- Email: [email protected]
Drape estimation vs. visual assessment for estimating postpartum hemorrhage
International Journal of Gynecology and Obstetrics (2006) 93, 220 — 224
www.elsevier.com/locate/ijgo
CLINICAL ARTICLE
Drape estimation vs. visual assessment for estimating postpartum hemorrhage A. Patel a,*, S.S. Goudar b, S.E. Geller c, B.S. Kodkany b, S.A. Edlavitch d, K. Wagh b, S.S. Patted b, V.A. Naik b, N. Moss e, R.J. Derman d a
John H. Stroger Jr. Hospital of Cook County, Chicago, IL, USA J. N. Medical College, Belgaum, Karnataka, India c Department of Obstetrics and Gynecology, University of Illinois at Chicago, Chicago, IL, USA d University of Missouri at Kansas City School of Medicine, Kansas City, MO, USA e Global Network for Women’s and Children’s Health Research, National Institute of Child Health and Human Development, Rockville, MD, USA b
Received 14 September 2005; received in revised form 10 February 2006; accepted 14 February 2006
KEYWORDS Postpartum hemorrhage; Maternal morbidity; Maternal mortality; Blood collection
Abstract Objective: To compare (1) visual estimation of postpartum blood loss with estimation using a specifically designed blood collection drape and (2) the drape estimate with a measurement of blood loss by photospectrometry. Methods: A randomized controlled study was performed with 123 women delivered at the District Hospital, Belgaum, India. The women were randomized to visual or drape estimation of blood loss. A subsample of 10 drape estimates was compared with photospectrometry results. Results: The visual estimate of blood loss was 33% less than the drape estimate. The interclass correlation of the drape estimate to photospectrometry measurement was 0.92. Conclusion: Drape estimation of blood loss is more accurate than visual estimation and may have particular utility in the developing world. Prompt detection of postpartum hemorrhage may reduce maternal morbidity and mortality in low-resource settings. D 2006 International Federation of Gynecology and Obstetrics. Published by Elsevier Ireland Ltd. All rights reserved.
1. Introduction * Corresponding author. E-mail address: [email protected] (A. Patel).
Postpartum hemorrhage (PPH) is the leading cause of maternal mortality worldwide. It accounts for
0020-7292/$ - see front matter D 2006 International Federation of Gynecology and Obstetrics. Published by Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.ijgo.2006.02.014
Drape estimation vs. visual assessment for estimating postpartum hemorrhage 25% of maternal deaths, and most of these deaths occur in rural areas of developing countries [1]. The risk of PPH is increased in the absence of prompt and appropriate care, which includes the active management of the third stage of labor (gentle umbilical cord traction, uterine massage, and use of uterotonics), as well as fluid infusion, blood transfusion, and surgical intervention. Maternal mortality rates in rural India are estimated to be 350 to 650 per 100,000 live births, accounting for the world’s highest number of maternal deaths per year [1]. Postpartum hemorrhage, often preceded by antepartum anemia, is the primary cause of maternal morbidity and mortality in India. In spite of national preventive efforts to administer supplemental iron during the antepartum period, hemorrhage, in conjunction with anemia, still accounts for approximately 25— 30% of these maternal deaths [2]. Most deliveries occur at home or at rudimentary health facilities, where many proven therapeutic interventions are not available. Inaccurate blood loss estimates often delay recognition of PPH and active intervention at these low-resource settings. The current worldwide standard practice of postpartum blood loss assessment is visual estimation. A minimally trained health care provider generally observes blood lost during delivery and makes a quantitative or semiquantitative estimate. This approach has been shown for decades to be inaccurate, visual estimates being up to 50% less than actual values for blood lost in controlled studies [3,4,5]. Methods to quantify postpartum vaginal blood loss include visual estimation, direct collection, venous blood sampling, dye dilution techniques for plasma volume measurement, and red blood cell and plasma volume determinations using radioactive tracer elements [3,4,5,6,7,8,9,10,11]. The most accurate measures include venous blood sampling for determination of hemoglobin concentration, with and without assessment of blood volume by red blood cell labeling or spectrometry [3]. However, the most accurate methods have not been widely adopted because they are neither practical nor affordable in most clinical settings [7]. A blood collection drape (Fig. 1) was specially designed to assist in estimating postpartum blood loss in low-resource settings. It consists of a funneled and calibrated collecting pouch attached to a plastic sheet that is placed under the woman’s buttocks immediately after delivery. Two belts attached to the upper end of the drape are tied around the woman’s abdomen to optimize blood collection, particularly for deliveries performed on
221
Figure 1 The BRASS-V Drape, a specially designed blood collection drape with a calibrated collection pouch.
the floor or other flat surface. Calibration levels indicate the volume of blood collected by the drape. This collection drape has the potential to provide an objective measurement of postpartum blood loss and allows for a more accurately diagnosis of PPH than does visual assessment [12,13]. The present pilot study was conducted to compare (1) visual estimates of blood loss to drape estimates and (2) drape estimates to the bgold standardQ of photospectrometry.
2. Materials and methods A randomized, controlled, hospital-based trial was conducted from September through December 2003 with 123 women undergoing vaginal delivery at the District Hospital in Belgaum, Karnataka, India—a hospital affiliated with Jawaharlal Nehru Medical College. The internal review board approved the study. All eligible women were approached when they were admitted to the labor and delivery ward, and all consented to participate in the study. The house officer or attending physician on duty determined eligibility using the following inclusion criteria: (1) the woman was scheduled for vaginal delivery; (2) she was able to give written or verbal consent to participating in the study; and (3) she had no
222 contraindications for vaginal delivery. Given the difficulty of obtaining informed consent from a woman who is in active labor (defined as labor with a cervical dilation N4 cm), women in active labor were not recruited. Written informed consent forms in local languages, Kannada and Marathi, were used. For semiliterate or illiterate women, oral consent with witnessed written confirmation was obtained. The women were randomly assigned to the visual estimation group or the drape estimation group. To ensure balanced randomization and to conceal the treatment assignment, random block size was used as well as a computer-generated randomization list with a sequence of letters V (visual) and B (drape).
A. Patel et al. 15 min. The OD was read with a Tracer 30 spectrophotometer at 546 nm after 30 min. The volume of blood lost was calculated according to the following equation: Volume of blood lost ¼
OD sample 2000 10 mL OD blood standard 100
2.4. Statistical analysis Comparisons between the study groups were performed using v 2 tests for categorical data and t-tests between means for continuous data. Univariate and multivariate analyses were performed using SAS software (SAS Institute, Cary, NC, USA).
2.1. Visual assessment group The estimation of postpartum blood loss began immediately after delivery and the clamping and cutting of the umbilical cord. The visual estimate of blood loss was recorded after 1 h. If bleeding persisted after 1 h, estimates of blood loss were recorded for a maximum of 2 h. The physician performing the delivery estimated blood loss, as is standard practice.
2.2. Drape group The study drape was placed under the buttocks of the woman immediately after delivery and the clamping and cutting of the umbilical cord. If bleeding stopped within 1 h, the blood loss estimate was recorded. If bleeding persisted after 1 h, estimates of blood loss were recorded for a maximum of 2 h. The physician performing the delivery estimated blood loss by recording the calibration on the drape.
2.3. Photospectrometry measurement Photospectrometry was performed for the first 10 patients in the drape group according to a previously described method [3,4]. The blood standard was prepared utilizing 0.1 mL of the patient’s peripheral blood in 9.9 mL of a 5% sodium hydroxide solution. The optical density (OD) was read at 546 nm [3] with a Tracer 30 spectrophotometer 30 min after the addition of sodium hydroxide. The contents of the drape were added to 2 L of 5% sodium hydroxide and blended for 15 min in an inert container. All stained pads and gauze strips were also placed in the container. After the blended material was filtered under suction, 1 mL of the filtrate was diluted 10 times in a 5% sodium hydroxide solution and left to stand for another
3. Results Data on postpartum blood loss were collected on the 123 women, 61 in the visual group and 62 in the drape group. Spectrometry data were collected on the first 10 women assigned to the drape group. The women’s demographics were similar in the two groups. The mean blood loss in the visual group was reported at 203 mL (range, 50—950 mL) and assessed as 304 mL (range, 50—975 mL) in the drape group, a mean difference of 101 mL (Table 1). The difference between visual and measured estimates of blood loss was significant ( P b.001). The values of blood loss reported using the drape method and photospectrometry are shown in Table 2. The mean blood loss with photospectrometry measured at 546 nm was 188 mL (range, 93— 286 mL) while the mean blood loss with using the drape method was 239 mL (range 100—350 mL), a mean difference of 51 mL. The Pearson correlation coefficient for the two methods was 0.928, demonstrating a high level of interclass correlation and a high-level accuracy for the drape method.
4. Discussion In November 2003, the International Federation of Gynecology and Obstetrics (FIGO), and International Confederation of Midwives (ICM) issued a Table 1 groups
Distribution of blood loss between study
Blood loss, mean F S.D. (range) (mL) *P = 0.0008.
Visual group, n = 61
Drape group, n = 62
203.11 F 147.49 (50—950)*
302.82 F 173.28 (50—975)
Drape estimation vs. visual assessment for estimating postpartum hemorrhage Table 2 Comparison of blood loss estimation of photospectrometry and drape estimation for 10 sequential womena Patient number
Spectrometry measurement
Drape estimate
1 2 3 4 5 6 7 8 9 10
93.19 107.97 116.16 176.95 185.88 188.54 208.88 248.78 266.08 285.98
100.00 100.00 100.00 200.00 200.00 250.00 350.00 300.00 300.00 350.00
a Values are given in milliliters. Means are 187.80 and 238.90 and standard deviations are 67.10 and 96.10, respectively, for spectrometry measurements and drape estimates.
Joint Statement declaring the reduction of postpartum hemorrhage an integral component of the Safe Motherhood Global Initiative [14]. In most of the world, prompt detection of PPH is essential to expedite transport of the woman for interventions not available at the site of delivery. In such areas, an inexpensive and easily readable blood loss assessment instrument such as the collection drape used in this study may be useful. It may provide a more accurate measurement of postpartum blood loss, help to detect acute PPH, and thus, facilitate the woman’s rapid transfer for appropriate management. In this study, visual assessment underestimated postpartum blood loss by 33% compared with the drape assessment (203 mL vs. 304 mL). This finding is consistent with findings of other studies indicating an underestimation of 25—50% when visual assessment is compared to a more objective measurement [4,15]. Although the difference among the methods was statistically significant, a limitation of this study is its inability to clinically demonstrate significant differences in blood loss, primarily because of the small sample size and a relatively narrow blood loss range. A reported difference of 33% may not be clinically meaningful in women sustaining a small blood loss (a mean difference of 101 mL in this study); however, with a greater blood loss of 1000 mL or more (acute severe PPH), the 33% underestimation of visual estimate may well have a significant impact on maternal survival. This study also assessed the accuracy of the drape estimate by comparing it with the bgold standardQ of photospectrometry [3], and found a high level of correlation between the two methods (r = 0.928). This value, though significant, may be
223
lower than values reported in other studies, possibly because of the small sample size and because OD readings were taken at 546 nm (owing to the type of spectrometer available), whereas readings were taken at 550 nm in the other studies [3,4]. This may result in a 5% difference. Although an ideal study design would have included a comparison of visual assessment with photospectrometry measurement, it would have been difficult to perform such a comparison. To measure photospectrometrically the volume of lost blood visually estimated would require the accurate collection of blood that may have spilled on the floor, on sheets, or on the delivery table, and the collection method itself would require validation. The collection drape provides a method of blood collection that is accurate, easy to use, and relatively inexpensive (b $1.00 per sterilized drape), and it may allow for earlier and more accurate detection of PPH. This pilot study verified the greater accuracy of the drape collection method compared with visual assessment. However, further studies conducted in other settings with larger sample sizes and broader ranges of volumes of blood lost would be warranted to fully validate the blood loss estimation using the drape. Should such studies demonstrate similar findings, the drape could be an important and cost-effective tool to identify acute PPH in low-resource settings, thereby allowing for more timely diagnosis and appropriate intervention.
References [1] World Health Organization, United Nations’ Children’s Fund, United Nations Population Fund. Maternal Mortality in 2000: Estimates Developed by WHO, UNICEF, UNFPA. Geneva, Switzerland: World Health Organization; 2004. WHO/RHRO1.9. [2] Motashaw N. Root causes of maternal mortality: infancy to motherhood. J Fam Welf 1997;43:4 – 7. [3] Chua S, Ho LM, Vanaja K, Nordstrom L, Roy AC, Arulkumaran S. Validation of a laboratory method of measuring postpartum blood loss. Gynecol Obstet Invest 1998;46: 31 – 3. [4] Duthie SJ, Ven D, Yung GL, Guang DZ, Chan SY, Ma HK. Discrepancy between laboratory determination and visual estimation of blood loss during normal delivery. Eur J Obstet Gynaecol Reprod Biol 1991;38:119 – 24. [5] Pritchard JA. Blood volume changes in pregnancy and puerperium. Am J Obstet Gynecol 1962;84:1271. [6] Haswell JN. Measured blood loss at delivery. J Ind State Med Assoc 1981;74:34 – 6. [7] Hill JA, Fadel HE, Nelson MC, Nelson RM, Nelson GH. Blood loss at vaginal delivery. South Med J 1986;79:188 – 92. [8] Nelson GH, Ashford C, Williamson R, Amburn SD. Method for calculating blood loss at vaginal delivery. South Med J 1981;74:550 – 2.
224 [9] Nelson GH. Consideration of blood loss at delivery as percentage of estimated blood volume. Am J Obstet Gynecol 1980;138:1117 – 9. [10] Quinlivan WL, Brock JA. Blood volume changes and blood loss associated with labor: correlation of changes in blood volume measured by I-131 albumin and Evans blue dye, with measured blood loss. Am J Obstet Gynecol 1970;106:843 – 9. [11] Robson SC, Boys RJ, Hunter S, Dunlop W. Maternal homodynamics after normal delivery and delivery complicated by postpartum hemorrhage. Obstet Gynecol 1989;74:234 – 9. [12] Geller SE, Patel A, Niak VA, et al. Conducting international collaborative research in developing nations. Int J Gynecol Obstet 2004;87:267 – 71.
A. Patel et al. [13] Kodkany BS, Derman RJ, Goudar SS, et al. Initiating a novel therapy in preventing postpartum hemorrhage in rural India: a joint collaboration between the United States and India. Int J Fert Women’s Med 2004;49:91 – 6. [14] International Federation of Gynecology and Obstetrics, International Confederation of Midwives. Joint Statement on the Management of the Third Stage of Labour to Prevent Post-partum Haemorrhage. Available at: http://www.mnh. jhpiego.org/updates/2004/figoicmMNH0504.pdf. [15] Prasertcharoensuk W, Swadpanich U, Lumbiganon P. Accuracy of the blood loss estimation in the third stage of labor. Int J Gynaecol Obstet 2000;71:69 – 70.
www.elsevier.com/locate/ijgo
CLINICAL ARTICLE
Drape estimation vs. visual assessment for estimating postpartum hemorrhage A. Patel a,*, S.S. Goudar b, S.E. Geller c, B.S. Kodkany b, S.A. Edlavitch d, K. Wagh b, S.S. Patted b, V.A. Naik b, N. Moss e, R.J. Derman d a
John H. Stroger Jr. Hospital of Cook County, Chicago, IL, USA J. N. Medical College, Belgaum, Karnataka, India c Department of Obstetrics and Gynecology, University of Illinois at Chicago, Chicago, IL, USA d University of Missouri at Kansas City School of Medicine, Kansas City, MO, USA e Global Network for Women’s and Children’s Health Research, National Institute of Child Health and Human Development, Rockville, MD, USA b
Received 14 September 2005; received in revised form 10 February 2006; accepted 14 February 2006
KEYWORDS Postpartum hemorrhage; Maternal morbidity; Maternal mortality; Blood collection
Abstract Objective: To compare (1) visual estimation of postpartum blood loss with estimation using a specifically designed blood collection drape and (2) the drape estimate with a measurement of blood loss by photospectrometry. Methods: A randomized controlled study was performed with 123 women delivered at the District Hospital, Belgaum, India. The women were randomized to visual or drape estimation of blood loss. A subsample of 10 drape estimates was compared with photospectrometry results. Results: The visual estimate of blood loss was 33% less than the drape estimate. The interclass correlation of the drape estimate to photospectrometry measurement was 0.92. Conclusion: Drape estimation of blood loss is more accurate than visual estimation and may have particular utility in the developing world. Prompt detection of postpartum hemorrhage may reduce maternal morbidity and mortality in low-resource settings. D 2006 International Federation of Gynecology and Obstetrics. Published by Elsevier Ireland Ltd. All rights reserved.
1. Introduction * Corresponding author. E-mail address: [email protected] (A. Patel).
Postpartum hemorrhage (PPH) is the leading cause of maternal mortality worldwide. It accounts for
0020-7292/$ - see front matter D 2006 International Federation of Gynecology and Obstetrics. Published by Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.ijgo.2006.02.014
Drape estimation vs. visual assessment for estimating postpartum hemorrhage 25% of maternal deaths, and most of these deaths occur in rural areas of developing countries [1]. The risk of PPH is increased in the absence of prompt and appropriate care, which includes the active management of the third stage of labor (gentle umbilical cord traction, uterine massage, and use of uterotonics), as well as fluid infusion, blood transfusion, and surgical intervention. Maternal mortality rates in rural India are estimated to be 350 to 650 per 100,000 live births, accounting for the world’s highest number of maternal deaths per year [1]. Postpartum hemorrhage, often preceded by antepartum anemia, is the primary cause of maternal morbidity and mortality in India. In spite of national preventive efforts to administer supplemental iron during the antepartum period, hemorrhage, in conjunction with anemia, still accounts for approximately 25— 30% of these maternal deaths [2]. Most deliveries occur at home or at rudimentary health facilities, where many proven therapeutic interventions are not available. Inaccurate blood loss estimates often delay recognition of PPH and active intervention at these low-resource settings. The current worldwide standard practice of postpartum blood loss assessment is visual estimation. A minimally trained health care provider generally observes blood lost during delivery and makes a quantitative or semiquantitative estimate. This approach has been shown for decades to be inaccurate, visual estimates being up to 50% less than actual values for blood lost in controlled studies [3,4,5]. Methods to quantify postpartum vaginal blood loss include visual estimation, direct collection, venous blood sampling, dye dilution techniques for plasma volume measurement, and red blood cell and plasma volume determinations using radioactive tracer elements [3,4,5,6,7,8,9,10,11]. The most accurate measures include venous blood sampling for determination of hemoglobin concentration, with and without assessment of blood volume by red blood cell labeling or spectrometry [3]. However, the most accurate methods have not been widely adopted because they are neither practical nor affordable in most clinical settings [7]. A blood collection drape (Fig. 1) was specially designed to assist in estimating postpartum blood loss in low-resource settings. It consists of a funneled and calibrated collecting pouch attached to a plastic sheet that is placed under the woman’s buttocks immediately after delivery. Two belts attached to the upper end of the drape are tied around the woman’s abdomen to optimize blood collection, particularly for deliveries performed on
221
Figure 1 The BRASS-V Drape, a specially designed blood collection drape with a calibrated collection pouch.
the floor or other flat surface. Calibration levels indicate the volume of blood collected by the drape. This collection drape has the potential to provide an objective measurement of postpartum blood loss and allows for a more accurately diagnosis of PPH than does visual assessment [12,13]. The present pilot study was conducted to compare (1) visual estimates of blood loss to drape estimates and (2) drape estimates to the bgold standardQ of photospectrometry.
2. Materials and methods A randomized, controlled, hospital-based trial was conducted from September through December 2003 with 123 women undergoing vaginal delivery at the District Hospital in Belgaum, Karnataka, India—a hospital affiliated with Jawaharlal Nehru Medical College. The internal review board approved the study. All eligible women were approached when they were admitted to the labor and delivery ward, and all consented to participate in the study. The house officer or attending physician on duty determined eligibility using the following inclusion criteria: (1) the woman was scheduled for vaginal delivery; (2) she was able to give written or verbal consent to participating in the study; and (3) she had no
222 contraindications for vaginal delivery. Given the difficulty of obtaining informed consent from a woman who is in active labor (defined as labor with a cervical dilation N4 cm), women in active labor were not recruited. Written informed consent forms in local languages, Kannada and Marathi, were used. For semiliterate or illiterate women, oral consent with witnessed written confirmation was obtained. The women were randomly assigned to the visual estimation group or the drape estimation group. To ensure balanced randomization and to conceal the treatment assignment, random block size was used as well as a computer-generated randomization list with a sequence of letters V (visual) and B (drape).
A. Patel et al. 15 min. The OD was read with a Tracer 30 spectrophotometer at 546 nm after 30 min. The volume of blood lost was calculated according to the following equation: Volume of blood lost ¼
OD sample 2000 10 mL OD blood standard 100
2.4. Statistical analysis Comparisons between the study groups were performed using v 2 tests for categorical data and t-tests between means for continuous data. Univariate and multivariate analyses were performed using SAS software (SAS Institute, Cary, NC, USA).
2.1. Visual assessment group The estimation of postpartum blood loss began immediately after delivery and the clamping and cutting of the umbilical cord. The visual estimate of blood loss was recorded after 1 h. If bleeding persisted after 1 h, estimates of blood loss were recorded for a maximum of 2 h. The physician performing the delivery estimated blood loss, as is standard practice.
2.2. Drape group The study drape was placed under the buttocks of the woman immediately after delivery and the clamping and cutting of the umbilical cord. If bleeding stopped within 1 h, the blood loss estimate was recorded. If bleeding persisted after 1 h, estimates of blood loss were recorded for a maximum of 2 h. The physician performing the delivery estimated blood loss by recording the calibration on the drape.
2.3. Photospectrometry measurement Photospectrometry was performed for the first 10 patients in the drape group according to a previously described method [3,4]. The blood standard was prepared utilizing 0.1 mL of the patient’s peripheral blood in 9.9 mL of a 5% sodium hydroxide solution. The optical density (OD) was read at 546 nm [3] with a Tracer 30 spectrophotometer 30 min after the addition of sodium hydroxide. The contents of the drape were added to 2 L of 5% sodium hydroxide and blended for 15 min in an inert container. All stained pads and gauze strips were also placed in the container. After the blended material was filtered under suction, 1 mL of the filtrate was diluted 10 times in a 5% sodium hydroxide solution and left to stand for another
3. Results Data on postpartum blood loss were collected on the 123 women, 61 in the visual group and 62 in the drape group. Spectrometry data were collected on the first 10 women assigned to the drape group. The women’s demographics were similar in the two groups. The mean blood loss in the visual group was reported at 203 mL (range, 50—950 mL) and assessed as 304 mL (range, 50—975 mL) in the drape group, a mean difference of 101 mL (Table 1). The difference between visual and measured estimates of blood loss was significant ( P b.001). The values of blood loss reported using the drape method and photospectrometry are shown in Table 2. The mean blood loss with photospectrometry measured at 546 nm was 188 mL (range, 93— 286 mL) while the mean blood loss with using the drape method was 239 mL (range 100—350 mL), a mean difference of 51 mL. The Pearson correlation coefficient for the two methods was 0.928, demonstrating a high level of interclass correlation and a high-level accuracy for the drape method.
4. Discussion In November 2003, the International Federation of Gynecology and Obstetrics (FIGO), and International Confederation of Midwives (ICM) issued a Table 1 groups
Distribution of blood loss between study
Blood loss, mean F S.D. (range) (mL) *P = 0.0008.
Visual group, n = 61
Drape group, n = 62
203.11 F 147.49 (50—950)*
302.82 F 173.28 (50—975)
Drape estimation vs. visual assessment for estimating postpartum hemorrhage Table 2 Comparison of blood loss estimation of photospectrometry and drape estimation for 10 sequential womena Patient number
Spectrometry measurement
Drape estimate
1 2 3 4 5 6 7 8 9 10
93.19 107.97 116.16 176.95 185.88 188.54 208.88 248.78 266.08 285.98
100.00 100.00 100.00 200.00 200.00 250.00 350.00 300.00 300.00 350.00
a Values are given in milliliters. Means are 187.80 and 238.90 and standard deviations are 67.10 and 96.10, respectively, for spectrometry measurements and drape estimates.
Joint Statement declaring the reduction of postpartum hemorrhage an integral component of the Safe Motherhood Global Initiative [14]. In most of the world, prompt detection of PPH is essential to expedite transport of the woman for interventions not available at the site of delivery. In such areas, an inexpensive and easily readable blood loss assessment instrument such as the collection drape used in this study may be useful. It may provide a more accurate measurement of postpartum blood loss, help to detect acute PPH, and thus, facilitate the woman’s rapid transfer for appropriate management. In this study, visual assessment underestimated postpartum blood loss by 33% compared with the drape assessment (203 mL vs. 304 mL). This finding is consistent with findings of other studies indicating an underestimation of 25—50% when visual assessment is compared to a more objective measurement [4,15]. Although the difference among the methods was statistically significant, a limitation of this study is its inability to clinically demonstrate significant differences in blood loss, primarily because of the small sample size and a relatively narrow blood loss range. A reported difference of 33% may not be clinically meaningful in women sustaining a small blood loss (a mean difference of 101 mL in this study); however, with a greater blood loss of 1000 mL or more (acute severe PPH), the 33% underestimation of visual estimate may well have a significant impact on maternal survival. This study also assessed the accuracy of the drape estimate by comparing it with the bgold standardQ of photospectrometry [3], and found a high level of correlation between the two methods (r = 0.928). This value, though significant, may be
223
lower than values reported in other studies, possibly because of the small sample size and because OD readings were taken at 546 nm (owing to the type of spectrometer available), whereas readings were taken at 550 nm in the other studies [3,4]. This may result in a 5% difference. Although an ideal study design would have included a comparison of visual assessment with photospectrometry measurement, it would have been difficult to perform such a comparison. To measure photospectrometrically the volume of lost blood visually estimated would require the accurate collection of blood that may have spilled on the floor, on sheets, or on the delivery table, and the collection method itself would require validation. The collection drape provides a method of blood collection that is accurate, easy to use, and relatively inexpensive (b $1.00 per sterilized drape), and it may allow for earlier and more accurate detection of PPH. This pilot study verified the greater accuracy of the drape collection method compared with visual assessment. However, further studies conducted in other settings with larger sample sizes and broader ranges of volumes of blood lost would be warranted to fully validate the blood loss estimation using the drape. Should such studies demonstrate similar findings, the drape could be an important and cost-effective tool to identify acute PPH in low-resource settings, thereby allowing for more timely diagnosis and appropriate intervention.
References [1] World Health Organization, United Nations’ Children’s Fund, United Nations Population Fund. Maternal Mortality in 2000: Estimates Developed by WHO, UNICEF, UNFPA. Geneva, Switzerland: World Health Organization; 2004. WHO/RHRO1.9. [2] Motashaw N. Root causes of maternal mortality: infancy to motherhood. J Fam Welf 1997;43:4 – 7. [3] Chua S, Ho LM, Vanaja K, Nordstrom L, Roy AC, Arulkumaran S. Validation of a laboratory method of measuring postpartum blood loss. Gynecol Obstet Invest 1998;46: 31 – 3. [4] Duthie SJ, Ven D, Yung GL, Guang DZ, Chan SY, Ma HK. Discrepancy between laboratory determination and visual estimation of blood loss during normal delivery. Eur J Obstet Gynaecol Reprod Biol 1991;38:119 – 24. [5] Pritchard JA. Blood volume changes in pregnancy and puerperium. Am J Obstet Gynecol 1962;84:1271. [6] Haswell JN. Measured blood loss at delivery. J Ind State Med Assoc 1981;74:34 – 6. [7] Hill JA, Fadel HE, Nelson MC, Nelson RM, Nelson GH. Blood loss at vaginal delivery. South Med J 1986;79:188 – 92. [8] Nelson GH, Ashford C, Williamson R, Amburn SD. Method for calculating blood loss at vaginal delivery. South Med J 1981;74:550 – 2.
224 [9] Nelson GH. Consideration of blood loss at delivery as percentage of estimated blood volume. Am J Obstet Gynecol 1980;138:1117 – 9. [10] Quinlivan WL, Brock JA. Blood volume changes and blood loss associated with labor: correlation of changes in blood volume measured by I-131 albumin and Evans blue dye, with measured blood loss. Am J Obstet Gynecol 1970;106:843 – 9. [11] Robson SC, Boys RJ, Hunter S, Dunlop W. Maternal homodynamics after normal delivery and delivery complicated by postpartum hemorrhage. Obstet Gynecol 1989;74:234 – 9. [12] Geller SE, Patel A, Niak VA, et al. Conducting international collaborative research in developing nations. Int J Gynecol Obstet 2004;87:267 – 71.
A. Patel et al. [13] Kodkany BS, Derman RJ, Goudar SS, et al. Initiating a novel therapy in preventing postpartum hemorrhage in rural India: a joint collaboration between the United States and India. Int J Fert Women’s Med 2004;49:91 – 6. [14] International Federation of Gynecology and Obstetrics, International Confederation of Midwives. Joint Statement on the Management of the Third Stage of Labour to Prevent Post-partum Haemorrhage. Available at: http://www.mnh. jhpiego.org/updates/2004/figoicmMNH0504.pdf. [15] Prasertcharoensuk W, Swadpanich U, Lumbiganon P. Accuracy of the blood loss estimation in the third stage of labor. Int J Gynaecol Obstet 2000;71:69 – 70.