Usefulness of a systematic exit survey for assessing residents’ satisfaction with postgraduate training: a pilot study

Context: Residents and chief-residents’ satisfaction assessment is important to promote quality training services in hospitals, and to adapt postgraduate training as well as working conditions. Exit interviews are helpful but are time consuming and not easy to manage in a large hospital. The usefulness of exit questionnaire has been demonstrated in large organisations. We wanted to assess the feasibility of implementing a routine exit questionnaire for physicians and the ability of the exit questionnaire to identify fields in need for quality improvement. Methods: From 2010 to 2014, we sent by post a self-reported questionnaire on satisfaction with working conditions, postgraduate training and supervision to residents and chief-residents leaving the hospital. Results: We obtained a response rate of 33%. The sample was representative of the resident population of our hospital considering gender and activity rate. Global satisfaction was good with 75.6% of respondents who agreed or strongly agreed with the proposed statements. The results of the survey were in line with what had been previously found using time-consuming exit interviews. Conclusion: The study demonstrated the feasibility and relevance to measure residents and chief-residents satisfaction using an exit questionnaire. The questionnaire used was easy to apply and provided relevant information for our teaching hospital in the Swiss context. The fields in need for improvement of postgraduate education were the appraisal system and individual career path counseling.


Introduction
In Switzerland, half of the physicians' workforce in hospitals are residents or chief-residents in postgraduate training (Hostettler & Kraft, 2016). They provide an important part of medical services while having specific training needs. To ensure a high quality medical service, hospitals have to offer a high quality medical training as well as an attractive workplace.
The university hospital of Lausanne (CHUV) is a large organisation with 10'000 employees, 18% of which are physicians. Residents (46%) and chief-residents (31%) together represent 77% of the total physician workforce of the hospital. Residents are physicians graduated from medical school who perform clinical work under supervision from the hospital medical staff. Chief-residents are experienced residents who have supervision and management tasks towards residents; they are certified specialists or close to obtaining a specialist title. Heads of services establish the residency program and conduct appraisal interviews. A postgraduate training plan is provided to residents and chief-residents in addition to their employment contract.
There are several reasons for measuring satisfaction with training and working conditions in residents: In Switzerland, a large part of front-line medical service is provided by residents and chief-residents in postgraduate training. More than 40% of residents and chief-residents meet criteria of emotional exhaustion (Biaggi, Peter, & Ulich, 2003) , which is the core dimension of burnout (Maslach & Leiter, 2008;Sonnentag, Kuttler, & Fritz, 2010) . As there is a lack of Swiss physicians, many specialities and hospitals recruit physicians abroad. In addition, today's residents having different priorities than previous generations of physicians about attitudes and life-style, it is necessary to adapt the profession and postgraduate training  . Young physicians (both men and women) value work-life balance more highly than financial or career advancement opportunities than in the past (Smith, Lambert, & Goldacre, 2015) . Support of colleagues and superiors, performance feedback, availability of postgraduate training possibilities, overtime, weekend duties and selected medical speciality are significant factors when deciding leaving clinical practice (Ochsmann, 2012). Therefore, regular assessments of residents' and chief-residents' satisfaction with working conditions and postgraduate training may be an important way to gain information on areas for quality improvement, in order to attract and retain physicians and, ultimately, to prevent them from leaving clinical practice. In addition to other employees' satisfaction surveys, exit interviews can offer precious information on reasons for leaving the workplace, but take a lot of time and need professional interview-trained staff. In comparison to exit interviews, exit questionnaires are easy to administrate to a large population and anonymous, which is important to promote honesty (Giacalone, Knouse, & Montagliani, 1997). They allow to learn about job satisfaction and to detect areas of improvement. Exit surveys are well-known management tools to gather information on job satisfaction, but we could not find published studies on their usefulness for residents in teaching hospitals.
Prior to this study, we conducted exit interviews to gain information on postgraduate training and working conditions in residents and chief-residents (Deriaz S, 2012). A sample of 36 physicians filled in an exit questionnaire and had an exit interview afterwards with a senior physician. The interviews allowed for learning details on specifics themes but were often redundant since -compared to other employees-residents have a high turnover because of mandatory rotations in different hospitals linked with their postgraduate training. As a result, we decided to pursue the use of the exit questionnaire without following interviews as a routine evaluation of physicians leaving the hospital workforce. The objective of this study was, first, to assess the feasibility of implementing a routine exit questionnaire for physicians and, second, to assess the ability of the exit questionnaire to identify fields in need of quality improvement. Favrod C, Koch N, Deriaz S, Wasserfallen J MedEdPublish https://doi.org/10.15694/mep.2016.000142 Page | 3

Material and Methods
From 2010 to 2014, we sent by post a self-reported questionnaire on satisfaction with postgraduate training, supervision as well as working conditions to residents and chief-residents leaving the hospital. Respondents did not have to mention their names and the data base was anonymous. Since the questionnaire was part of a hospital-based quality improvement measure, approval by the local ethics committee was not necessary. This study is in accordance with Helsinki Declaration revised in 2013.
The questionnaire was adapted from an institutional exit questionnaire used by human resource management of the hospital for other professionals, to which we added items focusing on postgraduate training. In addition to sociodemographic data like gender, position (resident or chief-resident) and activity rate, the questionnaire contained 16 items rated on a 5 point Likert-scale: eight questions addressed quality of postgraduate training and supervision and eight questions addressed working conditions. Data analysis: We tested the representativeness of our sample with the total population of the hospital's residents and chief-residents using chi-square analysis. Frequencies analyses were run for socio-demographics and the 16 satisfaction items. When more than 25% of respondents answered "strongly disagree"; "disagree" or "no opinion" to a question, we considered it as a field in need for improvement. We chose this cut-off since customers' satisfaction assessments frequently consider scores between 7 and 8 (over 10) as adequate in identifying fields in need for attention (b2binternational, 2016;HelpScout, 2016). To compare satisfaction between subgroups of physicians (male/female, residents/chief-residents and year leaving the hospital), we recoded the answers into 2 classes: "satisfied" combining answers "agree" and "totally agree", and "not satisfied" combining the answers "strongly disagree"; "disagree" and "no opinion". Chi-square tests were run to analyse differences between these subgroups (gender, position and year of leaving the hospital) on satisfaction items.

Results
We posted 1409 exit questionnaires from 2010 to 2014 and received 465 questionnaires in return (response rate 33%). Thirteen questionnaires were excluded because physicians were not leaving the hospital or were not part of our target population (errors in listing).
Our sample contained 258 women (57.1%) and 194 men (42.9%). The majority were residents (290, 64.2%) and three quarters of them worked full time (343,75.9%). The sample was representative of the working population of residents and chief-residents for gender and activity rate, but not for position, as residents were over-represented in our sample. For more details on sample characteristics see Detailed answers to the 16 questions are displayed in table 2. Globally, 75.6% of respondents agreed or strongly agreed with the statements proposed, with variations depending on the subject of the statement (minimum 60.2%; maximum 95.4%). Concerning postgraduate training and supervision, 3 items obtained more than 25% of nonsatisfaction (disagree, strongly disagree or no opinion): "Structured appraisal interviews occurs periodically", "The head of service actively offers help in achieving the desired career path" and "The clinic offers career prospects". Concerning the items on working conditions, 5 items obtained more than 25% of non-satisfaction: "The workload allocated to the job position is adapted", "The administrative support meets the needs", "The organisation of leaves (holidays, disease, maternity, training ...) is well managed by the clinic", "Achieving work-life balance is possible", and "Working schedules are correct, respected and adapted".
Chi-square analyses between subgroups are reported in tables 3, 4 and 5. The only significant difference between genders (table 3) was that women were more likely satisfied with opportunities for promotion offered than men (65.4% vs. 52.1%, p=0.005). Residents and chief-residents differed about 3 items (table 4): Residents were less likely than chief-residents to be satisfied with the help offered for achieving their career goals (58.0% vs. 69.1%, p=0.042), the workload allocated to their job position (70.7% vs. 84.5%, p=0.005) and the working schedule being correct, respected and adapted (68.0% vs. 82.7%, p=0.003). There was no significant difference between the different years of the study period (table 5).

Discussion
Our results are congruent with previous findings using exit interviews and other studies, and showed that implementing an exit questionnaire was feasible and inexpensive as well as able to identify relevant differences in the various fields of physician satisfaction.
The response rate of 33% was in line with what can be expected when questioning physicians (McAvoy & Kaner, 1996). The sample was representative of the working residents and chief-residents' population regarding gender and working rate. Residents were over-represented compared with the hospital population of physicians because they rotate more frequently between hospitals than chief-residents and are therefore more frequently in situation of leaving the hospital. In our study, more than 25% of respondents were not satisfied with 3 items (out of 8) concerning postgraduate training and supervision. They all concern physicians' appraisal system and career counselling (appraisal interview, help concerning career path and career prospects). Our previous findings using exit interviews also revealed a need for improving the appraisal system and career counselling skills of senior doctors (Deriaz S, 2012).
With regards to working conditions, more than 25% of respondents were not satisfied with 5 items (out of 8) concerning institutional organisation, while residents and chief-residents were satisfied with the 3 items concerning clinical work (premises and equipments, working atmosphere and degree of autonomy). These findings are in line with the exit interviews conducted previously, which had also indicated that organisation within the service was a frequent source of dissatisfaction (Deriaz S, 2012).
The gender comparison showed that women were more easily satisfied with the perspectives of promotion than men. This may seem surprising since women occupy senior positions less frequently than men, but was coherent with studies showing that in first year residents, women are more positive that men about their work (Goldacre, Davidson, & Lambert, 2003) and with the fact that women may be less motivated by career achievements than men .
When comparing residents with chief-residents, the former reported lower satisfaction with working schedules, allocated workload and the management of job absences. This confirmed previous Swiss studies which showed that chronic stress among residents was particularly high, while high satisfaction with career support, sense of coherence and occupational self-efficacy were the most important protective factors (Barbara Buddeberg-Fischer, Stamm, Buddeberg, . Compared to other physicians, residents had lower life satisfaction, higher chronic stress and more depressive symptoms (Klaghofer et al., 2011) .
Finally, analyses did not show any significant difference over time, which was not surprising since no major changes were implemented in physicians' conditions during the study period.
Our study has several limitations: First, our exit questionnaire was not validated formally. Secondly, with regards to methodology, the methods used to dichotomize answers into "satisfied" and "not satisfied" was conservative as we classified the response option "no opinion" as "not-satisfied in order to avoid missing any fields needing improvement. Third, we did not include questions on experience of stress related to working conditions.

Conclusion
Our study showed that it is feasible and inexpensive to measure resident's satisfaction using an exit questionnaire. The questionnaire used was easy to apply and provided relevant information for our teaching hospital in the Swiss context.
While residents and chief-residents were globally satisfied with their working conditions and postgraduate training, the exit survey identified a need for improving the appraisal system and individual career plan counseling. Also, working conditions (workload, work schedules, management of absences) are an important cause of low satisfaction. Systematic exit surveys are a helpful tool to identify fields in need for improvement in postgraduate education and work environment for residents and chief-residents and should be used routinely. Favrod C, Koch N, Deriaz S, Wasserfallen J MedEdPublish https://doi.org/10.15694/mep.2016.000142 Page | 6 Tables   Table 2:

Take Home Messages
Exit surveys of residents and chiefs-residents provide relevant information on satisfaction with postgraduate training and working environment. The survey is easy to use and less time-consuming than exit interviews. It may be a useful tool for continuous quality improvement in physician's postgraduate training and working conditions.