Acute Adverse Reactions to Nonionic Low-Osmolar Iodinated Contrast Media in Children: A Retrospective Study of 23,429 Injections Over 6 Years

Abstract

Objective

To evaluate the incidence and risk factors of acute adverse drug reactions (ADRs) to nonionic low-osmolar iodinated contrast media (LOCM) in a pediatric population.

Materials and Methods

This single-center retrospective study included consecutive data on nonionic LOCM injections in pediatric patients (≤18 years) between January 1, 2016, and June 30, 2023. The per-examination incidences of acute ADRs (physiologic and allergic-like reactions) were assessed along with their severities. Multivariable Poisson regression analysis with generalized estimating equations was used to explore the risk factors associated with the occurrence of each ADR type.

Results

Among 23,429 injections administered to 13,172 pediatric patients, acute allergic-like reactions occurred in 0.89% (208/23,429) of cases, most of which were mild (0.79%; 186/23,429). Acute physiologic reactions occurred in 0.55% (128/23,429), also predominantly mild (0.50%, 118/23,429). Both types of ADRs occurred slightly more frequently in older patients (allergic-like reactions: adjusted relative risk (RR) = 1.09 per 1-year increase [95% confidence interval {CI}: 1.06, 1.11], P < 0.001; physiologic reactions: adjusted RR = 1.04 per 1-year increase [95% CI: 1.01, 1.07], P = 0.008) and more frequently in those with a history of such reactions (allergic-like reactions = 3.62 [95% CI: 1.44, 9.09], P = 0.006; physiologic reactions = 7.44 [95% CI: 2.55, 21.70], P < 0.001). Acute allergic-like reactions occurred less frequently in inpatient/emergency settings than in outpatient settings (adjusted RR = 0.64 [95% CI: 0.48, 0.85], P = 0.002). Among the 226 examinations with prior allergic-like reactions, recurrent acute allergic-like reactions occurred in 7.26% (9/124) when re-exposed to the same generic LOCM and in 3.92% (4/102) when exposed to a different generic LOCM (P = 0.28). Moderate reactions occurred in three cases that were re-exposed to the same LOCM, whereas all four reactions with different LOCM were mild.

Conclusion

Acute allergic-like and physiologic reactions to nonionic LOCM occurred in 0.89% and 0.55% of cases, respectively. Older age and a history of prior reactions were significant risk factors. Particular care is warranted during contrast-enhanced CT in these high-risk groups.

INTRODUCTION

Iodinated contrast media (ICM) are widely used in CT to enhance the soft-tissue contrast. Although adverse reactions to contrast media are rare and generally mild, they can occasionally result in life-threatening complications, such as anaphylaxis [1,2]. These reactions are typically classified as either allergic-like or physiologic. Although the exact mechanism of allergic-like reactions is not fully understood, the direct release of histamine is a key factor in their pathogenesis. In contrast, physiologic reactions are associated with specific molecular properties that induce direct chemotoxicity, osmotoxicity, and molecular binding to certain activators [1]. The American College of Radiology (ACR) categorizes these adverse reactions into the following three severity levels: mild, moderate, and severe [1].

According to McDonald et al. [3], the incidences of acute allergic-like and physiologic reactions in the adult population were 0.3% and 0.2%, respectively. Identified risk factors for these reactions include a prior history of nonionic low-osmolar iodinated contrast media (LOCM)-related reactions, asthma, drug allergies, patient status, and age [4,5,6,7,8]. In the pediatric population, adverse reactions to ICM are less common. A study by Callahan et al. [4] involving 12,494 pediatric patients reported an incidence of 0.46%, while Dillman et al. [9] documented acute allergic-like reactions in 0.18% of 11,306 injections. However, the low incidence of such reactions makes it difficult to precisely estimate their incidence because of the relatively small sample size, and meaningful analyses of risk factors remain limited. Furthermore, it remains unclear whether the risk factors identified in adults, such as a prior history of adverse reactions, are directly applicable to pediatric patients.

To address these gaps, this study evaluated the incidence and risk factors of acute adverse drug reactions (ADRs) to nonionic LOCM in a pediatric population by analyzing 23,429 injections administered to 13,172 patients.

MATERIALS AND METHODS

This single-center retrospective study was approved by the Institutional Review Board of Asan Medical Center (IRB No. 2023-0884), which waived the requirement for written informed consent. This study was conducted in accordance with the ACR manual and the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines [10].

Study Cohort

Pediatric patients (aged ≤18 years) who received intravenous administration of nonionic LOCM for enhanced CT examinations at our institution between January 1, 2016, and June 30, 2023, were considered eligible for the study. Data on CT examinations of pediatric patients were extracted from electronic medical records (Fig. 1). Unenhanced CT examinations and those with missing data regarding the type of LOCM used were excluded from the analysis. Examinations in which contrast media were not injected intravenously (i.e., CT myelography, cisternography, dacrocystography, and arthrography) were also excluded from the analysis. If patients underwent two or more consecutive contrast-enhanced CT examinations with a single intravenous LOCM administration (e.g., consecutive chest and abdominal CT scans), only the first examination was included in the analysis.

Fig. 1. Patient selection flowchart.

Clinical Data Retrieval

Clinical data, including patient age, sex, height, weight, serum creatinine, patient status at examination (outpatient vs. inpatient or emergency), generic profile, concentration and osmolality of LOCM, iodine dose per body weight (calculated by multiplying the concentration of LOCM by the volume administered per body weight), premedication, and type of CT examination (abdomen and pelvis, heart, head and neck, chest, musculoskeletal, and multiple) were extracted from electronic medical records. The estimated glomerular filtration rate (eGFR) was calculated using the bedside Schwartz equation as follows: eGFR = 0.413 x height (cm)/serum creatinine (mg/dL). When calculating the eGFR, height was used if available within 3 months of the CT examination. Additionally, the following factors known to be associated with acute ADR to LOCM were evaluated: prior history of LOCM-associated physiologic or allergic-like reactions, allergic rhinitis, arrhythmia, asthma, atopic dermatitis or urticaria, diabetes mellitus, drug allergy, food allergy, heart failure, hypertension, and hyperthyroidism.

CT Examination Protocol and LOCM Administration

All studies were conducted using prespecified pediatric protocols. In patients weighing ≤5 kg, LOCM was administered at a dose of 1.5 mL/kg over 15 seconds. Subsequently, a 30% mixture of contrast media and normal saline was injected over 5 seconds, followed by normal saline until the end of the CT examination. All injections were delivered at the same flow rate as that of the preceding contrast injection. In patients weighing >5 kg, LOCM was administered at a dose of 1.2 mL/kg for chest CT or 1.5 mL/kg for other types of CT examinations, injected over 20 seconds. Subsequently, a 10% mixture of contrast media and normal saline was injected over 10 seconds at the same flow rate as that in the preceding contrast injection. For example, in a 20 kg patient undergoing abdominal CT, a total of 30 mL of LOCM (i.e., 1.5 mL/kg x 20 kg) was administered over 20 seconds at a flow rate of 1.5 mL/sec, followed by 15 mL of a 10% mixture of contrast media and normal saline, injected over 10 seconds at a flow rate of 1.5 mL/sec.

A circular region of interest was placed in the left ventricle for patients weighing ≤5 kg or undergoing abdominal dynamic CT, and in the descending aorta for other types of examinations. The region of interest was monitored until the attenuation exceeded 150 Hounsfield units. The scan delay time varied between 6 and 30 seconds, depending on the type of CT examination. Different kVp settings were applied according to the body region: 70 kVp for the heart, abdomen, and pelvis; 80 kVp for the chest and neck; 100 kVp for brain, head, or neck angiography, orbit, and paranasal sinus; and 120 kVp for the temporal bone.

Identification and Classification of Acute Reactions

At our institution, contrast-enhanced CT examinations were performed using one of six nonionic LOCMs: iobitridol (Xenetix, Guerbet, Villepinte, France), iohexol (Bonorex, Central Medical Service, Seoul, Republic of Korea; Omnipaque, GE Healthcare, Chicago, IL, USA), iomeprol (Iomeron, Bracco, Milan, Italy), iopamidol (Iopamiro, Bracco; Pamiray, Dongkook Pharmaceutical, Seoul, Republic of Korea), iopromide (Ultravist, Bayer Healthcare, Leverkusen, Germany), and ioversol (Optiray, Mallinckrodt Pharmaceuticals, Hazelwood, MO, USA).

Data on the occurrence of LOCM-associated ADRs, including their type (i.e., allergic-like vs. physiologic reactions) and severity, were extracted from prospectively collected databases dedicated to contrast media-associated ADRs. At our institution, data on contrast media-associated ADRs are systematically recorded in a dedicated database embedded in electronic medical records. Patients were monitored during contrast media administration and for up to 10 minutes afterwards. Nursing staff continued to observe patients for any acute ADRs until they left the radiology department. Any reactions, including the occurrence of ADRs and their detailed descriptions, were documented in accordance with the recommendations of the ACR manual, whenever feasible [1]. Premedication (antihistamines with or without steroids) was indicated for all patients with a history of ADRs associated with LOCM. During data extraction, a pediatric radiologist (P.H.K.; 9 years of experience in radiology) reviewed the records and, if necessary, revised them based on the ACR manual. In cases where the distinction between physiologic and allergic-like reactions was unclear, the reactions were classified as physiologic, unless the symptoms clearly indicated an allergic-like reaction. Severity was also classified by the same pediatric radiologist, based on the details of acute adverse events and management performed.

Outcomes and Statistical Analysis

Continuous variables are expressed as mean and standard deviation (SD), while categorical variables are expressed as percentages and fractions. A frequency analysis was conducted on the incidence of acute allergic-like and physiologic reactions according to the generic profile of the LOCM. Factors independently associated with the occurrence of LOCM-associated acute allergic-like reactions and physiologic reactions were investigated using multivariable Poisson regression analysis with backward elimination among the following variables: patient age, sex, body indices (height, weight, and body mass index), calculated eGFR values (≤60 mL/min/1.73 m²), type of LOCM used, concentration and osmolality of LOCM, iodine dose per body weight, type of CT examination, patient status, history of ICM-associated reactions, and other medical history (i.e., allergic rhinitis, arrhythmia, asthma, atopic dermatitis or urticaria, diabetes mellitus, drug allergy, food allergy, heart failure, hypertension, and hyperthyroidism). Missing data for continuous variables were handled using a single imputation with the mean value. Generalized estimating equations were applied to account for intra-patient correlations due to repeated measures. Variables with a P-value <0.05 in univariable analyses were entered into the multivariable analysis. Additionally, prior history of allergic-like reactions, which is known to be the most substantial risk factor for recurrent acute allergic-like reactions [1], was entered into the multivariable analysis for the investigation of factors associated with the occurrence of acute allergic-like reactions, regardless of its P-value in univariable analysis.

Further analysis investigated the incidence and severity of recurrent acute allergic-like reactions, comparing cases re-exposed to the same generic LOCMs and those exposed to a different generic LOCM, using the chi-square test. Additionally, the incidence of acute ADRs associated with LOCMs was stratified based on the use of premedication and its clinical indications.

Statistical analyses were performed using R Version 4.0.3 (R Foundation for Statistical Computing, Vienna, Austria). Statistical significance was set at P < 0.05.

RESULTS

Patients

During the study period, 48,883 CT examinations were performed on pediatric patients. After applying our eligibility criteria, 23,429 LOCM injections (iomeprol, n = 11,910; iopamidol, n = 6,217; iohexol, n = 4,221; ioversol, n = 989; iobitridol, n = 85; iopromide, n = 7) administered to 13,172 pediatric patients were selected for analysis, with these including instances of re-exposure to contrast-enhanced CT examination (Fig. 1).

The detailed characteristics of the examinations included are presented in Table 1. In brief, 13,028 injections (55.6%, 13,028 of 23,429) were administered to boys and 10,401 (44.4%, 10,401 of 23,429) to girls. Regarding patient status during CT examination, 65.3% (15,294 of 23,429) of the examinations were performed in inpatient or emergency settings, whereas the remaining 34.7% (8,135 of 23,429) were performed in outpatient settings. The most frequently performed type of CT examination was abdomen and pelvis CT (48.9%, 11,448 of 23,429), followed by heart CT (19.4%, 4,543 of 23,429).

Table 1. Baseline characteristics of 23,429 ICM injections in the pediatric population.

Characteristic		Value
Age, yrs		8.7 ± 6.5
Sex
	Boys	13,028 (55.6)
	Girls	10,401 (44.4)
Height, cm (n = 20,680)		121.5 ± 41.3
Weight, kg (n = 22,189)		33.0 ± 24.0
Body mass index, kg/m2 (n = 20,676)		18.1 ± 4.6
Calculated eGFR, mL/min/1.73 m2 (n = 6,962)		101.3 ± 36.9
LOCM used
	Iomeprol	11,910 (50.8)
	Iopamidol	6,217 (26.5)
	Iohexol	4,221 (18.0)
	Ioversol	989 (4.2)
	Iobitridol	85 (0.4)
	Iopromide	7 (0.02)
Concentration of LOCM, mgI/mL		319.7 ± 31.2
Iodine dose per body weight, mgI/kg		471.9 ± 54.5
Osmolality of LOCM, mOsm/kg water		630.5 ± 118.4
Patient status at CT examination
	Inpatient or emergency	15,294 (65.3)
	Outpatient	8,135 (34.7)
Type of CT examinations
	Abdomen and pelvis	11,448 (48.9)
	Heart	4,543 (19.4)
	Head and neck	3,566 (15.2)
	Chest	3,387 (14.5)
	Musculoskeletal	298 (1.3)
	Multiple	187 (0.8)
Prior ICM-associated allergic-like reaction		226 (1.0)
Prior ICM-associated physiologic reaction		138 (0.6)
Allergic rhinitis		633 (2.7)
Arrhythmia		60 (0.3)
Asthma		564 (2.4)
Atopic dermatitis or urticaria		933 (4.0)
Diabetes mellitus		187 (0.8)
Drug allergy		388 (1.7)
Food allergy		136 (0.6)
Heart failure		418 (1.8)
Hypertension		556 (2.4)
Hyperthyroidism		31 (0.1)

Data are presented as mean ± standard deviation or number of injections with percentage in parentheses.

ICM = iodinated contrast media, eGFR = estimated glomerular filtration rate, LOCM = low-osmolar iodinated contrast media

Age distribution and LOCM used differed according to the type of CT examination (Supplementary Table 1). Notably, patients who underwent heart CT (mean age ± SD, 3.5 ± 5.4 years) were younger than those undergoing other types of CT examinations. Additionally, iopamidol was the most frequently used LOCM for heart (87.6%, 3,981 of 4,543) and musculoskeletal (70.5%, 210 of 298) CTs, whereas iomeprol was the most frequently used LOCM for abdomen and pelvis (61.7%, 7,060 of 11,448), head and neck (55.2%, 1,968 of 3,566), and chest (72.2%, 2,446 of 3,387) CTs.

ICM-Associated Acute Allergic-Like Reactions

The incidences of acute allergic-like reactions according to the type of LOCM used are described in Table 2 and Figure 2. Among the 23,429 injections, 0.89% (208 of 23,429) resulted in allergic-like reactions, most of which were mild (0.79% of all injections, 186 of 23,429), followed by moderate (0.07%, 17 of 23,429), and severe reactions (0.02%, 5 of 23,429). Acute allergic-like reactions occurred in 0.96% of all injections of iomeprol (114 of 11,910), 0.53% of iopamidol (33 of 6,217), 1.16% of iohexol (49 of 4,221), and 1.21% of ioversol (12 of 989). Acute allergic-like reactions did not occur with iobitridol (0 of 85) or iopromide (0 of 7).

Table 2. Summary of acute allergic-like and physiologic reaction rates associated with LOCM in the pediatric population.

LOCM type	Acute allergic-like reaction				Acute physiologic reaction
	Total	Mild	Moderate	Severe	Total	Mild	Moderate	Severe
Iomeprol	0.96 (114/11,910) [0.79, 1.15]	0.82 (98/11,910) [0.67, 1.00]	0.12 (14/11,910) [0.06, 0.20]	0.02 (2/11,910) [0, 0.06]	0.62 (74/11,910) [0.49, 0.78]	0.59 (70/11,910) [0.46, 0.74]	0.03 (4/11,910) [0.01, 0.09]	0
Iopamidol	0.53 (33/6,217) [0.37, 0.74]	0.53 (33/6,217) [0.37, 0.74]	0	0	0.32 (20/6,217) [0.20, 0.50]	0.26 (16/6,217) [0.15, 0.42]	0.06 (4/6,217) [0.02, 0.16]	0
Iohexol	1.16 (49/4,221) [0.86, 1.53]	1.04 (44/4,221) [0.76, 1.40]	0.05 (2/4,221) [0.01, 0.17]	0.07 (3/4,221) [0.01, 0.21]	0.64 (27/4,221) [0.42, 0.93]	0.59 (25/4,221) [0.38, 0.87]	0.02 (1/4,221) [0, 0.13]	0.02 (1/4,221) [0, 0.13]
Ioversol	1.21 (12/989) [0.63, 2.11]	1.11 (11/989) [0.56, 1.98]	0.10 (1/989) [0, 0.56]	0	0.61 (6/989) [0.22, 1.32]	0.61 (6/989) [0.22, 1.32]	0	0
Iobitridol	0 (0/85) [0, 4.25]	0	0	0	1.18 (1/85) [0.03, 6.38]	1.18 (1/85) [0.03, 6.38]	0	0
Iopromide	0 (0/7) [0, 40.96]	0	0	0	0	0	0	0
Total	0.89 (208/23,429) [0.77, 1.02]	0.79 (186/23,429) [0.68, 0.92]	0.07 (17/23,429) [0.04, 0.12]	0.02 (5/23,429) [0.01, 0.05]	0.55 (128/23,429) [0.46, 0.65]	0.50 (118/23,429) [0.42, 0.60]	0.04 (9/23,429) [0.02, 0.07]	0.01 (1/23,429) [0, 0.02]

Data are shown as percentages, with numerators and denominators in parentheses, and 95% confidence intervals in brackets.

LOCM = low-osmolar iodinated contrast media

Fig. 2. Bar graph depicting the incidence of acute allergic-like reactions associated with different iodinated contrast media. Values in parentheses indicate numerators and denominators, and values in brackets indicate the 95% confidence intervals of the corresponding proportions.

Multivariable Poisson regression analyses (Table 3) showed that acute allergic-like reactions occurred slightly more frequently in older patients with adjusted relative risk [RR] = 1.09 per 1-year increase (95% confidence interval [CI]: 1.06, 1.11, P < 0.001). When stratified by age group, the incidence of allergic-like reactions was 0.2% (10 of 5,038) in patients <2 years old, 1.3% (109 of 8,439) in those aged 2–12 years, and 0.9% (89 of 9,952) in those >12 years old (Supplementary Table 2). Additionally, patients with a history of allergic-like reactions showed a significantly higher incidence of acute allergic-like reactions (adjusted RR = 3.62 [95% CI: 1.44, 9.09], P = 0.006). The incidence of acute allergic-like reactions varied according to the patient status during CT examination, with a lower incidence being observed in inpatient or emergency settings than in outpatient settings (adjusted RR = 0.64 [95% CI: 0.48, 0.85], P = 0.002).

Table 3. Results of Poisson regression for the risk of acute allergic-like reactions associated with LOCM.

Variables		Univariable		Multivariable
		Unadjusted RR	P	Adjusted RR	P
Age, yrs		1.09 (1.07, 1.12)	<0.001	1.09 (1.06, 1.11)	<0.001
Sex
	Boys	Reference category
	Girls	1.05 (0.79, 1.40)	0.75
Height (n = 20,680), cm*		1.017 (1.013, 1.021)	<0.001	Eliminated
Weight (n = 22,189), kg*		1.018 (1.015, 1.022)	<0.001	Eliminated
Body mass index (n = 20676), kg/m2*		1.06 (1.04, 1.08)	<0.001	Eliminated
Calculated eGFR, <60 mL/min/1.73m2		0.34 (0.11, 1.09)	0.07
LOCM used				Eliminated
	Iomeprol	Reference category
	Iopamidol	0.56 (0.38, 0.82)	0.003
	Iohexol	1.13 (0.79, 1.61)	0.51
	Ioversol	1.23 (0.69, 2.22)	0.48
	Iobitridol	Not estimable
	Iopromide	Not estimable
Concentration of LOCM, mgI/mL†		0.37 (0.22, 0.63)	<0.001	Eliminated
Iodine dose per body weight, mgI/kg†		0.57 (0.44, 0.74)	<0.001	Eliminated
Osmolality of LOCM, mOsm/kg water†		0.85 (0.75, 0.96)	0.006	Eliminated
Patient status at CT examination
	Outpatient	Reference category		Reference category
	Inpatient or emergency	0.53 (0.40, 0.70)	<0.001	0.64 (0.48, 0.85)	0.002
Type of CT examination
	Abdomen and pelvis	Reference category		Excluded
	Heart	0.33 (0.20, 0.57)	<0.001
	Head and neck	0.82 (0.55, 1.22)	0.32
	Chest	1.22 (0.84, 1.78)	0.30
	Musculoskeletal	1.66 (0.69, 3.99)	0.26
	Multiple	1.08 (0.28, 4.23)	0.92
Prior ICM-associated physiologic reaction		2.98 (1.06, 8.42)	0.04	Eliminated
Prior ICM-associated allergic-like reaction‡		3.06 (0.77, 12.08)	0.11	3.62 (1.44, 9.09)	0.006
Allergic rhinitis		1.67 (0.87, 3.22)	0.12
Arrhythmia		Not estimable
Asthma		1.05 (0.43, 2.54)	0.91
Atopic dermatitis or urticaria		1.37 (0.75, 2.48)	0.31
Diabetes mellitus		1.94 (0.66, 5.71)	0.23
Drug allergy		0.88 (0.28, 2.73)	0.82
Food allergy		0.82 (0.13, 5.23)	0.83
Heart failure		Not estimable
Hypertension		0.44 (0.12, 1.65)	0.22
Hyperthyroidism		Not estimable

^*Missing data were handled using single imputation with the mean value, ^†Concentration of LOCM, iodine dose per body weight, and osmolality of LOCM were divided by 100 to provide clinically informative results of RR, ^‡Of the 226 examinations in this category for which premedication was indicated, 172 were performed following premedication.

LOCM = low-osmolar iodinated contrast media, RR = relative risk, eGFR = estimated glomerular filtration rate, ICM = iodinated contrast media

Among the 23,429 examinations, 226 were indicated for premedication owing to a prior history of allergic-like reactions associated with LOCM. Of these, 172 examinations were performed after premedication. Acute allergic-like reactions occurred in 6.40% (11 of 172) of examinations performed with premedication and in 3.70% (2 of 54) of examinations for which premedication was indicated but not administered. The incidence of acute allergic-like reactions remained significantly higher in patients with a history of such reactions, even with premedication (6.40%), compared to that of those without such a history (0.84%, 195 of 23,203) (P < 0.001).

ICM-Associated Acute Physiologic Reactions

The incidences of acute physiologic reactions according to the type of LOCM used are described in Table 2 and Figure 3. The incidence of acute physiologic reactions among all LOCM injections was 0.55% (128 of 23,429), with the majority being mild (0.50% of all injections, 118 of 23,429), followed by moderate (0.04%, 9 of 23,429), and severe (0.01%, 1 of 23,429). Acute physiologic reactions occurred in 0.62% of iomeprol injections (74 of 11,910), 0.32% of iopamidol injections (20 of 6,217), 0.64% of iohexol injections (27 of 4,221), 0.61% of ioversol injections (6 of 989), and 1.18% of iobitridol injections (1 of 85). No acute physiologic reactions were observed with iopromide (0 of 7).

Fig. 3. Bar graph illustrating the incidence of acute physiologic reactions associated with different iodinated contrast media. Values in parentheses indicate numerators and denominators, and values in brackets indicate the 95% confidence intervals of the corresponding proportions.

Multivariable Poisson regression analyses (Table 4) showed that acute physiologic reactions occurred slightly more frequently in older patients (adjusted RR = 1.04 per 1-year increase [95% CI: 1.01, 1.07], P = 0.008). When stratified by age group, the incidence of physiologic reactions was 0.08% (4 of 5,038) in patients <2 years old, 0.60% (51 of 8,439) in those aged 2–12 years, and 0.73% (73 of 9,952) in those >12 years old (Supplementary Table 2). Additionally, acute physiologic reactions were more frequent in patients with a prior history of such reactions (adjusted RR = 7.44 [95% CI: 2.55, 21.70], P < 0.001). Heart CT examinations showed a lower incidence of acute physiologic reactions than abdomen and pelvis CT examinations (adjusted RR = 0.22 [95% CI: 0.09, 0.56], P = 0.002); however, this association was not statistically significant when stratified by age (Supplementary Table 3).

Table 4. Results of Poisson regression for the risk of acute physiologic reactions associated with LOCM.

Variables		Univariable		Multivariable
		Unadjusted RR	P	Adjusted RR	P
Age, yrs		1.06 (1.04, 1.09)	<0.001	1.04 (1.01, 1.07)	0.008
Sex
	Boys	Reference category
	Girls	0.89 (0.60, 1.30)	0.54
Height (n = 20,680), cm*		1.015 (1.010, 1.019)	<0.001	Eliminated
Weight (n = 22,189), kg*		1.018 (1.012, 1.025)	<0.001	Eliminated
Body mass index (n = 20,676), kg/m2*		1.08 (1.05, 1.11)	<0.001	Eliminated
Calculated eGFR <60 mL/min/1.73 m2		0.19 (0.03, 1.34)	0.09
	LOCM used			Eliminated
	Iomeprol	Reference category
	Iopamidol	0.51 (0.31, 0.84)	0.009
	Iohexol	0.96 (0.59, 1.55)	0.86
	Ioversol	0.92 (0.39, 2.21)	0.85
	Iobitridol	1.53 (0.24, 9.56)	0.65
	Iopromide	Nonestimable
Concentration of LOCM, mgI/mL†		0.29 (0.14, 0.60)	0.001	Eliminated
Iodine dose per body weight, mgI/kg†		0.57 (0.42, 0.77)	<0.001	Eliminated
Osmolality of LOCM, mOsm/kg water†		0.79 (0.67, 0.93)	0.004	Eliminated
Patient status at CT examination
	Outpatient	Reference category
	Inpatient or emergency	0.80 (0.55, 1.16)	0.23
Type of CT examination
	Abdomen and pelvis	Reference category		Reference category
	Heart	0.16 (0.07, 0.40)	<0.001	0.22 (0.09, 0.56)	0.002
	Head and neck	0.70 (0.41, 1.19)	0.19	0.72 (0.42, 1.23)	0.23
	Chest	1.00 (0.64, 1.59)	0.99	1.06 (0.67, 1.68)	0.80
	Musculoskeletal	0.54 (0.09, 3.29)	0.50	0.49 (0.07, 3.33)	0.47
	Multiple	2.14 (0.62, 7.30)	0.23	1.93 (0.57, 6.58)	0.29
Prior ICM-associated physiologic reaction‡		6.74 (1.57, 29.01)	0.01	7.44 (2.55, 21.70)	<0.001
Prior ICM-associated allergic-like reaction		1.63 (0.21, 12.76)	0.64
Allergic rhinitis		0.23 (0.04, 1.36)	0.10
Arrhythmia		Not estimable
Asthma		0.92 (0.27, 3.12)	0.89
Atopic dermatitis or urticaria		0.59 (0.20, 1.77)	0.35
Diabetes mellitus		Not estimable
Drug allergy		0.81 (0.18, 3.59)	0.78
Food allergy		1.51 (0.23, 9.85)	0.67
Heart failure		Not estimable
Hypertension		0.37 (0.04, 3.56)	0.39
Hyperthyroidism		5.77 (0.81, 41.19)	0.08

^*Missing data were handled using single imputation with the mean value, ^†Concentration of LOCM, iodine dose per body weight, and osmolality of LOCM were divided by 100 to provide clinically informative results of RR, ^‡Of the 138 examinations in this category for which premedication was indicated, 84 were performed following premedication.

LOCM = low-osmolar iodinated contrast media, RR = relative risk, eGFR = estimated glomerular filtration rate, ICM = iodinated contrast media

Among the 23,429 examinations, 138 were indicated for premedication because of a history of physiologic reactions associated with LOCM. Of these, 84 examinations were performed after premedication. Acute physiologic reactions occurred in 9.52% (8 of 84) of examinations with premedication and in 5.56% (3 of 54) of examinations in which premedication was indicated but not administered. The incidence of acute physiologic reactions remained significantly higher in patients with a history of such reactions, even with premedication (9.52%), compared to that of those without such a history (0.50%, 117 of 23,291) (P < 0.001).

Recurrent Acute Allergic-Like Reactions: Same vs. Different Generic LOCM

Among the 226 examinations with a prior history of allergic-like reactions, 124 were re-exposed to the same generic LOCM, whereas the remaining 102 were exposed to a different generic LOCM. Overall, recurrent acute allergic-like reactions occurred in 5.75% (13 of 226) of cases with a prior history of allergic-like reactions. Specifically, recurrent acute allergic-like reactions occurred in 7.26% (9 of 124) of cases re-exposed to the same generic LOCM and in 3.92% (4 of 102) of cases exposed to a different generic LOCM, with no significant difference between the two groups (P = 0.28). However, focusing on the severity of the reactions, three of the recurrent acute allergic-like reactions in case re-exposed to the same generic LOCMs were moderate, whereas all four recurrent acute allergic-like reactions in cases exposed to different generic LOCMs were mild.

DISCUSSION

To date, data on the safety of LOCM in the pediatric population have been limited to a few retrospective studies [4,9], and the data on the incidence of acute physiologic reactions in children is scarce. Furthermore, risk factors associated with the occurrence of LOCM-associated acute ADRs have rarely been analyzed. In this retrospective study, including 23,429 LOCM injections, acute allergic-like reactions occurred in 0.89% of the injections, and acute physiologic reactions occurred in 0.55%, with most reactions being mild (0.79% of all injections for allergic-like reactions; 0.50% for physiologic reactions).

Both acute allergic-like and physiologic reactions were more common in older patients, consistent with the results of previous studies. In a study conducted by Katayama et al. [2], the incidence of ADRs increased with age among patients under 20 years of age (<1 year: 0.44%, 1–9 years: 2.52%, 10–19 years: 4.51%). Similarly, Callahan et al. [4] reported that the mean age of patients who experienced contrast media reactions (mean age, 12.9 ± 4.3 years) was 3.4 years older than the mean age of all patients (9.5 ± 5.9 years) included in their study.

We also found that a prior history of allergic-like reactions was an independent risk factor for such reactions. Recurrent acute allergic-like reactions occurred in 13 examinations in patients with a history of allergic-like reactions (5.75%), and nine of these (69%) were re-exposed to the same generic LOCM that was associated with their prior allergic-like reactions. Three of these nine examinations were moderate in severity. Our study demonstrated that a prior history of ADRs to LOCM is a risk factor for recurrent ADRs in pediatric patients, which is limited to studies conducted on adult patients [1]. Thus, children with a prior history of acute allergic-like reactions require special care and preventive measures, including premedication, skin prick tests (especially in cases with a history of severe reactions), and close monitoring of their condition after contrast media administration [1,11,12,13].

Several studies have reported that substituting the originally used contrast agent with a different type is effective in preventing recurrent allergic-like reactions in adults [14,15,16,17]. Lim et al. [16] reported a 69% reduction in recurrent ADRs when the original contrast agent was substituted with a different type and an additional 35% reduction when using a generic LOCM lacking a common carbamoyl side chain. Similarly, in a study on switching to contrast media based on sharing an N-(2,3-dihydroxypropyl) carbamoyl side chain, Kim et al. [14] found that switching LOCMs was associated with a lower recurrence rate (switching, 10.4% vs. non-switching, 28.4%), with a further reduction when different side chain groups were used (from 7.8% to 5.0%). Umakoshi et al. [17] reported that switching LOCM reduced the risk of recurrent hypersensitivity reactions by 61%, whereas Kim et al. [15] found that using a different LOCM without a common side chain reduced recurrence by 49%. Although no statistically significant difference was observed in our study, recurrent acute allergic-like reactions occurred more frequently when patients were re-exposed to a contrast agent that was previously associated with an allergic-like reaction (7.3% vs. 3.9%, P = 0.28). Notably, one-third of cases were classified as moderate in severity. Therefore, substitution with a different generic LOCM may be effective in reducing the rate of recurrent acute allergic-like reactions in pediatric patients. Larger cohort studies are required to address this issue.

We also found that a prior history of physiologic reactions was a risk factor for these reactions. Unlike allergic-like reactions, the occurrence of physiologic reactions is influenced by the specific chemical properties of CT contrast media and is therefore often dose- and concentration-dependent. Although the exact cause is uncertain, some allergic-like reactions in children might be misclassified as physiologic reactions owing to their vague symptom descriptions. Alternatively, psychological factors such as fear and apprehension might act as risk factors [18], possibly contributing to recurrence in more sensitive pediatric patients. Therefore, emotional reassurance and careful monitoring may be required for children with a prior history of acute physiologic reactions.

Our study showed a higher incidence of acute allergic-like reactions in outpatient settings than in inpatient or emergency settings (outpatient, 1.29% [105 of 8,135] vs. inpatient or emergency, 0.67% [103 of 15,294]), which is in line with studies by Mortelé et al. [8] and Chatani et al. [6], who both identified outpatient status as a significant factor (odds ratio [OR] of 7.5, 95% CI: 4.3–11.2; and 2.08, 95% CI: 1.20–3.60, respectively). Additionally, Chen et al. [7] reported that the incidence of ADRs was higher in outpatients and emergency patients than in inpatients, both in 2014 (0.59% vs. 0.30%) and 2014–2015 (0.46% vs. 0.28%). Although the contrasting patient groups differed between studies, with our study comparing outpatients with inpatients/emergency patients and other studies comparing outpatients with inpatients or outpatients/emergency patients with inpatients, when factors such as the relatively small number of emergency patients are considered, these studies suggest that outpatient status could be a significant risk factor for ADRs. This raises the possibility of potential confounding factors between the two groups, such as variations in age distribution or differences in workflow and examination procedures, which could affect the identification of allergic-like reactions [7]. Additionally, because of the short interview time available for healthcare providers to understand the medical information of outpatients, important risk factors such as a history of allergic-like reactions may have been missed [6]. Nevertheless, radiology staff should pay more attention to outpatient examination procedures. Further studies are needed to elucidate the exact reasons for the differences in the prevalence of allergic-like reactions according to patient status.

In our study, institutional practice patterns related to the type of CT examination may have acted as confounders in the risk factor analysis for the occurrence of ICM-associated acute ADRs. Acute physiologic reactions were less frequent on heart CT than on abdomen and pelvis CT (adjusted RR = 0.22; P = 0.002); however, this association was not evident when stratified by age. This suggests that the type of CT examination might be a confounder in the occurrence of such reactions, as the mean age of the patients undergoing heart CT (3.5 years) was younger than that of those undergoing abdomen and pelvis CT (10.3 years). Similarly, in the univariable analysis, LOCM concentration, iodine dose per body weight, and osmolality were associated with a lower risk of both acute allergic-like and physiologic reactions. This might be attributed to the fact that higher concentrations of LOCMs (350–370 mg I/mL) are commonly used for heart CT, which in turn is more frequently used in younger patients. Among the various CT examination types, heart CT showed the highest mean concentration, iodine dose per body weight, and LOCM osmolality. Therefore, careful interpretation of these results is necessary.

This study has several limitations. First, nursing staff did not actively monitor all patients; instead, they introduced the potential side effects of contrast agents before the examination and relied on patients and caregivers to report their symptoms. Consequently, the incidence of adverse effects might have been underestimated. Second, the classification of the severity or type of ADRs may vary depending on the judgment of the nursing staff who documented it. To reduce the potential for misclassification, an experienced radiologist reviewed and revised the data, when necessary. Nevertheless, in cases where documentation was unclear or ambiguous, classification may depend on subjective judgment, which introduces the possibility of the misclassification of acute adverse events. Third, ADRs may have been misclassified or underestimated because it was difficult to accurately describe ADRs in children. Finally, the low incidence of acute ADRs presents a challenge in achieving statistical power in adult studies, even when approximately 25,000 examinations are included. Additionally, the infrequent use of certain contrast media, such as iopromide, restricted the accurate evaluation of the incidence of acute ADRs.

In conclusion, acute allergic-like reactions occurred in 0.89% of cases administered LOCM, and acute physiologic reactions occurred in 0.55%. Both acute allergic-like and physiologic reactions were slightly more common in older patients and were more frequent in those with a history of such reactions. Additionally, acute allergic-like reactions are less likely to occur in inpatient or emergency settings.

Supplement

The Supplement is available with this article at https://doi.org/10.3348/kjr.2025.0263.

Availability of Data and Material

The datasets generated or analyzed during the study are available from the corresponding author on reasonable request.