Validation of complex interventions in a low back pain trial: selective video analysis cross-referenced to clinical...

Auditing the interventions used in clinical trials to demonstrate that treatment protocols were adhered to and that discrete modalities were delivered, is essential in establishing internal validity. Pragmatic studies involving complex interventions provide a particular challenge. The value of case report forms which document individual treatment content relies upon accurate recall and recording by the clinician. This report describes a method to validate this type of data in a randomised clinical trial, by using video analysis on a sample of patient therapist interactions.

Six physiotherapists used standardised case report forms to record the treatment given to 402 patients recruited into the two treatment arms of a low back pain clinical trial. An audit checklist comprising the principal components of both management approaches was designed. Twelve treatment sessions were video recorded. Three independent clinicians viewed the recordings, using the checklist to identify the treatment content. These data were compared for agreement with the matching case report forms. The content of the interventions delivered to all of the trial patients was then audited.

Agreement between the video observers, and between the observers and the case report forms ranged from moderate to very good (kappa=0.45-0.82). When compared with the video observations, some under-reporting of treatment occurred on the case report forms. Overall, 80% of the patients received the treatment to which they were allocated.

This study has demonstrated that selective use of video recording, cross-referenced to clinical case notes may provide a feasible and relatively inexpensive means of establishing the internal validity of the interventions used in a pragmatic clinical trial.