Health app to be integrated into clinical trials
Questionnaires are a thing of the past, now it is becoming increasingly likely that information for clinical trials will be collected via apps in the future. A new research platform has been developed to prepare patient data transmitted via smartphone applications. Doctors who are carrying out research will thus have access to more accurate results that will also ultimately benefit patients.
A team headed up by Dr. Heinrich Lautenbacher from the IT division at the University Hospital in Tübingen is developing a digital research platform to acquire and access the data of patients who are participating in clinical trials. The IMeRa (Integrated Mobile Health Research Platform) project receives around 470,000 euros in funding from the Baden-Württemberg government. The data are collected by health apps specifically developed for use in clinical trials. Patients use the apps to collect and transfer information such as subjective health ratings and adverse effects of drugs they are taking. The app also processes vital parameters such as pulse, galvanic skin response and activity patterns. These parameters are recorded by sensors that the patients carry with them.
Lautenbacher is a medical doctor who has also trained in medical informatics. He knows from his own experience what his medical colleagues need. “Many people use smartphones and apps can be downloaded quickly and easily. In terms of follow-up care alone, we believe that we can reach significantly more patients with apps. Apps also help us collect information more frequently than with paper questionnaires. Oncological studies usually involve hundreds of people; other clinical trials can even involve several thousand. Sending clinical trial participants paper questionnaires is a time-consuming process, and not all volunteers are willing to fill out and return them every few days. We hope that the app will generate much better feedback,” said Lautenbacher. In addition, activity trackers also help collect activity patterns and body functions of clinical trial participants. Such data provide doctors with useful information when assessing a patient’s condition.
Harmonising data and providing system-compatible data
However, the two-year project goes far beyond the development of mobile apps. Its overall objective is to develop the complete infrastructure needed for handling mobile data of clinical trial patients. Amongst other things, the data will be used to gain new scientific insights into the effect of different forms of medical treatment, and thus help doctors make better therapeutic decisions. The researchers have developed a database that not only acquires and stores app data, but which is also able to merge data from apps with other data sources, process them and pass them on to specialised research software. “At hospitals we use many different systems to acquire data used for assessing a patient’s health – digital patient records, hospital laboratories, radiology, as well as mobile data. It all has to be integrated into one system. The project seeks to develop special software programmes to integrate all the data,” said Lautenbacher.
Although the project will be less than halfway through by summer 2017, the team from Tübingen is already ahead of schedule. A database prototype is already available and is being used in a clinical trial performed by the Department of Paediatric and Adolescent Psychiatry with school-age children. The team faced quite a few challenges in the first pilot trial. The health data of children are particularly sensitive and require the highest level of data security. Data protection involved three levels. Lautenbacher comments: “Data transfer can be insecure. However, our app developer in the Department of Paediatric and Adolescent Psychiatry solved the problem with state-of-the-art encryption methods. In addition, the data are collected and transferred in pseudonymised form, which means that names and other personal information of clinical trial patients are not shown at any point in the process. The patient only appears as a number that means nothing to unauthorised people. The person behind the number is known only to the principal investigator.” The second level is data storage. The group of developers at the University Hospital solved this particular problem by storing the data on one of the clinic’s servers that is maximally protected by cascaded firewalls. The third and most difficult task was to ensure data protection on mobile devices for which the patients were the end users.
Explaining and raising awareness of data protection
The children that participated in the pilot study were given smartphones with settings that could not be changed and which could not be used to download anything from the internet. As the pilot study only involved between 30 and 40 children, the cost of the phones was manageable. However, making phones available to thousands of patients enrolled in large-scale clinical trials is a completely different and much trickier issue. Patients would then have to use their own smartphones. This carries its own particular responsibility. “Although the app’s pseudonymisation function offers some protection, it cannot be ruled out that malware may obtain access to patients’ personal data. A very important aspect for us is to thoroughly inform clinical trial participants about potential risks, and they also have to give their express consent to using the health app,” said Lautenbacher.
Another challenge associated with clinical trials is data validation; the correctness and suitability of the data collected needs to be assessed. This applies to both “hard” data such as pulse, body temperature and galvanic skin response, as well as the “soft” data provided in patient questionnaires. “The validation of data occurs on three levels. The first level is the app, the second the database, where the data undergo plausibility tests to filter out inexplicable data, for example pulses outside plausible ranges. The third level is where the principal investigators validate the data according to their own criteria,” said Lautenbacher. He expects the app to enable a larger number of conclusive soft, subjective data to be collected. “Take a patient who has just undergone tumour surgery. We want to find out how he or she feels in everyday life, not just whether, and for how long, the patient survives. We want to analyse and evaluate whether and to what extent health problems have an influence on the patient’s life,” said Lautenbacher. This information is important for assessing whether therapy A is better than therapy B.
In order to demonstrate the practicality of the new research platform, six new clinical trials that will use the new platform are being prepared. These will also focus on any app developments that are required. The platform will be tested in different disciplines, for example oncology, gynaecology and paediatrics. The platform will not only be used by researchers and doctors from Tübingen. “We want to cover a clinical trial range that is as broad as possible and that also takes into account the requirements of other hospitals. We are already working with the Freiburg University Medical Centre to make the research platform suitable for application in orthopaedics. The platform software is an open source solution and is provided free of charge to other university hospitals. This was a requirement of the Baden-Württemberg ministry that sponsored the project,” said Lautenbacher.