Quality measurement at intensive care units: which indicators should we use?

OBJECTIVE: This study was conducted to develop a set of indicators that measure the quality of care in intensive care units (ICU) in Dutch hospitals and to evaluate the feasibility of the registration of these indicators.
METHODS: To define potential indicators for measuring quality, 3 steps were made. First, a literature search was carried out to obtain peer-reviewed articles from 2000 to 2005, describing process or structure indicators in intensive care, which are associated with patient outcome. Additional indicators were suggested by a panel of experts. Second, a selection of indicators was made by a panel of experts using a questionnaire and ranking in a consensus procedure. Third, a study was done for 6 months in 18 ICUs to evaluate the feasibility of using the identified quality indicators. Site visits, interviews, and written questionnaires were used to evaluate the use of indicators.
RESULTS: Sixty-two indicators were initially found, either in the literature or suggested by the members of the expert panel. From these, 12 indicators were selected by the expert panel by consensus. After the feasibility study, 11 indicators were eventually selected. "Interclinical transport," referring to a change of hospital, was dropped because of lack of reliability and support for further implementation by the participating hospitals in the study. The following structure indicators were selected: availability of intensivist (hours per day), patient-to-nurse ratio, strategy to prevent medication errors, measurement of patient/family satisfaction. Four process indicators were selected: length of ICU stay, duration of mechanical ventilation, proportion of days with all ICU beds occupied, and proportion of glucose measurement exceeding 8.0 mmol/L or lower than 2.2 mmol/L. The selected outcome indicators were as follows: standardized mortality (APACHE II), incidence of decubitus, number of unplanned extubations. The time for registration varied from less than 30 minutes to more than 1 hour per day to collect the items. Among other factors, this variation in workload was related to the availability of computerized systems to collect the data.
CONCLUSION: In this study, a set of 11 quality indicators for intensive care was defined based on literature research, expert opinion, and testing. The set gives a quick view of the quality of care in individual ICUs. The availability of a computerized data collection system is important for an acceptable workload.