This multi-level approach presents a roadmap to enhance the use of PROMs data within the healthcare system. It intends to support end-users in understanding the potential usages of this data, and getting the most value out of routinely collected PROMs data.
Many interrelated factors impact the collection and use of PROMs data within healthcare systems. The success of using PROMs data at the micro level depends on many factors including clinicians’ training in using PROMs and their buy-in, the clinical usefulness of the PROM, the ease and timeliness of collecting and reporting PROMs data, integrating PROMs into clinical workflows and electronic health records, ease of interpreting PROMs data, and guidance to support clinical actions based on PROMs data [16, 23]. The use of aggregate PROMs data at the meso and macro levels is limited. Greenhalgh and colleagues suggests that there are three main theories driving the use of aggregated PROMs data in quality improvement initiatives, largely based on the NHS PROMs program, including supporting patient choice (selection pathway), improving accountability and enabling providers to compare their performance with others (change pathway) [18]. While less evidence is available to support the former pathway, emerging evidence supports the “change pathway” [18].
It is important to note that much of the work that takes place to implement PROMs programs in real world healthcare settings does not get published, partly due to restrictions related to patient privacy and consent for data use for research purposes. In a recent international meeting of healthcare system PROMs users, various successful examples of PROMs programs within healthcare systems were presented and discussed, however, most of these programs—with the exception of the NHS PROMs program—have limited or no publications reporting them [24]. In Alberta, the use of PROMs data at the micro, meso and macro levels vary across programs and clinical settings. For example, at Cancer Care Alberta, patients complete a PROM at each visit, then data are entered into a dashboard that a clinician can access during the clinical encounter and use to evaluate change in PROMs scores of a given patient over time [25]. This micro level use of PROMs data has been facilitated by extensive training of clinicians, and integrating PROMs collection and reporting into the clinical workflow of oncology clinics. Alternatively, at the Alberta Bone and Joint Health Institute, patients undergoing hip or knee arthroplasty complete a PROM before their surgery, and at 3 and 12 months post-surgery [25]. PROMs data are entered into a database, and analyzed at the aggregate level to evaluate the effectiveness of these surgical interventions, and to compare across providers and surgical sites, while clinicians do not have access to this data during clinical encounters limiting the micro level use of this data. Detailed information about the use of PROMs data within each of these levels in various clinical areas, and the facilitators and barriers to their use are provided in subsequent papers in this supplement.
The transition of PROMs use from clinical effectiveness research into real world healthcare setting, with all its variants, imposes various practical and methodological challenges on users [11, 12, 14]. For instance, in clinical studies, there are defined time points for data collection and a specified mode of administering PROMs that is often managed by research staff; however, the timing, frequency and completion of PROMs measurements may be challenging to maintain in real-world healthcare settings given various clinical workflows and other contextual factors that vary across settings. Further, in research applications, PROMs data are used at the aggregate level for the whole study sample with a very clear analytical purpose; however, in clinical settings, PROMs data are used both at the individual patient level and at the aggregate level. These practical challenges could be partly addressed with standardization of PROMs measurement and incorporating it into existing measurement frameworks and electronic medical records.
Using real-world routinely collected PROMs data also imposes several methodological challenges that users need to consider. These include attrition and missing data, varying time points of PROM(s) measurement, lack of a control arm in comparative effectiveness analysis, large data pitfalls, sample representativeness, statistical significance versus clinical importance, issues imposed by response shift in long-term data collection especially in chronic diseases, and case-mix adjustment, among others.
Despite significant developments in research and application of PROMs in real world settings in many countries around the world, more evidence and guidance for proper implementation and use of PROMs data is needed. Standards for the selection, collection, interpretation, and reporting of PROMs data with other clinical or administrative datasets are essential to ensure meaningful use of this data for clinical care and policy-decision-making. Future research should focus on approaches of integrating PROMs data with other patient data to enhance its use at all levels within the healthcare system.