FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W2078584781> ?p ?o ?g. }

• W2078584781 endingPage "1375" @default.
• W2078584781 startingPage "1372" @default.
• W2078584781 abstract "To the Editor. The case-control study by Zerr et al1 is an attempt to ascertain whether an outbreak of necrotizing fasciitis (NF) among children with primary varicella was linked to ibuprofen use. The authors concluded that there was an association, requiring additional studies to determine whether the association was causal. The authors attempted to control for biases to which case-control studies are susceptible, but some may have occurred.All but 2 of the 9 cases who were administered ibuprofen used itafter the onset of symptoms of the secondary infection. Thus, ibuprofen could not have caused the infection; rather, the symptoms associated with the infection probably resulted in the ibuprofen use. Furthermore, the authors acknowledge that ibuprofen may have been more likely used as a result of a serious secondary infection. Because the serious infection is a predictor of NF, the resulting association of ibuprofen with NF is spurious.Recall bias occurs when cases have a higher likelihood than controls of recalling incidents. One way to minimize this bias is to conduct the interviews as soon as feasible. Although the authors conducted the majority of the interviews within 1 month of the event, others were conducted 2 to 12 months afterwards. Authorities issued an announcement to the local media cautioning parents about the use of ibuprofen in children with varicella during the conduct of this study, which could have increased this bias.Severity of varicella is a relevant factor in assessing an association between ibuprofen and NF. If ibuprofen use increases with the varicella severity, then naturally ibuprofen use will be more prevalent with NF if development of NF also increases with the varicella severity. Unfortunately, no information regarding the severity of the varicella is provided in the article, and the authors did not control for varicella severity in either the design or analysis of the study.One cannot conclude from this study that ibuprofen is causally related to NF. Recall bias may have inflated the association of ibuprofen with NF. Furthermore, ibuprofen use was very possibly confounded with the varicella severity, and more likely an effect of the secondary infection rather than the cause. Other epidemiologic studies to date have shown no significant causal link between NSAID use in varicella patients and the subsequent occurrence of group A streptococcal necrotizing fasciitis (GAS-NF).2–5We agree that a large prospective multicenter national study would be useful to yield a more definitive answer to the important question of risk factors associated with the varicella-associated NF in children.Above all, ibuprofen remains a safe and effective analgesic and antipyretic in children. Almost a decade of use in children clearly demonstrates this. Moreover, because pediatric ibuprofen gained over-the-counter status in 1995, there has been no increased incidence of varicella-NF. The Boston Fever Study6 also supports ibuprofen's excellent safety profile. Emphasizing the rarity of NF, there were no cases of NF reported in this large prospective study.To the Editor. After reading the recent report by Zerr et al1regarding the association of ibuprofen use with necrotizing fasciitis (NF) among children with varicella, one is struck by the odds ratio of 11.5 for ibuprofen use despite the authors' caution that this is not proof of causation. This strong association may influence pediatric practice, but should it? We would like to draw attention to several issues, which suggest that no causal inference should be drawn from these data.First, in trying to resolve the issue of causality, it is critical to know whether the exposure actually preceded the disease. Because most exposed cases (7 of 9) were given ibuprofen only after the onset of symptoms of NF, these data are most compatible with ibuprofen being used to treat symptoms, rather than it causing NF. Second, cases were more likely than controls to have taken diphenhydramine before hospitalization (although the difference did not reach statistical significance), suggesting that the cases were medicated with other drugs intended to increase comfort and may have been sicker than controls, placing them at greater risk for NF. Third, during the 14 months before a press release warning about use of ibuprofen, there were 11 cases of NF (.8 cases/month), whereas there were 8 cases in the 5 months after the press release (1.6 cases/month)—and none of the latter was exposed to ibuprofen. Assuming the population did not change dramatically, these data suggest that the entire cluster of cases described by Zerr et al occurred independently of ibuprofen use. Fourth, in the 3 years since study completion, only a single additional NF case has been identified despite the fact that pediatric ibuprofen is now available over-the-counter and public awareness of the press release has likely dissipated. If ibuprofen indeed caused or contributed to the number of cases of NF reported, it is unlikely there would have been such a dramatic dropoff in cases in an era when ibuprofen is widely used without physician consultation.Without a documented causal link between ibuprofen use and NF in children with varicella, how should the pediatric practitioner respond to this study? Some may choose to recommend that their patients not take ibuprofen because acetaminophen is available; others would prefer to have ibuprofen on hand as an alternate and effective medication. As noted by the authors themselves, the study by Zerr et al does not resolve this quandry, which requires further careful epidemiologic study and, if possible, controlled experiments in appropriate animal models of NF.In Reply. We thank Drs Ford and Waksman of Whitehall-Robins Healthcare, manufacturers of Advil and Orudis Kt, and Drs O'Brien et al for their response to our manuscript.1Both letters question the association between ibuprofen use and NF based on the chronology of exposure to ibuprofen relative to development of NF. Although it is true that in most patients ibuprofen use occurred after the onset of symptoms of the secondary infection, it is unknown whether one group was sicker than the other at that point in time because all cases and controls had 2 or more symptoms of secondary infection (fever, swelling, erythema, pain, and splinting) at the time of presentation. The precise pathophysiology of NF is unknown. It is possible that streptococcal cellulitis differentiates into NF secondary to a combination of factors, one of which could be the use of nonsteroidal antiinflammatory drugs (NSAIDs). Furthermore, we demonstrated that among those children with NF, ibuprofen use was greater in those children whose course was complicated by streptococcal toxic shock syndrome (STSS) and/or renal insufficiency. It is unlikely that these children consumed ibuprofen after they had STSS and renal insufficiency. Furthermore, we found that the children who consumed ibuprofen had a significantly longer duration of symptoms before hospitalization than the children who did not. Accordingly, one possible explanation for the association between ibuprofen and NF is that children who consumed ibuprofen may have experienced a delay in receiving appropriate therapy, perhaps attributable to symptomatic relief provided by the ibuprofen, leading to the development of NF or more complicated NF.Drs O'Brien et al state that cases of NF were more likely than controls to take diphenhydramine. True, but not significantly more likely based on statistical testing. We studied the use of 5 medications and their association with NF. The differences between groups regarding use of diphenhydramine (68% vs 50%, P = 22) as well as acetaminophen (79% vs 85%, P = .70), antibiotics (35% vs 23%, P = .52), and calamine lotion (56% vs 69%, P = .35) are actually relatively small, with a high likelihood of being attributable to chance alone. Ibuprofen was the only medication that had a very large difference between groups (42% vs 15%, P = .02), and statistically speaking, there is a small chance that this is attributable to chance alone.Recall bias, a concern raised by Drs Ford and Waksman, is always a concern in studies using interview data. We felt that recall bias was probably minimized in this study because all cases and controls were hospitalized for their secondary infection. That is, whether they had NF or a less serious infection, it was a highly significant event for all of the patients and parents involved in this study. Regarding concern about the announcement and its possible impact on recall bias, as we reported in the manuscript, the data were reanalyzed excluding those cases and controls who were hospitalized after the announcement. Results from this analysis were essentially unchanged (odds ratio for ibuprofen use among cases versus controls before the announcement, 12.4; 95% confidence interval 1.4–148.6).Drs O'Brien et al also raise the issue of the press release with regard to the presentation of cases of NF and their use of ibuprofen. First, we believe NF is most likely a multifactorial problem, and our data suggests that ibuprofen may be one of the factors associated with its development. The fact that there were cases of NF that did not use ibuprofen is not counter to this conclusion. NF has occurred in sporadic outbreaks across the country. It is likely that circulation of more virulent group A streptococcal (GAS) strains is a necessary factor in the development of invasive streptococcal disease, and it is possible that the development of invasive disease may be facilitated by NSAID use. The cessation of cases of invasive disease would therefore perhaps correlate more with the prevalence of invasive GAS strains than with that of any other cofactors. Second, without studying the question, it may not be reasonable to assume that ibuprofen use is widespread in this area of the country. It is unknown what effect this outbreak and the attendant publicity have had on the use of ibuprofen among children with varicella in this region.We agree with Drs Ford and Waksman that the severity of varicella may play a role in the development of superinfection. We found that information regarding this issue was difficult to obtain from the medical charts and parental interview. This would be an interesting variable to explore in a prospective fashion.Drs Ford and Waksman, citing four references, state that “other epidemiological studies to date have shown no significant causal link between NSAID use in varicella patients and the subsequent occurrence of GAS-NF.” We take issue with this statement. The first reference cited is a preliminary abstract of our case-control study that was presented at a national conference.2 Contrary to the doctors' suggestion, we did find and report an association between ibuprofen use and NF. The second citation is a letter from Dr Rosefsky of Wyeth-Ayerst Laboratories3 in response to the case series by Brogan and colleagues.4 Rosefsky, after reviewing the data collection sheets from the case series of children with NF (there were no controls), noted that 11 of 12 of the patients had received acetaminophen. We examined acetaminophen use in ourcase-control study and did not find an association with NF or NF complicated by STSS or renal insufficiency. The third reference5 utilized the Food and Drug Administration's Spontaneous Reporting System to review cases of NF where NSAIDs had been used for the purpose of identifying common features among these cases. Again, this was not a controlled study designed to examine a possible association between NF and NSAIDs. The fourth reference6 was a controlled study investigating the possible association between ibuprofen use and dermatologic superinfections among children with varicella. There were no cases of NF in this study.We were confused by Ford and Waksman's reference to the Boston Fever Study7 as evidence of ibuprofen's safety in this context. The Boston Fever Study was not designed to study ibuprofen use in the setting of varicella or serious soft-tissue infections. This study, which occurred in the outpatient setting of primary care providers, may have selected for a less sick population as physicians enrolled eligible patients “at their discretion (eg as workload permitted),” and patients were ineligible if they were >10% dehydrated. In fact, the vast majority of the patients had upper respiratory tract infection, otitis media, pharyngitis, lower respiratory tract infection, or gastrointestinal tract illnesses while approximately 13% of the children did not have a specified illness and <1% of the children had varicella. We agree that NF following varicella is a rare phenomenon, attributable to unclear determinants.In summary, we found an association between the use of ibuprofen and NF and the complications of NF. This adds to previously published data suggesting a possible association between NF and NSAIDs including case reports of NF associated with NSAID use in healthy adults8–12 and experimental evidence suggesting that leukocyte function may be adversely affected by NSAIDs.13–16 Although we cannot conclude that this is a causal relationship, we stand by our recommendation that based on the available evidence, clinicians caring for patients with varicella should exercise heightened vigilance for the development of secondary infections, especially in the context of NSAID use. Furthermore, clinicians may want to exercise prudence when recommending the use of ibuprofen in their patients with varicella, especially if soft-tissue infection is a concern, until additional information becomes available." @default.
• W2078584781 created "2016-06-24" @default.
• W2078584781 creator A5037622152 @default.
• W2078584781 creator A5088533032 @default.
• W2078584781 date "2000-06-01" @default.
• W2078584781 modified "2023-10-12" @default.
• W2078584781 title "Necrotizing Fasciitis During Primary Varicella" @default.
• W2078584781 cites W2001426216 @default.
• W2078584781 cites W2039200745 @default.
• W2078584781 cites W2070888339 @default.
• W2078584781 cites W2078707956 @default.
• W2078584781 cites W2285223599 @default.
• W2078584781 doi "https://doi.org/10.1542/peds.105.6.1372" @default.
• W2078584781 hasPubMedId "https://pubmed.ncbi.nlm.nih.gov/10877671" @default.
• W2078584781 hasPublicationYear "2000" @default.
• W2078584781 type Work @default.
• W2078584781 sameAs 2078584781 @default.
• W2078584781 citedByCount "13" @default.
• W2078584781 countsByYear W20785847812012 @default.
• W2078584781 countsByYear W20785847812015 @default.
• W2078584781 countsByYear W20785847812016 @default.
• W2078584781 countsByYear W20785847812017 @default.
• W2078584781 crossrefType "journal-article" @default.
• W2078584781 hasAuthorship W2078584781A5037622152 @default.
• W2078584781 hasAuthorship W2078584781A5088533032 @default.
• W2078584781 hasConcept C121332964 @default.
• W2078584781 hasConcept C1276947 @default.
• W2078584781 hasConcept C141071460 @default.
• W2078584781 hasConcept C16005928 @default.
• W2078584781 hasConcept C2777217379 @default.
• W2078584781 hasConcept C2777977315 @default.
• W2078584781 hasConcept C71924100 @default.
• W2078584781 hasConceptScore W2078584781C121332964 @default.
• W2078584781 hasConceptScore W2078584781C1276947 @default.
• W2078584781 hasConceptScore W2078584781C141071460 @default.
• W2078584781 hasConceptScore W2078584781C16005928 @default.
• W2078584781 hasConceptScore W2078584781C2777217379 @default.
• W2078584781 hasConceptScore W2078584781C2777977315 @default.
• W2078584781 hasConceptScore W2078584781C71924100 @default.
• W2078584781 hasIssue "6" @default.
• W2078584781 hasLocation W20785847811 @default.
• W2078584781 hasLocation W20785847812 @default.
• W2078584781 hasOpenAccess W2078584781 @default.
• W2078584781 hasPrimaryLocation W20785847811 @default.
• W2078584781 hasRelatedWork W1506200166 @default.
• W2078584781 hasRelatedWork W1995515455 @default.
• W2078584781 hasRelatedWork W2048182022 @default.
• W2078584781 hasRelatedWork W2080531066 @default.
• W2078584781 hasRelatedWork W2559834514 @default.
• W2078584781 hasRelatedWork W2748952813 @default.
• W2078584781 hasRelatedWork W2899084033 @default.
• W2078584781 hasRelatedWork W3031052312 @default.
• W2078584781 hasRelatedWork W3032375762 @default.
• W2078584781 hasRelatedWork W3108674512 @default.
• W2078584781 hasVolume "105" @default.
• W2078584781 isParatext "false" @default.
• W2078584781 isRetracted "false" @default.
• W2078584781 magId "2078584781" @default.
• W2078584781 workType "article" @default.