Can implementation failure or intervention failure explain the result of the 3D multimorbidity trial in general practice: mixed methods process evaluation

Objectives: During a cluster randomised trial, (the 3Dstudy) of an intervention enacting recommended carefor people with multimorbidity, including continuity ofcare and comprehensive biennial reviews, we examinedimplementation fidelity to interpret the trial outcome andinform future implementation decisions.
Design: Mixed-methods process evaluation usingcross-trial data and a sample of practices, clinicians,administrators and patients. Interviews, focus groupsand review observations were analysed thematically andintegrated with quantitative data about implementation.Analysis was blind to trial outcomes and examined context,intervention adoption, reach and maintenance, anddelivery of reviews to patients.
Setting: Thirty-three UK general practices in three areas.
Participants: The trial included 1546 people withmultimorbidity. 11 general practitioners, 14 nurses, 7administrators and 38 patients from 9 of 16 interventionpractices were sampled for an interview.
Results: Staff loss, practice size and different administrativestrategies influenced implementation fidelity. Practiceswith whole administrative team involvement and goodalignment between the intervention and usual care generallyimplemented better. Fewer reviews than intended weredelivered (49% of patients receiving both intended reviews,30% partially reviewed). In completed reviews >90% ofintended components were delivered, but review observationsand interviews with patients and clinicians found variation instyle of component delivery, from ‘tick-box’ to patient-centredapproaches. Implementation barriers included inadequateskills training to implement patient-centred care planning,but patients reported increased patient-centredness due tocomprehensive reviews, extra time and being asked abouttheir health concerns.
Conclusions: Implementation failure contributed to lack ofimpact of the 3D intervention on the trial primary outcome(quality of life), but so did intervention failure since modifiableelements of intervention design were partially responsible.When a decisive distinction between implementation failureand intervention failure cannot be made, identifying potentiallymodifiable reasons for suboptimal implementation isimportant to enhance potential for impact and effectiveness ofa redesigned intervention.