SGEM#334: In My Life there’s been Earache and Pain I don’t know if it’s IBI again – in an Afebrile Infant with Acute Otitis Media

Date: June 11th, 2021 Guest Skeptic: Dr. Dennis Ren is a paediatric emergency medicine fellow at Children’s National Hospital in Washington, DC. Reference: McLaren SH, et al. Invasive bacterial infections in afebrile infants diagnosed with acute otitis media. Pediatrics 2021 Case: You are working with a medical student at the emergency department when a 2-month-old boy is brought in by his parents for fussiness. They note that he has had upper respiratory symptoms for the past few days and fussier than usual. He has still continued to feed well and make wet diapers. He has not had any fever. Yesterday, they noticed that he seemed to be pulling at his right ear. On exam, he is afebrile, active, and alert. He cries and moves vigorously when you look into his ear. You see a bulging, red tympanic membrane. His left tympanic membrane is clear. The rest of his exam is unremarkable. You turn to the medical student and ask her what she would like to do for this patient. She replies that she thinks the patient has an acute otitis media (AOM) but given his age, she is also thinking about the possibility of an invasive bacterial infection (IBI) and would like to obtain some blood for labs and even consider a lumbar puncture for cerebral spinal fluid. How do you reply? Background: Acute Otitis media is the second most diagnosed illness in children and the most common indication for antibiotic prescription [1-2]. We have covered the of AOM twice on the SGEM: SGEM#132: One Balloon for Otitis Media with Effusion with Dr. Richard Lubell SGEM#278: Seen Your Video for Acute Otitis Media Discharge Instructions SGEMHOP with lead author Dr. Naveen Poonai? In 2013, the American Academy of Pediatrics (AAP) updated recommendations for the diagnosis and management of acute otitis media (AOM) for children older than 6 months. Unfortunately, there is limited guidance for patients younger than 6 months. The diagnosis of AOM becomes more complicated by the concern for concurrent invasive bacterial infections (IBI) in infants less than 3 months of age. Previous studies have demonstrated low prevalence of concurrent IBI in infants with AOM, but sample size has been small and included a mix of afebrile and febrile infants [3-4]. Additionally, the microbiology of pathogens causing AOM has shifted after the implementation of the pneumococcal conjugate vaccine with a higher proportion of patients having culture negative AOM [5]. This uncertainty has led to wide practice variation and controversy surrounding diagnostic testing (blood and cerebrospinal fluid testing), antibiotic administration (IV vs oral), and disposition (discharge vs admission) in infants with AOM. Clinical Question: What is the prevalence of invasive bacterial infections and adverse events in afebrile infants ≤ 90 days of age with acute otitis media? Reference: McLaren SH, et al. Invasive bacterial infections in afebrile infants diagnosed with acute otitis media. Pediatrics 2021 Population: Afebrile infants ≤ 90 days of age with clinically diagnosed acute otitis media across 33 pediatric emergency departments (29 USA, 2 Canadian and 2 Spanish EDs) from 2007 to 2017 Excluded: Temperature ≥38°C and <36°C in the ED or within 48 hours, antibiotic use (other than topical) within 48 hours of presentation, concurrent mastoiditis, evidence of focal bacterial infection, transferred to ED with previous diagnostic testing/antibiotics Intervention: Evaluation of invasive bacterial infections in blood or cerebrospinal fluid (CSF) Comparison: None Outcomes: Primary Outcome: Prevalence of IBI (bacterial meningitis and bacteremia) Secondary Outcomes: Variability in diagnostic testing for IBIincluding blood or cerebrospinal fluid (CSF), parenteral antibiotic administration, and hospitalization. Safety Outcome: AOM-associated adverse events Authors’ Conclusions: “Afebrile infants with clinician-diagnosed acute otitis media have a low prevalence of invasive bacterial infections and adverse events; therefore outpatient management without diagnostic testing may be reasonable.” Quality Checklist for Observational Study: Did the study address a clearly focused issue? Yes Did the authors use an appropriate method to answer their question? Yes Was the cohort recruited in an acceptable way? Unsure Was the exposure (diagnostic testing) accurately measured to minimize bias? Yes Was the outcome accurately measured to minimize bias? Unsure Have the authors identified all-important confounding factors? Unsure Was the follow up of subjects complete enough? Yes How precise are the results? Unsure Do you believe the results? Yes Can the results be applied to the local population? Unsure Do the results of this study fit with other available evidence? Yes Results: They screened 5,270 infants from 33 sites to establish a cohort of 1,637 who met inclusion/exclusion criteria. The median age was 68 days and 89% met the simplified American Academy of Pediatrics diagnostic criteria for AOM. Key Result: No infant with blood cultures had bacteremia and no infant was diagnosed with bacterial meningitis by CSF culture. Primary Outcome: Prevalence of Invasive Bacterial Infections 0/278 with blood cultures had bacteremia (95% CI 0%-1.4%) 0/102 with CSF cultures had bacterial meningitis (95% CI 0% to 3.6%) Two infants had adverse events in 30-day follow up and hospitalization history: Culture negative sepsis vs. severe dehydration from milk protein allergy Lymphadenitis (20 days after ED visit) after initial discharge on amoxicillin Secondary Outcomes: Diagnostic Testing, Antibiotic Administration, and Hospitalization Rates. One-fifth (21.7%, 355 of 1637) had ≥1 diagnostic test. One-third (34%, 34 of 100) of infants ≤28 days had a lumbar puncture. For infants with upper respiratory symptoms (n=1179), ~5% had a lumbar puncture, and 13.7% had blood cultures. ~10% received IV or IM antibiotic 90% received a prescription for oral antibiotic (most commonly amoxicillin followed by amoxicillin-clavulanate) 5% were discharged from the emergency department Almost half (47%) of infants 0-28 days were hospitalized 72-hour return rate was 63 (4.3%) of which 15 infants were hospitalized. Most common reason for hospitalization was bronchiolitis Older infants were less likely to have blood cultures, lumbar puncture, and hospitalization compared to younger infants. Ear discharge was significantly associated with diagnostic testing and hospitalization. 1. Diagnostic Uncertainty: The diagnosis of otitis media remains challenging. It is a clinical diagnosis and there is variability amongst clinicians. Here is the American Academy of Pediatrics (AAP) 2013 diagnostic criteria for AOM [6]. But even the AAP concedes that there is no gold standard for the diagnosis of AOM because it exists in a spectrum of findings as the disease process develops. This becomes challenging in a retrospective review as the diagnosis of AOM cannot be verified. The authors chose to use a simplified diagnostic criteria defined by presence of tympanic membrane erythema, bulging tympanic membrane, or otorrhea. Unfortunately, the presence of tympanic membrane erythema alone is not specific for AOM as it can be because the infant is crying. This may falsely increase the number of infants with AOM included in the analysis and underestimate the prevalence of IBI. Although, zero infants in the study had confirmed bacteremia or bacterial meningitis on cultures. 2. Primary Outcomes: They had two primary outcomes, prevalence of IBI (bacteremia and bacterial meningitis) and AOM-associated adverse events. We know from the movie Highlander…There can be only one! Primary outcome. I am uncertain that the two reported adverse events are truly associated with AOM. One patient who was initially discharged on amoxicillin presented 20 days later  with cervical lymphadenitis and perforated AOM. The other patient did grow Staphylococcus aureus from middle ear effusion and was hospitalized for culture-negative sepsis but was also thought to have a severe milk protein allergy. The authors note that this patient presented with severe illness during the index ED visit so was likely not going to be discharged home regardless. 3. Partial Verification Bias: There were several potential biases in this study that could have impacted the results. One was already mentioned in nerdy point number one that there is not a gold standard. This is called the imperfect gold standard bias or the copper standard bias. Another potential bias is partial verification bias. Not all infants in the study underwent the same testing for IBI. This can cause an underestimation of IBI. It is possible that some infants with bacteremia on whom blood cultures were not obtained were treated by oral antibiotics. The authors did try to perform follow up on patients within 30 days by reviewing medical records but acknowledge that this may not fully capture all patients with bacterial meningitis that was seen outside of the index ED or hospital system. 4. Selection Bias: As someone who works mainly in community hospitals, I’m concerned about selection bias. Are the infants who present to free-standing pediatric hospitals the same as those who present to community or rural EDs? Are those in large urban areas more likely or less likely to bring in a “fussy” infant? Are parents in the USA less likely to go to the ED because of co-pays, insurance status or other financial issues? It is unclear how any of these biases could alter the findings in this study. 5. Age: A last point to highlight is that there is a very low number of infants ≤28 days old (n=100, 6.1% of study population). This low number could be due to a variety of reasons.