Introduction
Since the advent of antibiotic treatment, there has been a drastic decline in the incidence of acute otitis media episodes progressing to acute mastoiditis, decreasing from 0.4 per cent in 1954 to 0.004 per cent in the 1980s. Previously, the frequency was reported to be up to 20 per cent in acute otitis media cases.Reference Nussinovitch, Yoeli, Elishkevitz and Varsano 1 – Reference Vassbotn, Klausen, Lind and Moller 3
Acute mastoiditis remains the most common intratemporal complication of otitis media in the paediatric population.Reference Mattos, Colman, Casselbrant and Chi 4 Peak incidence occurs at age two to three years, with twice as many children below the age of two years developing acute mastoiditis.Reference Kværner, Bentdal and Karevold 5 Younger children are vulnerable to infections because of immature immunological defences, predisposing them to otitis media risk factors, with an increased risk of progression to acute mastoiditis.Reference Kværner, Bentdal and Karevold 5 , Reference Spratley, Silveira, Alvarez and Pais-Clemente 6
There have been recent reports of increasing susceptibility of younger children to acute mastoiditis and a rise in hospital admission for paediatric acute mastoiditis cases.Reference Spratley, Silveira, Alvarez and Pais-Clemente 6 , Reference Luntz, Brodsky, Nusem, Kronenberg, Keren and Migirov 7 The emergence of antimicrobial resistance and the masking of clinical signs by prior antibiotic therapy are possible hypotheses raised by some authors on contributors to the rising incidence of acute mastoiditis.Reference Van den Aardweg, Rovers, De Ru, Albers and Schilder 2 , Reference Spratley, Silveira, Alvarez and Pais-Clemente 6 , Reference Kordeluk, Kraus and Leibovitz 8 A Dutch paper published in 2001 suggested an association between restricted antibiotic prescribing in paediatric acute otitis media cases and a higher incidence of paediatric acute mastoiditis.Reference Van Zuijlen, Schilder, Van Balen and Hoes 9 A later study analysing the rates of antibiotic prescription in paediatric acute otitis media between 2002 and 2003 in the Netherlands reiterated similar concerns, noting that under-prescription of antibiotics occurred in up to 11 per cent of acute otitis media consultations.Reference Akkerman, Kuyvenhoven, van der Wouden and Verheij 10 However, there has been conflicting evidence on the incidence of acute mastoiditis in recent years, and a clear trend presented as population-based information has not been demonstrated.Reference Kordeluk, Kraus and Leibovitz 8
In a systematic review published by van den Aardweg et al., only 26 studies included the diagnostic criteria of acute mastoiditis, and the criteria were divergent.Reference Van den Aardweg, Rovers, De Ru, Albers and Schilder 2 The most frequently reported diagnostic criteria were post-auricular swelling, erythema, tenderness and auricle protrusion.Reference Van den Aardweg, Rovers, De Ru, Albers and Schilder 2
There has been a lack of consensus regarding the diagnosis and management of acute mastoiditis. Treatment of paediatric acute mastoiditis differs between hospitals depending on presentation, severity, the presence of complications and surgeons’ expertise.Reference Van den Aardweg, Rovers, De Ru, Albers and Schilder 2 There is considerable disparity in the rates of conservative and surgical management of acute mastoiditis between different reports.Reference Vassbotn, Klausen, Lind and Moller 3 Most notably, rates of mastoidectomy for the treatment of paediatric acute mastoiditis vary significantly, ranging from 9 to 88 per cent between different studies.Reference Van den Aardweg, Rovers, De Ru, Albers and Schilder 2 Historically, simple mastoidectomy was the mainstay of surgical treatment. However, there has been a recent shift towards a conservative approach for acute mastoiditis and its complications.Reference Van den Aardweg, Rovers, De Ru, Albers and Schilder 2 , Reference Vassbotn, Klausen, Lind and Moller 3 , Reference Psarommatis, Voudouris, Douros, Giannakopoulos, Bairamis and Carabinos 11 Conservative treatment consists of antibiotics alone, antibiotics plus myringotomy with a ventilation tube (grommet), or grommet insertion and abscess drainage.
It is imperative to treat acute mastoiditis promptly and effectively in order to prevent potentially fatal intracranial complications that can arise from progression of the condition.Reference Psarommatis, Voudouris, Douros, Giannakopoulos, Bairamis and Carabinos 11 Possible intracranial and extracranial complications include subperiosteal abscess, suppurative labyrinthitis, facial paralysis, lateral sinus thrombophlebitis, meningitis, epidural or subdural abscesses, brain abscess, Bezold's abscess, and Gradenigo's syndrome.Reference Vassbotn, Klausen, Lind and Moller 3 , Reference Gliklich, Eavey, Iannuzzi and Camacho 12 The incidence of acute mastoiditis related intracranial complications remains high, at between 5 and 29 per cent, as seen in different studies.Reference Luntz, Bartal, Brodsky and Shihada 13 Ginsburg et al.Reference Ginsburg, Rudoy and Nelson 14 and Luntz et al.Reference Luntz, Brodsky, Nusem, Kronenberg, Keren and Migirov 7 concluded that despite the reduction in occurrence of acute mastoiditis, the rates of severe complications remain significant.
There is controversy regarding the best management of acute mastoiditis in children, and no study has reviewed the efficacy of individual treatment modalities. This systematic review aimed to investigate the most efficacious treatment modalities and their outcomes in paediatric acute mastoiditis by comparing and analysing current evidence in the literature.
Materials and methods
The Preferred Reporting Items for Systematic Reviews and Meta-Analyses (‘PRISMA’) statement was followed in this review.
PubMed, Web of Science and Cochrane Library databases were searched from the date of each database inception to 14 December 2016. The search was performed using the following combinations of keywords: ‘mastoiditis’ and ‘management’; ‘mastoiditis’ and ‘surgery’; ‘mastoiditis’ and ‘conservative’; ‘mastoiditis’ and ‘medical’; ‘mastoiditis’ and ‘antibiotic’; ‘mastoiditis’ and ‘myringotomy’; ‘mastoiditis’ and ‘tympanostomy’; ‘mastoiditis’ and ‘grommet’; ‘mastoiditis’ and ‘drainage’; and ‘mastoiditis’ and ‘mastoidectomy’. These were used in conjunction with the Medical Subject Headings ‘mastoiditis’ and ‘child’. No language restrictions were applied to the database searches. A search of references was performed to identify additional relevant articles.
The inclusion criteria were: abstracts, case reports, case series, literature reviews, retrospective analyses, clinical trials, randomised controlled trials, and systematic reviews that discussed treatment and the treatment outcomes of paediatric acute mastoiditis. Data were from children aged less than 18 years, obtained from articles available in full-text form.
We categorised treatment as: (1) medical therapy only (oral and parenteral antibiotic therapy); (2) medical therapy and myringotomy, with or without grommet insertion; (3) medical therapy and myringotomy plus subperiosteal abscess drainage; and (4) medical therapy and formal mastoidectomy, with or without myringotomy. Studies that did not report the outcomes based on the categories mentioned, those that failed to specify what ‘conservative’ or ‘surgical’ management encompassed, and studies involving patients with co-existing cholesteatoma, intracranial complications, immunological deficiency, chronic mastoiditis or cochlear implants, were excluded from the review.
Two reviewers screened titles and abstracts independently according to the predefined inclusion and exclusion criteria. The full texts of potentially relevant studies were subsequently examined. Disagreements between reviewers were resolved by discussion and examination of the full text. The following information was independently extracted by the reviewers: number of patients, study design, patient age, use of antibiotic therapy prior to hospital admission, treatment modality, complications and sequelae.
Results
A total of 520 studies were identified following a combined search of electronic databases, with an additional 2 studies identified from the retrieved articles. After removal of duplicates, 443 studies remained. Sixty-eight articles were selected for full-text assessment after abstract screening. Twenty-one articles were eventually selected for final inclusion and data extraction.Reference Enoksson, Groth, Hultcrantz, Stalfors, Stenfeldt and Hermansson 15 – Reference Moffett and Dalton 35 The Preferred Reporting Items for Systematic Reviews and Meta-Analyses flow diagram for article inclusion is shown in Figure 1.
Fig. 1 Flow diagram for article inclusion. AM = acute mastoiditis
Table I summarises the characteristics of the included studies.Reference Enoksson, Groth, Hultcrantz, Stalfors, Stenfeldt and Hermansson 15 – Reference Moffett and Dalton 35 Of the 21 studies included, 12 were cohort (retrospective) studies, 5 were case series and 4 were case reports. The studies comprised a total of 564 patients, whose ages ranged from 1 month to 16 years. Of this group, 253 cases were complicated with subperiosteal abscesses. (An indeterminate number of complicated acute mastoiditis cases reported in the study by Harley et al.Reference Harley, Sdralis and Berkowitz 27 (1997) were not included in this figure.)
Table I Summary of included studies
Mo = months; y = years; SPA = subperiosteal abscess; AM = acute mastoiditis; TM = tympanic membrane; med = medical treatment; CT = computed tomography; AOM = acute otitis media; MEE = middle-ear effusion; OM = otitis media
Medical management
All patients were started on parenteral antibiotics upon admission. The initial antibiotics prescribed in the reviewed studies were based on epidemiological knowledge along with treatment site guidelines (if any). Subsequent changes were made to the antibiotics prescribed based on culture and sensitivity results obtained.
A total of 122 patients (21.6 per cent) were managed medically with antibiotics alone (Table II). Of these, 117 (95.9 per cent) were treated successfully; no further otological problems were noted during follow up. Two patients required further treatment with myringotomy because of persistent middle-ear effusion, and another three patients required subsequent mastoidectomy because of insufficient improvement following medical management or the development of subperiosteal abscess during hospitalisation.
Table II Summary of treatment and follow-up outcomes
SPA = subperiosteal abscess; AM = acute mastoiditis; AOM = acute otitis media; VT = ventilation tube; OM = otitis media; I&D = incision and drainage
Surgical management
Surgical procedures were performed in 442 patients. Surgical procedures were categorised as either ‘conservative’ or ‘extensive’. ‘Conservative’ surgery refers to myringotomy, with or without grommet insertion, and/or subperiosteal abscess drainage (either via needle aspiration, or incision and drainage), while ‘extensive’ refers to mastoidectomy (simple/cortical or radical).
Conservative surgery was carried out in 268 patients (Table II). Of these, 160 underwent myringotomy, with or without grommet incision, and 108 underwent subperiosteal abscess drainage. Of the 160 patients who underwent myringotomy, 3 were complicated cases with a co-existing subperiosteal abscess, 1 was complicated with hearing impairment, 1 with mastoid abscess and 155 were uncomplicated cases. Of the patients who underwent conservative surgery, 258 were reported to be cured on follow up (96.3 per cent).
Of 174 patients who underwent extensive surgery (mastoidectomy), 155 were noted to have successful treatment (89.1 per cent cure rate).
Clinical outcome on follow up
In the studies reviewed, follow-up periods ranged from 1 month to 14 years (follow-up duration was not specified in 9 studies). All of the studies had a follow-up period of at least three months, except the study by Moffett and DaltonReference Moffett and Dalton 35 (1949) with no reason specified.
Of the patients, 530 (94.0 per cent) were considered to have been successfully treated, with no long-term sequelae noted. They were reported to have ‘no treatment failures’, to have ‘recovered’, to have been ‘cured’, to have ‘no recurrences’, to have ‘no further hearing trouble’ or to be ‘well and asymptomatic’.
The most frequently observed adverse outcome was recurrent acute mastoiditis (23 cases). Other adverse outcomes included: recurrent acute mastoiditis, recurrent acute otitis media, recurrent subperiosteal abscess, persistent hearing impairment, cholesteatoma, persistent tympanic membrane perforation and mastoid abscess development.
In summary, the cure rates of those who received medical therapy only, abscess drainage, myringotomy and mastoidectomy were 95.9 per cent, 94.4 per cent, 97.5 per cent and 89.1 per cent, respectively. This includes both uncomplicated and complicated acute mastoiditis cases. Twenty-four cases required mastoidectomy after the failure of improvement following initial therapy with either medical or conservative surgical treatment.
Secondary management
In the studies reviewed, 31 cases in total required subsequent management (Table II). The indications for additional management were: treatment failure within a threshold period of 48–72 hours, or the development of a new complication (i.e. subperiosteal abscess). Amongst these 31 cases, 25 required mastoidectomy as the definite treatment; in 15 cases this was because of initial treatment failure within 48–72 hours, and in 2 cases it was because of subperiosteal abscess development. Of the remaining cases that required subsequent mastoidectomy, one patient had a zygomatic abscess and seven had unspecified reasons for the progression to mastoidectomy.
Uncomplicated acute mastoiditis treatment
A total of 250 acute mastoiditis cases were uncomplicated (excluding cases from Harley et al.Reference Harley, Sdralis and Berkowitz 27 ); that is, they were simple acute mastoiditis cases, without associated complications (Table III). Medical therapy only was received by 104 patients; conservative therapy with myringotomy, with or without grommet insertion, was undertaken in 129 patients; and mastoidectomy was performed as the primary treatment in only 17 patients. Of the 104 patients in the medical therapy only group, 5 required subsequent therapy with either conservative surgery (myringotomy) or extensive surgery (mastoidectomy) because of persistent middle-ear effusion, subperiosteal abscess development or lack of clinical improvement within 48–72 hours.
Table III Treatment and follow-up outcomes in uncomplicated acute mastoiditis cases
*Total n = 250. SPA = subperiosteal abscess; AOM = acute otitis media; AM = acute mastoiditis; VT = ventilation tube; OM = otitis media
Six patients who received myringotomy, with or without grommet insertion, required further treatment with mastoidectomy. The reasons were: subperiosteal abscess development, zygoma swelling and poor clinical improvement.
No sequelae were reported during follow up in: 96.2 per cent of those who received medical therapy only (100 of 104 cases), 97.7 per cent of those who received conservative surgery (myringotomy) (126 of 129 cases) and 76.5 per cent of patients who received extensive surgery (mastoidectomy) (13 of 17 cases). The adverse follow-up outcomes experienced by patients with uncomplicated acute mastoiditis included: recurrent acute mastoiditis, recurrent acute otitis media, cholesteatoma, subperiosteal abscess and mastoid abscess development.
Subperiosteal abscess treatment
As mentioned previously, 253 of the reported acute mastoiditis cases were complicated with subperiosteal abscess (excluding cases from Harley et al.Reference Harley, Sdralis and Berkowitz 27 ). All of these patients underwent surgical management. A total of 111 patients received ‘conservative’ surgery for treatment of subperiosteal abscess. Of those 111 patients, 12 (1 who initially underwent myringotomy and 11 who initially underwent abscess drainage) received subsequent mastoidectomy because of the lack of clinical improvement within 48–72 hours; 4 patients (1 who initially underwent myringotomy and 3 who initially underwent abscess drainage) required subsequent or repeated abscess drainage because of persistent disease or newly developed subperiosteal abscess during their clinical course. The remaining 142 patients underwent primary treatment with mastoidectomy, with no reported treatment complications. During follow up, acute mastoiditis recurrence was the most frequent outcome noted (17 cases). Table IV summarises the treatment and follow-up outcomes of subperiosteal abscess cases.
Table IV Treatment and follow-up outcomes in subperiosteal abscess cases
*Total n = 253. VT = ventilation tube; AM = acute mastoiditis; AOM = acute otitis media
The cure rates for conservative and extensive surgery in cases complicated with subperiosteal abscess were 94.6 per cent and 90.1 per cent respectively. Adverse follow-up outcomes experienced by subperiosteal abscess patients included recurrent acute mastoiditis, recurrent acute otitis media and persistent tympanic membrane perforation.
Discussion
In uncomplicated acute mastoiditis cases, a conservative approach was much more frequently undertaken than extensive surgery (93.2 per cent vs 6.8 per cent). A conservative approach constituted either medical therapy only, or medical therapy plus conservative surgery with myringotomy (with or without ventilation tube insertion). Contrarily, in cases complicated with subperiosteal abscess, extensive surgery was more commonly performed than conservative surgery (56.1 per cent vs 43.9 per cent).
The majority of the reviewed patients (94.0 per cent) had a favourable clinical outcome, with no long-term sequelae on follow up. Of these 526 patients, 22.2 per cent (n = 117) received medical therapy only, 48.7 per cent (n = 256) received conservative surgery and 29.1 per cent (n = 153) underwent extensive surgery.
Correspondingly, conservative surgery was the most commonly selected treatment modality (47.3 per cent, 267 of 564). The extensive surgery treatment group had the lowest percentage of patients with a good clinical outcome (95.9 per cent medical only vs 96.3 per cent conservative surgery vs 89.1 per cent extensive surgery).
Of subperiosteal abscess patients, 14.4 per cent (16 of 111) who received conservative surgery required subsequent therapy, with either abscess drainage or mastoidectomy, because of persistent disease. All subperiosteal abscess patients who received mastoidectomy as the definitive treatment required no further therapy. Interestingly, this extensive surgery group also had a higher number of adverse follow-up outcomes (9.9 per cent) (Table IV). Similarly, all uncomplicated acute mastoiditis patients who received mastoidectomy as definitive treatment also had the highest number of adverse follow-up outcomes (23.5 per cent). Subsequent therapy with either myringotomy or mastoidectomy was required in 4.81 per cent (5 of 104) and 4.65 per cent (6 of 129) of uncomplicated acute mastoiditis patients who received medical treatment only and conservative surgery (myringotomy with or without grommet insertion) respectively.
There is a predilection for conservative therapy (medical therapy alone and conservative surgery) in the studies reviewed, and conservative therapy has been shown to be as effective as mastoidectomy in producing favourable long-term outcomes. However, we do not know what influenced the physicians’ decisions regarding treatment, as the majority of studies involved retrospective analyses. Furthermore, the lack of consensus regarding the diagnosis and management of acute mastoiditis could have contributed to the difference in surgical rates between the reviewed studies.Reference Van den Aardweg, Rovers, De Ru, Albers and Schilder 2 , Reference Psarommatis, Voudouris, Douros, Giannakopoulos, Bairamis and Carabinos 11 The cases that received extensive surgery may have been more severe on presentation, therefore possibly having more adverse follow-up outcomes.
In this review, studies were of a low evidence level (i.e. case reports, case series and retrospective studies). Although these articles contain valuable data, selection bias may be present.Reference Yu and Tse 36 , Reference Pannucci and Wilkins 37 Hence, this review may not provide the most accurate assessment of evidence-based medicine with regard to managing paediatric acute mastoiditis.
Moreover, some studies did not clearly report their methodology; some studies did not elaborate on what was considered to be ‘conservative’ management (i.e. antibiotics only vs conservative surgery). Such articles were excluded because of difficulties in extrapolating and comparing data. However, this also correlates to a reduction in potentially valuable data for our review. Additionally, taking into consideration our exclusion criteria (e.g. non-English articles, articles without full text, and cases of intracranial complication on presentation), other sources of valuable data would inevitably have been omitted as well.
To our knowledge, this is the first ever systematic review to concentrate on the management of paediatric acute mastoiditis. It focused particularly on subperiosteal abscess, the most prevalent complication of acute mastoiditis, and aimed to provide optimal comprehensive coverage of data from relevant studies.Reference Luntz, Brodsky, Nusem, Kronenberg, Keren and Migirov 7 , Reference Enoksson, Groth, Hultcrantz, Stalfors, Stenfeldt and Hermansson 15 , Reference Oestreicher-Kedem, Raveh, Kornreich, Popovtzer, Buller and Nageris 38
A notable shortcoming in this review would be the exclusion of cases with other equally important complications (i.e. intracranial complications). This exclusion allowed the streamlining of data for comparison, but also resulted in a narrower scope of review. Other improvements to establish a more well-rounded review include an analysis of the relationship between microbiology, antibiotic regimens and treatment outcomes. Additionally, there were insufficient data to compare admission time and acute mastoiditis recurrence between the treatment modalities, which would have provided valuable information for analysis.
Generally, the treatment outcomes of paediatric acute mastoiditis were favourable. The data extracted suggest that conservative treatment (medical therapy only, and myringotomy with or without grommet insertion) has high efficacy as a first-line treatment in uncomplicated acute mastoiditis cases. In contrast, in cases of acute mastoiditis complicated with subperiosteal abscess, conservative surgery provided similar cure rates when compared to those of extensive surgery. Hence, we conclude that although mastoidectomy may be the most definitive treatment available, medical therapy with or without conservative surgery is an appropriate and safe first-line management when treating acute mastoiditis.
Several studies specified a general timeframe of 48–72 hours for gauging treatment response. A failure to show improvements within this period, or the development of new complications, were the main reasons that warranted stepping up management in these studies. Given that the data did not seem to suggest any negative outcomes associated with receiving delayed surgery, 48–72 hours seems to be a reasonable threshold for gauging the initial treatment response.
Unfortunately, no statistical data could be provided in this paper to suggest the superiority of a single treatment modality over another. Furthermore, it is not possible to foresee which cases will result in a poor response and necessitate further treatment. Future meta-analyses could be conducted to address this issue, in view of the heterogeneity of treatment outcomes and methodologies across studies.
As mentioned, proper diagnostic and therapeutic criteria of acute mastoiditis are still unavailable. This remains an area for further research considering the potential benefits it could yield, such as providing a framework for conducting a well-designed prospective study. Future studies could include more detailed reporting of outcomes, and an in-depth analysis into the relationship of other variables (i.e. antibiotic regimens, other complications on presentation) and treatment outcomes.
Conclusion
Generally, the treatment outcome of paediatric acute mastoiditis was favourable. Mastoidectomy may be the most definitive treatment available; however, the reviewed data suggest that conservative treatment alone has high efficacy as first-line treatment in uncomplicated cases of acute mastoiditis, and conservative therapy may be an appropriate first-line management when treating acute mastoiditis.
