Table 2 Implications of Realist Assumptions for Evaluation of Lean Implementation in Saskatchewan

Social programs are an attempt to create some kind of change.

The implementation of Lean aims to improve the quality and efficiency of the Saskatchewan health system.

Programs ‘work’ by enabling participants to make different choices, understanding that choice making is constrained by participants’ previous experiences, beliefs and attitudes, opportunities and access to resources.

Different types of decision makers will influence Lean outcomes, from those in senior positions at the central level through unit administrators, team leaders, practitioners and patients. A wide variety of factors will affect the decisions that they make.

The context in which a program operates will make a difference to the outcomes the program achieves. The context features aspects such as social, economic and political structures, organizational context, program participants, program staffing, geographical and historical context, and so on.

Context can influence program mechanisms and outcomes in many different ways.

Organizational culture varies across regions and across types of health units. Professions and occupations have different norms. Culture, gender and socialization shape patterns of decision-making. Organization priorities may influence the ways in which, or the extent to which, particular Lean approaches are implemented, who it targets, who it reaches and so on.

Access to resources to implement decisions, and opportunities to implement decisions, can also influence reasoning itself, as well as whether or not desired choices can be put into action.

Because there is always an interaction between context and mechanism, that interaction is what determines the program’s impacts or outcomes: Context + Mechanism = Outcome.

Testing the ‘CMO hypotheses’ requires data about each element of the hypothesis; the context, mechanism, and outcome. It also requires analytic techniques that can identify the relationships between them.

Because programs work differently in different contexts and through different change mechanisms, programs cannot simply be replicated from one context to another and be expected to automatically achieve the same outcomes.

At the macro level, it cannot be assumed that Lean will work in the same ways, or to the same extents, in different kinds of health units or different regions.

At the micro level, it cannot be assumed that a solution generated in one setting will necessarily work in another setting.

Good understandings about ‘what works for whom, in what contexts, and how’ are, however, portable.

A realist evaluation should generate a deep understanding of how and why Lean, or Lean tools, works well in some contexts and less well in others. This may assist policy makers and administrators to adapt to their own contexts, and thus to improve outcomes. This is entirely consistent with the ‘local solutions’ principle of Lean itself.