Geriatric Hip Fracture Care: Fixing a Fragmented SystemMary E Anderson, MD; Kelly McDevitt, RN, MS, ONC; Ethan Cumbler, MD; Heather Bennett, MS, MBA; Perm J 2017;21:16-104 [Full Citation] https://doi.org/10.7812/TPP/16-104E-pub: 04/14/2017ABSTRACTContext: Fragmentation in geriatric hip fracture care is a growing concern because of the aging population. Patients with hip fractures at our institution historically were admitted to multiple different services and units, leading to unnecessary variation in inpatient care. Such inconsistency contributed to delays in surgery, discharge, and functional recovery; hospital-acquired complications; failure to adhere to best practices in osteoporosis management; and poor coordination with outpatient practitioners. INTRODUCTIONHip fractures are a major health burden on the geriatric population in the US. In 2010, an estimated 258,000 people aged 65 years and older were hospitalized with hip fractures.1 Hip fractures can be highly debilitating, resulting in loss of functional independence and a mortality rate approaching 30% at 1 year.2 Direct costs for hip fractures reach $18 billion per year (in 2002 dollars).3 With the aging population, hip fracture rates are projected to rise by 240% in women and 310% in men by 2050.4 Geriatric hip fracture care has historically suffered under fragmented care delivery systems. Patients may be admitted to numerous services and units in a single hospital, leading to unnecessary variation in inpatient care. Such inconsistency is associated with delays in surgery, longer hospitalizations, and higher readmission rates.5,6 More than half of all inhospital deaths among geriatric patients with hip fractures may be preventable.7 Both inpatient and outpatient practitioners often fail to diagnose osteoporosis in these patients and to initiate secondary fracture prevention.8-10 This “osteoporosis care gap” has been attributed in large part to a lack of coordination among orthopedic surgeons, hospitalists, primary care physicians, geriatricians, and other subspecialists.11 Recognition of the fragmentation in geriatric hip fracture care has prompted a call to action by major professional societies, hospital regulatory agencies, insurers, and national and international health organizations.3,12-20 Despite the strong evidence in favor of a population health management-based approach to geriatric hip fracture care, it can be challenging to introduce any change initiative at the local level. A step-by-step framework for successful systems redesign for this patient population is difficult to find. Our aim is to provide a practical how-to guide for implementing a comprehensive geriatric hip fracture program using experiences at our own institution. METHODSSettingThe University of Colorado Hospital is a 620-bed, urban, quaternary-care academic Medical Center in Aurora, CO. It carries a Level II trauma center designation. Geriatric patients with hip fractures have traditionally been admitted to either the Orthopedic Surgery Service or Internal Medicine (IM) services at the discretion of the emergency medicine physician and on-call orthopedic surgeon with older, more medically complex patients directed toward IM care. Patients admitted to the Orthopedic Surgery Service are cared for by orthopedic residents, a physician assistant, and the attending surgeon; patients receive subspecialty consultation or comanagement by the hospitalists on the Medicine Consult Service if requested. Patients older than age 75 years who are admitted to the IM services receive care from 1 of 2 “Acute Care for the Elderly” teams, which are staffed by rotating IM house staff and attending hospitalists, general internists, or geriatricians. Patients under age 75 years go to 1 of 8 general medicine teams under a “drip” admission algorithm; these teams are staffed by rotating IM house staff or advanced practice providers and attending hospitalists, general internists, or subspecialists. Patients can also be admitted to the Family Medicine Service or subspecialty services. Once discharged, geriatric patients with hip fractures are under the immediate care of practitioners at skilled nursing facilities, inpatient rehabilitation facilities, or primary care clinics, either internal or external to our system. Our system alone operates nine primary care clinics: four IM clinics, four Family Medicine clinics, and the Seniors Clinic. All patients have the ability to receive postoperative care in the Orthopedics Clinic. Stepwise Framework for ProgramOur stepwise approach to implementing a comprehensive geriatric hip fracture program is outlined in Figure 1. Step 1: Our institution saw greater than 20% growth in trauma patient volume from 2012 to 2013 because of the expansion of trauma services.21 It specifically witnessed a 40% increase in the number of geriatric patients with hip fractures after the hiring of 2 dedicated orthopedic trauma surgeons, outreach to emergency medical services, and the opening of a larger, state-of-the-art Emergency Department (ED) and second inpatient tower by Spring 2013. Health system leadership also designated the Department of Orthopedic Surgery as one of the priorities for targeted service line growth. Growing market share and the strategic imperative to solidify the organization’s reputation in orthopedic care converged on practitioners’ dissatisfaction with the fragmented state of geriatric hip fracture care to create a “sense of urgency for change,” the critical first step in change management according to Kotter.22 The need to deliver high-quality care and provide efficient throughput would drive our transformation efforts. Step 2: We first assembled a committed clinical leadership team consisting of the Director of Orthopedic Inpatient Medical Services, the hospitalist Director of the Medicine Consult Service, and the clinical nurse manager for the Orthopedics Unit. The Institute for Healthcare Quality, Safety, and Efficiency on campus provided access to a process improvement specialist, data analyst, and project coach, who rounded out our core working group. Other content experts and stakeholders who were engaged during the course of the initiative are listed in the Sidebar: Team Composition and Stakeholders for a Comprehensive Geriatric Hip Fracture Program. Step 3: We defined geriatric patients with hip fractures as those patients aged 65 years and older who sustained an acute hip fracture after minimal trauma such as a fall from standing height. We excluded patients with nonfragility hip fractures (eg, after a motor vehicle accident), nonhip femur fractures, periprosthetic fractures, or pathologic fractures (eg, those caused by malignancy). We identified the geriatric hip fracture population using the following procedural descriptions in our electronic medical record: closed reduction femur fracture/fracture with pinning/hip, intramedullary rodding femur antegrade, open reduction internal fixation femur/femur proximal/hip, pinning percutaneous hip/femur, trochanteric femoral nailing, and hemiarthroplasty hip. These descriptions were more reliable than either Current Procedural Terminology (CPT) or International Classification of Diseases, Ninth Revision (ICD-9) codes. We performed a manual audit after electronic data extraction to validate the data. We could not include the small proportion of patients receiving a total hip arthroplasty because it was prohibitive to manually screen out all the patients undergoing this procedure electively for osteoarthritis. We also could not capture the small proportion of nonoperative patients with this method. However, the population definitions were sufficient to move forward with our quality-improvement initiative. Step 4: We performed a preliminary data analysis to better characterize the current state of geriatric hip fracture care at our institution. In 2013, patients were admitted to 13 different teams on 8 different units. Among the patients on the Orthopedic Surgery Service, those off the Orthopedics Unit had a longer length of stay (LOS) compared with those on the unit (7.5 vs 4.9 days). The rate of osteoporosis evaluation and treatment was low (ie, only 38%, 66%, and 13% of patients, respectively, were discharged on a regimen of calcium, vitamin D, and an antiosteoporosis medication). Only 32% of patients with a primary care provider (PCP) internal to our system completed follow-up within 30 days of discharge. Certain elements of care also appeared to be worsening over time. For example, the average LOS for geriatric patients with hip fractures had progressively risen from 6.1 days in 2012 to 6.5 days in 2013. We also conducted “Voice of the Patient” interviews to gain a better understanding of the patient experience after a hip fracture. Overall, patients provided positive feedback about their interactions with hospital staff, but their responses reinforced the notion that osteoporosis tended to be overlooked in both the inpatient and outpatient settings (eg, “I don’t think we have heard much about bone density testing or osteoporosis” and “The primary care physician did not bring it up.”) The gaps in geriatric hip fracture care at our institution could be best summarized as 1) a lack of a cohesive interprofessional team taking ownership of this patient population along the continuum of care and 2) a lack of standardized, evidence-based care. Step 5: Our clinical leadership team articulated a vision of a comprehensive geriatric hip fracture program, with streamlined care from arrival to postdischarge follow-up. We communicated our vision in as many modalities as possible, including our hospital newsletter, grand rounds for both the Department of Orthopedic Surgery and the Hospital Medicine Group, NICHE (Nurses Improving Care for Health System Elders) Committee meetings, staff meetings on the Orthopedics Unit, informal hallway conversations with frontline practitioners, and one-on-one meetings with executive stakeholders. Using data from our gap analysis, we tailored our message to highlight those aspects of the case for change most motivating to each group of stakeholders. We drew attention to how the fragmented system of care affected both them and patients and encouraged their participation in the solution. Step 6: Given our broad vision for a comprehensive geriatric hip fracture program, we scoped our efforts into manageable steps. Conceptually we divided our program into 5 “buckets” along the continuum of care: ED care, preoperative care, intraoperative care, postoperative care, and postdischarge care. We did not focus on intraoperative care because the on-boarding of the 2 orthopedic trauma surgeons had already resulted in a streamlined on-call schedule, protected operating room time, and improved supply chain management. We decided to first address preoperative and postoperative care. Step 7: Our project charter served as an invaluable organizational tool (see Figure 2). It included the following elements: background, problem statement, aim statement, reporting measures, team members, operational definitions, project timeline, and scope and boundaries of the project. These items are also described in more detail in Steps 1, 2, 3, 6, and 8. In devising our initial project aim statement, we recognized that no single metric would perfectly encapsulate our overarching goal of comprehensive geriatric hip fracture care. We ultimately chose hospital LOS because this information was obtainable, would lend itself to the development of a business case, and would generally reflect coordinated and complication-free perioperative care. We aimed to reduce the average hospital LOS for geriatric patients with hip fractures at the University of Colorado Hospital by 0.5 days by January 1, 2015. Step 8: Metrics for evaluating geriatric hip fracture care have been proposed.3,15,23-25 Our primary outcome measure was hospital LOS, which we defined as time from admission to discharge. Our process measures included time to surgery, which we defined as the time from admission to surgical incision; the percentage of patients admitted to the Orthopedic Surgery Service; the percentage of patients discharged from the Orthopedics Unit; and the percentage of patients in which our new admission order set for geriatric hip fractures was used (see Step 10). Our primary balancing measures were the 30-day all-cause readmission rate to our institution and discharge disposition. We monitored several additional quality metrics: percentage of patients with a basic laboratory evaluation for secondary causes of osteoporosis (complete blood cell count, basic metabolic panel with calcium, hepatic function panel, and 25-hydroxyvitamin D level); percentage of patients discharged on a regimen of calcium, vitamin D, and an antiosteoporosis medication; percentage of patients with appointments scheduled before discharge with their PCP, the Metabolic Bone Clinic, and the Orthopedics Clinic; and percentage of patients with completed follow-up with their internal PCP, Metabolic Bone Clinic, and Orthopedics Clinic within 30 days of discharge. Step 9: Improving the cost of geriatric hip fracture care is integral to enhancing its value. One of the best ways to garner support from executive stakeholders is to demonstrate the business case for a project. We estimated the financial benefit of a LOS reduction using the equation in Figure 3. Cost savings and revenue estimates were based on the approximate contribution margin for medical-surgical hospital admissions used by the Institute for Healthcare Quality, Safety, and Efficiency and assumed the hospital was operating at capacity. The hospital contribution margin was thought to be a conservative estimate for patients with hip fractures. Step 10: Our gap analysis demonstrated the need for a cohesive interprofessional team and standardized, evidence-based care. To achieve our project aim, we developed the following three interventions: 1. Admission of all ward-status patients with hip fractures to the single Orthopedic Surgery Service with hospitalist comanagement, including nonoperative cases: Unlike the IM services at our institution, the Orthopedic Surgery Service was not constrained by age cutoffs, a “drip” admission algorithm, or team caps. It was also staffed by a smaller, more defined cohort of clinicians. The hospitalist-run Medicine Consult Service already had a strong relationship comanaging orthopedic patients and could provide 24/7 in-house coverage for preoperative assessment and subsequent care. We obtained buy-in for revising the admission criteria by emphasizing how the change would save emergency medicine, orthopedic, and IM practitioners the time of negotiating the admitting service on a case-by-case basis. 2. Geographic placement of patients with hip fractures on the Orthopedics Unit: The Orthopedics Unit served as the home unit for the Orthopedic Surgery Service and Medicine Consult Service. It offered nurses and physical and occupational therapists with orthopedic expertise, as well as a dedicated pharmacist, dietitian, social worker, case manager, and patient-resident liaison who scheduled follow-up appointments and arranged for durable medical equipment. It also had established interprofessional morning discharge planning rounds. We worked with hospital managers to assign geriatric patients with hip fractures priority placement on the unit. 3. Standardization of care: With interprofessional input, we designed electronic geriatric hip fracture admission, preoperative, and postoperative order sets (Table 1). The admission order set bundled best practices of geriatric care and contained automatic consults to Physical and Occupational Therapy for early postoperative mobilization and to Social Work/Case Management for early discharge planning. The order sets also standardized inpatient osteoporosis workup and treatment, as well as an automatic referral to the Metabolic Bone Clinic in the Endocrinology Division. This clinic’s role was to serve as a consistent layer of support for osteoporosis care across the diverse primary care practices. In designing the order sets, we adhered to the principles of human-factors engineering. We solicited feedback from the orthopedic residents who would be the end-users. We arranged order set elements in an intuitive order and preselected them when possible, making evidence-based care the default. In the process, we pared down the number of mouse clicks and free-text entries required for admission orders from a cumbersome 227 and 88 to 14 and 1, respectively. We rolled out our geriatric hip fracture program with these three interventions on October 29, 2014, to coincide with World Osteoporosis Day and International Orthopedic Nurses Day. We crafted an intentional communication plan in anticipation of the “go-live,” including an e-mail newsletter and in-person meetings with practitioners, staff, and other stakeholders. In keeping with the change management model by Kotter,22 we planned for and created short-term wins as a show of appreciation and to reinforce adherence to the interventions. For example, we celebrated with the orthopedic residents after they admitted the first geriatric hip fracture patient to the Orthopedic Surgery Service on the Orthopedics Unit using the new order sets. Step 11: Evaluate Outcomes and We monitored the primary outcome metric, process and balancing measures, and additional quality metrics on a monthly basis. We analyzed the data with a Student t-test for continuous variables, Fisher exact test for dichotomous variables, and c2 test for categorical variables, using January 1, 2012, through October 28, 2014, as the preintervention period and October 29, 2014, through March 31, 2016, as the postintervention period. A 2-sided p value of below 0.05 was considered statistically significant. Data analyses were performed using Minitab 17 statistical software (Minitab Inc, State College, PA). Step 12: Review Lessons Learned Recognizing that we could further improve transitions of care, we next focused on postdischarge care. Together with physician representatives from affiliated primary care clinics, we implemented a pathway in August 2015 to improve our 30-day PCP follow-up rates. We formalized the expectation that PCP appointments should be scheduled before discharge. For patients going home, the goal follow-up time was within 7 days of discharge. For patients going to a skilled nursing facility or inpatient rehabilitation facility, the goal follow-up time was 21 to 30 days from discharge, to account for US Centers for Medicare and Medicaid Services (Medicare) coverage rules. We aimed for 2-week follow-up in the Orthopedics Clinic and the Metabolic Bone Clinic. In February 2016, we introduced an electronic discharge order set that reiterated these goals and provided detailed discharge instructions for patients. Ethical ConsiderationsUsing the Colorado Multiple Institutional Review Board framework, this quality-improvement initiative did not meet the definition of human subjects research per US Department of Health and Human Services regulations.
RESULTSWe identified 271 geriatric hip fracture admissions among 267 patients during the study period, including 154 before and 117 after implementation of our program. The 4 patients who sustained bilateral hip fractures during the study period were hospitalized for their first hip fracture before implementation of our program. Patient characteristics before and after implementation were similar (Table 2). Mean hospital LOS significantly improved from 6.4 days before implementation to 5.5 days after implementation (p = 0.004), reversing the prior trend of rising LOS (Table 3). Mean time to surgery trended down from 29.0 to 26.5 hours (p = 0.168). From before to after implementation, the percentage of patients admitted to the Orthopedic Surgery Service increased from 65% to 96% (p < 0.001), the percentage discharged from the Orthopedics Unit increased from 67% to 85% (p < 0.001), and the percentage both on-service and on-unit increased from 55% to 82% (p < 0.001). Adherence to use of the admission order set was 96%. The 30-day all-cause readmission rate to our institution remained stable (3.2% vs 2.7%, p = 0.520). Patients were also just as likely to be discharged home rather than elsewhere before and after program implementation (21% vs 16%, p = 0.244). Results for the additional quality measures are also shown in Table 3. The percentage of patients receiving a basic laboratory evaluation for secondary causes of osteoporosis improved after implementation compared with before the program. Specifically, physician ordering of 25-hydroxyvitamin D levels increased from 68% to 89% (p < 0.001). The percentage of patients before and after implementation discharged on a regimen of calcium (55% vs 99%), vitamin D (70% vs 96%), and an antiosteoporosis medication (24% vs 85%) also improved significantly (p < 0.001 for all). Patients were more likely to have a follow-up appointment scheduled before discharge with their PCP (15% vs 28%, p = 0.006), the Metabolic Bone Clinic (3% vs 53%, p < 0.001), and the Orthopedics Clinic (82% vs 93%, p = 0.005). Completed follow-up within 30 days also improved for the Metabolic Bone Clinic (3% vs 28%, p < 0.001). DISCUSSIONOur comprehensive geriatric hip fracture program represents important progress in fixing the long-standing fragmentation in care at our institution. Through our stepwise framework for systems redesign, we have sustained an approximately 1-day reduction in LOS over 17 months, exceeding our aim of 0.5 days, and made significant gains in other quality metrics. Our findings compare favorably with national outcomes reported for geriatric patients with hip fractures in the US. According to Medicare data, the mean hospital LOS for geriatric patients with hip fractures in this country is 6.1 days.2 Our geriatric hip fracture program reduced LOS for our population from above to below the national average. We believe that we achieved more coordinated and proactive care by consolidating patients under a cohesive interprofessional team and standardizing care expectations. Given that mean time to surgery did not significantly decrease, most of the LOS improvements occurred postoperatively. It is unlikely that the LOS reduction resulted from premature discharges or from prioritization of throughput over functional outcomes. Although it is difficult to directly compare our 30-day all-cause readmission rate with the national average of 12.4% because we were able to track only readmissions to our own institution, there was no difference in readmissions from before to after implementation.2 We also preserved patients’ ability to return to independent living at discharge. More than 16% of our patients were discharged directly to home with or without home health care services, compared with 7.8% nationally.2 Other studies have found that a joint model of care with shared responsibility improves outcomes for geriatric patients with hip fractures. Integrated orthogeriatric comanagement models of care are now strongly recommended given their potential to improve LOS, time to surgery, perioperative complications, functional outcomes, short- and long-term mortality, and costs.26-42 Successful orthogeriatric comanagement models employing hospitalists or general internists, rather than geriatricians, have also been reported, with similar LOS reductions between 2.3 and 2.7 days.5,43 We opted for hospitalist comanagement for practical reasons, because hospitalists have a more robust inpatient presence at our institution. Other studies have also evaluated strategies for improving osteoporosis care. Coordinator-based fracture liaison services are widely promoted as an effective means of improving adherence to best practices in osteoporosis management.4,12,44-50 Such services have been shown to improve the rates of postfracture assessment and treatment from approximately 20% to 80% or 90%.12 We had no funding to hire a fracture liaison service coordinator at the outset of our initiative; however, we were able to achieve similar rates of osteoporosis evaluation and treatment with our program. Implementing a fracture liaison service may be a reasonable next step to further improve our transitions of care with PCPs and our Metabolic Bone Clinic. A more in-depth look at our referral process and access to outpatient care, as well as a process for scheduling all follow-up appointments for the same postdischarge day, may also improve patients’ ability to complete follow-up within 30 days of discharge. This quality-improvement initiative has several strengths. Our application of strategies from disciplines outside medicine, including change management and human-factors engineering, greatly facilitated our ability to lead and sustain change. The high level of adherence to our three interventions speaks to the buy-in of the orthopedic surgeons in particular, because they were responsible for accepting geriatric hip fracture admissions from the ED and for using the admission order sets; anecdotally, the emergency medicine physicians and other inpatient practitioners and staff also appreciated having a clear, standardized workflow. Our “population health management” approach aligns well with the Institute for Healthcare Improvement’s Triple Aim.51 In addition to improving the quality of care processes, our interventions also improved efficiency, thereby reducing the cost of geriatric hip fracture care. Using our actual LOS reduction of 0.9 days for 117 patients, we conservatively estimated the cost savings for our hospital to date at approximately $190,000. Our choice of interventions, including the electronic order sets, also positions us to be able to disseminate our initiative to other hospitals in our system. This quality-improvement initiative has several limitations. First, we did not adjust results for comorbid conditions or severity of illness. Second, given our multiple concurrent interventions, it was not possible to assess the impact of any one intervention in isolation. Third, we have not yet been able to formally evaluate patient or practitioner satisfaction or long-term outcomes such as functional status, secondary fracture rate, or one-year mortality. Our institution had no preexisting infrastructure for tracking these metrics in our patient population. Fourth, our initial interventions did not address the ED phase of care. We are currently working to develop an ED care pathway for geriatric patients with hip fractures. This pathway will incorporate fascia iliaca compartment blocks, with the goal to improve pain scores and the overall opioid burden for patients. Finally, this was a single-center initiative. Although this limits generalization to other institutions, our stepwise approach may serve as a guide for others interested in implementing comprehensive geriatric hip fracture care. CONCLUSIONOur comprehensive geriatric hip fracture program improved hospital LOS, osteoporosis care, and outpatient follow-up with potentially substantial cost savings by targeting the fragmented system of care at our institution. Our experience offers important insights for the larger US health care system as it attempts to meet the needs of a rapidly aging population. Disclosure StatementThe author(s) have no conflicts of interest to disclose. AcknowledgementsThe authors thank Jeffrey Glasheen, MD; Read Pierce, MD; and the rest of the faculty of the Institute for Healthcare Quality, Safety, and Efficiency for their instruction and support. They also thank Micol Rothman, JoAnn Young Marrs, and all the 8W practitioners and staff for their contributions to the University of Colorado Hospital Geriatric Hip Fracture Program. A modified abstract was presented in poster format at the Society of Hospital Medicine 2015 Annual Meeting on March 30, 2015, in National Harbor, MD. A modified abstract was also presented in storyboard format at the Institute for Healthcare Improvement’s 27th Annual National Forum on December 8, 2015, in Orlando, FL, where it was 1 of 20 winners of The Permanente Journal Service Quality Award. Kathleen Louden, ELS, of Louden Health Communications provided editorial assistance. How to Cite this ArticleAnderson ME, McDevitt K, Cumbler E, et al. Geriatric hip fracture care: Fixing a fragmented system. Perm J 2017;21:16-104. DOI: https://doi.org/10.1278/TPP/16-104. References1. Stevens JA, Rudd RA. The impact of decreasing U.S. hip fracture rates on future hip fracture estimates. Osteoporos Int 2013 Oct;24(10):2725-8. DOI: https://doi.org/10.1007/s00198-013-2375-9.
|
ETOC
Click here to join the eTOC list or text ETOC to 22828. You will receive an email notice with the Table of Contents of The Permanente Journal.
CIRCULATION
2 million page views of TPJ articles in PubMed from a broad international readership.
Indexing
Indexed in MEDLINE, PubMed Central, EMBASE, EBSCO Academic Search Complete, CrossRef, and SciVerse/Scopus.