HEMORRAGIA POSPARTO E INDICE DE SHOCK

1. Accepted
Article
This article has been accepted for publication and undergone full peer review but has not
been through the copyediting, typesetting, pagination and proofreading process, which may
lead to differences between this version and the Version of Record. Please cite this article as
doi: 10.1002/ijgo.12343
This article is protected by copyright. All rights reserved.
Article Type: Clinical Article
Case–control study of shock index among women who did and did not receive blood
transfusions due to postpartum hemorrhage
Anderson Borovac-Pinheiro, Rodolfo C. Pacagnella *, Carolina Puzzi-Fernandes, José G.
Cecatti
Department of Obstetrics and Gynaecology, School of Medical Sciences – Women’s
Hospital, University of Campinas, Campinas, Brazil
* Correspondence
Rodolfo C Pacagnella, University of Campinas, Rua Alexander Fleming, nº 101. Cidade
Universitária Zeferino Vaz , Campinas, São Paulo, 13083-881, Brazil.
Email: rodolfo@caism.unicamp.br
Keywords: Blood transfusion; Postpartum hemorrhage; Shock index
Synopsis: Shock index values measured after delivery could be used to identify women at
risk of severe postpartum hemorrhage and blood transfusion after vaginal delivery.

2. Accepted
Article
This article is protected by copyright. All rights reserved.
Abstract
Objective: To compare shock index (SI) values between women who required blood
transfusion due to postpartum hemorrhage (PPH) and women who did not.
Methods: In a case–control study, clinical data were assessed from the medical records of
women requiring blood transfusion for PPH at a center in Brazil between 2012 and 2015
(n=105). A control group was randomly selected from women who did not receive blood
transfusion (n=129).
Results: Compared with women who did not receive a transfusion after delivery, women
who did receive one had significantly higher SI values 10 minutes after delivery (0.81 ± 0.27
vs 0.72 ± 0.16; P=0.012), at 30 minutes (0.83 ± 0.26 vs 0.71 ± 0.15; P<0.001), and at
2 hours (0.84 ± 0.27 vs 0.70 ± 0.14; P=0.032). For vaginal deliveries, SI values were
significantly different at 30 minutes (0.88 ± 0.26 vs 0.71 ± 0.14; P<0.001) and 2 hours
(0.90 ± 0.23 vs 0.72 ± 0.14; P=0.001). No significant differences were found for cesarean
delivery.
Conclusion: The SI might be useful to identify early vital sign changes due to PPH.
Increased SI values were associated with need for transfusion in vaginal deliveries.
1 INTRODUCTION
Reducing maternal mortality is a global concern. Hemorrhage is the leading direct cause of
maternal mortality worldwide, followed by hypertension and sepsis [1]. In addition to being
an important cause of maternal mortality, postpartum hemorrhage (PPH) plays a significant
part in severe maternal morbidity and maternal near miss [2,3]. Furthermore, there is a
fourfold higher incidence of PPH-related maternal near miss as compared with PPH-related
maternal death [4].

3. Accepted
Article
This article is protected by copyright. All rights reserved.
As a result, it is important to assess why and how PPH occurs, and to identify better and
more effective methods for diagnosis and treatment that might be applied globally. The
current criterion for PPH diagnosis is an estimated blood loss of more than 500 mL. WHO
suggests visual estimation to identify PPH, but some studies have shown that this method
has low accuracy [5–7]. An ideal diagnostic method for PPH should adopt features that
guarantee its standardization, including simplicity, ease of application in daily practice, and
prioritization of both the lost blood volume and its clinical consequences [8].
Considering these points, there remains an effort to establish the best way of identifying
obstetric hemorrhage with clinical repercussions. A promising alternative is the shock index
(SI), which is the simple ratio of heart rate divided by systolic blood pressure [9]. The SI has
been used as a valuable tool to assist in diagnosis, guide treatment, and predict the
probability of death among patients with blood loss due to trauma [10–12].
The SI might be an important tool applicable to the early diagnosis of PPH. An increase in SI
has been related to severe maternal outcomes and maternal death [13,14], requirement for
massive transfusion [15], massive PPH [16], low fibrinogen levels, and transfusion of blood
volume and fresh frozen plasma [17]. In most studies, a cutoff point of 0.9 for the SI value
has been found to be associated with PPH and increased risk of maternal death, end-organ
failure, maternal morbidity, and critical interventions such as intensive care unit (ICU)
admission and massive transfusion [13,14,16].
The aim of the present study was to assess whether SI values vary significantly between
women who require blood transfusion in the postpartum period and those who do not require
transfusion, and to identify factors associated with the need for blood transfusion due to
PPH.

4. Accepted
Article
This article is protected by copyright. All rights reserved.
2 MATERIALS AND METHODS
In a retrospective case–control study, data were reviewed from pregnant women
admitted to the University of Campinas Women’s Hospital, Brazil, between January
1, 2012, and December 31, 2015. The Institutional Review Board (CAAE no.
42716814.5.0000.5404) approved the study protocol before the study begun. The
need for informed consent was waived given that the data were collected
retrospectively from clinical records and the women were not identified.
On the basis of a power of 90% and assuming that 3% of controls would have
increased SI values without clinical meaning [18], a sample size of 180 women (90
cases and 90 controls) was calculated. A larger number of controls were included to
compensate for missing data. The institutional electronic information system was
used to identify women who required blood transfusion due to PPH (study group).
Women who received blood products for other problems were not included. Control
women who delivered in the same period and required no blood transfusion were
identified in the electronic information system and randomly selected by assigning
random numbers in Excel 2013 (Microsoft, Redmond, WA, USA) and choosing
numbers from 1 onwards.
Clinical and obstetric data were obtained from medical records via a specific data
collection form. Cases and controls were classified by type of delivery (vaginal or
cesarean) and length of pregnancy (21–27 weeks, 28–32 weeks, 33–36 weeks, and
≥37 weeks). Vital signs were recorded at delivery and at 10, 30, and 120 minutes
after delivery.

5. Accepted
Article
This article is protected by copyright. All rights reserved.
Data were entered in an online platform and statistical analyses were performed with SAS
version 9.4 (SAS Institute, Cary, NC, USA) and SPSS version 20.0 (IBM, Armonk, NY,
USA). For the analysis, clinical and obstetric characteristics were compared between the
study group and control group with 2
, Student t, and Fisher exact tests used to assess
differences. Mean ± SD SI values were compared at delivery and at 10, 30, and 120 minutes
after delivery via the Mann–Whitney U test. A Spearman correlation between SI and the
number of blood units transfused was also performed.
Multivariate analysis of covariance was used to identify the risk of transfusion depending on
pregnancy length, body mass index, SI, anesthesia, type of delivery, and previous
morbidities. Only women with complete data were included in this analysis. A cutoff SI value
of 0.83 was used on the basis of the highest mean SI values in a previous study that
assessed standard reference values for SI during pregnancy [19]. P<0.05 was taken as
statistically significant. Using this cutoff point in a 2×2 table, sensitivity, specificity and
positive and negative likelihood ratios were calculated.
3 RESULTS
During the study period, 105 women needed blood transfusion after delivery owing to
PPH and formed the study group; 129 women who did not need blood transfusion
were randomly selected as the control group. In all PPH cases, the transfusion
occurred at 2 hours after delivery, meaning that it did not influence SI values.
The clinical and obstetric characteristics of the participants are presented in Table 1. There
were no differences between the study and control groups in these characteristics.

6. Accepted
Article
This article is protected by copyright. All rights reserved.
The mean SI values at delivery, and 10, 30, and 120 minutes after delivery were compared
between the study and control groups (Table 2). Among all deliveries (vaginal and
cesarean), mean SI values were significantly higher at 10 minutes, 30 minutes, and 2 hours
after delivery for women who received a blood transfusion than for control women (P=0.012,
P<0.001, and P=0.032, respectively). Values were also significantly higher in the study
group at 30 minutes and 2 hours after delivery among vaginal deliveries only (P<0.001 and
P=0.001, respectively). There were no significant differences in SI between the groups for
women who delivered by cesarean. The sample size obtained had a power of 97% to
identify differences between the two groups at 30 minutes and 2 hours after delivery.
In the study group, 37 (35.2%) of the 105 women received a transfusion of one blood unit,
53 (50.5%) received two blood units, and 15 (14.3%) received three or more blood units.
One woman received more than 10 units of blood. The Spearman correlation coefficient
showed a very weak correlation between SI and the number of blood units transfused
(during delivery: r=0.03; 10 minutes after delivery: r=0.08; 30 minutes after delivery: r=0.05;
2 hours after delivery: r=0.10) (data not shown).
Multivariate analyses to assess the risk of needing blood transfusion due to PPH showed
that the odds were not influenced by maternal age, pregnancy length, body mass index, or
mode of delivery (Table 3). By contrast, the presence of morbidities and an SI above 0.83 at
30 minutes after delivery increased the need for blood transfusion after PPH significantly
(P=0.001 and P=0.002, respectively). The sensitivity, specificity, and likelihood ratio of SI
values above 0.83 to predict PPH are presented in Table 4.

7. Accepted
Article
This article is protected by copyright. All rights reserved.
4 DISCUSSION
The present study found that SI values after delivery can identify women at risk of PPH: the
values were higher among women who underwent blood transfusion for PPH than among
those who did not. When the sample was stratified according to type of delivery, however,
the differences remained significant only for women who delivered vaginally.
At all timepoints, the mean SI was below 0.8 in the control group and below 0.9 in the study
group. The current SI values agree with those of other studies. A study on the relationship
between SI and massive PPH (defined as ≥30% loss of blood volume at delivery) [16] found
that the mean SI value was greater than 0.9 at 10 and 30 minutes after delivery. For women
who did not have massive transfusion, the mean SI value was 0.74 at 10 minutes after
delivery and 0.76 at 30 minutes after delivery. Another study of 26 patients who required
massive transfusion (defined as transfusion of ≥10 units of packed red blood cells within
24 hours of admission) due to primary PPH [15] found that SI and heart rate were
significantly associated with massive transfusion. The odds ratio (OR) for SI was 9.47 (95%
confidence interval [CI] 1.75–51.28), whereas that for heart rate was 1.06 (95% CI 1.02–
1.09). Overall, 67% of women with an SI value of 1.3 or higher required massive transfusion
[15].
The slight differences between the previous findings and the present data might be
explained by differences between the women studied: other studies focused only on women
with massive PPH, whereas the present study evaluated women who needed transfusion of
any amount of blood. In the present sample, at 30 minutes after delivery, 31.4% had an SI of
0.83 or higher, whereas 68.6% had an SI below 0.83. Elevated SI was shown to be
associated with the need for blood transfusion of any volume, which is important when
considering SI as an instrument for the early identification of women at risk of complications
due to PPH.

8. Accepted
Article
This article is protected by copyright. All rights reserved.
The current study showed that, by contrast with previous research, SI values can be used to
predict the need for transfusion of any amount of blood derivatives. Furthermore, in the
study group, almost all SI values after delivery were higher than 0.8, except for those among
women who delivered by cesarean; thus, SI values higher than 0.8 might indicate the
severity of the case and that specific treatment should be offered, including the need for
blood transfusion due to PPH. The present study also showed that women with SI values
above 0.83 at 30 minutes were five times more likely to require blood transfusion.
In addition to being a possible indicator for blood transfusion, SI might also identify women
with severe maternal morbidity requiring ICU admission. A study of 233 postpartum women
with PPH and blood loss greater than 1500 mL [13] found that an SI value of 0.9 or higher
more accurately identified women who required ICU admission after delivery and those who
received more than four units of blood products. This threshold for SI (0.9) has also been
identified as an important cutoff for predicting adverse maternal outcome [14]. After delivery,
values of SI greater than 0.9 had 100%, 95.2%, and 94% sensitivity in predicting maternal
death, severe maternal outcomes (death or severe end-organ failure maternal morbidity),
and critical intervention (ICU admission, blood transfusion >5 units, or emergency
hysterectomy), respectively [14].
It is clinically important to predict the need for transfusion as early as possible. The
management of PPH should be started as soon as PPH is suspected, ideally within the first
hour of delivery and preferably less than 30 minutes after delivery. Therefore, the present
study focused on the diagnosis of PPH before this period to enable prompt handling. It
seems that an SI value above 0.8 indicates that a woman might need blood transfusion or at
least rigorous surveillance, whereas an SI greater than 0.9 indicates a potentially severe
maternal outcome or the need for ICU admission. Previous studies of SI values in the

9. Accepted
Article
This article is protected by copyright. All rights reserved.
postpartum period have proposed different cutoff points to trigger timely action;
nevertheless, they all report an increased risk of adverse maternal outcomes with increasing
SI after delivery. The optimal thresholds and timepoints to measure SI remain unclear and
should be validated in prospective studies.
A previous study established standard reference values for SI during low-risk pregnancy,
reporting mean SI values from 0.75 to 0.83 [19]. The mean values increased during
pregnancy and decreased toward the end of pregnancy: for women at more than 37 weeks,
the mean SI was 0.79 [19]. The present study found lower mean SI values in the postpartum
period. One possible explanation is the physiological changes that occur during the
postpartum period: for example, there is a sustained increase in cardiac output for 1 hour
after delivery owing to the relief of inferior vena cava obstruction and the transfer of
extravascular fluid into the intravascular space. Cardiac output and peripheral vascular
resistance directly influence blood pressure. Because SI is the ratio of heart rate to systolic
blood pressure, it is inversely proportional to an increase in cardiac output; thus, it is
expected that SI values during the postpartum period would be lower than those at the end
of the pregnancy [20,21].
To the best of our knowledge, the present study is the first to evaluate SI values during
cesarean delivery. There was no difference in SI values between the study group and the
control group who underwent this type of delivery. One possible explanation is that SI values
are sensitive to small changes in their components (heart rate and systolic blood pressure)
and, during cesarean delivery, the anesthesiologists have more strict control over vital signs,
which would mask the increase in SI in women with PPH.

10. Accepted
Article
This article is protected by copyright. All rights reserved.
The present study has some limitations including its retrospective design and use of medical
records with some missing information. Nevertheless, the results show that SI might be used
as an additional tool to identify women at risk of blood transfusion due to PPH. However,
cutoff points will need to be validated in prospective controlled studies to ensure better-
standardized collection of data and blood-loss measurement to clarify the applicability of SI
values to the obstetric population.
The present findings might contribute to a rethinking of the diagnostic criteria for PPH. Do all
women with an estimated blood loss of 500 mL need intervention, irrespective of any clinical
aspects? SI might help us to find the answer.
Author contributions
AB-P, RCP, and JGC conceived and designed the study. AB-P and CP-F acquired the data.
All authors were involved in data analysis and interpretation, drafted the manuscript, and
revised and approved the final version of the manuscript.
Acknowledgments
WHO and FAEPEX (Fundo de Apoio ao Ensino, Pesquisa e Extensão)–UNICAMP funded
the study.
Conflicts of interest
The authors have no conflicts of interest.

11. Accepted
Article
This article is protected by copyright. All rights reserved.
References
1. Say L, Chou D, Gemmill A, et al. Global causes of maternal death: a WHO
systematic analysis. Lancet Glob Heal. 2014 Jun;2(6):e323-33.
2. Bateman BT, Berman MF, Riley LE, Leffert LR. The epidemiology of postpartum
hemorrhage in a large, nationwide sample of deliveries. Anesth Analg. 2010 May
1;110(5):1368–73.
3. Souza JP, Gülmezoglu AM, Vogel J, et al. Moving beyond essential interventions for
reduction of maternal mortality (the WHO Multicountry Survey on Maternal and Newborn
Health): a cross-sectional study. Lancet (London, England). 2013 May 18;381(9879):1747–
55.
4. Sheldon WR, Blum J, Vogel JP, et al. Postpartum haemorrhage management, risks,
and maternal outcomes: findings from the World Health Organization Multicountry Survey on
Maternal and Newborn Health. BJOG. 2014;121 Suppl:5–13.
5. Tunçalp Ö, Souza JP, Gülmezoglu M. New WHO recommendations on prevention
and treatment of postpartum hemorrhage. Int J Gynecol Obstet. 2013;123(3):254–256.
6. Toledo P, McCarthy RJ, Hewlett BJ, Fitzgerald PC, Wong CA. The accuracy of blood
loss estimation after simulated vaginal delivery. Anesth Analg. 2007 Dec;105(6):1736–40,
table of contents.
7. Patel A, Goudar SS, Geller SE, et al. Drape estimation vs. visual assessment for
estimating postpartum hemorrhage. Int J Gynecol Obstet. 2006 Jun;93(3):220–224.
8. World Health Organization. WHO recommendations for the prevention and treatment
of postpartum haemorrhage: evidence base. WHO Guidelines Approved by the Guidelines
Review Committee. Geneva: WHO Library; 2012. 48 p.
9. Pacagnella RC, Souza JP, Durocher J, et al. A Systematic Review of the

12. Accepted
Article
This article is protected by copyright. All rights reserved.
Relationship between Blood Loss and Clinical Signs. PLoS One. 2013 Jan;8(3):e57594.
10. Mutschler M, Nienaber U, Münzberg M, et al. The Shock Index revisited - a fast
guide to transfusion requirement? A retrospective analysis on 21,853 patients derived from
the TraumaRegister DGU®. Crit Care. 2013 Aug 12;17(4):R172.
11. Olaussen A, Blackburn T, Mitra B, Fitzgerald M. Review article: shock index for
prediction of critical bleeding post-trauma: a systematic review. Emerg Med Australas. 2014
Jun;26(3):223–8.
12. Sloan EP, Koenigsberg M, Clark JM, Weir WB, Philbin N. Shock index and prediction
of traumatic hemorrhagic shock 28-day mortality: data from the DCLHb resuscitation clinical
trials. West J Emerg Med. 2014 Nov;15(7):795–802.
13. Nathan H, El Ayadi A, Hezelgrave N, et al. Shock index: an effective predictor of
outcome in postpartum haemorrhage? BJOG An Int J Obstet Gynaecol. 2015 Jan
26;122(2):268–75.
14. Ayadi AM El, Nathan HL, Seed PT, et al. Vital sign prediction of adverse maternal
outcomes in women with hypovolemic shock: The role of shock index. PLoS One.
2016;11(2):1–12.
15. Sohn CH, Kim WY, Kim SR, et al. An increase in initial shock index is associated
with the requirement for massive transfusion in emergency department patients with primary
postpartum hemorrhage. Shock. 2013;40(2):101–5.
16. Le Bas A, Chandraharan E, Addei A, Arulkumaran S. Use of the “obstetric shock
index” as an adjunct in identifying significant blood loss in patients with massive postpartum
hemorrhage. Int J Gynecol Obstet. 2014;124(3):253–255.
17. Era S, Matsunaga S, Matsumura H, Murayama Y, Takai Y, Seki H. Usefulness of
shock indicators for determining the need for blood transfusion after massive obstetric

13. Accepted
Article
This article is protected by copyright. All rights reserved.
hemorrhage. J Obstet Gynaecol Res. 2015 Jan;41(1):39–43.
18. Rappaport LD, Deakyne S, Carcillo JA, McFann K, Sills MR. Age- and sex-specific
normal values for shock index in National Health and Nutrition Examination Survey 1999-
2008 for ages 8 years and older. Am J Emerg Med. 2013 May;31(5):838–42.
19. Borovac-Pinheiro A, Pacagnella RC, Morais SS, Cecatti JG. Standard reference
values for the shock index during pregnancy. Int J Gynecol Obstet. 2016;135(1):11–15.
20. Tan EK, Tan EL. Alterations in physiology and anatomy during pregnancy. Best Pract
Res Clin Obstet Gynaecol. 2013 Dec;27(6):791–802.
21. Ouzounian JG, Elkayam U. Physiologic changes during normal pregnancy and
delivery. Cardiol Clin 2012 Aug;30(3):317–29.

14. Accepted
Article
This article is protected by copyright. All rights reserved.
Table 1 Clinical and obstetric characteristics by group.
a
Characteristics Study group (n=105) Control group (n=129) P value
Age, y 0.568
b
<19 12 (11.4) 18 (14.0)
19–35 81 (77.1) 101 (78.3)
>35 12 (11.5) 10 (7.7)
Pregnancy length, wk 37.7 ± 3.3 37.6 ± 3.5 0.719
c
Body mass index
d
29.4 ± 6.7 30.1 ± 6.8 0.509
c
Anesthesia 0.181
b
No 17 (16.2) 25 (19.3)
Yes 88 (83.8) 104 (80.7)
Parity 0.685
b
0 53 (50.4) 61 (47.3)
1 32 (30.4) 37 (28.7)
≥2 20 (19.2) 31 (24.0)
Onset of labor
e
0.965
b
Spontaneous 61 (76.3) 75 (76.5)
Induced 19 (23.8) 23 (23.5)
Delivery 0.436
b
Vaginal 55 (52) 74 (57.3)
Cesarean 50 (48) 55 (42.7)
Previous comorbidities
f
0.67
g
Yes 20 (20.2) 6 (4.8)
No 79 (79.8) 120 (95.2)
a
Values are given as number (percentage) or mean ± SD, unless indicated otherwise.
b
By 
2
test.
c
By t test.
d
Calculated as weight in kilograms divided by the square of height in meters. Data missing for 64 (22
in study group; 42 in control group).

15. Accepted
Article
This article is protected by copyright. All rights reserved.
e
Data missing for 56 (25 in study group; 31 in control group).
f
Data missing for 9 (6 in study group; 3 in control group).
g
By Fisher exact test.
Table 2 Shock index values during the postpartum period by mode of delivery.
a
Delivery mode and
timepoint
Study group Control group P
value
b
No. of
women
Shock index No. of
women
Shock index
All delivery modes
Delivery 84 0.75 ± 0.22 105 0.73 ± 0.17 0.830
10 min after delivery 88 0.81 ± 0.27 106 0.72 ± 0.16 0.012
30 min after delivery 90 0.83 ± 0.26 123 0.71 ± 0.15 <0.001
2 h after delivery 96 0.84 ± 0.27 21 0.70 ± 0.14 0.032
Vaginal delivery
Delivery 36 0.76 ± 0.17 50 0.76 ± 0.15 0.611
10 min after delivery 39 0.82 ± 0.30 51 0.73 ± 0.15 0.060
30 min after delivery 46 0.88 ± 0.26 68 0.71 ± 0.14 <0.001
2 h after delivery 52 0.90 ± 0.23 17 0.72 ± 0.14 0.001
Cesarean
Delivery 47 0.74 ± 0.25 54 0.71 ± 0.18 0.675
10 min after delivery 48 0.80 ±0.25 54 0.71 ± 0.16 0.078
30 min after delivery 43 0.79 ± 0.25 54 0.71 ± 0.16 0.204
2 hours after delivery 43 0.76 ± 0.30 4 0.63 ± 0.10 0.598
a
Values are given as mean ± SD unless indicated otherwise.
b
By Mann–Whitney U test.

16. Accepted
Article
This article is protected by copyright. All rights reserved.
Table 3 Multivariate analysis for factors associated with risk of postpartum hemorrhage needing
transfusion.
a
Factor Odds ratio (95% confidence interval) P value
Maternal age, y
<19 0.57 (0.20–1.60) 0.285
19–35 Ref.
>35 0.59 (0.16–2.23) 0.438
Pregnancy length, wk
21–27 0.47 (0.05–4.52) 0.517
28–32 0.39 (0.03–4.89) 0.468
33–36 1.60 (0.40–6.48) 0.509
>37 Ref.
Body mass index (continuous) 0.97 (0.91–1.04) 0.381
Anesthesia
Yes 0.28 (0.02–4.93) 0.384
No Ref.
Delivery mode
Vaginal Ref.
Cesarean 0.57 (0.248–1.30) 0.179
Morbidity
Yes 6.40 (2.171–18.88) 0.001
No Ref.
SI at delivery
≥0.83 0.83 (0.311–2.23) 0.713
<0.83 Ref.
SI at 10 min after delivery
≥0.83 1.08 (0.35–3.41) 0.888
<0.83 Ref.
SI at 30 min after delivery
≥0.83 5.182 (1.77–15.13) 0.002
<0.83 Ref.

17. Accepted
Article
This article is protected by copyright. All rights reserved.
Abbreviations: SI, shock index.
a
Multiple regression using data from 151 complete records.
Table 4 Sensitivity, specificity, and likelihood ratio of SI values above 0.83 to predict postpartum
hemorrhage.
SI ≥0.83 Sensibility (95% CI), % Specificity (95% CI), % Positive likelihood
ratio
Negative
likelihood ratio
At delivery 0.30 (0.20–0.40) 0.74 (0.66–0.83) 1.16 0.95
10 min 0.40 (0.30–0.50) 0.77 (0.69–0.85) 1.76 0.78
30 min 0.47 (0.36–0.57) 0.79 (0.72–0.86) 2.21 0.68
Abbreviations: SI, shock index; CI, confidence interval.