Risk factors for perioperative blood transfusion in patients undergoing hysterectomy for benign disease in a teaching institution

The purpose is to identify risk factors for perioperative blood transfusion in patients undergoing hysterectomy for benign disease. This study is a retrospective chart review including all the patients who underwent hysterectomy for benign disease between January 1st 2018 and December 31st 2019. Patients who received perioperative blood transfusion were identified and compared to those who did not. The following risk factors for blood transfusion were analyzed: route of hysterectomy, BMI, presence of adhesions, history of cesarean section, uterine weight. Descriptive statistics was used to analyze the data. A total of 517 patients were identified and included in the study. Forty-seven patients (9.09%) received a perioperative blood transfusion. The abdominal hysterectomy route (TAH) was a significant risk factor for receiving blood transfusion (p = 0.012). Other identified risk factors for blood transfusion included: Body mass index above 33.0 (p = 0.002), and uterine weight (p = 0.002). There was no association between the presence of pelvic adhesions (p = 0.91) or a personal history of cesarean section (p = 0.89) and receiving perioperative blood transfusion. When analyzing only the patients who underwent TLH, the presence of pelvic adhesion was found as a risk factor for perioperative blood transfusion (p = 0.024). The abdominal hysterectomy route, the presence of a large uterus, and obesity are risk factors for receiving a blood transfusion. Early identification of the patient at risk of requiring perioperative blood transfusion provides better patient counseling and surgical preparation.


Introduction
In the United States, hysterectomy is one of the most common surgeries performed in gynecology [1]. As per the Centers for Disease Control and Prevention (CDC), approximately 600,000 hysterectomies are done each year, and by the age of 60, more than 1/3 of all women have undergone a hysterectomy [2]. Hysterectomy is a relatively safe procedure regardless of the selected surgical route [3,4]. While safe, hysterectomies, like other surgical procedures, carry the risk of blood transfusion, and these risks depend on several factors. It is important to identify these prior to surgery as perioperative blood transfusions should, when possible, be avoided. Despite the safeguards in place today, blood transfusions are not without risks such as acute transfusion reactions, medical errors, viral, and bacterial infections among others [5][6][7].
Researchers have examined the risk factors for transfusion following hysterectomy for benign indications. Several studies have demonstrated a low pre-operative hemoglobin and hematocrit to be a risk factor for blood transfusion [4]. A retrospective study in 2012 of 441 hysterectomies, 137 of which received a transfusion, concluded that a low preoperative hematocrit was associated with the need for perioperative transfusion, more so when the hematocrit is less than 30% [4]. Another study examining 1208 women found that large uterine size, higher estimated blood loss (EBL) during the procedure, and increased operative time were associated with receiving a blood transfusion [8]. This, however, appears to have some contradictory evidence. In a cohort of 263 patients, age of the patient, type of hysterectomy, duration of surgery and uterine weight were not associated with an increased risk of blood transfusion [9]. Hillis et al. demonstrated that women undergoing hysterectomy with a uterus weighing more than 500 g were 1.6 times more likely to develop operative or postoperative complications, specifically cuff cellulitis and the need for blood transfusion [4].
The goal of our study was to identify risk factors for perioperative blood transfusion in patients undergoing hysterectomy for benign disease in our teaching institution.

Materials and methods
This study was approved by the institutional review board. Using electronic medical records, a systematic query was made regarding all types of hysterectomies for benign disease performed at our institution from January 1st, 2018, to December 31st, 2019.
Electronic medical records for each patient were accessed to obtain the following data: age, race, body mass index (BMI), history of c-section, history of vaginal deliveries, pre-op hemoglobin; surgical indication, route of hysterectomy (laparotomy, vaginal or laparoscopic), surgical operative reports, presence of adhesions, estimated blood loss (EBL), and pathology report for final diagnosis and uterine weight. Patients with gynecologic cancer diagnosis were excluded from this study.
Estimated blood loss was compared between routes of hysterectomy. In addition, we examined the following factors to determine their association with the need for blood transfusion: BMI > 30, presence of adhesions (bowel, omental or both), uterine weight, history of cesarean section, and pre-operative hemoglobin. The decision to transfuse patients intraoperatively was made on a case-by-case basis and at the discretion of the attending surgeon. If the attending surgeon felt that the estimated blood loss as determined by the amount of blood suctioned into the cannister during the surgery was over 500 mL, anesthesia was asked to draw a point of care hematocrit. If the hematocrit was below 24%, the patient was transfused.
Data were stored in a password protected online database. Pearson chi-square analysis and descriptive statistics were used to analyze the data. A P value of less than 0.05 was considered statistically significant.

Results
A total of 517 cases were identified and included in the analysis. Tables 1 and 2 outline the demographic information of the patients, including body mass index (BMI) and pre-operative hemoglobin. Forty-seven patients (9.09%) received a perioperative blood transfusion. The need for blood transfusion according to route of hysterectomy was as follows: Total abdominal hysterectomy (TAH): 34/263 (12.92%), total laparoscopic hysterectomy (TLH): 5/119   Table 3). The mean estimated blood loss (EBL) in a total abdominal hysterectomy was 327.17 cc ± 306.30, which is 187.8 cc, 42 cc, and 128.8 cc greater than the mean EBL noted in TLH, LAVH, and VH, respectively. Hysterectomies performed with the abdominal laparotomy route was a significant risk factor for perioperative blood transfusion (p = 0.012) ( Table 3). The average uterine weight of those who did not receive a transfusion was 541.40 g ± 795.93 while the average uterine weight of those who required transfusion was 961.43 g ± 810.881. This difference was found to be statistically significant (p = 0.001) ( Table 4). The mean BMI of patients requiring blood transfusion was 32.82 kg/m 2 ± 7.20, while the BMI in the remainder of the cohort not receiving blood transfusion was 29.59 kg/m 2 ± 7.09. This difference was also statistically (p = 0.003) ( Table 4).
Total abdominal hysterectomy, BMI and uterine size were all found to be statistically significant risk factors associated with need for perioperative blood transfusions during hysterectomy for benign disease.
The pre-operative mean hemoglobin in patients who did not receive a blood transfusion was 11.76 g/dl ± 1.87, compared to the mean hemoglobin in those who did receive a transfusion (9.43 g/dl ± 1.92), which was statistically significant (p < 0.001).
There was a statistically significant difference between the presence of adhesions (p = 0.026) and the need of blood transfusion. There was no association between a personal history of cesarean section (p = 0.123) and receiving a blood transfusion.
We explored the correlation between uterine size and risk of receiving a blood transfusion. In the total abdominal hysterectomy group, the average uterine weight in those who did receive a transfusion was 1,196.63 g, while the average weight in those who did not receive a transfusion was 896.33 g. This was not found to be a statistically significant difference (p = 0.094). When analyzing the patients who had a minimally invasive approach (TLH + LAVH + VH) the average uterine weight of those that received a blood transfusion compared to those who did not was similar (p = 0.87). Therefore, uterine weight was not identified as an independent risk factor for blood transfusion when controlling for route of hysterectomy.
We also looked at the risk of having a blood transfusion based on patient's race when controlling for type of hysterectomy. Hispanic patients were 1.21 more likely to receive a blood transfusion compared to Black women (OR 1.12), however the difference was not statistically significant (p = 0.71) ( Table 5). When looking at risk of conversion to laparotomy of patients who had a laparoscopic approach (TLH + LAVH) based on race, it was noted that Hispanic women were 0.2 times more likely to undergo conversion to laparotomy, and this value was considered statistically significant (p = 0.01) ( Table 6).

Discussion
In this cohort, 9.09% of all women received a perioperative blood transfusion. The greatest risk factor for receiving a blood transfusion was the open abdominal hysterectomy route compared to a minimally invasive procedure. It is  well known that the open abdominal route involves a larger incision, more tissue manipulation and need for exposure and dissection. These factors put tissue integrity at risk and increases likelihood of blood loss and subsequent need of blood transfusion. However, a factor frequently used by surgeon when choosing the route of hysterectomy is the preoperative estimation of the uterine size. This leads to another risk factor for perioperative blood transfusion: uterine weight. The average uterine weight in patients receiving a transfusion was 961.43 g which was significantly higher than the average weight in those who did not receive a transfusion (541.40 g). Other studies have supported this finding by demonstrating that there is greater surgical blood loss in women who undergo hysterectomy with larger uterine size. Unger et al. demonstrated that the risk of receiving blood transfusion was twice as much in women with uterine size greater than 1000 g compared to those less than 500 g [5].
Reasoning for greater blood loss and need for blood transfusion in those with larger uterine size largely involves the mechanics of surgical procedure. Larger uteri often have fibroids which can distort anatomy, making surgical exposure and ligation of vessels technically harder. Larger uteri are also more vascular, further increasing risk of blood loss. Our study did not show a significant difference in uterine weight with patient who received blood transfusions when controlling for type of procedure. This is likely multifactorial. There is not a large range in terms of uterine size amongst all minimally invasive surgeries so a significance between small differences in uterine weight would not necessarily be expected. Additionally, as alluded to earlier, open abdominal surgical route is usually performed due to larger uterine size, thus, uterine weight and mode of procedure are confounding factors as size is a cause of decision for open procedure.
Interestingly, there was a higher risk of blood transfusion in patients with a higher BMI (p = 0.003). In patients who did not receive a blood transfusion, the mean BMI was 29.53 kg/m 2 , whereas the mean BMI of patients who received a transfusion was 33.01 kg/m 2 . This increased risk of transfusion in obese patients may be related to the increased complexity of performing surgery in obese patients or the increased amount of blood in adipose tissue dissected on entry into the peritoneum [10]. The larger surface area of subcutaneous fat exposed in obese patients may explain greater blood loss due to increased venous and arterial bleeding points, however this theory applies primarily to open procedures [10]. Another theory also explaining the higher risk of blood transfusion during hysterectomy of obese patients, has to do with altered cardiorespiratory physiology in obese patients which may account for greater blood loss and in turn, greater risk of requiring blood transfusion. Increased ventilation pressures required to overcome lower pulmonary compliance may result in higher venous pressures, therefore causing greater losses through venous oozing [10].
As expected, a lower pre-operative hemoglobin level was found to be a risk factor for perioperative blood transfusion with the mean hemoglobin of those getting a blood transfusion of 9.45 g/dl, as opposed to 11.76 g/dl in those who did not receive a transfusion (p < 0.001). This appears to be comparable to the pre-operative hemoglobin levels identified in other studies [1,11]. Knowing the value of pre-operative hemoglobin can decrease the risk of blood transfusion and can assist with surgical planning and risk reduction from transfusion related reactions for patients [12]. If possible, the surgeon should elect to medically optimize the patient before proceeding with the surgery [13].
We acknowledge the weaknesses of our study, including the retrospective nature and the biases that are naturally incurred when the data are collected retrospectively. Similarly, there is a selection bias when comparing surgical approaches, as an open procedure may be selected if there is reason to believe that the uterus is too large for a minimally invasive approach. Another weakness encountered is the omission of robotic hysterectomies. Despite debates about increased cost of surgery and time, robotic surgical technology has gained popularity over the last decade due to believed better outcomes and decreased blood loss [14,15]. Information regarding various comorbidities that may influence whether a patient requires a transfusion, such as hypertension and diabetes mellitus, was not collected for this study. A novel study to determine the effects of these comorbidities on the risk of blood transfusion can be performed in the future. Additionally, surgeon skills and comfort level are different across the various modalities, and our study included multiple surgeons with different levels of expertise. This is a natural limitation to performing a large study in a teaching institution, but future studies may be performed with a smaller, more homogenous group of surgeons and measuring outcomes in their hysterectomies.
One of the strengths from this study is the sample size. While an even larger number of patients would give our study increased power, we believe that this number of patients was sufficient for the analysis of the variables that we analyzed. In addition, we attempted to control for several

Conclusion
Efforts to avoid the need for perioperative blood transfusion should always be considered when performing hysterectomy for benign disease. The open hysterectomy route, a large uterus, obesity, the presence of adhesions, and low preoperative hemoglobin are identified risk factors for perioperative blood transfusion. Women who are at greater risk of receiving perioperative blood transfusion based on the aforementioned factors should be counseled accordingly and adequately optimized for surgery to avoid unnecessary interventions at the time of surgery.
Author contributions MS-N, AU, JC: project development, data analysis manuscript writing/editing. DT: project development, data analysis. MG: data collection, data analysis. ML: data Analysis, data collection.

Funding
The authors received no funding for this manuscript and have no financial interest in it.

Availability of data and materials Not applicable.
Code availability Not applicable.

Conflict of interest
The authors declare that they have no conflict of interest.
Ethical approval Institutional review board approval was determined not needed for this work. The patients signed informed consent allowing the de-identified publication of this manuscript.

Consent to participate Not applicable.
Consent for publication All authors consent to the publication of this manuscript. This work has not been presented or published previously in any format.