60 Chapter 4 the preventable AEs, the MRAEs, the preventable MRAEs the number of readmissions (RA) and the length of stay in the hospital (the dependent variables in the model). Multilevel analyses were used because the data had a hierarchical structure: patients (level 1) were clustered within hospital departments (level 2), and hospital departments were clustered within hospitals (level 3).18 Adjustments were made for the stratified sample to correct for over-representation of deceased patients and for hospital types.10 RESULTS From the 20 hospitals selected by NIVEL, 17 are measured in the EMRAM study, leaving 3436 patients (Table 2). Seven hospitals had an EMRAM-score 2, nine hospitals an EMRAMscore 5 and one hospital with EMRAM-score 6. As the EMRAM study took place in 20122014 only patients (3436) measured in 2011/2012 of the retrospective patient record review study were used. Table 2 - Number of patients per EMRAM score and per hospital type Type of Hospital Number of Hospitals EMRAM-score Total number of patients EMRAM 2 EMRAM 5 EMRAM 6 University 2 1 1 0 406 Tertiary teaching 7 3 3 1 1423 General 8 3 5 0 1607 Total number of hospitals 17 7 9 1 3436 Total number of patients 3436 1405 1829 202 It concerned 311 patients with at least one adverse event. That is 9%of 3.436 patients. On hospital level, no indication was found that a higher EMRAM score might lead to better patient outcomes (Table 3). The median values of the number of patients with at least one adverse event ranged from 4.4% in EMRAM group 2 to 7.2% in EMRAM group 5. For the three hospitals of which no EMRAM score is known, the median was 10.8%. In EMRAM group 6 only one hospital is present, which may be too small to draw conclusions from.
RkJQdWJsaXNoZXIy MjY0ODMw