A modified delphi consensus study of the screening, diagnosis, and treatment of tardive dyskinesia

Objective: A nominal group process followed by a modified Delphi method was used to survey expert opinions on best practices for tardive dyskinesia (TD) screening, diagnosis, and treatment and to identify areas lacking in clinical evidence. Participants: A steering committee of 11 TD experts met in nominal group format to prioritize questions to be addressed and identify core bibliographic materials and criteria for survey panelists. Of 60 invited experts, 29 (23 psychiatrists and 6 neurologists) agreed to participate. Evidence: A targeted literature search of PubMed (search term: tardive dyskinesia) and recommendations of the steering committee were used to generate core bibliographic material. Inclusion criteria were as follows: (1) review articles, metaanalyses, guidelines, or clinical trials; (2) publication in English between 2007 and 2017; (3) > 3 pages in length; and (4) publication in key clinical journals with impact factors ≥ 2.0. Of 29 references that met these criteria, 18 achieved a score ≥ 5 (calculated as the number of steering committee votes multiplied by journal impact factor and number of citations divided by years since publication) and were included. Consensus Process: Two survey rounds were conducted anonymously through electronic media from November 2017 to January 2018; responses were collected, collated, and analyzed. Respondent agreement was defined a priori as unanimous (100%), consensus (75%-99%), or majority (50%- 74%). For questions using a 5-point Likert scale, agreement was based on percentage of respondents choosing ≥ 4 ("agree completely" or "agree"). Round 1 survey included questions on TD screening, diagnosis, and treatment. Round 2 questions were refined per panelist feedback and excluded Round 1 questions with < 25% agreement and > 75% agreement (unless feedback suggested further investigation). Conclusions: Consensus was reached that (1) a brief, clinical assessment for TD should be performed at every clinical encounter in patients taking antipsychotics; (2) even mild movements in 1 body area may represent possible TD; (3) management requires an overall evaluation of treatment, including reassessment of antipsychotics and anticholinergics as well as consideration of vesicular monoamine transporter 2 (VMAT2) inhibitors; and (4) informed discussions with patients/caregivers are essential.