Diagnosis of Primary Ciliary Dyskinesia: Discrepancy according to different algorithms

Abstract

Diagnosis of primary ciliary dyskinesia (PCD) is challenging since no gold standard test is available. Evidence-based diagnostic guidelines exist from the European Respiratory (ERS) and the American Thoracic (ATS) Society. Their considerable differences have been described theoretically, but no study has yet compared them in a clinical setting. We compared the diagnostic outcome of patients referred to our diagnostic centre according to three different algorithms: ERS, ATS and our own algorithm (PCD-UniBe). Our algorithm is comparable to the ERS, but additionally we perform always cell culture and immunofluorescence staining. We included all patients diagnosed with PCD by at least one or without PCD by all three algorithms. We calculated agreement (Cohen’s kappa). From 108 referred patients, we included 54 and excluded 54 mostly due to missing data (e.g. nasal nitric oxide). For 46 patients (85%) the diagnosis was consistent (30x no PCD, 16x PCD). For 8 patients (15%) the outcome varied depending on the algorithm: 5 patients were diagnosed with PCD only by ATS and 1 only by PCD-UniBe; 1 patient by PCD-UniBe and ERS, but not ATS; 1 patient by ATS and PCD-UniBe, but not ERS. The agreement was substantial between ERS and ATS (κ=0.72, 95% Confidence Interval (CI) 0.53 – 0.92) and ATS and PCD-UniBe algorithms (κ=0.73, CI 0.53 – 0.92), and almost perfect between ERS and PCD-UniBe (κ=0.92, CI 0.80 – 1.00). Prevalence of incongruent PCD diagnoses was relative high in our study, highlighting the practical consequences of the differences between the existing diagnostic algorithms. An updated single international diagnostic guideline would be helpful to avoid algorithm-dependent differences in PCD diagnosis.