Objective. To test the hypothesis that short-term use of ibuprofen increases the risk of impaired renal function in children.
Study Design. Randomized, double-blind acetaminophen-controlled clinical trial. Children with a febrile illness were enrolled from outpatient pediatric and family medicine practices and randomly assigned to receive either acetaminophen suspension or one of two dosages of ibuprofen suspension (5 mg/kg or 10 mg/kg) for fever control.
Results. Mean blood urea nitrogen levels on admission among children admitted to hospital and assigned ibuprofen 5 mg/kg (n = 96), ibuprofen 10 mg/kg (n = 102), and acetaminophen 12 mg/kg (n = 87) were 4.1, 3.8, and 3.9 mmol/L, respectively. The corresponding creatinine levels were 43, 41, and 43 μmol/L, respectively. The prevalence of a creatinine level >62 μmol/L was 9.5% overall and did not vary by antipyretic assignment. Among 83 children hospitalized with dehydration, the mean creatinine level was 44 μmol/L, and the prevalence of an elevated creatinine was 14%; neither measure varied by antipyretic assignment.
Conclusion. Although renal failure in children has been reported after ibuprofen use, these data suggest that for short-term use the risk of less severe renal impairment, as reflected by blood urea nitrogen and creatinine levels, is small and not significantly greater than that after acetaminophen use.
- nonsteroidal antiinflammatory drugs
- renal function
- blood urea nitrogen
- clinical trial
Renal failure has been reported after nonsteroidal antiinflammatory drug (NSAID) use in adults,1 but only rarely among children exposed to ibuprofen.2-4 In a recently completed randomized clinical trial, we found that none of 55 785 children treated with ibuprofen for fever was hospitalized with renal failure.5 In the present analysis, we sought to determine whether the risk of less severe renal impairment, as reflected by elevated blood urea nitrogen (BUN) and creatinine levels, was associated with short-term use of ibuprofen suspension for the control of fever in children.
These data were obtained as part of the Boston University Fever Study, the methods and principal results of which have been previously described.5-7 Briefly, the study was a practitioner-based, double-blind clinical trial designed to assess the safety of ibuprofen suspension when used to treat fever in children. Patients were enrolled into the study from February 1991 through June 1993. The final data included information on 83 915 children treated for febrile illnesses who were randomly assigned to receive suspensions of either acetaminophen (12 mg/kg) or ibuprofen in one of two dosages (5 or 10 mg/kg) and followed for 4 weeks. The study protocol did not require laboratory measures of renal function to be systematically obtained for all study participants. However, for each child hospitalized during follow-up, we obtained the hospital record and abstracted the admission BUN and creatinine levels (or the first value of each reported within 24 hours of admission); abstractions were performed blindly with respect to antipyretic assignment. Patients who did not have BUN and creatinine levels obtained within 24 hours of admission were excluded from this analysis.
These data were evaluated using an intention-to-treat analysis (ie, participants were grouped according to the treatment assigned at study enrollment, and these groups were preserved throughout the analyses). The analyses were repeated excluding patients who had received none of the assigned medication (approximately 10% in each assignment group); the results were not materially different, and only the results of the intention-to-treat analysis are presented here.
Mean admission BUN and creatinine levels were compared using analyses of covariance to control for the potential confounding effects of age, sex, weight, and discharge diagnosis (comorbidity). Between-group differences in the prevalences of an abnormal BUN or creatinine level were assessed using the χ2 test. Post hoc power analyses were performed for both continuous and dichotomous variables comparing children assigned to treatment with acetaminophen to all children assigned to ibuprofen (ie, both ibuprofen dosage groups were combined).8
Overall, 795 (1%) of the study children were hospitalized in the 4-week follow-up period; hospitalization rates did not vary according to antipyretic assignment; most hospitalizations were for acute infections or their complications, and pneumonia was the most common discharge diagnosis.5 Renal function tests were performed at the discretion of the attending physician. Among hospitalized children, admission BUN or creatinine levels were available for 288 (36%); of these, 97, 103, and 88 children had been randomized to treatment with acetaminophen, ibuprofen (5 mg/kg), or ibuprofen (10 mg/kg), respectively. Descriptive characteristics of these children (including cause of fever at study enrollment) are shown in Table 1; the three treatment groups were generally similar. The children received a median of 7 doses of study medication throughout a median duration of 2 days; the number of doses received and duration did not vary by treatment group. Discharge diagnoses are shown in Table 2. One child randomized to receive acetaminophen was admitted with hypertension, edema, and hematuria; his discharge diagnosis was postinfectious glomerulonephritis, probably streptococcal.5
Mean BUN and creatinine levels and the prevalence of BUN values greater than 6.4 mmol/L (18 mg/dL) and creatinine values greater than 62 μmol/L (0.7 mg/dL) according to antipyretic assignment group are shown in Table 3. The mean BUN level was approximately 4 mmol/L and the mean creatinine level was 42 μmol/L in each group; neither measure varied significantly by treatment group after controlling for age, sex, weight, and discharge diagnosis (P > .5). The overall prevalences of BUN levels greater than 6.4 mmol/L (18 mg/dL) and creatinine levels greater than 62 μmol/L (0.7mg/dL) were 8.1% and 9.5%, respectively, and the proportion of children with elevated levels did not vary by antipyretic assignment. The prevalence of high creatinine levels did not vary significantly by antipyretic group when other creatinine reference values were used (eg, 35 μmol/L or 88 μmol/L). Admission BUN and creatinine values among the 108 children hospitalized with a concomitant diagnosis of dehydration are shown in Table4. The prevalence of BUN levels greater than 6.4 mmol/L (18 mg/dL) and creatinine levels greater than 62 μmol/L (0.7 mg/dL) were slightly higher among these children than among all hospitalized children. However, as was true in the larger group, neither mean level nor the prevalence of an elevated level of either BUN or creatinine varied significantly by antipyretic assignment.
These data have adequate power to detect small increases in BUN or creatinine. For example, the data had 80% power to detect either a difference of 8 μmol/L (0.1 mg/dL) in mean creatinine level or a 2.7-fold increase in risk of a creatinine level greater than 62 μmol/L among all children randomized to ibuprofen (ie, both dose groups combined) compared with those randomized to acetaminophen.
Renal impairment in adult NSAID users occurs primarily among patients with preexisting renal disease or other conditions associated with low intravascular volume or low cardiac output.1,9 Under these conditions, renal blood flow is regulated by processes involving prostaglandins, and the inhibition of prostaglandin synthesis by NSAIDs may result in reduced renal blood flow and glomerular filtration. Acute renal failure after ibuprofen overdose and interstitial nephritis after ibuprofen used in clinical doses have also been described in patients without preexisting renal disease.9,10
It is unclear how this adult experience applies to children, who may be more or less susceptible to renal injury by NSAIDs. Children may be at lower risk because they are less likely than adults to have other factors predisposing to acute renal disorders (eg, congestive heart failure, diuretic use, and chronic renal failure). Conversely, they may be at greater risk because some degree of dehydration is likely to accompany fever, the most common indication for ibuprofen use in children, and dehydration may increase the likelihood of renal complications.
A number of reports have linked ibuprofen use to renal complications in children, and three have described renal failure in children younger than 15 years of age after short-term treatment with ibuprofen in therapeutic doses.2-4 These include acute renal failure with edema in a 10 year old after ibuprofen use for traumatic pain,2 and episodes of acute flank pain and non-oliguric renal failure in adolescent girls after ibuprofen use.3,4In the last of these, interstitial nephritis was documented on renal biopsy.4 Renal failure has also been reported in a healthy two year old after an overdose with ibuprofen11 and in newborns after in utero exposure to NSAIDs (1 infant was exposed to ibuprofen).12 Case reports, however, neither prove causation (none of the affected children was rechallenged with ibuprofen) nor do they provide estimates of the rate at which renal failure occurs among otherwise healthy children receiving short-term treatment with ibuprofen. Spontaneous (case) reports and observational (ie, nonrandomized) studies cannot provide valid estimates of risk when the choice of treatment is influenced by illness severity (ie, when confounding by indication occurs).6,13 Because ibuprofen was until recently available for use in children only by prescription, children given this medication were likely to be more seriously ill than children treated with acetaminophen. Ibuprofen-treated children therefore may have been at greater risk of renal failure because of the severity of their underlying illness, unrelated to ibuprofen use. Because of this confounding, an unbiased estimate of the risk of renal failure requires that the choice of antipyretic be independent of the severity of the underlying febrile illness (ie, random).6The experience of 55 785 children in the present report provides one such estimate of the risk of complete renal failure after ibuprofen use in children. None of the participants randomized to ibuprofen was hospitalized with renal failure (the observed risk was 0 per 55 785), and the upper bound of the 95% confidence interval around this estimate allows one to conclude that the true risk is likely to be no greater than 5.4 per 100 000.5 However, this estimate does not allow one to assess the risk of lesser degrees of renal impairment, which may be predictive of more serious renal failure.
Randomized clinical trials designed to test the efficacy of ibuprofen in children have typically been too small to detect even modest differences in the risk of mild renal impairment. However, in one clinical trial involving 119 febrile children, increases in BUN were observed after administration of a single dose of ibuprofen (5 to 10 mg/kg); the authors describe the increases as being small and clinically unimportant.14 No differences were observed in creatinine levels and, when multiple testing was taken into account, the increases in mean BUN levels were not statistically significant. In a smaller multi-dose trial described in the same report, neither BUN nor creatinine levels increased significantly among children treated with ibuprofen (2.5 to 10 mg/kg/dose).11 As was the case for renal failure, the randomized, double-blind design of the present study permits an unbiased estimation of the risk of mild degrees of renal impairment (as reflected by elevated serum BUN and creatinine levels) after short-term ibuprofen use for fever in relatively unselected, otherwise healthy children. By virtue of its size, this study has power adequate to detect small differences in mean creatinine level (8 μmol/L) and can effectively rule out a 3-fold or greater relative risk of an elevated creatinine level among ibuprofen-treated children compared with those treated with acetaminophen.
Because renal function tests were not performed on all participants, we cannot estimate the absolute incidence (rate/1000 courses of treatment) of minor, transient, and undiagnosed impairment in renal function after ibuprofen use, nor can we estimate the relative risk of mild renal effects among children not hospitalized. However, it seems likely that clinically important episodes of renal impairment occurring among participants in the trial were not missed, and that those children at greatest risk for renal failure were hospitalized. As might be expected, elevated BUN and creatinine levels were somewhat more common among children with a discharge diagnosis of dehydration, but there was no evidence of a difference in risk related to antipyretic assignment among these children.
These data provide no information on the risk of renal impairment in children after long-term use of ibuprofen or among children ineligible for the clinical trial (eg, those with severe dehydration [≥10% of body weight], preexisting chronic renal, endocrine, or neoplastic disease). Further, it is possible that the children in this study are not representative of all patients seen by the participating physicians because these physicians may not have enrolled their sickest patients. However, the diagnoses at enrollment reflect the wide spectrum of febrile illness seen among children in pediatric practice, and because all study participants were seen by a physician, they were likely to have been more seriously ill than most children who are treated with ibuprofen obtained without a prescription.
These data do not exclude the possibility that ibuprofen may cause acute renal failure in some children, but they do suggest that the risk of acute renal complications after short-term use for fever reduction is small and cannot easily be differentiated from the risk among children given acetaminophen for the same indication.
This work was supported by McNeil Consumer Products Company, Fort Washington, Pennsylvania.
We wish to thank our Advisory Committee for their valuable advice and guidance provided throughout the study: Ralph E. Kauffman, MD (chair), Director of Medical Research, Children's Mercy Hospital, Kansas City, MO; Michael D. Bailie, MD, PhD, University of Illinois College of Medicine at Peoria, Peoria, IL; William Gerson, MD, private practice, Burlington, VT; Alan M. Leichtner, MD, Division of Gastroenterology and Nutrition, Children's Hospital, Boston, MA; Alan Leviton, MD, Neuroepidemiology Unit, Children's Hospital, Boston MA; and Frederick H. Lovejoy, Jr, MD, Children's Hospital, Boston, MA. The liaison to the Advisory Committee consists of: Sumner J. Yaffe, MD, National Institute of Child Health and Human Development, Bethesda, MD; Anthony R. Temple, MD, Medical Affairs, and Barbara H. Korberly, PharmD, Medical New Product Development, McNeil Consumer Products Co. We also wish to thank Richard Vezina, MPH, who coordinated the data collection and supervised our study staff; Mary Joan Denisco, PharmD (McNeil Consumer Products Co), for technical support; David P. Harrington, PhD, for statistical advice; and Samuel Shapiro, MB, FRCP(E), for his advice and guidance. We are especially grateful to the more than 1700 physicians/investigators without whose participation this study would not have been possible.
- Received March 25, 1997.
- Accepted May 12, 1997.
Reprint requests to (S.M.L.) The Slone Epidemiology Unit, Boston University School of Medicine, 1371 Beacon St, Brookline, MA 02146.
- NSAID =
- nonsteroidal antiinflammatory drug •
- BUN =
- blood urea nitrogen
- McIntire SC,
- Rubenstein RC,
- Gartner JC,
- Gilboa N,
- Ellis D
- Wattad A,
- Feehan T,
- Shepard FM,
- Youngberg G
- ↵Hintze JL. SOLO Statistical System: Power Analysis. Los Angeles, CA: BMDP Statistical Software; 1992
- Kelly MT,
- Walson PD,
- Hayes JR,
- Edge JH
- Copyright © 1997 American Academy of Pediatrics