- Email: [email protected]
New scoring system at admission to predict walking ability at discharge for patients with hip fracture
G Model
ARTICLE IN PRESS
OTSR-2117; No. of Pages 4
Orthopaedics & Traumatology: Surgery & Research xxx (2018) xxx–xxx
Contents lists available at ScienceDirect
Orthopaedics & Traumatology: Surgery & Research journal homepage: www.elsevier.com
Original article
New scoring system at admission to predict walking ability at discharge for patients with hip fracture Takayoshi Oba a,∗ , Hiroyuki Makita a , Yutaka Inaba b , Hayato Yamana c , Tomoyuki Saito b a Department of Orthopaedic Surgery, Kanagawa Prefectural Ashigarakami Hospital, 866-1 Matusda-souryou, Matsuda-machi, Ashigarakami-gun, 2580003 Kanagawa, Japan b Department of Orthopaedic Surgery, Yokohama City University School of Medicine, 3–9 Fukuura, Kanazawa-ku, Yokohama, 2360004 Kanagawa, Japan c Department of Clinical Epidemiology and Health Economics, School of Public Health, The University of Tokyo, 7–3–1 Hongo Bunkyo-ku, 1113033 Tokyo, Japan
a r t i c l e
i n f o
Article history: Received 8 January 2018 Accepted 19 July 2018 Available online xxx Keywords: Hip fracture Scoring system Albumin Walking ability Predictive factors
a b s t r a c t Introduction: A reliable scoring system that predicts the walking ability of hip fracture patients would be useful for clinicians. Here we developed a scoring system for hip fracture patients and evaluated its predictive ability. Hypothesis: We hypothesized that age, sex, presence of dementia, walking ability before the injury, fracture type, serum hemoglobin level, serum albumin level and interval in days between admission and surgery would be the predictive factors of the walking ability at discharge. Material and methods: Data from 409 patients who underwent hip fracture surgery were included. We analyzed factors that affected walking ability and developed a scoring system that predicts the probability of walking unaided or with a cane at discharge. Results: The mean age of the patients was 81.3 years. A total of 164 (40%) patients could walk unaided or with a cane at discharge. Multivariate logistic regression analysis showed that the obstructive factors for the ability to walk unaided or with a cane at discharge were older age (odds ratio [OR] = 0.962, p = 0.002), dementia (OR = 0.126, p < 0.001), use of a cane before injury (OR = 0.396, p < 0.001), trochanteric fracture (OR = 0.571, p = 0.027) and low serum albumin level (OR = 4.15, p < 0.001) at admission. The scoring system used the following formula: Score = 5 − 0.04 × age + albumin − 2 (with dementia) − 1 (with use of a cane before injury) − 1 (with trochanteric fracture). The C-statistics for the scoring system was 0.81 (95% confidence interval, 0.77–0.85). Discussion: This newly developed scoring system of information at admission predicted the discharge mobility of hip fracture patients. In addition to the previously known risk factors, serum albumin level at admission was detected as a new predictor for mobility at discharge. Level of proof: IV, retrospective study. © 2018 Elsevier Masson SAS. All rights reserved.
1. Introduction Patients who experience hip fracture are increasing, especially in countries with increasing populations of elderly people such as Asia and Latin America [1]. Hip fracture is associated with high mortality and morbidity rates. The ability to make early predictions of a patient’s postoperative outcomes would enable health care providers to choose a discharge location early and provide the family time to prepare for discharge. Previous studies investigated the risk factors of morbidity and mortality in hip fracture patients [2–9]. Some focused on the
economical aspect and developed scoring systems to predict the discharge location [10–14]. However, discharge locations would vary among countries, societies and hospitals. The purpose of this study was to clarify factors that affect the postoperative walking ability of hip fracture patients at the time of discharge. Furthermore, we developed a scoring system that predicts the probability of walking after a hip fracture surgery and evaluated its predictive ability. 2. Material and methods 2.1. Patients
∗ Corresponding author. E-mail address: [email protected] (T. Oba).
The study was approved by our hospital’s ethics committee. We retrospectively evaluated data from 593 patients with hip
https://doi.org/10.1016/j.otsr.2018.07.024 1877-0568/© 2018 Elsevier Masson SAS. All rights reserved.
Please cite this article in press as: Oba T, et al. New scoring system at admission to predict walking ability at discharge for patients with hip fracture. Orthop Traumatol Surg Res (2018), https://doi.org/10.1016/j.otsr.2018.07.024
G Model OTSR-2117; No. of Pages 4
ARTICLE IN PRESS T. Oba et al. / Orthopaedics & Traumatology: Surgery & Research xxx (2018) xxx–xxx
2
Table 1 Characteristics of the cohort stratified by mobility at discharge. Characteristics
Age, yearsa Male sexb Dementiab Use of a cane before injuryb Trochanteric fractureb Hemoglobin, g/dLa Albumin, g/dLa Time until surgery, daysa Length of hospitalization, daysa a b c
p-valuec
All patients
Mobility at discharge
(n = 409)
Unaided or with a cane(n = 164)
Unable to walk with a cane(n = 245)
81.3 (11) 78 (19) 91 (22) 171 (42) 205 (50) 11.8 (1.7) 3.6 (0.5) 4.5 (3.1) 31.7 (14)
76.9 (11) 30 (18) 10 (6.1) 44 (27) 61 (37) 12.2 (1.8) 3.8 (0.4) 4.3 (2.9) 32.8 (12)
84.1 (9.4) 48 (20) 81 (33) 127 (52) 144 (59) 11.5 (1.6) 3.5 (0.4) 4.7 (3.2) 30.9 (15)
<0.001 0.743 <0.001 <0.001 <0.001 <0.001 <0.001 0.160 0.157
Data presented as mean (SD). Data presented as n (%). Bivariate analysis.
fracture who were admitted to a single institution (Kanagawa, Japan) between January 2012 and December 2016. Of these 593 patients, 49 who were treated non-surgically because of their comorbidities were excluded. The 130 patients who were not able to walk even with a cane before the injury were also excluded. This was considered necessary because the probability of patients who could not walk prior to the injury being able to walk after the surgery is very low [15]. One patient in whom hip fracture was complicated by an ankle fracture and four for whom full blood testing results were not available were also excluded. The remaining 409 patients were included in this study.
2.2. Treatment Patients who had a femoral neck fracture were evaluated by using Garden classification [16]. Patients with Garden Type I or II fractures were treated using cannulated cancellous hip screws, while patients with Garden Type III or IV fractures were mostly treated with hemiarthroplasty. Patients with a trochanteric fracture were treated with proximal femoral nails. After surgery, rehabilitation was started in all patients immediately and continued until discharge. In all patients, full weight bearing was allowed from the next day of the surgery. The patients underwent a stepwise program to attain the ability to use a wheelchair and then walk using parallel bars, a walker, a cane, and finally unaided. Patients were discharged to home or an alternative location for further treatment if necessary.
2.3. Statistical analysis To analyze the factors that affected the walking ability at discharge, the following factors were chosen as predictor variables: age, sex, presence of dementia, walking ability before the injury (unaided or use of a cane), fracture type (trochanteric or femoral neck), serum hemoglobin level (g/dL), serum albumin level (g/dL) and interval in days between admission and surgery. The selection of these variables was based on previous studies and clinical experience [2–7,17]. All of this information could be obtained at the time of admission through patient/family interviews and blood tests. The ability to walk unaided or with a cane at discharge was chosen as the dependent variable. First, the variables above and the length of hospitalization were analyzed on bivariate analysis; then, the logistic regression model was developed. Using the method of Sullivan et al., a prediction scoring system was created by adding the points corresponding to the regression coefficient of the significant (p < 0.05) variables from the logistic regression model [18]. We added the number derived from the lowest score to the formula to adjust the scores to be positive.
C-statistics was used to evaluate the predictive ability of the model. We also stratified patients based on score and evaluated the observed probability of walking unaided or with a cane. All statistical analyses were performed using SPSS version 23 (IBM, Chicago, IL, USA). p-values < 0.05 were considered significant. 3. Results The mean patient age at the time of admission of the cohort was 81.3 ± 11 years. The ambulatory ability at discharge were the following: walk unaided, walk with a cane, walk with a walker, walk in parallel bars, wheel chair and bedridden (n [%]; 28 [6.9], 136 [33], 100 [24], 49 [12], 79 [19] and 17 [4.2], respectively). One hundred and sixty-four of the 409 (40%) patients could walk unaided or with a cane independently at discharge. The characteristics of the patients stratified by walking ability at discharge are shown in Table 1. On bivariate analysis, the patients who could not walk unaided or with a cane at discharge had the following characteristics: older age, presence of dementia, use of a cane before injury, higher proportion of trochanteric fractures and lower mean hemoglobin and serum albumin levels at admission. However, there were no significant differences in sex, days until surgery and length of stay between the patients who walked unaided or with a cane and those who did not (Table 1). Multivariable logistic regression analysis showed that the obstructive factors, at admission, for the ability to walk unaided or with a cane at discharge were older age regression coefficient ([ˇ] = − 0.039, odds ratio [OR] = 0.96, p = 0.002), presence of dementia (ˇ = − 2.069, OR = 0.126, p < 0.001), use of a cane before injury (ˇ = − 0.925, OR = 0.396, p < 0.001), trochanteric fracture (ˇ = − 0.560, OR = 0.571, p = 0.027) and low serum albumin level (ˇ = 1.423, OR = 4.15, p < 0.001) (Table 2). The prediction scoring formula was the following: score = − 0.04 × age + albumin − 2 (with dementia) − 1 (with use of a cane before injury) − 1 (with trochanteric fracture). The patients’ outcome with scores is shown in Table 3. The original scores ranged from − 4.7 to 2.5; thus, 5 points were added to make the scores positive. The receiver operating characteristic curve of the scoring system is shown in Fig. 1, and the C-statistics for the scoring system was 0.81 (95% confidence interval [CI], 0.77–0.85). The cut-off point for balancing sensitivity and specificity was best at 4.2, with 75.6% sensitivity and 74.3% specificity. 4. Discussion In this study we developed a scoring system usable at admission to predict the walking ability at discharge of patients with hip fracture, which demonstrated high predictive ability in a large cohort.
Please cite this article in press as: Oba T, et al. New scoring system at admission to predict walking ability at discharge for patients with hip fracture. Orthop Traumatol Surg Res (2018), https://doi.org/10.1016/j.otsr.2018.07.024
G Model
ARTICLE IN PRESS
OTSR-2117; No. of Pages 4
T. Oba et al. / Orthopaedics & Traumatology: Surgery & Research xxx (2018) xxx–xxx
3
Table 2 Result of multivariate logistic regression analysis for walking unaided or with a cane at discharge. Variable
Beta coefficient
Odds ratio
95% confidence interval
p-value
Age Male sex Dementia Use of a cane before injury Trochanteric fracture Hemoglobin Albumin Days until surgery
−0.04 −0.41 −2.10 −0.93 −0.56 −0.03 1.42 −0.02
0.96 0.67 0.13 0.40 0.57 0.97 4.15 0.98
0.94–0.99 0.36–1.23 0.06–0.27 0.24–0.65 0.35–0.94 0.83–1.14 2.19–7.85 0.91–1.06
0.002 0.196 <0.001 <0.001 0.027 0.737 <0.001 0.658
Table 3 Prediction scoring system for walking ability at discharge. Scoring formula
Outcome
Factors
Points
Total score
n
Patients unaided or with a cane
Observed probability(%)
Age, (years) Albumin, (g/dl) Dementia Use of a cane before injury Trochanteric fracture
×(−0.04) ×1 −2 −1 −1 +5a
0–0.9 1–1.9 2–2.9 3–3.9 4–4.9 5–5.9 6–6.9 7–7.9
11 26 69 97 95 66 38 7
0 1 5 27 53 42 30 6
0 3.8 7.2 28 56 64 79 86
Total score a
Five points were added for the convenience of deriving positive values.
Fig. 1. Receiver operating characteristic curve of the prediction score. The Cstatistics for the scoring system was 0.81 (95% confidence interval, 0.77–0.85).
The information required to calculate the score can be obtained at the time of admission through patient/family interviews and blood tests. Therefore, this scoring system could be widely utilized to predict the mobility outcomes of hip fracture patients. Predicting the discharge walking ability of hip fracture patients at admission will enable concerned patients and family members to be informed early, providing them with time to prepare for discharge. Furthermore, the prediction will help health care providers to appropriately choose the discharge location which differs among each societies. Several studies reported factors at admission that could predict mobility at discharge using multiple regression analysis. Kristensen et al. reported obstructive factors for attaining basic mobility after surgery in 280 hip fracture patients. Hagino et al.
reported obstructive factors for walking independently or with aids at discharge in 186 hip fracture patients [5,6]. Our study was based on a larger number of patients and also different in the outcome. The outcome in our study did not include those who could walk independently with only by a walker. The outcome was chosen based on the fact that the ability to walk unaided or with a cane at discharge was associated with a better prognosis in mobility at 12 months after surgery, reported by Fukui et al. in a prospective study with a cohort of 650 hip fracture patients [7]. Older age, presence of dementia, disability in walking before the injury and trochanteric fracture, which we detected as obstructive factors for walking ability at discharge, were similar in these studies [5,6]. However, we also identified low serum albumin level at admission as a new predictor for discharge mobility in hip fracture patients. The OR for not being able to walk unaided or with an aid at discharge was 4 per 1 g/dL decrease in serum albumin at admission. There are several studies about serum albumin relatable to our findings. Wakabayashi et al. reported that malnutrition, including low serum albumin levels, was associated with poor general rehabilitation outcomes for elderly inpatients [19]. Eneroth et al. reported that nutritional supplementation decreased hip fracture–related complications in their prospective randomized controlled trial [20]. The importance of nutrition status in hip fracture patients should be recognized. Mean age, proportion of males, and proportion of trochanteric fractures in our study were similar to those in several previous studies (81–83 years, 18–27%, and 47–61%, respectively) [5–7,11,12]. This suggests that our cohort was representative of hip fracture patients. The limitations of this study must be acknowledged. First, it was retrospective. Second, it was conducted in a single institution in Japan and all patients were Japanese. Hip fracture patients in Japan often undergo rehabilitation in the hospital in which they undergo surgery. The scoring system may be more applicable in countries in similar situations than in countries that discharge patients sooner. Thus, further research is required to validate the scoring system in other situations. Third, dementia was not assessed quantitatively. In previous studies, the Short Portable Mental Status Questionnaire was used and reported to be a useful measurement, which
Please cite this article in press as: Oba T, et al. New scoring system at admission to predict walking ability at discharge for patients with hip fracture. Orthop Traumatol Surg Res (2018), https://doi.org/10.1016/j.otsr.2018.07.024
G Model OTSR-2117; No. of Pages 4
ARTICLE IN PRESS T. Oba et al. / Orthopaedics & Traumatology: Surgery & Research xxx (2018) xxx–xxx
4
could have been appropriate [21]. However, scoring systems must be simple and quick to perform in daily practice. Finally, although we assessed numerous variables, the effects of unmeasured confounders cannot be ruled out. 5. Conclusion We report the development of a new scoring system used at admission for hip fracture patients that predicts mobility at discharge. Furthermore, we identified serum albumin level at admission as a new predictor for mobility at discharge. Future challenges include utilizing the scoring system and further evaluating its accuracy and usability. Disclosure of interest The authors declare that they have no competing interest. Funding None. Contribution All authors contributed to the study concept and design. T. Oba drafted the manuscript. T. Oba and H. Makita collected the data. T. Oba and H. Yamana performed the statistical analyses. H. Yamana, Y. Inaba, H. Yamana and T. Saito revised the manuscript critically for important intellectual content. References [1] Cooper C, Campion G, Melton LJ. Hip fractures in the elderly: a world-wide projection. Osteoporos Int 1992;2:285–9. [2] Parker MJ, Palmer CR. A new mobility score for predicting mortality after hip fracture. J Bone Joint Surg Br 1993;75:797–8. [3] Koval KJ, Skovron ML, Aharonoff GB, Meadows SE, Zuckerman JD. Ambulatory ability after hip fracture. A prospective study in geriatric patients. Clin Orthop Relat Res 1995;310:150–9.
[4] Cornwall R, Gilbert MS, Koval KJ, Strauss E, Siu AL. Functional outcomes and mortality vary among different types of hip fractures: a function of patient characteristics. Clin Orthop Relat Res 2004;425:64–71. [5] Hagino T, Ochiai S, Wako M, Sato E, Maekawa S, Senga S, et al. A simple scoring system to predict ambulation prognosis after hip fracture in the elderly. Arch Orthop Trauma Surg 2007;127:603–6. [6] Kristensen MT, Foss NB, Ekdahl C, Kehlet H. Prefracture functional level evaluated by the New Mobility Score predicts in-hospital outcome after hip fracture surgery. Acta Orthop 2010;81:296–302. [7] Fukui N, Watanabe Y, Nakano T, Sawaguchi T, Matsushita T. Predictors for ambulatory ability and the change in ADL after hip fracture in patients with different levels of mobility before injury: a 1 year prospective cohort study. J Orthop Trauma 2012;26:163–71. [8] Simon P, Gouin F, Veillard D, Laffargue P, Ehlinger M, Bel JC, et al. Femoral neck fractures in patients over 50 years old. Rev Chir Orthop Reparatrice Appar Mot 2008;94:108–32 [Epub 2008 Sep 19]. [9] Reina N, Bonnevialle P, Rubens DB, Adam P, Loubignac F, Favier T, et al. Internal fixation of intra-capsular proximal femoral fractures in patients older than 80 years: still relevant? Multivariate analysis of a prospective multicentre cohort. Orthop Traumatol Surg Res 2017;103:3–7 [Epub 2016 Dec 2]. [10] Thorngren KG, Ceder L, Svensson K. Predicting results of rehabilitation after hip fracture. A ten-year follow-up study. Clin Orthop Relat Res 1993;287:76–81. [11] Svensson O, Strömberg L, Öhlén G, Lindgren U. Prediction of outcome of hip fractures in elderly patients. J Bone Joint Surg Br 1996;78:115–8. [12] Hagino T, Ochiai S, Sato E, Watanabe Y, Senga S, Haro H. Prognostic prediction in patients with hip fracture: risk factors predicting difficulties with discharge to own home. J Orthop Traumatol 2011;12:77–80. [13] Vochteloo AJH, Tuinebreijer WE, Maier AB, Nelissen RGH, Bloem RM, Pilot P. Predicting discharge location of hip fracture patients; the new discharge of hip fracture patients score. Int Orthop 2012;36:1709–14. [14] Salar O, Baker PN, Forward DP, Ollivere BJ, Weerasuriya N, Moppett IK, et al. Predictors of direct home discharge following fractured neck of femur. Ann R Coll Surg Engl 2017;99:444–51. [15] Hay D, Parker MJ. Hip fracture in the immobile patient. J Bone Joint Surg Br 2003;85:1037–9. [16] Garden RS. Low-angle fixation in fractures of the femoral neck. J Bone Joint Surg 1961;43:647–63. [17] Hagino T, Ochiai S, Sato E, Maekawa S, Wako M, Haro H. The relationship between anemia at admission and outcome in patients older than 60 years with hip fracture. J Orthop Traumatol 2009;10:119–22. [18] Sullivan LM, Massaro JM, D’Agostino Sr RB. Presentation of multivariate data for clinical use: the Framingham Study risk score functions. Stat Med 2004;23:1631–60. [19] Wakabayashi H, Sashika H. Malnutrition is associated with poor rehabilitation outcome in elderly patients with hospital-associated deconditioning: a prospective cohort study. J Rehabil Med 2014;46:277–82. [20] Eneroth M, Olsson UB, Thorngren KG. Nutritional supplementation decreases hip fracture-related complications. Clin Orthop 2006;451:212–7. [21] Söderqvist A, Miedel R, Ponzer S, Tidermark J. The influence of cognitive function on outcome after a hip fracture. J Bone Joint Surg Am 2006;88:2115–23.
Please cite this article in press as: Oba T, et al. New scoring system at admission to predict walking ability at discharge for patients with hip fracture. Orthop Traumatol Surg Res (2018), https://doi.org/10.1016/j.otsr.2018.07.024
ARTICLE IN PRESS
OTSR-2117; No. of Pages 4
Orthopaedics & Traumatology: Surgery & Research xxx (2018) xxx–xxx
Contents lists available at ScienceDirect
Orthopaedics & Traumatology: Surgery & Research journal homepage: www.elsevier.com
Original article
New scoring system at admission to predict walking ability at discharge for patients with hip fracture Takayoshi Oba a,∗ , Hiroyuki Makita a , Yutaka Inaba b , Hayato Yamana c , Tomoyuki Saito b a Department of Orthopaedic Surgery, Kanagawa Prefectural Ashigarakami Hospital, 866-1 Matusda-souryou, Matsuda-machi, Ashigarakami-gun, 2580003 Kanagawa, Japan b Department of Orthopaedic Surgery, Yokohama City University School of Medicine, 3–9 Fukuura, Kanazawa-ku, Yokohama, 2360004 Kanagawa, Japan c Department of Clinical Epidemiology and Health Economics, School of Public Health, The University of Tokyo, 7–3–1 Hongo Bunkyo-ku, 1113033 Tokyo, Japan
a r t i c l e
i n f o
Article history: Received 8 January 2018 Accepted 19 July 2018 Available online xxx Keywords: Hip fracture Scoring system Albumin Walking ability Predictive factors
a b s t r a c t Introduction: A reliable scoring system that predicts the walking ability of hip fracture patients would be useful for clinicians. Here we developed a scoring system for hip fracture patients and evaluated its predictive ability. Hypothesis: We hypothesized that age, sex, presence of dementia, walking ability before the injury, fracture type, serum hemoglobin level, serum albumin level and interval in days between admission and surgery would be the predictive factors of the walking ability at discharge. Material and methods: Data from 409 patients who underwent hip fracture surgery were included. We analyzed factors that affected walking ability and developed a scoring system that predicts the probability of walking unaided or with a cane at discharge. Results: The mean age of the patients was 81.3 years. A total of 164 (40%) patients could walk unaided or with a cane at discharge. Multivariate logistic regression analysis showed that the obstructive factors for the ability to walk unaided or with a cane at discharge were older age (odds ratio [OR] = 0.962, p = 0.002), dementia (OR = 0.126, p < 0.001), use of a cane before injury (OR = 0.396, p < 0.001), trochanteric fracture (OR = 0.571, p = 0.027) and low serum albumin level (OR = 4.15, p < 0.001) at admission. The scoring system used the following formula: Score = 5 − 0.04 × age + albumin − 2 (with dementia) − 1 (with use of a cane before injury) − 1 (with trochanteric fracture). The C-statistics for the scoring system was 0.81 (95% confidence interval, 0.77–0.85). Discussion: This newly developed scoring system of information at admission predicted the discharge mobility of hip fracture patients. In addition to the previously known risk factors, serum albumin level at admission was detected as a new predictor for mobility at discharge. Level of proof: IV, retrospective study. © 2018 Elsevier Masson SAS. All rights reserved.
1. Introduction Patients who experience hip fracture are increasing, especially in countries with increasing populations of elderly people such as Asia and Latin America [1]. Hip fracture is associated with high mortality and morbidity rates. The ability to make early predictions of a patient’s postoperative outcomes would enable health care providers to choose a discharge location early and provide the family time to prepare for discharge. Previous studies investigated the risk factors of morbidity and mortality in hip fracture patients [2–9]. Some focused on the
economical aspect and developed scoring systems to predict the discharge location [10–14]. However, discharge locations would vary among countries, societies and hospitals. The purpose of this study was to clarify factors that affect the postoperative walking ability of hip fracture patients at the time of discharge. Furthermore, we developed a scoring system that predicts the probability of walking after a hip fracture surgery and evaluated its predictive ability. 2. Material and methods 2.1. Patients
∗ Corresponding author. E-mail address: [email protected] (T. Oba).
The study was approved by our hospital’s ethics committee. We retrospectively evaluated data from 593 patients with hip
https://doi.org/10.1016/j.otsr.2018.07.024 1877-0568/© 2018 Elsevier Masson SAS. All rights reserved.
Please cite this article in press as: Oba T, et al. New scoring system at admission to predict walking ability at discharge for patients with hip fracture. Orthop Traumatol Surg Res (2018), https://doi.org/10.1016/j.otsr.2018.07.024
G Model OTSR-2117; No. of Pages 4
ARTICLE IN PRESS T. Oba et al. / Orthopaedics & Traumatology: Surgery & Research xxx (2018) xxx–xxx
2
Table 1 Characteristics of the cohort stratified by mobility at discharge. Characteristics
Age, yearsa Male sexb Dementiab Use of a cane before injuryb Trochanteric fractureb Hemoglobin, g/dLa Albumin, g/dLa Time until surgery, daysa Length of hospitalization, daysa a b c
p-valuec
All patients
Mobility at discharge
(n = 409)
Unaided or with a cane(n = 164)
Unable to walk with a cane(n = 245)
81.3 (11) 78 (19) 91 (22) 171 (42) 205 (50) 11.8 (1.7) 3.6 (0.5) 4.5 (3.1) 31.7 (14)
76.9 (11) 30 (18) 10 (6.1) 44 (27) 61 (37) 12.2 (1.8) 3.8 (0.4) 4.3 (2.9) 32.8 (12)
84.1 (9.4) 48 (20) 81 (33) 127 (52) 144 (59) 11.5 (1.6) 3.5 (0.4) 4.7 (3.2) 30.9 (15)
<0.001 0.743 <0.001 <0.001 <0.001 <0.001 <0.001 0.160 0.157
Data presented as mean (SD). Data presented as n (%). Bivariate analysis.
fracture who were admitted to a single institution (Kanagawa, Japan) between January 2012 and December 2016. Of these 593 patients, 49 who were treated non-surgically because of their comorbidities were excluded. The 130 patients who were not able to walk even with a cane before the injury were also excluded. This was considered necessary because the probability of patients who could not walk prior to the injury being able to walk after the surgery is very low [15]. One patient in whom hip fracture was complicated by an ankle fracture and four for whom full blood testing results were not available were also excluded. The remaining 409 patients were included in this study.
2.2. Treatment Patients who had a femoral neck fracture were evaluated by using Garden classification [16]. Patients with Garden Type I or II fractures were treated using cannulated cancellous hip screws, while patients with Garden Type III or IV fractures were mostly treated with hemiarthroplasty. Patients with a trochanteric fracture were treated with proximal femoral nails. After surgery, rehabilitation was started in all patients immediately and continued until discharge. In all patients, full weight bearing was allowed from the next day of the surgery. The patients underwent a stepwise program to attain the ability to use a wheelchair and then walk using parallel bars, a walker, a cane, and finally unaided. Patients were discharged to home or an alternative location for further treatment if necessary.
2.3. Statistical analysis To analyze the factors that affected the walking ability at discharge, the following factors were chosen as predictor variables: age, sex, presence of dementia, walking ability before the injury (unaided or use of a cane), fracture type (trochanteric or femoral neck), serum hemoglobin level (g/dL), serum albumin level (g/dL) and interval in days between admission and surgery. The selection of these variables was based on previous studies and clinical experience [2–7,17]. All of this information could be obtained at the time of admission through patient/family interviews and blood tests. The ability to walk unaided or with a cane at discharge was chosen as the dependent variable. First, the variables above and the length of hospitalization were analyzed on bivariate analysis; then, the logistic regression model was developed. Using the method of Sullivan et al., a prediction scoring system was created by adding the points corresponding to the regression coefficient of the significant (p < 0.05) variables from the logistic regression model [18]. We added the number derived from the lowest score to the formula to adjust the scores to be positive.
C-statistics was used to evaluate the predictive ability of the model. We also stratified patients based on score and evaluated the observed probability of walking unaided or with a cane. All statistical analyses were performed using SPSS version 23 (IBM, Chicago, IL, USA). p-values < 0.05 were considered significant. 3. Results The mean patient age at the time of admission of the cohort was 81.3 ± 11 years. The ambulatory ability at discharge were the following: walk unaided, walk with a cane, walk with a walker, walk in parallel bars, wheel chair and bedridden (n [%]; 28 [6.9], 136 [33], 100 [24], 49 [12], 79 [19] and 17 [4.2], respectively). One hundred and sixty-four of the 409 (40%) patients could walk unaided or with a cane independently at discharge. The characteristics of the patients stratified by walking ability at discharge are shown in Table 1. On bivariate analysis, the patients who could not walk unaided or with a cane at discharge had the following characteristics: older age, presence of dementia, use of a cane before injury, higher proportion of trochanteric fractures and lower mean hemoglobin and serum albumin levels at admission. However, there were no significant differences in sex, days until surgery and length of stay between the patients who walked unaided or with a cane and those who did not (Table 1). Multivariable logistic regression analysis showed that the obstructive factors, at admission, for the ability to walk unaided or with a cane at discharge were older age regression coefficient ([ˇ] = − 0.039, odds ratio [OR] = 0.96, p = 0.002), presence of dementia (ˇ = − 2.069, OR = 0.126, p < 0.001), use of a cane before injury (ˇ = − 0.925, OR = 0.396, p < 0.001), trochanteric fracture (ˇ = − 0.560, OR = 0.571, p = 0.027) and low serum albumin level (ˇ = 1.423, OR = 4.15, p < 0.001) (Table 2). The prediction scoring formula was the following: score = − 0.04 × age + albumin − 2 (with dementia) − 1 (with use of a cane before injury) − 1 (with trochanteric fracture). The patients’ outcome with scores is shown in Table 3. The original scores ranged from − 4.7 to 2.5; thus, 5 points were added to make the scores positive. The receiver operating characteristic curve of the scoring system is shown in Fig. 1, and the C-statistics for the scoring system was 0.81 (95% confidence interval [CI], 0.77–0.85). The cut-off point for balancing sensitivity and specificity was best at 4.2, with 75.6% sensitivity and 74.3% specificity. 4. Discussion In this study we developed a scoring system usable at admission to predict the walking ability at discharge of patients with hip fracture, which demonstrated high predictive ability in a large cohort.
Please cite this article in press as: Oba T, et al. New scoring system at admission to predict walking ability at discharge for patients with hip fracture. Orthop Traumatol Surg Res (2018), https://doi.org/10.1016/j.otsr.2018.07.024
G Model
ARTICLE IN PRESS
OTSR-2117; No. of Pages 4
T. Oba et al. / Orthopaedics & Traumatology: Surgery & Research xxx (2018) xxx–xxx
3
Table 2 Result of multivariate logistic regression analysis for walking unaided or with a cane at discharge. Variable
Beta coefficient
Odds ratio
95% confidence interval
p-value
Age Male sex Dementia Use of a cane before injury Trochanteric fracture Hemoglobin Albumin Days until surgery
−0.04 −0.41 −2.10 −0.93 −0.56 −0.03 1.42 −0.02
0.96 0.67 0.13 0.40 0.57 0.97 4.15 0.98
0.94–0.99 0.36–1.23 0.06–0.27 0.24–0.65 0.35–0.94 0.83–1.14 2.19–7.85 0.91–1.06
0.002 0.196 <0.001 <0.001 0.027 0.737 <0.001 0.658
Table 3 Prediction scoring system for walking ability at discharge. Scoring formula
Outcome
Factors
Points
Total score
n
Patients unaided or with a cane
Observed probability(%)
Age, (years) Albumin, (g/dl) Dementia Use of a cane before injury Trochanteric fracture
×(−0.04) ×1 −2 −1 −1 +5a
0–0.9 1–1.9 2–2.9 3–3.9 4–4.9 5–5.9 6–6.9 7–7.9
11 26 69 97 95 66 38 7
0 1 5 27 53 42 30 6
0 3.8 7.2 28 56 64 79 86
Total score a
Five points were added for the convenience of deriving positive values.
Fig. 1. Receiver operating characteristic curve of the prediction score. The Cstatistics for the scoring system was 0.81 (95% confidence interval, 0.77–0.85).
The information required to calculate the score can be obtained at the time of admission through patient/family interviews and blood tests. Therefore, this scoring system could be widely utilized to predict the mobility outcomes of hip fracture patients. Predicting the discharge walking ability of hip fracture patients at admission will enable concerned patients and family members to be informed early, providing them with time to prepare for discharge. Furthermore, the prediction will help health care providers to appropriately choose the discharge location which differs among each societies. Several studies reported factors at admission that could predict mobility at discharge using multiple regression analysis. Kristensen et al. reported obstructive factors for attaining basic mobility after surgery in 280 hip fracture patients. Hagino et al.
reported obstructive factors for walking independently or with aids at discharge in 186 hip fracture patients [5,6]. Our study was based on a larger number of patients and also different in the outcome. The outcome in our study did not include those who could walk independently with only by a walker. The outcome was chosen based on the fact that the ability to walk unaided or with a cane at discharge was associated with a better prognosis in mobility at 12 months after surgery, reported by Fukui et al. in a prospective study with a cohort of 650 hip fracture patients [7]. Older age, presence of dementia, disability in walking before the injury and trochanteric fracture, which we detected as obstructive factors for walking ability at discharge, were similar in these studies [5,6]. However, we also identified low serum albumin level at admission as a new predictor for discharge mobility in hip fracture patients. The OR for not being able to walk unaided or with an aid at discharge was 4 per 1 g/dL decrease in serum albumin at admission. There are several studies about serum albumin relatable to our findings. Wakabayashi et al. reported that malnutrition, including low serum albumin levels, was associated with poor general rehabilitation outcomes for elderly inpatients [19]. Eneroth et al. reported that nutritional supplementation decreased hip fracture–related complications in their prospective randomized controlled trial [20]. The importance of nutrition status in hip fracture patients should be recognized. Mean age, proportion of males, and proportion of trochanteric fractures in our study were similar to those in several previous studies (81–83 years, 18–27%, and 47–61%, respectively) [5–7,11,12]. This suggests that our cohort was representative of hip fracture patients. The limitations of this study must be acknowledged. First, it was retrospective. Second, it was conducted in a single institution in Japan and all patients were Japanese. Hip fracture patients in Japan often undergo rehabilitation in the hospital in which they undergo surgery. The scoring system may be more applicable in countries in similar situations than in countries that discharge patients sooner. Thus, further research is required to validate the scoring system in other situations. Third, dementia was not assessed quantitatively. In previous studies, the Short Portable Mental Status Questionnaire was used and reported to be a useful measurement, which
Please cite this article in press as: Oba T, et al. New scoring system at admission to predict walking ability at discharge for patients with hip fracture. Orthop Traumatol Surg Res (2018), https://doi.org/10.1016/j.otsr.2018.07.024
G Model OTSR-2117; No. of Pages 4
ARTICLE IN PRESS T. Oba et al. / Orthopaedics & Traumatology: Surgery & Research xxx (2018) xxx–xxx
4
could have been appropriate [21]. However, scoring systems must be simple and quick to perform in daily practice. Finally, although we assessed numerous variables, the effects of unmeasured confounders cannot be ruled out. 5. Conclusion We report the development of a new scoring system used at admission for hip fracture patients that predicts mobility at discharge. Furthermore, we identified serum albumin level at admission as a new predictor for mobility at discharge. Future challenges include utilizing the scoring system and further evaluating its accuracy and usability. Disclosure of interest The authors declare that they have no competing interest. Funding None. Contribution All authors contributed to the study concept and design. T. Oba drafted the manuscript. T. Oba and H. Makita collected the data. T. Oba and H. Yamana performed the statistical analyses. H. Yamana, Y. Inaba, H. Yamana and T. Saito revised the manuscript critically for important intellectual content. References [1] Cooper C, Campion G, Melton LJ. Hip fractures in the elderly: a world-wide projection. Osteoporos Int 1992;2:285–9. [2] Parker MJ, Palmer CR. A new mobility score for predicting mortality after hip fracture. J Bone Joint Surg Br 1993;75:797–8. [3] Koval KJ, Skovron ML, Aharonoff GB, Meadows SE, Zuckerman JD. Ambulatory ability after hip fracture. A prospective study in geriatric patients. Clin Orthop Relat Res 1995;310:150–9.
[4] Cornwall R, Gilbert MS, Koval KJ, Strauss E, Siu AL. Functional outcomes and mortality vary among different types of hip fractures: a function of patient characteristics. Clin Orthop Relat Res 2004;425:64–71. [5] Hagino T, Ochiai S, Wako M, Sato E, Maekawa S, Senga S, et al. A simple scoring system to predict ambulation prognosis after hip fracture in the elderly. Arch Orthop Trauma Surg 2007;127:603–6. [6] Kristensen MT, Foss NB, Ekdahl C, Kehlet H. Prefracture functional level evaluated by the New Mobility Score predicts in-hospital outcome after hip fracture surgery. Acta Orthop 2010;81:296–302. [7] Fukui N, Watanabe Y, Nakano T, Sawaguchi T, Matsushita T. Predictors for ambulatory ability and the change in ADL after hip fracture in patients with different levels of mobility before injury: a 1 year prospective cohort study. J Orthop Trauma 2012;26:163–71. [8] Simon P, Gouin F, Veillard D, Laffargue P, Ehlinger M, Bel JC, et al. Femoral neck fractures in patients over 50 years old. Rev Chir Orthop Reparatrice Appar Mot 2008;94:108–32 [Epub 2008 Sep 19]. [9] Reina N, Bonnevialle P, Rubens DB, Adam P, Loubignac F, Favier T, et al. Internal fixation of intra-capsular proximal femoral fractures in patients older than 80 years: still relevant? Multivariate analysis of a prospective multicentre cohort. Orthop Traumatol Surg Res 2017;103:3–7 [Epub 2016 Dec 2]. [10] Thorngren KG, Ceder L, Svensson K. Predicting results of rehabilitation after hip fracture. A ten-year follow-up study. Clin Orthop Relat Res 1993;287:76–81. [11] Svensson O, Strömberg L, Öhlén G, Lindgren U. Prediction of outcome of hip fractures in elderly patients. J Bone Joint Surg Br 1996;78:115–8. [12] Hagino T, Ochiai S, Sato E, Watanabe Y, Senga S, Haro H. Prognostic prediction in patients with hip fracture: risk factors predicting difficulties with discharge to own home. J Orthop Traumatol 2011;12:77–80. [13] Vochteloo AJH, Tuinebreijer WE, Maier AB, Nelissen RGH, Bloem RM, Pilot P. Predicting discharge location of hip fracture patients; the new discharge of hip fracture patients score. Int Orthop 2012;36:1709–14. [14] Salar O, Baker PN, Forward DP, Ollivere BJ, Weerasuriya N, Moppett IK, et al. Predictors of direct home discharge following fractured neck of femur. Ann R Coll Surg Engl 2017;99:444–51. [15] Hay D, Parker MJ. Hip fracture in the immobile patient. J Bone Joint Surg Br 2003;85:1037–9. [16] Garden RS. Low-angle fixation in fractures of the femoral neck. J Bone Joint Surg 1961;43:647–63. [17] Hagino T, Ochiai S, Sato E, Maekawa S, Wako M, Haro H. The relationship between anemia at admission and outcome in patients older than 60 years with hip fracture. J Orthop Traumatol 2009;10:119–22. [18] Sullivan LM, Massaro JM, D’Agostino Sr RB. Presentation of multivariate data for clinical use: the Framingham Study risk score functions. Stat Med 2004;23:1631–60. [19] Wakabayashi H, Sashika H. Malnutrition is associated with poor rehabilitation outcome in elderly patients with hospital-associated deconditioning: a prospective cohort study. J Rehabil Med 2014;46:277–82. [20] Eneroth M, Olsson UB, Thorngren KG. Nutritional supplementation decreases hip fracture-related complications. Clin Orthop 2006;451:212–7. [21] Söderqvist A, Miedel R, Ponzer S, Tidermark J. The influence of cognitive function on outcome after a hip fracture. J Bone Joint Surg Am 2006;88:2115–23.
Please cite this article in press as: Oba T, et al. New scoring system at admission to predict walking ability at discharge for patients with hip fracture. Orthop Traumatol Surg Res (2018), https://doi.org/10.1016/j.otsr.2018.07.024