Chiropractor interaction and treatment equivalence in a pilot randomized controlled trial: an observational analysis of clinical encounter video-recordings

BACKGROUND: Chiropractic care is a complex health intervention composed of both treatment effects and non-specific, or placebo, effects. While doctor-patient interactions are a component of the non-specific effects of chiropractic, these effects are not evaluated in most clinical trials. This study aimed to: 1) develop an instrument to assess practitioner-patient interactions; 2) determine the equivalence of a chiropractor’s verbal interactions and treatment delivery for participants allocated to active or sham chiropractic groups; and 3) describe the perceptions of a treatment-masked evaluator and study participants regarding treatment group assignment. METHODS: We conducted an observational analysis of digital video-recordings derived from study visits conducted during a pilot randomized trial of conservative therapies for temporomandibular pain. A theory-based, iterative process developed the 13-item Chiropractor Interaction and Treatment Equivalence Instrument. A trained evaluator masked to treatment assignment coded video-recordings of clinical encounters between one chiropractor and multiple visits of 26 participants allocated to active or sham chiropractic treatment groups. Non-parametric statistics were calculated. RESULTS: The trial ran from January 2010 to October 2011. We analyzed 111 complete video-recordings (54 active, 57 sham). Chiropractor interactions differed between the treatment groups in 7 categories. Active participants received more interactions with clinical information (8 vs. 4) or explanations (3.5 vs. 1) than sham participants within the therapeutic domain. Active participants received more directions (63 vs. 58) and adjusting instrument thrusts (41.5 vs. 23) in the procedural domain and more optimistic (2.5 vs. 0) or neutral (7.5 vs. 5) outcome statements in the treatment effectiveness domain. Active participants recorded longer visit durations (13.5 vs. 10 minutes). The evaluator correctly identified 61% of active care video-recordings as active treatments but categorized only 31% of the sham treatments correctly. Following the first treatment, 82% of active and 11% of sham participants correctly identified their treatment group. At 2-months, 93% of active and 42% of sham participants correctly identified their group assignment. CONCLUSIONS: Our findings show the feasibility of evaluating doctor-patient interactions in chiropractic clinical trials using video-recordings and standardized instrumentation. Clinical trial design and clinician training protocols should improve and assess the equivalence of doctor-patient interactions between treatment groups. TRIAL REGISTRATION: This trial was registered in ClinicalTrials.gov as NCT01021306 on 24 November 2009.