Burnout among Dutch intensivists – a nationwide survey

Background Burnout in healthcare workers is deleterious for patients as well as for the healthcare workers themselves, with consequences ranging from less job satisfaction to increased mortality rates. Burnout in intensivists is reportedly high with 50% of intensivists having a burnout in some series. The present study was done to estimate incidence and prevalence of burnout in intensivists in the Netherlands and to identify risk factors for burnout. Methods Two online questionnaires were sent: one to all intensivists in the Netherlands and one to the medical directors of Dutch ICUs. Results A reply was received from 308 out of 664 intensivists (46.4%). Results of 272 intensivists (41%) were evaluable, 12 of whom (4.4%) were diagnosed with burnout at the time of the questionnaire. No association was found between working conditions or personal characteristics and burnout. Intensivists who reported conflict with the hospital management and having complaints filed against them had a significantly higher burnout rate. From the medical directors questionnaire it was found that 7.4% of intensivists suffered from burnout in 2013. The lower incidence and prevalence of burnout in Dutch intensivists compared with foreign series might be explained by a lower workload for intensivists as compared with the literature or by different weighing of the burnout test results by different authors. Conclusion Incidence and prevalence of burnout among Dutch intensivists were found to be unexpectedly low as compared with the literature. The burnout rate was significantly higher in intensivists involved in conflicts or complaints. Introduction Burnout was described by Maslach et al. as a response to chronic emotional and interpersonal stressors on the job.[1] It is characterised by emotional exhaustion, cynicism or depersonalisation, and inefficacy or diminished personal accomplishment.[1] Situational characteristics such as the demands of the job or the organisation as well as personal characteristics and coping styles may predispose to burnout.[2] The current understanding is that burnout is a result of imbalance between job demands and job resources.[3] In the Job DemandsResources model demands have negative consequences such as strain or stress while resources or rewards have positive consequences resulting in motivation and resilience. An imbalance between positive and negative consequences or between resources and demands may result in burnout. Burnout in healthcare workers has negative consequences for the workers themselves as well as for the patients they care for. The Dutch Bureau for Statistics found the burnout rate in the Dutch working population to be 14% (http://www.cbs.nl/nl-NL/ menu/themas/arbeid-sociale-zekerheid/publicaties/artikelen/ archief/2015/cbs-en-tno-een-op-de-zeven-werknemers-heeftburn-outklachten). Healthcare workers with burnout have been found to suffer from post-traumatic stress syndrome, depression, worse sleep quality, worse teamwork quality, less job satisfaction and higher mortality rates.[4] Their patients are at risk of medical errors and higher mortality; there is also a risk that a healthcare worker suffering from a burnout may be less mindful of the needs of families.[4-7] Burnout is also found in Netherlands Journal of Critical Care NETH J CRIT CARE VOLUME 24 NO 1 JANUARY 2016 13 Burnout among Dutch intensivists – a nationwide survey medical students.[8] Burnout in intensivists may be a significant threat to the health of both the intensivists and their patients.[9] The incidence of burnout in intensivists is reportedly high with burnout rates up to 51%, but the incidence and prevalence of burnout in Dutch intensivists is unknown.[10-14] The objective of this study was to estimate prevalence and incidence of burnout in Dutch intensivists and to identify risk factors for burnout. Methods Intensivists questionnaire On 9 September 2013, an email with a direct link to a questionnaire was sent by the Dutch Society for Intensive Care (NVIC) and on behalf of the investigators to all 664 intensivistmembers of the NVIC asking them to participate in the study and fill out the 45-item questionnaire. A reminder was sent on 17 October 2013. The questionnaire was anonymous and contained questions on personal circumstances, working conditions and arrangements for shifts, compensation time and the Utrechtse Burnout Schaal Contactuele beroepen (UBOS-C) questionnaire to diagnose burnout. Because longer questionnaires were expected to illicit less responses, the decision was made not to add extra questionnaires testing respondents for other diagnoses besides burnout (such as engagement or safety climate). The percentage of burnout among the intensivists from this questionnaire was regarded as an estimate of burnout prevalence. For research purposes burnout can be diagnosed with the Maslach Burnout Inventory (MBI).[15] The MBI is used by most authors on the subject. There are different versions of the MBI for different professional groups. The MBI for Human Services Survey (MBI-HSS) is appropriate for use in healthcare.[16] The MBI-HSS contains 22 questions on the three dimensions of burnout, namely emotional exhaustion (nine questions), depersonalisation (five questions) and personal accomplishment (eight questions). The word ‘burnout’ does not appear in any of the questions. Typical questions are formulated like ‘Do you experience......?’ with possible answers ranging from ‘never’ to ‘every day’ on a seven-point Likert scale. Questions are to be scored from 0 (never) to 6 (every day), resulting in separate scores for each of the three dimensions. High scores in emotional exhaustion and depersonalisation and low scores in personal accomplishment are indicative for burnout. Although most authors use the same MBI questionnaire, weighing of the results is not uniform.[17] The MBI manual defines the upper third part of the scoring range as abnormal, but some authors use different cut-off values.[15,18] The MBI manual does not give instructions on how to combine the three scores to diagnose burnout. Some authors combine scores to a single numerical value by adding all scores for emotional exhaustion and depersonalisation and subtracting all scores for personal accomplishment, using a cutoff value of -8, above which burnout is diagnosed. Others define burnout as having abnormal scores in all three dimensions together. The MBI has been validated extensively by comparing results of the test in subjects with or without burnout as established by psychiatrists or psychologists. The UBOS-C is the validated Dutch translation of the MBI-HSS and was used in the present study to define burnout.[16] The UBOS-C contains eight questions on emotional exhaustion, five on depersonalisation and seven on personal accomplishment, totalling 20 questions instead of the original 22 in the MBI-HSS. This is the result of extensive testing and validation, eliminating two redundant questions. Contrary to the original MBI-HSS, cut-off values to determine abnormal scores are not defined as the upper third (as in the MBI-HSS) but as the upper fourth or 75th percentile.[16] Instead of calculating the sum, mean scores for emotional exhaustion, depersonalisation and personal accomplishment are to be calculated. According to the UBOS-C a high score for emotional exhaustion is >2.38, a high score for depersonalisation is >1.80 for men and >1.60 for women and a low score for personal accomplishment is <3.70. Contrary to the MBI-HSS, the way burnout is diagnosed in the UBOS-C is unambiguously defined by a high score in emotional exhaustion (EE) in combination with either a high score in depersonalisation (DP) or a low score in personal accomplishment (PA) or both. To be able to compare our results with the literature we also reported on the three dimensions separately and we calculated the sum of all scores (EE+DP-PA) correcting for the fact that the UBOS-C has eight instead of nine questions on emotional exhaustion and seven instead of eight questions on personal accomplishment. This means that the sum was calculated as follows: (9/8*SumEE) + (SumDP) – (8/7*SumPA). Medical directors questionnaire On 24 March 2014, an email with a link to a questionnaire was sent to all medical directors of Dutch intensive care units asking them to participate in the study by answering questions about their unit and burnout in their staff. The questionnaire contained questions on the characteristics and size of the ICU and the final question was ‘How many of your intensivists were suffering from burnout in 2013 in your opinion?’ No diagnostic criteria or definition of burnout was given. A reminder email was sent on 7 April 2014. The percentage of burnout in the intensivists from this questionnaire was regarded as an estimate of burnout incidence in 2013. Ethics, consent and permission The need for ethical approval was waived by the local ethics committee (Medisch Ethische Toetsings Commissie Zuidwest Holland). Response to the questionnaires was regarded as consent and permission from the respondent to be included in the study. Netherlands Journal of Critical Care 14 NETH J CRIT CARE VOLUME 24 NO 1 JANUARY 2016 Burnout among Dutch intensivists – a nationwide survey Analyses Questionnaires were analysed when no more responses were submitted (about six weeks after the reminders had been sent), by using Microsoft Excel and SPSS (SPSS, Chicago, Ill, USA). Continuous variables were compared using t-tests when normally distributed. We used the median values of the results to dichotomise results whenever possible and necessary to obtain two categories of results which could hence be analysed by the Fisher’s exact test to obtain p-values for comparison. Results Intensivists questionnaire A response to the survey was received from 308 intensivists (46.4% of the 664 registered intensivists in the Netherlands). Twenty-one respondents were no longer practising intensive care and 15 responses were incomplete in such a way that it was impossible to conclude if the respondent was suffering from burnout, leaving responses of 272 individual intensivists (41.0%) to be studied. As judged by the UBOS-C, which requires a high score in emotional exhaustion and either a high score in depersonalisation or a low score in personal accomplishment or both to diagnose burnout, 12 intensivists out of 272 (4.4%) were suffering from burnout at the time of the questionnaire. UBOS-C scores are shown in table 1. When the corrected sum Table 1. Scores of 272 Dutch intensivist on the UBOS-C, the Dutch version of the Maslach burnout inventory. Emotional exhaustion (EE) Very Low (<= 0.37) 24 (9%) Low (0.38-0.99) 68 (25%) Average (1.00-2.37) 151 (56%) High (2.38-3.62) 25 (9%) Very high (>= 3.63) 4 1%) Depersonalization (DP) Very Low (<= 0.19) 30 (11%) Low (0.20-0.59) 101 (37%) Average (0.60-1.59 for women, 0.60-1.79 for men) 120 (44%) High (1.60-2.59 for women, 1.80-2.79 for men) 18 (7%) Very high (>= 2.60 for women, >= 2.80 for men) 3 (1%) Personal accomplishment (PA) Very Low (<= 2.99) 5 (2%) Low (3.00-3.70) 31 (11%) Average (3.71—4.70) 112 (41%) High (4.71-5.56) 107 (40%) Very high (>= 5.57) 17 (6%) Combined diagnosis of burnout (Very) High EE and (very) high DP and (very) low PA 5 (2%) (Very) High EE and (very) high DP 6 (2%) (Very) High EE and (very) low PA 1 (0.3%)