Acute otitis media (AOM) is a common childhood infection, affecting > 80% of children under 3 years of age.1 Although AOM is relatively uncommon in infants under 3 months of age, once the diagnosis is made, it is seen a paucity of evidence to guide subsequent clinical care.
There are no national recommendations for young infants with AOM since the current American Academy of Pediatrics (AAP) guidelines on diagnosis and management of AOM exclude infants younger than 6 months of age.2
An important clinical concern in infants younger than 3 months with AOM is whether they have concomitant invasive bacterial infections (IBIs). In previous studies, investigators have suggested a low prevalence of bacteremia among infants with AOM, but these studies were limited by small sample sizes and the inclusion of mixed populations of children with and without fever.3–8
The clinical conundrum of AOM in infants younger than 3 months may be more relevant for those without fever, since in them the appropriate diagnostic evaluation of IBI, if any, is unclear.
The relative lack of data on afebrile infants with AOM has likely resulted in variability in diagnostic approach among clinicians. If IBIs and adverse events are shown to be rare in a large sample of afebrile infants, clinicians may tend to order fewer diagnostic tests.
The authors’ primary objective was to determine the prevalence of IBIs and adverse events associated with AOM in afebrile infants ≤90 days of age with clinically diagnosed otitis. The secondary objective was to describe the patterns of diagnostic testing and the factors associated with such evaluations in these afebrile infants.
| Methods |
> Study design and population
A cross-sectional study was conducted in 33 pediatric emergency departments (EDs) associated with the AAP Pediatric Emergency Medicine Research Collaborative Committee. Participating sites included 29 DEs in the United States, 2 DEs in Canada, and 2 DEs in Spain. All were teaching hospitals, and 25 (76%) were freestanding children’s hospitals. The study was approved by the Board of Directors of the Collaborative Committee and by the research ethics committees of all participating sites.
Infants ≤90 days old who presented to a participating ED between January 2007 and December 2017 and who were clinically diagnosed with AOM were included. The study period varied by site depending on available data.
We excluded infants with documented temperatures of ≥ 38.0°C or < 36.0°C in the ED or 48 hours before arrival, use of antibiotics other than topical antibiotics within 48 hours of presentation to the ED, diagnosis mastoiditis, evidence of focal bacterial infection on ED examination, or who were referred to the ED with diagnostic testing completed and/or antibiotic administration initiated at an outside hospital.
We chose to exclude febrile infants from this study because the presence of fever in young infants, and not necessarily the diagnosis of AOM, is a primary driver for more extensive evaluation and/or empirical treatment of IBI.
Eligible infants were identified using two methods. The primary method was the use of an inclusive list of diagnostic codes for AOM from the International Classification of Diseases (ICD) 9th or 10th Revision. To evaluate infants with IBI who also had AOM but without documented ICD codes, the records of all infants evaluated in the ED were retrieved and reviewed to see if the presence of bacteria in the blood had been recorded in microbiology databases. or cerebrospinal fluid (CSF). T
All AOM diagnoses were verified by review of the clinical evaluation in the medical records. Because the AAP diagnostic criteria for AOM are for children ≥ 6 months and may be too strict to apply directly to this younger age group, we chose for the primary analysis to include all infants with documentation of ≥ 1 examination. abnormal middle ear as diagnostic support for AOM.2
However, a subanalysis was performed on infants who met simplified AAP diagnostic criteria, defined as the presence of tympanic membrane erythema, tympanic membrane bulging, or otorrhea not due to acute otitis externa.
> Study protocol
All study variables were defined a priori and described in the operations manual. Researchers and research coordinators received training in standardized data abstraction and entered their information electronically into REDCap, a secure web-based electronic database.
Data collected included demographic characteristics, medical history, presenting symptoms, physical examination in the ED, laboratory data, findings on evaluation by otolaryngology consultants when available, and hospital management of hospitalized infants.
To minimize bias associated with abstracting potentially more subjective descriptions from medical records, restrictive keywords were provided to guide the
determination of sick appearance, respiratory distress and clinical dehydration. These findings were categorized as present, not present, or unclear. To assess inter-rater reliability, a second abstracter at each site reviewed the medical records and
performed an independent assessment of clinical appearance, respiratory status, and hydration status for a 10% random sample of all infants, as well as for all infants with IBI or adverse outcomes.
> Outcome measures
The primary outcomes were IBIs and adverse events associated with AOM. IBIs included bacteremia and bacterial meningitis, which were defined as the growth of a pathogen in blood culture or CSF culture, respectively.9
Bacteria not predefined as contaminants or pathogens, or whose pathogenicity categorization by the treating physician was unclear or differed from the prespecified list, were reviewed by a pediatric infectious disease specialist co-investigator for pathogenicity.
For those infants who did not have CSF obtained for culture, the presence or absence of bacterial meningitis was determined through chart review. If the infant had a follow-up visit to the index hospital within 30 days of the ED visit without mention of bacterial meningitis, this result was considered negative.
AOM-associated adverse events were defined as any substantial complication resulting from or potentially associated with AOM, including, but not limited to, development of mastoiditis, sepsis, or death.
Urinary tract infection (UTI) was not considered an adverse event associated with AOM because the binding causal mechanism was unclear. However, information on urinalysis was collected because it is part of the evaluation of a severe bacterial infection in infants.
For children discharged from the ED, medical records were reviewed to identify ED follow-up visits within 72 hours. Additionally, medical records were reviewed for any outpatient physician visits or hospitalizations within 30 days of discharge from the ED. For any identified visits, records were evaluated for evidence of IBI or adverse events.
Secondary outcomes were (1) obtaining blood cultures in the ED (binary), (2) obtaining CSF cultures in the ED (binary), and (3) hospitalization (binary) due to the index ED visit. Each of these outcomes was stratified by age and site to assess variation in care.
> Sample size
Because IBI was expected to be a rare event, the sample size was determined based on the prevalence of blood culture testing, a secondary outcome. With targeted assessment of up to 20 independent variables associated with this outcome, we aimed to study at least 200 infants for whom blood cultures were obtained. Pilot trials at the main site showed that blood cultures were obtained for 30% of infants; therefore, the goal was to enroll a minimum of 667 children.
> Statistical analysis
To improve clinical interpretation, several continuous variables were categorized. Age was classified as 0 to 28, 29 to 56, and 57 to 90 days based on cutoffs generally used when stratifying the risk of IBI in febrile infants.10,11
Tests generally available during the ED stay were categorized as not performed, performed and within normal limits, and performed and abnormal. For complete blood count, white blood cell (WBC) count between 5 and 15 x 103 cells/µl was considered normal; values outside this range were considered abnormal.12
Urinalysis was considered abnormal if there was leukocyte esterase, nitrites, or > 5 WBCs per high-power field, similar to previous studies.13,14 Given the challenges in interpreting CSF leukocyte counts if the lumbar puncture was traumatic, 15 these results were not categorized.
The prevalence of IBI, adverse events, and diagnostic tests obtained were described with proportions and 95% confidence intervals [CI]. For all association tests regarding variation in practices (e.g., testing, hospitalization), generalized linear mixed effects models were used to account for clustering effects at the site level.
For each outcome, we first examined the association between the outcome and a potential explanatory variable, including infant characteristics, clinical symptoms, and physical examination findings, using a univariate generalized linear mixed-effect model.
For subsequent multivariable models, potential collinearity between predictor variables using generalized variance inflation factors was first assessed. Those predictors with univariate P value <0.1 and no evidence of multicollinearity were included in multivariable models to evaluate an association with the following outcomes as fixed effects:
(1) obtaining a blood culture,
(2) obtaining a CSF culture, and
(3) hospitalization. Hospital sites were used as random effects.
Interrater reliability for assessment of general appearance, respiratory distress, and dehydration was measured using unweighted Cohen k, a scale from 0 to 1 where 0 indicates lack of agreement and 1 indicates perfect agreement.16 All statistical analyzes were conducted using Statistical Program for the Social Sciences Version 26 (IBM SPSS Statistics, IBM Corporation, Armonk, NY) or R (R version 3.6.3, 2019).
| Results |
5270 infants were assessed for eligibility; of these, 1637 (31.1%) met the study criteria. Sites contributed 0 (1 site) to 386 cases, with a median of 30 cases, ranging from 5 to 11 years of review.
> Patient characteristics
The median age of included infants was 68 days (interquartile range [IQR]: 49–80 days). One thousand four hundred fifty-nine (89.1%) infants met the AAP simplified diagnostic criteria for AOM.
The remaining 178 children had ≥ 1 of the following examination findings reported as supporting evidence for AOM: opacification of the tympanic membrane (n = 113), pallor of the tympanic membrane (n = 57), decreased visualization of the structures of the middle ear (n = 25), middle ear effusion (n = 9), tympanic membrane perforation (n = 8), or decreased tympanic membrane mobility with insufflation (n = 5).
> Prevalence of IBI and adverse events associated with AOM
None of the 278 infants with blood cultures had bacteremia (0 of 278; 0%, 95% CI: 0% –1.4%). No infants were diagnosed with bacterial meningitis based on CSF culture results (0 of 102; 0%, 95% CI: 0% – 3.6%) or in combination with 30 days of follow-up (0 of 672; 0 %, 95% CI: 0% –0.6%).
Two of 645 infants (0.3%; 95% CI: 0.1%-1.1%) with a history of 30-day follow-up or hospitalization experienced adverse events. One of these infants was treated for culture-negative sepsis, an adverse event potentially related to AOM, although review of medical records suggested that the primary cause of his illness was likely severe dehydration secondary to milk protein allergy. . The other presented a diagnosis of lymphadenitis. Two infants were subsequently diagnosed with urinary tract infection.
> Management in the ED
Rates of diagnostic testing and hospitalization varied by site. One-fifth of all infants (355 of 1637, 21.7%) had ≥ 1 diagnostic test obtained for infectious disease.
One third of infants younger than 28 days underwent lumbar puncture (34/100; 34.0%). Of 1179 infants with symptoms of upper respiratory tract infection, 58 (4.9%) underwent lumbar puncture and 162 (13.7%) had blood cultures obtained.
One hundred and seventy-three infants (10.6%) received an intravenous or intramuscular antibiotic in the ED. Of the 1450 children discharged from the ED (88.6% of the total), 1311 (90.4%) were prescribed an oral antibiotic. Amoxicillin was the most frequently prescribed antibiotic (n = 1228; 93.7%), followed by amoxicillin-clavulanic acid (n = 56; 4.3%) .
The 72-hour readmission rate was 4.3%. Of the 63 neonates who returned, 15 (23.8%) were hospitalized for related reasons (median age: 49 days, IQR: 37–66). One patient returned 3 weeks after the initial ED presentation with lymphadenitis, as mentioned above.
> Factors associated with diagnostic tests and hospitalization
None of the potential predictor variables evaluated for inclusion in the effects models were collinear (generalized variance inflation factors < 3). Adjusting for covariates, older infants were less likely to have a blood culture, lumbar puncture, or be hospitalized compared to infants 0 to 28 days of age.
In all 3 models, history of ear discharge was significantly associated with testing for IBI and hospitalization. Decreased urination, respiratory distress, and obtaining a complete blood count (regardless of whether the results were normal or abnormal) were also independently associated with hospitalization.
For the subanalysis of infants meeting simplified AAP diagnostic criteria, minor differences were observed in the point estimates of the predictor variables for each model.
| Discussion |
In this international, multicenter study of afebrile infants ≤ 90 days of age with a clinical diagnosis of AOM, the prevalence of IBI and adverse events associated with AOM was low. Despite the low probability of IBI in this population, more than one-fifth underwent diagnostic testing for IBI and were hospitalized. This practice varied by site and was largely driven by age, with younger infants more likely to undergo invasive testing and hospitalization.
Using data from this study, the authors suggest that given the low rates of IBI and adverse events, outpatient management without testing for IBI is reasonable for most afebrile infants with a clinical diagnosis of AOM.
These data can be used to help guide the clinical management of afebrile infants with physician-diagnosed AOM, which is not included in the current AAP practice guideline for AOM.2 Variation in testing for IBI was expected and not surprising. that younger age seems to be the main driver leading to the evaluation of IBI either with blood or CSF tests.
Increased testing in younger children is likely related to the lack of data on how AOM modifies the risk of IBI, concern about possible IBI in younger infants based on data from the febrile infant population, and the reluctance to start antibiotics without any evaluation for IBI.
However, it is suggested by these data and previous information that the risk of IBI is low in afebrile infants with AOM.4-6 Furthermore, the prevalence of IBI in afebrile infants with AOM appears to be lower than the 0.8% prevalence. 2.5% described in the population of febrile infants with AOM.4,5
Of note, approximately three-quarters of the infants in this study had symptoms of upper respiratory tract infection, which can lead to viral AOM.17 Inclusion of these infants, who may have a lower likelihood of IBI than those with Bacterial AOM, could have led to an underestimation of the prevalence of IBI. However, existing data do not provide clarity regarding the ability of
distinguish viral from bacterial AOM without performing tympanocentesis. Given that >85% of older infants and children with a clinical diagnosis of AOM have identifiable bacterial otopathogens18 that could potentially spread to the blood or CSF, it is understandable why clinicians would treat infants with AOM conservatively, regardless of the presence of concurrent viral diseases.
A major challenge in any study of infants with AOM is to ensure that AOM is, in fact, present.
The diagnosis of AOM is unquestionably difficult.
However, in this young age group, physicians may be less inclined to make this diagnosis without clear evidence of AOM on examination, because once it is made, they must make difficult decisions regarding the need for invasive diagnostic testing. , systemic antibiotics, and hospitalization.
In previous studies, researchers have included examination by otolaryngologists, who often use surgical microscopes and/or tympanocentesis results to support the clinical diagnosis of AOM.3,4,6,7 However, these methods are not used consistently. routine in current clinical practice.19
Although examining the ears of young children is technically challenging, previous investigators have demonstrated the successful use of otoscopy to diagnose AOM in infants younger than 8 weeks, with 52% to 85% of middle ear effusion cultures producing growth. of true otopathogens after tympanocentesis.4,20
This study had limitations:
1. First, given its retrospective design, the authors were unable to ensure the integrity and accuracy of the clinical data. Interval agreement for the occurrence of dehydration was fair but was higher for general appearance and respiratory distress.
2. Second, not all infants had testing for IBI, potentially leading to an underestimation of its prevalence. However, the authors minimized misclassification of infants with meningitis by reviewing any available medical consultation within 30 days for discharged infants.
It is recognized that the use of follow-up visits and medical record review as surrogates for identifying those with bacterial meningitis is likely not to fully capture events occurring outside of EDs and hospital systems. However, most participating EDs are referral centers for the sickest children in their respective regions, making presentation to an alternative hospital less likely.
Additionally, all but one site had the ability to review some or all outpatient records from hospital-affiliated pediatric practices, increasing the ability to capture those infants presenting in the outpatient setting. Similar assumptions were not made regarding bacteremia, which may be transient or respond to standard treatment for AOM.
3. Third, infants whose discharge codes did not include AOM could have been missed, but this concern was addressed by screening all infants with positive blood or CSF cultures for the missed diagnosis of AOM.
4. Fourth, although factors associated with diagnostic testing and hospitalization were identified, these management decisions may have been driven by factors that cannot be captured retrospectively and not primarily by the AOM diagnosis.
This study included only a relatively small number of infants ≤28 days, which may reflect diagnostic challenges or the lower prevalence of AOM in this age group. Therefore, the conclusions of this study should be easily applied to infants older than 28 days. Finally, these findings are not generalizable to febrile infants.
| Conclusions |
In this cohort of afebrile infants with physician-diagnosed AOM, IBI and IBI-associated adverse events were rare. Rates of diagnostic testing and hospitalization varied by site and were substantial in contrast to the low prevalence of IBI and adverse events.
Based on these findings, outpatient management without diagnostic testing for IBI may be reasonable for most afebrile infants with AOM.
