Acute otitis media diagnosis in childhood: still a problem in 2023?

Our study points out that AOM diagnosis still represents a relevant issue for healthcare providers, with potential relevant consequences in terms of costs and unnecessary antibiotic treatments [10].

Our analysis showed that 27.6% of AOM diagnosis performed by paediatricians were not validated by our internal validator; this finding is in line with previous data, which hypothesized that AOM could be overdiagnosed in about 30% of children [11‐13].

The most common diagnostic pitfall detected was the misleading diagnosis of AOM in case of hyperaemic TM without bulging (32/37); the following most common mistake was the misleading diagnosis in case of opaque TM without bulging (5/37). While performing otoscopy in case of suspected AOM, particular attention should be paid to the identification of a bulging eardrum, as this finding better correlates with a bacterial infection in the middle ear and represents the most consistent sign of AOM [14‐16].

On the contrary, other otoscopic findings such as an isolated hyperaemia of the TM (frequent finding in the infant during crying), the loss of the bright triangle, the retraction of the TM and air-fluid levels should not be considered diagnostic criteria for AOM [1, 17, 18]. On the contrary, these otoscopic findings are in most cases suggestive for otitis media with effusion (OME), defined as the presence of middle ear fluid (MEE) without signs or symptoms of acute infection [19].

AOM misdiagnosis in case of OME represents one of the most common diagnostic pitfalls and still remains a frequent condition in which antibiotics are prescribed inappropriately [8, 18, 20]. An investigation by Shaikh et al. on 263 children documented how the bulging of the TM was identified in 92% of those with AOM, compared to 0% of those with OME; moreover, in the same paper the authors developed a diagnostic algorithm focused on the bulging of the TM as the starting point to discriminate between AOM and OME: if any bulging is not identified, AOM diagnosis should be set aside [21]. A more recent survey by Chiappini et al. was conducted in 2019 through a questionnaire administered to pediatricians, with a main focus on the therapeutic management of the disease: similarly to our results, only in 57.8% of cases a bulging TM was considered among the most relevant signs of AOM, while in 25.2% of cases a retraction of the eardrum was considered as a major sign [22].

Performing otoscopy can be challenging in younger children, especially in those who are scared and not cooperative; however, in our analysis, the age of the child did not influence the rate of error, as the mean age was almost equivalent in the group of correct diagnosis and in that of incorrect ones.

The rate of misdiagnosis was slightly higher among emergency room paediatricians than among primary care ones (34.2% vs. 25.2%). Compared to our results, an Australian retrospective cohort study reported lower rates of error and a good diagnostic accuracy in the emergency department, with about 13% of diagnoses classified as “unlikely”; however, it should be noted that the remainder 87% of cases included in this study was composed by those defined as “likely” (60%) plus those defined as “possible” (22%), hence the percentage of misdiagnosis could have been underestimated [23]. The slightly higher error rate in the emergency department could be at least in part related to the different setting, as children coming to hospital are often less cooperative, parents could be in distress and the physician is able to dedicate less time to the child and its family. Moreover, in our opinion, this finding could be also in part related to the mean younger age of healthcare providers in the emergency department. Previous studies reported how otoscopic skills relate to the examiner’s experience, showing significant differences between medical students, residents, paediatricians and experts [24, 25].

This issue is tightly connected to the scarce training towards otoscopy and paediatric ENT diseases in general in our country: otoscopy is indeed a learned clinical skill that requires a reference bank of “normal” otologic variations [20]. Accordingly, literature shows a certain grade of discordance in recognizing otologic conditions among medical students, trainees and practitioners [26, 27]. Unfortunately, as reported in a survey by Marchisio et al., most of the Italian paediatricians and ENTs do not receive adequate AOM education during medical school and residency, especially among pediatricians, while among ENT schools paediatric conditions are often considered a minor issue [7]. As shown by Paul et al., students who receive a specific paediatric otoscopy curriculum had significant improvements in their skills when compared those who received a routine learning program. It is therefore essential to implement training programs focused on this topic in our schools to improve diagnostic accuracy; moreover, it should be mentioned that acquired skills tend to diminish over time, emphasizing the need for a continued practice [28].

Concerning the diagnostic tools, pneumatic otoscopy is the optimal choice to identify AOM [18]; however, it requires a certain training and it is not always available in the office or in the emergency room and its use is still limited in common practice [7]. On the other hand, it should be noted that an AOM diagnosis is still possible in case of severe bulging of the eardrum associated with signs of inflammation [29]. Unfortunately, in our study we cannot provide data on the use of the pneumatic otoscope; however, in a recent Italian survey, only 9.6% of primary care paediatricians declared to use instrument [22]. Among other devices, tympanometry is surely helpful in improving diagnostic accuracy and is perceived as easy to learn and to interpret by paediatricians [30]; however, it is quite expensive and is prevalently used by ENTs, while paediatricians are resorting to this tool only in 3.9% of cases [22].

A point of strength of our study is the evaluation of AOM diagnostic matching through the direct analysis of eardrum description provided by paediatricians on clinical records, rather than the administration of a questionnaire, thus giving an accurate report on this topic. One limitation is related to the inclusion of data extracted only from the Milan area; moreover, the diagnostic concordance has been evaluated on reports, rather than with a direct evaluation of physicians’ skills.