General principles of implementation

First of all, it is important to summarise the main general principles of implementation science.10 11 Without this educational basis, it is not possible to put the implementation of a single or a set of recommendations into practice successfully. These general principles include the phases of implementation that will be described in detail.

Figure 1 outlines the general principles of implementation: (1) the multilevel approach, (2) the need to prioritise and adapt, (3) the implementation team, (4) the nature of the implementation process, (5) the need for resources and (6) the phases of implementation.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

Principles of implementation and its phases.

Connected to the multilevel approach, recommendations can influence three levels (macro, meso, micro), all of which might have an impact on implementation. The macro-level is the policy level. Depending on the country, health policymakers might decide, for example, which biological therapies are available nationally, or provide financial support in case of implementing specific recommendations.12 National societies of rheumatology would be at this macro-level as well. The meso-level (primary care, regional organisations, patient charities or hospitals) addresses decentralisation, common in many health systems worldwide, and organisational aspects.11 At this level, clinical protocols and pathways may ‘encourage or promote’ specific treatment alternatives over others and decisions on human resources allocation are also made (eg, nurses specialised in RMDs). The micro-level corresponds to the clinicians, healthcare professionals and patients, who will eventually decide, for example, which type of exercises is more appropriate for individual patients with RMD or which joints to examine.

Implementation can be determined through prioritisation and local adaptations. Prioritisation refers to the selection of recommendations to put into practice, usually based on feasibility, potential for impact, patient and population need, etc. The adaptation of recommendations to local needs might be necessary, and how it is implemented may vary in different health systems where there may be different professional roles, access to drugs, etc. A recommendation can propose an intervention, for example, a joint education programme provided by occupational therapists, but in a specific setting, where occupational therapists are not available, this task can be offered by a specialised nurse or physiotherapist.

The implementation team is necessary at the local level and should be multidisciplinary, ideally with guidance from those who developed the recommendations and could vary depending on the recommendations to implement (eg, one may need a politician, another a pharmacist). Besides a team, other resources necessary for implementation can include time, financial support, patient and public involvement and engagement, and digital innovation.

Implementation requires specific knowledge mobilisation skills and training, not only the implementation team but also the clinical guideline developers. A minimum implementation knowledge includes the basis, methodology, and processes of implementation science and the practical application of theory.

Although implementation is better apprehended in its phases (table 1), it is critical to acknowledge that many processes and actions will run in parallel and circles based on immediate feedback from the field; as implementation is an iterative and dynamic process.

A final educational point is the terminology used, which will be new to many. The Effective Practice and Organisation of Care of the Cochrane Collaboration provides terms and definitions.13 Here, for example, ‘continuity of care’ is defined as ‘Interventions to reduce fragmented care and undesirable consequences of fragmented care, for example, by ensuring the responsibility of care is passed from one facility to another so the patient perceives their needs and circumstances are known to the provider’.

Implementation phases

Regarding the phases of implementation (table 1), the implementation of any recommendation starts with an implementation plan. Usually, implementation planning starts upon guideline completion.14 However, implementation is more successful if planning occurs concurrently rather than consecutively to recommendations development, or even before sometimes so that the recommendations issued are clear and usable, target users are primed for adoption, and their needs and preferences are taken into account.15 Implementation plan templates are abundant on the internet, most of which only highlight the actions and actors involved. It is important to determine in this plan which is the recommendation’s implementation objective (eg, to increase uptake of core treatment, to implement exercise in spondyloarthritis or having rheumatologists perform synovial fluid aspiration in patients with undiagnosed inflammatory arthritis).

An analysis of context will afterwards assess the organisational, community and individual readiness for change.16 This analysis should identify the care level/s and their relationships (eg, at what level are specific decisions related to the recommendation taken), the organisational culture and climate (eg, whether the national societies have the power to homogenise behaviours), which teams will be likely involved in the implementation (eg, whether a primary care physician should be included), and which are the human, material, economic and time resources available, including a precise description of the information systems. The latter will be critical to both evaluate and ensure that the recommendation is implemented. The analysis of the context requires accurate knowledge of current clinical practice in the setting.16 For example, in the recommendations dealing with the transition of care from paediatrics to adult rheumatology, the age at which children become adults in the different health systems varies across countries.17

The following phase is the identification of barriers and facilitators. These are factors that hinder or facilitate, totally or partially, the implementation of a change in clinical practice, which are related to health professionals, social (including patients) and organisational context or to the recommendations.18 19 Many techniques can be used to identify them, such as Delphi, nominal groups, qualitative interviews, surveys, communities of practice, etc.20 The Eumusc.net project identified several facilitators and barriers in European rheumatology.21

Next is the design or selection of implementation strategies, that is, the interventions that will facilitate the implementation of recommendations.22 23 Implementation needs to be adjusted for the various target populations and organisations and to offer educational and practical tools. Therefore, strategies include economic, organisational, or regulatory tools, actions and activities focused on clinicians, health professionals and patients. A non-exhaustive list includes leaflets, courses, clinical sessions, local consensus documents, decision rules, checklists, standards of care, electronic medical records or decision-making programmes.22 24–26 However, the efficacy of these strategies is variable.25 26

The evaluation of the implementation is the subsequent step,27 and is not only related to the outcome of the implementation but also the implementation process. Selected recommendations can be transformed into quality measures (ie, indicators and standards of indicators), which are observed before and after the implementation (eg, waiting list, time to access rheumatologist, time to remission).28 There are examples of quality indicators in rheumatology.4 28–30 The whole implementation process can also be evaluated with checklists.

The final phase is the review or replanning. This phase includes taking into consideration the evaluation of the whole implementation process and, if necessary, to redesign or redefine a new implementation plan or even de-implement strategies that do not produce the expected outcome.