Background: Because the diagnosis of restless legs syndrome (RLS) depends on clinical features ascertained by interview, it is important to have structured diagnostic instruments that can guide a diagnostician to an accurate diagnosis. With this aim in mind, the RLS Center at Johns Hopkins has been developing the Hopkins telephone diagnostic interview (HTDI). A previous validation was performed on a patient group. In the current report, we have extended that validation to a non-patient group drawn from on ongoing family study.
Methods: Family members from a case-control RLS study were diagnosed by telephone. Once all available family members in a given family had been interviewed, those who lived locally and had responded were asked to come to Johns Hopkins-Bayview Medical center and had dual clinical interviews to ascertain RLS diagnosis. We then compared the results between the two clinical interviews and between the HTDI and the clinical interviews.
Results: Diagnostic agreement between two expert clinicians was 93-96% (kappa 0.87-0.92). Compared to those subjects on whom the clinicians agreed, the HTDI agreed at least 90% of the time; sensitivity was at least 0.90, specificity 0.91, positive predictive value 0.86, and negative predictive value 0.94.
Conclusion: The HTDI managed a high level of diagnostic accuracy, showing only slightly less agreement than the two clinical interviewers. Because of a mean 12-month period between HTDI and clinical interview, this result also indicates that the subjects' reports of symptoms are consistent and stable. The HTDI should be useful for confirming questionnaire diagnoses or screening subjects to enter basic or therapeutic trials.