Impacts of Organizational Context on Quality Improvement: don't forget internal management !

To the Editor, In a recent article, Glasgow et al assessed the impacts of organizational context on quality improvement (QI). In their study of organizational characteristics of 100 hospitals participating in a national QI collaborative, they did not find any predictive associations with QI performance. Using a different approach, we conducted a study of 505 French hospitals that are subject to mandatory annual measurements of quality indicators in a national program aimed at promoting continuous quality improvement (CQI). This program is driven by the French Ministry of Health, the French Health Authority, and our research group (COMPAQ-HPST). From an analysis of the literature, we developed a theoretical frame that ranged potential factors of success for a CQI program along 2 axes. One concerned contextual factors (eg, staff stability, QI information), which covered the same organizational characteristics as described by Glasgow et al. The other included managerial factors (eg, managers use quality indicator results to define the hospi-tal's strategy, managers favor multiprofessional commitment when analyzing quality indicator results), which do not seem to have been considered in that article. Seventeen factors were identified and considered to be relevant by a panel of 22 health care professionals, using a consensus method. The relevance of these 17 factors was thereafter confirmed in 505 hospitals by assessing the correlation between the presence/absence of these factors and the results of 3 quality indicators (quality of patient medical records, pain assessment, and delay in sending information at patient discharge). A factorial analysis showed that the managerial axis included 6 factors and the contextual axis included 11. Cronbach α for the 2 axes were .78 and .82, respectively. A multivariate analysis showed a link between managerial score and the results of an indicator assessing the quality of patients' medical records (P = .02) and an indicator measuring the delay in sending information at patients' discharge (P = .03). We believe that managerial capacity is an important factor in the success of QI programs and should be analyzed more precisely.