|Year : 2019 | Volume
| Issue : 1 | Page : 25-30
Comparison of visual estimation of intra-operative blood loss with haemoglobin estimation in patients undergoing caesarean section
Sampson Uzo Anya, Fidelis Anayo Onyekwulu, Elias Chukwujioke Onuora
Department of Anaesthesia, University of Nigeria Teaching Hospital, Enugu, Nigeria
|Date of Web Publication||12-Mar-2019|
Fidelis Anayo Onyekwulu
Department of Anaesthesia, University of Nigeria Teaching Hospital, Ituku Ozalla, Enugu
Source of Support: None, Conflict of Interest: None
Background: Obstetrical haemorrhage is a potentially preventable cause of maternal morbidity and mortality; and measurements of surgical blood loss (BL) are often inaccurate. Accurate BL estimation is paramount as it may substantially alter the timing of interventions to control haemorrhage. The study compares the assessment of intra-operative BL by visual estimation with BL calculated from haemoglobin estimation using the HemoCue®201+. Materials and Methods: A total of 60 pregnant patients at term undergoing elective caesarean section under spinal anaesthesia were enrolled into the study. In the theatre, the patients' haemoglobin level was determined before and after the surgery using the HemoCue®201+; and a modified Gross formula was used to calculate the BL. BL was also visually estimated and documented by counting the blood-soaked abdominal mops and gauze pieces and multiplying them by the estimated volume of blood each would hold; fixed size mops and gauzes were used. Statistical analysis was performed to compare both methods using SPSS Version 17. To compare BL assessment, Pearson's correlation and the Bland and Altman's method of assessing agreement between two methods of clinical measurement were used. Results: The mean visually estimated BL (EBL) and HemoCue calculated BL (CBL) were 470 ± 221 ml and 563 ± 204 ml, respectively (P = 0.125). The bias (mean difference between both methods) was negligible (45.25 ml), and the limit of agreement between both methods was −222.20–275.43 ml. The discrepancy between the two methods increased when BL was ≥500 ml. Conclusion: This study showed that visually EBL was closely related to HemoCue CBL when the quantity of BL was <500 ml.
Keywords: HemoCue®201+, intraoperative blood loss, visual estimation
|How to cite this article:|
Anya SU, Onyekwulu FA, Onuora EC. Comparison of visual estimation of intra-operative blood loss with haemoglobin estimation in patients undergoing caesarean section. Niger Postgrad Med J 2019;26:25-30
|How to cite this URL:|
Anya SU, Onyekwulu FA, Onuora EC. Comparison of visual estimation of intra-operative blood loss with haemoglobin estimation in patients undergoing caesarean section. Niger Postgrad Med J [serial online] 2019 [cited 2020 May 27];26:25-30. Available from: http://www.npmj.org/text.asp?2019/26/1/25/253974
| Introduction|| |
Over the years, different methods have been used for the estimation of intraoperative blood loss (BL), which is an important aspect of peri-operative management of the surgical patient. Visual estimation has been the most commonly used method and sometimes, the only method available for assessing intraoperative BL simply because it is easy, quick, and convenient. Accurate assessment of intraoperative BL is an important aspect of peri-operative management of patients undergoing caesarean section where BL is often dispersed and mixed with the amniotic fluid. Anaesthetists in developing countries may not have the luxury of point-of-care monitoring devices and may have to rely on visually estimated BL (VEBL) in making this decision.
Accurate BL estimation during caesarean section is crucial as it may substantially alter the timing of blood transfusion and interventions to control haemorrhage since morbidity and mortality can result from both delayed recognition and denial of the occurrence of significant bleeding. Several studies carried out within Sub-Saharan Africa using visual estimations of BL for estimating intraoperative BL reveal a high transfusion rate of up to 25.2% among women undergoing caesarean section.,, Appreciable numbers of these patients were transfused unnecessarily. Inaccurate BL estimation can lead to either over or under transfusion.
Clinical methods of BL assessment used include counting the blood-soaked mops, gauze pieces (and multiplying them by the estimated volume of blood they carry), measuring BL to suction bottles and estimating that, which was in and around the operating field. Blood transfusion may be associated with adverse immunological transfusion reactions, transmission of infections, increased cost, increased perioperative morbidity and mortality, and delayed recovery from anaesthesia., Major haemorrhage, if left unattended to, continues to be one of the most common causes of direct maternal death in obstetric practice.
The HemoCue (AB Medical Inc., Angelholm, Sweden) haemoglobin estimation is a point-of-care testing which is relatively inexpensive and faster than the laboratory testing, with result readily available within 30–45 s; and when it is used with the modified Gross formula can be used in estimation of BL. Therefore, assessing the accuracy of visual estimation of BL compared to estimation using a point-of-care device like HemoCue will be very relevant in optimal obstetric care.
| Materials and Methods|| |
The study was carried out among pregnant women at term undergoing elective caesarean section at the University of Nigeria Teaching Hospital Enugu, Nigeria. Ethical clearance for the study was obtained from the Hospital's Health Research Ethics Committee (HREC), and informed written consent was also obtained from each patient recruited for the study. The research was approved on 30th May, 2013 and a protocol number NHREC/05/01/2008B was assigned to the study. Data collection spanned over 10 months from 5 June 2013 to 12 April 2014.
Sample size calculation was based on a previous study by Ashraf Aly and Ramadani where VEBL was compared to haematocrit calculated blood loss (HCBL) during caesarean section. In their study, VEBL and HCBL were 528 ± 143.2 ml and 627 ± 180.6 ml, respectively. The difference in blood loss between both methods is 99ml; therefore a 10% increase in the difference (DIFF) =108.9ml. The sample size was calculated using the formula: N = (A + B)2 × 2× SD2/(DIFF)2 with a significance level of 5% and power of the study of 80%. Sample size (2N) =2 × 27 = 54. Using 10% attrition of 5.4, sample size = 54 + 5.4 = 59.4 ~ 60.
A total of 60 pregnant women at term presenting for elective Caesarean section under subarachnoid block who were American Society of Anaesthesiologists physical status II were enrolled into this double-blind, prospective, nonrandomized, controlled comparative study, in which each patient acted as her own control. Excluded were patients with pre-operative anaemia or haemoglobin ≤8 g/L, the patient at the risk of massive intraoperative haemorrhage, for example, major placenta previa, and patients with comorbidities.
The researcher (anaesthetist) and the patients were not aware of the CBL using the HemoCue before the time of making a visual estimate of intraoperative BL. However, the researcher had the responsibility to alert the attending physician whenever the haemoglobin level falls ≤7 g/L using the HemoCue.
All patients were seen and reviewed evening before the surgery. The patient's age, weight, indication for caesarean section, parity, and gestational age were taken and recorded. The patients' preoperative laboratory (lab) haemoglobin was documented, and the preoperative haemoglobin (hi) obtained before preloading (intravenous administration of 15 ml/kg of 0.9% normal saline to patients before the induction of spinal anaesthesia) and after preloading (hp) using the HemoCue (HM) was also documented.
The laboratory haemoglobin measurement was done using 2 ml of venous blood taken from the non-cannulated arm and using an autoanalyser GEM premier 3000 (IL Werfen Inc., Benford USA) in the side laboratory. The laboratory samples were taken before the HemoCue samples.
The blood samples for HemoCue estimation were taken from a skin prick on the thumb of the non-cannulated arm. All capillary samples were taken by the researcher. The capillary samples were taken strictly according to the manufacturer's recommendation, which includes choosing a suitable digit for sampling, wiping it with a disinfectant chlorhexidine 1% solution, making a skin puncture that will allow a ladybird size drop of blood to be expelled, drawing up the blood sample with the cuvette ensuring that the entire chamber is filled (not only the circular section), and wiping off the excess on absorbent material taking care not to siphon out the content. This sample was then placed into the HemoCue® 201+ (AB Medical Inc., Angelholm, Sweden) device. To eliminate bias, an assistant took the readings (results) which were displayed on the screen in less than a minute.
Spinal anaesthesia was induced with 2.5 ml of 0.5% hyperbaric bupivacaine at the L3/L4 lumbar interspace using the iliac crest as a landmark. The surgery was allowed to start after a block height of T4–T6 was achieved. After skin closure, the researcher visually estimated the BL and documented it as vf, and this was followed by haemoglobin estimation by the researcher using HemoCue which was also documented as hf. The researcher estimated the BL visually by counting the blood-soaked abdominal mops and gauze pieces and multiplying them by the estimated volume of blood each would hold; fixed size mops and gauzes were used. A fully soaked and dripping abdominal swab (10 inch × 10 inch) was taken as containing 100 ml of blood, while a piece of gauze (4 inch × 4 inch) was assumed to contain 10 ml of blood. Blood lost to suction bottles and that lost in and around the operating field mixed with amniotic fluid were also estimated. Abdominal mops and gauze pieces not fully soaked were also estimated using the experience and expertise of the researcher. The VEBL was designated as estimated BL (EBL), while the BL calculated from the HemoCue haemoglobin was designated as actual BL (ABL). The ABL was calculated from a modification of the Gross formula given below.
ABL = BV (Hct(i)- Hct(f))/Hct(m). Where ABL = Actual BL; BV = blood volume calculated from the patient weight in Kg (65 ml/kg). Hct(i)= haematocrit initial; Hct(f)= haematocrit final; Hct(m)= the mean of the initial and final haematocrits. Haematocrit (Hct) derived from HemoCue = 2.953 × Hb g/dl., The constant 2.953 is the relation between Hct and Hb as expressed in the equation.
CBL using HemoCue = BV (2.953 × hi¯2.953 × hf)
2.953 × hm
The average BL (AVG-BL) derived from each patient which was the mean of the actual and EBL.
Statistical Package for Social Sciences version 17.0 (SPSS Inc., Chicago, IL, USA) was used for data entry and statistical analysis. The mean, standard deviation (SD), and range were used to analyse basic demographics. Results were presented in frequency, percent, and charts. To compare BL assessment using the visual estimation and the HemoCue test, paired t-test, Pearson's correlation, and the Bland and Altman's method of assessing agreement between two methods of clinical measurement were used. The relationship between laboratory haemoglobin concentration and that of HemoCue haemoglobin concentration was determined using Pearson's correlation.
| Results|| |
A total of 60 pregnant women for elective caesarean section were recruited for this study. The ages of the women were between 21 and 37 years with a mean age of 27.6 ± 3.1 years. Their body weights ranged from 62 to 95 kg with a mean weight of 80.2 ± 5.7 kg, and their gestational age was 37–41 weeks and a mean gestational age of 36.4 ± 1.4 weeks. Majority of the patients were nulliparous 45 (75%), whereas 15 (25%) were multiparous. The indications for caesarean section are as shown in [Figure 1].
When the laboratory haemoglobin results were compared to HemoCue haemoglobin results. The result suggests that the HemoCue result has a strong positive correlation with a correlation coefficient of 0.89 (P = 0.0001) when compared with the autoanalyser results [Figure 2]. Further comparison of the values obtained from the two methods of haemoglobin estimation was carried out using a paired t-test and no significant difference was observed (P = 0.07) [Table 1]. The patients' mean haemoglobin using HemoCue before preloading with normal saline was 11.3 ± 0.7 g/dl (range of 10.0–12.7 g/dl), and after preloading was 11.2 ± 1.1 g/dl (range of 9.8–12.3 g/dl). There was a slight drop in the mean haemoglobin level but the drop was not statistically significant (P = 0.14; 95% confidence interval [CI] 0.9060–1.1514).
|Figure 2: A simple scatter plot of laboratory haemoglobin and HemoCue haemoglobin. Lab Hb: Laboratory haemoglobin g/dl, HM Hb: HemoCue haemoglobin g/dl|
Click here to view
|Table 1: Comparison of Laboratory haemoglobin versus HemoCue haemoglobin|
Click here to view
The mean visual EBL for caesarean section was 470.32 ± 221.66 ml with a range of 200–1100 ml, while the mean CBL using HemoCue was 563.47 ± 203.53 ml with a range of 180.7–1074.5 ml. Visual estimation was less than HemoCue CBL (underestimation). The difference between VEBL and CBL was not statistically significant with a P = 0.125; 95% CI −13.65 to −102.34. A Pearson product-moment correlation coefficient was computed to assess the relationship between EBL and CBL using HemoCue. There was a strong positive statistically significant correlation between both methods (r = 0.66; P = 0.002) [Figure 3].
|Figure 3: A simple scatter plot of estimated blood loss and calculated blood loss. EBL@skin clos – Visually estimated blood loss (ml). calculated blood loss – HemoCue calculated blood loss (ml)|
Click here to view
Further analysis was done using Bland and Altman plot. EBL and CBL were used to determine the agreement in difference in BL (DIFF-BL). The AVG-BL (derived from the mean of both methods of estimation) was 456.22 ± 177.46 ml [Table 2]. The bias (mean difference between both methods) was negligible (+45.25 ml) and the limit of agreement (mean difference ± 2 SD) between both methods was −222.20–275.43 ml. The bias and limit of agreement between both methods of assessing BL was small and not significant enough to cause error in clinical judgement in this group of patients.
|Table 2: Measurement of agreement between difference in blood loss and average blood loss|
Click here to view
To obtain the Bland-Altman plot, the difference in BL was plotted on the Y-axis with the AVG-BL on the X-axis [Figure 4]. From the plot it was observed that at AVG-BL between 200–500 ml, the values were around the “0” point on the Y-axis, which connotes good level of agreement between both methods of determining intraoperative BL but this deviates further away around the “0” point at AVG-BL of >500 ml which implies that as BL increases above 500 ml the error margin between VEBL and HemoCue CBL widens.
|Figure 4: Bland and Altman plot for comparison of difference in blood loss and average blood loss (comparison of visually estimated blood with HemoCue calculated blood loss). Diff in EBL/CBL: Difference between visually estimated blood loss and the HemoCue calculated blood loss (ml), Average blood loss: mean of both visually estimated blood loss and the HemoCue calculated blood loss (ml). The middle horizontal line represents the bias (mean difference)|
Click here to view
| Discussion|| |
This study revealed that visual estimation of intraoperative BL by the anaesthetist did not significantly differ from HemoCue CBL. The AVG-BL using visual estimation in this study was 470 ± 221 ml with a range of 200–1100 ml. This finding agrees with the results of previous studies., Duthie et al. in their study, observed a mean EBL of 425 ml with a range of 100–1300 ml during caesarean section using visual estimation. In the same way, Fauzia et al. in their work observed that the AVG-BL estimated by anaesthetist was 498 ± 176 ml, while VEBL by obstetrician was 592 ± 222 ml. In their study, the EBL by the anaesthetists was surprisingly lower than those of the obstetricians in contrast to the general belief that anaesthetists often tend to “overestimate” BL.
There is a paucity of report on the use of HemoCue in determining intraoperative BL during caesarean section globally. In this study, the HemoCue CBL was 563 ± 207 ml with a wide range of 181–1075 ml. The result is in tandem with previous work done using laboratory haematocrit values, where the measured BL was within the range of 66–1290 ml during caesarean section with a mean of 627 ml.
For this study, there was no significant difference between the mean visual EBL and the mean HemoCue CBL (P = 0.125). The relationship between both methods of estimating BL was assessed using Pearson's correlation, and the result showed a high correlation between both methods (r = 0.66). This finding is in contrast with Eipe and Ponniah who compared visually EBL with laboratory CBL and found poor correlation between both methods with interclass correlation coefficient of 0.34. Their study was done on general surgery patients, neurosurgical and orthopaedic patients. In addition, more than one anaesthetist was responsible for estimation of BL and this could account for the disparity in their result compared to the index study.
However, the interpretation of test of significance and correlation coefficient when comparing two methods could be misleading as data, which seem to be poor in agreement can produce quite high correlations. Therefore, further analysis to measure the degree of agreement between both methods of clinical measurement recommended by Bland and Altman was also employed. The result showed good agreement between the two methods of assessing BL with a bias of +45.25 ml. The level of agreement was better at BL of between 200 and 500 ml. At BL >500 ml the difference between both methods was between 500 and 750 ml, which was too wide and represent an unacceptable margin if used solely for clinical judgement. These findings corroborate previous studies where BL was visually estimated with reasonable accuracy among anaesthetists and the error in clinically EBL was typically higher if the measured BL was ≥600 ml., Furthermore, previous studies have found greater error in the estimation of larger blood losses by health professionals which was also in keeping with this study.
In this study, the parturients' laboratory haemoglobin and HemoCue haemoglobin were compared and a very high correlation was observed (r = 0.89). The observation was in keeping with several other studies which suggested that HemoCue is an accurate device in the assessment of haemoglobin in obstetric patients.,, Sanchis-Gomar et al. in their study compared the HemoCue Hb 201 system with the reference method according to the International Council For Standardization in Haematology and found a correlation of 0.99. This result reflects a near perfect relationship between both methods, and further validates the accuracy of HemoCue in assessing haemoglobin level in obstetric patients. In addition, no statistically significant change in womens' haemoglobin level before and after preloading with normal saline was noted (P = 0.14). The result aligns with the previous study where it was observed that the associated changes in blood volume and packed cell volume are very transient and insignificant when volumes of 10–20 ml/kg are administered as a single loading dose as seen during caesarean section.
Caesarean section is known to be associated with varying degrees of BL; however, visual estimation in this study was found to be reasonably accurate with marginal error, and a greater imprecision was found with higher losses of >500 ml. In this study, mathematical calculation using modified Gross formula was adopted to determine measured BL with HemoCue. This is considered as a limitation in this study as reports suggesting that all theoretical relationship between BL and changes in haematocrit often leads to “overestimation” of BL.
| Conclusion|| |
In conclusion, VEBL was closely related to HemoCue CBL when the quantity of BL was <500 ml; as BL increases above 500 ml, the error margin between the two methods widens. Although VEBL was less than HemoCue CBL, the margin was not significant and could not have led to error in clinical judgement such as over transfusion or unrecognised massive maternal haemorrhage. We, therefore, recommend that HemoCue should be used during caesarean section to complement visual estimation of intra-operative BL, especially in obstetric cases where large BL is anticipated.
Financial support and sponsorship
Conflicts of interest
There are no conflicts of interest.
| References|| |
Vimala N, Mittai S, Kumar S. Sublingual misoprostol versus oxytocin infusion to reduce blood loss at caesarean section. Int J Obstet Gynecol 2005;90:322-9.
Desalu I, Dada OI, Ahmed RA, Akin-Williams OO, Ogun HA, Kushimo OT, et al.
Transfusion trigger – How precise are we? Intraoperative blood transfusion practices in a tertiary centre in Nigeria. Transfus Med 2008;18:211-5.
Doctorvaladan SV, Jelks AT, Hsieh EW, Thurer RL, Zakowski MI, Lagrew DC, et al.
Accuracy of blood loss measurement during cesarean delivery. AJP Rep 2017;7:e93-100.
Faponle AF, Makinde ON. Caesarean section: Intra-operative blood loss and its restitution. East Afr Med J 2007;84:31-4.
Anorlu RI, Orakwe CO, Abudu OO, Akanmu AS. Uses and misuse of blood transfusion in obstetrics in Lagos, Nigeria. West Afr J Med 2003;22:124-7.
Ozumba BC, Ezegwui HU. Blood transfusion and caesarean section in a developing country. J Obstet Gynaecol 2006;26:746-8.
Eipe N, Ponniah M. Perioperative blood loss assessment- How accurate? Indian J Anaesth 2006;50:35-8. [Full text]
Practice guidelines for blood component therapy: A report by the American Society of Anesthesiologists task force on blood component therapy. Anesthesiology 1996;84:732-47.
Imarengaiye CO, Enosoleaseb ME, Iribhogbec PE. Intraoperative blood transfusion among adult surgical patients in a tertiary hospital in Benin City, Nigeria. J Med Biomed Res 2002;1:17-22.
Esler MD, Douglas MJ. Planning for hemorrhage. Steps an anesthesiologist can take to limit and treat hemorrhage in the obstetric patient. Anesthesiol Clin North America 2003;21:127-44, vii.
Gross JB. Estimating allowable blood loss: Corrected for dilution. Anesthesiology 1983;58:277-80.
Ashraf Aly H, Ramadani HM. Assessment of blood loss during caesarean section under general anaesthesia and epidural analgesia using different methods. Alexandria J Anaesth Intensive Care 2006;9:25-34.
Miller RD, editor. Transfusion therapy. In: Anaesthesia. 5th
ed. Philadelphia: Churchill-Livingstone; 2000. p. 1613-7.
Nijboer JM, van der Horst IC, Hendriks HG, ten Duis HJ, Nijsten MW. Myth or reality, haematocrit and haemoglobin differ in trauma. J Trauma 2007;62:1310-2.
Mark EE, Douglas DR, Lawrence DE. Correlation of transfusion volume to change in haematocrit. Am J Hematol 2006;81:145-6.
Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet 1986;1:307-10.
Duthie SJ, Ghosh A, Ng A, Ho PC. Intra-operative blood loss during elective lower segment caesarean section. Br J Obstet Gynaecol 1992;99:364-7.
Fauzia AK, Mueenullah K, Asif A, Ursula C. Estimation of blood loss during caesarean section. J Pak Med Assoc 2006;12:572-5.
Bose P, Regan F, Paterson-Brown S. Improving the accuracy of estimated blood loss at obstetric haemorrhage using clinical reconstructions. BJOG 2006;113:919-24.
Duthie SJ, Ven D, Yung GL, Guang DZ, Chan SY, Ma HK, et al.
Discrepancy between laboratory determination and visual estimation of blood loss during normal delivery. Eur J Obstet Gynecol Reprod Biol 1991;38:119-24.
Kavle JA, Khalfan SS, Stoltzfus RJ, Witter F, Tielsch JM, Caulfield LE, et al.
Measurement of blood loss at childbirth and postpartum. Int J Gynaecol Obstet 2006;95:24-8.
Glover P. Blood loss at delivery: How accurate is your estimation? Aust J Midwifery 2003;16:21-4.
Sanchis-Gomar F, Cortell-Ballester J, Pareja-Galeano H, Banfi G, Lippi G. Hemoglobin point-of-care testing: The HemoCue system. J Lab Autom 2013;18:198-205.
Richards NA, Boyce H, Yentis SM. Estimation of blood haemoglobin concentration using the HemoCue during caesarean section: The effect of sampling site. Int J Obstet Anesth 2010;19:67-70.
Skelton VA, Wijayasinghe N, Sharafudeen S, Sange A, Parry NS, Junghans C, et al.
Evaluation of point-of-care haemoglobin measuring devices: A comparison of radical-7™ pulse co-oximetry, hemoCue(®) and laboratory haemoglobin measurements in obstetric patients. Anaesthesia 2013;68:40-5.
Iijima T, Brandstrup B, Rodhe P, Andrijauskas A, Svensen CH. The maintenance and monitoring of perioperative blood volume. Perioper Med (Lond) 2013;2:9.
Fransen EJ, de Jong DS, Hermens WT, Maessen JG. Is postoperative blood loss, loss of blood? A pilot study in cardiac surgical patients. Perfusion 2001;16:301-8.
[Figure 1], [Figure 2], [Figure 3], [Figure 4]
[Table 1], [Table 2]