The ethical climate in paediatric oncology—A national cross‐sectional survey of health‐care personnel

Abstract Objective To describe health‐care personnel's (HCP's) perceptions of the ethical climate at their workplace in paediatric oncology. Methods A cross‐sectional survey was conducted using the Swedish version of the shortened Hospital Ethical Climate Survey (HECS‐S). HCP at all six paediatric oncology centres (POCs) in Sweden were invited to participate. Analysis included descriptive statistics, the Mann‐Whitney U test (differences between groups) and Spearman's rank correlation. Informed consent was assumed when the respondents returned the survey. Results A high response rate was achieved as 278 HCP answered the questionnaire. Medical doctors perceived the ethical climate to be more positive than registered nurses and nursing assistants. At the POC with the significantly lowest values concerning immediate manager, no significant correlation with the other items was found. At the POC with the poorest ethical climate, HCP also had the lowest perception of the possibility of practicing ethically good care. Conclusions Differences between centres and professional groups have been demonstrated. A negative perception of the immediate manager does not necessarily mean that the ethical climate is poor, but the manager's ability to provide the conditions for an open dialogue within the health‐care team is key to achieving an ethical climate.


| Instrument
The paper survey included demographic questions, as well as the abovementioned Swedish HECS-S, an instrument for assessing HCP's perceptions of the ethical climate. The Swedish HECS-S includes all 14 items from the original HECS-S as well as items added to accomplish a multi-professional instrument relevant for paediatrics (Table 1). 16 Respondents were asked to state how often they perceived the statements to be consistent with the situation at their workplace on a fivepoint Liker-type scale with the labels "Almost never-Almost always."

| Data collection
Data collection was conducted during meetings/training sessions arranged by each of the centres. The local coordinator subsequently invited HCP who were unable to attend these sessions to answer the survey and sent those by mail. Data collection was conducted from February to September 2016.

| Data analysis
All statistical analyses were conducted using the Statistical Package for Social Sciences version 25.0. Descriptive statistics (frequencies, mean values, and SD) were calculated, and differences in distribution between groups were tested using the Mann-Whitney U test. Correlations between per person mean values of (groups of) items were tested using Spearman's rank correlation. P values less than 0.05 were considered statistically significant. Up to 10% missing items were considered acceptable. 18

| Ethical considerations
The research team provided oral information at the sessions and written information together with the survey. The information included the aim of the study and when the respondents returned the survey informed consent was assumed. The total number of respondents per centre will not be disclosed, and the results are reported in a way that protects the integrity of the centres. In its advisory statement, the Regional Ethical Review Board in Stockholm had no ethical objections (D-no: 2015/1782-31/5).

| Respondents
HCP (n = 278) from six POCs (henceforth referred to with random letters) answered the questionnaire, response rate 89%. The mean number of respondents from each of the POCs was 47, range 29 to 74.
The demographic characteristics of respondents is presented in Table 2.
The overall highest and lowest scored items and differences between groups are presented below.

| Overall highest and lowest scored items in the HECS-S
The five items with the highest overall values included three items concerning team interactions and the two items concerning patients/parents. The five items with the lowest overall values were the two hospital items, two items on team interactions and, finally, the item on MDs asking RNs for their opinions regarding treatment (Table 3).

| Differences between professional groups
MDs scored significantly higher than RNs and NAs on 10 items (

| Gender differences
No gender differences were identified except for two items. Male respondents scored significantly (P = 0.041) higher (mean 3.63) than female (mean 3.30) on the item on physicians asking nurses for their opinions. This difference was also present in the group of MDs in which male respondents scored significantly (P = 0.002) higher (mean 3.97) than female respondents (mean 3.23). Female respondents scored significantly (P = 0.034) higher (mean 3.03) than male (mean 2.67) on the item concerning hospital guidelines. However, there were no significant differences between genders in the professional groups on this item.

| Differences between years of experience
In three items, differences were identified between the groups with different levels of experience of paediatrics. Respondents with less than 5 years of experience scored significantly (P = 0.010) higher (mean 3.21) on the item concerning being helped by hospital guidelines than those with more experience (mean 2.86). This difference was also significant (P ≤0.001) when looking at the group of NAs but not at the group of RNs.
The item concerning giving attention to ethical problems scored significantly (P = 0.049) higher for those with 5 years of experience or more (mean 3.97) than those with less experience (mean 3.73).
Likewise, the item concerning talking about different ways of dealing with ethical issues scored significantly (P = 0.017) higher for those with 5 years of experience or more (mean 3.78) than those with less experience (mean 3.46). However, the difference in these two items was not significant in the groups of NAs and RNs. However, it had significantly higher scores on four items concerning the relationship between MD/RN (n = 2), team interactions (n = 1), and patient wishes (n = 1) ( Table 4).

| Differences between POCs
Another centre (B) scored significantly higher on the three items regarding the immediate manager and the item about conflicts being dealt with openly. In addition to these four items, this centre (B) had significantly higher scores on one item and no significantly lower values (Table 4).
A third centre (C) had significantly lower scores than the other POCs on 10 items, including items concerning the relationship between HCP (n = 5), team interactions (n = 1), patient wishes (n = 1), the possibility of practicing care as they think it should be practiced (n = 1), and identifying and dealing with ethical issues (n = 2). This centre did not have significantly higher values on any of the items (Table 4) and had a significantly (P ≤ 0.001) lower total score (mean 3.61) than the other centres.

| DISCUSSION
The In this study the item on MDs asking RNs for their opinions regarding treatment was one of the items that had the lowest overall value. This item had been removed from the HECS-S but was reintroduced in the Swedish HECS-S. 16  perspectives emerged as an ethical concern, and nurses felt that they could not influence medical decisions. 2 MDs stand out in this study as the group with higher scores compared with the other two groups as they scored significantly higher values than RNs and/or NAs on 12 out of the 21 items.
Significant differences between NAs and RNs were only identified in  3 Mean value and SD on all items for the whole group and for each of the professional groups, and P values when significant differences were found between groups with colleagues and the multidisciplinary team and rating this collaboration as being more favourable than RNs. 12 Moreover, support from colleagues has been identified as a main contributory factor for MDs' resilience. 19 This could also explain why, in a Swedish paediatric oncology study, MDs had a significantly lower total moral distress score compared with RNs. 20 The only significant difference identified between years of experience was that NAs with less than 5 years of experience scored higher on the item on "guidelines help me" than NAs with more experience.

Mean (SD) Mean (SD) Mean (SD) Mean (SD) P Value
Initially, the results indicated a difference between years of experience on the items concerning identifying and dealing with ethical issues.
However, this difference rather related to the significantly higher scores on these items amongst MDs, of whom almost all had more than 5 years of experience. These high scores could be explained by the MDs being "forced" to deal with ethical issues because studies have shown that they assume a considerable amount of responsibility 3 and also feel great uncertainty in difficult decisions. 2 In the present study, one of the centres (A) had significantly lower values on the three items regarding the immediate manager, with no significant correlation to the other items. This indicates that a negative perception of the immediate manager does not necessarily mean that the ethical climate is poor. Furthermore, another centre (C) had supported routines that enhanced collaboration and dialogue between coworkers. As previously suggested, management that supports ethical behaviour generates positive organisational outcomes. 5 The centre (C) that had significantly lower scores on 10 items also had the lowest overall score. This could relate to the fact that many items in the HECS-S concern the relationship between the different professions and team interaction. However, we would argue that the fact that this centre (C) also had the lowest values regarding the possibility of practicing care as it should be practiced could be seen as an indication that the ethical climate at this centre will negatively influence patient care. Also, in a previous study, a positive perception of interprofessional trust was related to the possibility of practicing ethically good care. 13 Obvious strengths of this study are the nationwide and multiprofessional coverage, as well as the high response rate. The latter limits the risk of nonresponse bias and increases the generalisability, not only to international paediatric oncology but also to other highly specialised paediatric settings.

| CONCLUSIONS
Because paediatric oncology entails difficult ethical issues, an ethical climate is crucial for preventing moral distress and staff turnover, as

| Study limitations
A potential limitation of the HECS-S is that the items concerning the immediate manager do not fully capture this aspects of the ethical climate. It is reasonable to assume that rather than discussing patientcare issues, the role of the immediate manager is to provide opportunities for HCP to deliberate on ethical issues and collaborate interprofessionally. Furthermore, many statistical tests were performed, increasing the risk of mass significance.

| Clinical implications
The understanding of the hospital ethical climate from a multidisciplinary perspective in paediatric cancer care could facilitate the formulation of plans for organisational improvments. The knowledge from this study motivates actions to promote a good ethical climate by supporting interprofessional collaboration and providing ethics support in identifying and dealing with ethical issues, especially for RNs and NAs. Furthermore, it is important to deal with the issues in the interprofessional collaboration due to the descrepancies in perceptions on MDs asking RNs for their opinions on treatment.