Table 3.

Primary and secondary outcomes by group

	Current guidelines (control) (n=133)	User-tested guidelines (n=140)	Multivariate analyses*
Number of simulations with one or more moderate-severe IMG-related errors (% within group)	79 (59)	68 (49)	RR: 0.82†‡ (95% CI 0.66 to 1.02)
Number of simulations with one or more moderate-severe non-IMG-related errors (% within group)	109 (82)	124 (89)	RR: 1.09†§ (95% CI 0.98 to 1.20)
Number of simulations without any IMG-related errors (% within group)	26 (20)	67 (48)	RR: 2.46‡ (95% CI 1.68 to 3.60)
Number of simulations without any non-IMG-related errors (% within group)	11 (8)	6 (4)	RR: 0.45§ (95% CI 0.19 to 1.30)
Median minutes taken to prepare and set up infusion (95% CI)	13.0 (12.2 to 13.8)	11.4 (10.5 to 12.2)	HR: 1.45‡ (95% CI 1.11 to 1.89)
Median modified decisional conflict score (IQR)¶	27 (18–32)	16 (13–18)	MRC: −9.83‡ (95% CI −11.36 to −8.32)

*Multivariate analyses compare outcomes for the user-tested guidelines relative to outcomes for the current guidelines. They were adjusted for participant characteristics: first language (English or not), experience administering intravenous medicines (more or less than 5 years), current use of the IMG (regular or not) and National Health Service (NHS) trust.

†Only NHS trust included as covariate. Other participant characteristics did not improve fit of the model.

‡NHS trust included in model as random effect.

§NHS trust included in model as fixed effect.

¶The modified decisional conflict score (mDCS) can range from 8 to 40, higher numbers represent greater uncertainty.

IMG, Injectable Medicines Guide; MRC, multiple regression coefficient; RR, risk ratio.