Relationship Between Surgery Timing and Hidden Blood Loss in Proximal Femoral Nailing of Pertrochanteric Fractures in Elderly Patients: A Cross-sectional Study*

Objectives. Pertrochanteric fractures, commonly incurred by the elderly from low-energy falls, require surgical management. Visible blood loss in the perioperative period is inconsistent with the big drop in hemoglobin seen postoperatively. This discrepancy is attributed to the hidden blood loss (HBL), which must be anticipated in anemia management. This study aimed to determine if the amount of HBL in elderly patients with pertrochanteric femur fractures treated with short proximal femoral anti-rotation nailing (PFNa) is affected by the delay in time to surgery. Methodology. This is a detailed cross-sectional study from a single institution. Two hundred and ten patients admitted and operated on from January 2017 to December 2019 at an orthopedic specialty hospital were included in the study. Patient’s age, sex, AO classification, weight, height, operative time, and hematocrit levels on admission, within 7 days before surgery and immediately postoperatively; visual blood loss and blood transfused were reviewed and retrieved from medical records. Cases were grouped into early (<30 days) or late (≥30 days) surgery groups based on the time from injury to surgery. Total blood loss and hidden blood loss were computed based on the data. Results. There was no significant difference in the demographic and clinical characteristics of patients in both groups. Mean HBL was 113.65 mL (±99.25 mL) in the early surgery group and 95.32 mL (±111.79 mL) in the late surgery group. Mean HBLs were 31.47% and 27.88% of the total blood loss computed for the early and late surgery groups, respectively. Using an independent t-test, we noted no significant difference in the HBLs between the groups ( p = 0.22). Conclusion


INTRODUCTION
Pertrochanteric femur fractures are a common injury in the elderly population.Due to a globally aging population, hip fractures are estimated to number around 2.56 million by the year 2050. 1 Innovations in technology accompany this upsurge in cases.Most fractures are treated using either a proximal femoral nail anti-rotation (PFNa) or a dynamic hip screw (DHS).][4] However, this amount does not correlate with the postoperative drop in hemoglobin. 5Sehat et al., -in an experiment comparing visible blood loss and total blood loss in arthroplasty cases -proposed that the difference is accounted for by hidden blood loss (HBL). 6Other sources of loss include those from the initial trauma, intraoperative extravasation into the soft tissues, or other blood loss incurred during hospitalization.In recent years, there has been growing interest in defining HBL, especially in the aged population.If left untreated, HBL may aggravate pre-op comorbidities and lead to anemia. 5Most studies have focused on quantifying HBL 7 or determining its associated risk factors, 8 but to date none have studied time from trauma to surgery as a variable.
A systematic review and meta-analysis by Klestil et al., 9 noted that hip fractures treated within 48 hours had a lower risk for mortality and perioperative complications.This is, however, not observed in our local setting.Delays to surgery are common due to socio-economic factors and scant resources (including blood products).Whatever the reason for the delay, we still push for surgery to improve hip function and reduce complications when compared to nonoperative treatment. 10omputing for the HBL may help in estimating the amount of blood products needed to manage postoperative anemia. 11his study aims to determine if HBL in elderly patients with pertrochanteric fractures treated by PFN is affected by delay in time to surgery.

METHODOLOGY
This was a detailed cross-sectional study.Exemption from ethics review was granted by the institution's ERB before starting data gathering (POCERB 2020-01-002).The data analyzed was limited to non-identifiable figures, waiving the need for informed consent.The STROBE cross-sectional reporting guidelines 12 were used as a checklist in making this study.
A review of the surgical census and medical records was done to collect data about patients with pertrochanteric fractures treated by closed reduction and short proximal femoral antirotation nailing (PFNa).Patients aged 60 years old and above who were seen at the ER, admitted, and operated on from January 2017 to December 2019 were included.Excluded from the study were patients with 1) fractures of other extremities surgically treated during the same operative time, 2) pathologic fractures, 3) patients with hematologic or GI comorbidities, and 4) patients also taking blood thinners or were given Tranexamic acid.
The sample size was calculated based on the estimation of population proportion for pertrochanteric femur fractures.Assuming that the incidence of hip fractures in the elderly population is 18%, 13 and using a level of significance equal to 5%, the sample size required was 196 patients.
The following data were obtained from the medical and radiographic records of patients: age, sex, height, weight, AO classification, time from initial trauma to surgery, total operative time, intraoperative visible blood loss from the operative record, and complete blood counts done on admission, within 7 days before surgery and up to 72 hours post-surgery.Patients were classified into either early (time from injury to surgery <30 days) or late (≥30 days) surgery group.The patient's total blood volume (PBV) was computed using the formula as reported by Nadler et al., 14 as follows: PBV, in liters = K1×h 3 + K2×w + K3 where h = height in meters, and w = weight in kg, Men: K1 = 0.3669, K2 = 0.03219, K3 = 0.6041, Women: K1 = 0.3561, K2 = 0.03308, K3 = 0.1833 After computing the total blood volume, we computed the estimated blood loss, and the peri-operative hidden blood loss, using the formula by Gross, 15 as follows: Estimated blood loss (mL) = PBV × [(Hct adm -Hct pod1 )/ Hct ave ] × 1000 Total perioperative blood loss (mL) = estimated blood loss + blood transfused Perioperative HBL (mL) = total perioperative blood loss -intraoperative visible blood loss Where Hct adm was the hematocrit count done during admission, Hct pod1 was the hematocrit count on the first postoperative day, and Hct ave was the average of all hematocrit counts done from admission to discharge.
The data was encoded in Microsoft Excel and analysis was done using SPSS 20 software.Data was described as mean ± standard deviation.Comparison of the demographic and clinical characteristics was performed using the independent sample t-test.A confidence interval of 95% was used to determine statistical significance.

RESULTS
A total of 210 patients were included in the study.One hundred and thirty-five patients were categorized in the early surgery group, while 75 patients were categorized in the late surgery group.
The mean age for the early surgery group was 77 years old and 76 years old for the late surgery group (Table 1).Most patients were women.Most of the patients were classified under AO 31-A1.2(2-part pertrochanteric fracture).The operative duration averaged 73.17 minutes in the early surgery group and 72.64 minutes in the late surgery group.The mean visual blood loss for the early surgery and late surgery groups were 247.37 mL and 245.33 mL, respectively.The averages of hematocrit levels taken on admission, pre-operatively within 7 days, and immediately postoperatively are listed in the table.There were no significant differences between the two groups in terms of age, sex, AO classification, operative duration, HCT levels, intraoperative visual blood loss, postoperative blood transfused, and total blood loss.
Our results showed that there was no significant difference noted between the two groups in terms of peri-op HBL (p = 0.22) (Table 2).The mean peri-operative HBL of patients with delayed time to surgery was 95.32 ml compared to 113.65 ml of patients with early time to surgery.

DISCUSSION
Increased perioperative blood loss may lead to many complications and a poor prognosis, especially in the elderly population.Therefore, it is important to identify the causes of perioperative blood loss.Hidden blood loss accounts for a high percentage of the total perioperative blood loss in patients with pertrochanteric fractures. 5If HBL is disregarded, it may lead to anemia or low blood volume.Patient-blood management programs resulted in positive outcomes for patients under orthogeriatric care.It showed a reduction in complication and mortality rates, as well as improvement in functional recovery in the transfused patient group. 16Foss 17 also reported that uncorrected anemia is associated with decreased mobility and an increased 30-day mortality and length of hospital stay in surgically treated hip fracture patients.Higher hemoglobin levels post-op were also shown to be related to a shorter hospital stay and lower odds of readmission. 18This fueled the increased interest in quantifying hidden blood loss and studying its risk factors.
Hidden blood loss was computed to be nearly 75% of total blood loss according to Yu et al., 19 or 86.8% to 89.4% according to Li et al., 20 using the same formula.In this paper, we computed the mean HBL of 210 patients to be 107.10mL comprising 30.27% of the mean total blood loss.While published literature reports figures twice as large as ours, there is no denying that HBL makes up a sizeable amount of total blood loss.One of the limitations of this study is the inconsistency in measuring visible blood loss, given that it is a retrospective study.The estimation of visible blood loss may vary from person to person, thus one of the recommendations for a future prospective paper is the creation of a protocol for standardized measurement.
Most available literature on hidden blood loss in surgically treated pertrochanteric fractures excluded patients whose surgeries are delayed.][8]11 This delay is brought about by many factors, including patient socio-economic factors and limited hospital resources.
Our results cannot therefore be compared side-by-side to available studies.Yu et al. 16 noted an HBL of 272.2 mL (± 7.6 mL) in their subset of patients who underwent fixation using a PFNa.A similar study by Yang et al., 21 computed HBL of 787.7 mL (± 250.9 mL) in the group fixed with a PFNa.Tian et al., 11 computed HBL to be around 326 to 596 mL, depending on the need for postoperative blood transfusion.Our computed HBLs were 113.65 mL (± 99.25 mL) for the early surgery and 95.32 mL (± 111.79 mL) for the late surgery groups, with no significant difference between the two groups.In theory, the absence of fracture hematoma, resolution of third-spacing post-injury, and optimized medical status may account for the lower computed HBL.However, we do not have supporting literature for this, owing to the lack of studies on delayed treatment.
This study has its limitations, such as the study design, the lack of an electronic medical database, and the limited resource setting of the institution.The authors suggest pursuing a prospective study to control confounding factors, such as measurement of visible blood loss, and uniform date of blood extraction for CBC results up to the third day postoperatively.Future studies may reveal a relationship between time to surgery and the amount of blood loss.

CONCLUSION
In the setting of delayed treatment of elderly patients with pertrochanteric femur fractures fixed with short proximal femoral antirotation nailing, the amount of hidden blood loss is not significantly affected by the timing of surgery.However, this hidden blood loss still makes up almost a third of the total blood loss and should be anticipated in the management of postoperative anemia.

Table 1 .
Demographic and clinical characteristics of patients by time from injury to surgery (n=210)