Subgroup Analysis of the 230 ICU Nurses

We further examined the responses of the 230 ICU nurses who represented 38% of the 611 ICU nurses in our system. Overall, 21% recalled having nightmares, and 17% had severe anxiety or panic related to experiences while working in the ICU. The ICU nurses who tested positive for symptoms of PTSD had significantly higher scores in all 10 domains on the PTSS-10 when compared with the ICU nurses who tested negative for symptoms of PTSD (p

There were several demographic differences between the ICU nurses with and without symptoms of PTSD. Regarding their work environment, ICU nurses with positive symptoms of PTSD were more likely to work evening or night shifts and were less likely to have taken the role as charge nurse (Table 2). There was no correlation between the overall mortality rate in the specific ICUs and the prevalence of symptoms of PTSD in the nurses who worked in those units. In a multivariable analysis, working on the evening/night shift was the only variable that remained significantly associated with having symptoms of PTSD (OR, 1.47; 95% CI, 1.23-1.71; p = 0.026).

ICU Nurses in the Atlanta Area

To determine whether these results were generalizable to other ICU nurses, questionnaires were mailed to 415 local AACN members in June 2005. A total of 195 (47%) of the questionnaires were returned. Fifty-four of the respondents did not complete the questionnaire because most of them no longer worked in an ICU. The remaining 140 ICU nurses, who had worked in an ICU during the previous 12 months, were included in the final analysis (33.7% of those questionnaires that were originally mailed). The demographics of the nurses who did not complete the questionnaire were not significantly different regarding age, gender, marital status, race, number of children, or whether they provided the primary income for their family.

The general demographics of these ICU nurses are displayed in Table 1. Of the 65 (46%) respondents who included the name of their primary hospital of employment, nurses who work in at least 16 different hospitals are represented in this cohort. Of the 114 (81%) respondents who listed the type of ICU where they primarily worked, a variety of ICUs are represented, including combined medical/surgical (24%), medical ICUs (14%), cardiac ICUs (18%), surgical ICUs (7%), and other (37%).

Overall, 28% of these ICU nurses had nightmares, and 16% had severe anxiety or panic related to their experience working as a nurse. Twenty-nine percent (95% CI, 22-37) of these nurses tested positive for symptoms of PTSD. The median PTSD scores were higher in the metro-Atlanta ICU nurses when compared with the first cohort of ICU nurses (24 vs. 19; p = 0.014); however, the percentage of nurses that were positive for PTSD symptoms was not different between the two groups (29 and 24%; p = 0.22). The distribution of symptoms in these ICU nurses who were positive for symptoms of PTSD was similar to the first cohort of nurses (Figure E2). The PTSS-10 maintained a high internal consistency reliability with a Cronbach's α of 0.90 and 0.87 for each of the two ICU nursing cohorts, respectively. Using the HADS, 20% (28/140) of these ICU nurses had symptoms consistent with possible anxiety, and 26% (36/140) had symptoms consistent with depression. When compared with the ICU nurses who work at our affiliated hospitals, the HADS scores were not different (p = 0.63 and 0.60, respectively).

Subgroup Analysis of Nurses with Positive Symptoms of PTSD

Of the 41 ICU nurses with positive symptoms of PTSD, 24 included contact information on their questionnaire. We were able to contact and formally interview 18 of these nurses. The events that were listed as traumatic episodes related to their work as an ICU nurse are included in Table 3. Seventeen of the 18 (94%) nurses have been having symptoms consistent with PTSD for at least 1 month, and 13 (72%) have had symptoms for more than 3 months.