The implementation of menus based on nutrition guidelines in UK care homes proved challenging, although some changes were successfully embedded in routine practice (e.g., substituting saturated with polyunsaturated margarine in baking). It proved difficult to build collective understanding of and commitment to the study, resulting in inconsistent implementation; similar issues with lack of compliance with nutritional interventions in care homes have previously been reported . The four key constructs of NPT  proved useful in understanding the barriers to implementation. Most previous studies using NPT have focused primarily on collective action [45, 49]; our work highlights the importance of the remaining constructs, in particular, the critical role of coherence. In home 2, where the cooks refused to implement the modified menus, our experience can be conceptualized as a recursive loop, whereby the failure of the intervention to make sense (coherence) and to engage staff (cognitive participation) resulted in some staff acting out their resistance (collective actions) and bringing about outcomes that fulfilled their expectations, in a self-fulfilling prophecy (reflexive monitoring).
While we used the constructs of NPT to understand the findings, the barriers identified are largely consistent with previous work on guideline implementation. The priority given to personal knowledge over scientific evidence by care staff [58, 59] led to some staff contesting the value of the nutrition guidelines. Issues relating to role conflict and perceived incompatibility with other goals [60–63] have undermined the implementation of guidelines on lifestyle management [64, 65] and nutrition in other contexts [33, 66]. Given this uncertainty over the legitimacy of the nutrition guidelines, the concept of relative advantage was key [61–63]. However, staff who were satisfied with existing menus were more attuned to the potential risks of implementation, particularly given the status of food and mealtimes as the “highlight of the day” . In this context, the lack of observable benefits was a significant barrier. These factors individually and collectively undermined the coherence or sense of the intervention for many staff, leading to a lack of investment in the nutrition guidelines. The situation was exacerbated in most homes by the absence of strong leadership, which is well-established as a facilitator of guideline implementation (e.g., [33, 61, 68].
Previous initiatives to improve nutrition have often provided additional staff to deliver aspects of the intervention (e.g., [15, 69]). Although the study dietitian provided training and facilitated the process of menu development, the day-to-day implementation of the modified menus had to be achieved within existing resources. Managers’ commitment to the nutrition guidelines did not extend to using scarce resources to facilitate implementation. Despite the importance of supporting guideline implementation with additional financial and human resources [15, 32, 58, 61, 70], we were unable to secure these. The lack of nutritional knowledge and reliance on personal knowledge documented in previous studies [58, 71] were also identified in the present study; furthermore, the limited training provided, while valued, was insufficient to enable cooks to modify menus and recipes without the continued support of the study dietitian.
The value of Normalization Process Theory
The process of using NPT to identify real and ideal conditions for implementation  was useful in identifying potential strategies to address the barriers identified. One possible area for further development of NPT would be to link the theoretical constructs of NPT to specific behavior-change techniques; this would increase the practical utility of the theory. NPT highlighted barriers related to the work of implementing the nutrition guidelines; using an alternative theoretical framework, such as the Promoting Action on Research Implementation in Health Services (PARiHS) framework , might have directed our attention more to the process of facilitation, in particular, the skills and attributes required for facilitation (including understanding, nurturing staff, and support for learning ), but would not necessarily have enabled us to identify so clearly issues relating to the lack of coherence of the nutrition guidelines.
Implications for wider implementation of the nutrition guidelines
Strategies that may facilitate implementation of nutrition guidelines include:
ensuring that all staff are well briefed on the rationale for, and short- and long-term benefits of, the nutrition guidelines (coherence);
facilitating ownership of the modified menus (to the degree preferred by staff) and focusing on working towards rather than implementing the nutrition guidelines (cognitive participation);
providing ongoing training in the principles underlying the nutrition guidelines, menu analysis, and strategies for adapting recipes (collective action);
agreeing on outcome measures and a process for collecting information to review the impacts of the modified menus (reflexive monitoring).
An implementation team with a broad range of skills is needed to effectively implement these strategies, in addition to adequate resources. While not explored in the present study, policies on procurement of ingredients merit further exploration, since the most successful and enduring changes resulted from simple substitution of ingredients .
Limitations of the study
We studied five care homes in the North East of England. Many of the factors influencing implementation of the nutrition guidelines were identified in all of the homes. The emergence of some new factors in Homes 4 and 5, however, suggests that data saturation may not have been achieved. While the sample of homes was diverse in terms of organization and socioeconomic status, they were public sector homes in one geographical region. Additional factors influencing implementation may emerge in privately run homes and those catering to more diverse client groups.
Facilitation was primarily provided by the study dietitian, who typically worked with individuals or groups at the contemplation or action stage in the cycle-of-change model . In the present study, many staff were not at this stage; a greater emphasis on facilitation activities targeted at planning for change  might usefully have addressed staff reservations about the nutrition guidelines.