Purpose of the study: To investigate whether video analysis, in addition to self-reported paper audit, could elucidate expected differences in the content of two interventions.
Background: We have completed a randomised clinical trial comparing two types of physiotherapy for subacute low back pain (“hands on” physiotherapy versus a pain management programme). An essential component in conducting clinical trials is to audit the interventions to check for compliance with the protocol. We use two approached:
self complete proforma
video recording.
Methods: i) Treatment content was recorded on a proforma by the physiotherapists after each session.
ii) A check-list of treatment modalities was constructed from this proforma. Twelve sessions were recorded on video (one new and one review patient for each therapist). The recordings were rated by 3 blinded, independent observers using the checklist. These were compared with the self-report audit forms relating to the same physiotherapy session.
Results: Analysis of the videos showed good levels of agreement (67%) between the 3 observers. Agreement between the video content and paper audit was also good (84%, _ = 0.59). The complete paper audit revealed clear differences between the treatment arms. Patients undergoing the “handson” treatment received manual therapy, whereas patients in the pain management group had specific issues addressed in the course of the consultation.
Conclusions: Feasible, reliable methods of confirming the content of interventions delivered in pragmatic trials are difficult to achieve. Self report paper audits are simple but rely upon the honesty and accuracy of the completer, and may not pick up subtle differences in approach. Video recording is time consuming, may be threatening to the treating practitioner and patient, and is difficult to analyse. A compromise approach involving sample video recordings along with paper self complete audit was able to validate the content of the treatments delivered.
