Frequency of Transfusions and Risk Factor for Bleeding Risk to Guide a Blood-Sparing Program during Hip Arthroplasty in Gabon ()
1. Introduction
Arthroplasty is a common intervention in orthopedic surgery. It is an intervention with a high hemorrhagic risk, it is often accompanied by blood transfusion. Despite improvements in transfusion safety in recent years, allogeneic blood transfusion still leads to complications, including allergic reactions, infections, circulatory overload and lung damage [1] . In addition, studies have shown que blood transfusions can be responsible for prolonged hospitalization, mortality and high postoperative morbidity during hip and knee arthroplasty [2] [3] . Hip arthroplasty has been performed in Gabon since 2011 at the Omar Bongo Ondimba Army Training Hospital in Libreville (HIAOBO). This surgical activity was accompanied by a blood transfusion which seemed important to us and did not benefit from a specific protocol concerning transfusion, unlike the prevention of the risk of infection which was the subject of a protocol with antibiotic prophylaxis, research urinary, dental, sinus and pulmonary infection. No study had been carried out on transfusion practice during hip arthroplasty in Gabon. The aim of the study was to look for risk factors for transfusion during hip arthroplasty at the Omar Bongo Ondimba Army Training Hospital in Libreville. (HIAOBO) to guide a blood saving program
2. Material and Method
Retrospective, descriptive, analytical study. Performed in the Anesthesia Resuscitation department of HIAOBO over 11 years, from January 2011 to June 2021.
Inclusion criteria: Patients who underwent prosthetic hip prosthesis surgery such as total hip prosthesis (THA) or intermediate hip prosthesis (IPH) were included.
Non-inclusion criteria: patients who underwent DHS (dynamic hip screw) or Gamma nail were not included.
Study variables: The variables studied were sociodemographic (age, sex, weight, height, BMI, history), comorbidities (arterial hypertension, diabetes, renal failure, HIV, sickle cell disease), ASA score. Biologicals (NFS, urea, creatinine, blood ionogram, coagulation assessment, GsRh, RAI, bilirubin, troponin). Operating variables (anesthesia, type of prostheses, tranexamic acid, operating time, blood loss), per and post complications operations, per and post-operative blood transfusion.
The variables studied were obtained from medical records (sociodemographic variables, comorbidities), surgical records (postoperative clinical and biological data) and anesthetic records (technical and operative data). All these variables were listed on a survey sheet separated into 3 periods (pre, per and post operative). After discharge from hospital, patients were systematically seen by the surgeon on D14, D21 and D30 as part of the follow-up to obtain late data. All the operated patients carried out an infectious assessment in search of a urinary, dental, sinus infection and antibiotic prophylaxis during the surgery. A cardiological evaluation by electrocardiogram and echocardiography was performed in all patients.
Statistical analysis: data were entered on computer and analyzed using Epi Info 7.2 software, IBM SPSS version 21 and Excel 2016. For data comparison, we used the Chi2 test, the Pearson test and the significance level < 5%. The quantitative descriptive variables were expressed on average with the standard deviation. Qualitative variables were described as percentages. Nonparametric Kruskal Wallis tests for independent samples were used to compare quantitative variables by linear regressions. This made it possible to link together different quantitative variables with determination of the Pearson correlation coefficient. The difference was significant for p values less than 0.05.
3. Results
• Demography
The sample consisted of 276 patients with an average age of 64.3 ± 15.7 years with extremes of 17 and 95 years. There were 163 (59.3%) female patients and 112 (40.7%) male patients, a sex ratio of 0.69. The most represented comorbidities were arterial hypertension (42.7%) and diabetes (10.9%). The mean hemoglobin level was 11.9 ± 1.9 g/dl with a minimum of 5.7 and a maximum of 18.6. The population had an ASA score around 2. Patient characteristics are given in Table 1(a).
The population was heterogeneous with two types of patients. Patients who benefited from total hip arthroplasty (THA) (64.9%) and another from intermediate hip arthroplasty (IHA) (35.1%) patients. Table 1(b) compares the two populations.
• Perioperative data
Spinal anesthesia was mainly performed. The operating time was relatively long. The mean hemoglobin level was 11.9 ± 1.9 g/dl with a minimum of 5.7 and a maximum of 18.6. Table 2 summarizes the intraoperative data.
• Transfusion aspects
The average number of RBCs transfused was 1 with a maximum of 3 per patient and an average blood loss of 528.8 ± 405.8 ml and a maximum of 1900 ml. Transfusions were preferentially performed postoperatively with 96 (34.8%) patients for a total of 157 (56.9%) patients transfused. Tranexamic acid was used in 45 (16.3%) patients. No significant association between transfusion and tranexamic acid (Table 3).
• Univariate analysis: Comparison of transfused and non-transfused patients shows a significant difference in age, values hemoglobin, hematocrit and ASA score (Table 4).
(a) (b)
Table 1. (a) Demographics; (b) Demographies.
Table 4. Analysis of transfused and non-transfused populations.
• Multivariate analysis
After multivariate analysis, only hemoglobin level and Total Hip Arthroplasty remained significantly related to transfusion. All other factors (anesthetic technique, age, sex, operating time) were not related to transfusion (Table 5).
• Postoperative complications
Complications were mainly cardiovascular with 5 (1.8%) patients, respiratory and infectious complications had the same rate with 2 (0.7%) patients The number of deaths was 4 (1.4%) patients, all related to coronary ischemia in patients patients admitted to intensive care (Table 6).
Table 5. Risk factors associated with transfusion by logistic regression.
Table 6. Post operative complications.
4. Discussion
The aim of this study was to investigate the factors affecting perioperative hemoglobin loss and the frequency of transfusions in patients undergoing hip arthroplasty.
• Epidémiology
Study carried out on a heterogeneous population composed of patients having benefited from intermediate hip prosthesis (PIH) and patients benefiting from total hip prosthesis (THP). PIH is performed for fracture of the femoral neck, a pathology most often affecting elderly female subjects, while patients most often benefiting from THA for hip osteoarthritis are younger. Tables 1a and 1b summarize the characteristics of the two populations.
• Transfusion
The incidence of transfusion is high in our study population (56.9%), with relatively modest blood loss. Total hip arthroplasty (THA) is one of the most performed orthopedic procedures in the world. Studies show that up to 46% of patients require red blood cell transfusions during or after hip or knee replacement surgery [4] [5] [6] . As a risk factor for transfusion, it appeared the preoperative hemoglobin level and the installation of THA. Preoperative anemia is common in orthopedic arthroplasty surgery and is prevalent in approximately 25% of patients [2] [3] [7] [8] . Low preoperative Hb has been identifed as an independent risk factor for postoperative transfusions after hip or knee arthroplasty [9] [10] [11] Ryan and al. reported a Hb of 12.5 g/dL as an optimal cutoff for predicting postoperative transfusion with a specificity 76.4% and a sensitivity of 84.8% in knee arthroplasty [12] . This anemia may be associated with high morbidity and mortality and a high incidence of red blood cell transfusions [7] . Our study finds an overall average of 11.9 g/dl including 11.5 g/dl for patients receiving PIH and 12.2 g/dl for patients receiving PTH, these preoperative hemoglobin levels partly explain the high rate of transfusion.
The total hip prosthesis was found to be a risk factor for transfusion, in fact, the intermediate hip prosthesis is known to be less hemorrhagic than the total prosthesis, which would be explained by the fact that the surgical procedure during the total prosthesis is more heavy with more bleeding according to some studies [13] [14] . Liodakis and al found an increase in the number of transfusions in the total hip arthroplasty versus intermediate hip arthroplasty [15]
Tranexamic acid has been used very little in our practice (45 patients) mainly for reasons of availability, its low impact on the reduction of bleeding is certainly linked to selection bias and the small sample of patients who have benefited from this treatment. Since a long time Tranexamic acid has been proved to be effective in preventing blood transfusions as well as reducing perioperative blood loss. Meta-analysis of Sadigursky (7 randomized clinical trials, 948 patients) and Rajiv Gandhi (33 clinical trials, 403 patients) reported a reduced blood loss and lower transfusion rates for patients receiving tranexamic acid, without an increased risk for deep vein thrombosis, pulmonary embolism or other complications [16] [17] . These findings are consistent with our results. However, some outcomes indicate that the avoidance of suction drains might be as important as the use of tranexamic acid for the reduction of perioperative blood loss. There are many areas for improvement in order to reduce transfusion: this is how perioperative blood optimization programs have been developed in the last years to reduce transfusion frequency and to increase patient safety. These programs include the pre-operative optimization of the hemoglobin level and the optimal hemostasis obtained during surgery.
Various studies have shown a significant lower blood loss by using the minimal invasive surgery compared to the standard approach due to smaller skin and muscle incisions [18] [19] . This technique is not yet available to us, hip arthroplasty is a recent surgery in our structure, since 2013
- Optimal hemostasis during surgery was maintained by avoiding hypothermia and optimization of pH level. A meta-analysis reported that even mild hypothermia significantly increases blood loss and the risk for transfusion by 22% [20] . In addition, RBC transfusions were administered restrictively only in patients with Hb ≤ 6 g/dL or in symptomatic patients with Hb ≤ 8 g/dL
-The coadministration of erythropoietin has shown to further reduce blood transfusions in patients undergoing hip surgery [21] . Several studies have reported that blood transfusions are associated with an increased perioperative mortality, more complications, especially surgical site infections and prolonged hospitalization [4] [7] [22] [23] . In our study there were 12 complications between Day 7 and Day 21 mainly coronary ischemia, 2 infections and 04 deaths (Table 6).
- Some studies show that anesthetic technique can influence bleeding and transfusion during hip arthroplasty, particularly total arthroplasty. According to Owen and al, patients who received spinal anesthesia had lower mean pain scores throughout length of their hospitalizations, and required fewer blood transfusions (OR, 0.7), length of stay was not significantly different between the groups, and there was no difference in readmissions at 30 or 90 days or venous thromboembolic events at 90 days [24] . The Memtsoudis meta-analysis (94 studies) also shows a reduction in bleeding and transfusion in patients undergoing spinal anesthesia [25] .
5. Study Limitations
1) The present study has several limitations, mainly due to its retrospective design. Secondly, patients have been operated by different surgeons, which might have influenced bleeding. Thirdly, the criteria for insertion of suction drains were systematic, thus, surgeons’ expertise and patient risk factors may have influenced the decision on the use of suction drains.
2) Not taking into account blood loss through the drains. The insertion of drains after knee or hip arthroplasty is still a matter of controversy [26] [27] [28] . tranexamiqueThe insertion of a suction drain has no scientifically proven benefits and drains are not used routinely anymore. Using enhanced recovery protocols in the arthroplasty leads to a changing role for drains, particularly with the use of tranexamic acid. Most literature is from the pre-tranexamic era and there are only few studies on the use of drains in combination with tranexamic acid. In our study, this aspect was not taken into account although in our practice, the installation of drains is very frequent and depends on the surgeons.
3) The absence of iron balance to look for iron deficiency anemia, probably because iron deficiency anemia has been known to be endemic in Africa for a very long time [29] . Iron deficiency anemia is the most common type of anemia in patients undergoing hip and knee replacement surgery and can be accompanied by increased length of hospital stay, high readmission rate at 90 days and postoperative complications [30] [31] .
6. Conclusions
Preoperative anemia and Total Hip Arthroplasty are the main risk factors for transfusion during hip arthroplasty. The establishment of a blood-saving program centered on modifiable factors (anemia, iron deficiency) associated with better use of tranexamic acid could reduce the incidence of transfusion, improve postoperative rehabilitation, save resources, reduce costs and improve patient safety.
Studies have shown the effectiveness of this kind of program. Polanco-García et al showed that a blood-sparing program reduced the frequency of transfusions from 41% to 16% by emphasizing the correction of preoperative anemia by the erythropoietin association and iron intake and intraoperative tranexamic acid [32] , Kopanidis and al. achieved a lower blood transfusion rate and higher postoperative hemoglobin values after implementing a patient blood management program [33] , Pinilla-Gracia and al. also showed lower transfusion rates, shorter length of stay and a corrected preoperative anemia in 79% of cases [34] . In a financial aspect, Fenelon and al reported on a 46% reduction of cross-matched blood and an annual cost saving of ?4,375 after introduction of an enhanced recovery program [35] .
In this multidisciplinary blood saving program, the anesthesia team plays a key role in coordinating and planning the various procedures
Authors’ Contributions
G. Edjo Nkilly: principal investigator, drafting the manuscript.
R. Okoue Ondo: inclusion and follow-up of patients.
A. Matsanga: inclusion and follow-up of patients.
PC. Nze Obiang: inclusion and follow-up of patients.
S. Oliveira: inclusion and follow-up of patients.
L. Nguiabanda: surgeon, inclusion and follow-up of patients.
JM Mandji-Lawson: reading final manuscript.
R. Tchoua: reading and final approval of the manuscript.