. The E3 does not require finger-stick confirmation of readings and has a sensor that is placed under the skin allowing for freedom of movement and activity, including exposure to water. The E3 communicates with your SmartPhone and does not require an external transmitter to communicate readings. Your CGM readings are held in the E3 application and can be accessed by up to five (5) caregivers involved in your diabetic care.
The Cons of the Dexcom G6 and the Eversense E3 CGM Systems
– The Dexcom G6 CGM system cons encompass the external wearing of a sensor with a shorter sensor life of 10 days and the requirement for a large hand-held transmitter to be used with the system with a limit of 20 unobstructed feet between the sensor and the receiver. The sensor measures 1.8″ x 1.2″ x 0.6″ or about the size of a standard thumb drive. The receiver is 4.2″ x 2.5 ” x 0.6″ or is about the size of a small cell phone. Additionally, the receiver weighs about 4 or more ounces, depending on the cover case.
– The Eversense E3 CGM system cons require every 12 hours finger-stick blood glucose calibration, and the limitations of the accuracy of a result lower than 40 mg/dl and higher than 400 mg/dl, similar to standard finger-stick blood glucose meters. Additionally, only a healthcare provider can insert the E3 sensor under the skin, necessitating office visits and the limitation of up to five (5) caregivers to access the blood sugar reading data via the application.
– Closing – How does this information affect patients managing or considering managing their glucose readings with CGM devices?
For ease of use, a system with a small sensor and transmitter is preferred. Larger sensors inserted into the skin and worn with an adhesive dressing can be cumbersome, limit movement, get caught on clothing, prevent participation in water-based activities, and become sore or irritated with movement. Providers report patients with a small sensor and small transmitter are more compliant with wearing the device all the time. However, patients preferring a simplistic setup with fewer office visits might be amenable to the larger sensor and receiver as these options are more important to them.
CGMs provide more accurate readings than traditional finger-stick methods as the sugar readings use interstitial fluid, or the plasma of the blood, to determine results. Interstitial fluid readings represent not only the amount of sugar in the blood but also the amount of sugar absorbed into the cells, making these readings more accurate than whole-blood finger sticks. Choosing a CGM that best matches your lifestyle is important in managing your overall health.
Only you and your provider decide which method of blood-glucose or sugar monitoring is best for you. With all of the choices available on the market today, choosing a method that is convenient, accurate, affordable, accessible, and comfortable is important to consider. For more information about the CGMs reviewed today, please visit the ADA website and read the full reviews.
As illustrated in the examples, the ChatGPT output lacks evidence-based content. For example, information from the American Diabetes Association (ADA), the governing body in standards of care in diabetes, is neither reviewed nor identified. The nurse writer, based on licensure, education, and experience, knew to review the product recommendations of the ADA. ChatGPT is not capable of this higher-level decision-making.
The ChatGPT content documents superficial information available to anyone on the web and lacks the synthesis of the information to help healthcare consumers make decisions about their diabetic care. Additionally, ChatGPT lacks easy-to-understand information aimed at patient education and safety. The nurse-authored article explains key terms, identifies the pros and cons of both devices and reminds the consumer to discuss their care with a provider to make the safest care plan decisions before investing time or money into a device.
Consumers of healthcare content want credibly written, evidence-based material to engage with their providers. Providers want patients to be involved in their plans of care as research demonstrates patients who are engaged in decision-making have better overall health. To bridge this information gap, only evidence-based healthcare writing is acceptable, something AI, like ChatGPT just cannot provide.
What do you think? Could AI be the future of healthcare content? Why or Why Not?
Visit our website, www.NurseWritersGroup.com, and click on “Blog Submission” to send your thoughts and feedback. We’d love to hear from you!
