DISCUSSION
NOE presents a challenge both in terms of diagnosis and treatment. It is an uncommon condition which makes large studies difficult to conduct and, as such, management options remain poorly defined. Most clinicians agree that intravenous antibiotics and meticulous glycaemic control are essential, but the role of surgical intervention, amongst other factors, remains controversial. Indeed, a survey of over 200 UK otolaryngologists reveals surprising disparities in management (5).
Evidence from orthopaedic literature concerning osteomyelitis at subsites other than the external auditory canal has suggested that short courses of antimicrobial therapy may be suitable for particular patient groups (4). Short course may offer distinct advantages over long course antimicrobial therapy; such as fewer drug side effects, lower financial cost, shorter length of hospital admission, and is more convenient for patients. There are also advantages to limited surgical debridement, including de-bulking microbial load, facilitating microbiological testing, and permit histological analysis for exclusion of malignancy. Indeed other groups have suggested that deep tissue cultures may be more useful than superficial swabs (6).
To our knowledge this has not been investigated in NOE until the present study. In this series allocation to short course +/- surgery vs long course antimicrobial therapy was determined based on a historical database of outcomes in NOE treated with a standard long course of antimicrobial therapy at SRH. We previously identified (unpublished) positive prognostic factors, and where these are present the patients were allocated to short course antimicrobial therapy +/- surgery (see methods). All other patients were allocated to long course antimicrobial therapy.
Most patients in this series were elderly and most were male. The majority had extensive comorbidities - as measured by the Charleston comorbidity index; in particular diabetes mellitus, which was often poorly controlled; predisposing them to infection. Interestingly, despite the classical association with NOE, recent data suggests that diabetes mellitus is not an independent risk factor for 30-day readmission, prolonged length of stay, or discharge to a rehabilitation facility (7). Many patients in this series grew bacteria typical for NOE, such as P. Aeruginosa (8). Few demonstrated no bacterial growth, which may be a function of having received treatment prior to microbiological analysis.
Most patients in this series presented with advanced disease, as measured by otoscopy and CT findings. To be able to compare findings between patients it is important to have a scoring system to objectively describe extent of disease. We therefore developed a simple and intuitive distinct grading system for each, as outlined in methods, which offer a common language to clinicians that is currently missing in NOE practice.
Although most patients receiving short course antimicrobial therapy +/- surgery had positive outcomes, this series is small and not statistically powered to definitely compare treatment arms. Importantly, however, we identified that all but one patient in this series survived and had good outcomes. The patient who did not survive died of related comorbidities, rather than from complications of NOE. This suggests a larger trial comparing short course antimicrobial therapy +/- surgery versus long course antimicrobial therapy for non-inferiority would be feasible and may well offer distinct advantages for selected patients.