FOUR OR TWELVE MONTHS FOLLOW-UP IN THE EVALUATION OF FUNCTIONAL OUTCOME AFTER HIP FRACTURE SURGERY?

Transcription

1 Scandinavian Journal of Surgery 94: 59 66, 2005 FOUR OR TWELVE MONTHS FOLLOW-UP IN THE EVALUATION OF FUNCTIONAL OUTCOME AFTER HIP FRACTURE SURGERY? T. Heikkinen, P. Jalovaara University of Oulu University Hospital of Oulu Department of Orthopaedic and Trauma Surgery, Finland ABSTRACT Background and Aims: As a rule, follow-up for at least one year is recommended for fracture studies. This is considered the shortest reliable interval. Still, in the case of hip fractures of the elderly, shorter follow-up might be more practical, since the life expectancy of these patients is often short. The aim of this study was to see if a short four months follow-up period would be acceptable in hip fracture surveys. Material and Methods: Information on 196 consecutive non-pathological hip fracture patients aged 50 years or over (mean 79 years) was collected using a standardised hip fracture audit concentrating on functional measurements at admission and at four and twelve months follow-ups. Results: 167 patients were alive at four months and 152 and at one year. The patients who died between four and twelve months had poorer functional capacity in the fourmonth evaluation than those who survived one year. The analysis of repeated measures, including only the patients alive at the last follow-up, showed that residential status, use of walking aids and 6 out of 10 and ADL variables (bathing, toileting, shopping, household activities, doing laundry, banking) did not change significantly. Walking ability and the rest 4 ADL variables (dressing, eating, food preparation, use of transportation) improved and pain decreased. Conclusions: Due to high mortality and age-related deterioration of functioning, no steady state i.e. final result is ever reached after hip fracture in the elderly. Four-month follow-up is justified as the shortest possible period, because the socioeconomically most important variable, i.e. place of living, and most of the ADL functions do not change significantly after that. Key words: Activities of daily living; aged; hip fractures; recovery of function; follow-up studies; treatment outcome INTRODUCTION Correspondence: Pekka Jalovaara M.D. Department of Orthopaedic and Trauma Surgery University of Oulu FIN University of Oulu, Finland pekka.jalovaara@oulu.fi Sufficiently long follow-up, usually for at least one or two years, is a common requirement in most orthopaedic journals for studies on the outcome of treatment of fractures. This requirement is difficult to apply to studies on hip fractures in the elderly because mortality is high, and the number of cases available for analysis thus decreases rapidly over time. Furthermore, the functional abilities of hip fracture patients, which are often markedly poor even before the time of the fracture, deteriorate rapidly further due to natural age-related causes. This age impairment of function will distort the interpretation of the long-term outcome. In the studies on opera-

2 60 T. Heikkinen, P. Jalovaara tive methods and rehabilitation routines applied to hip fracture patients, follow-up times have varied from a few months to ten years. There is no consensus on the optimal follow-up time for hip fracture surveys. Earlier studies on hip fracture treatment, which have used several follow-up points with the same functional variables, suggest that a substantial proportion of excess mortality occurs in the first two to six months following fracture (1 6), and that the main recovery of functions in the activities of daily living has been achieved by four to six months after the fracture (7 13). Our aim was to evaluate if short four months follow-up is adequate in studies on the treatment and rehabilitation of hip fractures using function, residence and activities of daily living as the main outcome variables. TABLE 1 Prefracture patient characteristics (total number of patients 196). Sex Males 44 Females 152 Side of fracture Right 96 Left 100 Type of fracture Undisplaced intracapsular 27 Displaced intracapsular 97 Basocervical 3 Trochanteric two-fragment 30 Trochanteric multi-fragment 33 Subtrochanteric 6 Living alone Yes 73 No 73 Institutional care 50 ASA grade I 3 II 27 III 124 IV 37 Primary operation Three screws 31 Single screw with slide plate 13 Intramedullary nail 61 Hemiarthroplasty 76 Total hip arthroplasty 13 Conservative 2 Discharged to Own home 13 Institutional care 1 Nursing home 2 Permanent hospital inpatient 11 Rehabilitation unit 69 Health centre hospital 98 Died 2 Age Mean 79 Median 80 Min-max Length of stay in orthopedic ward (days) Mean 7.7 Median 6 Min-max 2 64 MATERIAL AND METHODS The study involved 196 consecutive patients aged 50 years and over with non-pathological hip fractures. All patients were residents of the primary catchment area of Oulu University Hospital and were admitted between August 1, 1998 and July 31, The data were collected using standardised multicentre hip fracture study forms (Standardised Audit for Hip Fractures in Europe, SAHFE) (14 16). On admission, a study nurse recorded the baseline personal information, such as age, sex, place of residence, walking ability and use of walking aids by interviewing the patient and, if necessary, the patient s relatives or the staff of the relevant institution. The activities of daily living (ADL) inquiry was recorded similarly. It contains four questions on basic ADL (dressing, bathing, feeding, toileting) and six questions on independent ADL (shopping, household chores, doing laundry, cooking, banking and use of transportation). Types of fracture, treatment methods and ASA grades (American Society of Anesthesiologists) (17) were drawn from patient files and operation registers. There were 44 males and 152 females with a mean age of 79.0 years at fracture (Table 1). Most of the patients were admitted from their own homes and were able to walk alone outdoors without any walking aids (Table 2). The majority of fractures were cervical and were operated on with osteosynthesis or hemi- or total hip arthroplasty (Table 1). Gamma nail was the most common treatment for trochanteric fractures, and sliding hip screw was used by preference of the surgeon. Only 3 patients were in ASA category I, and most were in ASA category III. Most patients were discharged from the primary hospital to rehabilitation units or health centre hospitals, and only 13 patients were primarily discharged into their own homes (Table 1). Two patients died in the orthopaedic ward. The mean duration of hospitalisation on the orthopaedic ward was 7.7 days. The four-month follow-up was performed using the same questions on ADL functions as on admission. The forms were filled in by the patient or, if the patient was unable to do it, by the study nurse based on a telephone interview. The completeness and validity of the forms completed by the patients were checked by the study nurse from the patient files or operation registers or by talking to the patients themselves or their relatives or the staff of the relevant institution over the telephone. The twelvemonth follow-up was carried out using the same inquiries and procedures as at four months. Mortality data were obtained from the Finnish Census Register using the unique identification codes assigned to each Finnish citizen, and the coverage was 100 % complete for a one-year period. Reoperations were recorded on a special form. Statistical analyses were performed by a statistician using the SPSS statistical software (SPSS for Windows, SPSS inc., Standard Version Release 9.0.1, 1999). The comparison of the different checkpoints was carried out using the analysis of repeated measures, which excludes data on patients not alive at the last follow-up. Statistical significances were calculated using a marginal homogeneity test for categorical variables and Wilcoxon s signed ranks test for continuous variables. A two-sided probability level of 0.05 or less was deemed to indicate statistical significance. RESULTS By one year after the fracture, 23 (15 %) patients had undergone a reoperation, of whom 18 had been reoperated within the first four months (p =0.063). By

3 Four or twelve months follow-up in the evaluation of functional outcome after hip fracture surgery? 61 TABLE 2 Comparison of patients functional capacity. Survivors: patients who survived for 12 months. Prefracture 4 months 12 months p-values for survivors All Survivors All Survivors Survivors pre vs. 4 pre vs vs. 12 months months months Residential status p=0.002 p=0.003 p=0.385 Own home Sheltered housing Institutional care Health centre hospital Permanent hospital inpatient Rehabilitation unit Emergency hospital Walking ability p<0.001 p<0.001 p=0.004 Alone outdoors Outdoors only if accompanied Alone indoors but not outdoors Indoors only if accompanied Unable to walk Walking aids p<0.001 p<0.001 p=0.323 No aids One aid Two aids Frame Wheelchair/bedbound Pain in the injured hip Severe and spontaneous, even at rest Severe when walking and prevents all activity Tolerable, permitting limited activity Occurs only after some activity, disappears quickly at rest Slight or intermittent, alleviated in normal activity No hip pain Unable to answer p<0.001 the four-month follow-up, 29 patients (14.8 %) had died, and by the one-year follow-up, as many as 44 patients (22.4 %) had died (p <0.001). The patients who died between four and twelve months had poorer functional capacity at four months than those who survived for twelve months; the surviving patients were more often living at their own homes (p =0.035), had better walking ability (p =0.004), used less often walking aids (p =0.005) and managed significantly better in some ADL tasks (dressing p=0.013, toileting p=0.028, transportation p=0.030) than the patients who died during the longer followup (Tables 2 and 3). The change in the residential status of the survivors between four months and one year was not statistically significant; all the patients who were able to return to their own homes had done so by four months (Table 2). The patients walking abilities were significantly (p =0.004) restored at the population level between four months and one year, but the changes in individual cases were not always positive (Table 2). Ninety patients (59 %) retained the level of walking ability achieved at four months up to one year; forty patients (26 %) improved their walking ability and the walking ability of twenty-two patients (14 %) deteriorated. The use of walking aids did not decrease significantly after four months (Table 2). Generally, pain in the injured hip became less severe (p <0.001) between the four-month and oneyear follow-ups (Table 2). This was independent of the type of fracture (cervical/trochanteric) and the operation method used. Of the ADL functions, the abilities to dress, eat and cook improved and the need for special transportation services declined marginally but statistically significantly after four months of follow-up, but the abilities to bath, shop, do laundry and manage toileting, household chores and money matters remained unchanged (Table 3). DISCUSSION Findings from the previous studies indicate that most of the recovery of the patients functions takes place by four to six months (Table 4). Magaziner et al. observed only slight additional recovery in a few functional domains at the population level between six and twelve months, while most recovery occurred by six months (12). In a later study he noted that recuperation times were specific to the area of function, ranging from approximately four months for depressive symptoms, upper extremity function and cognition to almost a year for lower extremity function (11). Ceder et al. found that the proportion of patients

4 62 T. Heikkinen, P. Jalovaara TABLE 3 Comparison of patients functional capacity. Survivors: patients who survived for 12 months. Prefracture 4 months 12 months p-values for survivors All Survivors All Survivors Survivors pre vs. 4 pre vs vs. 12 months months months Dressing p <0.001 p<0.001 p=0.013 Able to dress completely without help Needs some help with buttons or zippers Needs assistance with shoes and stockings Needs assistance with up to 3 items Needs to be dressed by others Bathing or showering p<0.001 p<0.001 p=0.601 Able to bath or shower Needs some help in washing a single part of the body, such as back or feet, or needs a bystander Needs assistance in getting in and out of the bathtub Needs assistance in washing one or several parts of the body Always needs to be bathed by others Eating p<0.001 p=0.024 p=0.001 Able to cut food and eat without help Needs help from others to cut hard food Needs assistance in handling food, e.g. buttering bread Needs a lot of help to eat Has to be completely fed by others Toileting Able to get to the toilet, get on and off, manage clothing, etc p<0.001 p<0.001 p=0.877 Needs assistance in getting to and from toilet Needs assistance in getting on and off the toilet and adjusting clothing Needs assistance in cleaning organs of excretion Wears pads or uses a catheter or bedpan at all times Shopping p<0.001 p<0.001 p=0.540 Able to do all shopping without assistance Needs assistance in getting to or returning from shops, can only shop independently for small purchases or is able to shop but gets someone else to do it Needs assistance with selecting shopping, is unsure what he/she needs to buy or must always be accompanied due to physical, psychological or visual impairment Needs help with two or more tasks associated with grocery shopping Completely unable to shop Household chores p<0.001 p<0.001 p=0.477 Able to manage housekeeping alone or with only occasional assistance Able to perform all home maintenance tasks but needs some assistance with, e.g., lifting or is able to do housework but has someone else do it Able to perform only light daily tasks Needs assistance with light household duties Unable to do housework Laundry p <0.001 p<0.001 p=0.089 Able to do laundry Needs assistance in doing or hanging up laundry or is able to do it but somebody else does it or would be able to do laundry if there were a machine at home Able to wash delicates and personals by hand or needs some assistance in loading or unloading the machine Needs a lot of help to do laundry Unable to do laundry To be continued

5 Four or twelve months follow-up in the evaluation of functional outcome after hip fracture surgery? 63 Continued Prefracture 4 months 12 months p-values for survivors All Survivors All Survivors Survivors pre vs. 4 pre vs vs. 12 months months months Preparation of meals p <0.001 p<0.001 p=0.030 Able to prepare meals Able to prepare meals but someone else does it Able to prepare a small meal or sandwich if supplied with ingredients Able only to reheat meals Must have all meals prepared Banking/finances p<0.001 p<0.001 p=0.272 Able to manage all financial matters Needs assistance in going to bank or does billing and banking by mail or cannot go to bank but is able to perform all other financial tasks or is able to do it but someone else does it Able to manage day-to-day purchases, but needs assistance with banking and major purchases Needs to be taken to the bank and requires someone to handle the transactions and all other financial needs Unable to handle financial matters Use of transportation p<0.001 p<0.001 p<0.001 Able to travel independently on public transportation or drive a car Arranges his/her own travel by taxi but does not use bus or train Must always be accompanied due to physical, psychological or visual impairment Travels in taxi or car only with assistance Unable to travel living at their own homes and managing the basic ADL or household functions did not markedly increase between four months and twelve months after the fracture (8). In the study of Koval et al., 84 % of patients, who were home-dwelling before the fracture, had been independent in all basic activities of daily living before the hip fracture (10). The proportion of patients who recovered to the baseline level increased notably between three and six months but only marginally between six and twelve months. Borgquist et al. concluded that ADL, walking ability and household activities remained at the level already achieved within four months after the fracture during the ten-year period for which they prospectively followed up patients admitted from their own homes (7). Our data also suggest that most of the recovery after a hip fracture occurs during the first four months. The proportion of patients living at their own home at the four-month follow-up has varied in different reports from 40 to 82 % (18 22) and the corresponding percentage at one-year follow-up from 44 to 86 % (21 23). Our figures, 58 % and 56 %, fall between these ranges. The marked variation is probably due to cultural differences and the availability of institutional care in different countries. The changes noted in patients residential status after four months in the previous studies are not systematic. Kitamura et al. and Ceder et al. found that the proportion of patients living at their own homes increased slightly but insignificantly between four months and one year (8, 21). On the other hand, in the studies of Thorngren et al., Holmberg et al. and Keene et al., the proportion of patients living at their own homes decreased somewhat between four months and one year (19, 22, 23). Our finding of a 3%decrease is in agreement with the latter reports. Thus, it can be concluded that four months seems to be a sufficient follow-up time for residential status as a functional outcome variable in view of the fact that mortality changes the outcome over time. Comparison of the changes in walking ability between four and twelve months with the previous studies is difficult due to the variation in scoring methods and follow-up points. However, walking ability seems to improve slightly in most reports (Table 4), (7, 11, 12, 21), as it also did in this study. After all, it seems that the majority of patients recover their pre-fracture walking ability within four to six months following the surgery. The use of walking aids is not frequently used as an outcome variable in hip fracture surveys. This variable is strongly influenced by the local and national health care system and resources. Magaziner reported that 26 % of patients were able to walk unaided at two months and 54 % at twelve months (12). In the study of Rogmark et al., the use of walking aids decreased significantly between four and twelve months but increased again between twelve and twenty-four months to the same level as at four months (24). It is notable that, while the present patients walking ability improved between four and

6 64 T. Heikkinen, P. Jalovaara TABLE 4 Changes in patients functionality during the first postoperative year. Walking Eating Toileting Transferring / transportation Dressing Bathing Money / banking Meals / food preparation Shopping Cleaning / doing laundry 0 Follow-up time point (months) Study Ceder b Hardy c Kitamura d Magaziner e Magaziner f Rogmark g Magaziner e Magaziner e Magaziner e Ceder b Decline Kitamura d Magaziner e Magaziner e Magaziner e Ceder b Magaziner e Ceder b Magaziner e Ceder b a Adapted from table 4in [2]. b Adapted from figures 3and 4in[3]. c Adapted from figure 4in [6]. d Adapted from table 2in[13]. e Adapted from figures 2 and3in [21]. f Adapted from figure 1 in [20]. g Adapted from figure 2 in [24].

7 Four or twelve months follow-up in the evaluation of functional outcome after hip fracture surgery? 65 twelve months at the population level, the use of walking aids did not decrease equally. This may imply that the patients become somehow accustomed to their walking aids, that the use of walking aids is not an equally good indicator of function as walking ability, or that the change in walking ability after four months is only marginal. In our study, the patients surviving for one year regained four out of ten ADL functions between four months and one year. Ceder et al. found no remarkable changes in ADL functions between four and twelve months (Table 4) (8). Koval et al. reported that 59 % had recovered to their pre-fracture level of basic ADL at three months, 71 % at six and 73 % at twelve months (10). Their respective percentages in instrumental ADLs were 53 %, 42 % and 48 % (Table 4). Magaziner et al. noted an increase in the proportion of patients who attained their pre-fracture level of functioning between two and six months following discharge in most tasks. Later, between six and twelve months, no change was observed in any task (Table 4) (12). In his later study, recuperation times were 4.3 months for upper extremity physical ADLs and 11 months for lower extremity physical and instrumental ADLs (11). However, recovery appeared to be concentrated in the first six months, with only minor gain observed thereafter (11). According to Borgquist et al., ADL functions remained fairly stable from four months postoperatively to up to ten year (Table 4) (7). Thus, it seems that most ADL abilities are regained during the four to six months, and that only minor further improvement may occur hereafter, which is in agreement with our results. Pain in the fractured hip is a relatively rarely reported outcome variable in hip fracture surveys. While it is a very subjective experience and therefore difficult to study and interpret, it is important for frail, elderly patients even when it prevents activity, thereby impeding rehabilitation. In our study, the proportion of painless patients increased after four months and the proportion of patients who experienced pain limiting their activity decreased similarly by one year. Likewise, Rogmark et al. found that the percentage of patients reporting pain on walking decreased steadily after four months up to twenty-four months (24). The proportion of painless patients at one year in our study is in line with the % reported earlier (25). At any rate, pain in the injured hip seems to subside slowly after the operation. However, it does not seem to reflect in patient s function to great extent. It is noteworthy that, in most of the previous studies using several checkpoints, the results are given as fractions of surviving patients at each follow-up point. Only Koval et al. and Magaziner et al. seem to have evaluated at each checkpoint the function of only those patients who were alive at the end of the study, as done here (10 12). We observed that the patients who died between four and twelve months were among those with the poorest functional capacity in the four-month evaluation. If this fact is overlooked, it may seem that the functional capacity of the patient population improves markedly over time. Therefore, attention should be paid to the use of an adequate statistical method, i.e. analysis of repeated measures, which excludes data on patients not alive at the last follow-up and reporting methods used accurately. Functional outcome is very important in the elderly. It can be assessed using inquiries or telephone interviews, which helps to avoid stressful visits to the outpatient clinic by the physically impaired patient. Here, we performed the follow-up using questionnaires completed by the patient or by a research nurse during a telephone interview. A similar approach has been used in many previous studies (15, 16, 26 28). There is no evidence to suggest that a personal interview at a follow-up visit and a clinical examination and radiographs would improve the reliability of the functional outcome measurements used in this study. It is concluded that residential status and 6 out of 10 ADL domains do not change, while the remaining 4 ADL functions improve slightly after the first four postoperative months. Walking ability also improves marginally; a few patients walking only indoors at four months also walked out of doors at twelve months. When the outcomes at the different time points are compared, only the patients alive at the last follow-up point should be included in the analysis, and analysis of repeated measures should be used. Due to high mortality and deterioration of functioning, both caused by advanced age, no steady state in function is ever reached, and four-month follow-up is justified as the shortest feasible alternative. ACKNOWLEDGMENTS We thank Mrs. Eila Haapakoski for her persistence and honest work with the data collection, Mr. Mikko Simonaho for help with the statistical analysis and The Finnish Medical Society Duodecim for giving a grant to TH. REFERENCES 0 1. Dahl E: Mortality and life expectancy after hip fractures. Acta Orthop Scand 1980;51(1): Forsén L, Søgaard AJ, Meyer HE, Edna TH, Kopjar B: Survival after hip fracture: Short- and long-term excess mortality according to age and gender. Osteoporos Int 1999;10: Holmberg S, Conradi P, Kalen R, Thorngren KG: Mortality after cervical hip fracture patients followed for 6 years. Acta Orthop Scand 1986;57(1): Kreutzfeldt J, Haim M, Bach E: Hip fracture among the elderly in a mixed urban and rural population. Age Ageing 1984; 13(2): Magaziner J, Simonsick EM, Kashner TM, Hebel JR, Kenzora, JE: Survival experience of aged hip fracture patients. Am J Public Health 1989;79(3): Wolinsky FD, Fitzgerald JF, Stump TE: The effect of hip fracture on mortality, hospitalization, and functional status: a prospective study. Am J Public Health 1997;87(3): Borgquist L, Ceder L, Thorngren KG: Function and social status 10 years after hip fracture: Prospective follow-up of 103 patients. Acta Orthop Scand 1990;61(5): Ceder L, Thorngren KG, Wallden B: Prognostic indicators and early home rehabilitation in elderly patients with hip fractures. Clin Orthop 1980;152: Koot VCM, Peeters PHM, de Jong JR, Clevers GJ, van der

8 66 T. Heikkinen, P. Jalovaara Werken C: Functional results after treatment of hip fracture: A multicentre, prospective study in 215 patients. Eur J Surg 2000;166: Koval KJ, Skovron ML, Aharonoff GB, Zuckerman JD: Predictors of functional recovery after hip fracture in the elderly. Clin Orthop 1998;348: Magaziner J, Hawkes W, Hebel JR, Zimmerman SI, Fox KM, Dolan M, Felsenthal G, Kenzora J: Recovery from hip fracture in eight areas of function. J Gerontol Med Sci 2000;55A(9): M Magaziner J, Simonsick EM, Kashner TM, Hebel JR, Kenzora, JE: Predictors of functional recovery one year following hospital discharge for hip fracture: a prospective study. J Gerontol 1990;45(3):M101 M Thorngren KG: Optimal treatment of hip fractures. Acta Orthop Scand Suppl 1991;241: Parker M J, Currie C T, Mountain J A, Thorngren KG: Standardised Audit for Hip Fractures in Europe (SAHFE). Hip Int 1998;8: Rosell PA, Parker MJ: Functional outcome after hip fracture. A 1-year prospective outcome study of 275 patients. Injury 2003;34: Scheerlinck T, Opdeweegh L, Vaes P, Opdecam P: Hip fracture treatment: outcome and socio-economic aspects. A oneyear survey in a Belgian University Hospital. Acta Orthop Belg 2003;69: American Society of Anaesthesiologists: New classification of physical status. Anaesthesiology 1963;24: Heikkinen T, Parker M, Jalovaara P: Hip fractures in Finland and Great Britain a comparison of patient characteristics and outcomes. Int Orthop 2001;25: Heikkinen T, Wingstrand H, Partanen J, Thorngren KG, Jalovaara P: Hemiarthroplasty or osteosynthesis in cervical hip fractures: matched-pair analysis in 892 patients. Arch Orthop Trauma Surg 2002;122: Jalovaara P, Berglund-Rödén M, Wingstrand H, Thorngren KG: Treatment of hip fracture in Finland and Sweden: Prospective comparison of 788 cases in three hospitals. Acta Orthop Scand 1992;63(5): Kitamura S, Hasegava Y, Suzuki Sasaki R, Iwata H, Wingstrand H, Thorngren KG: Functional outcome after hip fracture in Japan. Clin Orthop 1998;348: Thorngren KG, Ceder L, Svensson K: Predicting result of rehabilitation after hip fracture. A ten-year follow-up study. Clin Orthop 1991;(287): Keene GS, Parker MJ, Pryor GA: Mortality and morbidity after hip fracture. BMJ 1993;307: Rogmark C, Carlsson Å, Johnell O, Sernbo I: A prospective randomized trial of internal fixation versus arthroplasty for displaced fractures of the neck of the femur: Functional outcome for 450 patients at two years. J Bone Joint Surg (Br) 2002; 84-B(2): Ravikumar KJ, Marsh G: Internal fixation versus hemiarthroplasty versus total hip arthroplasty for displaced subcapital fractures of femur 13 year results of a prospective randomised study. Injury 2000;31(10): Heikkinen T, Willig R, Hänninen A, Koskinen K, Mannismäki P, Alavaikko A, von und zu Fraunberg M, Jalovaara P: Hip fractures in Finland a comparison of patient characteristics and outcomes in six hospitals. Scand J Surg 2004;93(3): Partanen J, Jalovaara P: Functional comparison between uncemented Austin-Moore hemiarthroplasty and osteosynthesis with three screws in displaced femoral neck fractures a matched-pair study of 168 patients. Int Orthop 2004;28(1): Partanen J, Saarenpää I, Heikkinen T, Wingstrand H, Thorngren KG, Jalovaara P: Functional outcome after displaced femoral neck fractures treated with osteosynthesis or hemiarthroplasty: a matched-pair study of 714 patients. Acta Orthop Scand 2002;73(5): Received: May 24, 2004 Accepted: December 27, 2004