National assessment of validity of coding of acute mastoiditis: a standardised reassessment of 1966 records

Objectives To investigate the internal validity of the diagnosis code used at discharge after treatment of acute mastoiditis. Design Retrospective national re-evaluation study of patient records 19932007 and make comparison with the original ICD codes. Setting All ENT departments at university hospitals and one large county hospital department in Sweden. Participants A total of 1966 records were reviewed for patients with ICD codes for in-patient treatment of acute (529), chronic (44) and unspecified mastoiditis (21) and acute otitis media (1372). Main outcome measures ICD codes were reviewed by the authors with a defined protocol for the clinical diagnosis of acute mastoiditis. Those not satisfying the diagnosis were given an alternative diagnosis. Results Of 529 records with ICD coding for acute mastoiditis, 397 (75%) were found to meet the definition of acute mastoiditis used in this study, while 18% were not diagnosed as having any type of mastoiditis after review. Review of the in-patients treated for acute media otitis identified an additional 60 cases fulfilling the definition of acute mastoiditis. Overdiagnosis was common, and many patients with a diagnostic code indicating acute mastoiditis had been treated for external otitis or otorrhoea with transmyringeal drainage. Conclusions The internal validity of the diagnosis acute mastoiditis is dependent on the use of standardised, well-defined criteria. Reliability of diagnosis is fundamental for the comparison of results from different studies. Inadequate reliability in the diagnosis of acute mastoiditis also affects calculations of incidence rates and statistical power and may also affect the conclusions drawn from the results.