Optimizing the patient care pathway through data-driven insights

Share this article

The patient’s journey in a healthcare setting is influenced by factors beyond the mere medical aspects of the case. Local variations in care procedures among hospitals and caregivers enable benchmarking and adaptation of practices to optimize outcomes. For effective process modeling, it’s essential to collect data and extract insights. To optimize this process efficiently, collaborative efforts should explore alternative data sources and implement novel tools.

Evidence-based process mining to improve health outcomes

The diagnosis of a condition usually marks the start of a patient journey in a healthcare setting. Various processes, such as clinical investigations and treatment plans, follow and necessitate ongoing decision-making by healthcare professionals and patients. In general, the decision-making process is based on individual patient assessments and is dependent on the hospital and treating physician. However, by using insights extracted from previously collected data, one can benchmark hospitals and procedures. David Smeets, Real World Evidence Manager at Johnson & Johnson Innovative Medicine, explains how this type of process mining can lead to better health outcomes.

“By combining several types of data, we can map the full care pathway of our patients, test hypotheses, and predict which decisions are likely to be most effective.” – David Smeets, Johnson & Johnson Innovative Medicine

The Dutch TripleAiM1 project is illustrative in this respect. This nationwide registry of de novo metastatic hormone-sensitive prostate cancer aims to extract insights on diagnostics and treatment effectiveness, from real-world data. “Some hospitals deploy new techniques for the diagnosis of prostate cancer,” Smeets explains. These techniques are currently more expensive and are not a first-line diagnostic aid in the majority of health centers. “Interestingly, the Triple AIM1 project can give more insights on the patient outcomes and overall costs related to these new techniques, comparing them to the more traditional options.” Christel Meertens, HealthCare Project Manager at imec, emphasizes the value of data-based decision support in the patient care pathway. “Decision support tools can act as advisors, joining healthcare providers at the table. They offer valuable input that can be consulted and considered during the decision-making process, alongside other sources of information”.

Exploring alternative data sources for patient care insights

To enable proper process modeling, high quality and relevant data are key. Apart from clinical and claims data, PROMs – patient-reported outcome measures – need to be included as well, as these reflect the overall impact of healthcare interventions on the patients’ wellbeing. As Smeets illustrates, “Addressing seemingly minor details, like scheduling multiple consultations in a single day instead of requiring the patient to visit the hospital on consecutive days, contributes to overall patient satisfaction and alleviates the stress and fatigue often associated with the medical care process.” Telemonitoring can play an important role. “Lifestyle and sentiment can severely impact a patient’s health status and recovery process, and can be monitored remotely,” says Meertens. “By focusing on these aspects, we can move towards a healthcare system that aims to prevent as well as to cure.” Yet, to do so, steps towards standardization need to be taken even further. “Similar as is already the case for medical devices, we need to move towards more harmonized and normalized data in this area,” emphasizes Meertens.

Telemonitoring can also fill in data gaps in other domains. For instance, when assessing the effectiveness or side effects of a drug, a need for more longitudinal information arises. The current practice is to only monitor patient parameters starting from the moment they receive the drug. “The risk here is that certain concerning trends in parameters may be wrongly attributed to the drug when, in fact, they could be inherent to the patient’s overall health,” says Smeets. “A better option would be to monitor patient parameters before as well as during treatment.” Using medically verified wearables, such as smart watches, allows for the non-intrusive recording of vital parameters. The data is subsequently sent to the healthcare provider for in-depth analysis.

“Looking towards the future, algorithms have the potential to predict optimal times for administering specific therapies, utilizing individual patient data for more personalized and effective treatment planning.” – David Smeets, Johnson & Johnson Innovative Medicine

Innovation and collaboration will pave the way

The patient pathway can be mapped as a digital twin, or virtual representation of the patient based on relevant health data sources. This does not only allow for improvement of the health and quality of life of specific patients but also offers benefits that extend beyond the individual level. It allows us to reduce the cost for all parties involved, and other stakeholders, such as pharma companies, to deploy their treatments in the most effective way. “By gathering insights derived from patient data and presenting these to the government, we can enhance and refine guidelines, ultimately benefiting a larger number of patients,” Smeets says. Although many care professionals and hospitals are willing to implement these innovative tools to optimize their patient care trajectories, certain challenges can be identified. Patient privacy is such an aspect that might appear as an obstacle but doesn’t necessarily have to be one, as technologies to secure privacy are readily available. “Privacy preserving techniques such as federated learning are key,” says Meertens. “It allows patient data to remain local, at the hospital site or with the telemonitoring device, while only aggregated insights are shared among stakeholders, ensuring  higher levels of patient privacy.”

Tools to protect patient data, but also to collect and assess it are being developed by several local, national and international institutions. To further optimize these tools, and evaluate their value in the care process, collaboration is essential.

“We have to convince stakeholders, such as hospitals, pharma and medtech companies, patient organizations, and knowledge institutions to collaborate on concrete use cases.” – Christel Meertens, imec

The Save Data project, a joint initiative between Vito, imec, and MEDVIA, supported by Flanders innovation & entrepreneurship (VLAIO) aims to catalyze exactly that. “With Save Data, we offer technological building blocks that can be used by industrial partners to build their business case,” says Meertens. “As such facilitating the conversion from knowledge development to practical implementation.” Both interviewees agree that establishing local collaborations among like-minded parties, focusing on specific use cases, is crucial to demonstrate the potential value of data in optimizing healthcare. “Small-scale community-based initiatives can evolve into larger platforms, where collaboration is ingrained structurally, and more stakeholders become convinced of its value,” says Smeets. OHDSI, short for Observational Health Data Sciences and Informatics, is one of these international initiatives which provides a platform to collaboratively generate evidence for data-driven solutions. “OHDSI is one of those community initiatives with a clear vision and appropriate funding, both of which are essential if one wants to achieve success,” concludes Meertens. Progress will necessitate a change in mindset for stakeholders. However, the advantages for both the patient and society are undeniably evident.