Impact of gender parity on preoperative anaemia prevalence and Patient Blood Management practice

Abstract

Background

Anaemia is a common finding in the preoperative setting, affecting around one-third of patients for whom major surgery is programmed. Moreover, preoperative anaemia has been shown to worsen patient outcome and increase length of hospital stay and costs. In the field of preoperative anaemia correction, a recent Consensus statement suggested reviewing the classic World Health Organization (WHO) criteria in adults by aligning the haemoglobin cut-off to 13 g/dL for both genders. The aim of our study was to assess the differences in terms of prevalence, transfusion rate, transfusion trigger, and blood losses according to gender in a mixed population of surgical patients.

Material and methods

We reviewed data of 610 consecutive patients undergoing elective major surgery at a tertiary care hospital during a 9-month period. Transfusion rate and transfusion triggers were recorded, analysed and stratified by haemoglobin class, with a particular focus on the 12.0–12.9 g/dL range.

Results

Since the anaemia threshold was redefined at 13 g/dL for both genders, its prevalence rose from 26.4 to 39.5% (161/610 vs 241/610; p<0.001) in the overall population and from 22.7 to 49.3% (68/300 vs 148/300; p<0.001) in women. Eighty women (26.7%) fell in the haemoglobin 12.0–12.9 g/dL range, and this category was the most represented among transfused women (34.0%). There was no statistical difference in transfusion triggers or overall transfusion rate between genders. Subjects of both genders were transfused at the same haemoglobin level (8.1 g/dL), but women reached the transfusion trigger after less red cell mass loss than men, i.e. 377 mL (249–472 mL) vs 528 mL (356–717 mL), respectively (p<0.001).

Discussion

Treatment of pre-surgical anaemia is one of the core principles of Patient Blood Management. Aligning the haemoglobin threshold between genders in the management of pre-surgical anaemia may result in a lower transfusion rate, but in an increased workload for medical staff in the preoperative phase.

INTRODUCTION

Anaemia is a frequent finding in the preoperative setting^1,2 and has been shown to have a detrimental effect on patient outcome^1,3–6, length of hospital stay⁷ and costs^8,9. Moreover, optimisation of preoperative haemoglobin (Hb) is one of the main principles of Patient Blood Management (PBM)^10,11.

According to the World Health Organization (WHO) definition¹², adult males and adult non-pregnant women are considered anaemic when their haemoglobin level is below 13.0 g/dL and 12.0 g/dL, respectively. In recent years, the application of different parameters according to gender to correct preoperative anaemia has been questioned^13–15. The lower Hb threshold could preclude a large proportion of women undergoing surgery from receiving an effective treatment¹⁶. On the other hand, the introduction of a new gender parity criterion for Hb levels may raise additional issues in setting PBM policy.

The aim of our study was to assess the differences in terms of prevalence, transfusion rate, transfusion trigger and blood losses according to gender in a mixed population of surgical patients, focusing on the Hb class cohort that the WHO classifies as not anaemic.

MATERIALS AND METHODS

Patient population and data collection

Data of 610 patients scheduled for major surgery and consecutively admitted to the Preoperative Assessement Unit (POAU) during a 9-month period at Legnano General Hospital were collected. Patients programmed for cardiac surgery did not follow the POAU path and their data were recorded at the time of hospital admission. A specific local non-shared database was created and updated daily.

The following pre-surgery data were recorded for each patient: age, gender, haemoglobin level, haematocrit, weight, height, body mass index (BMI), red blood cell mass (RBCM), and type of surgery. The amount of red cell loss was also individually estimated.

Haemoglobin level at transfusion and the number of pre-storage leukodepleted packed red blood cell units (PRBC) transfused from surgery to discharge were also recorded.

Prevalence of anaemia in the two gender groups was compared by considering the traditional WHO classification for anaemia (i.e. female Hb<12.0 g/dL; male<13.0 g/dL) and the newly suggested value (i.e. both males and females considered anaemic with Hb<13.0 g/dL), referred to hereon as PBM-Hb.

Transfusion rates according to type of surgery and stratified by Hb classes were also calculated. We focused in particular on the transfusion trigger and transfusion rate in women transfused in the Hb range 12.0–12.9 g/dL, i.e. the Hb class gap not taken into account by the WHO classification.

Calculation of red cell mass and surgical blood loss

Nadler’s formula¹⁷ was used to calculate total blood volume. After correction of haematocrit of peripheral blood sampling by applying a factor of 0.91 in accordance with Chaplin et al.¹⁸, we assessed the patient median and interquartile range (IQR) of RBCM by multiplying the total blood volume with the corresponding baseline haematocrit level. The estimated RBCM loss until transfusion was calculated by multiplying the total blood volume with the haematocrit at transfusion, then subtracting the value from the baseline RBCM.

Statistical analysis

Data are presented as percentage or median ± IQR. In fact, continuous variables taken into consideration were all non-normally distributed (Shapiro Wilk test for normality). Pearson’s χ2 test with Yates correction or Fisher exact test were performed when indicated to assess the differences among categorical variables. Mann-Whitney U test was used in case of non-normally distributed continuous variables. Data were analysed using IBM SPSS software (version 20, IBM Corp., Armonk, NY, USA). p<0.05 was considered statistically significant.

RESULTS

Baseline data and type of surgery are listed in Tables I and II. The PBM-Hb classification led to a general increase in anaemia prevalence compared to the WHO classification, from 26.4 to 39.5% (161/610 vs 241/610, +49.7%; p<0.001) in men and from 22.7 to 49.3% (68/300 vs 148/300, +117%; p<0.001) in women (Figure 1). In the gynaecological surgery group, the new gender parity classification would have doubled the number of patients diagnosed as anaemic.

Table I.

Patients’ baseline data

	Total n. 610	Males n. 310	Females n. 300
Age - years (range)	66 (52–74)	70 (63–76)	66 (52–74)
Weight - kg (range)	73 (63–83)	78 (70–86)	66 (58–77)
Height - cm (range)	168 (160–173)	172 (169–178)	160 (158–166)
BMI - kg/m 2 (range)	25.8 (23.1–29.0)	26.1 (24.0–28.7)	25.2 (22.2–29.4)
RBCM - mL (range)	1,617 (1,364–1,922)	1,901 (1,659–2,118)	1,399 (1,229–1,568)

p<0.001 for all comparisons between continuous variables except for body mass index (BMI) (p=0.030). n: number; RBCM: red blood cell mass calculated according to Nadler et al.¹⁷ and Chandler et al.¹⁸. All data are shown as interquartile range.

Table II.

Type of surgery

Total n. of surgical interventions	610
Cardiac surgery, total n.	166
Annuloplasty	48
Aortic valve substitution	11
Mitral valve substitution	103
Tricuspid valve substitution	4
Orthopaedics, total n.	129
Hip replacement	46
Knee replacement	39
Spinal stabilisation	44
General surgery*, total n.	125
Digestive	74
Liver	6
Lung	34
Others	11
Gynaecology, total n.	86
Hysterectomy	75
Pelvic mass	5
Other	6
Urology, total n.	48
Nephrectomy	23
Cystectomy	14
Radical prostatectomy	8
Others	3
Vascular surgery, total n.	42
Aneurysm surgery	40
Other	2
Otolaryngology, total n.	14
Laryngectomy	2
Basicranial	2
Others	10

^*Oncologic surgery (data are shown as system/organ involved). n.: number

Figure 1. Prevalence of preoperative anaemia.

WHO: World Health Organization anaemia criteria (anaemia thresholds: 13 g/dL for males; 12 g/dL for women).

PBM: Patient Blood Management guidelines; Hb: haemoglobin; PBM-Hb: anaemia classification according to PBM (anaemia threshold: Hb<13.0 g/dL for both sexes). Numbers at top of each column represent absolute numbers of patients.

As expected, men had a preoperative Hb value higher than women: 14.1 g/dL (12.5–14.9 g/dL) vs 13.0 g/dL (12.0–13.8 g/dL), respectively (p<0.001). According to the PBM-Hb classification, at the preoperative assessment, 30.0% of males and 49.3% of females were anaemic (p<0.001). After stratification by Hb class, anaemia prevalence differed significantly by gender in the range 12.0–12.9 g/dL (men 11.9% vs women 26.7%; p<0.001) (Table III).

Table III.

Preoperative data

Variable	Male	Female	P
n. of patients	310	300	--
PreOp haemoglobin - g/dL (range)	14.1 (12.5–14.9)	13.0 (12.0–13.8)	< 0.001
n. of patients with haemoglobin <13.0 - g/dL (%)	93 (30.0)	148 (49.3)	< 0.001
Haemoglobin class - g/dL	n. (%)	n. (%)	p
< 9.0	5 (1.6)	9 (3.0)	0.298
9.0–9.9	9 (2.9)	10 (3.3)	0.819
10.0–10.9	18 (5.8)	24 (8.0)	0.396
11.0–11.9	24 (7.7)	25 (8.3)	0.788
12.0–12.9	37 (11.9)	80 (26.7)	<0.001
13.0–13.9	50 (16.1)	88 (29.3)	<0.001
14.0–14.9	91 (29.4)	49 (16.3)	<0.001
≥15.0	76 (24.5)	15 (5.0)	<0.001

PreOp: preoperative; n.: number.

Overall, 104 out of 610 (17.0%) patients were transfused (Table IV). Females received transfusions at the same rate as males (17.1 vs 17.0%, respectively; p=1.000). In accordance with the PBM-Hb classification, among transfused patients, preoperative anaemia affected 67.9% of males and 82.3% of females (p=0.141). Transfused patients had the same median preoperative Hb, irrespective of gender (12.0 g/dL for both sexes).

Table IV.

Preoperative (PreOp) and transfusion data in transfused patients

Variable	Male	Female	P
n. patients transfused/all patients (%)	53/310 (17.1)	51/300 (17.0)	1.000
PreOp haemoglobin - g/dL (range)	12.0 (10.5–13.7)	12.0 (10.7–12.8)	0.450
n. of patients with haemoglobin <13.0 - g/dL (%)	36 (67.9)	42 (82.3)	0.141
Haemoglobin class (g/dL)	n. (%)	n. (%)	p
< 9.0	3 (5.7)	2 (5.6)	1.000
9.0–9.9	6 (11.3)	5 (11.3)	1.000
10.0–10.9	9 (17.0)	8 (15.1)	1.000
11.0–11.9	8 (15.1)	8 (15.1)	1.000
12.0–12.9	6 (11.3)	18 (34.0)	<0.008
13.0–13.9	9 (17.0)	7 (13.2)	0.851
14.0–14.9	7 (13.2)	2 (3.8)	0.161
≥15.0	5 (9.4)	1 (1.9)	0.205
RBCM losses until transfusion - mL (range)	528 (356–717)	377 (249–472)	<0.001
Haemoglobin at transfusion - g/dL (range)	8.1 (7.5–8.5)	8.1 (7.5–8.7)	0.628
n. PRBC units transfused (range)	2 (2–3)	2 (1–3)	0.497

n.: number; RBCM: red blood cell mass; PRBC: packed red blood cell; PreOp: preoperative.

A third of transfused women fell into the Hb class that was not considered anaemic according to the WHO classification. In the same Hb class (12.0–12.9 g/dL), females were transfused significantly more than males (34.0 vs 11.5%, respectively; p=0.008).

Women reached the Hb trigger point after an estimated RBCM of 28.6% less than men (Table IV): 377 mL (249–472 mL) vs 528 mL (356–717 mL) (p<0.001). This reflects the 26.4% difference in RBCM between genders at baseline (Table I).

Haemoglobin levels at transfusion were the same for both genders: males 8.1 g/dL (7.5–8.5 g/dL), females 8.1 g/dL (7.5–8.7). The median number of transfused PRBC units was also the same: males 2 (2–3) and females 2 (1–3).

The transfusion rate was the same for both genders, irrespective of type of surgery (Table V).

Table V.

Transfusion data by gender and type of surgery

	Total	Male	Female
All surgical interventions	104/610 (17.0)	53/310 (17.1)	51/300 (17.0)
Cardiac surgery	46/166 (27.7)	21/91 (23.1)	25/75 (33.3)
Orthopaedics	15/129 (11.6)	4/62 (6.5)	11/67 (16.4)
General surgery	25/125 (20.0)	18/76 (23.7)	7/49 (14.3)
Gynaecology	1/86 (1.2)	==	1/86 (1.2)
Urology	10/48 (20.8)	7/36 (19.4)	3/12 (25.0)
Vascular surgery	4/42 (9.5)	2/37 (5.4)	2/5 (40.0)
Otolaryngology	3/14 (21.4)	1/8 (12.5)	2/6 (33.3)

Data are shown as number patients transfused/all patients (% transfused). p=not significant for all comparisons between genders.

DISCUSSION

In this study, we evaluated the changes in anaemia prevalence in a mixed surgical population using a new approach which did not take into consideration the gender disparity promoted by expert opinion^13,15,19 and advocated by a recent international Consensus Statement¹⁴. We further focused on evaluating transfused patients by gender, and analysing transfusion rates, RBCM losses and transfusion triggers.

Our series confirms the differences in Hb levels between males and females. Several factors are involved, i.e. the effect of sex hormones on erythropoiesis²⁰, menstruation and lower baseline ferritin values.

As well pointed out by Muñoz et al.¹⁵, the misconception that the gender-related haemoglobin levels, as stated in the WHO classification, also fit women in apresurgical setting denies access to anaemia correction to a large proportion of anaemic patients scheduled for elective surgery. In general, women have a lower BMI and can not adequately respond to the same amount of blood loss as men, making them at greater risk of requiring transfusion¹⁴. In our series, the PBM-Hb classification leads to a more than a 2-fold increase in the number of women considered to be anaemic.

Our observations are in agreement with previous studies evaluating the prevalence of anaemia in women. In a mixed surgical population of 3,342 patients², 36.2% were anaemic according to the PBM-Hb classification versus 39.5% in our cohort. Interestingly, in the same study, the Authors found a slightly similar prevalence of absolute iron deficiency among women with Hb<13.0 g/dL than among those with Hb<12.0 g/dL, thus reinforcing the need to provide preoperative evaluation for those women who fall into the Hb range excluded from the WHO classification. Preoperative treatment of this well-represented patient category may improve haematopoietic response both during the hospital stay and after discharge²¹.

In a study specifically focused on mild anaemia in women undergoing cardiac surgery¹⁶, the Authors found a prevalence of 29.4% of patients in the Hb class 12.0–13.0 g/dL. A similar prevalence was observed in our series (26.7%). In the above-mentioned study, borderline anaemia was correlated with an increased length of hospital stay and a higher chance of transfusion in comparison with non-anaemic women.

Some studies evaluating gender-related transfusion rates have, however, reported conflicting results. In a study performed on 6,530 surgical patients in Austria²², women were significantly more frequently transfused than men and the same results were reported by Stammers et al. in a very large series of 54,122 patients undergoing coronary artery bypass graft. These gender-related differences were, however, not confirmed by others^23,24.

In our study, men and women were transfused at the same rate, but about one-third of transfused women fell into the Hb class not covered by the WHO criteria, i.e. 12.0–12.9 g/dL. There was no difference in transfusion trigger between genders, but this was reached after lower RBCM loss in women.

The baseline Hb physiological gap between men and women has not been taken into account in the context of transfusion, and this reinforces the need to end gender discrimination when correcting preoperative anaemia, although women seem to tolerate lower Hb levels better²⁵.

A moderate anaemia does not seem to affect the need for RBC transfusion in the mid-term, or re-hospitalisation or mortality²⁶. As pointed out by Shander and Goodnough²⁷, however, the global quality of life of discharged anaemic patients may be temporarily poorer, and this may be taken into account by the attending physicians.

Since 2015, under Italian law, PBM is mandatory²⁸. This obligation leads to medico-legal disputes in case of acute and long-term mortality or morbidity triggered by transfusion in an anaemic patient who had not undergone preoperative evaluation^29,30.

Despite the fact that the “Italian Regulatory Guidelines for the Implementation of Patient Blood Management - 2017” still indicate that anaemia according to WHO criteria should be corrected, in the face of the more recent indications such transfusion triggers appear outdated and an upgrade seems advisable³¹.

In our study, application of a gender equality approach to preoperative patient management would have led to a big increase in the workload of medical staff, implying the need for changes in the allocation of resources by hospital management to sustain the burden. Despite the fact that the absolute number of patients requiring preoperative anaemia treatment may be relatively low, screening the large population of patients who are candidates for surgery is a challenging and time-consuming activity. This is something that must be taken into account, especially given that, in Italy, the cost of implementing PBM is not sustained by specific national resources.

The main limit of our study is the relatively low number of patients. Furthermore, we cannot exclude the possibility that pre-transfusion Hb could be affected by intraoperative haemodilution applied in the context of an acute blood loss.

CONCLUSIONS

Equalising haemoglobin gender disparity can significantly improve the efficacy of programmes aimed at correcting pre-surgical anaemia, but this does create an additional workload. The upcoming guidelines should take into account the evidence accumulated in the more recent studies.