2.1 Literature search and article selection

The systematic review was registered on PROSPERO on 23 January 2020 (PROSPERO ID: CRD42020128957).¹² A comprehensive search of Medline, Embase, Cochrane Library and Web of Science was performed on 8 November 2019, then repeated on 30 April 2021, using the search strategy in Appendix S1.

Titles and abstracts of identified articles were independently screened by two reviewers using the inclusion and exclusion criteria in Table 1A, with a third reviewer making the final decision if there was disagreement. For SBO, articles were only included if cases were explicitly secondary to EAC pathology. The selected full-text articles were reviewed by a single reviewer to confirm inclusion.

Final analysis was restricted to articles of higher quality, with an explicit case definition and ≥6 cases (Table 1B). Articles were further subdivided into ‘general’, describing unselected NOE cases, or ‘specific’, describing particular subpopulations. ‘Specific’ articles were only considered in subsections related to clinical investigations or management.

2.2 Data extraction and analysis

Key variables, shown in Appendix S2, were extracted by a single reviewer into REDCap®, a secure electronic data capture tool hosted by the University of Oxford.^{13, 14} Articles were classified by study type using definitions in Appendix S3 and by level of evidence using the Oxford Level of Evidence.¹⁵ Data analysis was conducted in Microsoft Excel (version 16.46), Prism (GraphPad Software, Version 9.0.2), Stata (version 17.0) and RStudio (version 1.3.959). The Joanna Briggs Institute (JBI) checklists were used to assess methodological quality and risk of bias; JBI was chosen due to the availability of tools for all study types included in our review (Appendix S4).¹⁶ Results were synthesised and reported according to PRISMA guidelines (Appendix S5).¹⁷ No ethical approval was required for this study.

3.1 Study characteristics

A summary of the search is presented in the PRISMA diagram (Figure 1).¹⁷ Of 1429 articles identified, 422 (30%) articles of 16 528 patients were included in the initial synthesis. The majority were single case reports (49%, 207/422) or formal case series (28%, 120/422; Figure S1A). Since the first article in 1968,³ a bimodal pattern in the number of articles published per year was seen, with an increase in the last 5 years; however, there remains a paucity of high-quality articles that are not case reports or case series (Figure 2).

Seventeen percent (70/422) of articles of 2274 patients were included in the final synthesis (Table 2), with a median of 23 patients included per article (IQR 14.8–36.0). Seventy-three percent (51/70) were formal, consecutive retrospective case series (Figure S1B) and increased in number over time (Figure S2), while quasi-experimental studies tended to be earlier. Most case series (75%, 42/56) were methodologically high quality with low risk of bias; in contrast, only 29% (4/14) of other study types had low risk of bias (Appendix S3). The majority of articles were from Israel (31%, 22/70), the USA (16%, 11/70) and the UK (7%, 5/70; Figure 3), while none were from South America, Sub-Saharan Africa, South-east Asia, or Oceania.

3.2 Diagnostic criteria

No consistent case definition was applied across 51 ‘general’ articles (Figure 4). 18% (9/51) used the case definition proposed by Cohen and Friedman⁸⁷ and 6% (3/51) used a modification. A minority (22%, 11/51) included risk factors in their case definition, commonly diabetes, or immunodeficiency.

The commonest obligatory case definition criterion was ‘failure of symptoms to improve with outpatient therapy’ (80%, 41/51); however, the definition of ‘failure to improve’ varied in duration and route of antimicrobial therapy (Table S1). Otalgia was the commonest obligatory symptom (78%, 40/51), followed by the presence of granulation tissue (77%, 39/51). Only four articles (8%, 4/51) included minimum symptom duration in the case definition and this duration varied. 31% (16/51) included P. aeruginosa isolation as an obligatory criterion.

Only 59% (30/51) of articles included radiological evidence of NOE as an obligatory criterion. Choice of modality varied, and inclusion criteria for a cohort often referenced radiological evidence from multiple modalities (Figure 4D), which may identify pathology at different stages of the disease and thus represent heterogeneous populations. Only 53% (16/30) of articles included radiological evidence on computed tomography (CT) as an obligatory criterion, despite how widely CT is used to investigate suspected NOE. Radiological criteria for diagnosis were only clearly stated in 63% (19/30) (Table S2) and varied across articles, but commonly referenced bony erosion.

3.3 Demographics and risk factors

Mean patient age was 69.2 years (range 19–101). Sixty-eight percent (1206/1776) of patients were male. Diabetes was the commonest cited risk factor, reported in 84% (1400/1668) of patients where diabetes was not part of the case definition (Table 3). While older age was an often-cited risk factor, only six articles defined a cut-off, ranging from >65 to >80 years. Six percent (61/994) of patients were immunosuppressed for reasons other than age or diabetes and 10% (109/1130) had no immunosuppressive risk factor. In one ‘specific’ case series of seven patients with AIDS, the patient age was notably younger (mean 31.8 years, range 2–46) and no patients were diabetic.⁷⁰ Another ‘specific’ case series of non-diabetic patients identified age, ischaemic heart disease and hypertension as risk factors.⁶⁹ In one ‘specific’ case series of temporal bone osteomyelitis, 27% (15/55) of patients had prior ear surgery and presented more atypically with younger ages, lower diabetes prevalence and lack of otalgia, EAC granulation or isolation of P. aeruginosa.⁷⁴

3.4 Clinical presentation

Not all ‘general’ articles reported signs and symptoms and findings were biased according to the case definition (Table 3). Acknowledging this, the commonest presenting symptom was otalgia (96%, 1249/1307) followed by otorrhoea (78%, 972/1255). Fever was infrequently reported (7%, 28/416). Granulation tissue (69%, 918/1332) and oedema/swelling (76%, 754/987) were the commonest clinical signs.

Twenty-one percent (371/1741) of patients had a facial nerve (VII) palsy and 5% (73/1447) had two or more cranial nerves affected. In two ‘specific’ case series of SBO secondary to NOE, cranial nerve palsies were more common, affecting 57% (8/14)⁷¹ and 100% (12/12)⁸¹ of patients. In one of these series, 21% (3/14) of patients also presented with neurological signs including dysarthria, dizziness, gait disturbance and hemiparesis.⁷¹

3.5 Imaging

Fourteen percent (7/51) of ‘general’ articles did not describe radiological assessment; for the remainder, imaging modalities and findings were poorly reported. CT was the commonest modality, used in 95% (42/44) of articles, followed by nuclear imaging (59%, 26/44) and magnetic resonance imaging (MRI) (32%, 14/44). Plain radiography was only used up to 1991, while MRI has been in use since 2010. No articles compared the sensitivity or specificity of different modalities. Twenty-seven percent (12/44) of articles described performing follow-up imaging, most commonly Technetium 99m and Gallium-67 scanning (four articles each), followed by CT and MRI (three articles each); however, the time interval between scans or indication for follow-up imaging was poorly described.

3.6 Microbiology

The commonest method of microbiological sampling across 40 ‘general’ articles (including 1112 patients) was a superficial EAC swab (58%, 23/40). Twenty-five (10/40) cultured EAC discharge, 13% (5/40) obtained an EAC biopsy or cultured EAC granulation tissue, and 3% (1/40) obtained deep samples under general anaesthesia.⁵¹ Only 55% (22/40) of articles reported antimicrobial susceptibility testing.

P. aeruginosa was the commonest pathogen, identified in 62% (603/967) of patients (Table 4 and Table S3); where reported, quinolone resistance was found in 8% (19/253) of isolates. Fungi were responsible for 9% (84/967) of cases and S. aureus for 6% (62/967). A causative organism was not identified in 14% (131/967). Co-infection with more than one pathogen was reported in 7% (65/967) of patients, with P. aeruginosa and S. aureus the commonest combination (seven cases). The reported incidence of culture-negative and fungal NOE appeared to increase over time, which may reflect increasing awareness that pathogens other than Pseudomonas spp. may cause NOE; indeed, this change coincided with fewer articles including Pseudomonas as part of the case definition for NOE (Table S4).

For the majority of articles, it was not possible to determine patient-level detail about pathogens isolated, risk factors for NOE, clinical manifestations of disease and outcomes in order to allow meta-analysis; articles which reported these factors were generally limited by extremely small sample sizes. While some articles reported that fungi or ciprofloxacin-resistant P. aeruginosa were associated with more severe disease,^{45, 51} others found no relationship between pathogen and patient outcome.^{34, 49, 53, 65}

3.7 Antimicrobial therapy

3.8 Surgical management

Twenty-one percent (365/1779) of patients in 71% (36/51) of ‘general’ articles underwent surgery (not including removal of granulation tissue or microsuction). The proportion of patients undergoing surgery varied from 0% to 91% per article (Figure S3) and declined over time, from 47% (173/368) in articles pre-2000 to 13% (172/1365) in articles post-2000 (Figure S4). Local debridement of necrotic bone and cartilage (60%, 208/345) was most frequently performed followed by mastoidectomy (31%, 108/345), which ranged from a simple cortical approach to canal wall down, modified radical mastoidectomy (Figure S5). Less frequently, other major surgical procedures such as lateral temporal bone resections (2%, 7/345) and facial nerve decompression (4%, 13/345) were described. The proportion of procedures considered major (including mastoidectomy, petrosectomy, TMJ excision or soft tissue debridement around the skull base/infratemporal fossa) has increased over time, from 23% and 10% of procedures, respectively, in the 1980s and 1990s, to 44%, 51% and 49%, respectively, in articles in the decades post-2000 (Figure S6). Commonest indications for surgery included failure to respond to conservative management; clinical deterioration including the development of lower cranial nerve palsies; source control (e.g., removal of sequestered bone); or debridement of the severe and advanced stage of the disease.

3.9 Hyperbaric oxygen therapy

Twenty percent (10/51) of ‘general’ articles from Israel,^{39, 43, 45, 53, 56, 60} the USA^{18, 20} or Turkey,^{58, 67} described the use of hyperbaric oxygen therapy (HBOT) (Figure S7). Indications included the failure of conservative treatment or last resort therapy (36%, 15/42); adjunctive treatment (38%, 16/42); or to assess efficacy of HBOT (26%, 11/42). There was no clear evidence to support HBOT use in NOE.

3.10 Outcomes

Eighty percent (41/51) of ‘general’ articles reported mortality data but varied in the time interval for mortality assessment (unknown to 5 years post-diagnosis). All-cause mortality was recorded for 24% (12/51) of articles, with an estimated all-cause 1-year mortality of 7% (34/471). Disease-specific mortality was recorded for 28% (14/51) of articles, with an estimated disease-specific 1-year mortality of 2% (14/589).

Many ‘general’ articles used terms such as ‘resolution’, ‘relapse’, or ‘treatment failure’, but were only clearly defined in 22% (11/51), 8% (4/51) and 22% (11/51), respectively, and definitions varied (Table S5). Only 4% (2/51) of ‘general’ articles defined ‘aggressive disease’ and 6% (3/51) defined ‘persistent disease’. Twenty-two percent (274/1224) of patients in 42 ‘general’ articles experienced treatment failure, when defined as ‘failure to improve on initial therapy requiring escalation of antimicrobial therapy, surgical intervention, use of HBOT, or death during treatment’. Seven percent (73/1053) of patients in 30 ‘general’ articles experienced relapse of disease, when defined as ‘disease recurrence following a period off antimicrobial therapy’.