Introduction

There is growing national interest in improving the value of total joint arthroplasties (TJA) of the hip and knee [6, 28]. Nearly one million TJAs are performed annually [21], and the frequency of these procedures increases by 10% to 15% per year [12, 20]. Despite the high volume of TJAs, there is substantial variation across facilities in adherence to evidence-based care processes, operative times, length of stay, discharge disposition, complication rates, patient-reported outcomes, and episode costs [9, 15, 19, 26, 30, 31].

Several tools have been developed to standardize care delivery and support modeling of care for different patient groups. For instance, evidence-based clinical practice guidelines produced by the American Academy of Orthopaedic Surgeons focus on improving TJA safety, including venous thromboembolism (VTE) prophylaxis and surgical site infection prevention [1–3, 17]. TJA clinical practice guidelines produced in Canada and the United Kingdom focus on care from initial assessment through outpatient rehabilitation and followup [5, 23]. A set of quality indicators has been developed to guide surgical practice [29], numerous care pathways target inpatient care [4], and some institution-specific pathways target the care continuum. For instance, the Dartmouth “GreenCare” model [11] implemented in 2011 addresses care from initial surgeon referral through 1 year after surgery, with attention to role-task alignment of providers involved in delivering care, measurement of compliance with clinical evidence for each patient, incorporation of patient-reported outcomes into clinical decisions, use of formal shared decision-making, and per appointment and per case cost reduction.

Despite development of best practice guidelines, there has been limited attention to developing guidelines that consider patient-centered care processes [27]; lean consumption (eg, redesigning care processes to meet consumers’ demands without wasting time, effort, or resources) [35]; or processes to improve communication across settings [34]. As value-based payment reforms emphasize the shift from volume and intensity of services toward high-value, patient-centered care [6], care pathways that can consistently guide reliable delivery of safe, effective, efficient, and patient-centered care are required.

Recognizing this need, our study expands on prior pathway development efforts (including Dartmouth’s development of GreenCare) with the purposes of (1) developing a generalizable clinical care pathway for primary TJA using inputs from clinical, academic, and patient stakeholders; and (2) identifying system- and patient-level processes that may provide safe, effective, efficient, and patient-centered care for patients undergoing TJA.

Materials and Methods

We used a combination of quantitative and qualitative methods [10] to develop a care pathway for elective TJA that begins at the presurgical office visit and continues through the first year of postoperative followup. This approach included (1) identification of high-performing hospitals using available data and author opinion; (2) semistructured interviews with interdisciplinary care teams from these hospitals and independently selected patients undergoing TJA; (3) drafting the care pathway; and (4) consolidation by a multistakeholder panel (Fig. 1).

Fig. 1
figure 1

The process used to develop the care pathway is shown. SCIP = Surgical Care Improvement Project.

Full size image

Population Studied

We used the Premier Healthcare Alliance quality improvement database (n = 234 hospital members) to identify 16 high-performing hospitals with at least 150 primary THA discharges and 300 primary TKA discharges during a 2-year period. We recruited 10 of 16 selected hospitals, including six of eight teaching hospitals and four of eight nonteaching hospitals. We were unable to contact four hospitals and two declined, stating lack of interest. To identify high-performing hospitals, we calculated standardized z-scores for (1) 30-day readmission rates and (2) inpatient costs for patients discharged with a primary TKA or THA between October 1, 2009, and September 30, 2011; and (3) hospital-level surgical care improvement project measures available through the Centers for Medicare & Medicaid Services’ (CMS) Hospital Compare (Surgical Care Improvement Project [SCIP] VTE metrics, hip replacement score, and knee replacement score; collected April 1, 2010, to March 31, 2011). We ranked the top 25 teaching and 25 nonteaching hospitals in the database by adding z-scores across the three dimensions, with double weighting of cost. Hospitals were chosen to prioritize the highest overall rank (≤ 10 in teaching/nonteaching categories), minimize inclusion of multiple hospitals from the same healthcare system, and allow inclusion of two high-volume hospitals (rank, 20, 22) and two hospitals from geographically unrepresented areas (rank, 12, 16). Participating hospitals had similar characteristics, but slightly higher performance than those that could not be contacted or declined (mean z-score: participants, 2.8; nonparticipants, 1.8).

To expand our sample beyond the 234 hospitals in the Premier database and identify other potential best practices, three of us (KJB, KMK, AMD) identified 14 hospitals that were nationally recognized as high performers in evidence-based care and patient- and family-centered care. Two hospitals overlapped with hospitals identified from the Premier database; one accepted the invitation and is included in the above sample. We successfully recruited six of 12 remaining hospitals. We were unable to contact two hospitals, three declined stating lack of interest, and one could not be scheduled within the timeframe. Data were not available to compare hospitals that participated with those that declined.

Organizations were diverse in geographic region, teaching status, quality recognition, number of inpatient beds, and annual surgical volume (Table 1). Hospitals selected using the Premier database and author opinion were compared with median performance benchmarks from Premier and CMS comparators; all hospitals exceeded benchmarks in at least 50% of reported metrics (Fig. 2). Owing to differing data sources (database versus self-report), it is difficult to ascertain true differences between hospitals from the two selection methods.

Table 1 Characteristics of hospitals in the sample
Full size table
Fig. 2A–D
figure 2

The distributions of performance among participating hospitals relative to the median performance (green line) among members of the Premier Healthcare Alliance database are shown for (A) inpatient cost, (B) 30-day readmission rate, and (C) length of stay and for (D) CMS Hospital Compare database (SCIP metrics). < A = median performance of hospitals selected from the Premier database; < B = median performance of hospitals selected using author opinion. a = source was 234-Premier Healthcare Alliance member hospitals in the quality improvement database with greater than 150 THAs and greater than 300 TKAs in a 2-year period (October 1, 2009, to September 30, 2011); b = CMS Hospital Compare Surgical Care Improvement Project (SCIP) data, from January 1, 2011, to December 31, 2011; c = data for all surgical types, not specific to TKA and THA; VTE = venous thromboembolism; VTE-1 is a measure of surgery for patients with recommended venous thromboembolism prophylaxis ordered. VTE-2 is a measure of surgery for patients who received appropriate venous thromboembolism prophylaxis within 24 hours before surgery to 24 hours after surgery.

Full size image

Semistructured Interviews

We conducted semistructured telephone interviews with members of interdisciplinary care teams from selected hospitals and with independently selected patients between March and September 2012. Interdisciplinary care team interviews included nurses from the surgical practice, operating room, postanesthesia care unit, inpatient unit, and home health settings (n = 15); quality improvement personnel (n = 11); midlevel leaders (n = 9); surgeons (n = 9); TJA program coordinators (n = 7); physical therapists (n = 5); care managers (n = 3); senior-level leaders (n = 3); an anesthesiologist (n = 1); and a pharmacist (n = 1). Interviews included one to 13 team members per hospital (median, 3). Patient interviews included one male and one female with a TKA in the previous 2 years. Patients were identified by two of us (AVC, BO) and did not receive care from an interviewed hospital.

We selected interview topics to identify care processes that may contribute to safe, effective, efficient, and patient-centered care. Interdisciplinary care team discussions included (1) typical care experience for a patient and their family; (2) greatest program successes; (3) factors that lead to success; (4) strategy for measuring and tracking care processes and outcomes; and (5) plans to improve care and efficiency. Average duration of interviews was 58 minutes (SD, 15; range, 35–109 minutes). Patient-level discussions were designed to validate concepts identified by care teams and included pleasing and disappointing features of care; factors that contributed to safety, efficiency, or patient and family experience; and advice for providers. Patient interviews lasted 58 and 80 minutes.

We recorded interviews, summarized site attributes, and completed a site checklist to document processes that contributed to a safe, effective, efficient, and patient- and family-centered care experience. We also collected relevant documentation from sites (eg, process and outcome data, TJA care protocols and pathways, and education materials). For each site and patient interview, we mapped flow of care between patients and an interdisciplinary team of providers.

Care Pathway Development

Potential best practices were distilled into a care pathway that starts when a patient has decided to have surgery and has a presurgical visit with their surgeon and ends 12 months after surgery. The care pathway is organized into four care periods: (1) preoperative surgical visit; (2) preoperative preparation and planning for surgery; (3) hospital admission for surgery through hospital discharge; and (4) postdischarge care.

The care pathway highlights processes that apply across the care continuum and in each care period. For each period, it suggests: (1) processes for providing safe, effective, efficient, and patient- and family-centered care; techniques to reduce waste; and techniques to improve communication. It includes system-level processes that apply to how the system of care is designed and patient-level steps that may be applied to most patients; (2) process and outcome measures to monitor; (3) a description of how system-level and patient-level suggestions are mapped to flow of care and provider responsibilities; and (4) additional resources, including links to how-to guides, clinical practice guidelines, meta-analyses, and selected scientific literature.

Multistakeholder Panel Review

After developing a draft care pathway, we convened a 32-member multistakeholder panel to participate in a 1-day workshop focused on reviewing and refining the care pathway. The workshop was sponsored by the Chief Medical Officer of Premier Healthcare Solutions, Inc (RAB); Chief Medical and Scientific Officer of the Institute for Healthcare Improvement (DAG); Vice President of the American Association for Hip and Knee Surgeons (JRL); and Chair of the Council on Research and Quality of the American Academy of Orthopaedic Surgeons (KJB). Sponsors defined meeting objectives, deliverables, and boundaries; selected and recruited participants; and set meeting expectations. Participants were selected based on leadership positions in orthopaedic (n = 4) or anesthesia (n = 1) specialty societies; involvement in organizations specializing in safety and high reliability care (n = 3), lean production/consumption of care (n = 3), and patient experience of care (n=3); membership in an interdisciplinary care team of a hospital selected for interviewing (surgeons, nurses, TJA program directors, and a physical therapist; n = 8); recent receipt of a TJA (n = 1); and participation in the pathway development team (n = 9).

The facilitated workshop included small group breakout sessions with report-outs and opportunities for feedback and consensus building. Sessions focused on gaining consensus on high-level processes and flow; identifying a starter set of metrics to track clinical outcomes, patient experience, cost, and pathway adherence; and developing a demonstration plan to test the impact and feasibility of implementing the care pathway.

The multistakeholder panel input resulted in four classes of revisions to the care pathway. First, language was changed to emphasize general care principles instead of specific care practices to improve generalizability across settings. For example, a recommendation that patients participate in a preoperative education “class” was changed to reflect participation in a preoperative education “process (eg, books, online, video, didactic, class)”. Second, several recommendations were advanced to earlier care periods. For instance, the process of identifying, evaluating, and developing a plan to mitigate surgical risk factors was advanced from the preoperative testing period to the presurgical office visit. Third, recommendations were added to improve specificity in areas that had been insufficiently addressed in semistructured interviews (eg, anesthesiology, postdischarge care). Finally, processes to improve patient engagement and communication were strengthened.

Final Revision

Interview participants, multistakeholder panel members, and three individuals who were unable to attend the multistakeholder panel reviewed the revised care pathway. After further revisions, the proposed care pathway was finalized and made publicly available for testing and context-specific modification [25].

Results

The care pathway suggests 40 processes to improve care, 37 techniques to avoid waste, and 55 techniques to improve communication (total suggestions, 132; Fig. 3; Appendix 1). Overall, 43% (n = 57) of suggestions are aimed at processes that apply to how the system of care is designed, and 57% (n = 75) are steps that may be applied to most patients. A subset (n = 17) of these suggestions applies across all care periods (Table 2), whereas the remainder (n = 115) apply to discrete periods of care (Table 3). The care pathway emphasizes three central themes: standardization and process improvement, interdisciplinary communication and collaboration, and patient and family engagement and education.

Fig. 3
figure 3

The distribution of 132 system- and patient-level suggestions across discrete care periods is shown.

Full size image
Table 2 Seventeen system- and patient-level suggestions* that apply across all care periods
Full size table
Table 3 Fifteen sample suggestions from discrete care pathway periods (among 132 total)
Full size table

Standardization and Process Improvement

Among the 17 suggestions that apply across all care periods, six focus on improving standardization and process improvement. These include standardizing care protocols and staff roles; aligning information flow with patient and process flow; following a risk identification, evaluation, and mitigation process to stratify patients into appropriate care levels; and using registries and electronic systems to track patient outcomes and improve quality. For example, an organization might standardize care across all providers and routinely update protocols to reflect best practices.

Communication and Collaboration

Five of the 17 suggestions that apply across all care periods focus on improving communication and collaboration. These include identifying a role accountable for care delivery and communication; establishing financial arrangements between hospitals and physicians that encourage high-value care; using checklists and scripts to manage communication and care transitions; and using an electronic health record or web portal to facilitate critical element communication. For example, organizations may use electronic transmission of information to enable interdisciplinary communication in and across care settings.

Patient and Family Engagement and Education

The final subset (n = 6) of the 17 suggestions that apply across all care periods focuses on improving patient and family engagement and education. These suggestions include using appropriate health literacy levels and culturally sensitive communication; managing patient and provider expectations for care and recovery, standardizing who delivers information to patients and what information is conveyed, documenting and communicating the patient’s goals for TJA in a care plan that follows the patient across the care continuum, and engaging patients and their families in value-based discussions of care options. For example, an organization might establish a program to engage informal caregivers as active members of the care team.

Discussion

Nearly one million TJAs are performed each year in the United States [21] and substantial variation exists across facilities with respect to quality and efficiency [9, 15, 19, 26, 30, 31]. As payment reforms emphasize high-value, patient-centered care [6], guidelines and care pathways are needed that can consistently guide reliable delivery of safe, effective, efficient, and patient-centered care. The proposed care pathway endeavors to balance safety and effectiveness with dimensions of care not addressed in most other care pathways, such as lean consumption [35], patient-centered care [27], and communication and coordination of care [32, 33]. It addresses care during a period of approximately 14 months (from the presurgical office visit through 12 months after discharge), addresses care delivered by a cross section of providers in multiple settings, and includes 132 suggestions for providing high-value care for primary TJA.

There are several limitations to this care pathway. First, the overall care pathway represents a proposal that has not been tested in a clinical setting. Although some suggestions are evidence-based practices, others may not be supported by an evidence base. This compilation of potential best practices has not been validated as effective in its full form or implemented by any one group. It is unknown whether the pathway can be fully implemented, will improve the quality and value of care, or will generalize to diverse care settings. We have attempted to offset these limitations by surveying different kinds of hospitals, creating a diverse multistakeholder panel, and providing reference to a select set of evidence-based practices. We placed recommendations in the context of established principles of the science of improvement [13, 14, 22]; however, organizations may require coaching to prioritize and implement suggestions. Verbal scripts, checklists, standard order sets, and other materials for patients, providers, and administrators may strengthen the ability of an organization to operationalize and implement suggestions. Moreover, we recognize that implementing a large number of suggestions will be challenging and therefore have structured the pathway to include discrete improvement categories and care periods that allow providers to implement suggestions that correspond to their priorities without undertaking full implementation of the pathway. To identify implementation priorities, an organization may find it valuable to complete a matrix that couples care pathway suggestions with expected ease and cost of implementation; impact on clinical, safety, satisfaction, and cost metrics; and alignment with strategic objectives. Finally, we have not evaluated cost of implementation. Although our process for selecting hospitals targeted those that had achieved high-quality care at a low cost, testing is needed to determine the cost associated with implementing these suggestions.

There are also limitations to the methods used to develop the pathway. First, although we used several mechanisms to elicit input from patients and patient advocates (eg, interviews, multistakeholder panel, and ad hoc advisory discussions), there would be value in having greater participation of patients who had recently undergone TJA (we included three patients) and in establishing a formal advisory role of one or more patients throughout the project. Second, identification of high-performing organizations was limited by available data. There are no national databases available to monitor quality and cost associated with TJA programs over the episode of care. As such, we used author opinion to supplement available data from the Premier and CMS Hospital Compare databases. Finally, we encountered some resistance from hospitals regarding sharing proprietary information, however most organizations welcomed the opportunity to contribute to a pathway that would be freely available to the orthopaedic community.

Common themes in the care pathway align with recommendations found in the literature [4, 7, 8, 13, 14, 16, 18, 24, 27, 33] and include standardization and process improvement, communication and collaboration, and patient engagement and education. Research has shown that standardization and use of established process improvement methods can improve clinical outcomes, safety, and efficiency of TJA [4, 7, 8, 13, 14, 16, 33]. For instance, process standardization and adherence to evidence-based guidelines can shorten lengths of hospital stay, improve clinical outcomes, and reduce negative outcomes (including death, readmission, reoperation, or surgical complications) independent of hospital or surgeon procedure volume [7]. Communication across members of interdisciplinary care teams has been recognized as a critical element for successful care transformation, yet this area has been largely neglected in care pathways [33]. Finally, patient-centeredness is increasingly recognized as a necessary attribute of healthcare quality [18, 27], and patient and family engagement can lead to improved TJA clinical outcomes [13, 14, 24].

TJA processes are evolving, and progressive health systems are actively testing initiatives to improve delivery of high-value care. We used a multistakeholder approach to develop a TJA care pathway that outlines suggestions that might improve care. Care pathway suggestions are designed to be transferable to diverse settings, and suggestions are specifically broad to accommodate local characteristics, culture, and resource availability. The pathway we proposed should be evaluated in high- and low-volume settings to determine its effectiveness, feasibility, cost of implementation, and need for context-specific adaptation. To ensure high-value services across the care continuum, TJA programs should endeavor to standardize care processes and may draw on suggestions identified herein.