Marjon Borgert

81 A flowchart for building evidence-based care bundles aim to improve quality of care, the possibility exists that bundled elements have unexpected negative e ects on other care processes. This issue is not well described in the literature but should not be neglected. Therefore, moments for evaluations were incorporated in the bundle design process intended to identify unexpected risks (Fig. 2). Strengths and weaknesses To our knowledge, this is the rst study that reported about the di erent methodologies used in literature to develop new evidence-based care bundles. Our systematic review has several limitations. A description of the bundle development process is not often reported in detail nor described in abstracts. Therefore, we might have missed some relevant articles. We searched for bundles that were developed for ICUs, while methods used in other hospital areas might be relevant and valid as well. However, the rst developed bundles of the IHI were also designed for adult ICUs. Furthermore, the complexity in ICU care is not comparable with other hospital wards. We screened the titles and abstracts of the articles in two steps. During the rst, step one author screened all titles and abstracts. However, predetermined unambiguously clear exclusion criteria were applied. In case there was any uncertainty, the study was included for the second step in this screening process. In the second step the titles and abstract were screened by two authors independently as recommended in the PRISMA-statement. 37 The quality assessment of the articles were conducted by two persons independently. However, the interrater reliability was not calculated. Although the outline of the owchart is based on the IHI approach, the order of the process phases and incorporating the methodologies in each process phase was conducted by opinions of the expert group. However, we have used a validated consensus method to overcome this issue. By combining both IHI and additional methods, we created a owchart on how to develop new evidence- based ICU care bundles. We only searched for studies that described the methods used for bundle development and we incorporated these methods into the owchart (Fig. 2). However, other methods might also be applicable that were not identi ed in our literature search. For instance, in step 1 (Fig. 2) other risk assessment tools might be e ective instead, such as a BowTie analyses 38 or using the analysis from incident reporting systems 39,40 or ’lean management’. 41