Table 2.
Epochs (epoch2:epoch1) | Day of admission (weekend:weekday) | Weekend:weekday | ||||
Between epochs | ||||||
RR (OR)* (confidence limits) | P value | RR (OR)* (confidence limits) | P value | RR (OR)* (confidence limits) | P value | |
Errors | ||||||
Assessment, investigation or diagnosis | 0.71 (0.61 to 0.83) | 0.93 (0.79 to 1.10) | 1.14 (0.77 to 1.69) | |||
Treatment and management | 0.74 (0.63 to 0.87) | 0.97 (0.83 to 1.13) | 1.18 (0.80 to 1.73) | |||
Communication | 0.84 (0.66 to 1.07) | 1.08 (0.87 to 1.34) | 1.44 (0.93 to 2.23) | |||
Medication | 0.65 (0.52 to 0.81) | 0.92 (0.72 to 1.19) | 0.86 (0.50 to 1.47) | |||
Monitoring | 0.78 (0.55 to 1.11) | 0.94 (0.70 to 1.27) | 0.89 (0.36 to 2.21) | |||
Resuscitation | 0.82 (0.37 to 1.81) | 2.61 (1.15 to 5.91) | 1.14 (0.16 to 8.03) | |||
Infection | 2.35 (0.71 to 7.76) | 0.73 (0.24 to 2.26) | 1.23 (0.13 to 11.5) | |||
Invasive procedures | 0.46 (0.20 to 1.04) | 1.63 (0.70 to 3.77) | 1.15 (0.21 to 6.27) | |||
Other | 0.52 (0.30 to 0.92) | 1.21 (0.78 to 1.86) | 1.20 (0.44 to 3.29) | |||
All errors | 0.78 (0.70 to 0.87) | <0.0001 | 0.96 (0.86 to 1.07) | 0.4922 | 1.14 (0.91 to 1.45) | 0.2566 |
Adverse events | 0.48 (0.33 to 0.69) | 0.0001 | 0.89 (0.57 to 1.38) | 0.5991 | 1.29 (0.54 to 3.08) | 0.5663 |
Global quality of care | 0.78 (0.70 to 0.87) | <0.0001 | 0.98 (0.86 to 1.10) | 0.6904 | 1.03 (0.81 to 1.31) | 0.7973 |
Process indicators | ||||||
Location not appropriate | 0.91 (0.64 to 1.30) | 0.6170 | 1.00 (0.71 to 1.42) | 0.9961 | 1.16 (0.54 to 2.46) | 0.7051 |
Incomplete vital signs | 1.25 (1.07 to 1.47) | 0.0056 | 0.80 (0.71 to 0.91) | 0.0009 | 0.99 (0.74 to 1.32) | 0.9267 |
NEWS not recorded | 0.75 (0.65 to 0.88) | 0.0003 | 0.81 (0.71 to 0.93) | 0.0026 | 1.06 (0.79 to 1.43) | 0.6805 |
Specialist review <14 hours not documented | 0.82 (0.72 to 0.93) | 0.0026 | 0.95 (0.81 to 1.11) | 0.5142 | 0.86 (0.68 to 1.09) | 0.2194 |
Adjusted for age, sex, ethnicity and hospital trust.
*Rate ratios for errors (from mixed effects negative binomial models); ORs for adverse events and process indicators (from mixed effects binary logistic models); proportional ORs for global quality of care (from mixed effects ordinal logistic regression).
OR, odds ratio; RR, rate ratio.