The research sample included participants who completed both the pre- and post-training surveys, which totaled 114 internal medicine residents. The sample consisted of 38 first year, 36 second year, and 40 third year residents. The gender composition was 54 females and 60 males, with no participants identifying Not Applicable.
Residents’ Perception of Preparedness for Goals of Care Discussions – Pre-training
Residents’ perceptions of preparedness are based on three survey questions asked before training and descriptive statistics of responses are shown in Table 1.
Table 1
Residents’ Pre-Training Perception of Preparedness for GOC Discussions by Year of Residency and Gender (Means and SD)
Measure
|
Residency Year 1
|
Residency Year 2
|
Residency Year 3
|
Male (N = 18)
|
Female (N = 20)
|
Male (N = 25)
|
Female
(N = 11)
|
Male (N = 17)
|
Female (N = 23)
|
GOC Training Prior to Residency
|
2.72 (.67)
|
3.00 (.86)
|
2.72 (.91)
|
3.36 (.67)
|
3.06 (.83)
|
2.65 (.94)
|
GOC Discussion Practice
|
3.78 (.65)
|
3.15 (.88)
|
3.76 (.60)
|
4.00 (.45)
|
4.18 (.81)
|
4.09 (.42)
|
Recommendations Aligned to Patient Goals
|
3.83 (.79)
|
3.85 (.75)
|
3.60 (.65)
|
3.91 (.70)
|
4.06 (.66)
|
3.57 (.59)
|
The two-way ANOVA tests show the interaction effect between gender and residency year was statistically significant with regards to prior GOC training before residency, F (2, 108) = 3.147 p = .047. There was no significant main effect for year of residency, F (2, 108) = .774 so no further post-hoc comparisons were needed. The main effect for gender, F (1, 108) = .790 p = .376, did not reach statistical significance.
The second two-way ANOVA test shows the interaction effect between gender and residency year was statistically significant for prior practice in GOC discussions, F (2, 108) = 3.861 p = .024. There was a statistically significant main effect for year of residency, F (2, 108) = 10.114, p = < .001. Post-hoc comparison using the Tukey HSD test indicated that the mean score for the first-year residents (M = 3.45, SD = .828) was significantly different (and lower) from second-year residents (M = 3.83, SD = .561) and the third-year residents (M = 4.13, SD = .607). The main effect for gender, F (1, 108) = 1.561 p = .214, did not reach statistical significance.
In the question regarding baseline preparedness to provide a recommendation aligned with patient goals, the interaction effect between gender and residency year was statistically significant, F (2, 108) = 3.115 p = .048. There was no significant main effect for year of residency, F (2, 108) = .870 so no further post-hoc comparisons were needed. The main effect for gender, F (1, 108) = .178 p = .674, did not reach statistical significance.
Figure 1 shows a visual depiction of the mean scores for the three measures of GOC discussions preparedness as perceived by residents prior to the training session.
Overall Perceived Effectiveness of Conducting GOC Discussions
The second research question of the study focused on the perceived effectiveness of GOC discussions, hypothesized to change after the communication training.
Gender Effect. Table 2 shows the descriptive statistics comparing pre- and post-training mean scores by gender.
Table 2. Overall Perceived Effectiveness of Conducting GOC Discussions by Gender (Means and SD)
Variable
|
Male (N=60)
|
Female (N=54)
|
All (N=114)
|
Effectiveness Pre-training
|
3.69 (.56)
|
3.53 (.47)
|
3.56 (.52)
|
Effectiveness Post-training
|
4.02 (.51)
|
3.99 (.57)
|
4.01 (.53)
|
First mixed ANOVA test shows there was no significant interaction between gender and time (training), F(1, 112) = 0.018, p = .90. There was a significant difference between pre- and post-training means, F(1, 112) = 39.303, p = < .001, with both gender groups increasing their perceptions of GOC skill effectiveness after the communication training. The main effect for gender, F(1, 112) = 0.740. p = .480, did not reach statistical significance.
Year of Residency Effect. Table 3 shows the descriptive statistics comparing pre- and post-training mean scores by year of residency.
Table 3. Overall Perceived Effectiveness of Conducting GOC Discussions by Residency Year (Means and SD)
Variable
|
Residency Year 1 (N=38)
|
Residency Year 2 (N=36)
|
Residency Year 3 (N=40)
|
Effectiveness Pre-training
|
3.47 (.52)
|
3.62 (.47)
|
3.60 (.56)
|
Effectiveness Post-training
|
4.02 (.51)
|
3.95 (.51)
|
4.05 (.45)
|
Second mixed ANOVA test shows there was no significant interaction between time and year of residency, F(1, 111) = 0.740, p = .480. There was a significant difference between pre- and post-training means, F(1, 112) = 39.303, p = < .001,with all resident groups increasing their perceptions of GOC effectiveness after the training. The main effect for year, F(1, 111) = 0.436. p = .648, did not reach statistical significance.
Finally, descriptive statistics data show the overall perceptions of GOC discussion effectiveness increase after the communication training for all gender-residency year groups (Figure 2). The increase is particularly noticeable for male residents in year 1 and year 3, and for female residents in year 2. The least change in perceptions after the training is shown by the male year 2 residents.