A realistic evaluation of the impact of a computerised information system on clinical practice : the nurses' perspective.

The aim of this study is to explore nurses' perceptions of the impact in clinical practice of the use of a computerised hospital information system. Design: A realistic evaluation design based on Pawson and Tilley's work has been used across all the phases of the study. It is a theory driven approach and focuses evaluation on the study of what works, for whom and in what circumstances. These relationships are constructed as context-mechanisms-outcomes configurations (CMO). Methods: A self-developed questionnaire containing both closed and open-ended questions was piloted and distributed to all nurses working in in-patient units of the Clinica Universitaria of Navarra (n= 227). Quantitative data were analysed using SPSS 13.0. Descriptive statistics were used for an overall overview of nurses' perception. Inferential analysis, including both bivariate and multivariate methods (path analysis), were used for cross-tabulation of variables searching for CMO relationships. Content analysis of open-ended questions was used to identify major themes in nurses' responses. Results: 179 nurses participated in the study (78.8% response rate). Overall satisfaction with the IT system is positive and only 7.5% of the nurses wanted to go back to paper records. Mechanisms and outcomes are highly correlated. Comparisons with context variables show how users' characteristics, except attitude towards the introduction of technology, did not have a significant influence on perceptions while the nursing unit context had greater influence. Path analysis illustrated that the influence of unit context variables are on outcomes and not on mechanisms. Six main themes emerged from open-ended questions: information, communication, patient care, documentation processes, work dynamics and running of the program. Some differences in relation to the unit context were observed