Effect of AI on the healthcare professional-patient relationship

Besides the influence artificial intelligence may have on healthcare professionals individually, it could also affect the relationship they have with their patients.

This influence can take place at several moments during the patient’s journey. AI can be present in the prevention stage, the diagnosis stage, during therapy, and during the monitoring stage. Let’s look at some scenarios in each of these stages to see how the relationship could be affected.

Prevention

Even before a patient decides to contact a healthcare professional, they now have the possibility to discuss their situation with AI first. For example, they could consult a chatbot or use an application on their phone that tracks a particular aspect of their health for a longer period of time. These AI systems could already provide tips and guidelines on how to deal with certain problems, reducing the need to consult the healthcare professional right away.

If necessary, the AI system could also refer the patient to a particular healthcare professional, along with the required data. Wearables are nowadays able to register lung function, heart frequency and rhythm, temperature, and movements at IC-level accuracy (e.g., Auepanwiriyakul et al., 2020). Some are able to register and analyse sleep patterns or even the release of certain biochemicals. All of this data can be made available to the treating healthcare practitioner, who can then have a look at the data (and even analyse it with the use of AI) before the patient comes in. Besides making more information available, this also provides the healthcare professional with another advantage: it makes the contact with the patient more efficient.

Diagnosis

With the use of the data acquired in the prevention stage, the healthcare professional can continue to diagnose the patient.

Imagine the patient now coming to the professional’s office. There is no need for the healthcare professional to actively fill in the patient’s data in the Electronic Patient Record (EPR) anymore. Instead, an AI-based voice system recognizes who is talking and places all of the information in the correct location in the EPR without any human intervention. The AI is even able to recognize the patient’s mood, and whether or not they understood everything. This allows the healthcare professional to really focus on the conversation with the patient and to make the correct diagnosis, supported by another AI that provides diagnosis suggestions based on the collected information.

This is of course a very promising outlook. However, we could also consider a situation in which the healthcare professional is focusing on the AI systems too much.

For example, the large amounts of data generated by imaging modalities can also be analysed with the support of AI. The system can make predictions and give advice on what to do next. The downside of this, however, is that the healthcare professional can become too focused on the AI system instead of thinking critically themselves. The negative consequences of deskilling and overreliance, which have been explained before, can negatively affect the relationship with the patient. How would a patient feel if they realized the healthcare professional does not know what to do if an AI tool shuts down? Or even worse, if it turns out that the healthcare professional blindly accepted an incorrect AI diagnosis, leading to the wrong treatment? The patient might also simply not trust AI, resulting in them not trusting the healthcare professional who uses this tool.

Therapy

AI can also play a significant role in medication. It can not only help with the development and personalization of the compositions of the drugs, but also with automatic administration. This can be done using wearable devices or implants. In combination with sensors, a closed loop can be implemented, in which continuous monitoring allows AI to decide when the medication should be administered.

Surgery can be supported by AI too. In part by the optimization of the planning and preparation processes, but also by AI-based robots taking over (part of) the surgical procedures.

These applications also reduce the presence of the healthcare professional along the patient’s journey. This can be considered both a positive and a negative consequence. However, it also allows the healthcare professional to provide more efficient and more effective treatments, and to focus on other aspects in which the patient could require more support. This could ultimately strengthen the relationship between them.

Monitoring

Finally, for the monitoring stage of the patient’s journey, specific devices can be provided to the patient which allow for monitoring the course of the disease as well as the patient’s rehabilitation. In the case of chronic diseases, the combination of sensors and automatic administration of medication can offer a better life quality.

This means that the patient would not have to visit the healthcare professional as regularly anymore. The data could be sent to the professional automatically, who could intervene only when necessary. This too can, on the one hand, enhance the relationship between the healthcare professional and the patient and, on the other hand, diminish it. One way to deal with this issue would be to ask the patients for their preferences. How do they feel toward AI replacing the tasks of the healthcare professional? Do they trust it enough to put their fate in its hands? How often would they still want a regular appointment with their physician? Considering the patient’s feelings is one of the most important aspects when transitioning towards an AI-based system.