Comparison of an Organized Geriatric Fracture Program to United States Government Data

Abstract

Objective: This study describes the financial impact of an organized hip fracture program for elderly patients age 65 years and older. Methods: This is a retrospective study of 797 fractures in 776 consecutive patients over a 50-month period (May 2005 to July 2009) treated in an organized hip fracture program for the elderly identified from a quality management database. Financial, demographic, and quality-of-care data were collected. The length of hospital stay, in-hospital complications, and Charlson comorbidity scores were collected from patient records, and all data were evaluated using standard statistical methods. Setting: 261-bed community-based, university-affiliated teaching hospital in an urban setting with a catchment area of approximately 1 million persons. This is a level 3 trauma center. Results: The average total net revenue per hip fracture was $12 159, with an average total cost to hospital of $8264. Physicians' fees consisted of fees collected by surgeons, anesthesiologists, medical specialty consultants, and consulting geriatricians and averaged $2024 per case. Thus, the average hospital charge to payers was $15 188. Compared to Agency for Healthcare Research and Quality average inpatient hospital costs in 2005 of $33 693, a savings of more than $18 000 was realized per patient. The average length of stay was 4.6 days, markedly less than the national average of 6.2 days. Conclusions: This organized geriatric fracture care model with geriatrics comanagement resulted in significant cost savings over a 50-month period, with associated increased quality. With an estimated 330 000 hip fractures annually in the United States, a large cost savings could potentially be realized if this model were more widely applied.

Hip fractures are the most costly fall-related fracture.^1,2 In 1984, the inpatient cost of hip fractures in the United States was estimated at $7.2 billion and rose to $10.1 billion by 2004. The overall cost of hip fracture care is projected to reach $240 billion by 2040.^1–3 The burden of health care costs for the elderly is becoming a national crisis. It is critical to identify opportunities to decrease the cost of health care for the elderly while preserving or improving the quality of care.

Hip fractures are a common event for older adults and are associated with significant morbidity and mortality. Most hip fractures are the result of falls.⁴ An estimated 330 000 hospitalizations for hip fractures occur each year in the United States, with a 1-year mortality of 20% to 25%.^5–10 Persons older than age 65 years sustain 86% of hip fractures, and the US Census Bureau predicts this segment of the population will double over the next 25 years.⁷ By 2030, 71 million Americans will be older than 65 years of age, accounting for 20% of the US population.^2,8,11,12 Estimates for the year 2040 predict more than 500 000 US hip fracture patients per year.^11,13,14

Most geriatric hip fracture patients are admitted to the hospital and are treated surgically.¹³ Several studies have shown inpatient hospital costs to be the largest single cost associated with hip fracture care, averaging between 44% and 57% of total costs.^3,9,15 These costs include fees to hospitals, doctors, and other professional services provided, paid by patients or insurers. Reducing hospital costs may result in substantial financial savings.

In 1995, Barrett-Connor⁷ reported that the cost for hospitalization, including physician fees, of a patient with hip fracture was $18 000 or 44% of medical costs of care. In 1996, Beck and coworkers¹⁶ reported that the average hospital cost per hip fracture patient, excluding physician fees, was $16 072. In 2003, the American Academy of Orthopaedic Surgeons (AAOS)^1,10 estimated that the cost of inpatient hip fracture care averaged $26 912. Medicare Standard Analytic Files (SAFs) from 1999-2003 show a mean inpatient cost of $26 884 per hip fracture patient.¹⁷ The Agency for Healthcare Research and Quality (AHRQ) reported mean US inpatient costs per hip fracture to be $32 310 in 2004¹ and $33 693 in 2005.⁶ Several studies have shown similar numbers.^{8,13,16,18,19}

The strongest predictors of total hospital charges for hip fractures repeatedly have been shown to be length of stay (LOS), in-hospital complications, and patient comorbidities.^13,16,18 It has previously been demonstrated that a comanaged, protocol-driven hip fracture program for the elderly reduces LOS and inpatient complications.^20,21 Early results from 193 consecutive patients showed significant in-hospital cost savings in nearly all expenditure categories.²² This study reports the in-hospital cost of 797 consecutive hip fractures over 50 months treated in such a program. An improvement in patient outcomes that simultaneously decreases costs of care creates high value for patients and health care systems.

Materials and Methods

Study Type: Retrospective Case Review

The Geriatric Fracture Program was implemented gradually beginning in 1995 and, since 2004, has used a comprehensive, standardized approach to each hip fracture patient starting at admission to the hospital. This program is co-led by an orthopaedic surgeon and a geriatrician. Approximately 10 attending orthopaedic surgeons and 10 attending geriatricians participate in the program.

Standard order sets are used for each patient. Most patients are admitted to the orthopaedic service with a geriatrician consulted immediately. An emphasis on efficient but thorough preoperative geriatric assessment ensures each patient is optimized for surgery in a timely fashion. Patients are risk stratified as well. A preprinted standardized geriatric consultation form is used. Early surgery for medically appropriate patients is a key principle, with most cases being completed within 24 hours of presentation. Patients are comanaged by an orthopaedic surgeon and a geriatrician throughout their stay at the center; both teams share “ownership” of the patients. Equality of services and a collegial environment are fundamental elements of the program. Demographic, surgical, and medical data are recorded for each fracture in a quality improvement database (SQL based; administered by the university’s General Clinical Research Center). Details of this program and the quality improvement database have been published previously.^20,21

Study Population and Surgical Procedures

Included in this study were 797 consecutive native hip fractures in 776 elderly patients surgically treated at a single community teaching hospital. The hospital is a level 3 trauma center affiliated with a university medical center. Pathological, periprosthetic, high-energy, multiple, and nonsurgical fractures were excluded, as were patients younger than age 65 years. A hip fracture was defined as a fracture of the proximal femur occurring anywhere from the femoral neck to the subtrochanteric region 5 cm distal to the lesser trochanter.

After approval from the University Research Subjects Review Board was obtained, research data were extracted from the program’s quality improvement database for all native hip fractures in patients 65 years and older discharged between May 1, 2005, and June 30, 2009. Financial data for each study patient were extracted from the hospital billing system by the finance department.

Nine different surgical procedures were used to operatively repair the hip fractures in study patients based on the fracture pattern and individual surgeon judgment and preference. The procedure and implants were selected by the surgeon and were not formally standardized, although there is an evidence-based guideline provided by the program’s surgical director. The implants and procedures used were blade plate fixation, cannulated screw fixation, sliding helical hip device, dynamic hip screw, dynamic condylar screw, hemiarthroplasty, and intramedullary nail fixation (IM nail). One Girdlestone resection was performed. All nondisplaced femoral neck fractures were treated with in situ cannulated screw fixation. Displaced femoral neck fractures were treated with hemiarthroplasty with cemented or uncemented prostheses. One total hip replacement was performed. Stable pertrochanteric fractures (AO type 31-A1 and some AO type 31-A2) were treated with dynamic hip screws or a sliding helical hip device according to surgeon preference. Unstable pertrochanteric fractures (some AO types 31-A2 and all 31-A3) were treated using long trochanteric entry IM nails, dynamic condylar screw, or blade plate fixation, depending on surgeon preference.

Data Acquisition and Variables

In total, 797 fractures (776 patients) met the inclusion criteria listed above, and their records were identified for analysis. Twenty-one patients suffered fractures of both native femurs at different times during the study period, resulting in 21 more individual hip fractures than patients. Financial, demographic, and outcome data for each patient were obtained from the hospital finance department and quality improvement database.

Costs of care were subdivided into several categories, including total hospital cost to payers, total net revenue, net margin, fixed cost, operating room costs, surgical type costs, and others (please see appendix for definitions). Physician service fees were obtained for each procedure and service and were included in this analysis. The average orthopaedic surgeon fee was $1310. The geriatrician’s fees were computed from averages received by the department for one level 5 consultation and three level 3 follow-up consultation visits. Because this information was obtained retrospectively, it was not possible to separate individual receipts for each patient; this methodology modestly overestimates the actual service fees received for the geriatrician’s service. The approximate average geriatrician consultation fee was no more than $382 per case. The anesthesiologists participating in this program were all members of the full-time faculty anesthesia group and used a unified billing system. The average anesthesiologist fee for care of a hip fracture was ascertained by the department billings to be $304 for 14 units of service. The cost of other specialty consultations averaged to $28 per case as these were rarely obtained. The average sum of physician fees was therefore $2024 per case.

Patient demographics (age, race, and gender) were available from the quality management database. The database also includes Charlson comorbidity scores, LOS, discharge destination, and mortality.²³ The LOS was defined as the number of days spent in the hospital and included the day of admission but not discharge (standard midnight census methodology). Mortality data were obtained by accessing the hospital data system, telephone calls, and query of the Social Security Death Index.

Average US inpatient hospital costs for hip fracture patients were determined using Medline and PubMed literature searches, Medicare’s SAFs, and AHRQ Healthcare Cost and Utilization Project (AHRQ-HCUP) data. SAFs are large databases on different chronic conditions, including hip fractures. The data are collected on samples of Medicare patients over specified time periods and present mean direct costs for hip fracture patients. Mean costs are broken down into acute hospitalization, skilled nursing facility (SNF), hospice, home health, outpatient, and total services. The AHQR is a division of the US Department of Health and Human Services. Data were available for direct inpatient charges from 1999 to 2003 and 2004 and 2005. The United States reports health care charges in publically accessible reports every year. This project is known as the AHRQ-HCUP.⁶ The figures reported in the HCUP Nationwide Inpatient Sample Comparison report are hospitals' charges for care. The actual payments to hospitals are not reported in the HCUP report.

The National Inpatient Discharge Survey is done every 2 years and was available from 2006. This survey includes the number of hospitalizations for hip fracture, the length of stay, and rates of discharge.⁵ Data from this survey were used to compare LOS to national data.

Statistical Methods

Patient demographic and financial data were analyzed using Microsoft Excel software. Characteristics of the study population are presented using descriptive statistics. Total hospital costs to payers are presented and compared with previously published data from centers in the United States. Overall costs are compared to expected values for the population in the United States using data from national databases (Medicare and AHRQ-HCUP). In addition, the cost analysis is provided based on the specific surgical procedure the patient underwent.

Results

The study population consisted of 78% women with a mean age of the 85.6 ± 7.7 years. Ninety-five percent of the patients were white. Average Charlson score was 3.0 ± 2.2. About one third came from nursing homes and half from their own home. Almost half had dementia. Most patients had surgery with 1 day of presentation. Table 1 displays a summary of the patient demographics. Table 2 is a summary of the financial data results presented in this article.

Table 1.

Patient Demographics for Geriatric Fracture Program, 2005-2009

Total no. of patients	776
Total no. of fractures	797
Gender (% female)	78
Ethnicity (% white)	95
Age, mean ± SD, y	85.6 ± 7.7
Preadmission residence, %  Home  SNF  ALF  Unknown	47 37 13 3
Patients with dementia, %	49
Average case mix index	2.1
Average Charlson score	3.0 ± 2.2
Average length of stay, d	4.6
Acute hospital mortality, %	2.6
Average time to surgery, h	22.18 ± 15.5

ALF, assisted living facility; SNF, skilled nursing facility.

Table 2.

Inpatient Costs of Care of a Hip Fracture (in US Dollars)

Average total hospital cost to payers	15 188
Average total net revenue	12 159 ± 5198
Average net margin	3990 ± 3634
Average fixed cost	3270 ± 2382
Average variable cost	4897 ± 3212
Total hospital costs	8264 ± 6234

These costs are the acute hospitalization costs for hip fracture patients, both including and excluding physician fees. The total hospital cost to the payer for caring for a hip fracture in this program was $15 188, which represented costs for inpatient hospital services. The hospital’s average total net revenue per case was $12 159. The hospital’s average total cost per hip fracture in the study population was $8264. Average net margin to the hospital was $3990.

Table 3 separates the hip fracture patients by their type of surgical procedure. The total number of procedures performed, their average length of stay, average total net revenue, and average net margin are shown. Hip hemiarthroplasty (both cemented and uncemented) was performed on approximately 35% of patients, 34% had dynamic hip screws, 9% had helical hip devices, 12% had femoral IM nails, 9% had cannulated screws, and less than 1% had blade plate fixation and dynamic condylar screws. Total net revenue was highest for those patients who had a femoral blade plate, averaging $15 313, but average net margin was lowest for these procedures because of their high cost to the hospital for 1 outlier case. The highest net margin of $4730 was realized for patients who underwent hemiarthroplasty for displaced femoral neck fracture.

Table 3.

Financial Analysis by Surgical Fixation Type

Type of Surgical Fixation	Number of Procedures (N = 797), n (%)a	Average Length of Stay, d	Average Total Net Revenue, US $	Average Total Net Margin, US $
Blade plate	7 (0.9)	4.4	15 313	–448
Hemiarthroplasty	278 (35.0)	4.9	13 472	4730
Dynamic hip screw	271 (34.1)	4.3	11 127	3608
Cannulated screws	73 (9.2)	4.5	11 287	4604
Intramedullary nail	94 (11.8)	4.6	11 474	2564
Dynamic helical hip device	70 (8.8)	4.4	12 507	4325
Dynamic condylar screw	2 (0.3)	3.5	12 208	3537

^a One total hip replacement and one Girdlestone arthroplasty not included in the table.

Table 4 is a breakdown of acute hospital costs per patient by individual year of the study. Although inflation is a reality in hospital costs, these numbers show minimal increases over the past 4 years despite substantial increases in medical care costs nationally. Average total net revenue and net margin are minimally changed from 2006 to 2009. The average costs per hip fracture were also calculated for individual surgeons involved in the study. Although the differences are not large, there is a general trend for the surgeons performing more cases to have higher net revenue averages. However, overall, the numbers are similar between surgeons. Table 5 describes inpatient cost drivers and the program’s countermeasures to reduce these costs. Table 6 describes features of the program in general to allow the reader to better see these features and some of their potential benefits.

Table 4.

Financial Analysis by Year

Year	No. of Patients	Average Total Net Revenue, US $	Average Total Net Margin, US $
2006	221	11 904	3913
2007	172	11 847	4570
2008	227	12 234	4455
2009	177	12 675	3880

Table 5.

Techniques for Inpatient Cost Reduction

Inpatient Cost Driver	Program Countermeasure
Length of stay	Early surgery
	Discharge planning when admitted
	Consistent team goal setting
Complications	Daily geriatric comanagement
	Standardized order sets
	Daily team communication
Delirium	Early surgery and careful pain  management
	Standardized avoidance of harmful  meds
Arrhythmia	Low-dose standard beta blockade
Acute renal failure	Early aggressive hydration when  admitted
Excessive diagnostic testing	Avoidance of tests by protocol
Multiple medical specialty  consultations	Geriatricians order any needed  consultations

Table 6.

Geriatric Fracture Program Features

Program Features of Hip Fracture Program	Benefit
Standardized orders in  emergency department	Reduced length of stay in  emergency
Transfer envelopea	Improved transfer of essential  documents
Standardized admit orders	Improved patient care, reduced  errors
Standard geriatric  consultation	Improved documentation of  condition
Standard postoperative  orders	Reduced variation in care,  reduced errors
Standard discharge packet	Improved outpatient care at  handoff
Osteoporosis treatment  recommendations	Improved osteoporosis care  after fracture
Standard data collection	Continuous quality improvement
Standard surgical  treatment poster	Point-of-service tool to assist  surgeon
Standard nursing care plan	Coordinated care
Standardized implant  selection	Reduced costs to hospital
All patients comanaged  daily	Improved quality of care
Community outreach to  nursing homes	Improved feedback in both  directions
Academic program	Residents and student  suggestions lead to improved  program features

^a Transfer envelope is a large preprinted envelope with a checklist on the front of essential items required to care for the patient in the hospital: medication list, advanced directives, family contact information and most recent history and physical exam, and so on.

Comparison costs are shown in Figure 1 . Published studies from the 1990s showed total hospital cost to payers of hip fracture care to be $11 000 to $18 000.^7,16,24 The Medicare estimate for the average inpatient hospital cost of each hip fracture was $26 884 in the years 1999-2003.² The AHRQ reports that the average direct inpatient charges per hip fracture in the United States in 2004 were $32 310.¹ This figure increased to $33 962 in the AHRQ’s HCUP report for 2005.⁶ For the study population in this program, cost savings exceeding $18 000 per fracture are demonstrated compared with these national cost averages.

Figure 1.

Medicare, American Academy of Orthopaedic Surgeons (AAOS), and Agency for Healthcare Research and Quality–Healthcare Cost and Utilization Project (AHRQ-HCUP) data represent charges. Other comparison figures represent total hospital cost to payers. SAF, Standard Analytic File.

The average length of stay was 4.6 days in the study group, compared to 6.2 days according to the 2004 and 2006 National Hospital Discharge Surveys.^1,5 The average Charlson comorbidity score was 3.0 in this study population. Patients with Charlson scores ≥4 are at higher risk for in-hospital complications due to more serious medical comorbidities at admission.^21,25 In this study population, 35% of the patients had a Charlson score ≥4, and this group of patients had an increased LOS of 5 days.

Average time to surgery was calculated to be 22.18 ± 15.5 hours. There were only 7 intensive care unit (ICU) stays out of 797 fracture admissions during this study. Twenty-seven patients (3.3%) required reoperation between 14 days and 17 months of the initial hip fracture surgery. In-hospital mortality was 2.6% in this study population. The 30-day readmission rate was 10.3%. Of the readmissions, 14 (1.8%) were surgical complications, and 68 (8.5%) experienced medical complications such as pneumonia, congestive heart failure, gastrointestinal problems, and renal failure. When the place of residence prior to admission was considered, 13% of home dwellers died at 1 year; 23% of assisted living dwellers and 30% of skilled nursing home patients had died by 1 year (a separate mortality analysis and discussion has been accepted for publication). These numbers compare very favorably to other published literature on hip fracture mortality rates.

Discussion

Hip fractures in the geriatric population produce immense human and economic costs. By 2030, the US Census Bureau predicts that 20% of the US population will be older than age 65 years, and this group sustains more than 85% of hip fractures.^2,12 The US health care system is currently burdened financially by the cost of care of older adults, and this problem has no foreseeable solution. A recent Centers for Disease Control and Prevention (CDC) statement warns that US health care spending on fall-related injuries in the geriatric population will increase dramatically by 2020, approaching $54.9 billion yearly.²⁶ Therefore, physicians, hospitals, and the US government will need to work together to meet the demands of an aging population while containing costs and improving quality.

Cost analysis of hip fracture patients is a complex and detailed undertaking. Several studies have shown that inpatient hospital costs are the largest single medical cost for patients who have sustained a hip fracture.^{7,9,13,15,27,28} This study evaluates inpatient hospital costs for 797 hip fractures (776 patients) treated in a comanaged protocol-driven program. The results show significantly lower costs than those found for US national averages. The average cost to payers for acute hospitalization was $15 188, dramatically lower than national 2005 averages.⁶ The most recent US national data available for review are from 2005, and thus the savings offered by this model of care may be underestimated if more current data were available for study, given anticipated yearly increases. Savings appear to be sustained throughout the study period.

The use of a lean business flow model when constructing this program likely has a considerable effect on the cost savings and quality measures. The lean business flow model looks at the process of care from admission to discharge with a goal of eliminating wasteful and harmful processes that do not add value to the patient’s care. Lean business principles are used widely in industry but infrequently in health care.²⁹

Multiple studies have shown that the strongest predictors of total hospital charges are LOS, in-hospital complications, and comorbidities.^13,27 Despite a high level of comorbidity in the patient population, this program results in decreased LOS compared to national averages by approximately 1.6 days. Also, the comanaged fracture care model has been shown by several studies to decrease postoperative complications, inpatient mortality, and ICU bed-days.^14,20,21,30 Table 5 describes techniques used to reduce costs associated with LOS, complications, and comorbidities.

The patients in comanaged care models benefit from the expertise of both geriatricians and orthopaedic surgeons. There is sufficient evidence that an interdisciplinary approach to geriatric care decreases morbidity and mortality. Perhaps the most impressive aspect of this model is the ability to generate cost savings while providing patients with high-quality care and better outcomes.

Other contributors to the cost savings produced by this geriatric fracture model relate to the utilization of resources. The standard protocols use generic medications in all cases. Lower costs on implants have been negotiated and contribute to cost savings. These areas are too broad to address in this study. However, it is likely that they have contributed to the cost savings experienced by this program.

The close interaction between geriatricians and orthopaedic surgeons leads to better utilization of hospital resources. Prompt geriatric assessment of hip fracture patients decreases time to surgery as well as overall LOS.^14,30 Standardized order sets and consultation forms ensure that necessary labs and radiographs are routinely obtained to optimize patients for surgery. These standardized forms also help to decrease unnecessary tests that add to overall costs but do not improve patient care or outcomes.

Preoperative risk stratification for surgery helps to decrease the number of surgical cases that are delayed or cancelled by the anesthesiologist. Surgical delays and cancellations are costly to the patients and hospitals. In this model of care, medical subspecialty consults (ie, cardiology) are rarely required, leading to expedited surgery and decreased hospital costs.

Nurses and social workers play an integral role in this fracture program. All staff members in this program are familiar with the standard protocols and begin to implement them immediately upon hospital admission. The nursing care plan precisely matches the standardized order sets at each step. Social work involvement begins when the patient is admitted, which facilitates the determination of individual patient discharge needs. This leads to reduced LOS and reduced readmission rates.

Limitations of This Study

This study focuses only on inpatient hospital costs for hip fracture patients. The costs of skilled nursing and rehabilitation facilities, home care, outpatient care, and loss of work or productivity or opportunity were not evaluated.^{7–9,15,19,28} We chose to focus on the acute hospital costs because they are most directly related to the comanagement model at our institution. Another limitation is that the data are compared to average costs to Medicare and do not adjust for patient characteristics. The patient population studied in this program is older, has more comorbidity, and has greater functional impairment than average hip fracture populations, so it is likely that cost savings are underestimated.

Another limitation of this study is the ability to generalize these findings. Not all hospitals have the level of resources that the study hospital has, and this may limit the ability to replicate this model of care.

Summary

This model of care is highly cost-effective while simultaneously providing the patients with excellent care. A lean business flow model combined with geriatric comanagement is described above. Cost savings for the acute hospital stay are reduced by a shortened LOS, reduction of harmful and wasteful processes, and decreased in-hospital complications. It is suggested that the adoption of protocol-driven, comanaged models at other institutions could result in comparable cost savings and quality outcomes.

Further studies at multiple centers are necessary to validate these findings. Such investigations are being planned. With inflation and the aging of the population, the cost of caring for geriatric hip fractures will increase dramatically over the next 30 years. An organized comanaged program for geriatric hip fractures described in this article could save the US health system more than $1 billion per year while enhancing the quality of care to frail elders.

Appendix

Financial Terminology

• Case mix index (CMI): Case mix groupers categorize patients into statistically and clinically homogeneous groups based on the collection of clinical and administrative data. Adjusting for patients of different levels of acuity forms the basis for health care organization comparisons and case mix–adjusted resource utilization. Over the years, these grouping methodologies and their accompanying indicators have been used by health care facilities to effectively plan, monitor, and manage the services they provide.

• Total hospital cost to payers: The costs to the patients and insurers for the inpatient hospital stay, not including physician and professional fees

• Total net revenue: Total receipts for the hospital; does not include physician’s fees

• Total cost to hospital: Fixed plus variable costs

• Net margin: Total net revenue minus the fixed and variable costs (ie, the profit)

• Fixed cost: Fixed overhead expense, including facility costs, depreciation, utilities, housekeeping, administration, and so on. Fixed costs vary with LOS.

• Variable costs: All expenses related to variable use of services and materials, including ICU cost, pharmacy, laboratory, radiology, operating room costs, preanesthesia, cardiology, emergency department, and supplies