The Orthopaedic Forum Strategies for Effective Implementation of Patient-Reported Outcome Measures in Arthroplasty Practice

Abstract

Patient-reported outcome measures (PROMs) will be an important component of real-world evidence, but best practices for capture and integration are not yet defined. While digital tools support patients, clinicians, and researchers to collect PROMs, PROM capture in clinical care remains challenging. We synthesized PROM implementation strategies that are successfully used by hundreds of arthroplasty surgeons and early PROM-adopter clinical systems. This information can guide health systems that are preparing to implement PROMs to inform clinical care, drive quality-improvement activities, and support reporting for payer-sponsored incentives. Specific information is included to guide each step in the implementation process, including selecting PROMs, redesigning office procedures, programming of information technology, and informing interpretation and clinical use. While no one solution exists for successful PROM implementation in total hip and total knee replacement, referred to as total joint replacement, these guidelines can inform optimal PROM deployment and the capture of complete PROM data. In addition, we outline future research that is needed to define methods for optimal patient engagement, technology infrastructure, and operational systems to seamlessly integrate PROM collection in clinical care.

The National Academy of Medicine Leadership Consortium Evidence Mobilization Action Collaborative promotes best practice for real-world data collection so that clinicians and health systems can provide personalized care to patients and generate data to improve research and the quality of health care¹. Patient-reported outcome measures (PROMs) are a critical piece of this real-world evidence vision, yet best practices for capture and integration are not defined. Digital tools exist to support patients, clinicians, and health systems to collect PROMs, but integrating PROM capture into clinical care remains challenging. The focus on PROM capture for patients who undergo total hip or total knee replacement, referred to as total joint replacement (TJR), is important because TJR is a common and costly procedure that is designed to improve pain and physical function, which are health domains that are best measured with PROMs. Recognizing their importance, the Centers for Medicare & Medicaid Services (CMS) offers incentives in its bundled payment program for PROM reporting before and after TJR². Other private payers have followed this model. Thus, complete longitudinal PROM capture with TJR is a priority among orthopaedic practices.

The purpose of this paper is to synthesize PROM implementation strategies that are successfully used by hundreds of arthroplasty surgeons who participate in the U.S. TJR registries and early PROM-adopter clinical systems. This information can guide health systems that are preparing to implement PROMs to inform clinical care, drive quality-improvement activities, and support reporting for payer-sponsored incentives. Specific information is included to guide each step in the implementation process, including selecting PROMs, redesigning office procedures, programming of information technology (IT), and informing interpretation and clinical use. While no single solution exists for successful PROM implementation with elective TJR, these guidelines can inform optimal PROM deployment and the capture of complete PROM data.

Selecting PROMs for TJR

Optimal implementation begins with multistakeholder discussions to understand how PROMs will be used within a health system. Uses include guiding clinical care, supporting quality-improvement activities, reporting to regulatory bodies, and research^3,4. If clinicians will use PROMs to inform care at patient visits, previsit collection of PROMs is a priority. If PROMs will be reported for payment incentives, specific PROM domains and time frames may be required. Once the intended uses are clearly defined, PROM domains and measures, time intervals, and collection strategies can be determined.

Internationally, the most common PROM domains that are measured in patients who undergo TJR are physical function, pain interference, and pain intensity. Universal (“generic”) measures, such as the Patient-Reported Outcomes Measurement Information System (PROMIS)⁵, and disease-specific legacy measures reliably capture these domains. The CMS bundled payment program endorses reporting universal measures (e.g., PROMIS Global, Veterans RAND-12 [VR-12]) and disease and joint-specific measures. The relative advantages and disadvantages of both universal and disease-specific measures should be considered before final measure selection.

PROMIS Measures with TJR

The National Institutes of Health developed the PROMIS to address the shortcomings of legacy PROMs, namely limited measurement precision and range, the inability for use across diverse clinical conditions, and excessive burdens on patients and practices⁵. The PROMIS measures are brief and precise, and they demonstrate validity across diseases, conditions, and procedures. The PROMIS measures are universal, meaning that they capture health domains across clinical conditions. Such measures are appropriate for use across all patient conditions, including orthopaedic procedures, and for patients with multiple comorbidities. This is an important consideration for implementation within a health-care system because patient burden is reduced when the same measure can be used across clinical settings to minimize repetition. The PROMIS can precisely measure patient outcomes and symptoms across the full range of severity and precisely capture change (improvement or deterioration), which is a critical benefit when PROMs are used to evaluate the quality of TJR care^6,7.

The PROMIS measures can monitor TJR outcomes because it provides independent scores for physical function and pain. Although physical function and pain are correlated, each domain may follow independent longitudinal trajectories and treatment implications⁸. With TJR, approximately 10 PROMIS items assess physical function and pain interference with the use of computer adaptive tests (CATs), or 15 items with the use of static forms, which require ≤4 minutes to complete⁹. All of the measures have been translated into multiple languages and are available in multiple electronic health records (EHRs) (see Appendix 1).

Clinicians may also consider global PROMs to assess general health status. The PROMIS Global scale (10 items) is a widely used measure that provides a summary of one’s physical and mental health. Global scores are helpful in comparing groups of individuals but are not ideal for measuring change in an individual patient. Because global scores are composites of multiple domains like physical function, pain, and fatigue, it is difficult to translate them into specific clinical action¹⁰. Improvement or deterioration over time is difficult to interpret for a single patient because it is not clear what symptom or function is driving that change. Global measures are prone to being less sensitive to change in single aspects of health.

Disease-Specific Legacy Measures

Beyond universal and global measures, disease and joint-specific outcome measures may be collected. Commonly used TJR measures include the Hip disability and Osteoarthritis Outcome Score (HOOS) or the Knee injury and Osteoarthritis Outcome Score (KOOS)¹¹, which have items from the Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC)¹². Because the KOOS and HOOS assessments include >40 items, clinicians often prefer shorter versions. The 12-item versions retain the ability to estimate the pain, function (activities of daily living), and quality-of-life subscores of the full-length assessment^13,14. In contrast, the HOOS or KOOS Joint Replacement (JR) score is a summary score and does not generate domain-specific scores^15,16 (see Appendix 2 ).

Office Procedures for PROM Implementation

Office staff participation in the implementation of PROMs is essential for successful patient engagement and the collection of complete patient-reported data. Patient engagement is improved when staff emphasize the connection between the patient’s responses and treatment decisions during the clinical visit. Providing staff with training and suggested scripts will help them explain the value of PROMs to patients (i.e., that this is not only paperwork but generates visit information). Supporting staff to engage patients can help maximize PROM collection¹⁷.

PROM collection that is integrated into routine office workflow enables complete data capture. High completion rates are achieved using multimodal strategies that provide the opportunity to collect PROMs at flexible times based on patient convenience rather than limiting completion to the time of office visits. For example, PROM collection can be integrated into previsit reminders, linked to the validation of billing information on clinic arrival, or distributed as a post-visit follow-up¹⁸. Once PROM collection is integrated into the workflow, trained and accountable staff are essential.

Previsit patient engagement strategies may include mailing a letter or sending an email from the clinician to encourage patients to complete the assessment prior to their visit^19,20. For patients who did not complete the previsit PROMs, an office clerk may be able to support patients in completing PROMs on the day of the visit. Relying on patients to log into the patient portal independently has proven insufficient to ensure complete PROM capture. Not all patients maintain EHR portal accounts, and many require technical assistance. As patients become more consistently technology-savvy and easy-to-use portal interfaces expand, electronic (e)-PROM collection will become standard.

While PROMs can be automatically distributed prior to each visit, incremental staff time may be required to support patients who did not complete the automated PROM. Low-volume offices may find that the front-office staff can distribute technology and monitor the completion of PROMs. In busy specialty offices where all patients are requested to complete PROMs, an additional clerk or front-desk staff member may be required to manage the technology and the completion of PROM reports to ensure complete collection.

Beyond the staff, office-based PROM capture requires a dedicated space, IT resources (e.g., web access or tablets) to capture the data, and IT programming to make PROM scores available to the clinical team. In the COVID-19 era, health systems are minimizing patient touch-screen technologies and increasing the importance of previsit PROM capture and EHR patient portal technology.

Clinicians should define routine intervals for PROM completion whether or not the patient has a scheduled visit. Post-TJR data capture is particularly important for quality outcome and regulatory incentives. Office staff can utilize multiple methods to contact patients (e.g., an EHR portal, a telephone prompt by the staff, or an office visit). Including a message from the clinician that explains why the post-treatment PROM is helpful in assessing care impact and monitoring the quality of care will emphasize the importance¹⁸. When physicians encourage patients to complete a PROM assessment and then use the scores when reviewing symptoms at the clinic visit, patients are more likely to value the PROM and be willing to complete future assessments. Staff responsibilities should be clearly outlined regardless of whether patients are in the office or at home.

IT for PROM Implementation

IT is critical to enable PROM capture within the EHR patient portal before and during clinic visits and at specified intervals. This is facilitated by native EHR applications (e.g., the PROMIS CATs application for Epic).

PROMs that are captured before or at the time of a clinic visit can be utilized in the visit. Fixed assessment intervals (e.g., 9 months after TJR) facilitate clinician ability to monitor treatment outcomes and support payer requirements. Working with clinical teams, IT will establish assessment windows (e.g., a previsit 2 weeks before surgery) and schedule reminders¹⁸. PROM management reports should be programmed to identify any missing assessments.

In addition, the EHR should be programmed to capture contextual variables that are essential to interpreting PROMs and reporting to regulatory programs. For example, height and weight for calculating body mass index, concurrent pain in the other knee or hip (in the case of a total knee or hip replacement) or the lower back, and opioid use are variables that inform TJR care. Because these data are inconsistently documented, the CMS recommends that bundled payment participants capture this information from the patient at the time that PROMs are captured^2,21,22.

In order to use PROMs in clinical care, they must be scored and accessible to the clinician in an interpretable format. IT can program scoring algorithms or use the native EHR PROM-scoring functions. In collaboration with clinicians and stakeholders, IT can determine PROM-report content and develop graphical summaries to report scores over time so that clinicians can identify trends.

Beyond individual patient PROM summaries, PROM implementation requires management reports to guide staff and clinicians to achieve complete data capture. Beyond lists of incomplete previsit and follow-up assessments for office-staff use, IT can generate clinical management reports with a dashboard featuring completion rates by groups (e.g., clinic or surgeons) as well as descriptive statistics of PROM completers by assessment time point (e.g., 1 year after TJR) or by patient attributes (e.g., age and primary language) to refine the implementation procedures. To enable clinician and quality outcome use, the IT group should also generate reports to summarize aggregate PROM scores such as average pain or function before and after treatment (e.g., before and at 1 year after TJR) for all patients in the clinic.

The IT time that is required to program PROM collection through the EHR, as well as the monitoring of reports, will vary by system. However, we estimate approximately 40 hours of programming time at the initiation of PROM collection for each clinic. Over time, we estimate that 1 day per month will be needed to manage reports and any IT modifications.

Using PROM Scores for Patient Care

PROMs can be a useful tool in clinical encounters. Arthroplasty visits routinely include an assessment of knee or hip pain and physical function. Reviewing PROM scores can guide the clinical interview (e.g., “It looks like pain is a substantial challenge.”). The clinician can verify the PROM score interpretation, and the patient can provide additional information to facilitate treatment decisions. After TJR surgery or nonoperative treatment, changes in scores on repeated PROMs provide information regarding the magnitude of improvement or decline in outcomes, helping to identify when other treatments should be considered. Clinicians should also keep comorbid conditions in mind while interpreting PROM scores. For example, many conditions can influence physical function scores, such as pulmonary conditions or pain in untreated weight-bearing joints or the lower back. Clinician interpretation of individual PROMs will involve consideration of each patient’s unique history.

Diverse PROMs are scored using varied metrics. The PROMIS Pain Interference and Physical Function measures T-scores. In the U.S., a T-score is a standard score with a mean of 50 and a standard deviation of 10 in the general population; the score usually ranges from 20 to 80. The PROMIS Pain Intensity nominal rating scale ranges from 0 (no pain) to 10 (worst imaginable pain). For all 3 of the described measures, high PROMIS scores indicate “more” of each domain: high pain interference means more pain that disrupts usual activities, high pain intensity means more severe pain, and high physical function means better function with more mobility. In contrast, the KOOS and HOOS measures are generally scored on a 0 to 100 scale, with lower scores reflecting poorer function or greater pain.

PROMs for Quality Improvement and Value-Based Payment Programs

PROMs are aggregated to evaluate care quality and are reported to payers in value-based payment programs. Comparable patient data are particularly important when comparisons will be made across health systems to generate quality rankings²³. It is important to ensure that complete and consistent PROM data are collected from the majority of patients in order to minimize bias. For example, PROM collection is likely to include all patients with complications who schedule clinician visits after TJR. It is also important to ensure that PROMs are collected from patients who are healing without problems and may not visit the clinician for follow-up. Thus, PROM collection methods should prioritize representative PROM collection from patients in the 9 to 12-month interval after surgery. Additionally, comparative outcome analyses will adjust for the case mix using administrative data that reflect comorbid conditions. The CMS and other payers specify which administrative codes and/or clinical and demographic variables should be used when adjusting for the case mix before comparing outcome data^2,24,25. Health systems should ensure that these contextual variables are reported with PROM data.

Future Research to Refine Implementation Strategies

As more arthroplasty surgeons and health systems adopt PROMs to guide clinical care and assess quality outcomes, best practices will emerge to guide efficient and effective PROM implementation. This article offers PROM implementation strategies that have effectively supported PROM integration in hundreds of U.S. orthopaedic practices²⁶. While foundational steps to implement PROMs in clinical care are emerging, additional research is needed to refine patient engagement, health-care IT systems, and operational procedures to optimize complete PROM capture both before office visits and at defined post-treatment intervals²⁷. Internationally, hundreds of thousands of patients have demonstrated the ability to complete PROMs to inform quality outcomes. Of note, many countries centralize PROM collection in national registries and do not rely on clinics to collect the data. These countries offer lessons for collecting PROMs directly from patients at fixed intervals, independent of clinic visits.

While broad patient acceptance of PROM completion has been documented, disparities among TJR patients in the U.S. heighten the importance for clinicians to ensure that PROM collection methods engage diverse patients²⁸. Research is needed to clarify whether age, health literacy, or primary language influence the completion of PROMs in order to ensure that all patient data are captured²⁹. Further research should also determine strategies that effectively engage diverse, representative patients in PROM capture in order to evaluate care quality.

Health systems also may evaluate the value of PROM collection and its impact on office resources and clinical decisions. While office-staff time is required to monitor PROM completion, varied IT solutions may assist with capture and reporting. In addition, the value of the PROM data for clinical decisions and quality monitoring must be compared with the resources that are consumed by implementation. The Patient-Centered Outcomes Research Institute has funded dozens of research studies on shared decision-making, and most include patient-reported measures³⁰. Emerging evidence from this research may help refine PROM use and collection protocols for clinical care, as well as support the assessment of PROM value in care.

The CMS and international TJR registries are actively evaluating the use of aggregate PROM feedback for hospitals and health systems as one mechanism to assess quality of care. Historically, the absence of post-TJR complications and readmissions has defined quality of care. Arthroplasty surgeons have the opportunity to be leaders in defining how pain relief and functional gain, as measured by PROMs, can be integrated into quality monitoring. This broader definition of TJR quality can incorporate both health status improvement and the absence of postoperative complications.

Summary

The PROMIS and other PROMs are readily available for integration into arthroplasty clinical practice. Since 2004, >8,000 research papers have reported the use of the PROMIS. While clinical adoption lags behind research, 2 factors are accelerating the use of the PROMIS in clinics: (1) national and international PROM recommendations (namely, the Center for Medicare & Medicaid Innovation [CMMI] bundled payment for TJR and the International Consortium for Health Outcomes Measurement [ICHOM] standard method sets), and (2) EHRs that integrate the PROMIS in the health system. Arthroplasty practices have the opportunity to generate evidence to refine health-system procedures that optimally engage patients in complete PROM reporting and to inform how to interpret PROM data when making individual patient treatment plans and aggregating data for quality comparisons. In the spirit of learning health systems, we encourage orthopaedists to share best practices in office procedures and PROM reporting to accelerate the generation of best practices for PROM implementation and use.

The following content was supplied by the authors as supporting material and has not been copy-edited or verified by JBJS.

Appendix 1: PROMIS Measures for TJR Practice

1. PROMIS Physical Function CAT (4–12 items) or PROMIS Physical Function Short Form 10a (10 items)

2. PROMIS Pain Interference CAT (4–12 items) or PROMIS Pain Interference Short Form 4a (4 items)

3. PROMIS Numeric Rating Scale - Pain Intensity 1a (1 item)

Total assessment time: 1.5 to 4 minutes (9 to 25 items)

Appendix 2: Commonly used Joint-specific Outcome Measures

• HOOS-12 and KOOS-12 capture symptoms in the hip and knee, respectively, but allow the clinician to assess the sub-score domains separately. The measures have 12 items and generate pain, function (or ADL), and quality of life scores that are comparable to the full HOOS and KOOS surveys. [13,14]

• HOOS-JR and KOOS-JR capture symptoms in the hip or knee, respectively, but do not address broader health impact. As with PROMIS Global, the JR scores are aggregate and do not allow the user to assess specific domains of physical function, pain, or fatigue, a concern if using the scores in a clinical encounter [11,15,16].