Likelihood of speaking up in traditional versus professionalism-related patient safety vignettes by level of hierarchy, among all respondents and respondents who perceived high potential for patient harm
| All respondents | Respondents that perceived high potential for patient harm* | |||||
|---|---|---|---|---|---|---|
| Item | Traditional safety vignette % (n/N)† | Professionalism vignette % (n/N)† | p Value‡ | Traditional safety vignette % (n/N) | Professionalism vignette % (n/N) | p Value‡ |
| Would speak up to§ | ||||||
| Nurse | 89% (747/836) | 36% (298/836) | <0.001 | 94% (286/304) | 53% (162/304) | <0.001 |
| Intern | 95% (792/836) | 55% (460/836) | <0.001 | 98% (326/333) | 70% (232/333) | <0.001 |
| Resident | 89% (742/836) | 43% (362/836) | <0.001 | 95% (329/347) | 60% (209/347) | <0.001 |
| Attending | 64% (537/836) | 9% (78/836) | <0.001 | 71% (179/251) | 20% (49/251) | <0.001 |
*Perceived potential for harm to patients was dichotomised very low/low/moderate (low) versus high/very high (high). In the professionalism vignette (ie, speaking up to an inattentive team member), the number of respondents perceiving a high potential for harm (the denominator) varied by the role of the inattentive team member (ie, nurse, intern, resident and attending physician).
†One respondent did not complete all the vignette items (N=836).
‡p Value for McNemar's test comparing within-respondent differences in speaking up in the traditional and professionalism-related safety vignettes.
§For our analyses, we categorised very or completely likely to speak up as ‘speaking up’ and not at all, slightly and moderately likely as ‘reticence’.