- Email: [email protected]
Predicting the need for blood transfusions in elderly patients with pertrochanteric femoral fractures
G Model
JINJ-5853; No. of Pages 6 Injury, Int. J. Care Injured xxx (2014) xxx–xxx
Contents lists available at ScienceDirect
Injury journal homepage: www.elsevier.com/locate/injury
Predicting the need for blood transfusions in elderly patients with pertrochanteric femoral fractures Guojin Hou 1, Fang Zhou *, Yun Tian, Hongquan Ji, Zhishan Zhang, Yan Guo, Yang Lv Department of Orthopaedic Surgery, Peking University Third Hospital, No 49, North Garden Rd, HaiDian District, Beijing 100191, China
A R T I C L E I N F O
A B S T R A C T
Article history: Accepted 17 August 2014
Objective: The need exists for perioperative blood management measures aimed at improving patient outcomes and reducing the risks of allogeneic blood transfusion (ABT). Our study aim is to discuss an algorithm to predict the need for perioperative blood transfusion in old patients with pertrochanteric femoral fractures. Methods: We retrospectively analysed the data from 220 elderly patients with pertrochanteric femoral fractures with regard to the probability of receiving an ABT within 72 h after surgery. The patients were divided into ABT and non-ABT groups. A univariate analysis was used to compare between-group differences with regard to 13 variables. A logistic regression analysis and a probability algorithm to predict the need for an ABT based on independent predictors were used. Results: The non-ABT group included 131 patients (55 males and 76 females), with an average age of 77.2 6.8 years; the ABT group included 89 patients (29 males and 60 females), with an average age of 79.7 6.6 years. The total volume of transfused blood was 276 Units; the actual average blood transfusion was 3.1 1.47 Units. Significant between-group differences (P < 0.05) were observed with regard to age, duration of operation, haemoglobin (Hb) at admission, intra-operative blood loss, type of fracture and type of anaesthesia. The mean volume of transfused blood in the proximal femoral nail anti-rotation (PFNA) and Gamma3 group was larger than that of the dynamic hip screw (DHS) group (P < 0.05). A logistic regression analysis revealed that patients with pertrochanteric femoral fractures who were elderly (>81 years), had lower Hb levels at admission (124 g/L), longer duration of operations (t > 85 min), underwent intramedullary fixation (Gamma3 and PFNA) and had more intra-operative blood loss were more likely to need an ABT. This regression model predicted 74.1% of the transfused cases. Conclusions: An algorithm was devised to predict and manage the need for an ABT within 72 h after surgery in patients with pertrochanteric femoral fractures. A reasonable transfusion program might reduce the complications caused by anaemia and effectively avoid the risks associated with ABTs. ß 2014 Elsevier Ltd. All rights reserved.
Keywords: Blood transfusion Allogeneic Hip fractures Pertrochanteric fractures Surgery Internal fixation Perioperative period The elderly
Introduction Pertrochanteric femoral fracture occurs primarily in the elderly, approximate 3–4% of all the fractures in the general population. With the aging process of the society, the number of this fracture has been increasing year by year. Such fractures constitute an important cause of hospital admissions and length of stay in the elderly. Moreover, they are associated with functional
* Corresponding author. Tel.: +86 010 82267380; fax: +86 010 82267380. E-mail addresses: [email protected] (G. Hou), [email protected] (F. Zhou). 1 Reprint contact: Guojin Hou, Department of Orthopaedic Surgery, Peking University Third Hospital. No 49, North Garden Rd, HaiDian District, Beijing 100191, China.
impairments and major disabilities and result in high rates of institutionalisation and mortality. Surgery with stable internal fixation, which has been recommended as the first choice of treatment [1], allows early mobilisation and reduces complications. With the advances in techniques of orthopaedic surgery and mechanical improvements of implants, most pertrochanteric femoral fractures can be treated with minimally invasive surgery; however, post-operative anaemia is still a common phenomenon in elderly patients with pertrochanteric femoral fractures. Lower postoperative haemoglobin (Hb) level has been associated with a worse early functional recovery, longer length of stay, and higher rates of readmission and mortality [2,3]. Cross-matching is the routine preparation to prevent postoperative anaemia and complications in these patients. Furthermore, surgery for hip
http://dx.doi.org/10.1016/j.injury.2014.08.033 0020–1383/ß 2014 Elsevier Ltd. All rights reserved.
Please cite this article in press as: Hou G, et al. Predicting the need for blood transfusions in elderly patients with pertrochanteric femoral fractures. Injury (2014), http://dx.doi.org/10.1016/j.injury.2014.08.033
G Model
JINJ-5853; No. of Pages 6 G. Hou et al. / Injury, Int. J. Care Injured xxx (2014) xxx–xxx
2
fractures ranks second in the total number of blood units administered to patients according to diagnoses groups [4]. Allogeneic blood transfusion (ABT) is also a known risk factor for immunosuppression and increases the risk of infection; moreover, over-transfusion influences patient health and has economic implications [5–7]. Therefore, many scholars have recommended a ‘‘restrictive transfusion policy’’ [8–10]. However, the indications for ABTs in elderly patients with pertrochanteric femoral fracture have not been standardised and remain controversial. Previous attempts to improve the efficiency of blood ordering and transfusion in elderly patients have advocated policies for all patients with hip fractures [1]. Few studies have focused on pertrochanteric femoral fracture. The purpose of the current study was to assess the epidemiological characteristics of elderly patients requiring ABT admitted to a hospital for pertrochanteric femoral fractures. Using these data, an algorithm was built to predict the need for ABTs. We hypothesised that we would be able to predict which of the patients who were admitted for pertrochanteric femoral fracture would need an ABT. By better managing this process, patients are more likely to receive optimal medical care without straining hospital resources, thereby reducing the cost of care. Patients and methods Patients This retrospective database study was based on data collected from the hospital electronic medical record (EMR) system and the blood bank database. Approval for conducting this study was received from the local research committees. A total of 408 patients were admitted for pertrochanteric femoral fractures and underwent surgical treatment between 2007 and 2012; however, only 220 of these patients met the inclusion criteria (Table 1). We excluded all patients who were treated without surgery. Risk factors for an ABT that were assessed included patient demographics (i.e., age and gender), fracture type, American Society of Anaesthesiology (ASA) grade, anaesthesia type, time between injury and surgery, laboratory values (i.e., Hb at admission, mean corpuscular volume [MCV], international normalised ratio [INR], transaminase and creatinine), the use of anticoagulants, and surgical details (i.e., internal fixation type, duration of operation, intra-operative blood loss, and postoperative drainage). Complications such as infections (pneumonia and surgical site infection), cardiac events, cerebro-vascular events and 30-day mortality were also recorded, as well as the number of units of transfused blood. Monitoring of Hb and transfusion protocol At admission to hospital, all patients had a complete blood count (CBC) analysis; two samples of blood were taken for the ABO Table 1 Study criteria. Inclusion criteria (1) older than 60 years (2) low energy injury without concomitant fractures or injuries that might require an ABT (3) treated with close reduction and minimally invasive surgery (4) blood count measured at admission, post-operative days one and three, and post-transfusion Exclusive criteria (1) pathologic fracture due to malignancy and high energy injury (2) history of blood system diseases and Hb <80 g/L at admission (3) severe liver or kidney dysfunction
Rh group type and antibody screening and two units of packed red blood cells (RBCs) were cross-matched and ordered. An additional cross-match was carried out thereafter as necessary because the previous sample had expired according to blood bank safety regulations. A CBC was taken on post-operative days one and three, when the patients presented signs of symptomatic anaemia (i.e., tachycardia, weakness and palpitations) and after blood transfusion. Our institute currently adheres to a ‘‘restrictive transfusion policy’’. According to our transfusion protocol, patients were preoperatively transfused to an Hb level of 90 g/L. The patients were transfused post-operatively when their Hb level dropped below 80 g/L. This ‘‘transfusion trigger’’ was raised to 90 g/L in clinically symptomatic patients (i.e., extreme weakness, chest pain, extreme paleness or major bleeding), when vital signs were abnormal (i.e., tachycardia, heart rate [HR] >100 bpm, and systolic blood pressure [SBP] <90 mmHg); and in those with a history of heart disease (i.e., coronary, valvular problems and arrhythmia) or cerebro-vascular disease. Surgical procedure All patients were treated on a traction table under the guidance of G-arm fluoroscopy. All of the procedures were performed by 4 senior surgeons in our institution using modified techniques, who are experienced in the management of pertrochanteric fractures. Moreover, duration of operation in this study was measured from the time of close reduction to incision closure. Surgeons did not perform surgery directly under imaging intensify in order to avoid excess of X-ray exposure. When the image intensifier was used to assess the reduction of fracture, the locations of the guide wire and internal fixations, the surgeons were behind the tinplate. Forty patients received a dynamic hip screw (DHS, Synthes, Switzerland) fixation; 84 patients received a proximal femoral nail anti-rotation (PFNA, Synthes, Switzerland) fixation, and 96 patients received a Gamma3 nail (Stryker, MI, USA) fixation. Statistic analyses Statistical analyses were conducted using SPSS 13.0 (SPSS Inc., USA). Univariate analyses were used to compare the group that received an ABT within 72 h after surgery with the group that did not with regard to 13 variables. Continuous, normally distributed variables were analysed using a two-sample Student’s t test; continuous, non-normally distributed variables were analysed using the Wilcoxon-rank-sum test; Pearson’s chi-square test and Fisher’s exact test were used to compare the groups with respect to categorical variables; a one-way ANOVA was used to compare differences in groups, and the LSD method was used to compare differences between groups. A P value of <0.05 was considered to indicate statistical significance. A multivariate logistic regression was applied to identify the significant independent predictors of ABT. The full regression model included 13 risk factor candidates. Model selection methods such as Wald-backward elimination were used to identify significant factors from the 13 explanatory variables. Results The data of all 220 patients were included in the final analysis. The patients were divided into non-ABT and ABT groups. Table 2 shows the general characteristics and the univariate analysis of the study population. A total of 89 patients (40.5%) received an ABT within the first 72 h after surgery, 5 patients were transfused to an Hb level of 90 g/L pre-operative, 76 patients were transfused when their Hb levels dropped below 80 g/L post-operative, and 8 patients
Please cite this article in press as: Hou G, et al. Predicting the need for blood transfusions in elderly patients with pertrochanteric femoral fractures. Injury (2014), http://dx.doi.org/10.1016/j.injury.2014.08.033
G Model
JINJ-5853; No. of Pages 6 G. Hou et al. / Injury, Int. J. Care Injured xxx (2014) xxx–xxx
3
Table 2 General characteristics and univariate analysis of the risk factors for ABT within 72 h after surgery in 220 patients. Variable
Total
Non ABT group
ABT group
P value
Number Gender (male/female) Age (years) mean SD AO fracture type (A1/A2/A3) ASA grade ( / )§ Anaesthesia type (general/regional) Time between injury and surgery (days)mean SD Duration of operation (min) mean SD Hb at admission (g/L) mean SD Internal fixation type (DHS/PFNA/Gamma3) Anti-coagulation (preoperative Yes/No) Anti-coagulation (postoperative Yes/No) Intra-operative blood loss (mL)mean SD Postoperative drainage (mL)mean SD Average volume of ABT (mL) mean SD Actual average volume of ABT (mL) mean SD
220 84/136 78.2 6.8 65/144/11 128/92 79/141 4.5 3.9 76.7 27.6 113.6 16.1 40/84/96 124/96 194/26 107.0 93.0 24.7 60.3 250.9 357.5 –
131 55/76 77.2 6.8 50/77/4 79/52 39/92 4.3 3.6 72.7 23.4 119.5 14.7 30/49/52 71/60 114/17 92.0 63.0 25.8 54.2 – –
89 29/60 79.7 6.6 15/67/7 49/40 40/49 4.7 4.2 82.6 32.1 104.9 14.1 10/35/44 53/36 80/9 129.2 121.7 23.1 68.5 – 620.2 294.0
– 0.159¥ 0.008* 0.002¥ 0.439¥ 0.021¥ £ 0.946 £ 0.025 0.000* 0.075¥ 0.432¥ 0.518¥ £ 0.041 £ 0.316 – –
ABT, allogeneic blood transfusion; AO, Arbeitsgemeinschaft fu¨r Osteosythese; ASA, American Society of Anaesthesiology; Hb, haemoglobin; DHS, dynamic hip screw; PFNA, proximal femoral nail anti-rotation. The values are given as the mean and standard deviation for continuous variables and as the number of patients for categorical variables. * Two sample t-test. ¥ Chi-square test. £ Wilcoxon-rank-sum test. § Group contains ASA I and II, whereas group contains ASA III and IV because separate groups of patients classified as ASA I (n = 4) or ASA IV (n = 2) were too small to be analysed separately.
were transfused post-operatively when clinical symptoms were detected. The total volume of blood transfused was 276 units, and the actual average blood transfusion was 3.1 1.47 units. Pneumonia developed in 5 patients who had fully recovered during the follow-up; fat liquefaction occurred in one patient who healed after a continued dressing change; and wound infection occurred in 2 patients after surgery, but the wounds healed after antibiotic treatment and extensive debridement. The 30-day mortality rate was 3.6%. An additional analysis demonstrated that ABT was not associated with post-operative complications such as infections (i.e., pneumonia or surgical site infection), cardiac and cerebro-vascular events or 30-day mortality (Table 3). Risk factors for ABT within 72 h after surgery were assessed with a univariate analysis (Table 2). Significantly different factors (P < 0.05) included age (79.7 6.6 years vs. 77.2 6.8 years), Hb level at admission (104.9 14.1 g/L vs. 119.5 14.7 g/L), duration of surgery (82.6 32.1 min vs. 72.7 23.4 min), blood loss during operation (129.2 121.7 mL vs. 92.0 63.0 mL), more severe injuries (AO fracture type A2 and A3 vs. A1), and anaesthesia type (regional vs. general anaesthesia). The mean volumes of blood transfused in the PFNA and Gamma3 groups were 269.1 mL and 285.5 mL, respectively, which were greater than that transfused in the DHS group (130.0 mL, P < 0.05). A multivariate logistic regression was performed to simulate a decision analysis situation when admitting a patient with a pertrochanteric femoral fracture. Five out of 13 independent variables were found to have a statistically significant effect on the rate of ABT within 72 h after surgery: age, Hb level at admission, Table 3 Comparisons of the post-operative complications between patients with and without ABT. Post-operative complications
Non ABT group
ABT group
P value
Pneumonia Surgical site infection Cardiac related complications Cerebro-vascular events 30-day mortality
3/131 1/131 2/131 2/131 5/131
2/89 1/89 4/89 2/89 3/89
1.000* 0.647¥ 0.366* 1.000* 1.000*
* ¥
Chi-square test. Fisher’s exact test.
type of internal fixation, duration of operation and blood loss during operation. Regression coefficients, likelihood ratios, P values, adjusted odds ratios, and 95% confidence intervals were determined (Table 4). This regression model predicted 74.1% of the transfused cases. A receiver operating characteristic (ROC) curve analysis was used to evaluate the predictive performance of the logistic regression model and its ability to predict ABT within 72 h after surgery (Fig. 1). The area under the curve was 0.824, which demonstrated an acceptable diagnostic performance. A predicted probability (Fig. 2) P (ABT 72 h) of 0.373 was selected as an optimal cut-off point to best differentiate patients who should receive ABT within 72 h after surgery from those who should not. This cut-off point has the highest sensitivity and specificity rates; and it has a false negative rate of 20.2% and a false positive rate of 29.0%. To further characterise the patients, they were divided into two groups stratified by age, Hb level at admission and duration of operation (cut-off point: P = 0.5). A significant difference was observed between the two groups (Table 5). Patients who were elderly (>81years), had lower Hb levels at admission (Hb 124 g/ L) and longer duration of operation (t > 85 min), were considered as the significant factors for the need for blood transfusions. Discussion Patients with pertrochanteric femoral fractures who are treated operatively frequently require the utilisation of ABT. However,
Table 4 Results of the final logistic regression model to predict ABT within 72 hours after surgery in patients with pertrochanteric femoral fractures. Multivariate predictor Age Hb at admission Type of internal fixation Duration of operation Blood loss during operation
Regression coefficient 0.012 0.086 0.810 0.021 0.006
Odds ratio
95% Confidence intervals
P value
1.012 0.917 0.445 1.021 1.006
0.958–1.069 0.893–0.942 0.275–0.720 1.004–1.038 1.001–1.010
0.034 0.000 0.001 0.013 0.016
Please cite this article in press as: Hou G, et al. Predicting the need for blood transfusions in elderly patients with pertrochanteric femoral fractures. Injury (2014), http://dx.doi.org/10.1016/j.injury.2014.08.033
G Model
JINJ-5853; No. of Pages 6 G. Hou et al. / Injury, Int. J. Care Injured xxx (2014) xxx–xxx
4
Table 5 Comparison of the two patient groups stratified by age, haemoglobin at admission and duration of operation.
ABT group Non-ABT group
Age (years)
Hb at admission (g/L)
Duration of operation (min)
81
124
85
>81
54 35 96 35 P = 0.049*
>124
82 7 82 49 P < 0.001*
>85
57 32 99 32 P = 0.045¥
ABT, allogeneic blood transfusion; Hb, haemoglobin. * Chi-square test. ¥ Fisher’s exact test.
patients older than 81 years was obviously higher than that in younger patients (P < 0.05). Swain [17] also reported a significantly greater transfusion need for patients aged 80 years and above. This finding was most likely related to the lower Hb level at admission in the elderly group compared with their younger counterparts (109.2 g/L vs. 115.3 g/L). Comorbid disease, organ dysfunction and poor compensatory ability all lead to a high incidence of ABT in the elderly. Hb at admission Fig. 1. A ROC curve analysis was used to evaluate the predictive performance of the logistic regression model to predict ABT within 72 h after surgery.
patients with normal preoperative Hb values may undergo surgery without ABTs. The early identification of these patients might save unnecessary blood cross-matching. The purpose of the current study was to define a method to predict the need for an ABT in elderly patients admitted with pertrochanteric femoral fractures. This study is important because the ability to predict the need for ABT is the first step in managing blood transfusions in pertrochanteric femoral fracture patients. Managing these transfusions has direct implications for patients who do not benefit from a transfusion, furthermore, it has economic implications for the institutional blood bank. A reasonable transfusion program might reduce the complications caused by anaemia and effectively reduce the economic expense, avoid the risks of blood transfusion [11–13] and improve the perioperative function state [3,14]. This study focused on predictors of ABT and the algorithms that predict the need for such transfusions based on available institutional data and transfusion protocols. Logistic regression analysis revealed that patients with pertrochanteric femoral fractures who are elderly (>81 years), have lower Hb levels at admission (Hb 124 g/L), undergo longer surgeries (t > 85 min) and intramedullary fixation (Gamma3 and PFNA) and experience more intra-operative blood loss are more likely to need an ABT. The results of the present study are instructive with regard to certain aspects of the ABT requirements in this population: Age Old age can predicts the need for perioperative blood transfusions in patients with pertrochanteric femoral fractures [15,16]. This study found that the ABT group was older on average than the non-ABT group. Moreover, the incidence of ABT in
The average Hb level at admission of the non-ABT group was higher than that of the ABT group (119.5 g/L vs. 104.9 g/L). Furthermore, this study found that values of 124 g/L proved as best predictor of blood transfusion, approaching a 7-fold risk for transfusion (OR = 7.0, P < 0.001). This result is in accordance with previous observations [15,16,18,19]. Robbins [20] also reported that only 1 in 9 patients with Hb levels >11 g/dL received a transfusion. Kurdy and Hokan [21] advised that cross-matching should be routinely ordered for patients with Hb levels <12 g/dL. Adunsky [1] reported that patients with normal Hb levels (>12 g/ dL) could undergo surgery after typing without cross-matching or transfusion. Shokoohi [16] indicated that patients with anaemia at admission were approximately six times more likely to have a transfusion compared with non-anaemic patients. Type of internal fixation Patients treated with PFNA and Gamma3 had transfusion indexes of 1.35 units and 1.43 units, respectively, compared with 0.65 units for the DHS group. These results are in accordance with previous observations [1,15,18,22]. As orthopaedic surgery techniques have advanced and mechanic implants have improved (especially after the introduction of minimally invasive surgery), only small volumes of gross blood loss have occurred during operations. However, Foss reported large volumes of hidden blood [23]. The opening of the medullary cavity and proximal reaming might increase hidden blood loss, which made a higher incidence of ABT in patients treated with intramedullary fixation devices (i.e., PFNA and Gamma3). Duration of operation The average operation time in the transfusion group was longer than that of the non-transfusion group (82.6 min vs. 72.7 min), and
Fig. 2. A predictive formula based on the five significant risk factors model used to predict the need for ABT within 72 h after surgery in patients with pertrochanteric femoral fractures. ABT 72 h, ABT within 72 h after surgery; Hb, haemoglobin; DHS, dynamic hip screw; D, duration of operation; V, volume of blood loss during operation.
Please cite this article in press as: Hou G, et al. Predicting the need for blood transfusions in elderly patients with pertrochanteric femoral fractures. Injury (2014), http://dx.doi.org/10.1016/j.injury.2014.08.033
G Model
JINJ-5853; No. of Pages 6 G. Hou et al. / Injury, Int. J. Care Injured xxx (2014) xxx–xxx
5
Fig. 3. An algorithm designed to predict the need for ABT. Hb, haemoglobin; CBC, complete blood count; t, duration of operation; post-op, post-operative.
a significantly higher incidence of ABT was observed after surgeries longer than 85 min, comparing with the counterparts. To our knowledge, this result has not been previously reported. The repeated mobility of the fracture parts, long duration of the medullary cavity opening and regional hemorrhage [24] all increased the likelihood of postoperative ABT. Blood loss during operation The average volume of blood loss during operation in the ABT group was greater than that in the non-ABT group (129.2 mL vs. 92.0 mL), and more blood loss during operation might predict the need for blood transfusion. Morritt and Foss also reported this result [22,23]. Insufficient haemostasis [25], repeated mobility of the fracture parts, and long duration of medullary cavity opening might all increase the volume of blood loss during operation. Furthermore, a post-operative ABT was required when the Hb levels dropped below 80 g/L. The current study should be interpreted in the light of its limitations. This retrospective study was based on prospectively collected database information; and mathematical model may not be able to fully explain the clinical problems; therefore, randomised controlled trials are necessary to confirm its results. Data such as radiographs were not reviewed. Moreover, the ASA grade rather than co-morbidities was used as a risk factor. Conclusion The current study recommends the ‘‘restrict transfusion policy’’, and an algorithm (Fig. 3) was designed to predict the need for an ABT. Cross-matching should be routinely ordered for patients with pertrochanteric femoral fractures who are older than 81years or have Hb levels below 124 g/L at admission; changes of patients’ CBC should be closely monitored when operation times exceed 85 min, when an intramedullary fixation device is used and when a large volume of blood is lost. Additionally, ABT is needed for patients whose Hb levels drop below 80 g/L, who have clinical symptoms (i.e., extreme weakness, chest pain, extreme paleness or major bleeding), or abnormal vital signs (i.e., tachycardia, HR >100 bpm, and SBP <90 mmHg). Conflicts of interests and disclosure of funding No funding was received for this paper. We declare that there are no conflicts of interests concerning this paper.
Acknowledgements The authors thank Dr. Ru Chen and Huijun Zhang for their help in designing the statistical analysis. References [1] Adunsky A, Lichtenstein A, Mizrahi E, et al. Blood transfusion requirements in elderly hip fracture patients[J]. Arch Gerontol Geriatr 2003;36(1):75–81. [2] Halm EA, Wang JJ, Boockvar K, et al. The effect of perioperative anemia on clinical and functional outcomes in patients with hip fracture. J Orthop Trauma 2004;18:369–74. [3] Lawrence VA, Silverstein JH, Cornell JE, et al. Higher Hb level is associated with better early functional recovery after hip fracture repair. Transfusion 2003;43:1717–22. [4] Friedman BA, Burns TL, Schork MA. A study of blood utilisation by diagnosis, month of transfusion, and geographic region of the United States. Transfusion 1979;19:511–25. [5] Carson JL, Altman DG, Duff A, et al. Risk of bacterial infection associated with allogeneic blood transfusion among patients undergoing hip fracture repair. Transfusion 1999;39(7):694–700. [6] Levi N, Sandberg T. Blood transfusion and postoperative wound infection in intracapsular femoral neck fractures. Bull-Hosp Joint Dis 1998;57(2):69–73. [7] Koval KJ, Rosenberg AD, Zuckerman JD, et al. Does blood transfusion increase the risk of infection after hip fracture? J Orthop Trauma 1997;11(4):260–5. [8] Carson JL, Hill S, Carless P, et al. Transfusion triggers: a systematic review of the literature. Transfus Med Rev 2002;16(3):187–99. [9] Khan AM, Mushtaq N, Giannakas K, et al. Cross-match protocols for femoral neck fractures—finding one that can work. Ann R Coll Surg Engl 2004;86(1):11–4. ˜ oz M, Leal-Noval SR. Restrictive transfusion triggers in major orthopaedic [10] Mun surgery: effective and safe? Blood Transfus 2013;11(2):169–71. [11] Bierbaum BE, Callaghan JJ, Galante JO, et al. An analysis of blood management in patients having a total hip or knee arthroplasty. J Bone Joint Surg 1999;81(1):2–10. [12] Goodnough LT, Shander A, Brecher ME. Transfusion medicine: looking to the future. Lancet 2003;361(9352):161–9. [13] Dodd RY. Current estimates of transfusion safety worldwide. Dev Biol 2004;120:3–10. [14] Foss NB, Kristensen MT, Kehlet H. Anaemia impedes functional mobility after hip fracture surgery. Age Ageing 2008;37(2):173–8. [15] Kadar A, Chechik O, Steinberg E, et al. Predicting the need for blood transfusion in patients with hip fractures. Int Orthop 2013;1–8. [16] Shokoohi A, Stanworth S, Mistry D, et al. The risks of red cell transfusion for hip fracture surgery in the elderly. Vox Sang 2012;103(3):223–30. [17] Swain DG, Nightingale PG, Patel JV. Blood transfusion requirements in femoral neck fracture. Injury 2000;31(1):7–10. [18] Bajs ID, Cvjetko I, Vicic VK, et al. Blood transfusion in patients with pertrochanteric femoral fracture. Acta Med Croatica 2006;60(1):7–10. [19] Cuenca J, JA GE, Martinez AA, et al. Preoperative blood test results and type of fracture as transfusional risk factors in patients older than 65 years with trochanteric hip fracture. Rev Esp Anestesiol Reanim 2004;51(9):515–22. [20] Robbins J, Steingold RF. Blood use in urgent operations for patients with fractures of the femoral neck. Injury 1986;17(4):265–6. [21] Kurdy NMG, Hokan R. A cross-matching policy for fractures of the proximal third of the femur. Injury 1993;24(8):521–4.
Please cite this article in press as: Hou G, et al. Predicting the need for blood transfusions in elderly patients with pertrochanteric femoral fractures. Injury (2014), http://dx.doi.org/10.1016/j.injury.2014.08.033
G Model
JINJ-5853; No. of Pages 6 6
G. Hou et al. / Injury, Int. J. Care Injured xxx (2014) xxx–xxx
[22] Morritt DG, Morritt AN, Kelley SP, et al. Blood ordering protocol based on proposed surgical implant in fractured neck of femur patients. Ann R Coll Surg Engl 2005;87(6):445–8. [23] Foss NB, Kehlet H. Hidden blood loss after surgery for hip fracture. J Bone Joint Surg BrVol 2006;88(8):1053–9.
[24] Zhang PX, Xue F, An S, et al. Clinical analysis of obvious and hidden blood loss in inter-trochanter fracture patients treated with proximal femoral nail antirotation and dynamic hip screw. Beijing Da Xue Xue Bao 2012;6:891–4. [25] Smith GH, Tsang J, Molyneux SG, et al. The hidden blood loss after hip fracture. Injury 2011;42(2):133–5.
Please cite this article in press as: Hou G, et al. Predicting the need for blood transfusions in elderly patients with pertrochanteric femoral fractures. Injury (2014), http://dx.doi.org/10.1016/j.injury.2014.08.033
JINJ-5853; No. of Pages 6 Injury, Int. J. Care Injured xxx (2014) xxx–xxx
Contents lists available at ScienceDirect
Injury journal homepage: www.elsevier.com/locate/injury
Predicting the need for blood transfusions in elderly patients with pertrochanteric femoral fractures Guojin Hou 1, Fang Zhou *, Yun Tian, Hongquan Ji, Zhishan Zhang, Yan Guo, Yang Lv Department of Orthopaedic Surgery, Peking University Third Hospital, No 49, North Garden Rd, HaiDian District, Beijing 100191, China
A R T I C L E I N F O
A B S T R A C T
Article history: Accepted 17 August 2014
Objective: The need exists for perioperative blood management measures aimed at improving patient outcomes and reducing the risks of allogeneic blood transfusion (ABT). Our study aim is to discuss an algorithm to predict the need for perioperative blood transfusion in old patients with pertrochanteric femoral fractures. Methods: We retrospectively analysed the data from 220 elderly patients with pertrochanteric femoral fractures with regard to the probability of receiving an ABT within 72 h after surgery. The patients were divided into ABT and non-ABT groups. A univariate analysis was used to compare between-group differences with regard to 13 variables. A logistic regression analysis and a probability algorithm to predict the need for an ABT based on independent predictors were used. Results: The non-ABT group included 131 patients (55 males and 76 females), with an average age of 77.2 6.8 years; the ABT group included 89 patients (29 males and 60 females), with an average age of 79.7 6.6 years. The total volume of transfused blood was 276 Units; the actual average blood transfusion was 3.1 1.47 Units. Significant between-group differences (P < 0.05) were observed with regard to age, duration of operation, haemoglobin (Hb) at admission, intra-operative blood loss, type of fracture and type of anaesthesia. The mean volume of transfused blood in the proximal femoral nail anti-rotation (PFNA) and Gamma3 group was larger than that of the dynamic hip screw (DHS) group (P < 0.05). A logistic regression analysis revealed that patients with pertrochanteric femoral fractures who were elderly (>81 years), had lower Hb levels at admission (124 g/L), longer duration of operations (t > 85 min), underwent intramedullary fixation (Gamma3 and PFNA) and had more intra-operative blood loss were more likely to need an ABT. This regression model predicted 74.1% of the transfused cases. Conclusions: An algorithm was devised to predict and manage the need for an ABT within 72 h after surgery in patients with pertrochanteric femoral fractures. A reasonable transfusion program might reduce the complications caused by anaemia and effectively avoid the risks associated with ABTs. ß 2014 Elsevier Ltd. All rights reserved.
Keywords: Blood transfusion Allogeneic Hip fractures Pertrochanteric fractures Surgery Internal fixation Perioperative period The elderly
Introduction Pertrochanteric femoral fracture occurs primarily in the elderly, approximate 3–4% of all the fractures in the general population. With the aging process of the society, the number of this fracture has been increasing year by year. Such fractures constitute an important cause of hospital admissions and length of stay in the elderly. Moreover, they are associated with functional
* Corresponding author. Tel.: +86 010 82267380; fax: +86 010 82267380. E-mail addresses: [email protected] (G. Hou), [email protected] (F. Zhou). 1 Reprint contact: Guojin Hou, Department of Orthopaedic Surgery, Peking University Third Hospital. No 49, North Garden Rd, HaiDian District, Beijing 100191, China.
impairments and major disabilities and result in high rates of institutionalisation and mortality. Surgery with stable internal fixation, which has been recommended as the first choice of treatment [1], allows early mobilisation and reduces complications. With the advances in techniques of orthopaedic surgery and mechanical improvements of implants, most pertrochanteric femoral fractures can be treated with minimally invasive surgery; however, post-operative anaemia is still a common phenomenon in elderly patients with pertrochanteric femoral fractures. Lower postoperative haemoglobin (Hb) level has been associated with a worse early functional recovery, longer length of stay, and higher rates of readmission and mortality [2,3]. Cross-matching is the routine preparation to prevent postoperative anaemia and complications in these patients. Furthermore, surgery for hip
http://dx.doi.org/10.1016/j.injury.2014.08.033 0020–1383/ß 2014 Elsevier Ltd. All rights reserved.
Please cite this article in press as: Hou G, et al. Predicting the need for blood transfusions in elderly patients with pertrochanteric femoral fractures. Injury (2014), http://dx.doi.org/10.1016/j.injury.2014.08.033
G Model
JINJ-5853; No. of Pages 6 G. Hou et al. / Injury, Int. J. Care Injured xxx (2014) xxx–xxx
2
fractures ranks second in the total number of blood units administered to patients according to diagnoses groups [4]. Allogeneic blood transfusion (ABT) is also a known risk factor for immunosuppression and increases the risk of infection; moreover, over-transfusion influences patient health and has economic implications [5–7]. Therefore, many scholars have recommended a ‘‘restrictive transfusion policy’’ [8–10]. However, the indications for ABTs in elderly patients with pertrochanteric femoral fracture have not been standardised and remain controversial. Previous attempts to improve the efficiency of blood ordering and transfusion in elderly patients have advocated policies for all patients with hip fractures [1]. Few studies have focused on pertrochanteric femoral fracture. The purpose of the current study was to assess the epidemiological characteristics of elderly patients requiring ABT admitted to a hospital for pertrochanteric femoral fractures. Using these data, an algorithm was built to predict the need for ABTs. We hypothesised that we would be able to predict which of the patients who were admitted for pertrochanteric femoral fracture would need an ABT. By better managing this process, patients are more likely to receive optimal medical care without straining hospital resources, thereby reducing the cost of care. Patients and methods Patients This retrospective database study was based on data collected from the hospital electronic medical record (EMR) system and the blood bank database. Approval for conducting this study was received from the local research committees. A total of 408 patients were admitted for pertrochanteric femoral fractures and underwent surgical treatment between 2007 and 2012; however, only 220 of these patients met the inclusion criteria (Table 1). We excluded all patients who were treated without surgery. Risk factors for an ABT that were assessed included patient demographics (i.e., age and gender), fracture type, American Society of Anaesthesiology (ASA) grade, anaesthesia type, time between injury and surgery, laboratory values (i.e., Hb at admission, mean corpuscular volume [MCV], international normalised ratio [INR], transaminase and creatinine), the use of anticoagulants, and surgical details (i.e., internal fixation type, duration of operation, intra-operative blood loss, and postoperative drainage). Complications such as infections (pneumonia and surgical site infection), cardiac events, cerebro-vascular events and 30-day mortality were also recorded, as well as the number of units of transfused blood. Monitoring of Hb and transfusion protocol At admission to hospital, all patients had a complete blood count (CBC) analysis; two samples of blood were taken for the ABO Table 1 Study criteria. Inclusion criteria (1) older than 60 years (2) low energy injury without concomitant fractures or injuries that might require an ABT (3) treated with close reduction and minimally invasive surgery (4) blood count measured at admission, post-operative days one and three, and post-transfusion Exclusive criteria (1) pathologic fracture due to malignancy and high energy injury (2) history of blood system diseases and Hb <80 g/L at admission (3) severe liver or kidney dysfunction
Rh group type and antibody screening and two units of packed red blood cells (RBCs) were cross-matched and ordered. An additional cross-match was carried out thereafter as necessary because the previous sample had expired according to blood bank safety regulations. A CBC was taken on post-operative days one and three, when the patients presented signs of symptomatic anaemia (i.e., tachycardia, weakness and palpitations) and after blood transfusion. Our institute currently adheres to a ‘‘restrictive transfusion policy’’. According to our transfusion protocol, patients were preoperatively transfused to an Hb level of 90 g/L. The patients were transfused post-operatively when their Hb level dropped below 80 g/L. This ‘‘transfusion trigger’’ was raised to 90 g/L in clinically symptomatic patients (i.e., extreme weakness, chest pain, extreme paleness or major bleeding), when vital signs were abnormal (i.e., tachycardia, heart rate [HR] >100 bpm, and systolic blood pressure [SBP] <90 mmHg); and in those with a history of heart disease (i.e., coronary, valvular problems and arrhythmia) or cerebro-vascular disease. Surgical procedure All patients were treated on a traction table under the guidance of G-arm fluoroscopy. All of the procedures were performed by 4 senior surgeons in our institution using modified techniques, who are experienced in the management of pertrochanteric fractures. Moreover, duration of operation in this study was measured from the time of close reduction to incision closure. Surgeons did not perform surgery directly under imaging intensify in order to avoid excess of X-ray exposure. When the image intensifier was used to assess the reduction of fracture, the locations of the guide wire and internal fixations, the surgeons were behind the tinplate. Forty patients received a dynamic hip screw (DHS, Synthes, Switzerland) fixation; 84 patients received a proximal femoral nail anti-rotation (PFNA, Synthes, Switzerland) fixation, and 96 patients received a Gamma3 nail (Stryker, MI, USA) fixation. Statistic analyses Statistical analyses were conducted using SPSS 13.0 (SPSS Inc., USA). Univariate analyses were used to compare the group that received an ABT within 72 h after surgery with the group that did not with regard to 13 variables. Continuous, normally distributed variables were analysed using a two-sample Student’s t test; continuous, non-normally distributed variables were analysed using the Wilcoxon-rank-sum test; Pearson’s chi-square test and Fisher’s exact test were used to compare the groups with respect to categorical variables; a one-way ANOVA was used to compare differences in groups, and the LSD method was used to compare differences between groups. A P value of <0.05 was considered to indicate statistical significance. A multivariate logistic regression was applied to identify the significant independent predictors of ABT. The full regression model included 13 risk factor candidates. Model selection methods such as Wald-backward elimination were used to identify significant factors from the 13 explanatory variables. Results The data of all 220 patients were included in the final analysis. The patients were divided into non-ABT and ABT groups. Table 2 shows the general characteristics and the univariate analysis of the study population. A total of 89 patients (40.5%) received an ABT within the first 72 h after surgery, 5 patients were transfused to an Hb level of 90 g/L pre-operative, 76 patients were transfused when their Hb levels dropped below 80 g/L post-operative, and 8 patients
Please cite this article in press as: Hou G, et al. Predicting the need for blood transfusions in elderly patients with pertrochanteric femoral fractures. Injury (2014), http://dx.doi.org/10.1016/j.injury.2014.08.033
G Model
JINJ-5853; No. of Pages 6 G. Hou et al. / Injury, Int. J. Care Injured xxx (2014) xxx–xxx
3
Table 2 General characteristics and univariate analysis of the risk factors for ABT within 72 h after surgery in 220 patients. Variable
Total
Non ABT group
ABT group
P value
Number Gender (male/female) Age (years) mean SD AO fracture type (A1/A2/A3) ASA grade ( / )§ Anaesthesia type (general/regional) Time between injury and surgery (days)mean SD Duration of operation (min) mean SD Hb at admission (g/L) mean SD Internal fixation type (DHS/PFNA/Gamma3) Anti-coagulation (preoperative Yes/No) Anti-coagulation (postoperative Yes/No) Intra-operative blood loss (mL)mean SD Postoperative drainage (mL)mean SD Average volume of ABT (mL) mean SD Actual average volume of ABT (mL) mean SD
220 84/136 78.2 6.8 65/144/11 128/92 79/141 4.5 3.9 76.7 27.6 113.6 16.1 40/84/96 124/96 194/26 107.0 93.0 24.7 60.3 250.9 357.5 –
131 55/76 77.2 6.8 50/77/4 79/52 39/92 4.3 3.6 72.7 23.4 119.5 14.7 30/49/52 71/60 114/17 92.0 63.0 25.8 54.2 – –
89 29/60 79.7 6.6 15/67/7 49/40 40/49 4.7 4.2 82.6 32.1 104.9 14.1 10/35/44 53/36 80/9 129.2 121.7 23.1 68.5 – 620.2 294.0
– 0.159¥ 0.008* 0.002¥ 0.439¥ 0.021¥ £ 0.946 £ 0.025 0.000* 0.075¥ 0.432¥ 0.518¥ £ 0.041 £ 0.316 – –
ABT, allogeneic blood transfusion; AO, Arbeitsgemeinschaft fu¨r Osteosythese; ASA, American Society of Anaesthesiology; Hb, haemoglobin; DHS, dynamic hip screw; PFNA, proximal femoral nail anti-rotation. The values are given as the mean and standard deviation for continuous variables and as the number of patients for categorical variables. * Two sample t-test. ¥ Chi-square test. £ Wilcoxon-rank-sum test. § Group contains ASA I and II, whereas group contains ASA III and IV because separate groups of patients classified as ASA I (n = 4) or ASA IV (n = 2) were too small to be analysed separately.
were transfused post-operatively when clinical symptoms were detected. The total volume of blood transfused was 276 units, and the actual average blood transfusion was 3.1 1.47 units. Pneumonia developed in 5 patients who had fully recovered during the follow-up; fat liquefaction occurred in one patient who healed after a continued dressing change; and wound infection occurred in 2 patients after surgery, but the wounds healed after antibiotic treatment and extensive debridement. The 30-day mortality rate was 3.6%. An additional analysis demonstrated that ABT was not associated with post-operative complications such as infections (i.e., pneumonia or surgical site infection), cardiac and cerebro-vascular events or 30-day mortality (Table 3). Risk factors for ABT within 72 h after surgery were assessed with a univariate analysis (Table 2). Significantly different factors (P < 0.05) included age (79.7 6.6 years vs. 77.2 6.8 years), Hb level at admission (104.9 14.1 g/L vs. 119.5 14.7 g/L), duration of surgery (82.6 32.1 min vs. 72.7 23.4 min), blood loss during operation (129.2 121.7 mL vs. 92.0 63.0 mL), more severe injuries (AO fracture type A2 and A3 vs. A1), and anaesthesia type (regional vs. general anaesthesia). The mean volumes of blood transfused in the PFNA and Gamma3 groups were 269.1 mL and 285.5 mL, respectively, which were greater than that transfused in the DHS group (130.0 mL, P < 0.05). A multivariate logistic regression was performed to simulate a decision analysis situation when admitting a patient with a pertrochanteric femoral fracture. Five out of 13 independent variables were found to have a statistically significant effect on the rate of ABT within 72 h after surgery: age, Hb level at admission, Table 3 Comparisons of the post-operative complications between patients with and without ABT. Post-operative complications
Non ABT group
ABT group
P value
Pneumonia Surgical site infection Cardiac related complications Cerebro-vascular events 30-day mortality
3/131 1/131 2/131 2/131 5/131
2/89 1/89 4/89 2/89 3/89
1.000* 0.647¥ 0.366* 1.000* 1.000*
* ¥
Chi-square test. Fisher’s exact test.
type of internal fixation, duration of operation and blood loss during operation. Regression coefficients, likelihood ratios, P values, adjusted odds ratios, and 95% confidence intervals were determined (Table 4). This regression model predicted 74.1% of the transfused cases. A receiver operating characteristic (ROC) curve analysis was used to evaluate the predictive performance of the logistic regression model and its ability to predict ABT within 72 h after surgery (Fig. 1). The area under the curve was 0.824, which demonstrated an acceptable diagnostic performance. A predicted probability (Fig. 2) P (ABT 72 h) of 0.373 was selected as an optimal cut-off point to best differentiate patients who should receive ABT within 72 h after surgery from those who should not. This cut-off point has the highest sensitivity and specificity rates; and it has a false negative rate of 20.2% and a false positive rate of 29.0%. To further characterise the patients, they were divided into two groups stratified by age, Hb level at admission and duration of operation (cut-off point: P = 0.5). A significant difference was observed between the two groups (Table 5). Patients who were elderly (>81years), had lower Hb levels at admission (Hb 124 g/ L) and longer duration of operation (t > 85 min), were considered as the significant factors for the need for blood transfusions. Discussion Patients with pertrochanteric femoral fractures who are treated operatively frequently require the utilisation of ABT. However,
Table 4 Results of the final logistic regression model to predict ABT within 72 hours after surgery in patients with pertrochanteric femoral fractures. Multivariate predictor Age Hb at admission Type of internal fixation Duration of operation Blood loss during operation
Regression coefficient 0.012 0.086 0.810 0.021 0.006
Odds ratio
95% Confidence intervals
P value
1.012 0.917 0.445 1.021 1.006
0.958–1.069 0.893–0.942 0.275–0.720 1.004–1.038 1.001–1.010
0.034 0.000 0.001 0.013 0.016
Please cite this article in press as: Hou G, et al. Predicting the need for blood transfusions in elderly patients with pertrochanteric femoral fractures. Injury (2014), http://dx.doi.org/10.1016/j.injury.2014.08.033
G Model
JINJ-5853; No. of Pages 6 G. Hou et al. / Injury, Int. J. Care Injured xxx (2014) xxx–xxx
4
Table 5 Comparison of the two patient groups stratified by age, haemoglobin at admission and duration of operation.
ABT group Non-ABT group
Age (years)
Hb at admission (g/L)
Duration of operation (min)
81
124
85
>81
54 35 96 35 P = 0.049*
>124
82 7 82 49 P < 0.001*
>85
57 32 99 32 P = 0.045¥
ABT, allogeneic blood transfusion; Hb, haemoglobin. * Chi-square test. ¥ Fisher’s exact test.
patients older than 81 years was obviously higher than that in younger patients (P < 0.05). Swain [17] also reported a significantly greater transfusion need for patients aged 80 years and above. This finding was most likely related to the lower Hb level at admission in the elderly group compared with their younger counterparts (109.2 g/L vs. 115.3 g/L). Comorbid disease, organ dysfunction and poor compensatory ability all lead to a high incidence of ABT in the elderly. Hb at admission Fig. 1. A ROC curve analysis was used to evaluate the predictive performance of the logistic regression model to predict ABT within 72 h after surgery.
patients with normal preoperative Hb values may undergo surgery without ABTs. The early identification of these patients might save unnecessary blood cross-matching. The purpose of the current study was to define a method to predict the need for an ABT in elderly patients admitted with pertrochanteric femoral fractures. This study is important because the ability to predict the need for ABT is the first step in managing blood transfusions in pertrochanteric femoral fracture patients. Managing these transfusions has direct implications for patients who do not benefit from a transfusion, furthermore, it has economic implications for the institutional blood bank. A reasonable transfusion program might reduce the complications caused by anaemia and effectively reduce the economic expense, avoid the risks of blood transfusion [11–13] and improve the perioperative function state [3,14]. This study focused on predictors of ABT and the algorithms that predict the need for such transfusions based on available institutional data and transfusion protocols. Logistic regression analysis revealed that patients with pertrochanteric femoral fractures who are elderly (>81 years), have lower Hb levels at admission (Hb 124 g/L), undergo longer surgeries (t > 85 min) and intramedullary fixation (Gamma3 and PFNA) and experience more intra-operative blood loss are more likely to need an ABT. The results of the present study are instructive with regard to certain aspects of the ABT requirements in this population: Age Old age can predicts the need for perioperative blood transfusions in patients with pertrochanteric femoral fractures [15,16]. This study found that the ABT group was older on average than the non-ABT group. Moreover, the incidence of ABT in
The average Hb level at admission of the non-ABT group was higher than that of the ABT group (119.5 g/L vs. 104.9 g/L). Furthermore, this study found that values of 124 g/L proved as best predictor of blood transfusion, approaching a 7-fold risk for transfusion (OR = 7.0, P < 0.001). This result is in accordance with previous observations [15,16,18,19]. Robbins [20] also reported that only 1 in 9 patients with Hb levels >11 g/dL received a transfusion. Kurdy and Hokan [21] advised that cross-matching should be routinely ordered for patients with Hb levels <12 g/dL. Adunsky [1] reported that patients with normal Hb levels (>12 g/ dL) could undergo surgery after typing without cross-matching or transfusion. Shokoohi [16] indicated that patients with anaemia at admission were approximately six times more likely to have a transfusion compared with non-anaemic patients. Type of internal fixation Patients treated with PFNA and Gamma3 had transfusion indexes of 1.35 units and 1.43 units, respectively, compared with 0.65 units for the DHS group. These results are in accordance with previous observations [1,15,18,22]. As orthopaedic surgery techniques have advanced and mechanic implants have improved (especially after the introduction of minimally invasive surgery), only small volumes of gross blood loss have occurred during operations. However, Foss reported large volumes of hidden blood [23]. The opening of the medullary cavity and proximal reaming might increase hidden blood loss, which made a higher incidence of ABT in patients treated with intramedullary fixation devices (i.e., PFNA and Gamma3). Duration of operation The average operation time in the transfusion group was longer than that of the non-transfusion group (82.6 min vs. 72.7 min), and
Fig. 2. A predictive formula based on the five significant risk factors model used to predict the need for ABT within 72 h after surgery in patients with pertrochanteric femoral fractures. ABT 72 h, ABT within 72 h after surgery; Hb, haemoglobin; DHS, dynamic hip screw; D, duration of operation; V, volume of blood loss during operation.
Please cite this article in press as: Hou G, et al. Predicting the need for blood transfusions in elderly patients with pertrochanteric femoral fractures. Injury (2014), http://dx.doi.org/10.1016/j.injury.2014.08.033
G Model
JINJ-5853; No. of Pages 6 G. Hou et al. / Injury, Int. J. Care Injured xxx (2014) xxx–xxx
5
Fig. 3. An algorithm designed to predict the need for ABT. Hb, haemoglobin; CBC, complete blood count; t, duration of operation; post-op, post-operative.
a significantly higher incidence of ABT was observed after surgeries longer than 85 min, comparing with the counterparts. To our knowledge, this result has not been previously reported. The repeated mobility of the fracture parts, long duration of the medullary cavity opening and regional hemorrhage [24] all increased the likelihood of postoperative ABT. Blood loss during operation The average volume of blood loss during operation in the ABT group was greater than that in the non-ABT group (129.2 mL vs. 92.0 mL), and more blood loss during operation might predict the need for blood transfusion. Morritt and Foss also reported this result [22,23]. Insufficient haemostasis [25], repeated mobility of the fracture parts, and long duration of medullary cavity opening might all increase the volume of blood loss during operation. Furthermore, a post-operative ABT was required when the Hb levels dropped below 80 g/L. The current study should be interpreted in the light of its limitations. This retrospective study was based on prospectively collected database information; and mathematical model may not be able to fully explain the clinical problems; therefore, randomised controlled trials are necessary to confirm its results. Data such as radiographs were not reviewed. Moreover, the ASA grade rather than co-morbidities was used as a risk factor. Conclusion The current study recommends the ‘‘restrict transfusion policy’’, and an algorithm (Fig. 3) was designed to predict the need for an ABT. Cross-matching should be routinely ordered for patients with pertrochanteric femoral fractures who are older than 81years or have Hb levels below 124 g/L at admission; changes of patients’ CBC should be closely monitored when operation times exceed 85 min, when an intramedullary fixation device is used and when a large volume of blood is lost. Additionally, ABT is needed for patients whose Hb levels drop below 80 g/L, who have clinical symptoms (i.e., extreme weakness, chest pain, extreme paleness or major bleeding), or abnormal vital signs (i.e., tachycardia, HR >100 bpm, and SBP <90 mmHg). Conflicts of interests and disclosure of funding No funding was received for this paper. We declare that there are no conflicts of interests concerning this paper.
Acknowledgements The authors thank Dr. Ru Chen and Huijun Zhang for their help in designing the statistical analysis. References [1] Adunsky A, Lichtenstein A, Mizrahi E, et al. Blood transfusion requirements in elderly hip fracture patients[J]. Arch Gerontol Geriatr 2003;36(1):75–81. [2] Halm EA, Wang JJ, Boockvar K, et al. The effect of perioperative anemia on clinical and functional outcomes in patients with hip fracture. J Orthop Trauma 2004;18:369–74. [3] Lawrence VA, Silverstein JH, Cornell JE, et al. Higher Hb level is associated with better early functional recovery after hip fracture repair. Transfusion 2003;43:1717–22. [4] Friedman BA, Burns TL, Schork MA. A study of blood utilisation by diagnosis, month of transfusion, and geographic region of the United States. Transfusion 1979;19:511–25. [5] Carson JL, Altman DG, Duff A, et al. Risk of bacterial infection associated with allogeneic blood transfusion among patients undergoing hip fracture repair. Transfusion 1999;39(7):694–700. [6] Levi N, Sandberg T. Blood transfusion and postoperative wound infection in intracapsular femoral neck fractures. Bull-Hosp Joint Dis 1998;57(2):69–73. [7] Koval KJ, Rosenberg AD, Zuckerman JD, et al. Does blood transfusion increase the risk of infection after hip fracture? J Orthop Trauma 1997;11(4):260–5. [8] Carson JL, Hill S, Carless P, et al. Transfusion triggers: a systematic review of the literature. Transfus Med Rev 2002;16(3):187–99. [9] Khan AM, Mushtaq N, Giannakas K, et al. Cross-match protocols for femoral neck fractures—finding one that can work. Ann R Coll Surg Engl 2004;86(1):11–4. ˜ oz M, Leal-Noval SR. Restrictive transfusion triggers in major orthopaedic [10] Mun surgery: effective and safe? Blood Transfus 2013;11(2):169–71. [11] Bierbaum BE, Callaghan JJ, Galante JO, et al. An analysis of blood management in patients having a total hip or knee arthroplasty. J Bone Joint Surg 1999;81(1):2–10. [12] Goodnough LT, Shander A, Brecher ME. Transfusion medicine: looking to the future. Lancet 2003;361(9352):161–9. [13] Dodd RY. Current estimates of transfusion safety worldwide. Dev Biol 2004;120:3–10. [14] Foss NB, Kristensen MT, Kehlet H. Anaemia impedes functional mobility after hip fracture surgery. Age Ageing 2008;37(2):173–8. [15] Kadar A, Chechik O, Steinberg E, et al. Predicting the need for blood transfusion in patients with hip fractures. Int Orthop 2013;1–8. [16] Shokoohi A, Stanworth S, Mistry D, et al. The risks of red cell transfusion for hip fracture surgery in the elderly. Vox Sang 2012;103(3):223–30. [17] Swain DG, Nightingale PG, Patel JV. Blood transfusion requirements in femoral neck fracture. Injury 2000;31(1):7–10. [18] Bajs ID, Cvjetko I, Vicic VK, et al. Blood transfusion in patients with pertrochanteric femoral fracture. Acta Med Croatica 2006;60(1):7–10. [19] Cuenca J, JA GE, Martinez AA, et al. Preoperative blood test results and type of fracture as transfusional risk factors in patients older than 65 years with trochanteric hip fracture. Rev Esp Anestesiol Reanim 2004;51(9):515–22. [20] Robbins J, Steingold RF. Blood use in urgent operations for patients with fractures of the femoral neck. Injury 1986;17(4):265–6. [21] Kurdy NMG, Hokan R. A cross-matching policy for fractures of the proximal third of the femur. Injury 1993;24(8):521–4.
Please cite this article in press as: Hou G, et al. Predicting the need for blood transfusions in elderly patients with pertrochanteric femoral fractures. Injury (2014), http://dx.doi.org/10.1016/j.injury.2014.08.033
G Model
JINJ-5853; No. of Pages 6 6
G. Hou et al. / Injury, Int. J. Care Injured xxx (2014) xxx–xxx
[22] Morritt DG, Morritt AN, Kelley SP, et al. Blood ordering protocol based on proposed surgical implant in fractured neck of femur patients. Ann R Coll Surg Engl 2005;87(6):445–8. [23] Foss NB, Kehlet H. Hidden blood loss after surgery for hip fracture. J Bone Joint Surg BrVol 2006;88(8):1053–9.
[24] Zhang PX, Xue F, An S, et al. Clinical analysis of obvious and hidden blood loss in inter-trochanter fracture patients treated with proximal femoral nail antirotation and dynamic hip screw. Beijing Da Xue Xue Bao 2012;6:891–4. [25] Smith GH, Tsang J, Molyneux SG, et al. The hidden blood loss after hip fracture. Injury 2011;42(2):133–5.
Please cite this article in press as: Hou G, et al. Predicting the need for blood transfusions in elderly patients with pertrochanteric femoral fractures. Injury (2014), http://dx.doi.org/10.1016/j.injury.2014.08.033