Successes and failures in improving osteoporosis care after fragility fracture: Results of a multiple-site clinical improvement project


  • J. Timothy Harrington,

    Corresponding author
    1. University of Wisconsin School of Medicine and Public Health, Madison
    • University of Wisconsin School of Medicine and Public Health, One South Park Street, Madison, WI 53715
    Search for more papers by this author
    • Dr. Harrington has received consulting fees (more than $10,000 per year) from the Alliance for Better Bone Health, and is the inventor of Stop Osteoporosis software (copyrighted by the Wisconsin Alumni Research Foundation). Dr. Deal has received consulting fees (more than $10,000 per year each) from Eli Lilly and GlaxoSmithKline/Roche.

  • Chad L. Deal

    1. Cleveland Clinic Foundation, Cleveland, Ohio
    Search for more papers by this author
    • Dr. Harrington has received consulting fees (more than $10,000 per year) from the Alliance for Better Bone Health, and is the inventor of Stop Osteoporosis software (copyrighted by the Wisconsin Alumni Research Foundation). Dr. Deal has received consulting fees (more than $10,000 per year each) from Eli Lilly and GlaxoSmithKline/Roche.



To improve osteoporosis diagnosis and treatment of fragility fracture patient populations because osteoporosis care is provided infrequently to those patients, leaving them vulnerable to further fractures and increasing debility.


Osteoporosis experts from 11 US health systems participated in a clinical improvement project based on previously described successful osteoporosis care process redesigns. Participants were taught rapid cycle process improvement methods that are widely used in clinical improvement projects, and were supported in their efforts by the program coordinator. Measures of successful process development included establishing reliable referral from orthopedic fracture care to osteoporosis diagnosis and treatment, nurse coordination and monitoring of osteoporosis care, and use of process management software for registering patients and organizing work.


Four sites were able to establish these critical referral and osteoporosis management processes. Two sites were partially successful in increasing orthopedic referrals to consultative care, but otherwise continued traditional care processes. Five were unsuccessful due to inability to implement 1 or more of these key process improvements.


Reliable osteoporosis care for fracture patients is possible if traditional practice processes are replaced with more effective, well-recognized approaches to chronic disease management.


It is widely recognized that patients with a fragility fracture have a high risk of further fractures (1–4), but many publications have documented that few of these patients are being evaluated or treated for osteoporosis, either before or after they fracture (5–10). Three reports of more reliable postfracture osteoporosis care have now been published, 2 from Europe and 1 from the US (11–13). This information supports the recently published opinion of Andrade et al that “To close this care gap and improve the quality of care for these patients … innovative, multifaceted interventions need to be developed to address the barriers related to system, physician, and patient that stand in the way of best practice” (7).

Health systems reporting successful osteoporosis care have implemented very similar, nontraditional delivery processes and population-based care principles to more reliably link fracture patients to osteoporosis diagnosis and treatment, and to sustain this treatment over time. The key factors for achieving these breakthrough improvements have included one or more strongly committed physician osteoporosis champions; cooperation among orthopedic surgeons, primary physicians, and osteoporosis experts; a more dependable process than visit-related referrals for connecting fracture patients to a subspecialty-based osteoporosis care service; a patient population registry; nurse management to coordinate and monitor care and to provide patient education; and system or practice willingness to provide financial support for required resources.

This article summarizes a multiyear, multiple-site effort to implement these process changes in additional US health care environments in order to provide effective osteoporosis care for fragility fracture patients and further verify the impact of clinical process redesign.


Champions and health systems.

Osteoporosis champions representing several different medical specialties and a variety of health care environments agreed to participate in this improvement project (Table 1). The majority of champions were members of the Alliance for Better Bone Health Regional Osteoporosis Boards, an industry-sponsored consulting organization. Others identified their interest in participation directly to the project coordinator (JTH). The participants included in this study made a sustained effort to implement an improvement process within their practices and/or health systems and agreed to use and evaluate the process management software program developed by the project coordinator.

Table 1. Summary of the improvement projects*
OutcomePractice typeChampion specialty
  • *

    Successful = developing a dependable referral process from orthopedic fracture care for central bone density measurement, providing effective care for willing and able patients, and using the software program to register patients and facilitate care; Partially successful = increasing referrals from orthopedic surgery to osteoporosis consultation with other aspects of care remaining traditional; Unsuccessful = unable to achieve improvement.

SuccessfulAcademic health systemRheumatology
SuccessfulAcademic health systemRheumatology
SuccessfulCommunity practiceOrthopedic surgery/physical medicine and rehabilitation
SuccessfulCommunity practiceOrthopedic surgery
Partially successfulAcademic health systemEndocrinology
Partially successfulAcademic health systemEndocrinology
UnsuccessfulMultispecialty group practiceEndocrinology
UnsuccessfulMultispecialty group practiceRheumatology
UnsuccessfulAcademic health systemEndocrinology
UnsuccessfulMultispecialty group practiceInternal medicine
UnsuccessfulMultispecialty group practiceGeriatrics

Osteoporosis management process improvements.

The process changes needed to more reliably link referral from orthopedic fracture care to osteoporosis care and to sustain effective treatment over time have been previously described (11–13) and are outlined above.

Process management software.

Software programs are used to manage work and monitor outcomes in many industries, but are seldom used in health care (14, 15). Computer-based patient population registries and task management programs based on care algorithms are required for effective chronic disease management over time. The software program for managing osteoporosis care has been developed and used by the project coordinator as previously described (13). The process algorithm used to create this software is shown in Figure 1.

Figure 1.

Algorithm for managing osteoporosis after fragility fracture. DXA = dual energy x-ray absorptiometry; BMD = bone mineral density; Vit D = vitamin D.

Champion and site preparation.

Each champion was provided with information about how to perform a successful clinical improvement project based on Institute for Healthcare Improvement methods (16) and how to install and use the management software. In most cases the project coordinator visited the health system and provided a seminar to encourage orthopedic, primary physician, and management support. The nurses or others identified to manage the care process and use the software were provided separate information and phone support by the University of Wisconsin (UW) osteoporosis nurse manager. As the UW pilot program progressed, process redesigns were shared with other participants. This project represents clinical improvement rather than clinical research (17), and the UW Health project has been exempted by the UW Institutional Review Board.

Monitoring and assessing improvement experiences.

The project has been active from February 2002 to the present. Champions have been queried periodically, problems have been discussed and noted as they have arisen, and exit interviews have been performed with unsuccessful champions to determine why they were unable to establish a successful program. Measures of success include: developing a dependable referral process for the fracture population from orthopedic fracture care for central bone density measurement by dual energy x-ray absorptiometry (DXA), providing effective care for willing and able patients over time based on nurse management principles, and using the software program to register patients and facilitate care.


Eleven medical practices including the UW Health pilot site participated in this project. The types of practices and champion specialties are summarized in Table 1, divided into those that have been successful, partially successful, or unsuccessful, based on the 3 criteria indicated above. Four systems have been successful in developing a reliable referral process from fracture to osteoporosis care, establishing an osteoporosis management program for a fragility fracture patient population, and using the software registry and task management features. In some cases, the population consisted of hospitalized fracture patients, and others included a broader population of fragility fracture patients identified from orthopedic practice billing data. Success was not associated with any one type of practice or champion specialty, although the success of 2 orthopedic practices is of interest. In all cases, gaining a commitment from involved physicians and support for a program coordinator, implementing new processes, and learning to use the software program took considerable time, commonly up to 2 years.

Two partially successful practices achieved improved referrals for fracture inpatients through cooperation between orthopedic surgery and endocrinology; however, the numbers of patients managed were relatively small, the overall fracture population was not defined, care processes were otherwise traditional, barriers to more significant improvement were encountered, and the software was not used to develop a patient registry and population-based care.

The other 5 champions were unable to create a successful program for one or more of the following reasons: 5 sites (including 1 partially successful site) were unable to gain support for a nurse manager; 1 site had a champion who left the practice; 1 site had an orthopedic surgery group unwilling to participate; and 1 site had a practice administration that was unwilling to install the software. Champions were not lacking in resolve; however, lack of cooperation from referring orthopedic surgeons, inability to coordinate care across specialties and/or the inpatient/outpatient interface, and the lack of support for a coordinator led each project to founder. Software use was the limiting factor in only 1 practice where information systems regulations did not permit its use.


Our experiences in this osteoporosis clinical improvement project illustrate the widely recognized barriers to effective care of chronic diseases, including traditional health care processes, loosely organized health systems, provider resistance to change, and fragmented health care financing (18, 19). Our successful champions had to overcome these barriers to establish an effective program. This required work, innovative approaches, persistence, and time, usually years, as explained more fully in a previous article (13).

Other champions were not able to succeed in spite of substantial efforts. The most common barrier was the unwillingness of practice leadership at 5 sites (including 1 partially successful site) to support a nurse coordinator. Without this person, several functions critical to successful management of chronic diseases cannot be provided reliably, including coordination and monitoring of care and patient education (20–26).

The success of 2 orthopedic groups and failure of 1 other site that lacked an orthopedic interaction suggest the importance of a commitment from orthopedic surgeons and the possible advantage of fracture and osteoporosis care within the same practice. Fortunately, orthopedic leaders have been among those advocating for a commitment from their specialty to arranging osteoporosis care for fracture patients (6, 27).

Our experiences also suggest an advantage of integrated multispecialty practice entities over health care environments in which fracture care, consultative and primary care services, and DXA are provided by different or multiple practices. Independent practice environments should, however, be able to develop successful programs by identifying fracture patients from orthopedic practice billing data and providing this information with a consultation request to an osteoporosis care service. Many physicians are also unaware that the Health Insurance Portability and Accountability Act regulations allow sharing of such data between practice entities for the purpose of improving patient care (28). We believe it is not the independent practice structure per se that prevents coordinated care across specialties, but the more common competitive and financial barriers to integrating care in such environments.

For process management software and other clinical decision support systems to be useful, health care providers and health system leadership must first embrace reliable referral and population management programs (29). Once these commitments are achieved, process management software becomes a requirement, because traditional paper records and even electronic prompt systems are not able to support such programs dependably (7, 8, 12, 13, 30, 31).

Lesser process changes are unlikely to resolve the deficiency of osteoporosis care after fracture. Another US based 3-system improvement project has reported more limited successes while describing the barriers to improvement that were encountered (8, 32). This effort was similar in the limited extent of its process redesigns to our partially successful sites. Two Canadian systems have studied the effects of a postfracture reminder letter program, and reported higher DXA utilization compared with traditional care, but lesser improvements in antiresorptive treatment (33, 34). In contrast, ours and other more successful improvement projects, after experiencing similar initial failure of physician prompt programs, then gained greater success by implementing more fundamental process changes.

A momentum is building to redesign the US health system, and the greatest need exists in the care of chronic diseases, including osteoporosis (18). These examples suggest that it should be possible to improve both the effectiveness of care and the impact of this disease on our society. Projects such as ours that focus on process improvements and report process-based measures of success should be viewed as essential steps toward longer-term improvements in disease outcomes, such as fracture reduction. Without the former, the latter will not happen. We hope our experiences will encourage other osteoporosis specialists to advocate for and implement clinical process redesign, to learn the methods critical to this work (16), to build more effective interdisciplinary relationships with their orthopedic and primary care colleagues, and to teach these approaches to future physicians.


The project champions appreciate the excellent work of our successful program coordinators Lynn M. Heindl, MS, RN, Debora Bork, MFA, Louise Ricker, RN, JoEllen Lease, RN, and Sherri Day, RN. Participating practices and champions were Aspen Orthopedic and Rehabilitation Specialists (Milwaukee, WI) Susan M. Larson, MD, Jeffrey E. Larson, MD, Lee M. Tyne, MD; Aspen Medical (St. Paul, MN) Michael Gonzalez-Campoy, MD, PhD (current address: Minnesota Center for Obesity, Metabolism, and Endocrinology, Eagan, MN); Central Maine Orthopaedics (Auburn, ME) A. Ingrid Erickson, MD; Cleveland Clinic Foundation (Cleveland, OH) Chad L. Deal, MD; Columbia University College of Physicians and Surgeons, New York (NY) Carolyn Becker, MD; Franciscan Health System (Tulsa, OK) Kenneth Piper, MD; Geisinger Medical Center (Danville, PA) Eric D. Newman, MD; Richland Medical Center (Richland Center, WI) Robert P. Smith, MD, Philomena Poole, RN, GNP; University of Maryland School of Medicine (Baltimore, MD) Elizabeth A. Streeten, MD, Denise Orwig, PhD; University of Cincinnati College of Medicine (Cincinnati, OH) Nelson Watts, MD; University of Wisconsin School of Medicine and Public Health (Madison, WI) J. T. Harrington, MD.