What benefits can AI offer patients?

1. Early detection and diagnosis

AI systems can help detect diseases early and make accurate diagnoses by analyzing large amounts of patient data and recognizing patterns that might be missed by doctors.

Early detection is also desirable in cases of alpha-1 antitrypsin deficiency, as the diagnosis is often made far too late in many patients, and sometimes not at all. Computer systems that analyze existing symptoms and present possible causes could be helpful in the future. For example, in ChatGPT, when asked: "I have pulmonary emphysema and have never smoked. My father died of liver disease. What could I have?", the answer is: "...various diseases could be possible, including genetic disorders such as alpha-1 antitrypsin deficiency, which can affect both the lungs and the liver...". This information and the inclusion of alpha-1 in the possible causes of the symptoms could be crucial for early diagnosis.

2. Personalized Medicine

By analyzing genetic information, health data, and other relevant information, AI systems can create personalized treatment plans tailored to the individual needs and characteristics of a patient.

3. Medication management:

AI can help patients manage their medications more effectively by sending medication reminders, monitoring drug interactions, and even suggesting alternative treatment options. This is particularly important for patients with chronic illnesses, such as COPD, who often have multiple comorbidities and frequently require a variety of medications. AI systems can help physicians optimize prescriptions by minimizing the number of prescriptions and more efficiently identify potential drug interactions.

Health management and lifestyle counseling:

AI-powered health and wellness apps can support patients in improving their lifestyles by offering nutritional and fitness advice, monitoring sleep patterns, and helping with stress management. In sleep medicine, AI-supported remote monitoring is already in use and has proven effective. Findings and current parameters can be transmitted without delay, allowing the patient's situation to be monitored and treatment measures to be continuously adapted.

Another example in healthcare management is the digital health application (DIGA) "kaia" for COPD patients, which can be prescribed free of charge by their pulmonologist. Using this app, daily home workouts can be monitored and optimized. An integrated motion scanner, for example, provides real-time correction recommendations, helping with the correct execution of exercises and preventing errors.

What advantages can AI offer doctors?

„"The crucial point is that humans retain control over AI! It is important that doctors are always able to recognize potential errors in AI and their sources. This, in turn, requires that doctors can understand how AI is developed and how the applied algorithm arrives at its results," states the Association of Scientific Medical Societies (AWMF) – one of the most important expert bodies in Germany – and calls for more teaching on AI applications already in medical studies.

This required foundation now needs to be put into practice!

Numerous studies already show where AI can support the work of doctors in the future, although by no means always with an advantage for AI; the following are two examples:

• AI apps can detect skin cancer as reliably as doctors. However, in its treatment recommendations, AI applications tend to remove more benign lesions than experts would. If this is taken into account, AI applications can certainly be used. It should be considered, however, that indiscriminate use could lead to too many false-positive results requiring further investigation. (Source: Menzies SW et al. Lancet Digital Health 2023)
• AI could support clinicians in diagnosing hospitalized patients – or so the hypothesis goes. However, systematic biases in AI models can worsen clinicians' diagnostic accuracy (Source: Jabbour S et al. JAMA 2023).

AI and Alpha-1 Antitrypsin Deficiency

There are also research projects that directly address alpha-1 antitrypsin deficiency. Two current projects are described below: A combined machine learning (ML) model for predicting the clinical outcomes of patients with alpha-1 antitrypsin deficiency and associated liver diseases, based on the British biobank, has two main objectives: firstly, to develop a better understanding of disease progression, and secondly, to better identify the appropriate time to initiate specific therapeutic measures.

Another scientific study examines the collaboration between artificial intelligence and pulmonologists to improve the accuracy of interpreting lung function tests. The pulmonologists participating in this study also used AI feedback to inform their decisions and consistently improved their assessments when the AI provided accurate predictions.

In conclusion, it appears that collaboration between a pulmonologist and AI in the interpretation of lung functions is more optimized than the assessment by the individual pulmonologist without AI support.

Furthermore, AI is becoming increasingly established in the field of radiology, which may be significant for improving emphysema diagnostics and also for the early detection of exacerbation risk (risk of acute deterioration).

Summary

Artificial intelligence is accompanied by a rapid development of new technological options, and AI is transforming medicine. The various AI models are continuously evolving and improving, which simultaneously leads to an expansion of their potential applications – even to the point where AI is being considered as a "development aid" for new medications.

Nevertheless, AI should not be overestimated, because where there is light, there is always shadow. Doctors and patients should be aware of the opportunities, but also the risks and limitations of AI. It is crucial to engage with the topic of AI – and to consider both its advantages and disadvantages. AI cannot and will not replace a doctor's visit; AI always needs feedback from an experienced physician!

Virtual doctor's appointments have become almost part of everyday life since the pandemic.

The following is stated on the website of the Federal Ministry:

„"The Federal Ministry of Health has recognized this and, in recent years, has created the framework to allow the many advantages of digitalization to be utilized. Important steps have been taken, always with the aim of improving healthcare in Germany. These include the introduction of the electronic patient record (ePA), the introduction of the electronic prescription (e-prescription), the creation of a new access point for digital health applications (DiGA or 'app on prescription'), the advancement of telemedicine, and the establishment of the Health Research Data Center at the Federal Institute for Drugs and Medical Devices."„

• We already reported on the introduction of the ePA in Journal 1/2021 – unfortunately, from our point of view, not much more has happened and far too few people are taking advantage of these opportunities.
• According to the Federal Ministry of the Interior, the path to the electronic prescription was also very long: "Because from September 1, 2022, pharmacies throughout Germany will be able to redeem electronic prescriptions and bill the health insurance companies." Only that late, you think? We also thought it was already possible.
• What about so-called DiGAs? A DiGA is a digital health application, a medical device, e.g. in the form of an app.

At the Federal Institute for Drugs and Medical Devices (BfArM) Here you will find everything you need to know, especially the list of DiGAs (Digital Health Applications) available by prescription. Currently, 33 apps are listed in the directory. However, only a few doctors have prescribed such apps so far. As also reported in our journals, you can find good apps by looking for the PneumoDigital seal. A list and further information can be found on the [website/link]. Pages of the Respiratory League.

Ultimately, we, as affected individuals, hope that digital possibilities in healthcare will provide assistance in everyday life and a well-informed doctor who may even have more time for us and our problems through the use of these tools.

Returning to the question: How can we advance all these digital tools – which are actually intended to support patients? Can we influence digitalization?

Our answer: Yes! Because digitalization is coming, no matter what we do. But should health insurance companies, doctors, and/or pharmaceutical companies decide what we patients need? It's surely better to be involved from the outset and to say how we envision a digital future and how we can benefit from it.

We will keep you updated on details and our projects as best we can.