Marlot Kuiper

274 Connective Routines the medical domain, a focus on top-down implementation of standards proved dominant (Close et al., 2017; Mahmood et al., 2019; Sendlhofer et al., 2015) thereby going by on what is actually happening on the work floor. As fieldwork is very time consuming, I decided to focus on medical specialists as ‘archetypical professionals’ (Etzioni, 1969; Fox, 1992; Freidson, 1970) in particular, as the dynamics between professional and organisational logics are assumed to become most visible here. Through the observations however, I got to see their interactions with other (semi-)professionals like nurse anaesthetists and scrub nurses, and I was able to converse with them and include their perspective in the analysis as well. I purposively sampled professionals that were known as ‘resistant’, and professionals that were as ‘frontrunners’ involved with or took a lead role in developing the Surgical Safety Checklist at their institute. This latter ‘type’ sometimes had a formal role in the ‘checklist project’, and generally had more believe in the checklist. Logically, people are inclined to get involved in things in which they see value. The hybridization of professional roles creates avenues for future research, as it might be worthwhile to more explicitly delve into the hybridization of professional- and organisational roles, and how this affects working with checklists. The limitations of an analytical framework Secondly, I used the initial routine model by Feldman and Pentland (2005) to construct an analytical framework. This frame proved useful to study practices , and helped me to gather and analyse data in a systematic way. It also caused some difficulties. In the complex reality of professional work, the ‘boundaries’ between abstract ideas and behaviour are blurred, and their representation therefore to a certain extent always involves (artificial) categorizations by the researcher. The three empirical chapters each focused on a specific part of the analytical frame (dynamics, interactions, artefacts), but these analytical ‘parts’ showed interrelated in many complex ways. Artefacts already made their entrance in the chapter on routine dynamics, and in order to understand the routine dynamics, findings on routine interactions proved indispensable. Further, the routine model explicitly focuses on the collective level, which makes questions about the relation between the individual and collective level particularly relevant. As discussed in chapter 5, both within and across professional groups, abstract ideas about the checklist proved rather heterogenous. Hence, the idea of ‘the ostensive dimension’ does not do justice to the plurality on this dimension. Especially the finding that actors occupy different hierarchical positions which implies differences in agency, makes questions about the relationship between routines and individual skills and competencies all the more relevant. Work