Pediatric Anesthesia ISSN 1155-5645

ORIGINAL ARTICLE

Development and validation of a risk score to predict the

probability of postoperative vomiting in pediatric patients:
the VPOP score
Nathalie Bourdaud1, Jean-Michel Devys2, Jocelyne Bientz3, Corinne Lejus4, Anne Hebrard5,
Olivier Tirel6, Damien Lecoutre7, Nada Sabourdin8, Yves Nivoche9, Catherine Baujard10 &
Gilles A. Orliaguet1
1 Department of Anesthesiology and Critical Care Medicine, Centre Hospitalier Universitaire Necker – Enfants Malades, AP-HP, University
Paris Descartes, Paris, France
2 Department of Anesthesiology and Intensive Care Unit, Fondation Ophtalmologique A. de Rothschild, Paris, France
3 Department of Anesthesiology and Critical Care Medicine, Centre Hospitalier Universitaire Hautepierre, Strasbourg, France
4 Department of Anesthesiology and Critical Care Medicine, Centre Hospitalier Universitaire Hotel-Dieu, Nantes, France
5 Department of Anesthesiology and Critical Care Medicine, Centre Hospitalier Universitaire Jeanne de Flandres, Lille, France
6 Department of Anesthesiology and Critical Care Medicine, Centre Hospitalier Universitaire Pontchaillou, Rennes, France
7 Department of Anesthesiology, Hospital Saint-Vincent de Paul, Lille, France
8 Department of Anesthesiology and Critical Care Medicine, Centre Hospitalier Universitaire Armand Trousseau, AP-HP, University Pierre et
Marie Curie, Paris, France
9 Department of Anesthesiology and Critical Care Medicine, Centre Hospitalier Universitaire Robert Debre, AP-HP, University Paris Diderot,
Paris, France
10 Department of Anesthesiology and Critical Care Medicine, Centre Hospitalier Universitaire de Bicetre, AP-HP, University Paris Sud, Le
Kremlin Bicetre, France

Keywords
pediatrics; postoperative nausea and
vomiting; risk factors
Correspondence
Gilles A. Orliaguet, Department of
Anesthesiology and Critical Care Medicine,
Centre Hospitalier Universitaire Necker –
Enfants Malades, 149 rue de Sevres, Paris
75743, France
Email: gilles.orliaguet@nck.aphp.fr
Section Editor: Jerrold Lerman
Accepted 9 April 2014
doi:10.1111/pan.12428
Summary
Background: Few data are available in the literature on risk factors for post-
operative vomiting (POV) in children.
Objective: The aim of the study was to establish independent risk factors for
POV and to construct a pediatric specific risk score to predict POV in chil-
dren.
Methods: Characteristics of 2392 children operated under general anesthesia
were recorded. The dataset was randomly split into an evaluation set
(n = 1761), analyzed with a multivariate analysis including logistic regression
and backward stepwise procedure, and a validation set (n = 450), used to
confirm the accuracy of prediction using the area under the receiver operating
characteristic curve (ROCAUC), to optimize sensitivity and specificity.
Results: The overall incidence of POV was 24.1%. Five independent risk fac-
tors were identified: stratified age (>3 and <6 or >13 years: adjusted OR 2.46
[95% CI 1.75–3.45]; ≥6 and ≤13 years: aOR 3.09 [95% CI 2.23–4.29]), dura-
tion of anesthesia (aOR 1.44 [95% IC 1.06–1.96]), surgery at risk (aOR 2.13
[95% IC 1.49–3.06]), predisposition to POV (aOR 1.81 [95% CI 1.43–2.31]),
and multiple opioids doses (aOR 2.76 [95% CI 2.06–3.70], P < 0.001). A sim-
plified score was created, ranging from 0 to 6 points. Respective incidences of
POV were 5%, 6%, 13%, 21%, 36%, 48%, and 52% when the risk score ran-
ged from 0 to 6. The model yielded a ROCAUC of 0.73 [95% CI 0.67–0.78]
when applied to the validation dataset.
Conclusions: Independent risk factors for POV were identified and used to
create a new score to predict which children are at high risk of POV.

© 2014 John Wiley & Sons Ltd 945

Pediatric Anesthesia 24 (2014) 945–952
Postoperative vomiting in children: the VPOP score
Introduction
Over the last decades, life-threatening complications
associated with anesthesia have decreased encouraging
anesthesiologists and patients to focus attention on
postoperative comfort. Postoperative nausea and vomit-
ing (PONV) is one of the most frequent adverse effects
and a main source of discomfort after general anesthesia
(1). In adults, several scores have been developed to
assess the risk of PONV to adapt PONV prevention on
the basis of an individual risk. However, risk scores vali-
dated in adults are not applicable to children (2). More-
over, as children often cannot complain about nausea,
only postoperative vomiting (POV) is usually studied in
children. Thus, risk factors specific for the pediatric
population are needed and a risk model is required to
facilitate the prediction of POV in children. At present,
only one risk score is available to predict the probability
of POV in children (POVOC score) (3). This score
includes strabismus surgery, age ≥3 years, duration of
surgery >30 min, and history of POV in the child or of
POV/PONV in its relatives (3). However, this score pre-
sents limitations. For example, this score theoretically
composed of four risk factors relies most of the time
upon three items, because the only surgery it includes is
strabismus correction (4,5), decreasing the discrimina-
tive power of the score.
Therefore, we conducted a prospective study in chil-
dren to determine independent risk factor for POV and
to create a new risk score of POV that could be used in a
wide pediatric patient population.
Methods
Study design and patient population
In this prospective observational study, patients were
recruited in France from 11 pediatric anesthesia depart-
ments of nine University Hospital and two general
hospitals. All patients aged from 0 to 16 years and man-
aged for any surgery (elective or emergency) or proce-
dures under general anesthesia were eligible for this
study. Patients who required preoperative antiemetic
prophylaxis because of severe potential consequences of
vomiting were not included. Exclusion criteria were the
administration of any antiemetic in the pre- or intraop-
erative period, the need for maintenance of tracheal
intubation and/or sedation for more than 1 h postopera-
tively, as well as if the primary end point (occurrence of
POV) cannot be found.
A total of 2392 children were included between
August 2007 and July 2008. One hundred and eighty-
one (8%) patients were excluded because of mainte-
946
N. Bourdaud et al.
nance of postoperative endotracheal tube more than 1 h
(n = 9), intraoperative antiemetic prevention (n = 89),
missing data regarding vomiting issue (n = 71), and
other miscellaneous reasons (n = 11). The sample size
used for statistical analysis was 2211 patients, including
1761 children in the evaluation set and 450 children in
the validation set (Figure 1).
Data collection
Preoperatively, the following data were obtained from
the parents or the children: history of PONV or motion
sickness in the child, history of PONV in the relatives,
and smoking status of the child and the parents. Preop-
erative characteristics and details of patients’ medical
and surgical history were also recorded.
For scheduled surgery, all children were fasted 2 h for
clear fluids and 6 h for milk and solids. In case of
emergency procedure, rapid sequence induction was per-
formed with intravenous hypnotic (propofol or thiopen-
tal) and succinylcholine. For elective surgery, oral
premedication was administered 20–60 min before sur-
gery. The type of premedication was recorded. Because
more than eight different preoperative treatments were
used, including many different combinations, this vari-
able was simplified into three classes: midazolam alone
or in combination, hydroxyzine alone or in combina-
tion, and any combinations of other medications. The
anesthesia technique was not standardized because of
the observational character of this study with no change
in daily practices. All drugs administered for induction
and maintenance of anesthesia were recorded. Hypno-
tics for induction were stratified into six classes includ-
ing: (1) propofol alone, (2) propofol in combination, (3)
sevoflurane alone, (4) sevoflurane in combination, (5)
nitrous oxide and (6) others. No prophylactic antiemet-
ics were given.
Postoperatively, the patients received supplemental
oxygen to maintain pulse oximetry (SpO2) >95% and
pain relief was provided using paracetamol and/or non-
steroidal antiinflammatory drugs (NSAIDs) and/or
morphine, according to the surgical procedures, pain
assessment, and local practices. The need for postopera-
tive opioids was left to the discretion of the anesthesiolo-
gist in charge of the patient, and the dose was adjusted
according to clinical needs. From the first episode of
POV, treatment was instituted with ondansetron, drop-
eridol, or dexamethasone, according to local procedures.
All postoperative drugs administered to the children
were recorded.
To simplify further analyses, several transformations
were performed following the initial univariate analysis.
Because the relationship between the age of the patient
© 2014 John Wiley & Sons Ltd
Pediatric Anesthesia 24 (2014) 945–952
N. Bourdaud et al.
Figure 1 Flow chart.
and the risk of POV was not linear, age was stratified
into a categorical variable with three groups (class 1:
≤3 years; class 2: >3 years and <6 years or >13 years;
and class 3: ≥6 years and ≤13 years). Duration of anes-
thesia was defined as <45 min or ≥45 min. Surgery at
risk included tympanoplasty, tonsillectomy, and strabis-
mus surgery. Similarly, to reduce the number of vari-
ables, we decided to combine ‘personal history of POV,’
‘motion sickness,’ and ‘familial history of POV’ into a
composite variable called ‘predisposition to POV.’ In
addition, we merged ‘intraoperative opioids re-injection’
and ‘postoperative use of opioids’ into a single variable
called ‘multiple opioids doses.’
POV assessment
All patients were admitted in the postanesthesia care
unit (PACU) following anesthesia, except those requir-
ing hospitalization in the pediatric intensive care unit.
POV was assessed in the PACU by specially instructed
nurses or anesthesiologists. Early POV were defined as
POV occurring during the stay in the PACU. All POV
episodes occurring during the first 24 h after anesthesia
were noted. For patients having surgery on an outpa-
tient basis, parents were asked to fill out a prefilled letter
and to send it to the Clinical Research Unit on the
second day after surgery. In case of nonreception of the
letter, the parents were interviewed by phone the follow-
ing day. Because nausea is a subjective phenomenon
and small children are not able to describe it, only POV
were evaluated.
© 2014 John Wiley & Sons Ltd
Pediatric Anesthesia 24 (2014) 945–952
Postoperative vomiting in children: the VPOP score
The primary end point of this study was the propor-
tion of POV during the first postoperative 24 h.
Sample size
The sample size of 1000 children was calculated to dem-
onstrate a 10% difference in POV incidence between
unexposed (POV incidence = 25%) and exposed children
(POV incidence = 35%) with a statistical power of 90%
and alpha type I error of 5%, assuming a 20% prevalence
of exposure to a specific risk factor. An independent sam-
ple was added for score validation (n = 500). The total
sample size was 1500, and it was reached before the
expected date.
Statistical analysis
The evaluation dataset and the validation dataset com-
prised 1761 and 450 participants, respectively, with com-
pleted data. Descriptive and univariate analysis included
the Student t-test or Wilcoxon rank test for quantitative
variables or v2 test for intergroup comparisons.
Variables associated with POV (P value <0.10) were
then computed in multiple logistic regression models
using a backward stepwise procedure to evaluate the
impact of potential risk factors on POV. For each pre-
dictor, crude and adjusted ORs with their 95% confi-
dence interval [95% CI] were calculated using regression
coefficients. The performances of the model were
assessed on the evaluation set by calibration and dis-
crimination. We plotted observed outcome by decile of
947
Postoperative vomiting in children: the VPOP score
predictions, which makes the plot a graphical illustra-
tion of the Hosmer–Lemeshow goodness-of-fit test. The
internal validation consisted in determining the discrimi-
native ability of the model by measuring areas under the
receiver operating characteristics curve (ROCAUC).
Internal validation was completed by applying the opti-
mal model to validation set. We built POVOC score
applying the customization approach as described by
Engel et al. (6). A comparison of ROCAUC between
vomiting in the post operative period (VPOP) and cus-
tomized POVOC models was performed on the valida-
tion set using bootstrap method which included a 2000-
fold sampling without replace. For clinical practice, a
score was proposed, based on regression coefficients.
Statistical analysis was performed with R program (Bos-
ton, MA, USA www.r-project.org) (7) and specific pack-
ages (EPICALCTM, PROC, EPITOOLS, PREDICTABEL; R Core
Team, R Foundation for Statistical Computing, Vienna,
Austria).
Ethics
This prospective, multicenter study was registered at
the Research Ministry (CCTIRS – French Consulta-
tive Committee for Information Management con-
cerning participants for medical research, permission
no 07.108) and at the French National Commission
for Information Technology and Civil Liberties (per-
mission no 907124). The institutional review board
(IRB) (i.e., Comit
e de Protection des Personnes Ile-
de-France III) of Tarnier-Cochin Hospital (University
Paris Descartes, 75005 Paris, France) approved the
study for all participating study centers. Waiver of
informed consent was authorized by the IRB because
of the design of the study: prospective observational
study, without changes in the usual daily management
of the children. The patients and their parents were
informed about the trial and were asked to sign the
information form. In case the parents or the child
refuse to participate in the study, the child was not
included, as prescribed by the IRB.
Results
Descriptive epidemiology
Pre-, intra-, and postoperative characteristics of children
did not differ between the evaluation and the validation
set (see Table S1 and S2). The ages ranged from birth to
16 years, with over three-quarters of the patients
<9 years of age. There were more boys than girls (sex
948
N. Bourdaud et al.
ratio = 1.4). Premedication concerned about 82% of
children, with half of the participants receiving midazo-
lam. The main types of surgery were orthopedics (24%),
urology (16%), ophthalmology (14%), and ENT sur-
gery (12%). Three-quarter of participants received opi-
oids during induction, with half of them receiving
sufentanil. Maintenance of anesthesia was performed
with sevoflurane in 80% of the children. Maintenance
with opioid occurred in a half of the subjects, with a
third of them receiving sufentanil. Among the 373 chil-
dren who received muscle relaxants, 36 were reversed
with a combination of neostigmine and atropine. Dura-
tion of anesthesia ranged from 9 to 520 min, with a
median value of 60 min.
For ambulatory patients, the final completion rate for
parental reports (prefilled letter or phone call) was 88%.
The global incidence of POV was 24.1%, with 414
patients (23.5%) in the evaluation set and 118 patients
(26.2%) in the validation set (P = 0.25).
Postoperative vomiting determinants: univariate
analysis
Factors significantly associated with POV in univariate
analysis were age, predisposition to POV, premedication,
surgery at risk, induction with muscle relaxant or opi-
oids, intraoperative reinjection of opioids, multiple doses
of opioids, duration of anesthesia, and postoperative
analgesia with nalbuphine or IV morphine (Table 1).
Postoperative vomiting risk factors: multivariate
analysis
Weight was discarded from the multivariate analysis
because of colinearity with age, as was regional anesthe-
sia because of colinearity with surgery. Premedication
with midazolam was not included in the model mainly
because it was strongly associated with some types of
surgery, in particular types of surgery that are risk
factors for POV (tympanoplasty, tonsillectomy, or stra-
bismus surgery).
The multivariate analysis included initially nine inde-
pendent risk factors for POV in children: age, duration
of anesthesia, predisposition to POV, surgery at risk,
use of muscle relaxant or of opioid during induction, use
of opioid for maintenance, and postoperative use of nal-
buphine or of IV morphine. The stepwise procedure
returned a simplified model including five predictors:
age in three classes, duration of anesthesia >45 min,
predisposition to POV, surgery at risk, and multiple
opioids doses (Table 2).
© 2014 John Wiley & Sons Ltd
Pediatric Anesthesia 24 (2014) 945–952
N. Bourdaud et al. Postoperative vomiting in children: the VPOP score

Table 1 Characteristics of vomiters and nonvomiters (n = 1761 –

univariate analysis) Internal validation

Nonvomiters Vomiters P The agreement between observed and predicted POV

n = 1347 n = 414 value reported a nonsignificant Hosmer–Lemeshow goodness-
of-fit test (P = 0.53). The ROC curve of the VPOP
Characteristics
model (AUC = 0.73 [95% CI 0.67–0.78]) displayed a
Age (years) 5.7  4.5 7.4  3.9 <0.001
Age by class <0.001 larger AUC than the POVOC model (0.66 [95% CI
≤3 503 (37) 56 (14) 0.61–0.71]) (P < 0.001) (Figure 2).
>3 and <6 or ≥13 397 (29) 141 (34)
>6 and <13 447 (33) 217 (52)
Sex (boys) 800 (59) 232 (56) 0.25 VPOP score
History
The score varies from 0 to 6 points (Table 3). In the vali-
History of motion sickness 184 (14) 94 (23) <0.001
History of POV 75 (6) 59 (14) <0.001
dation set, the incidences of POV were, respectively,
Familial history of motion 209 (16) 90 (22) 0.067 5%, 6%, 13%, 21%, 36%, 48%, and 52% for a score of
sickness or POV 0, 1, 2, 3, 4, 5, and 6. Therefore, patients with 0 or 1
Predisposition to POVa 390 (29) 192 (46) <0.001 point score can be considered with low risk of POV,
Home tobacco smoke 394 (29) 135 (33) 0.19 patients with 2 or 3 points have a moderate risk of POV,
exposure
and patients with a VPOP score >4 have a high risk of
Preoperative data
Preoperative medicationb 1068 (79) 373 (90) <0.001
POV.
Midazolam 530 (39) 187 (45) 0.04
Hydroxyzine 388 (29) 108 (26) 0.31
Discussion
Clorazepate or diazepam 27 (2) 10 (2) 0.75
Other 124 (9) 68 (16) <0.001 In this multicenter prospective study, five independent
Gastric tube 373 (28) 126 (30) 0.30
risk factors for POV in children were identified,
Full stomach 33 (2) 8 (2) 0.68
Surgery at riskc 82 (6) 70 (17) <0.001
including age, predisposition to POV, surgery at risk,
Intraoperative data duration of anesthesia >45 min, and multiple opioid
Hypnotic agent for induction 0.70 doses; and a pediatric risk score of POV was created,
Propofol 241 (18) 81 (20)
Sevoflurane 328 (24) 87 (21)
Nitrous oxide 583 (43) 195 (45)
Table 2 Risk factors for postoperative vomiting (POV) in the evalua-
Other 16 (1) 5 (1)
tion set (n = 1761)
Induction with 223 (17) 92 (22) <0.001
muscle relaxant P
Muscle relaxant antagonist 33 (2) 6 (1) 0.31 Crude OR Adjusted OR (Wald’s
Induction with opioid 977 (73) 353 (85) <0.001 (95% CI) (95% CI) test)
Multiple opioid doses 629 (47) 278 (67) <0.001
Regional anesthesia 368 (27) 91 (22) 0.04 Age: ref. = ≤3 years 1 1
Duration of 77  57 97  72 <0.001 >3 and <6 or 2.95 (2.13, 4.1) 2.46 (1.75,3.45) <0.001
anesthesia (min) >13 years
Duration of anesthesia 951 (71) 344 (83) <0.001 ≥6 and ≤13 years 4.03 (2.94, 5.52) 3.09 (2.23, 4.29) <0.001
≥45 min Duration of 2.05 (1.54, 2.71) 1.44 (1.06, 1.96) 0.019
Postoperative analgesia 1045 (78) 342 (83) 0.03 anesthesia >45 min
Paracetamol 819 (61) 240 (58) 0.33 Surgery at riska 3.14 (2.23, 4.41) 2.13 (1.49, 3.06) <0.001
Nalbuphine 349 (26) 144 (35) <0.001 Predisposition 2.12 (1.69, 2.66) 1.81 (1.43,2.31) <0.001
Nonsteroidal 219 (16) 63 (15) 0.67 to POVb
antiinflammatory drugs Multiple doses 3.27 (2.49, 4.31) 2.76 (2.06,3.70) <0.001
IV morphine 161 (12) 80 (19) <0.001 of opioidsc
Oral morphine 16 (1) 7 (2) 0.59
Other 9 (<1) 4 (<1) 0.77 Log-likelihood = 855.180; No. of observations = 1761; AIC
Gastric tube maintenance 24 (2) 8 (2) 0.99 value = 1724.3608; Hosmer–Lemeshow statistics: v² = 7.027;
df = 8; P value = 0.53.
Data are mean  SD or number of cases (%). a
Surgery at risk: tympanoplasty, tonsillectomy, and strabismus sur-
a
Personal history of postoperative vomiting (POV) or motion sickness gery.
or familial history of POV. b
Predisposition to POV: personal history of POV or motion sickness
b
Total >100%, many children having a combination of drugs. or familial history of POV.
c c
Surgery at risk includes tympanoplasty, tonsillectomy, and strabis- Multiple doses of opioids: injection was performed either during
mus surgery. induction or maintenance of anesthesia or in the postoperative period.

© 2014 John Wiley & Sons Ltd 949

Pediatric Anesthesia 24 (2014) 945–952
Postoperative vomiting in children: the VPOP score
Figure 2 Comparison of the area under the ROC curve (AUC) based
on the VPOP and the customized POVOC model (n = 450). P value
refers to the comparison between two paired ROC curves according
to the bootstrap method.
Table 3 Clinical risk score for postoperative vomiting in the evalua-
tion set (simplified model)
ranging from 0 to 6 points. With this score, children
with a score of 0–1 point have a low risk of POV,
children with a score of 2–3 have a moderate risk, and
children with a score of 4 or more have a high risk. The
logistic model demonstrated good performances which
gives a reasonable confidence to the VPOP score. In the
950
N. Bourdaud et al.
validation dataset, the score obtained from our model
had a greater discriminative ability to predict POV than
the other previously published pediatric score (3).
Some risk factors for POV in children found in
our study are similar to those found previously (3);
however, our study provides more accurate results
for some other variables. Until now, only age
>3 years was considered as a risk factor for POV
(8,9). We have demonstrated that there was no linear
relationship between age and risk of POV but that
between 6 and 13 years, the risk becomes twice that
for periods between 3 and 6 years or beyond age 13.
This particular relationship had never been demon-
strated, although many studies have shown that the
incidence of POV is greater in the prepubertal popu-
lation (10–12). We recorded duration of anesthesia
and not duration of surgery, as it had been reported in
others studies (3,13). The duration of anesthesia corre-
lates with duration of surgery (usually no more than
15 min longer, except in case of major surgery). How-
ever, in pediatric patients, venous access may be diffi-
cult and the exposure to volatile anesthetics may last a
long period of time, even for a surgery of very short
duration. Therefore, we preferred recording the dura-
tion of anesthesia instead of the duration of surgery.
Anyway, the threshold value of 45 min of anesthesia
that we found is in agreement with previous findings
(3).
The role of different types of surgery remains contro-
versial with its inclusion only into two of the seven
more frequently used scoring systems for adults (14,15).
In the study by Eberhart et al. (3), only strabismus sur-
gery was found to be an independent risk factor for
POV in children. In our study, tympanoplasty and ton-
sillectomy in addition to strabismus surgery were found
to be independent risk factors for POV in children.
Therefore, combining three types of surgery, our score
becomes more discriminative for predicting POV.
We, as other have published, found that personal his-
tories of POV and of motion sickness were independent
risk factors for POV in children (16,17). In addition, we
found that familial history of POV was an independent
risk factor, but with less influence as compared to per-
sonal history of POV or of motion sickness (3). Combin-
ing these three variables, the resulting odds ratio was
similar and we decided to group these factors into one
called ‘predisposition to POV.’
The use of opioids during induction of anesthesia is
not by itself an independent risk factor for POV. Use of
opioids becomes a risk factor for POV only when opi-
oids are administrated again during surgery or in the
postoperative period. This finding is consistent with
previous studies in adult patients (18). To simplify the
© 2014 John Wiley & Sons Ltd
Pediatric Anesthesia 24 (2014) 945–952
N. Bourdaud et al.
score and because respective odds ratio were close, we
merged these variables into a new one called ‘multiple
opioids doses.’
Premedication with midazolam was a significant risk
factor for POV in children. However, the few data avail-
able in the literature indicate an antiemetic effect of
midazolam in children (19–22). Because of the surprising
nature of this result, we decided to perform further
analysis to eliminate an unanticipated potential bias in
our population. It appeared that midazolam was
strongly associated with some types of surgery, in partic-
ular types of surgery that are risk factors for POV (tym-
panoplasty, tonsillectomy, or strabismus surgery).
Therefore, and also because Apfel et al. (23) showed
that the inclusion of more than five risk factors did not
lead to a clinical improvement in the prediction of
PONV, we decided to restrict the number of risk factors
to 5 and we chose to exclude premedication with
midazolam from our score.
Our study presents limitations. Our study was a
prospective observational study, without changes in
the usual daily practice; therefore, anesthesia man-
agement was not standardized but represented the
usual practices of each institution. Unlike in adults
(24), the type of anesthetics (volatile or intravenous)
did not influence the occurrence of POV in our
study. For maintenance of anesthesia, the number of
patients that received total intravenous anesthesia
was too small to conclude that propofol could have
a protective effect toward POV. In half of the
patients, nitrous oxide was used for induction and/or
maintenance of anesthesia. We did not find any sig-
nificant effect of nitrous oxide on POV in children,
as has already been observed in adult patients. For
ambulatory patients, we were eventually able to
obtain a completion rate of 88% for parental
reports. However, we do think that the few lacking
data do not modify the results of the study. Finally,
we are not able to definitively eliminate that other
factors, in addition to those shown to be significant
in our study, may influence the incidence of POV in
children. However, we do think that they do not
play a central role in the occurrence of POV in
children.
In conclusion, we found specific pediatric risk factors
for POV (stratified age, predisposition to POV, duration
of anesthesia, surgery at risk, and multiple opioids
doses) and we created a simplified score, based on five
items ranging from 0 to 6, that allows prediction of the
risk of POV for each child.
© 2014 John Wiley & Sons Ltd
Pediatric Anesthesia 24 (2014) 945–952
Postoperative vomiting in children: the VPOP score
Acknowledgments
The authors thank Sophie Gacia, M.D. (H^ opital
d’Instruction des Arm
ees Legouest, Metz, France) for
technical support, as well as Vijay Acharya (Embassy of
the United States of America, American Embassy in
Paris, Paris, France) for kindly reviewing the manu-
script.
Steering Committee of the VPOP Trial Group: Chairs:
N. Bourdaud, G. Orliaguet; Biostatisticians: P.Y. Ancel,
J.P. Jais, N. Nikasinovic (Necker – Enfants Malades
University Hospital of Paris, Biostatistic Unit, Paris,
France and School of Medicine, University Paris
Descartes, Paris, France); Data Management: C. Tour-
te, V. Jolaine (Necker – Enfants Malades University
Hospital of Paris, Clinical Reseach Unit Department,
Paris, France and School of Medicine, University Paris
Descartes, Paris, France); Project Coordinator: N.
Bourdaud.
VPOP Trial Investigators: N. Bourdaud, S. Gacia, G.
Orliaguet, J. Bientz, C. Populaire (H^ otel-Dieu Univer-
sity Hospital of Nantes, Anaesthesia and Intensive Care
Unit Department, Nantes, France), Anne Hebrard,
Jean-Michel Devys, O. Tirel, D. Lecoutre, A. Humblot
(Armand Trousseau University Hospital of Paris,
Anaesthesia and Intensive Care Unit Department, Paris,
France), N. Sabourdin, Y. Nivoche, C. Baujard.
Funding
This work was supported, in part, by a Public Health
Service Award from the Programme Hospitalier de
Recherche Clinique national 2007 (PHRC national
2007). Direction G
en
erale de l’Offre de Soins, Ministere
charge de la Sant
e, Paris, France.
Conflicts of interests
No conflicts of interest declared.
Supporting information
Additional Supporting Information may be found in the
online version of this article:
Table S1 Preoperative data of the evaluation
(n = 1761) and validation (n = 450) set. Data are mean
 SD or number of cases (%).
Table S2 Intra- and postoperative data of the evalua-
tion (n = 1761) and validation (n = 450) set. Data are
mean  SD or number of cases (%).
951
Postoperative vomiting in children: the VPOP score
References
1 Murat I, Constant I, Maud’huy H. Perioper-
ative anaesthetic morbidity in children: a
database of 24,165 anaesthetics over a 30-
month period. Pediatr Anesth 2004; 14:
158–166.
2 Eberhart LH, Morin AM, Guber D et al.
Applicability of risk scores for postoperative
nausea and vomiting in adults to paediat-
ric patients. Br J Anaesth 2004; 93:
386–392.
3 Eberhart LH, Geldner G, Kranke P et al.
The development and validation of a risk
score to predict the probability of postopera-
tive vomiting in pediatric patients. Anesth
Analg 2004; 99: 1630–1637.
4 Friedman DS, Repka MX, Katz J et al.
Prevalence of amblyopia and strabismus in
white and African American children aged 6
through 71 months the Baltimore Pediatric
Eye Disease Study. Ophthalmology 2009;
116: 2128–2134 e2121–2122.
5 Robaei D, Rose KA, Ojaimi E et al. Causes
and associations of amblyopia in a popula-
tion-based sample of 6-year-old Australian
children. Arch Ophthalmol 2006; 124: 878–
884.
6 Engel JM, Junger A, Hartmann B et al.
Performance and customization of 4
prognostic models for postoperative onset of
nausea and vomiting in ear, nose, and
throat surgery. J Clin Anesth 2006; 18:
256–263.
7 Ihaka R, Gentleman R. R: a language for
data analysis and graphics. J Comput Graph
Stat 1996; 5: 299–314.
952
8 Lerman J. Surgical and patient factors
involved in postoperative nausea and vomit-
ing. Br J Anaesth 1992; 69: 24S–32S.
9 Oddby E, Englund S, Lonnqvist PA. Postop-
erative nausea and vomiting in paediatric
ambulatory surgery: sevoflurane versus
spinal anaesthesia with propofol sedation.
Paediatr Anaesth 2001; 11: 337–342.
10 Cohen MM, Cameron CB, Duncan PG.
Pediatric anesthesia morbidity and mortality
in the perioperative period. Anesth Analg
1990; 70: 160–167.
11 Karlsson E, Larsson LE, Nilsson K. Postan-
aesthetic nausea in children. Acta Anaesthesi-
ol Scand 1990; 34: 515–518.
12 Murray DJ, Schmid CM, Forbes RB. Anes-
thesia for magnetic resonance imaging in
children: a low incidence of protracted post-
procedure vomiting. J Clin Anesth 1995; 7:
232–236.
13 Apfel CC, Laara E, Koivuranta M et al. A
simplified risk score for predicting postopera-
tive nausea and vomiting: conclusions from
cross-validations between two centers. Anes-
thesiology 1999; 91: 693–700.
14 Gan TJ. Risk factors for postoperative nau-
sea and vomiting. Anesth Analg 2006; 102:
1884–1898.
15 Sinclair DR, Chung F, Mezei G. Can postop-
erative nausea and vomiting be predicted?
Anesthesiology 1999; 91: 109–118.
16 Thomas M, Woodhead G, Masood N et al.
Motion sickness as a predictor of postopera-
tive vomiting in children aged 1–16 years.
Pediatr Anesth 2007; 17: 61–63.
N. Bourdaud et al.
17 Busoni P, Sarti A, Crescioli M et al. Motion
sickness and postoperative vomiting in chil-
dren. Paediatr Anaesth 2002; 12: 65–68.
18 Apfel CC, Philip BK, Cakmakkaya OS et al.
Who is at risk for postdischarge nausea and
vomiting after ambulatory surgery? Anesthe-
siology 2012; 117: 475–486.
19 Elhakim M, Abd-Elfattah H, El-Din DN
et al. Midazolam as an antiemetic in patients
receiving epidural morphine for postopera-
tive pain relief. J Opioid Manag 2009; 5:
189–195.
20 Riad W, Altaf R, Abdulla A et al. Effect of
midazolam, dexamethasone and their combi-
nation on the prevention of nausea and vom-
iting following strabismus repair in children.
Eur J Anaesthesiol 2007; 24: 697–701.
21 Riad W, Marouf H. Combination therapy in
the prevention of PONV after strabismus
surgery in children: granisetron, ondanse-
tron, midazolam with dexamethasone. Mid-
dle East J Anesthesiol 2009; 20: 431–436.
22 Splinter WM, MacNeill HB, Menard EA
et al. Midazolam reduces vomiting after ton-
sillectomy in children. Can J Anaesth 1995;
42: 201–203.
23 Apfel CC, Kranke P, Greim CA et al. What
can be expected from risk scores for predict-
ing postoperative nausea and vomiting? Br
J Anaesth 2001; 86: 822–827.
24 Apfel CC, Korttila K, Abdalla M et al. A
factorial trial of six interventions for the pre-
vention of postoperative nausea and vomit-
ing. N Engl J Med 2004; 350: 2441–2451.
© 2014 John Wiley & Sons Ltd
Pediatric Anesthesia 24 (2014) 945–952