The study sample included 21 patients and their parents out of 29 who were eligible (unwillingness to be filmed was the overriding reason for nonparticipation) and eight surgeons (five men and three women). All surgeons who assessed inguinal hernia cases during the study period participated. Table 1 shows study participants' background and characteristics of the consultation. The average age of children in the study was under four years (M = 3.8, SD = 3.2). The global ratings between coders were significantly correlated and indicated acceptable inter-rater agreement. Spearman correlation coefficients ranged between .61 to .71 for information, .50 to .79 for understanding, .44 to .76 for respect for integrity, and .86 to .88 for degree of participation in decision making among the three raters. These coefficients indicate that agreement was moderate to strong. Further, we noted that raters were no more than one point in disagreement with each other, which indicates acceptable agreement. Figure 1 shows the frequency distributions for each of the ethical components based on the means of the three raters. The shape of the distributions for information, understanding, and respect for integrity were somewhat positively skewed. The distribution for involving parents in decision making showed that many parents did not participate actively.
The occurrence of nonverbal behaviors varied widely between consultations and ranged from 3 to 66 (M = 28, SD = 19). The most frequently occurring behavior was speaking clearly.
Time varied across consultations from as little as one minute to as much as 6 minutes before the physical examination. Examination times ranged between 1 to 4 minutes. Over 52% of the cases were examined in 1 minute or less and only one case took 4 minutes. Post-examination time ranged between 1 to 10 minutes with more than 85% of the cases taking 5 minutes or less. The total time for consultations ranged between 3 to 19 minutes (M = 8.2, SD = 3.4). More than 85% of the consultations lasted 10 minutes or less.
Correlation analyses showed that the occurrence of physician nonverbal behaviors directed towards the parents was related to all ethical components we measured, as seen in Table 2. The length of time spent on the consultation was also positively related to ratings of ethical aspects; however, the time spent post-physical examination was clearly most important.
Total time of the consultation was related to the positive nonverbal behaviors directed towards parents, r = 40, p
We explored whether physicians were "on their best behavior" when medical students were present and found no significant correlations. Child age was not related to the ethical components, but the trend suggests that as age increased ethical ratings were more positive. Consultations in which physicians were women had significantly higher ratings on understanding and decision making than when physicians were men.
In order to discern the amount of independent variance explained by physician initiated nonverbal behaviors and consultation time on the ethical components, we ran a series of multiple regression analyses controlling for potential confounders, i.e. variables that were found to be significantly correlated with the outcome (see Table 2). As seen in Table 3, total consultation time was significantly associated with higher ratings on all aspects of ethical practice we investigated and accounted for a substantial amount of independent variance. Positive nonverbal behaviors initiated by physicians were associated with higher ratings of respect for integrity. Nonverbal behaviors seemed to be marginally related to the other ethical components; however, it may be that the small sample size precluded significance from being reached. None of the confounders (from Table 2) were significant with the exception of uncertainty of diagnosis in relation to decision making. Physicians were more inclined to set the stage for joint decision making when they were uncertain if the child had hernia or not.
The physical examination was visibly stressful for many children. Observers rated 53% of the children as calm, 29% somewhat upset, and 18% as very upset. We analyzed whether positive behaviors by physicians were important for respect for children's integrity using Pearson correlation analyses. We found a strong correlation between positive behaviors during the examination and respect for children's integrity, r = .70, p β =.49), total consultation time (t = 2.63, p β =.35) and child age (t = 2.27, p β =.31) significantly contributed to respect for children's integrity ratings and together explained 68% of the variance.