NRSG370 Case Study: Managing Postpartum Hemorrhage after LUCS

This assignment presents a case study focusing on the relationship between blood loss and clinical signs, particularly in the context of postpartum hemorrhage (PPH). It includes a systematic review analyzing various studies that explore the association between blood loss and clinical parameters such as heart rate (HR), systolic blood pressure (SBP), and shock index. The review highlights the variability in these relationships, making it challenging to establish definitive cut-off points for clinical interventions. The case study involves a 42-year-old woman undergoing a lower uterine caesarean section (LUCS) with a history of placenta previa, gestational diabetes, depression, and anxiety. The assignment assesses the initial post-operative observations, including vital signs and blood loss, and emphasizes the importance of early identification and management of PPH to prevent severe maternal morbidity and mortality. The study also points out the limitations of visual estimation of blood loss (VEBL) and the potential utility of the shock index as an indicator of cardiovascular compensation due to blood loss.

Pacagnella, RC; Souza, JP; Durocher, J; Perel, P; Blum, J; Winikoff,
B; Gulmezoglu, AM (2013) A Systematic Review of the Relationship
between Blood Loss and Clinical Signs.PLoS One, 8 (3).ISSN 1932-
6203 DOI: https://doi.org/10.1371/journal.pone.0057594
Downloaded from:http://researchonline.lshtm.ac.uk/856652/
DOI: 10.1371/journal.pone.0057594
Usage Guidelines
Please refer to usage guidelinesat http://researchonline.lshtm.ac.uk/policies.htmlor alterna-
tively contact [e-mail].
Available under license:http://creativecommons.org/licenses/by/2.5/

A Systematic Review of the Relationship between Blood
Loss and ClinicalSigns
Rodolfo Carvalho Pacagnella1*, Joa˜o Paulo Souza2, Jill Durocher 3, Pablo Perel4, Jennifer Blum3,
Beverly Winikoff3, Ahmet Metin Gu¨lmezoglu2
1 FederalUniversity ofSa˜o Carlos,Sa˜o Carlos,Brazil,2 Department ofReproductive Health and Research,World Health Organization,Geneva,Switzerland,3 Gynuity
Health Projects,New York,New York,United States of America,4 London Schoolof Hygiene and TropicalMedicine,London,United Kingdom
Abstract
Introduction:This systematic review examines the relationship between blood loss and clinical signs and explores its use
trigger clinicalinterventions in the management of obstetric haemorrhage.
Methods:A systematic review of the literature was carried out using a comprehensive search strategy to identify studies
presenting data on the relationship of clinical signs & symptoms and blood loss. Methodological quality was assessed u
the STROBE checklist and the generalguidelines of MOOSE.
Results:30 studies were included and five were performed in women with pregnancy-related haemorrhage (other studie
were carried in non-obstetric populations).Heartrate (HR),systolic blood pressure (SBP)and shock index were the
parameters most frequently studied.An association between blood loss and HR changes was observed in 22 out of24
studies,and between blood loss and SBP was observed in 17 out of 23 studies.An association was found in allpapers
reporting on the relationship ofshock index and blood loss.Seven studies have used Receiver Operating Characteristic
Curves to determine the accuracy of clinical signs in predicting blood loss. In those studies the AUC ranged from 0.56 t
for HR,from 0.56 to 0.79 for SBP and from 0.77 to 0.84 for shock index.In some studies,HR,SBP and shock index were
associated with increased mortality.
Conclusion:We found a substantial variability in the relationship between blood loss and clinical signs, making it difficult
establish specific cut-off points for clinical signs that could be used as triggers for clinical interventions.However, the shock
index can be an accurate indicator of compensatory changes in the cardiovascular system due to blood loss.Considering
that most of the evidence included in this systematic review is derived from studies in non-obstetric populations,further
research on the use of the shock index in obstetric populations is needed.
Citation: Pacagnella RC,Souza JP,Durocher J,Perel P,Blum J,et al.(2013) A Systematic Review of the Relationship between Blood Loss and Clinical Signs.PLoS
ONE 8(3):e57594.doi:10.1371/journal.pone.0057594
Editor: Shannon M.Hawkins,Baylor College of Medicine,United States of America
Received August 6,2012;Accepted January 24,2013;Published March 6,2013
Copyright: ß 2013 Pacagnella et al.This is an open-access article distributed under the terms of the Creative Commons Attribution License,which permits
unrestricted use,distribution,and reproduction in any medium,provided the originalauthor and source are credited.
Funding: Work on this systematic review was financially and technically supported by Gynuity Health Projects and the World Health Organization.The funders
had no role in study design,data collection and analysis,decision to publish,or preparation of the manuscript.
Competing Interests: The authors have declared that no competing interests exist.
* E-mail:[e-mail]
Introduction
All women giving birth lose some amount ofblood during the
immediate postpartum period.In the majority ofwomen,the
postpartum blood loss is welltolerated.In some women excessive
bleeding occurs and is associated with severe maternalmorbidity
and mortality. Postpartum haemorrhage (PPH) is one of the major
causesof maternaldeathsaround the world and itsunderlying
causes include uterine atony,genitaltracttears and retention of
placental tissue [1]. Depending on the rate of blood loss and other
factors such as pre-existing anaemia,untreated PPH can lead to
hypovolemic shock,multi-organ dysfunction and maternaldeath
within 2 to 6 hours[2,3].Therefore,early identification and
treatmentof women with PPH isa key factorfor maternal
survival.
The diagnosis ofPPH is largely based on the identification of
excessive blood loss in the postpartum period. In 1990, the World
Health Organization adopted the definition of PPH after vaginal
delivery as the loss of500 mlor more ofblood from the genital
tract after delivery ofa baby.Primary PPH is usually defined as
excessive blood loss thatoccurs within 24 hours after birth and
a blood loss of 1000 mlor more is defined as severe PPH [4].In
caesarean sections,a higher threshold for diagnosing PPH (e.g.
750–1000 ml)is generally accepted.
Direct measurement is the idealmethod for quantifying blood
loss after birth. The majority of PPH-related maternal deaths take
place in under-resourced settings and the use ofdirectmethods
(e.g.gravimetric or photometric)for quantifying blood loss in all
births is not realistic [5]. Visual estimation of blood loss (VEBL) is
the method most frequently used around the world in the diagnos
of PPH which is based on clinicaljudgmentby the provider via
visualestimation ofblood loss.However,the use ofVEBL has
been associated with underestimation of the amount of blood loss
[6]. Considering these limitations,other methodsfor estimating
blood loss have beenproposed(e.g.hematocrit/hemoglobin
PLOS ONE |www.plosone.org 1 March 2013 |Volume 8 |Issue 3 | e57594

assessment)together with alternative PPH definitions(e.g.10%
drop in hematocrit/hemoglobin)[7–9].Nevertheless,the added
benefitof these alternative methodsin comparison with VEBL
seemsto be minimaland their applicability in under-resourced
settings is limited.
Clinicalsigns have been used as a surrogate for blood loss in
non-obstetricpopulations,particularlywhen quantification of
blood loss is not feasible (e.g.trauma and occult bleeding).More
importantly,clinicalsignshavebeen used to guidefluid re-
placementin trauma patientswith hypovolemic shock due to
haemorrhage [10].By analogy,some authors have suggested the
use ofclinicalsigns and symptoms ofhypovolemia as markers of
PPH [8,11]. Signs and symptoms such as pallor, light-headedness,
weakness,palpitations,tachycardia,diaphoresis,restlessness,
confusion,air hunger,syncope,fatigue and oliguria have been
associated with blood loss [11].However,none ofthese clinical
signs and symptoms has been properly correlated with different
degrees of hypovolemia in obstetric populations and there has not
been any systematic review to assessthe relationship between
blood lossand clinicalsignsand symptoms.Other relevant
uncertaintiesrelateto the amountof blood lossthat should
indicate a diagnosisof PPH and whatclinicalconsequencesof
blood loss are ofgreatest importance in predicting consequences
for women experiencing excessive blood loss.
This systematicreview aimsat assessingthe relationship
between blood loss and clinicalsigns and explores the potential
of using such clinical findings to trigger clinical interventions in the
management of PPH.
Methods
We conducted a systematic review of the literature following the
Meta-AnalysisOf ObservationalStudies in Epidemiology
(MOOSE) guidelinesand the Preferred ReportingItemsfor
Systematic reviews and Meta-Analyses (PRISMA)statement [12].
The primaryfocusof this systematicreview is postpartum
haemorrhage and other pregnancy-related bleeding. We included
studiespresenting data on the clinicalsignsand symptomsin
relation to blood loss estimations in order to assess the diagnostic
accuracy of clinicalsigns for a specific amount of blood loss. Due
to the anticipated paucity ofdata from obstetric populations,we
also included studiesconducted in other populations.Papersin
which the relationship between blood loss and clinicalsigns was
not clear or could not be determined were excluded (Figure 1).
An electronic search wasconducted in February 2012 using
internationalstudy databasesincludingMedline, EMBASE,
Lilacs,Scielo,ISI and Google Scholar.We did notrestrictthe
search strategy to exclude papers published in other languages, o
studiesof specificpopulationsor studydesign.The search
strategiesused in each database are available in Appendix S1.
All citations identified through the electronic search had their title
and/or abstracts examined.All potentially relevantpapers were
retrieved and assessed in detail.All manuscriptsthatwere fully
retrieved had theirreferencelistsscreened to identifyother
potentially relevant papers.The finalset of papers to be included
in the review was determined by consensus by two reviewers (RC
and JPS).
A pre-designed form was used independently by two reviewers
(RCP and JPS) to conductstudy eligibility assessment,critical
Figure 1. Inclusion and exclusion criteria.
doi:10.1371/journal.pone.0057594.g001
Relationship between Blood Loss and ClinicalSigns
PLOS ONE |www.plosone.org 2 March 2013 |Volume 8 |Issue 3 | e57594

appraisal and data collection. Data on the following variables were
collected: type of study, population, blood loss assessment method,
clinicalsignsdata (i.e.heartrate,systolic blood pressure,pre-
hospitalsystolicblood pressure(PSBP),mean arterialblood
pressure,diastolic blood pressure,body temperature,respiratory
rate,pulse pressure,shock index,diuresis,Glasgow coma score),
clinical-sign assessment method and statistical method used. These
reviewers assessed the methodologicalquality ofstudies indepen-
dentlyusing thechecklistof essentialitemsdescribed in the
STROBE (Strengthening the Reporting ofObservationalstudies
in Epidemiology)[13] statementand the generalguidelinesof
MOOSE.
The included studieswere classified intothreecategories,
according to the mode ofblood loss estimation:directmeasure-
ment(i.e.using drapes,drains,suctions,or visualestimation),
indirectmeasurement(e.g. weighingsponges,hemodynamic
Figure 2. Flow diagram of identification and retrieval of examined studies.
doi:10.1371/journal.pone.0057594.g002
Relationship between Blood Loss and ClinicalSigns
PLOS ONE |www.plosone.org 3 March 2013 |Volume 8 |Issue 3 | e57594

monitoring,blood losssimulation or proxies)and simulation of
blood loss in healthy subjects.In each study,the presence ofan
association between blood loss and the occurrence ofchanges in
vitalsignswasdetermined and classified aspresentor absent.
Anotheranalysiswas performed usingpopulation categories,
accordingto the effectcondition(pregnancyrelatedstudy
population,trauma population and healthy population).
The MicrosoftExcelsoftware was used in the tabulation and
analysis of the abstracted data. Since a meta-analysis would not be
appropriate due to the variation in study designs,we were only
able to perform a qualitative analysis ofthe correlation between
clinicalsigns and symptoms and the estimation of blood loss.
Results
A totalof 4023 citations was identified by the electronic search
and 75 manuscripts were retrieved for full-text assessment. Revie
of the referencelistsof the selected articlesresulted in the
identification of6 additionalstudies.In total,30 studieswere
included in the systematic review (Figure 2).
Only five studies were performed in women with pregnancy-
related haemorrhage:one study evaluated women with PPH and
four included women with ectopicpregnancy (Table1). The
majority of the studies (19 out of 30)was related to haemorrhage
due to trauma.Five studies were experimentaland consisted of
simulation ofblood loss through the use ofLow-Body Negative
Pressureor Tilt-testin healthy malestudy subjects.In eight
studies,methods enabling direct blood loss estimation were used.
Table 1. Characteristics of included studies.
Study Population Blood loss estimation method Country Participants
Direct Blood Loss estimation
Birkhahn (2002) [33] Ectopic Pregnancy Qualitative Ruptured ectopic pregnancies USA 280
Birkhahn (2003) [34] Ectopic Pregnancy Quantitative Ruptured EP and visualestimation of hemoperitoneum USA 52
Hick (2001) [27] Ectopic Pregnancy Quantitative Aspiration of the abdominalcavity USA 50
Jaramillo (2010) [35] Ectopic Pregnancy VEBL Visualestimation of hemoperitoneum USA 65
Robson (1989) [28] Postpartum VEBL Clinically Estimation .500 ml England 40
Birkhahn (2005) [31] Healthy Subjects Quantitative Blood donation of 450 ml USA 46
Baron (2004) [36] Trauma Qualitative Chest tube drainage,intraoperative blood loss,and
radiographic evidence of bleeding
USA 108
Scalea (1990) [37] Trauma Quantitative Drainage of cavity OR Ht modification USA 26
Indirect Blood Loss estimation
Brasel (2007) [22] Trauma Proxies Intervention to stop bleeding USA 10,825
Bruns (2007) [38] Trauma Proxies Intervention to stop bleeding USA 404
Bruns (2008) [39] Trauma Proxies Requiring blood transfusion USA 16,365
Cancio (2008) [40] Trauma Proxies Requiring blood transfusion USA 536
Chen (2007) [41] Trauma Proxies Requiring blood transfusion and bleeding trauma USA 492
Chen (2008) [42] Trauma Proxies Requiring blood transfusion and bleeding trauma USA 358
Edelman (2007) [43] Trauma Proxies Requiring blood transfusion USA 2,071
Guly (2011) [19] Trauma Proxies Clinicalestimation based on blood
loss in specific injuries
England/
Wales
199,657
Hagiwara (2010) [44] Trauma Proxies Requiring blood transfusion . = 4 un Japan; 261
Luna (1989) [45] Trauma Proxies Requiring blood transfusion . = 5 un USA 116
McLaughlin (2009) [46] Trauma Proxies Requiring blood transfusion USA 302
Opreanu (2010) [47] Trauma Proxies Intervention to stop bleeding USA 388
Parks (2006) [48] Trauma Proxies Base deficit estimation as a marker of shock USA 117,686
Vandromme (2010) [49] Trauma Proxies Requiring Blood Transfusion .6 Un/24 h USA 787
Vandromme (2011a) [30] Trauma Proxies Requiring Blood Transfusion .10 Un/24 h USA 8,111
Vandromme (2011b) [50] Trauma Proxies Requiring blood transfusion .10 un/24 h USA 514
Zarzaur (2008) [51] Trauma Proxies Requiring blood transfusion .4 un/48 h USA 16,077
Simulation
Convertino (2006) [52] Healthy Subjects Simulation LBNP USA 10
Convertino (2009) [53] Healthy Subjects Simulation LBNP USA 10
Rickards (2008) [54] Healthy Subjects Simulation LBNP USA 12
Secher (1984) [55] Healthy Subjects Simulation Tilt-test Denmark 6
Ward (2010) [56] Healthy Subjects Simulation LBNP USA 20
VEBL – Visualestimation of blood loss;LBNP – Low body negative pressure.
doi:10.1371/journal.pone.0057594.t001
Relationship between Blood Loss and ClinicalSigns
PLOS ONE |www.plosone.org 4 March 2013 |Volume 8 |Issue 3 | e57594

Indirect methods and simulation methods were used in 17 and 5
studies,respectively.Moststudies were conducted in the United
Statesof America (26 outof 30)and none were conducted in
developing countries.The studies’sample sizes ranged from 6 to
20 in the simulation group,from 116 to 199,657 participants
(median = 404)in the indirectblood loss estimation group and
from 26 to 280 in the direct measurement group.
An overallassessmentof the methodologicalquality ofthe
included studiesis presented in the Table 2.Nine studieswere
considered ofhigh quality.Detailed description ofthe study
methodswere lacking in mostof the studiesincluded in this
review.For instance,21 studiesdid not describeor provide
sufficientdetailof the study population,the health status ofthe
population or the inclusion criteria.
The majority ofstudies did not provide information regarding
the method ofassessmentof clinicalsigns.Of the 11 studies in
which this information is available, only one performed the clinical
evaluationusing manualdevicesand ten performedsuch
evaluations with automatic devices.
Table 3 summarizesthe findingsrelated to the association
between clinicalsigns and blood loss.Heartrate,systolic blood
pressure and the shock index were the clinical signs or clinical-sig
derivatives most frequently studied. An association between blood
loss and heart rate changes was observed in 22 out of 24 studies,
and an association between blood loss and systolic blood pressure
wasobserved in 17 outof 23 studies.One study showed an
association between pre-hospitalsystolic blood pressure changes
(i.e.measurementtaken before reaching the hospital)and blood
loss.
A statistically significant association was found in all10 papers
reporting on the relationship ofshock index (SI)and blood loss.
Fewer studies evaluated the relationship between blood loss and
other clinicalsigns:mean arterialpressure (4 outof 6 found an
association), diastolic pressure (5/8), pulse pressure (4/6) and bod
temperature (2/2).Respiratory rate,diuresis and Glasgow coma
scale were notassociated with blood loss.In the subgroup of
studies including only women with pregnancy-related blood loss,
associations between blood loss and the shock index,heartrate
and systolic blood pressure were found.
Several approaches were used to assess the relationship betwe
clinicalsignsand blood lossin the included studies(Table 4).
Seven studiesused Received Operator Characteristic Curves to
determine the accuracy of clinical signs in predicting blood loss. In
those studies,the Area Under Curve (AUC)ranged from 0.56 to
0.74 for heart rate,from 0.56 to 0.79 for SBP and from 0.77 to
0.84 for shock index (Table 5).Seven studies (Table 6)provided
informationon the relationshipbetweenclinical signsand
mortality.Of those,one study found HR and SI associated with
highermortality,and all of them found thatlow SBP was
associated with higher mortality.
Discussion
This systematic review identified 30 scientific papers reporting
on the relationship between blood loss and clinicalsigns. Overall,
these studies found a substantialvariability in the relationship of
blood loss and clinical signs, making it difficult to establish specifi
cut-offpoints for clinicalsigns thatcould be used as triggers for
clinicalinterventions.However,the shock index seemsto be
a promising indicator of the severity of blood loss.
To the best of our knowledge this is the first systematic review o
studies assessing the relationship of blood loss and clinical signs i
the contextof pregnancyand childbirth.This review was
conducted following the mostrecentmethodologicalguidelines
for reviews of this kind and did not have any restrictions in terms
of language and source ofdata.Nevertheless,some limitations
need to be noted.First,thisreview intendsto inform decisions
concerning pregnancy-related haemorrhage.Due to paucity of
data on obstetric populations,most of the studies included in this
review come from non-obstetric populations.This is a relevant
issue considering that women experience substantialphysiological
changesduring pregnancy (e.g.increase in the maternalblood
volume and cardiac output,reduction ofcardiovascular reserve)
[14,15]. A second issue that needs to be considered is that most o
the women experiencing severe complicationsrelated to post-
partum haemorrhage are in developing countries. Anemia during
pregnancy due to iron deficiency or other factors(e.g.malaria)
affectsup to 55% of pregnantwomen from low and middle
income countries compared with around 20% or less from high
income countries [16,17].Anemia may impair the physiological
response to blood lossand worsen maternalprognosis.So, the
evidencegeneratedby this systematicreview needsto be
considered in the contextof indirectnessdue to differencesin
population and setting.
Table 2. Criticalappraisalof included studies.
Study Type of study Quality
Birkhahn (2002) Diagnostic Test Accuracy High
Convertino (2009) Experimental(simulation) High
Birkhahn (2005) Prospective Cohort High
McLaughlin (2009) Prospective Cohort High
Chen (2007) Cross-Sectional High
Edelman (2007) Cross-Sectional High
Guly (2011) Cross-Sectional High
Hagiwara (2010) Cross-Sectional High
Vandromme (2010) Cross-Sectional High
Birkhahn (2003) Diagnostic Test Accuracy Moderate
Bruns (2007) Diagnostic Test Accuracy Moderate
Opreanu (2010) Diagnostic Test Accuracy Moderate
Convertino (2006) Experimental(simulation) Moderate
Rickards (2008) Experimental(simulation) Moderate
Ward (2010) Experimental(simulation) Moderate
Jaramillo (2010) Prospective Cohort Moderate
Luna (1989) Prospective Cohort Moderate
Robson (1989) Prospective Cohort Moderate
Vandromme (2011a) Prospective Cohort Moderate
HICK JL (2001) Retrospective Cohort Moderate
Vandromme (2011b) Retrospective Cohort Moderate
Baron (2004) Cross-Sectional Moderate
Brasel (2007) Cross-Sectional Moderate
Cancio (2008) Cross-Sectional Moderate
Parks (2006) Cross-Sectional Moderate
Zarzaur (2008) Cross-Sectional Moderate
Secher (1984) Experimental(simulation) Low
Scalea (1990) Prospective Cohort Low
Bruns (2008) Cross-Sectional Low
Chen (2008) Cross-Sectional Low
doi:10.1371/journal.pone.0057594.t002
Relationship between Blood Loss and ClinicalSigns
PLOS ONE |www.plosone.org 5 March 2013 |Volume 8 |Issue 3 | e57594

Anotherlimitation isthat differentmethodsof blood loss
measurementwere used acrossthe studies.Ideally,the use of
directmethods would be desirable in this kind ofresearch.The
majority of studies we reviewed used proxies to define the severity
of blood loss,which may introduce a considerable biasin the
analysis. A proxy measure for evaluating blood loss based on red-
celltransfusion is influenced by other factors,including provider
and patient behaviors and attitudes towards transfusion, as well as
the availability of blood at some hospitals,thus altering the blood
loss estimation.In addition, only few studiesdescribedthe
methodsusedto assessclinicalsigns.The use of different
techniquesto obtain data on blood pressure,heart hate,
respiratoryrate, pulsepressureand otherclinicaldata may
increase the heterogeneity ofresults.Furthermore,clinicalsigns
may also be affected by other factors such as the use ofcaffeine
and alcoholor even by labour per se which can enhance heart
rate,mean arterialpressure and cardiac outputduring contrac-
tions [18].The experimentalstudies controlled for such factors,
but we found no evidence of controlling for this potentialbias in
the observationalstudies reviewed.
In spite of theselimitations,the studiesincludedin this
systematic review did show an association between blood loss
and changesin clinicalsignsin the non-obstetricpopulation.
However,there is substantialvariation in the response ofclinical
signs to blood loss,which limits their applicability in diagnosing
haemorrhage orguiding itsmanagement.Guly and colleagues
[19]found an association between high heartrate,low systolic
blood pressure and the amount ofblood loss in seriously injured
patients but not to the degree suggested by the classification of th
American College ofSurgeonsin the Advanced Trauma Life
Support(ATLS) program [20].Other authorshave found that
tachycardia (defined as a HR over 90 bpm) is neither sensitive nor
specific for the diagnose of hypotension and amount of blood loss
[21]. For Braselet al. [22] tachycardia (defined as a pulse greater
Table 3. Association between blood loss and changes in vitalsigns.
Study HR SBP SI PSBP MAP DBP PP BT RR
Pregnancy related
population
Birkhahn (2002) N N N N
Birkhahn (2003) N # N
HICK JL (2001) # #
Jaramillo (2010) N
Robson (1989) N # #
Healthy subjects Birkhahn (2005) N N N #
Convertino (2006) N N # N N
Convertino (2009) N N N N
Rickards (2008) N N N #
Secher (1984) N N #
Ward (2010) N N
Trauma patients Baron (2004) N # N #
Scalea (1990) N *N *N
Brasel(2007) N
Bruns (2007) N N
Bruns (2008) # #
Cancio (2008) N N N N # #
Chen (2007) N N N # N #
Chen (2008) N
Edelman (2007) N
Guly (2011) N N #
Hagiwara (2010) N N N N N
Luna (1989) # # #
McLaughlin (2009) N N N N
Opreanu (2010) N N
Parks (2006) N
Vandromme (2010) N N
Vandromme (2011a) N
Vandromme (2011b) N N
Zarzaur (2008) N N N
BLE - Blood loss estimation; HR - Heart Rate; SBP - Systolic Bood Pressure; SI - Shock Index; PSBP - Prehospital Systolic Blood Pressure; MAP - Mean Arterial Pr
Diastolic Blood Pressure;PP - pulse pressure;BT - Body Temperature;RR - respiratory rate;
Nthere is an association between blood loss and changes in the vitalsign;#: there is no association between blood loss and changes in the vitalsign;
*N:no specification if systolic or diastolic.
doi:10.1371/journal.pone.0057594.t003
Relationship between Blood Loss and ClinicalSigns
PLOS ONE |www.plosone.org 6 March 2013 |Volume 8 |Issue 3 | e57594

than 100 bpm)had poor sensitivity and specificity (less than 37%
and 79% respectively)in identifyingsubstantialblood loss.
Although the ATLS classification system for hypovolemic shock
is widely used,the proposed cut-offvalues for clinicalsigns have
been challenged.SBP values that are higher than what is usually
considered as ‘‘hypotension’’have been associated with increased
morbidityand mortality[19,23].It has been suggested that
hypotension should be redefined using a highercut-offblood
pressure than actually used in the generalpopulation [23–26].
The physiologicalchanges in the cardiovascular system during
pregnancyand postpartum mayhinderearlyrecognition of
hypovolemia and delay treatment.In a first-trimester pregnancy
population, the correlation between vitalsigns and the amount of
blood in the peritonealcavity was shown to be poor.Hick and
colleagues [27] did notfind any association ofclinicalsigns and
hemoperitoneum.The authorsassume thatif surgicaldecisions
were made based on clinical signs, more than one third of patients
might be treated inappropriately. Another study using an obstetric
populationin late pregnancyfound similar data with no
correlation between blood loss and blood pressure [28].During
late pregnancy and the postpartum period,physiologicalchanges
in the cardiovascular system are even more substantial. In the cas
of PPH, some variables have been suggested to improve clinical
judgmentfor PPH treatment(e.g.clinicalsignsand symptoms,
visualestimation ofblood loss,and the blood loss rate)but none
have been sufficiently tested.Some authors suggest changing the
blood-lossbased definition ofPPH to a system ofsignsand
symptomsof hypovolemia.A hypovolemicshock classification
system was proposed using classes of hemorrhage correlating sig
and symptomsto the amountof blood lostand to a fluid
replacement procedure [8,11,29].According to this classification,
a compensated shock occurs with a blood loss of less than 1000 m
and no changeor slightchangein clinicalsigns.Substantial
changesin heartrate and blood pressure would be seen after
a blood loss ofmore than 1000 ml.Hypotension with significant
tachycardia and rise in respiratory rates would occur after a loss o
25–35% of blood volume and profound shock occurs after a 40%
blood loss. However, the use of clinical signs may lack accuracy in
Table 4. Clinicalsigns assessment and blood loss estimation method.
Study Clinical signs assessment Blood loss estimation method
Hick (2001) Automatic Aspiration of the abdominalcavity
Robson (1989) Automatic Clinically Estimation .500 ml
Convertino (2006) Automatic LBNP
Convertino (2009) Automatic LBNP
Rickards (2008) Automatic LBNP
Ward (2010) Automatic LBNP
Chen (2007) Automatic Requiring blood transfusion and bleeding trauma
Chen (2008) Automatic Requiring blood transfusion and bleeding trauma
Birkhahn (2002) Automatic Ruptured ectopic pregnancies
Secher (1984) Automatic Tilt-test
Birkhahn (2005) Manual Blood donation of 450 ml
Parks (2006) Non Available Base deficit estimation as a marker of shock
Baron (2004) Non Available Chest tube drainage,intraoperative blood loss,and radiographic evidence of
bleeding
Guly (2011) Non Available Clinicalestimation based on blood loss in specific injuries
Scalea (1990) Non Available Drainage of cavity OR Hct modification
Brasel (2007) Non Available Intervention to stop bleeding
Bruns (2007) Non Available Intervention to stop bleeding
Opreanu (2010) Non Available Intervention to stop bleeding
Bruns (2008) Non Available Requiring blood transfusion
Cancio (2008) Non Available Requiring blood transfusion
Edelman (2007) Non Available Requiring blood transfusion
McLaughlin (2009) Non Available Requiring blood transfusion
Vandromme (2011a) Non Available Requiring Blood Transfusion .10 Un/24 h
Vandromme (2011b) Non Available Requiring blood transfusion .10 un/24 h
Hagiwara (2010) Non Available Requiring blood transfusion . = 4 un
Luna (1989) Non Available Requiring blood transfusion . = 5 un
Zarzaur (2008) Non Available Requiring blood transfusion .4 un/48 h
Vandromme (2010) Non Available Requiring Blood Transfusion .6 Un/24 h
Birkhahn (2003) Non Available Ruptured EP and visualestimation of hemoperitoneum
Jaramillo (2010) Non Available Visualestimation of hemoperitoneum
VEBL – Visualestimation of blood loss;LBNP – Low body negative pressure.
doi:10.1371/journal.pone.0057594.t004
Relationship between Blood Loss and ClinicalSigns
PLOS ONE |www.plosone.org 7 March 2013 |Volume 8 |Issue 3 | e57594

the assessment of hypotension and needs further testing in order to
help guide the management of PPH.
Overall, our review findings suggest that blood loss is associated
with changes in clinicalsigns but it is difficult to establish robust
cut-offsthat could guidethe managementof womenwith
pregnancy-related haemorrhage.On the otherhand,when it
comes to a clinicalsign derivative – the shock index – our review
findings are more encouraging.The shock index is calculated as
the heart rate divided by the systolic blood pressure and this simple
calculation maytransform unstableparametersinto a more
accurate predictor of hypovolemia.According to studies included
in this review,the shock index may identify hypovolemia even in
patients who otherwise would be considered with no hypotension
[30,31].In addition,the shock index has been recently suggested
as a tool to predict mortality due to hypovolemic shock in trauma
patients. The use of the shock index in the early identification and
assessmentof bleeding is considered promising even in obstetric
populations[32]. Birkhahn and colleagues[33] studied first-
trimester pregnantwomen with abdominalpain and found that
a shock index .0.85 washighly suggestive ofthe presence of
hemoperitomeum dueto rupturedectopicpregnancy.This
parameter wasfound to be a better predictor ofbleeding than
HR or SBP only [34].Similar findings were obtained by other
authorssuggesting thatshock index may be a good criteria for
early diagnosis of haemorrhage [35].
Conclusion
This systematic review found a substantialvariability in the
relationship between blood loss and clinicalsigns,making it very
difficultto establish specific cut-offpointsfor clinicalsignsthat
could be used as triggers ofclinicalinterventions.However,the
shock index was found to be an accurate indicator ofcompen-
satory changesin the cardiovascularsystem due to blood loss.
Considering that most of the evidence included in this systematic
review is derived from studies in non-obstetric populations, furthe
studies on the use of the shock index in obstetric populations are
needed.
Supporting Information
Appendix S1 Preliminary Search Strategy in Medline.
(DOCX)
Checklist S1 PRISMA Checklist.
(DOC)
Acknowledgments
Disclaimer
The viewsexpressed aresolelythoseof the authorsand do not
necessarily reflectthe decisionsor stated policy ofthe World Health
Organization.
Author Contributions
Conceived and designed theexperiments:JPS RCP JB BW AMG.
Performed the experiments:JPS RCP. Analyzed the data:JPS RCP.
Contributed reagents/materials/analysis tools: JPS RCP. Wrote the paper:
JPS RCP JD PP JB BW AMG.
Table 5. Clinicalsigns statisticalanalysis.
Study Statistic HR SBP PSBP MAP DPBP PP SI cuttoff SI
Birkhahn (2002) AUC 0.74 0.7 0.63 0.84 0.85
Brasel (2007) AUC 0.56–0.59
Chen (2007) AUC 0.66 0.71 0.58 0.73 0.77
Opreanu (2010) AUC 0.59 0.56
Vandromme (2010) AUC 0.6 0.61
Vandromme (2011b) AUC 0.65 0.79
Zarzaur (2008) AUC 0.73 0.71 0.78 0.83
Range 0.56–0.74 0.56–0.79 0.61 0.63 0.58 0.73 0.77–0.84
Parks (2006) Pearson coef 0.28
Scalea (1990) Pearson coef 0.199 0.004* 0.005
Birkhahn (2003) Pearson coef 0.50 20.34 0.69 0.7
Hick (2001) R2 0.04 0.1
Jaramillo (2010) R2 0.57
Range 0.04–0.199 20.34–0.28 0.005 0.57–0.69
AUC –Area under curve;HR - Heart Rate;SBP - Systolic Bood Pressure;SI - Shock Index;PSBP - Prehospital Systolic Blood Pressure;MAP - Mean Arterial Pressure;DPBP -
Diastolic Blood Pressure;PP - pulse pressure;BT - Body Temperature;RR - respiratory rate;*BP – no specification if systolic or diastolic.
doi:10.1371/journal.pone.0057594.t005
Table 6. Clinicalsigns associated to mortality in included
studies.
Study Mortality
Zarzaur (2008) HR/SBP/SI
Bruns (2008) SBP
Cancio (2008) SBP
Edelman (2007) SBP
Luna (1989) SBP
Parks (2006) SBP
Vandromme (2010) SBP
Victorino (2003) SBP+TACHYCARDIA
doi:10.1371/journal.pone.0057594.t006
Relationship between Blood Loss and ClinicalSigns
PLOS ONE |www.plosone.org 8 March 2013 |Volume 8 |Issue 3 | e57594

References
1. Khan KS, Wojdyla D,Say L,Gu¨lmezoglu AM,Van Look PFA (2006)WHO
analysis ofcauses ofmaternaldeath:a systematic review.Lancet367:1066–
1074.
2. GanatraBR, Coyaji KJ, Rao VN (1998) Too far, too little, too late:
a community-based case-controlstudy ofmaternalmortality in ruralwest
Maharashtra,India.Bulletin of the World Health Organization 76:591–598.
3. Maine D (1991)Safe Motherhood Programs:Options and Issues.New York:
Center for Population and Family Health,Columbia University.
4. World Health Organization (1990)The Prevention and Managementof
Postpartum Haemorrhage.Reportof a TechnicalWorking Group,Geneva,
3–6 July,1989.Geneva.
5. Schorn MN (2010) Measurement of blood loss: review of the literature. Journal
of midwifery & women’s health 55:20–27.
6. Stafford I,Dildy GA, Clark SL, BelfortMA (2008)Visually estimated and
calculated blood lossin vaginaland cesarean delivery.American journalof
obstetrics and gynecology 199:519.e1–7.
7. Quin˜ onesJN, Uxer JB, Gogle J, Scorza WE, Smulian JC (2010)Clinical
evaluation during postpartum hemorrhage.Clinicalobstetrics and gynecology
53:157–164.
8. Coker A,Oliver R (2006)Definitions and classifications.In: B-Lynch C,Keith
LG, Lalonde AB, Karoshi M, editors. A Textbook of Postpartum Hemorrhage.
Duncow,UK: Sapiens Publishing.11–16.
9. Mousa HA, Alfirevic Z (2007) Treatment for primary postpartum haemorrhage.
Cochranedatabaseof systematicreviews(Online):CD003249.Available:
http://www.ncbi.nlm.nih.gov/pubmed/17253486.Accessed14 December
2011.
10. Baskett PJ (1990)ABC of major trauma.Management of hypovolaemic shock.
BMJ (Clinicalresearch ed)300:1453–1457.
11. BonnarJ (2000)Massive obstetric haemorrhage.BestPractice & Research
ClinicalObstetrics & Gynaecology 14:1–18.
12. Moher D, Liberati A, Tetzlaff J, Altman DG (2009) Preferred reporting items for
systematic reviews and meta-analyses:the PRISMA statement.PLoS medicine
6: e1000097.
13. von Elm E,Altman DG,Egger M,Pocock SJ,Gøtzsche PC,et al.(2007)The
Strengtheningthe Reportingof ObservationalStudiesin Epidemiology
(STROBE) statement:guidelinesfor reporting observationalstudies.Bulletin
of the World Health Organization 85:867–872.
14. Ueland K (1976)Maternalcardiovascular dynamics.VII. Intrapartum blood
volume changes.American journalof obstetrics and gynecology 126:671–677.
15. Pritchard JA (1965)Changesin the blood volumeduring pregnancy and
delivery.Anesthesiology 26:393.
16. World Health Organization (2008)Worldwide prevalence ofanaemia 1993–
2005: WHO global database on anaemia. Geneva: World Health Organization.
17. Lee AI, Okam MM (2011) Anemia in pregnancy. Hematology/oncology clinics
of North America 25:241–59,vii.
18. Robson SC,Dunlop W, BoysRJ, HunterS (1987)Cardiac outputduring
labour.British medicaljournal(Clinicalresearch ed)295:1169–1172.
19. Guly HR, Bouamra O, Spiers M, Dark P, Coats T, et al. (2011) Vital signs and
estimated blood loss in patients with major trauma:testing the validity ofthe
ATLS classification of hypovolaemic shock.Resuscitation 82:556–559.
20. American College of Surgeons Trauma Committee (2008) Advanced trauma life
support for doctors.8th ed.Chicago:American College of Surgeons.
21. Victorino GP, Battistella FD, Wisner DH (2003) Does tachycardia correlate with
hypotension after trauma? Journalof the American College ofSurgeons 196:
679–684.
22. Brasel KJ, Guse C, Gentilello LM, Nirula R (2007) Heart rate: is it truly a vital
sign? The Journalof trauma 62:812–817.
23. Eastridge BJ,Salinas J,McManus JG,Blackburn L,Bugler EM,et al. (2007)
Hypotension begins at110 mm Hg:redefining ‘‘hypotension’’with data.The
Journalof trauma 63:291–7;discussion 297–9.
24. Mcgee S,Abernethy III WB,SimelDL (1999)Is This PatientHypovolemic?
JAMA. 281:1022–1029.
25. EdwardsM, Ley E, Mirocha J,HadjibashiAA, MarguliesDR, et al. (2010)
Defining hypotension in moderate to severely injured trauma patients:raising
the bar for the elderly.The American surgeon 76:1035–1038.
26. Guly HR, Bouamra O,Little R, Dark P, CoatsT, et al. (2010)Testing the
validity ofthe ATLS classification ofhypovolaemic shock.Resuscitation 81:
1142–1147.
27. Hick JL, Rodgerson JD,Heegaard WG,SternerS (2001)Vital signsfail to
correlatewith hemoperitoneum from rupturedectopicpregnancy.The
American journalof emergency medicine 19:488–491.
28. Robson SC,BoysRJ, Hunter S,Dunlop W (1989)Maternalhemodynamics
afternormaldelivery and delivery complicated by postpartum hemorrhage.
Obstetrics and gynecology 74:234–239.
29. BenedettiT (2002)Obstetric haemorrhage.In: Gabbe S,NiebylJ, Simpson J,
editors.A Pocket Companion to Obstetrics.New York:ChurchillLivingstone.
217–248.
30. Vandromme MJ,Griffin RL, Kerby JD, McGwin G, Rue LW, et al. (2011)
Identifying risk for massive transfusion in the relatively normotensive patient:
utility of the prehospitalshock index.The Journal of trauma 70:384–8;
discussion 388–90.
31. Birkhahn R,Gaeta T,Terry D,Bove J,TloczkowskiJ (2005)Shock index in
diagnosing early acutehypovolemia.The American Journalof Emergency
Medicine 23:323–326.
32. Cannon CM, Braxton CC,Kling-Smith M,Mahnken JD,Carlton E,et al.
(2009) Utility of the shock index in predicting mortality in traumatically injured
patients.The Journalof trauma 67:1426–1430.
33. Birkhahn RH, Gaeta TJ, Bei R, Bove JJ (2002) Shock index in the first trimester
of pregnancy and itsrelationship to ruptured ectopic pregnancy.Academic
emergency medicine:officialjournalof the Society for Academic Emergency
Medicine 9:115–119.
34. Birkhahn RH,Gaeta TJ,Van Deusen SK,TloczkowskiJ (2003)The ability of
traditionalvitalsigns and shock index to identify ruptured ectopic pregnancy.
American Journalof Obstetrics and Gynecology 189:1293–1296.
35. Jaramillo S,BarnhartK, Takacs P (2011)Use ofthe shock index to predict
rupturedectopicpregnancies.Internationaljournal of gynaecologyand
obstetrics:the officialorgan ofthe InternationalFederation ofGynaecology
and Obstetrics 112:68.
36. Baron BJ,SinertR, ZehtabchiS, Stavile KL,Scalea TM (2004)Diagnostic
Utility ofSublingualPco2 for Detecting Hemorrhage in Penetrating Trauma
Patients. The Journal of Trauma: Injury, Infection, and Critical Care 57: 69–74.
37. Scalea TM,HartnettRW, Duncan AO,Atweh NA,Phillips TF,et al. (1990)
Central venous oxygen saturation: a usefulclinicaltoolin trauma patients. The
Journalof trauma 30:1539–1543.
38. Bruns B,Lindsey M,Rowe K, Brown S,Minei JP, et al. (2007)Hemoglobin
drops within minutes ofinjuries and predicts need for an intervention to stop
hemorrhage.The Journalof trauma 63:312–315.
39. Bruns B, GentilelloL, Elliott A, Shafi S (2008)Prehospitalhypotension
redefined.The Journalof trauma 65:1217–1221.
40. Cancio LC, Wade CE, West SA, Holcomb JB (2008) Prediction of mortality and
of the need formassive transfusion in casualtiesarriving atcombatsupport
hospitals in Iraq.The Journalof trauma 64:S51–5;discussion S55–6.
41. Chen L, Reisner AT, McKenna TM, Gribok A, Reifman J (2007) Diagnosis of
hemorrhage in a prehospitaltrauma population using linearand nonlinear
multiparameteranalysisof vital signs.Conferenceproceedings:Annual
InternationalConference ofthe IEEE Engineering in Medicine and Biology
Society IEEE Engineering in Medicine and Biology Society Conference 2007:
3748–3751.
42. Chen L, Gribok A,Reisner AT,Reifman J (2008)Exploiting the existence of
temporalheart-rate patterns for the detection oftrauma-induced hemorrhage.
Conferenceproceedings:Annual InternationalConferenceof the IEEE
Engineering in Medicine and Biology Society IEEE Engineering in Medicine
and Biology Society Conference 2008:2865–2868.
43. Edelman DA,White MT, TyburskiJG, Wilson RF (2007)Post-traumatic
hypotension:should systolic blood pressure of90–109 mmHg be included?
Shock (Augusta,Ga)27:134–138.
44. Hagiwara A,Kimura A, Kato H, Mizushima Y,Matsuoka T,et al. (2010)
Hemodynamic reactions in patients with hemorrhagic shock from blunt trauma
after initialfluid therapy.The Journalof trauma 69:1161–1168.
45. Luna GK, Eddy AC, Copass M (1989) The sensitivity of vital signs in identifying
major thoracoabdominalhemorrhage.American journalof surgery 157:512–
515.
46. McLaughlin DF,NilesSE, SalinasJ, PerkinsJG, Cox ED, et al. (2008)A
predictivemodelfor massivetransfusion in combatcasualty patients.The
Journalof trauma 64:S57–63;discussion S63.
47. Opreanu RC, Arrangoiz R, Stevens P, Morrison CA, Mosher BD, et al. (2010)
Hematocrit, systolic blood pressure and heart rate are not accurate predictors for
surgery to controlhemorrhage in injured patients.The American surgeon 76:
296–301.
48. Parks JK, Elliott AC, Gentilello LM, Shafi S (2006) Systemic hypotension is a late
marker ofshock after trauma:a validation study ofAdvanced Trauma Life
Support principles in a large national sample. American journal of surgery 192:
727–731.
49. Vandromme MJ, Griffin RL, Weinberg JA, Rue LW, Kerby JD (2010) Lactate is
a betterpredictorthan systolicblood pressurefor determiningblood
requirement and mortality:could prehospitalmeasures improve trauma triage?
Journalof the American College of Surgeons 210:861–867,867–869.
50. Vandromme MJ, Griffin RL, McGwin G, Weinberg JA, Rue LW, et al. (2011)
Prospective identification of patients at risk for massive transfusion: an imprecise
endeavor.The American surgeon 77:155–161.
51. Zarzaur BL, Croce MA, Fischer PE, Magnotti LJ, Fabian TC (2008) New vitals
after injury:shock index for the young and age x shock index for the old.The
Journalof surgicalresearch 147:229–236.
52. Convertino VA, Cooke WH, Holcomb JB (2006) Arterial pulse pressure and its
association with reduced stroke volume during progressive central hypovolemia.
The Journalof trauma 61:629–634.
53. Convertino VA, Rickards CA, Lurie KG, Ryan KL (2009)Hyperventilation in
response to progressive reduction in centralblood volume to nearsyncope.
Aviation,space,and environmentalmedicine 80:1012–1017.
54. RickardsCA, Ryan KL, Cooke WH, Romero SA,Convertino VA (2008)
Combatstressor hemorrhage?Evidence fora decision-assistalgorithm for
remote triage.Aviation,space,and environmentalmedicine 79:670–676.
Relationship between Blood Loss and ClinicalSigns
PLOS ONE |www.plosone.org 9 March 2013 |Volume 8 |Issue 3 | e57594

55. Secher NH, Sander Jensen K, Werner C, Warberg J, Bie P (1984) Bradycardia
during severe but reversible hypovolemic shock in man.Circulatory shock 14:
267–274.
56. Ward KR, Tiba MH, Ryan KL, Filho IPT, Rickards CA, et al. (2010) Oxygen
transportcharacterization ofa human modelof progressivehemorrhage.
Resuscitation 81:987–993.
Relationship between Blood Loss and ClinicalSigns
PLOS ONE |www.plosone.org 10 March 2013 |Volume 8 |Issue 3 | e57594