Postpartum hemorrhage is a leading cause of maternal morbidity and death worldwide, occurring in approximately 10.5% of deliveries and causing approximately 140,000 deaths annually.1 The National Inpatient Sample reported that postpartum hemorrhage occurred in 2.7% of deliveries in the United States and increased 26% from 1994 to 2006.2 A study of pregnancy-related mortality in the United States reported that obstetric hemorrhage caused 12% of pregnancy-related deaths.3
Physiologic adaptations to pregnancy including an increase in plasma volume and hemodilution provide a reserve of circulating volume and may delay signs of hemorrhage.4–6 Late recognition of significant blood loss may delay intervention until hemorrhage has occurred. Medical providers are inaccurate when estimating blood loss. Simulation studies have shown underestimation by up to 50% for volumes larger than 1,000 mL.7–9 At larger volumes, the tendency to underestimate is most pronounced.10 In contrast, providers tend to overestimate blood loss when the actual volume is small.7 This can lead to unnecessary interventions and costs.
In response to the poor ability to accurately assess blood loss by providers, researchers have studied educational programs to improve blood volume estimation. These interventions have demonstrated success.7–10 However, such programs using didactics and specific exercises are not necessarily transferable to other institutions, and deviations from the intervention studied may not be similarly effective.
In this study we aimed to assess accuracy of estimated obstetric blood loss by provider type and years of experience, and to determine if a reproducible visual aid can improve accuracy without formal didactics.
MATERIALS AND METHODS
We created a visual aid containing images of measured volumes of artificial blood on common obstetric materials on a single laminated pocket card (Fig. 1). This visual aid served as our study intervention. Our institution's simulation department created the artificial blood, which is similar in texture and color to real blood and is used regularly by this department for high-fidelity medical simulations.
All medical personnel who work in the labor, antepartum, and postpartum departments at our institution were invited to participate in this study, which occurred from August 2011 to June 2012. We collected demographic information for all participants by written questionnaire, and this information was validated by two authors (L.C.Z. and H.S.L.). Participants signed a confidentiality agreement before participation, agreeing not to discuss their experiences with other potential participants. One author (L.C.Z.) was present for all administrations of the tests and ensured no discussion or communication between participants throughout participation. At the conclusion of their participation, we did not provide feedback about performance accuracy, nor did we reveal to any participants the actual volumes used in our stations. We did not exclude any providers and all available providers willingly participated. This study was deemed exempt from Institutional Review Board approval as educational research.
We created six stations with known volumes of artificial blood on materials common to standard delivery kits (Fig. 2). The stations included the following: tail sponge, 30 mL; Chux pad A, 100 mL; Chux pad B, 200 mL; peri pad A, 100 mL; peri pad B, 250 mL; and delivery drape, 500 mL. Participants recorded estimated blood loss for each station before and after receiving our visual aid. Participants were not informed of their accuracy after the initial test, nor did they undergo didactics between tests. Instead, they were simply given the visual aid after initial assessment of the stations and were instructed to repeat their assessments on the same stations with the pocket card available for reference.
We assessed the effects of clinical role and years of experience on the accuracy of estimated blood loss. Years of experience was evaluated as a categorical variable with participants divided into the following three groups: less than 10 years of experience; 11 to 20 years of experience; and more than 20 years of experience. For blood volume estimation, we prospectively categorized percent error as underestimation (less than −20%), accurate assessment (−20% to +20%), and overestimation (more than +20%). To compare preintervention and postintervention accuracy of estimated blood loss, we used the McNemar test to assess whether there was significant improvement from inaccurate estimation (overestimation or underestimation) to accurate assessment among participants at each station. For comparison of three or more groups, we used the Wilcoxon signed rank test. Statistical tests were performed using SAS 9.3 statistical software and P<.05 was considered significant. We prospectively determined that with 150 participants and a probability of a type 1 error set at 0.05, this study would have 80% power to detect a 15% improvement in the percentage of accurate assessments, allowing for a 5% increase in assessments that were overestimations or underestimations when we compared posttests with pretests.
One hundred fifty-one maternity care providers participated in this study. Participants included obstetric and anesthesia physicians, certified nurse midwives, registered nurses, and students (including medical and nursing students) (Table 1). Years of experience ranged from 0 to 40 (median, 8). Provider type significantly affected accuracy of estimated blood loss before intervention in two of the six stations, peri pad A (P=.01) and delivery drape (P=.03), with nurses performing best in the peri pad A station and community attending physicians demonstrating best accuracy with the delivery drape. This difference persisted in one station, delivery drape, after intervention (P<.01). Before intervention, years of experience did not significantly affect accuracy in five of the six stations tested (P>.05), because this variable only affected accuracy in the peri pad A station (P=.003), where providers with more than 20 years of experience performed best. This difference did not persist postintervention. Using our predetermined categorization for underestimation, accurate assessment, and overestimation, we categorized all participants by percent error of estimated blood loss before and after receiving our visual aid. We used the McNemar test to evaluate preintervention and postintervention accuracy. We found a significant improvement in the frequency of accurate assessments of blood loss for all provider types after intervention across four of the six volumes tested (P<.001) (Table 2). Additionally, in a posttest survey, 90% of participants (n=136) responded affirmatively to the question, “Do you believe your values were more accurate with the second quiz?,” indicating that they felt that the visual aid did, in fact, improve their ability to accurately assess blood loss.
Our study demonstrates that a visual aid depicting known volumes of blood on common obstetric materials significantly improves obstetric provider accuracy in blood loss estimation. This benefit appears independent of provider type or years of experience.
Historically, pictorial representations of blood volumes have been used in the field of gynecology, specifically for assessing menstrual blood flow. A study published in 1990 by Higham et al11 reported that women's use of a “menstrual pictogram” had a sensitivity of 80% for diagnosing menorrhagia. Another study found that women were able to attain a high level of accuracy by utilizing a pictogram to determine menstrual blood volume.12
Researchers have previously reported on didactic programs that improve provider accuracy of estimated blood loss.8–10 However, deviations from the exact intervention studied may not be similarly effective. Also, poor knowledge retention argues against the use of a single didactic initiative. For example, Toledo et al13 demonstrated that both Web-based and live seminars on estimating blood loss improved obstetric provider estimated blood loss accuracy; however, a follow-up study published by this group in 2012 showed significant decay in improvement after 9 months. Our study is unique because our intervention lacks formal didactics and, instead, relies on a tangible visual aid that could be available any time.
Although we used relatively small blood volumes in our study, we believe that knowledge of blood absorbency characteristics of commonly used materials such as peri pads and Chux pads is clinically important. In our institution, these items are used continuously with obstetric patients and are often changed by nurses and patients without necessarily alerting providers of frequency or degree of saturation. Our study demonstrates that before use of our visual aid obstetric providers significantly underestimated the blood volume in a saturated peri pad and underestimated that three saturated peri pads contain blood loss of almost 1,000 mL. Interestingly, the Chux pad stations were the only stations where our visual aid did not significantly improve accuracy. In fact, as demonstrated in Table 2, the Chux pad B station demonstrated an impressive shift from preintervention underestimation of 22.5% of participants to only 2% underestimation in the postintervention group; however, the majority of this shift was to overestimation rather than accurate assessment. It is possible that these pads do not show predictable or easily assessable blood absorbency and therefore render estimated blood loss assessment more challenging than other materials. We would therefore consider other methods of assessing clinically significant bleeding when Chux pads are used, such as weighing the pads and closely monitoring vital signs as clinically indicated to better-assess bleeding. Our results suggest a potential area of improvement in these products, such as markers for saturation when the materials render accurate visual estimation otherwise challenging.
Certain limitations of our study must be acknowledged. Our simulated stations only represent a few blood volumes, and it is difficult to assess whether the information presented on the visual aid can be generalized to scenarios not depicted on the card. Our stations contained some of the same materials and blood volumes as the images in the visual aid, although we did demonstrate improvement in accuracy in stations not depicted on our visual aid, such as the delivery drape. Future studies assessing our visual aid with varying blood volumes will support the clinical utility of our visual aid. Although the visual aid illustrates individual blood collection products and our testing stations specifically assessed the ability of the providers to accurately assess blood volumes for those individual products, we recognized that clinical assessment would include an estimate of the sum of all of the collected blood loss. We assumed the ability of the providers to sum visible blood loss across different blood collection devices. We also acknowledged that a unique aspect of obstetric blood loss is the potential for dilution with amniotic and other bodily fluids, which we did not account for in this study. This is something that providers should remain aware of. Additionally, we used materials common to delivery kits in our institution. The blood absorption characteristics of various surgical materials and brands used elsewhere may create different visual cues and appearances for a given volume of blood; therefore, our images may not be applicable to all clinical settings. We plan to study our visual aid in different institutions and with varying blood volumes. Finally, our data do not allow us to ascertain whether the improvement in accuracy will positively affect clinical outcomes or reduce morbidity associated with postpartum hemorrhage.
Because obstetric hemorrhage is a significant contributor to maternal morbidity and mortality, interventions that improve diagnosis and expedient management are clinically valuable. There is currently a paucity of data regarding simple educational tools used to improve obstetric outcomes. With the increased use of Smartphone applications and computer programs to assist with clinical care, such educational tools are important to create and study for efficacy.
We demonstrate in this study that a visual aid depicting measured volumes of artificial blood on common obstetric materials improves provider accuracy of estimating blood loss. We postulate that such visual aids could improve maternity care personnel's assessment of blood loss in situations of antepartum bleeding and in estimating intrapartum and postpartum blood loss.
1. AbouZahr C. Global burden of maternal death and disability. Br Med Bull 2003;67:1–11.
2. Callaghan WM, Kuklina EV, Berg C. Trends in postpartum hemorrhage: United States, 1994-2006. Am J Obstet Gynecol 2010;202:353.e1–6.
3. Berg CJ, Callaghan WM, Syverson C, Henderson Z. Pregnancy-related mortality in the United States, 1998-2005. Obstet Gynecol 2010;116:1302–9.
4. Carbillon L, Uzan M, Uzan S. Pregnancy, vascular tone, and maternal hemodynamics: a crucial adaptation. Obstet Gynecol Surv 2000;55:574–81.
5. Lund CJ, Donovan JC. Blood volume during pregnancy. Am J Obstet Gynecol 1967;98:393.
6. Pritchard JA. Anesthesiology 1965;26:393.
7. Yoong W, Karavolos S, Damodaram M, Madgwick K, Milestone N, Al-Habib A, et al.. Observer accuracy and reproducibility of visual estimation of blood loss in obstetrics: how accurate and consistent are health-care professionals? Arch Gynecol Obstet 2010;281:207–13.
8. Dildy GA, Paine AR, George NC, Velasco C. Estimating blood loss: can teaching significantly improve visual estimation? Obstet Gynecol 2004;104:601–6.
9. Toledo P, McCarthy RJ, Burke CA, Goetz K, Wong CA, Grobman WA. The effect of live and web-based education on the accuracy of blood-loss estimation in simulated obstetric scenarios. Am J Obstet Gynecol 2010;202:400.e1–5.
10. Maslovitz S, Barkai G, Lessing JB, Amitai Z, Many A. Recurrent obstetric management mistakes identified by simulation. Obstet Gynecol 2007;109:1295–300.
11. Higham JM, O'Brien PM, Shaw RW. Assessment of menstrual blood loss using a pictorial chart. Br J Obstet Gynaecol 1990;97:734–9.
12. Wyatt KM, Dimmock PW, Wlaker TJ, Phil M, O'Brian PM. Determination of total menstrual blood loss. Fertil Steril 2001;76:125–31.
13. Toledo P, Eosadul ST, Goetz K, Wong CA, Grobman WA. Decay in blood loss estimation skills after web-based didactic training. Simul Healthc 2012;6:18–21.