Comparison of negative predictive values of single- and two-day provocation tests with suspected nonsteroidal anti-inflammatory drug and paracetamol allergy in children

Abstract

Background:

Drug provocation tests (DPT) are the criterion standard method for diagnosing nonsteroidal anti-inflammatory drugs (NSAID) and paracetamol hypersensitivity reactions in children. However, there is no consensus in the literature with regard to the duration of DPTs.

Objective:

The objective was to compare the negative predictive values (NPV) of single- and 2-day DPTs for NSAID and paracetamol hypersensitivity diagnosis in pediatric patients.

Methods:

We retrospectively evaluated children (ages 1–18 years) with a history of NSAID and paracetamol hypersensitivity. The patients were categorized into two groups based on drug provocation duration: short (single-day test) and extended (test continued on the second day at home). Patients with negative DPT results for the suspected agent were contacted to determine whether they reused the drug and, if so, whether there was a reaction. The NPVs of the DPTs performed for both groups were calculated.

Results:

The DPT results of 104 patients (53.8% boys) were negative for 116 suspected agents: 67 (57.7%) tested with short DPT and 49 (42.2%) with extended DPT. No significant differences in age, sex, reaction type, or comorbidities were observed between the two groups. In the follow-up, 114 DPTs were performed for 102 patients, of whom 93 used the suspected drug(s) after the tests but none developed a reaction. The NPV was the same for both groups: 100%.

Conclusion:

To the best of our knowledge, this is the first study to compare the NPVs of single- and 2-day DPTs for children who present with suspected NSAID and paracetamol hypersensitivity. Our results indicate that both approaches have the same NPV and suggest that single-day DPT is sufficient to exclude suspicion of NSAID and paracetamol hypersensitivity in children.

Nonsteroidal anti-inflammatory drugs (NSAID) and paracetamol are safe and widely used worldwide by children for their antipyretic, analgesic, and anti-inflammatory effects.^1,2 Accordingly, they rank first after β-lactam antibiotics in terms of drug-related hypersensitivity reactions, and, in the general pediatric population, the prevalence rate of NSAID hypersensitivity (NSAID-H) is estimated to be 0.6%–5.7%.^3,4 However, among children with asthma, the prevalence ranges from 0% to 33%, depending on the study and population examined.^5,6

NSAID-H reactions present diversely, ranging from maculopapular exanthema and nonimmediate urticaria to life-threatening conditions, e.g., anaphylaxis, with children predominantly experiencing cutaneous symptoms, such as angioedema and urticaria.^7–9 According to the European Academy of Allergy and Clinical Immunology/European Network of Drug Allergy (ENDA) guidelines,¹⁰ NSAID-H reactions are classified into two main groups: cross-intolerant (nonimmunologic) and non–cross-intolerant (immunologic reactions. Cross-intolerant reactions encompass non-immunologic conditions such as NSAID-exacerbated respiratory disease, NSAID-exacerbated cutaneous disease and NSAID-induced urticaria/angioedema. These conditions are triggered by multiple, chemically unrelated NSAIDs via Cyclooxygenase-1 (COX-1) inhibition. By contrast, non–cross-intolerant reactions are immunologically mediated (immunoglobulin E or T-cell mediated) and occur in response to a specific NSAID or a chemically related group. Non–cross-intolerant reactions include single NSAID-induced urticaria/angioedema or anaphylaxis as well as single NSAID-induced delayed hypersensitivity reactions. The current criterion standard diagnostic method for children is the drug provocation test (DPT), and a strong history of NSAID-H with different phenotypes supports the diagnosis.¹¹ There is no consensus with regard to the optimal approach to DPT in children with NSAIDs. Specifically, there is ambiguity with regard to whether these tests should be performed with a suspected agent or an alternative drug as well as on the appropriate doses, dose intervals, and duration of testing.¹² In a study conducted by Mori et al.¹³ in Europe with multicenter pediatric patients, the test duration varied among countries, which indicated that a consensus on this issue remains elusive.

This study aimed to compare the negative predictive values (NPV) of single- and 2-day DPT performed on patients admitted to our clinics with suspicion of NSAID-H (including paracetamol).

METHODS

This study included all patients who had undergone DPTs for suspected NSAID-H (including paracetamol) and had a negative DPT result for the suspected agent in the Department of Pediatric Allergy and Immunology, Selcuk University, and Department of Pediatric Allergy and Immunology, Erzurum City Hospital, between September 1, 2021, and September 31, 2024. Patients who underwent DPT with alternative drugs were excluded from the study. Age, sex, the presence of concomitant asthma, allergic rhinitis, atopy, and chronic urticaria were recorded from patient records. A detailed history was taken, including the spectrum and timing of symptoms, the dose and route of administration, the history and timing of previous reactions, and previously tolerated NSAIDs.

DPT

DPT was performed in a hospital setting by an experienced allergist (I.K.C.) according to the European Academy of Allergy and Clinical Immunology/ENDA’s recommendations.¹⁰ The tests were conducted 4 to 6 weeks after the patients’ index reactions. Before DPT, all drugs that could affect the test were stopped and the suspected drug was administered orally until a cumulative dose close to the age- and weight-adjusted daily dose of the drug was achieved. Forced expiratory volume in 1 second was measured by spirometry at the beginning of the test and before each dose. Informed consent for DPT was obtained from the patients’ parents and/or carers. DPT was performed by using two different protocols:

1. Short protocol (single day): DPT was performed with the suspected NSAID (including paracetamol). The test was started with 1/10 of the daily dose, followed by 2/10 and 7/10 at 1-hour intervals for a total of three doses. The patients were closely monitored during the test and up to 3 hours after the last dose. No further doses were administered at home.

2. Extended protocol (2 days): patients who underwent the short DPT protocol and had negative results were instructed to take a single therapeutic dose of the drug the following day at home. After the test, the patients were contacted and asked if they had a reaction to the dose they received at home.

Patients were randomly assigned to either the short or the extended DPT protocol by the clinician. The DPT result was considered positive in the presence of any of the following objective signs and symptoms related to a hypersensitivity reaction: cutaneous symptoms (urticaria, angioedema, and/or conjunctivitis), upper or lower respiratory tract symptoms (rhinorrhea, sneezing, nasal congestion, cough, and/or dyspnea), >15% decrease in forced expiratory volume in 1 second during DPT, and cardiovascular symptoms (change in heart rate and hypotension). Those with positive drug challenge results on day one were excluded from the study.

Children with negative DPT results were contacted by telephone at least 3 months later or at a follow-up visit to check their tolerance to NSAIDs in case of reuse. We asked the patients or their caretakers whether they used the tested drug again and, if so, whether there was a reaction. If yes, then we inquired about the type of reaction and, if not, the reason for not using the drug. Based on the findings from this reevaluation, the NPV was calculated separately for each protocol.

Ethics Considerations

Written consent was obtained from the parents. This study was approved by the ethics committee of Selcuk University Faculty of Medicine (2025/87).

Statistical Analysis

The data obtained were evaluated by using the Statistical Package for the Social Sciences software, version 22.0 (IBM SPSS Corp., Armonk, NY). Numbers and percentages are reported for discrete variables, and means ± standard deviations are reported for continuous variables. Normally distributed data are presented as means ± standard deviations whereas nonnormally distributed data are presented as medians and interquartile ranges (IQR). The χ² test was used to compare nonparametric data; the Mann-Whitney U test was used in comparison of nonnormally distributed continuous data, and independent samples t-test was used for normally distributed continuous data. P-values < 0.05 were considered statistically significant.

RESULTS

A total of 143 DPTs with NSAIDs (including paracetamol) in 115 patients were performed during the study period. Fourteen patients with positive DPT results and 13 patients tested with alternative drugs were excluded from the study. The DPT results were negative for 116 suspected agents in 104 patients, 56 of whom were boys (53.8%). Ninety-two patients underwent DPT with one drug, whereas 12 underwent DPT with two different suspected NSAIDs. A minimum interval of 4 weeks was maintained between DPT with different drugs. The median (IQR) age of the patients was 8 years (4.8–12.2 years) at the time of DPT. Sixty patients (57.6%) and 67 suspected agents were tested with short DPT, and 44 patients (42.3%) and 49 suspected agents were tested with extended DPT. The sex distribution was similar between the short (51.7% boys, 48.3% girls) and extended DPT (56.8% boys, 43.12% girls) groups (p = 0.603). The median (IQR) age at the time of reaction was also similar between the short (5.6 years [2.7–10.4 years]) and extended DPT (8.4 years [4.5–13.5 years]) groups (p = 0.121). Similarly, the median (IQR) age at DPT was 6.8 years (4–11 years) for the short DPT group and 8.5 years (5–13.7 years) for the extended DPT group (p = 0.163). There were no significant differences between the groups in the prevalence of comorbidities such as asthma, chronic urticaria, and allergic rhinitis (p = 0.36, p = 0.82, and p = 0.38, respectively).

The clinical characteristics of the initial reactions to NSAIDs reported by the patients are listed in Table 1. The most common presenting symptom was urticaria, which occurred in 41.7% of patients in short DPT cases and 59.1% of patients with extended DPT cases. Urticaria with angioedema was the second most common symptom in both groups (short, 25%; extended, 22.7%). There was no significant difference between the groups for either symptom (p = 0.079 and p = 0.78, respectively). Paracetamol was the most frequently suspected drug in both groups (46.7% patients in the short DPT group and 44.4% patients in the extended DPT group), followed by ibuprofen (31.7% in the short DPT group and 42.2% in the extended DPT group) (Table 1).

Table 1.

Demographic characteristics of the patient

DPT = Drug provocation test; IQR = Interquartile range; min-max = minimum-maximum.

Of the 116 negative DPT results, 102 patients with 114 results (∼98.2%) could be reached. A total of 105 NSAIDs (90.5%) were used at least once after DPT. The median (IQR) time between the DPT and the first subsequent reuse of the drug was 2 months (1–4 months). Among the patients reached, four in the short DPT group did not reuse the drug (three did not need to use it, and one did not because he or she was afraid). In the extended group, three patients did not need to reuse the suspected drug and two did not use the drug because they were afraid (Fig. 1). In both groups, none of the patients reported a reaction during reexposure, and the NPV for both the short and the extended DPT groups was 100%.

Figure 1.

An algorithm of the study.

DISCUSSION

This study evaluated the outcomes of DPTs in pediatric patients with a history of suspected NSAID-H by comparing the NPVs of short and extended DPT protocols. During the study period, we determined that 143 DPTs were performed for 115 patients. A hundred and sixteen DPT results (97.2%) with the suspected drugs were negative. However, 14 tests (9.7%) were excluded because the results were positive, and 13 (9%) were excluded because they were performed with alternative drugs. For both DPT groups, we reached 102 patients, with a total of 114 tests (98.2%), by telephone. Of the patients who could be contacted, 93 had reused a total of 105 tested drugs (90.5%) and none developed reaction on reexposure. Analysis of our results suggests that both short and extended DPT protocols are reliable in confirming drug tolerance, with an NPV of 100% for both approaches.

Although DPT is the criterion standard when investigating hypersensitivity reactions to NSAIDs in children, there is no consensus on the optimal length of the DPT protocol. In a retrospective study, the ENDA group evaluated the total length of DPTs with NSAIDs performed in six different centers.¹³ Four centers performed DPT in a single day; Madrid extended the test to 2 days and Belgrade extended the test to 3 days.^14–16 In the two centers that extended the test, it was performed in the hospital on the first day, whereas, for the other days, patients received the therapeutic dose at home. According to this study, it was found that they were similar in terms of confirming the diagnosis regardless of the approach.¹³ In our study, we found that the NPV of the tests of different lengths performed was the same, at 100%. To the best of our knowledge, our study is the first to compare the NPVs of DPTs performed at different lengths for NSAID-H in children.

A limited number of studies have evaluated the NPV of DPT against NSAIDs.^12,17–19 In an adult study, the NPV for NSAIDs was 98.6%¹²; another adult study that involved patients with urticaria and/or angioedema reported an NPV of 96.97%.¹⁸ In our study, 90.5% of the total number of tested drugs were reused in both groups. Most of the drugs not used after a negative DPT result were NSAIDs such as metimazole, nimesulide, dexketoprofen, and ketoprofen, which are not typically the first preferred option for use in pediatric populations (Table 2). Topal et al.¹⁷ found that the reuse rate of paracetamol after a negative DPT result was the highest, and the NPV was calculated as 100%. According to a recent study in children, 52.3% of patients reused the suspected drug, most commonly paracetamol, after a negative test result.¹⁹ They calculated the NPV of DPTs for NSAIDs to be 94.5%.¹⁹ Similarly, in our study, only one patient in the extended group did not reuse paracetamol because of fear. The reason for the higher rate in our study compared with adult studies in the literature is that ibuprofen and paracetamol are often needed in pediatric patients because of their antipyretic and analgesic properties.^18,20

Table 2.

Characteristics of patients who did not use the suspected drug again after negative DPT (n = 9)

DPT = Drug provocation test.

The symptom distribution in our study also aligns with what is described in previous research.^20,21 Urticaria was the most commonly reported reaction, followed by urticaria with angioedema, which is consistent with earlier studies in both children and adults.^21–23 These findings reinforce the well-established understanding that cutaneous manifestations are the predominant presentation of NSAID-H. In both groups, five patients (three in the short group and two in the extended group) presented with a history of anaphylaxis. Three of the patients had a history of grade 1 anaphylaxis, whereas two had a history of grade 2 anaphylaxis.²⁴ It is interesting to note that all of these patients were able to safely use the suspected drug at home after the negative test result and did not experience any adverse reactions. This highlights the importance of DPTs in removing false drug delabeling.

The rate of NSAID-H confirmation when using DPT varies widely among different pediatric populations, with recent studies reporting values that range from 4% to 68%.^{5,8,14,19,23,25–27} This variation may be attributed to differences in study populations, diagnostic criteria, and protocols used for DPT. In our study, we found that 12.2% of 115 patients with suspected NSAID-H were definitively diagnosed by DPT. The reason for the high negative DPT results in our study is probably viral infections that can mimic allergic reactions because the main indication for the NSAID use in children is as an antipyretic for infections.

One of the strengths of our study was that we were able to reach most of the patients tested, with a high proportion reusing the tested drug at least once. This shows that parents have a high level of confidence in the drug tests performed at the clinic. The limited number of patients is one of the potential limitations of this study. Also, follow-up was conducted via telephone interviews, which carry a risk of recall bias despite efforts to standardize data collection.

CONCLUSION

Our study demonstrates that both short and extended DPT protocols are effective in confirming NSAID and paracetamol tolerance, with no significant differences in outcomes between the two approaches. This suggests that, to rule out suspected NSAID-H in children, a single day of DPT is sufficient. The absence of reactions on reexposure underscores the reliability of negative DPT results and supports its use in routine clinical practice. These findings indicate that DPT is a reliable tool for confirming NSAID tolerance, ultimately improving patient management and reducing unnecessary drug avoidance. Future studies with larger cohorts and longer follow-up periods could further validate these findings.