Publicatie

Publication date

A nation-wide transition in patient safety culture: a multilevel analysis on two cross-sectional surveys.

Verbeek-van Noord, I., Zwijnenberg, N.C., Spreeuwenberg, P., Wagner, C. A nation-wide transition in patient safety culture: a multilevel analysis on two cross-sectional surveys. International Journal for Quality in Health Care: 2019, 31(8), p. 627-632.
Download the PDF
Quality problem or issue
Patient safety is an important topic within healthcare systems. A favourable safety culture might promote safety. We examined whether a nation-wide patient safety programme (PSP) improved patient safety culture.

Initial assessment
We initially measured patient safety culture among 3779 healthcare providers in 45 hospitals in the Netherlands, using the Hospital Survey on Patient Safety Culture.

Choice of solution
A PSP based on two pillars: the implementation of a safety management system and the focus on 10 themes in which harm to patients appeared highly preventable. Elements of the safety management system were safety management, safety culture, risk assessments and continuous safety improvements.

Implementation
Implementation was nation-wide.

Evaluation
After implementation of the programme, we assessed patient safety culture among 6605 healthcare providers in 24 Dutch hospitals and compared the scores with the initial measurement. We hypothesized that after the programme (1) scores on safety culture dimensions improved, (2) safety culture became more homogeneous within and between hospitals and (3) relative influence of hospitals on safety culture increased. A three-level mixed model for continuous responses was fit for 11 safety culture dimensions. We calculated average individual means, between-department variances, between-hospital variances and total variances per dimension.

Lessons learned
In general, a more favourable safety culture was seen after the PSP. However, hospitals and departments did not become more homogeneous, except for ‘frequency of event reporting’. The variety in responses amongst departments and hospitals increased for several dimensions, implying that not all of them improved.