anafilaxis 2024 bja

Download as pdf or txt
Download as pdf or txt
You are on page 1of 7

British Journal of Anaesthesia, 133 (3): 584e590 (2024)

doi: 10.1016/j.bja.2024.05.025
Advance Access Publication Date: 3 July 2024
Clinical Investigation

PAEDIATRIC ANAESTHESIA

Incidence of and risk factors for paediatric perioperative


anaphylaxis in the USA
Ifat Z. Krase1,2 , Christine R. F. Rukasin2 , Keith Sacco2, Matthew A. Rank1,2,
Gerald W. Volcheck3 and Alexei Gonzalez-Estrada1,*
1
Division of Allergy, Asthma, and Clinical Immunology, Department of Medicine, Mayo Clinic, Scottsdale, AZ,
USA, 2Section of Allergy and Immunology, Division of Pulmonology, Phoenix Children’s Hospital, Phoenix, AZ, USA and
3
Division of Allergic Diseases, Department of Internal Medicine, Mayo Clinic, Rochester, MN, USA

*Corresponding author. E-mail: [email protected]

Abstract
Background: Perioperative anaphylaxis is a serious and often life-threatening immediate hypersensitivity reaction. There
are few published data on paediatric perioperative anaphylaxis (pPOA). We evaluated the incidence of and risk factors
involved in the occurrence of pPOA within a large US national database.
Methods: Deidentified data from the US Nationwide Inpatient Sample from 2005 to 2014 were used to identify pPOA cases
and to conduct a retrospective multivariate analysis of preselected independent variables.
Results: Among 3,601,180 surgeries and procedures in children aged 0e18 yr, 297 pPOA cases were identified for an
incidence of one in 12,125 surgeries and procedures. Compared with controls, pPOA cases had an increased median
length of stay (6 vs 2 days; P<0.001) and median hospital cost ($54 719 vs $5109; P<0.0001). The age groups between 6 and
12 yr (odds ratio [OR] 7.1; 95% confidence interval [CI] 3.9e12.9; P<0.001) and 13 and 17 yr (OR 8.5; 95% CI 4.7e15.2; P<0.001)
were associated with increased odds of pPOA. Transplant (OR 46.3; 95% CI 20.8e102.9; P<0.001), cardiac (OR 16.4; 95% CI
7.5e35.9; P<0.001), and vascular (OR 15.2; 95% CI 7.5e30.7; P<0.001) procedures posed the highest risk for pPOA. Chronic
pulmonary disease, coagulopathy, and fluid and electrolyte disorders were also associated with pPOA (OR 2.2; 95% CI
1.5e3.3; P<0.001).
Conclusions: The incidence of pPOA was one in 12,125 cases. Risk factors included age, procedure type, and
comorbidities.

Keywords: anaphylaxis; drug allergy; hypersensitivity; paediatric anaesthesia; perioperative outcomes

Editor’s key points  These data highlight the incidence of paediatric


perioperative anaphylaxis and identify patients at an
 Perioperative anaphylaxis is rare but potentially life- increased risk based on baseline characteristics and
threatening; it is understudied worldwide, particu- comorbidities.
larly in the paediatric population.
 The incidence of paediatric perioperative anaphy-
laxis in a large US database was one in 12,125 cases; Perioperative anaphylaxis (POA) is a serious and often life-
main predisposing risk factors included age 6e17 yr, threatening immediate hypersensitivity reaction.1 Despite
transplant, cardiac and vascular procedures, and adequate resuscitation, per-case mortality was estimated at
pulmonary disease. one in 26.6 cases in the UK with a similar result observed in

Received: 28 August 2023; Accepted: 9 May 2024


© 2024 British Journal of Anaesthesia. Published by Elsevier Ltd. All rights are reserved, including those for text and data mining, AI training, and similar
technologies.
For Permissions, please email: [email protected]

584
Paediatric perioperative anaphylaxis in the USA - 585

France for mortality associated with anaphylaxis to neuro- unspecified adverse effect or drug) or E930-E949 (drugs, me-
muscular blocking agents.1 The estimated incidence of pae- dicinal, and biological substance causing adverse effects in
diatric perioperative anaphylaxis (pPOA) varies widely from therapeutic use); (2) 458.9 (hypotension); and (3) either 519.11
one in 2100 to one in 37 000 procedures.2e4 Several studies of (bronchospasm) or 786.1 (stridor). Criterion B is a stricter
pPOA have been performed in Europe, including the Anaes- anaphylaxis criterion consistent with current US anaphylaxis
thesia Practice In Children Observational Trial (APRICOT) guidelines.13 Controls were identified from the database as all
performed across 33 countries, the Sixth National Audit Proj- paediatric patients who underwent a surgery or procedure and
ect (NAP6) in the UK, and an 8-yr national multicentre French did not develop pPOA during the study time frame.
study.3,5,6 The incidence of pPOA in the USA is unknown.
There is limited research demonstrating risk factors associ-
Variables
ated with pPOA, although history of asthma, food allergy,
multiple operations, and a family history of atopy have been We analysed pPOA cases by categorising them into four age
identified.7,8 The aims of this study are to assess the incidence groups (<1 yr, 1e5 yr, 6e12 yr, and 13e17 yr), sex, race and
of pPOA in the USA using a large national administrative ethnicity (White, Black, Hispanic, Asian or Pacific Islander,
database, and to evaluate characteristics of patients who Native American, and other), US region, non-teaching or
experienced pPOA to identify potential underlying risk factors. teaching hospital, elective or non-elective admission, length of
stay, procedure type, and mortality. US region categories
represented standardised geographic divisions as defined by
Methods the US Census Bureau. Procedures were categorised following
Study design and data source a previous methodology by organ system as: breast/skin; car-
diac; ear, nose, throat (ENT); general surgery-gastrointestinal
A retrospective analysis of data from the US Nationwide
(GI); genitourinary (GU)-urology; haematology; musculoskel-
Inpatient Sample (NIS) for 2005e14 was performed.9 The NIS is
etal; nervous system; obstetrics and gynaecology (OBGYN);
a part of the Healthcare Cost and Utilization Project (HCUP),
ophthalmic; thoracic; transplant; or vascular.11 Comorbidity
which comprises the largest collection of longitudinal hospital
measures included acquired immunodeficiency syndrome
care databases within the USA and includes discharge data
(AIDS), chronic blood loss anaemia, chronic pulmonary dis-
from all payers from over 1200 hospitals. Data from HCUP
ease, coagulopathy, congestive heart failure, depression, dia-
comprise a stratified sample of ~20% of all US hospitals. These
betes mellitus, drug abuse, fluid and electrolyte disorders,
data are deidentified thus eliminating the need for approval
hypertension, hypothyroidism, iron-deficiency anaemia, liver
from an institutional review board and patient-informed
disease, lymphoma, metastatic cancer, other neurological
consent. All authors involved were trained in the use of
disorders, obesity, paralysis, peripheral vascular disorders,
HCUP and adhered to HCUP policies. This study adhered to the
psychoses, pulmonary circulation disorders, renal failure,
reporting guidelines of Strengthening the Reporting of Obser-
rheumatoid arthritis/collagen vascular disorders, solid
vational Studies in Epidemiology.10
tumour without metastasis, valvular disease, and weight
loss.9,11 Data on teaching hospital status were not available in
Participants the NIS database for 2012e4. A total of 36 variables were
examined in the univariate analysis; 22 variables identified as
Patients were identified from the NIS database if they had a
statistically significant (P<0.05) were then further analysed
primary admission diagnosis for a surgical or procedural
using multivariate analysis. As per the HCUP data use agree-
indication. We utilised clinical classifications software codes,
ment, observation numbers with n<11 were not directly re-
which were used to classify Current Procedural Terminology
ported to minimise the risk for individual identification.
(CPT) codes into meaningful categories to identify all proced-
ures based on a previous methodology in an adult POA popu-
lation.11 Clinical classifications software codes were Statistical analysis
categorised according to organ system (e.g. nervous system),
Trends over time or between groups for categorical variables
anatomical localisation (e.g. thoracic), or surgical subspecialty
were analysed using c2 tests. Continuous variables were
(e.g. vascular surgery). Codes that would not potentially be
evaluated using KruskaleWallis tests. Potential risk factors for
associated with development of anaphylaxis were excluded
anaphylaxis were evaluated using logistic regression models.
(e.g. codes for fetal monitoring, chest radiography), as were
The odds ratios (ORs) and 95% confidence intervals (CIs) for
codes known to be associated with drug hypersensitivity re-
those ORs were presented from those models. Trends over the
actions owing to a clear causative agent (e.g. hypersensitivity
years in rates of pPOA were also evaluated using logistic
reactions that can occur as a result of certain treatments such
regression models. The analysis was completed using Statis-
as cancer chemotherapy). Radiologic procedures were not
tical Analysis System version 9.4 (SAS Institute, Cary, NC,
included in our analyses. However, organ-specific procedures
USA).
requiring radiocontrast media that require sedation were
included in our analyses. Anaphylaxis was identified using
International Classification of Diseases, Ninth Revision, Clinical Results
Modification (ICD-9-CM) codes and previously validated
Population
criteria.12 Participants were included under criterion A if they
had either one of two ICD-9-CM codes, 995.0 (other anaphy- During the 10-yr study period, 297 (one in 12,125) surgeries and
lactic shock) or 999.4 (anaphylactic reaction to serum) docu- procedures were associated with pPOA from a total of
mented during their admission. Criterion B identified 3,601,180 surgeries and procedures. The majority of patients
participants if they had the ICD-9-CM codes of symptom were identified under criterion A (288 patients), and the
combination: (1) 995.3 (allergy unspecified) or 995.2 (other remainder under criterion B (nine patients). The median age of
586 - Krase et al.

patients who experienced pPOA was 11 yr (interquartile range median length of stay (6 vs 2 days; P<0.0001), and had an
5e15), 54% were male, 55% were White, and 21% were Hispanic increased median hospital cost ($54,719 vs $5109; P<0.0001)
(Table 1). Elective procedures accounted for 66% of pPOA cases, (Table 2).
and 84% of pPOA cases occurred in teaching hospitals
(Table 1). The largest proportion of pPOA was among patients Time trends of paediatric perioperative anaphylaxis
aged 6e12 yr (40%) and occurred in the Southern region of the cases
US (37%) (Table 1). The mortality of pPOA was one in 59 cases
(1.7%). Further detail regarding mortality such as actual cases The proportion of pPOA cases in teaching hospitals overall
per year or procedure type are not reported to remain in remained stable from 2005 to 2011; however, information on
compliance with the HCUP data use agreement regarding pa- teaching hospital status was missing from the NIS database
tient privacy. for 2012e4, hence there are not enough data to examine the
trend through the entire study period and determine signifi-
cance. There were no statistically significant changes over
Characteristics of patients with paediatric time from 2005 to 2014 in the proportion of pPOA cases based
perioperative anaphylaxis on age, sex, race, US region, admission type, or mortality
Compared with controls, patients who experienced pPOA were (Table 2).
older (mean age 9.8 vs 3.3 yr; P<0.0001), had an increased
Risk factors for paediatric perioperative anaphylaxis
Patients aged 6e12 yr (OR 7.1; 95% CI 3.9e12.9; P<0.001) and
Table 1 Characteristics of perioperative anaphylaxis patients 13e17 yr (OR 8.5; 95% CI 4.7e15.2; P<0.001) both had an
and controls from 2005 to 2014. IQR, interquartile range; PA, increased risk for pPOA compared with patients aged <1 yr.
perioperative or periprocedural anaphylaxis; SD, standard de-
Females were at an increased risk for pPOA compared with
viation. *Values less than 11 (for both n and percentage).
males in our univariate analysis, but this was not evident in
multivariate analysis (Table 3). There was no significant dif-
Characteristic Controls PA cases P-
(n¼3,600,883) (n¼297) (%) value ference in pPOA cases based on admission type. Teaching
(%) hospital status was a significant risk factor in the univariate
and multivariate analyses; however, we lacked sufficient data
Age (yr) <0.001 for a complete analysis of this variable owing to missing in-
Median (IQR) 0 (0e4) 11 (5e15)
formation for 2012e4.
Groups <0.001
<1 2,505,194 (70) 24 (8)
1e5 254,163 (7) 60 (20) Risk factors for paediatric perioperative anaphylaxis
6e12 304,083 (8) 119 (40) by procedure type
13e17 537,443 (15) 94 (32)
Sex <0.001 Patients undergoing transplant procedures had the highest
Male 2,811,872 (78) 161 (54) risk of pPOA (OR 46.3; 95% CI 20.8e102.9; P<0.001), followed by
Female 789,011 (22) 136 (46) those undergoing cardiac (OR 16.4; 95% CI 7.5e35.9; P<0.001),
Race 0.09
vascular (OR 15.2; 95% CI 7.5e30.7; P<0.001), and haematologic
White 1,712,370 (60) 138 (55)
Black 449,905 (16) 30 (12) (OR 10.4; 95% CI 3.8e28.5) procedures (Table 3). Other less
Hispanic 435,781 (15) 53 (21) significant procedure categories included breast-skin, general
Asian or Pacific 92,788 (3) 12 (5) surgery-GI, musculoskeletal, and nervous system. The OBGYN
Islander and thoracic procedure categories showed significance in the
Native 22,287 (0.8) * (*) univariate analysis, but not in the multivariate analysis.
American
Other 154,069 (5) 14 (6)
Missing 733,498 (20) 48 (16) Risk factors for paediatric perioperative anaphylaxis
US region 0.003 by comorbid conditions
Northwest 632,176 (18) 58 (20)
Midwest 948,208 (26) 61 (21) Patients with comorbid AIDS had a significantly elevated risk
South 1,430,438 (40) 109 (37) for pPOA (OR 12.3; 95% CI 1.6e94.1; P¼0.02), as did those with
West 590,061 (16) 69 (23) chronic pulmonary (OR 2.2; 95% CI 1.5e3.3; P<0.001), coagulo-
Teaching <0.001 pathic (OR 2.2; 95% CI 1.3e3.6; P¼0.003), or fluid and electrolyte
hospital status
disorders (OR 1.9; 95% CI 1.3e2.8; P¼0.001). Pulmonary circu-
Non-teaching 1,164,311 (44) 32 (16)
Teaching 1,498,733 (56) 164 (84) lation disorders and renal failure were significant comorbid-
Admission type <0.001 ities in the univariate analysis, but not in the multivariate
Elective 3,137,120 (87) 197 (66) analysis. None of the comorbidities were highly correlated
Non-elective 450,805 (13) 100 (34) with one another, so the potential for collinearity was low.
Missing 12,920 (0.4) 0 (0)
Length of stay <0.001
Mean (SD) 4.5 (10.8) 12.1 (18.2) Discussion
Median (range) 2 (0e365) 6 (0e134)
Total charges <0.001 This is the first nationwide study on pPOA in the USA, and
Mean (SD) 26,069.6 130,584.0 showed an incidence of one in 12,125 procedures. Identifiable
(93,480.01) (207,206.83) risk factors for pPOA included age (6e17 yr), procedure type
Median (range) 5109 (26.0 54,719 (4470.0 (transplant, cardiac, and vascular procedures), and comorbid
e4,999,975.0) e1,367,626.0)
chronic pulmonary disease. Children with pPOA tended to be
older (11 yr and older), have longer hospital stays (7.6 more
Paediatric perioperative anaphylaxis in the USA - 587

Table 2 Overall cases and characteristics of patients with perioperative or periprocedural anaphylaxis from 2005 to 2014. Values
represent case numbers. *Values less than 11. NA indicates that data is not available because of small numbers.

Characteristic Year P-value

2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 Total

Overall cases 33 24 32 21 21 41 25 22 39 39 297


Age (yr) 0.02
<1 * * * * * * * * * * 24
1e5 * * * * * * * * * * 60
6e12 14 * * * * 15 * * 13 18 94
13e17 * 19 15 * * 17 11 * 17 12 119
Sex 0.08
Male 18 13 20 12 11 14 * 15 23 26 161
Female 15 11 12 * * 27 16 * 16 13 136
Race 0.26
White 16 10 11 * 13 14 17 * 22 18 138
Black * * * * * * * * * * 30
Hispanic * * * * * * * * * * 53
Asian or Pacific Islander * * * * * * * * * * 12
Native American * * * * * * * * * * *
Other * * * * * * * * * * 14
Missing * * * * * * * * * * 48
US Region 0.02
Northwest * 11 * * * * * * * * 58
Midwest * * * * * * * * 12 11 61
South * * 17 * * 16 14 * 13 13 109
West * * * * * 13 * * * * 69
Teaching hospital status NA
Non-teaching * * * * * * * NA NA NA 32
Teaching 25 22 25 19 18 35 20 NA NA NA 164
Admission type 0.97
Elective 20 16 23 13 14 28 15 17 25 26 197
Non-elective 13 18 * * * 13 * * 14 13 100
Mortality * * * * * * * * * * *

days), and incur higher inpatient costs ($49,610 more) than at greater risk of pPOA because of their higher exposure rates
those without pPOA. In the US adult population previously to previous anaesthetic and perioperative allergens compared
evaluated using the NIS database, similar findings were re- with younger patients aged <5 yr.
ported regarding longer length of hospital stay and higher total Interestingly, we found no sex preference for pPOA, con-
inpatient costs in patients with POA.11 trary to what has been seen in the adult population.11 In
Data regarding the incidence of pPOA are few, and pub- general, drug allergy is sex driven in adults, with more events
lished rates are quite variable. Toh and colleagues2 estimated occurring in females, the hypothesis being that female sex
an incidence of four in 10,000 cases based on 15 events of pPOA hormones play a role in development of immediate hyper-
in a retrospective review of 35,361 paediatric anaesthesia sensitivity.14,15 As a portion of our female patients were pre-
cases in a tertiary paediatric hospital in Singapore. Similarly, pubertal, a sex difference might not have been detectable. We
the 2017 European APRICOT study reported an incidence of expect prepubertal children of male and female sex to have
one in 10,000 cases.5 The NAP6 in the UK described an even similar hormonal regulation. Given there was not a higher
lower incidence of one in 37,000 cases.3 A French study of 2516 incidence of pPOA in males, we do not believe there is a hor-
patients from three combined databases over an 8-yr period monal switch, such as that seen in asthma epidemiology
included 266 paediatric patients and found 122 cases of where a greater burden of disease is noted in prepubertal
immunoglobulin E (IgE)-mediated POA, for an incidence of one males than in postpubertal females.16e21
in 20,000.6 Our study found an incidence of pPOA of approxi- The US region was found to be a significant factor in pPOA
mately one in 12,000 cases, which is closer to that reported by risk, with most cases occurring in the Southern region. The
Toh and colleagues2 and APRICOT.5 specific factors involved in more pPOA in the South, such as
Compared with other studies where the focus has been on more chronic lung disease, were skewed towards an older
clinical signs, physiologic variables, and the causative agents population, or underwent more high-risk procedures (i.e.
of anaphylaxis, the focus of our efforts was to identify risk transplant, cardiac, vascular), are unknown, but this would be
factors for surgical patients and categories of procedures that an interesting area for future investigation. Notably, there was
place them at higher risk of having pPOA. In general, older no geographic regional significance determined in the US adult
paediatric patients (aged 6e17 yr) were at greatest risk for population previously detected.11
pPOA. A previous study with combined data from the UK, USA, Transplant, cardiac, and vascular procedures were associ-
and France also found a significant difference in age of pa- ated with the highest risk for pPOA. A possible explanation is
tients with pPOA compared with controls (11 yr vs 4 yr, that these procedures are more complex and lengthy of all
respectively).4 We hypothesise that older patients are possibly procedure types evaluated, and typically require general
588 - Krase et al.

Table 3 Univariate and multivariable predictors of perioperative anaphylaxis. AIDS, acquired immunodeficiency syndrome; CI, con-
fidence interval; ENT, ears, nose, and throat; OBGYN, obstetrics and gynaecology; OR, odds ratio. *Data not available.

Characteristic Univariate Multivariable

OR (95% CI) P-value OR (95% CI) P-value

Age (yr)
<1 Reference Reference
1e5 24.64 (15.35e39.56) <0.001 4.56 (2.44e8.52) 0.45
6e12 32.27 (20.61e50.52) <0.001 7.09 (3.9e12.92) <0.001
13e17 23.13 (14.91e35.83) <0.001 8.46 (4.7e15.24) <0.001
Sex
Female 3.01 (2.40e3.78) <0.001 1.22 (0.91e1.63) 0.18
Male Reference Reference
Race * *
White 0.84 (0.65e1.08) 0.167
Non-White Reference
Teaching hospital status
Teaching 3.98 (2.73e5.82) <0.001 1.5 (1.01e2.23) 0.045
Non-teaching Reference Reference
Admission type
Elective 3.53 (2.78e4.49) <0.001 1.15 (0.83e1.58) 0.4
Non-elective Reference Reference
Procedure type
Breast/Skin 3.12 (2.11e4.60) <0.001 8.87 (4.33e18.18) <0.001
Cardiac 5.53 (3.55e8.62) <0.001 16.43 (7.52e35.87) <0.001
ENT 1.21 (0.62e2.35) 0.573 * *0
General surgery-gastrointestinal 1.94 (1.43e2.64) <0.001 3.54 (1.76e7.09) <0.001
Genitourinary-urology 0.02 (0.01e0.04) <0.001 0.73 (0.23e2.34) 0.6
Haematology 10.48 (6.01e18.28) <0.001 10.41 (3.8e28.54) <0.001
Musculoskeletal 2.81 (2.02e3.90) <0.001 4.5 (2.22e9.12) <0.001
Nervous system 2.42 (1.76e3.32) <0.001 6.75 (3.3e13.83) <0.001
OBGYN 0.25 (0.10e0.61) 0.002 * *
Ophthalmic 6.81 (2.54e18.27) <0.001 7.17 (0.92e55.82) 0.06
Thoracic 2.74 (1.46e5.14) 0.002 * *
Transplant 42.02 (27.44e64.35) <0.001 46.27 (20.8e102.9) <0.001
Vascular 4.68 (3.38e6.48) <0.001 15.15 (7.49e30.67) <0.001
Comorbidity measures
AIDS 73.01 (10.21e521.90) <0.001 12.28 (1.6e94.12) 0.02
Chronic blood loss anaemia 0.38 (0.05e2.67) 0.327 * *
Coagulopathy 13.62 (9.34e19.87) <0.001 2.15 (1.3e3.56) 0.003
Congestive heart failure 8.31 (2.66e25.92) <0.001 2.65 (0.81e8.63) 0.11
Chronic pulmonary Disease 6.27 (4.54e8.66) <0.001 2.25 (1.53e3.33) <0.001
Depression 2.70 (0.87e8.42) 0.087 * *
Diabetes mellitus 1.73 (0.24e12.34) 0.583 * *
Drug abuse 2.63 (0.65e10.56) 0.173 * *
Fluid and electrolyte disorders 7.24 (5.49e9.55) <0.001 1.89 (1.28e2.77) 0.001
Hypertension 11.30 (7.39e17.30) <0.001 1.49 (0.86e2.6) 0.16
Hypothyroidism 2.37 (0.59e9.50) 0.225 * *
Iron-deficiency anaemia 8.38 (6.09e11.55) <0.001 2.05 (1.33e3.15) 0.001
Liver disease 14.41 (7.13e29.09) <0.001 1.13 (0.4e3.15) 0.82
Lymphoma 20.86 (5.19e83.89) <0.001 2.26 (0.31e16.36) 0.42
Metastatic cancer 5.04 (1.26e20.27) 0.023 * *
Other neurological disorders 5.45 (3.43e8.68) <0.001 1 (0.52e1.98) 0.98
Obesity 3.43 (1.62e7.27) 0.001 0.85 (0.31e2.3) 0.82
Paralysis 3.77 (2.06e6.88) <0.001 1.18 (0.58e2.4) 0.65
Peripheral vascular disease 4.32 (0.61e30.78) 0.144 * *
Psychoses 9.71 (4.59e20.54) <0.001 2.02 (0.74e5.51) 0.17
Pulmonary circulation disorders 4.77 (1.19e19.17) 0.028 * *
Renal failure 4.96 (1.59e15.47) 0.006 * *
Rheumatoid arthritis/collagen 13.34 (4.28e41.62) <0.001 2.33 (0.51e8.6) 0.31
vascular disease
Solid tumour without metastasis 2.36 (0.33e16.72) 0.390 * *
Valvular disease 6.70 (2.77e16.23) <0.001 1.22 (0.3e5.1) 0.77
Weight loss 7.32 (4.11e13.04) <0.001 1.38 (0.64e2.9) 0.42
Paediatric perioperative anaphylaxis in the USA - 589

anaesthesia as compared with lower risk procedures that important limitation is the inability of this type of study to
utilise local or regional anaesthesia. Another explanation is allow identification of therapeutic agents or causative factors
that these procedures often require additional medications of anaphylaxis. Consequently, it can be difficult to identify
either in the immediate perioperative or intraoperative period, index cases of pPOA from recurrent episodes in the same pa-
such as contrast media, antibiotics, chemotherapeutics, and tient. Because of the nature of the NIS database, we cannot
other adjuvant drugs (beta blockers used in cardiac and determine whether a patient had multiple episodes of pPOA
vascular procedures, and diuretics used in renal transplants). during a single inpatient course, as each code is only entered
Chronic pulmonary disease was found to be a risk factor for once during a single hospital admission. We were also unable
pPOA, although we are unable to determine which specific to determine if a serum tryptase concentration was obtained
diagnoses constitute the highest risk. In children, a large or what other allergy evaluation might have taken place,
proportion of pulmonary disease is as a result of underlying either during the same admission or afterwards as an outpa-
asthma, with a smaller contribution from other conditions tient. Similarly, our study did not examine data to look for
such as cystic fibrosis. With respect to asthma, more than the causal links between anaphylaxis and other inpatient out-
presence of disease itself, uncontrolled asthma predisposes to comes as a result of the intrinsic limitations of the dataset. The
development of anaphylaxis and increased symptom severity, link between the comorbid conditions and pPOA might be
and this likely contributes to pPOA.22,23 Patients with severe influenced by the duration of hospitalisation and therapies
pulmonary disease might have developed cardiovascular and necessary for organ failure not directly linked to the periop-
respiratory compromise intraoperatively owing to their un- erative period. Finally, our study accounts for pPOA that
derlying comorbidity, but we were unable to separate this occurred in an inpatient setting only, and we cannot account
population from those experiencing true anaphylaxis. Further for cases occurring in outpatient centres.
epidemiological studies are needed to investigate asthma and
other pulmonary disorders as independent risk factors for
Conclusions
pPOA.
AIDS, coagulopathy, and fluid and electrolyte disorders As the first US national study of POA focusing solely on pae-
were also independent risk factors for pPOA. For coagulop- diatric patients, we identified the incidence of pPOA, signifi-
athy, this could be related to the underlying inflammatory cant patient characteristics, and higher risk procedures that
state present in coagulopathic patients which could lower the should help clinicians recognise susceptible patients. Our re-
threshold for mast cell and basophil-derived mediator release. sults suggest that certain populations at risk can be identified
This assumption lies in the well-established relationship be- during a preprocedural evaluation and might warrant more
tween mast cell activation and subsequent activation of the cautious intraoperative monitoring. Notification of potential
coagulation system.24,25 With fluid and electrolyte disorders, risk to the surgical team could be useful for better manage-
there might also be a connection between impaired haemo- ment should anaphylaxis occur. Accurate labelling of pPOA is
dynamics and a lowered threshold for activation of the me- important for epidemiologic studies, and it would be of inter-
diators of anaphylaxis. These comorbidities in relation to est to conduct a similar study using ICD-10 coding. Further
pPOA should be independently studied in future prospective large, multicentre, nationwide studies looking specifically at
studies. The link between AIDS and increased risk is not as paediatric populations would improve our understanding of
clear; however, a unifying factor in all of these comorbidities the epidemiology, risk factors, causative agents, and clinical
associated with higher risk of pPOA is their potential outcomes for pPOA.
compromise of physiologic accommodation to the effects of
anaphylaxis; although they might not have a direct effect on
Authors’ contributions
mast cell and basophil mediator release, they might interfere
with the ability to compensate once mediator release occurs. Study design: all authors
Our findings are consistent in many respects with the Data collection: IZK, AGE
findings of a previous retrospective analysis of POA in the Data analysis: all authors
adult population from the NIS over the same time period.11 Writing the first draft of the manuscript: IZK, AGE
This is important as the causative agents for anaphylaxis are Revising the manuscript critically for important intellectual
known to change over time.26 The areas of transplant, cardiac, content: all authors
and vascular surgery posed a greater risk of POA in adults, as All authors have revised and approved the final version of the
did chronic pulmonary disease.11 Adult patients with POA also manuscript for submission.
had longer median hospital stays and higher median inpatient
costs.4 Most notably, unlike the adult population, we did not
Declaration of interest
find that female sex posed a significant increased risk for
pPOA. The role of sex hormones on allergic sensitisation is not The authors declare that they have no relevant conflicts of
a variable we were able to analyse in our study, although this is interest.
an important area of future research.
There are several important limitations of our work, which
Funding
are intrinsic to retrospective observational studies. Firstly, it is
difficult to account for confounding variables that cannot be Mayo Clinic Foundation (MAR).
measured: timing of when anaphylaxis occurred, or whether
this was in the immediate perioperative period or at another
References
time point during the admission. Secondly, use of data from a
large database can lead to a lack of standardisation in terms of 1. Mertes PM, Ebo DG, Garcez T, et al. Comparative epide-
how diagnoses are coded (between healthcare systems and miology of suspected perioperative hypersensitivity re-
even between individual coders in the same system). Another actions. Br J Anaesth 2019; 123: e16e28
590 - Krase et al.

2. Toh TS, Foo SY, Loh W, et al. Perioperative anaphylaxis: a 13. Golden DBK, Wang J, Waserman S, et al. Anaphylaxis: a
five-year review in a tertiary paediatric hospital. Anaesth 2023 practice parameter update. Ann Allergy Asthma
Intensive Care 2021; 49: 44e51 Immunol 2024; 132: 124e76
3. Harper NJN, Cook TM, Garcez T, et al. Anaesthesia, sur- 14. Thong BY, Tan TC. Epidemiology and risk factors for drug
gery, and life-threatening allergic reactions: epidemiology allergy. Br J Clin Pharmacol 2011; 71: 684e700
and clinical features of perioperative anaphylaxis in the 15. Clark S, Wei W, Rudders SA, Camargo Jr CA. Risk factors
6th National Audit Project (NAP6). Br J Anaesth 2018; 121: for severe anaphylaxis in patients receiving anaphylaxis
159e71 treatment in US emergency departments and hospitals.
4. Khaleva E, Franz A, Garvey LH, et al. Perioperative J Allergy Clin Immunol 2014; 134: 1125e30
anaphylaxis in children: etiology, time sequence, and 16. Fuseini H, Newcomb DC. Mechanisms driving gender
patterns of clinical reactivity. Pediatr Allergy Immunol 2020; differences in asthma. Curr Allergy Asthma Rep 2017; 17: 19
31: 85e94 17. Vink NM, Postma DS, Schouten JP, Rosmalen JG,
5. Habre W, Disma N, Virag K, et al. Incidence of severe Boezen HM. Gender differences in asthma development
critical events in paediatric anaesthesia (APRICOT): a and remission during transition through puberty: the
prospective multicentre observational study in 261 hos- TRacking Adolescents’ Individual Lives Survey (TRAILS)
pitals in Europe. Lancet Respir Med 2017; 5: 412e25 study. J Allergy Clin Immunol 2010; 126: 498e504. e1e6
6. Mertes PM, Alla F, Trechot P, Auroy Y, Jougla E, Groupe 18. Rao CK, Moore CG, Bleecker E, et al. Characteristics of
d’Etudes des Reactions Anaphylactoides P. Anaphylaxis perimenstrual asthma and its relation to asthma severity
during anesthesia in France: an 8-year national survey. and control: data from the Severe Asthma Research Pro-
J Allergy Clin Immunol 2011; 128: 366e73 gram. Chest 2013; 143: 984e92
7. Stepanovic B, Sommerfield D, Lucas M, von Ungern- 19. Han YY, Forno E, Celedon JC. Sex steroid hormones and
Sternberg BS. An update on allergy and anaphylaxis asthma in a nationwide study of U.S. adults. Am J Respir
in pediatric anesthesia. Paediatr Anaesth 2019; 29: Crit Care Med 2020; 201: 158e66
892e900 20. DeBoer MD, Phillips BR, Mauger DT, et al. Effects of endog-
8. Tam J, Kelly K, Pongracic J. A seventeen-year review of enous sex hormones on lung function and symptom con-
peri-operative anaphylaxis in a pediatric hospital. J Allergy trol in adolescents with asthma. BMC Pulm Med 2018; 18: 58
Clin Immunol 2011; 127: AB247 21. Chen W, Mempel M, Schober W, Behrendt H, Ring J.
9. Healthy People 2030. The Healthcare Cost and Utilization Gender difference, sex hormones, and immediate type
Project. National (Nationwide) Inpatient Sample (NIS). hypersensitivity reactions. Allergy 2008; 63: 1418e27
Available from: https://health.gov/healthypeople/ 22. Tanno LK, Gonzalez-Estrada A, Olivieri B, Caminati M.
objectives-and-data/data-sources-and-methods. Asthma and anaphylaxis. Curr Opin Allergy Clin Immunol
[Accessed April 23, 2024] 2019; 19: 447e55
10. Vandenbroucke JP, von Elm E, Altman DG, et al. 23. Shaker MS, Wallace DV, Golden DBK, et al. Anaphylaxis-a
Strengthening the Reporting of Observational Studies in 2020 practice parameter update, systematic review, and
Epidemiology (STROBE): explanation and elaboration. Ann Grading of Recommendations, Assessment, Development
Intern Med 2007; 147: W163e94 and Evaluation (GRADE) analysis. J Allergy Clin Immunol
11. Gonzalez-Estrada A, Carrillo-Martin I, Renew JR, Rank MA, 2020; 145: 1082e123
Campbell RL, Volcheck GW. Incidence of and risk factors 24. Guilarte M, Sala-Cunill A, Luengo O, Labrador-Horrillo M,
for perioperative or periprocedural anaphylaxis in the Cardona V. The mast cell, contact, and coagulation sys-
United States from 2005 to 2014. Ann Allergy Asthma tem connection in anaphylaxis. Front Immunol 2017; 8: 846
Immunol 2021; 126: 180e186 e3 25. Khan BQ, Kemp SF. Pathophysiology of anaphylaxis. Curr
12. Walsh KE, Cutrona SL, Foy S, et al. Validation of anaphy- Opin Allergy Clin Immunol 2011; 11: 319e25
laxis in the Food and Drug Administration’s Mini- 26. Tacquard C, Serrier J, Viville S, et al. Epidemiology of
Sentinel. Pharmacoepidemiol Drug Saf 2013; 22: 1205e13 perioperative anaphylaxis in France in 2017e2018: the
11th GERAP survey. Br J Anaesth 2024; 132: 1230e7

Handling Editor: Hugh C Hemmings Jr

You might also like