Around 20% of patients who have total knee replacement find that they experience long-term pain afterwards. There is a pressing need for better treatment and management for patients who have this kind of pain but there is little evidence about how to improve care. To address this gap we are developing a complex intervention comprising a clinic to assess potential causes of a patient's long-term pain after knee replacement and onwards referral to appropriate, existing services. The Medical Research Council recommends that development of complex interventions include several stages of development and refinement and involvement of stakeholders. This study comprises the penultimate stage in the comprehensive development of this intervention. Earlier stages included a survey of current practice, focus groups with healthcare professionals, a systematic review of the literature and expert deliberation.

Healthcare professionals from diverse clinical backgrounds with experience of caring for patients with long-term pain after knee replacement were sent a study information pack. Professionals who wished to participate were asked to return their signed consent form and completed study questionnaire to the research team. Participants rated the appropriateness of different aspects of the assessment process and care pathway from 1–9 (not appropriate to very appropriate). Data were collated and a document prepared, consisting of anonymised mean appropriateness ratings and summaries of free-text comments. This document was then discussed in 4 facilitated meetings with healthcare professional held at the future trial centres. A summary report and revised care pathway was then prepared and sent to participants for further comments.

28 professionals completed the questionnaire and/or attended a meeting. Participants included surgeons, physiotherapists, nurses, pain specialists and rheumatologists. Mean appropriateness scores ranged from 6.9 to 8.4. Taking a score of 7–9 as agreement, consensus was achieved that the assessment should be performed at 3 months post-operative by an extended scope practitioner/nurse, treatment be guided by a standardised assessment of pain, and treatment individualised. There was also agreement that referrals in the care pathway to surgical review, GP and pain clinics were appropriate. Nurse-led/self-monitoring was rated lower (6.9) because of considerations about the need to ensure that patients receive appropriate support, follow-up and referral to other services.

This work demonstrates the research methods that can be used to refine the design of a complex intervention. The process and findings enable refinement of an intervention for patients with long-term pain after knee replacement. The next stage of intervention development will assess the acceptability and reliability of the assessment process, and the usability of the intervention's standard operating procedures. The intervention will then be evaluated by a larger research team in a multi-centre randomised controlled trial, starting in late 2016.