|| Checking for direct PDF access through Ovid
To investigate how a terminal illness may affect the health-care providers’ resuscitation preferences.We conducted a cross-sectional survey in 9 health-care institutions located in 4 geographical regions in North and Central America, investigating attitudes toward end-of-life practices in health-care providers. Statistical analysis included descriptive statistics and χ2 test for the presence of associations (P < 0.05 being significant) and Cramer V for the strength of the association. The main outcome measured the correlation between the respondents’ present code status and their preference for cardiopulmonary resuscitation (CPR) in case of terminal illness.A total of 852 surveys were completed. Among the respondents, 21% (n = 180) were physicians, 36.9% (n = 317) were nurses, 10.5% (n = 90) were medical students, and 265 participants were other staff members of the institutions. Most respondents (58.3%; n = 500) desired “definitely full code” (physicians 73.2%; n = 131), only 13.8% of the respondents (physicians 8.33%; n = 15) desired “definitely no code” or “partial support,” and 20.9% of the respondents (n = 179; among physicians 18.4%; n = 33) had never considered their code status. There was an association between current code status and resuscitation preference in case of terminal illness (P < .001), but this association was overall quite weak (Cramer V = 0.180). Subgroup analysis revealed no association between current code status and terminal illness code preference among physicians (P = .290) and nurses (P = .316), whereupon other hospital workers were more consistent (P < .01, Cramer V = .291).Doctors and nurses have different end-of-life preferences than other hospital workers. Their desire to undergo CPR may change when facing a terminal illness.