A recent review (1) published in the most influential journal in the field of oncology examined the usefulness of patient-generated health data (PGHD) in cancer care. While concluding that PGHD have potential benefits which will make them increasingly likely to be integrated into oncology research and clinical care, there are some challenges to overcome before routine integration into clinical care can happen, namely:
- Medical record integration
- Analysis of large and complex biometric data sets
- Potential clinic workflow redesign
Designed in collaboration with clinicians, the Vitaly solution for Multidisciplinary Team Meetings already addresses the problems identified by the review, thus providing a scalable and sustainable platform for support of outpatient management of patients with cancer (and also other diseases).
Medical record integration
Built on top of the interoperable Vitaly eHealth Platform (HL7® FHIR®) the data generated by patients via the Coordinated Care solution can easily be written to (as well as read from) interoperable health data silos of patients. Therefore, health care professionals (HCPs) can see the results of PGHD in the hospital’s electronic health record (EHR). Not only that, while using Vitaly, also patients can access their documents that are stored in the EHRs (thus having an insight into their health status).
However, EHRs usually do not have capabilities for an appropriate presentation of PGHD. In fact, a recent study about the use of PGHD (specifically patient-reported outcome measurements (PROMs)) in a large radiation-oncology clinic (2) has found that 60 % of users had difficulties accessing the data in the EHR, thus limiting the usefulness of PROMs. Not only patients but also healthcare professionals can use Vitaly during their routine clinical work (i.e. for accessing/viewing PGHD). Due to the user-centric design of the solution, the use is intuitive with data displayed in a user-friendly and understandable way. This all translates to timesaving and increased satisfaction when interacting with the solution.
Analysis of large and complex biometric data sets
In the heart of the solution lies a Care Map template (set of tasks, goals, care team member roles etc.), – a concept that allows for implementation of practically any disease management protocol including care for patients with cancer together with a collection of numerous PGHD (i.e. PROMs, blood pressure measurements, activity measurements etc.). As described above, because of the interoperability and user-centric solution design, the collection of the PGHD is automatic (integration with medical devices and gadgets) and intuitive (when manual input is needed). Since the collection of PGHD is usually periodic and frequent, this results in the generation of large data-sets of PGHD; manual analysis of such large sets would be time-consuming putting additional strain on already overloaded HCPs. Therefore, the Coordinated Care solution offers two complementary functionalities to assist in data analysis, namely:
- Data analysis with machine learning (ML) support – with the use of the disease agnostic ML algorithm (input parameters and outcomes can be defined for any specific disease/measurement) the decision support for disease management of interest is not based solely on the expert opinion defined thresholds (please see clinical workflow redesign part below) but is also data-driven. As humans are poor at discerning hidden patterns in big data sets (3), the threshold-based prediction probably under-utilizes the prediction power hidden in the data (4,5). Therefore, machine-learning support further adds to improved management of patients (i.e. by extracting all the potential of PGHD in patients with cancer).
- Automatic data aggregation and analysis, according to thresholds – covered in the following chapter.
Potential clinic workflow redesign
As mentioned above, one of the functionalities Vitaly includes is automatic data aggregation with threshold-based analysis. Instead of looking at every particular data point of the PGHD (i.e. every single PROM result, every single measurement of blood pressure/temperature etc.), the professionals can pre-set data aggregation rules/thresholds (i.e. weekly median value of blood pressure < 140 mmHg, difference between 2 consecutive PROM values < 2 (0-4 scale) etc.) and when the value of the observed parameter is out of the pre-defined goal range, a system automatically triggers notification informing the responsible care teams of the pathological result. As already pointed out, this functionality enables standard clinical workflow redesign by allowing clinicians to focus only on important information and not wasting their time on analysis of data with little added value (i.e. all PGHD as compared to only pathological values).
Conclusions
Because of the significant health-related benefits that they provide (6-9), PGHD are slowly becoming an essential part of the management of patients with cancer. Interoperable, user-friendly solutions that enable clinical workflow automation and machine-learning support are a must for this transition to happen swiftly and effectively in order to offer patients with cancer best possible care.
Julij Šelb, MD PhD
Parsek’s Medical consultant
References
1.) Jim HSL, Hoogland AI, Brownstein NC, et al. Innovations in research and clinical care using patient-generated health data. CA Cancer J Clin. 2020;
2.) Rotenstein LS, Agarwal A, O’neil K, et al. Implementing patient-reported outcome surveys as part of routine care: lessons from an academic radiation oncology department. J Am Med Inform Assoc. 2017;24(5):964-968.
3.) Rajkomar A, Dean J, Kohane I. Machine Learning in Medicine. N Engl J Med. 2019;380(14):1347-1358.
4.) Orchard P, Agakova A, Pinnock H, et al. Improving Prediction of Risk of Hospital Admission in Chronic Obstructive Pulmonary Disease: Application of Machine Learning to Telemonitoring Data. J Med Internet Res. 2018;20(9):e263.
5.) Sanchez-morillo D, Fernandez-granero MA, Leon-jimenez A. Use of predictive algorithms in-home monitoring of chronic obstructive pulmonary disease and asthma: A systematic review. Chron Respir Dis. 2016;13(3):264-83.
6.) Basch E, Deal AM, Dueck AC, et al. Overall Survival Results of a Trial Assessing Patient-Reported Outcomes for Symptom Monitoring During Routine Cancer Treatment. JAMA. 2017;318(2):197-198
7.) Denis F, Lethrosne C, Pourel N, et al. Randomized Trial Comparing a Web-Mediated Follow-up With Routine Surveillance in Lung Cancer Patients. J Natl Cancer Inst. 2017;109(9).
8.) Denis F, Basch E, Septans AL, et al. Two-Year Survival Comparing Web-Based Symptom Monitoring vs Routine Surveillance Following Treatment for Lung Cancer. JAMA. 2019;321(3):306-307
9.) Cardoso F, Senkus E, Costa A, et al. 4th ESO-ESMO International Consensus Guidelines for Advanced Breast Cancer (ABC 4)†. Ann Oncol. 2018;29(8):1634-1657.