BACKGROUND AND AIMS: Quality monitoring has great relevance in stroke care. The Project "How to guarantee adherence to effective interventions in stroke care" aimed to estimate adherence to acute-phase guidelines in stroke care in Italy. METHODS: A prospective observational study was performed in 27 hospitals of 13 Italian Regions. Adherence to 15 process indicators was evaluated, comparing also stroke units (SU) with conventional wards. An overall score of care, defined as the sum of achieved indicators, was calculated. A multilevel hierarchical model described performance at patient, hospital and regional level. RESULTS: Overall, 484 consecutive stroke patients (mean age, 73.4 years; 52.7 % males) were included. Total score ranged from 2 to 15 (mean 8.5 ± 2.4). SU patients were more often evaluated with the National Institutes of Health Stroke Scale (NIHSS) within 24 h, had more frequently an assessment of pre- and post-stroke disability, and a CT scan the same or the day after admission. Regional-hospital- and patient-level variability explained, respectively, 25, 34, and 41 % of total score variance. In multivariate models, patients >80 years vs. younger showed a change in total score of -0.45 (95 % CI -0.79 to -0.12), and those with NIHSS ≥14 vs. ≤5 of -0.92 (95 % CI -1.53 to -0.30). A negative change means a worse adjusted average adherence to process indicators. SU admission increased total score of 1.55 (95 % CI 0.52-2.58). CONCLUSIONS: Our data confirm the need of quality monitoring in stroke care. Although SU patients showed a better adherence to quality indicators, overall compliance was unsatisfactory.
BACKGROUND AND AIMS: Quality monitoring has great relevance in stroke care. The Project "How to guarantee adherence to effective interventions in stroke care" aimed to estimate adherence to acute-phase guidelines in stroke care in Italy. METHODS: A prospective observational study was performed in 27 hospitals of 13 Italian Regions. Adherence to 15 process indicators was evaluated, comparing also stroke units (SU) with conventional wards. An overall score of care, defined as the sum of achieved indicators, was calculated. A multilevel hierarchical model described performance at patient, hospital and regional level. RESULTS: Overall, 484 consecutive strokepatients (mean age, 73.4 years; 52.7 % males) were included. Total score ranged from 2 to 15 (mean 8.5 ± 2.4). SU patients were more often evaluated with the National Institutes of Health Stroke Scale (NIHSS) within 24 h, had more frequently an assessment of pre- and post-stroke disability, and a CT scan the same or the day after admission. Regional-hospital- and patient-level variability explained, respectively, 25, 34, and 41 % of total score variance. In multivariate models, patients >80 years vs. younger showed a change in total score of -0.45 (95 % CI -0.79 to -0.12), and those with NIHSS ≥14 vs. ≤5 of -0.92 (95 % CI -1.53 to -0.30). A negative change means a worse adjusted average adherence to process indicators. SU admission increased total score of 1.55 (95 % CI 0.52-2.58). CONCLUSIONS: Our data confirm the need of quality monitoring in stroke care. Although SU patients showed a better adherence to quality indicators, overall compliance was unsatisfactory.