The Journals of Gerontology Series A: Biological Sciences and Medical Sciences 58:M542-M547 (2003)
© 2003 The Gerontological Society of America
A Comparative Study of Rehabilitation Outcomes of Elderly Hip Fracture Patients: The Advantage of a Comprehensive Orthogeriatric Approach
Abraham Adunsky1,4,
Ayala Lusky2,
Marina Arad1 and
Raphael J. Heruti2,3,4
1 Department of Geriatric Medicine, The Orthogeriatric Unit, Sheba Medical Center, Tel-Hashomer, Israel.
2 The Health Service Research Unit, Israel Ministry of Health, Tel Aviv, Israel.
3 Department of Physical Medicine and Rehabilitation, Reuth Medical Center, Tel-Aviv, Israel.
4 The Sackler School of Medicine, Tel-Aviv University, Israel.
 |
Abstract
|
|---|
Background. The optimal setting and design of a comprehensive treatment for elderly hip fracture patients is unknown. The aims of this study were to compare the functional outcomes associated with an innovative comprehensive single-step (geriatric-based) and classic double-step (orthopedic-based) treatment of such patients.
Methods. This is a partially concurrent prospective study, taking place in the orthogeriatric unit of a department of geriatric rehabilitation, in a large urban academic freestanding hospital. The participants were 336 consecutive elderly people, admitted for surgry of extracapsular or intracapsular hip fracture. 204 patients were admitted to the department of orthopedic surgery and transferred, shortly after surgery, to the orthogeriatric ward for ongoing rehabilitation (ORT group). Another 116 patients were admitted directly to a geriatric-based orthogeriatric ward, and they received surgical, medical, and rehabilitation treatment within the same facility (ORTGER group). The main outcome measures were length of stay, absolute functional gains of the motor functional independence measure (FIM), and relative functional motor gains according to the Montebello rehabilitation factor score. Succesful rehabilitation was defined as relative functional gain >.5.
Results. 320 patients were included in the final analysis. The two groups were similar, yet ORT-group patients were somewhat younger (p <.02) and were cognitively better preserved (p <.01). Admission cognitive FIM was the strongest positive predictive factor associated with successful rehabilitation (odds ratio = 2.45, 95% confidence interval 1.89–3.31, and p <.001). Significant improvement of total FIM scores occurred during rehabilitation in both groups. The relative functional gain was smilar in the two groups, but total hospital length ofstay was 5 days shorter in the ORTGER group (p <.01). After the effect of age, sex, length of stay, fracture type, and cognition at onset of the rehabilitation period were adjusted for, patients of the ORTGER group had a twofold chance for successful rehabilitation, compared withORT-group patients (odds ratio = 1.97, 95% confidence interval 1.09–3.65, and p =.03).
Conclusions. The functional outcome of elderly hip fracture patients is better for those treated in the orthogeriatric setting, as compared with the common two-steps model of orthopedic surgery followed by transfer to a geriatric rehabilitation facility.
HIP fractures constitute a leading cause of hospital admissions and length of stay among the elderly population. The economic burden (1) of such fractures, combined with limited financial sources for rehabilitation programs, calls for the investigation of innovative new forms of organization and integration of medical, surgical, and rehabilitation services for elderly hip fracture patients (EHFP). The common in-hospital course of such patients includes admission to an orthopedic ward, liaison with internal or geriatric medicine physicians, and then transfer to a rehabilitation setting. The establishment of such orthogeriatric wards has been described as a success by many authors (2–6). However, most of these units reflected a limited degree of geriatric and orthopedic liaison, and their effectiveness has been questioned (7,8).
We have recently established a unique form of orthogeriatric ward, designed to take care of EHFPs throughout their hospital stay. The nature and characteristics of this orthogeriatric facility have already been described in detail (9). Briefly, this is a geriatrics-based ward, admitting EHFPs directly from the emergency department. The ward integrates a multidisciplinary staff, and care is taken of patients' surgical, medical, and rehabilitation needs in a single setting, from admission to discharge. Although the feasability and operational beneficial aspects of this ward have been described, data regarding its efficiency, in terms of rehabilitation outcomes, have yet to be determined.
The purpose of this report was to compare the functional outcomes of EHFPs treated in a single-step orthogeriatric facility, with the classic pathway, in which patients are operated on in an orthopedic ward and then transferred to a rehabilitation facility. The results should assist in decision-making processes, concerning the most efficient utilization of services, as well as the facilitation of effective treatment strategies and potential improvement of the deployment and organization of health care systems.
|
Methods
|
|---|
Design
The basic hypothesis was that direct admission to an orthogeriatric facility could reduce length of stay and improve the functional outcomes of the rehabilitation among EHFPs. Therefore, the design served to compare results of rehabilitation between the two mentioned modalities of care. This is a partially concurrent prospective study (also termed "historial-prospective," with the two arms of study partially overlapping with time) evaluating the relationship between type of hospitalization of EHFPs and the rehabilitation outcomes. The study was conducted between 1998 and 2000 in a geriatric ward, which is a 30-bed facility of a major tertiary hospital, comprising an orthogeriatric unit with 15 beds. This ward utilizes an interdisciplinary team approach (10), in which medical personnel (two certified geriatricians, i.e., Adunsky and Arad, an orthopedic consultant, nurses, physiotherapists, occupational and speech therapists, social workers, and psychologists) meet twice a week to evaluate the status of each patient. During these meetings, a treatment plan is established and monitored. The patients undergo, on average, 5 hours of physical therapy and 5 hours of occupational therapy per week (5 days a week). Additional rehabilitative services are given whenever needed. According to Israeli standards, the type of patients, length of stay (LOS), and treatment intensity, this facility is considered an acute rehabilitation setting.
Study Population
Three-hundred thirty-six consecutive patients, 65 years of age or older, were included in the study. There was no preselection of patients on clinical grounds. The decision to admit patients to the orthogeriatric ward was based solely on availability of beds. When no beds were available at the orthogeriatric ward, patients were referred to the orthopedic ward. This process was independent of other factors, such as type of fracture, general condition, comorbidities, sex, or age.
Inclusion criteria included age ≥ 65 years, admission after pertrochanteric (extracapsular) or subcapital (intracapsular) hip fracture, and stable medical status enabling active rehabilitation treatment. Exclusion criteria included a rehabilitation period shorter than 7 days (assuming that the extent of rehabilitation in such a short period is limited and could distort results), presence of other acute disabilities (such as multiple trauma), postoperative unstable (non-weight-bearing) hip fractures, medical conditions preventing active rehabilitation (such as cardiac failure with functional capacity stage III–IV of the New York Heart Association classification, or severe chronic lung disease necessitating a constant use of oxygen), transfer to acute-care departments as a result of complications, or death. These exclusion criteria enabled the screening out of patients with either medical or functional conditions, which would limit rehabilitation potential in advance.
Of the 336 patients entering the study, 320 met the aforementioned criteria and were included in the final analysis. Two groups were compared. One group consisted of 204 patients, admitted soon after surgery from the orthopedic department (ORT group) for further rehabilitation to the orthogeriatric ward. A second group consisted of 116 patients who were admitted directly from the emergency ward to the orthogeriatric ward (ORTGER group). These patients received all surgical, medical, and rehabilitation therapy in a single setting.
Assessment of Functional and Cognitive Status
We used the motor functional independence measure (FIM) for the assessment of motor functions (11), and the cognitive FIM for cognitive assessment (12). These tests are routinely used to assess patients admitted to the geriatric ward. Patients were evaluated within 72 hours following admission by the rehabilitation team (composed of a nurse, a physiotherapist, an occupational therapist and other team members, as needed). The same team tested patients' function (using FIM) 3 days preceding discharge. The team administering the FIM did not include the treating physicians. None of the 204 ORT-group patients got any interdisciplinary team meetings before being transferred from the orthopedic department to the orthogeriatric ward. In addition, none of these patients had the same nurses or therapists before they were transferred. The rehabilitation team members doing assessments were not "blinded" on intention, though practically most of them were unaware of the patients' referral source.
We determined functional gains by using absolute and relative parameters calculated from the motor FIM: (a) the absolute functional gain (AFG), which is the motor FIM gain, reflecting the change in motor FIM scores (discharge FIM - admission FIM), and (b) the absolute functional efficiency (AFE), reflecting the absolute daily functional gain (AFG/LOS). In addition to these two absolute parameters, we used the Montebello rehabilitation factor score (MRFS) (13) to calculate relative gains. Regardless of its name, the MRFS is not actually a score but rather a method for calculating a relative gain, based on existing validated scores (such as FIM). According to this method, the basis for calculating the relative gain is the patient's specific potential for improvement (maximal possible FIM - actual admission FIM). Using the MRFS helps us to overcome the "ceiling effect," that is, the fact that the gain that patients with high admission scores can achieve is limited compared with the gain that patients with low admission scores can achieve. The relative functional gain (RFG) is calculated according to the MRFS method as AFG/(maximal possible FIM - actual admission FIM). The relative functional efficiency (RFE) is calculated according to the MRFS method as RFG/LOS. Using this model, we find that RFG and RFE scores of patients who benefit from rehabilitation will range from 0 to 1. The motor RFG and motor RFE are measured in the same way, with the use of the motor FIM score (instead of the total FIM score).
Other Variables
Other variables that were examined included sex, age, time interval from admission to hospital to admission to rehabilitation interval (relevant only for the ORT group), type of fracture, and LOS. Dependent variables included absolute and relative gains, and efficiencies (daily gains). Independent variables included setting, cognitive state on admission, age, sex, fracture type, and LOS.
Statistical Analysis
We tested differences of demographic and clinical characteristics between the two groups, in categorical variables, by using the chi-square test for proportions. In case of continuous variables, we applied an unpaired Student's t-test for normally distributed variables. Otherwise, we used the Wilcoxon–Mann–Whitney test. Results are presented as means ± SD, as well as by median values. We defined a satisfactory functional gain (success of rehabilitation) as motor RFG >.50, and an unsatisfactory one as RFG <.50. We used the RFG score, as this score was previously found (14,15) to serve as the most important parameter associated with success of rehabilitation, compared with other functional gain scores. In contrast with the RFE (relative daily gain), the RFG does not depend on LOS, which frequently depends on administrative or other nonmedical considerations. Because the distribution of RFG ranges between 0 and 1, we used a cutoff value of.5, representing an improvement of at least 50% of the possible potential. We also used a chi-square test to compare the percent of rehabilitation success by each factor.
We performed a logistic regression analysis to assess the association between type of department and success of rehabilitation, adjusting for all other factors affecting it. We included in this model the type of department and all other factors found to be significant at the univariate analysis at p ≤.25. We used the.25 level as a criterion for inclusion of variables in the model as a way to identify important variables that fail to be expressed at the more traditional level, such as p <.05 (16). Results are presented as odds ratio (OR) and 95% confidence interval (95% CI). We performed all statistical tests by using SAS statistical software package Version 8.2 (SAS Institute, Inc, Cary, NC).
|
Results
|
|---|
Subject Participation and Baseline Characteristics
A total number of 320 patients met the aforementioned criteria and have completed surgery followed by a rehabilitation period. None of these patients presented with major medical comorbidities, which would make assessment of motor function impossible or prevent them from being treated by the rehabilitation team. Patients' demographic characteristics are presented in Table 1. Average number of days between admission and day of surgery was 3.5 and 3.6 for respective groups. Compared with ORTGER patients, the ORT-group patients were younger (p =.02), suffered more frequently subcapital rather than pertrochanteric fractures (statistically nonsignificant), and had a shorter rehabilitation stay (p <.01), but a significantly longer total hospital LOS (p <.01).
Functional Outcome
ORT patients presented with higher total FIM, cognitive FIM, and motor FIM scores on admission to the orthogeriatric unit (Table 2). Both groups showed statistically significant increases of total, motor, and cognitive FIM scores during the rehabilitation phase, similar to the increase in motor FIM scores, suggesting that FIM gains were mostly motor. Absolute total FIM gains and motor FIM gains were somewhat better in the ORTGER group (p =.03 vs p =.08, respectively). The resulting motor FIM daily gains (AFE) and motor RFG were similar in both groups, however, motor RFE results were significantly higher in the ORT group (2.42 vs 1.85, p <.001). In both groups, the values of these variables varied greatly among patients, indicating a large intersubject variability concerning the patients' potential for rehabilitation and the ability to materialize this potential.
A univariate analysis to identify variables associated with success of rehabilitaion (Table 3) showed that age and cognitive FIM at admission are highly correlated with success of rehabilitation. Paradoxically, longer LOS were associated with lower rates of successful rehabilitation, as measured by motor RFG. This may reflect delayed accomplishment of rehabilitation and discharge, resulting from the intention to further benefit such patients.
We have reevaluated the data by using a stratified analysis to test for the effect of service on the outcome, within strata of cognition. No interaction was observed between the strata. We also included in the logistic regression model an interaction term of cognition at baseline and service, which also gave a statistically nonsignificant result (p =.29). Furthermore, we have reanalyzed the data by strata of age and by type of fracture, and we found no interaction there as well.
To assess the independent effect of type of setting on success of rehabilitation, adjusting for independent variables, we used a logistic regression analysis. Table 4 shows that admission cognitive FIM has the strongest association with successful rehabilitation, when we adjust for all other variables in the model. Each increment of five units of cognitive FIM increases the chance for successful rehabilitation by 2.45 (OR = 2.45, 95% CI 1.89–3.31, p <.001). Because age and cognitive status are highly correlated (once cognitive FIM is entered into the model), the age variable did not have an additional independent effect on rehabilitation success. After adjusting for all previously defined independent variables, we found that patients of the orthogeriatric pathway had a twofold higher chance of successful rehabilitation, compared with patients in the orthopedic pathway (OR = 1.97, 95% CI 1.09–3.65, p =.03).
To evaluate the appropriateness of this model, we used Hosmer-Lemenshow goodness-of-fit statistics. These showed that the applied model is appropriate and fits well (p =.68).
|
Discussion
|
|---|
The present study compared the functional outcomes of a classic two-step pathway with an innovative single-step orthogeriatric management. The study is important with regard to several aspects. First, the AFGs were quite similar in both groups. This occurred despite the fact that ORT patients were younger, suffered more frequently from subcapital fractures (which usually require a shorter period of rehabilitaion), presented with higher total FIM, cognitive FIM, and motor FIM scores on admission to the rehabilitation, and had a longer total LOS. These are expected to put ORT patients in a much better starting position and result in better functional outcomes. However, the rehabilitation success (measured by motor RFG) did not differ between the groups (Table 3) and was even better in ORTGER patients after we adjust for the other variables (Table 4). This is important because it has been previously shown (14,15) that RFG is the most accurate parameter in evaluating the success of rehabilitation. In contrast with RFE (relative daily gain), RFG does not depend on LOS, which may be influenced by other than proper medical considerations. These results add further validation to this type of treatment and favor the adoption of the single-step orthogeriatric management, showing that an integrated approach is not only feasible for older people with hip fractures, but also results in favorable functional outcomes. Combined with previous data (9) taking into account other relevant parameters of the project, such as low rates of morbidity and mortality, the application of the ORTGER model seems to benefit EHFPs a great deal.
The orthogeriatric approach abolishes the need to relocate older patients to various departments of the hospital, which may be harmful for this frail population. Moreover, the patients benefit from early rehabilitation and close supervision of the same multidisciplinary staff members during the entire hospital stay. This mode of organization helps to resolve the logistical problem faced by a tertiary referral center of a shortage in acute care beds, needed for multitrauma cases requiring immediate surgery. It also utilized no additional facilities and was the result of a reorganization of existing resources. Such a geriatric-based facility minimizes the use of general orthopedic beds by large numbers of EHFPs and permits optimization of the orthopedic service by hospitalizing other complicated orthopedic multitrauma patients for better acute care. Finally, it creates and stimulates professional interest in a specific group, which is frequently considered too complex and too tedious by many medical and paramedical professionals.
An important finding of this study is the association between cognitive status and functional outcome, in both groups. In fact, this bore the highest OR (2.45) and the highest probability (<.001) with regard to successful rehabilitation. This is in accordance with previous reports comparing functional improvement of demented and cognitively intact patients (14,17,18), but it is in contrast with a recently published study showing that cognitive level may not be important for motor gain after hip fracture (19). Age was also significantly associated with success of rehabilitation (p <.01); however, because of a strong correlation with cognitive FIM, this variable did not have an additional effect on success once cognitive FIM was already included in the model.
Several limitations of this study warrant consideration. First, there is the potential bias that may be related to its design and the nonrandomization of patients. Moreover, potential bias might emerge from the multifactorial background of the patients and the noninclusion, in the analysis, of some already well-known predictors of outcome such as associated comorbidities, prefracture ambulating status, and prefracture level of activities of daily living (20–22). The variable of time interval from surgery to onset of rehabilitation period was not analyzed, because a delay in a patient's transfer (as well as timing of surgical intervention) may depend on other medical and nonmedical factors. Furthermore, a specific adjustment for coexisting comorbidities was not performed, because patients with significant comorbidities, which could affect rehabilitation outcome, were excluded a priori. Finally, we are unable to give any details regarding general costs and cost effectiveness, because our purpose was to look at functional rehabilitation outcomes rather than at associated expenses. Despite such limitations, our data provide clinical evidence supporting the implementation of this model of comprehensive orthogeriatric care. This implementation could also help in skillful use of economic resources and the facilitation of effective treatment strategies.
Conclusion
In conclusion, rehabilitation functional outcomes of EHFPs are better for those treated in the orthogeriatric setting, when compared with the common two-phase model of surgery followed by transfer to a geriatric rehabilitation facility. The present model of a comprehensive orthogeriatric ward is a practical and feasible service that covers the various needs of these patients, results in shorter LOS, and more efficient rehabilitation. This study underscores the need to reconsider the economical issues confronting policymakers involved with the geriatric hip fracture population.
|
Acknowledgments
|
|---|
We give special thanks to the rehabilitation team of the Department of Geriatric Medicine, Sheba Medical Center, for assistance in testing the patients and collecting the data.
Address correspondence to Abraham Adunsky, MD, Division of Geriatric Medicine, Sheba Medical Center, Tel-Hashomer 52621, Israel. E-mail: adunger1{at}sheba.health.gov.il
Received August 5, 2002
Accepted October 14, 2002
|
REFERENCES
|
|---|
- Schneider EL, Guralnik JM. The ageing of America: impact on health care costs. JAMA.. 1990;263:2335-2340.[Abstract]
- Capparella O, Bernabei R, Sanguinetti C, et al. Continuous geriatric care in orthopaedic wards: a valuable alternative to orthogeriatric units. Aging (Milano) Clin Exp Res.. 1993;5:207-216.
- Gilchrist WJ, Newman RJ, Hamblen DL, et al. Combined orthopaedic geriatric care. Lancet.. 1985;1:349-350.[Medline]
- Boyd RV, Compton E, Hawthorne J, et al. Orthogeriatric rehabilitation ward in Nottingham: a preliminary report. Br Med J.. 1982;285:937-938.
- Incalzi RA, Bernabei R, Carbonin P. The orthopeadic geriatric unit: a new model of hospital care. Ann Ital Med Int.. 1995;10:49-52.[Medline]
- Hicks KA, Currie CT. Patterns of geriatric orthopaedic liaison in Scotland. Health Bull (Edinb).. 1996;54:2-10.
- Gilchrist WJ, Newman RJ, Hamblen DL, et al. Prospective randomized study of an orthopaedic geriatric inpatient. Br Med J.. 1988;297:1116-1119.
- Hempsall VJ, Robertson DRC, Campbell MJ, et al. Orthopedic geriatric care—is it effective? A prospective population based comparison of outcome in fractured neck of femor. J R Coll Phys (London).. 1990;24:47-50.
- Adunsky A, Levi R, Cecic A, et al. The "Sheba" model of comprehensive orthogeriatric care for elderly hip fracture patients: a preliminary report. J Isr Med Assoc.. 2002;4:259-261.
- Heruti RJ, Ohry A. The rehabilitation team: commentary. Am J Phys Med Rehab.. 1995;74:466-468.[Medline]
- Lincare JM, Heinemann AW, Wright BD, et al. The structure and stability of the functional independence measure. Arch Phys Med Rehabil.. 1994;75:127-132.[Medline]
- Adunsky A, Fleissig M, Levenkrohn S, et al. A comparative study of mini-mental test, clock drawing task and cognitive-FIM in valuating functional outcome of elderly hip fracture patients. Clin Rehabil.. 2002;16:455-460.
- Drubach DA, Kelly MP, Taragano FE. The Montebello Rehabilitation Factor Score. J Nurs Rehabil.. 1994;8:92-96.
- Heruti RJ, Lusky L, Barell V, et al. Cognitive status at admission: does it affect the rehabilitation outcome of elderly patients with hip fracture? Arch Phys Med Rehabil.. 1999;80:432-436.[Medline]
- Heruti RJ, Lusky L, Dankner R, et al. Rehabilitation outcome of elderly patients after a first stroke: effect of cognitive status at admission on the functional outcome. Arch Phys Med Rehabil.. 2002;83:742-749.[Medline]
- Mickey J, Greenland S. A study of the impact of confounder-selection criteria on effect estimation. Am J Epidemiol.. 1989;129:125-137.[Abstract/Free Full Text]
- Rozzini R, Firsone GB, Barbisini P, et al. Dementia does not prevent the restoration of safe gait after hip fracture. J Am Geriatr Soc.. 1997;45:1406-1407.
- Goldstein FC, Strasser DC, Woodard JL, et al. Functional outcome of cognitively impaired hip fracture patients on a geriatric rehabilitation unit. J Am Geriatr Soc.. 1997;45:35-42.[Medline]
- Beloosesky Y, Grinblat J, Epelboym B, et al. Functional gain of hip fracture patients in different cognitive and functional groups. Clin Rehabil.. 2002;16:321-328.[Abstract/Free Full Text]
- Koval KJ, Skovron ML, Aharonoff GB, et al. Ambulatory ability after hip fracture. A prospective study in geriatric patients. Clin Orthop Related Res.. 1995;310:150-159.[Medline]
- Cheng CL, Lau S, Hui PW, et al. Prognostic factors and progress for ambulation in elderly patients after hip fracture. Am J Phys Med Rehab.. 1989;68:230-233.[Medline]
- Barnes B, Dunovan K. Functional outcomes after hip fracture. Phys Ther.. 1987;67:1675-1679.
Top
Abstract
Methods
