The Whole System Demonstrator was specifically designed to provide more robust evidence of the effectiveness and cost-effectiveness of telehealth and telecare in the management of patients with long-term health conditions. In doing so, it was seen as a way of stimulating faster implementation of remote care. The RCT element of the evaluation means that questions about benefits evidence should in part be answered. However, the other aim of the programme, with telehealth and telecare acting as a driver for new levels of integration, and providing lessons for wide-scale future implementation, was less successful.
The original WSD evaluation design attempted to combine the need for better evidence, using an RCT, with lessons about organisational change and strategic decision making, to better inform decisions about scaling up and embedding the technology into mainstream services. The emerging results support the recent suggestion for less of a focus on conducting RCTs in health, that the quest for improved evidence does not necessarily align with an RCT approach
[5, 6, 8], with more research needed which looks at ‘wicked’
 problems of the implementation and appropriateness of telecare/telehealth.
The components and constituent parts of the RCT did not exist in isolation from local implementation needs, and as the RCT was given priority these needs often became eroded or skewed. As with any pragmatic evaluation of this size, the final proposed design meant a compromise between methodological rigour and realism. A definitive RCT was required to produce the robust data policy makers and commissioners have been calling for. In contrast to other research designs that focus on knowledge in action
, in order to produce robust, reliable and universal evidence, the RCT had to be largely implemented as originally planned and detached from the complexities of the environment in which the organisations were embedded. We know that organisational implementation and working practices are influenced by a wide range of complex local, political and individual processes
[5, 6, 8]. Wide-scale rollout was not simply a question of increasing user numbers within a given domain and time span, and requires significant degrees of organisational learning about what works and what fails to work. Unfortunately, the artificiality of randomisation and associated levels of standardisation made a ‘learn, reflect and adopt’ approach very difficult. The RCT protocol was not well aligned with generating scalability lessons, iterative and participative modes of learning, and developing new levels of service integration. Action was often not possible until the trial was complete. Consequently, much of the iterative learning from the implementation process was not implemented.
Remote care technologies have often been positioned as a ‘cure all’, in both helping address the western world’s demographic time-bomb by filling gaps within the care system, and as a vehicle for new levels of system integration
[36, 37]. The White Paper from which the WSD programme originated states that it should ‘highlight the many barriers to realising this vision.’
 The idea that remote care implementation could lead to new levels of whole system working was always ambitious, with how this aim would be achieved was left ambiguous. As others have noted
, while policy support may increase an organisation’s capacity to adopt an innovation, it will not change its strategic and cultural predisposition. This WSD stands out in being the largest trial of its type ever undertaken. Despite the resources deployed, and a comprehensive and rigorous site selection process of cases with past experience of telecare and telehealth, we found that in actuality the sites were not ready to implement this level of organisational change within the timeframe given. It appears that despite a rigorous site selection process, remote care was not a sufficiently powerful driver to significantly reduce fragmentation and discontinuities across the system, and push system integration significantly forward. Nor did engaging in the WSD programme create new levels of organisational readiness as originally envisioned. Even in Cornwall, where they started off with more flexibility, in that they had less devolved and of remote care working, telecare and telehealth did not lead to greater system integration. People from different parts of the health and social care system may have worked together more than previously, but the underlying structure of the systems in place remained largely unchanged
. What this study shows is that telehealth and telecare was successfully implemented locally without significant levels of system integration occurring and without the perceived need for this to happen. If work from the programme is to be locally sustained, perhaps the original goals of integration need to be redrawn in favour of more organic notions of incremental scaling-up, that pay attention to the evolving needs of the service-in-context as it grows
. While it seems unlikely that care system redesign on any significant scale could be implemented within the relatively short two year time-frame of the WSD, like previous research
[6, 12], these results suggest that even without the constraints imposed by the RCT, the ‘holy grail’ of health and social care integration driven by remote care, is still a very long way off. What is clear is that during the WSD programme, the RCT protocol led to implementation models that were not sufficiently sensitised to local contextual differences. An important lesson is that the development of a remote care service cannot occur in a contextual vacuum treated as a pre-wrapped generic implementation package to be adapted later. Locally sensitive levers and incentives must be factored in and co-designed both from inception and along the way.