This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Email this article to a colleague Similar articles in this journal Similar articles in Web of Science Alert me to new issues of the journal Download to citation manager Request Permissions Citing Articles Citing Articles via HighWire Citing Articles via Web of Science (3) Citing Articles via Google Scholar Google Scholar Articles by Toledo, P. Articles by Wong, C. A. Search for Related Content PubMed Articles by Toledo, P. Articles by Wong, C. A. Related Collections Obstetrics Blood

---

OBSTETRIC ANESTHESIOLOGY

The Accuracy of Blood Loss Estimation After Simulated Vaginal Delivery

From the Department of Anesthesiology, Northwestern University Feinberg School of Medicine, Chicago, Illinois.

Address correspondence and reprint requests to Cynthia A. Wong, MD, Department of Anesthesiology, 251 E. Huron St., F5-704, Chicago, IL 60611. Address e-mail to c-wong2{at}northwestern.edu.

Abstract

BACKGROUND: Visual blood loss estimation often underestimates blood loss. In this study we sought to determine the effect of calibrated drape markings on blood loss estimation in a simulated vaginal delivery.

METHODS: Subjects were randomized to estimate simulated blood loss (300, 500, 1000, and 2000 mL) in calibrated or noncalibrated vaginal delivery drapes and then crossover.

RESULTS: Visual blood loss estimation with noncalibrated drapes underestimated blood loss, with worsening accuracy at larger volumes (16% error at 300 mL to 41% at 2000 mL). The calibrated drape error was <15% at all volumes.

CONCLUSIONS: Calibrated vaginal delivery drapes improve blood loss estimation.

Postpartum hemorrhage, defined as blood loss more than 500 mL after a vaginal delivery, is a major cause of maternal morbidity and mortality (1,2). Delay in the diagnosis and treatment of postpartum hemorrhage may place the parturient at increased risk of adverse outcome. Clinicians typically diagnose postpartum hemorrhage by visual estimation of blood loss in the vaginal drapes at delivery. The drapes are conical in shape and collect blood, urine, and sponges used during the delivery. Studies have shown that visual assessment of estimated blood loss (EBL) can underestimate postpartum blood loss by 33%–50% compared with the "gold standard," photospectrometry (3). The magnitude of underestimation increases as the amount of blood loss increases; however, previous studies have not examined volumes >1000 mL (4–6). To our knowledge, the vaginal delivery drapes currently marketed do not have any volume calibrations. We hypothesized that adding calibrations to the vaginal delivery drapes would improve visual EBL assessment.

METHODS

The study was approved by Northwestern University's IRB. Participants recruited were obstetric and anesthesia attending and resident physicians, and obstetric nurses. Written informed consent was obtained before participation.

The study was conducted on two separate dates in a labor and delivery room. The bed was assembled as it would be for a vaginal delivery and the delivery drapes were suspended from the foot of the bed (Fig. 1). There were eight study stations, each containing one vaginal delivery drape. Two types of drapes were used: drapes with and without volume calibrations. Each vaginal delivery drape contained a known volume of blood, urine, and a number of surgical sponges (Table 1). The calibrated drapes had volume markings beginning at 500 mL with 500 mL increments to a total of 2500 mL. Drape calibrations were made using 0.9% saline and a 500 mL graduated cylinder (TD/TC ± 5%). Expired packed red blood cells were diluted with 0.9% saline to a hematocrit of 33% to simulate whole blood. Urine was simulated with 100 mL of 0.9% saline.

View larger version (94K): [in this window] [in a new window]   Figure 1. Photograph of noncalibrated drape containing 500 mL of blood and calibrated drape containing 1000 mL of blood. In addition to blood, drapes also contained 100 mL of simulated urine and surgical sponges.

Participants were randomized in blocks, by provider type to view either the four calibrated stations or the four noncalibrated stations, and then crossover. The order of the volumes within each set of stations was randomized. Subjects were not informed that the volumes in the two groups were the same. Participants received a data card for each station on which they wrote the volume estimate. After each station, the data card was collected. Alterations to the estimations were not allowed once the answers were recorded.

The primary outcome variable was accuracy of EBL in the calibrated versus noncalibrated drapes. Secondary outcome variables were the effects of provider type, level of training, and number of years of experience. The study sample (n = 100) was estimated to achieve 90% power to detect an intraclass correlation of 0.90 under the alternative hypothesis and, assuming four observations per subject when the intraclass correlation under the null hypothesis is 0.70, using an F-test with = 0.05.

The difference between the visual EBL and the actual blood volume was compared between the groups that initially estimated calibrated versus the noncalibrated drapes using ANOVA with repeated measures. The type of provider, level of training, and years of experience were covariates. Post hoc analysis was conducted using Bonferroni adjustment. Bland and Altman analysis curves were constructed by comparison of the differences of the estimated versus the actual blood volume. P < 0.05 was used to reject the null hypothesis.

RESULTS

One hundred six subjects participated. There were no differences in gender, level of training, or years of experience between groups (Table 2). The differences in EBL from actual volumes were larger in subjects viewing noncalibrated drapes compared with the subjects who viewed the calibrated drapes first (P = 0.002) (Fig. 2). In addition, the accuracy of EBL worsened with increasing blood volume in the former group (P < 0.05).

View larger version (14K): [in this window] [in a new window]   Figure 2. Accuracy of visual blood loss estimations in subjects that viewed noncalibrated (top) and calibrated (bottom) conical shaped vaginal delivery drapes first. Each circle represents a single participant. Slope and intercept of regression lines are –0.46 and 166 for noncalibrated drapes and –0.09 and 11 for the calibrated drapes.

The accuracy of the EBL in the second four stations viewed by each group is shown in Figure 3. The error in the EBL was reduced to <15% at all volumes when the group that saw noncalibrated drapes first viewed the calibrated drapes. The overall error in EBL was <15% at all volumes when noncalibrated drapes were viewed second.

View larger version (13K): [in this window] [in a new window]   Figure 3. Accuracy of visual blood loss estimations in subjects that viewed calibrated (top) and noncalibrated (bottom) conical shaped vaginal delivery drapes second. Each circle represents a single participant. Slope and intercept of regression lines are –0.11 and 25 for calibrated drapes and –0.15 and 98 for the noncalibrated drapes. There were no differences between the groups.

There was no difference in accuracy of EBL between groups based on provider type, level of training, or years of experience. Marginal mean differences adjusted for covariates are shown in Table 3.

DISCUSSION

The important finding of this study was that the addition of calibration markings significantly improved the accuracy of EBL in conical vaginal delivery drapes. We also found EBL underestimation was proportional to the volume and was as high as 41% at 2000 mL with noncalibrated drapes, but was reduced to between 9% and 11% at this volume with calibrated drapes. Our study confirmed the findings of previous investigators who have demonstrated in actual vaginal deliveries that visual estimation of blood loss is inaccurate when compared with quantitative methods (3,6). This has important clinical implications as delayed diagnosis and treatment of postpartum hemorrhage may contribute to maternal deaths (1,2).

There are several quantitative methods to estimate blood loss; however, most are impractical in the delivery suite. The gold standard of photospectrometry, which has a 10% error rate, requires special equipment (7). A reference blood standard is prepared by mixing the patient's blood with 5% sodium hydroxide solution. The contents of the drapes, including sponges, are soaked in sodium hydroxide solution and then the blended material is filtered and the filtrate is again diluted. Finally, the standard and sample solutions are read using a spectrophotometer and blood loss is calculated using a standard equation. Radioisotope dilution, which also has an error rate of approximately 10%, is technically difficult and requires special equipment and disposal (8). Gravimetric methods require weighing all materials and may take many hours to complete (9). Hence, clinicians estimate blood loss based on visual assessment of the contents of the vaginal delivery drapes.

The use of calibrated drapes as a means of improving blood loss estimates after vaginal deliveries has been shown to correlate (r = 0.92) with blood loss measured by photospectrometry (3). The use of calibrated drapes resulted in EBL volumes 33% more than those obtained by visual estimation. In the aforementioned study, however, blood loss determinations by photospectrometry were made only in 10 of the 163 patients studied. In addition, the range of blood loss volume and absolute volumes were small (93–285 mL). No previous clinical study has compared blood loss volumes consistent with hemorrhage to standardized methods of blood loss estimation or compared calibrated drapes to noncalibrated drapes.

One interesting finding of this study was that there was no difference between providers (anesthesiologists, obstetricians, or nurses) in the accuracy of blood loss estimation, nor was there an association between accuracy and years of training or experience. In a study examining blood loss estimation using simulated clinical scenarios, Bose et al. (5) found that anesthesiologists more accurately estimated blood loss (median overestimate of 4%) compared to other health care providers, all of whom underestimated blood loss (median estimate –11% to –32%). However, only a small number of obstetric anesthesiologists participated in the study (9 of 103 participants). In the scenarios where there was blood spilled onto the floor, both the anesthesiologists and the other obstetric team members underestimated blood loss by 35%–50%, similar to the errors in estimation found at large volumes in our study. These investigators did not examine the level of experience of the participants.

We found that there was a learning effect in that, participants who viewed the calibrated drapes, made better estimations of EBL in the noncalibrated drapes. Our study was not designed to test the long-term retention of this information. A limitation to our study is that we asked participants to estimate blood loss based on inspection of the conical drape only. We did not ask them to estimate blood loss in hidden places, e.g., blood spills on the bed, sheets, or floor. In the situation where there is hidden blood loss, experienced clinicians may be more accurate at blood loss estimation. An additional limitation of this study is that this was not an actual delivery. We did, however, include common distractors, such as urine and surgical sponges in the drape, but did not include amniotic fluid. As there was no associated clinical scenario in our study, it is not clear whether knowledge of commonly available clinical clues (e.g., vital signs, inspection of the perineum) would have improved the accuracy of blood loss estimation.

In conclusion, we demonstrated that visual EBL using noncalibrated conical delivery drapes is inaccurate and that this inaccuracy is increased at clinically significant volumes of blood loss. Our study demonstrated that the addition of calibrations to vaginal delivery drapes can improve the accuracy of EBL, to an extent similar to other quantifying methods such as photospectrometry, without the added cost or limitations. Further study is warranted to determine whether the use of calibrated vaginal delivery drapes could prevent delay in diagnosis and treatment of postpartum hemorrhage.

Footnotes

Accepted for publication August 6, 2007.

Dr. Cynthia A. Wong, Section Editor for Obstetric Anesthesiology, was revised from all decisions related to this manuscript.

Bradley J. Hewlett was sponsored by the Foundation for Anesthesia Education and Research Medical Student Anesthesia Research Fellowship.

REFERENCES

1. Berg CJ, Atrash HK, Koonin LM, Tucker M. Pregnancy-related mortality in the United States, 1987–1990. Obstet Gynecol 1996;88:161–7[Web of Science][Medline]
2. Why mothers die 2000–2002: the confidential enquiries into maternal deaths in the United Kingdom. London (United Kingdom): HMSO, 2004
3. Patel A, Goudar SS, Geller SE, Kodkany BS, Edlavitch SA, Wagh K, Patted SS, Naik VA, Moss N, Derman RJ. Drape estimation vs. visual assessment for estimating postpartum hemorrhage. Int J Obstet Gynecol 2006;93:220–4
4. Prasertcharoensuk W, Swadpanich U, Lumbiganon P. Accuracy of blood loss estimation in the third stage of labor. Int J Obstet Gynecol 2000;71:69–70
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[Web of Science][Medline]
6. Dildy GA, Archana PR, George NC, Velasco C. Estimating blood loss: can teaching significantly improve visual estimation? Obstet Gynecol 2004;104:601–6[Web of Science][Medline]
7. 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;36:31–3
8. Quinlivan WLG, Brock JA. Blood volume changes and blood loss associated with labor. Am J Obstet Gynecol 1970;106:843–9[Web of Science][Medline]
9. Nelson GH, Ashford C, Williamson R, Amburn SD. Method for calculating blood loss at vaginal delivery. South Med J 1981;74:550–2[Web of Science][Medline]

This article has been cited by other articles:

				M. Tebruegge, I. Misra, A. Pantazidou, A. Padhye, S. Maity, B. Dwarakanathan, S. Donath, N. Curtis, and V. Nerminathan Estimating Blood Loss: Comparative Study of the Accuracy of Parents and Health Care Professionals Pediatrics, October 1, 2009; 124(4): e729 - e736. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Email this article to a colleague Similar articles in this journal Similar articles in Web of Science Alert me to new issues of the journal Download to citation manager Request Permissions Citing Articles Citing Articles via HighWire Citing Articles via Web of Science (3) Citing Articles via Google Scholar Google Scholar Articles by Toledo, P. Articles by Wong, C. A. Search for Related Content PubMed Articles by Toledo, P. Articles by Wong, C. A. Related Collections Obstetrics Blood