Mental_Heatlh_of_a_Team

Soc Psychiatry Psychiatr Epidemiol (2009) 44:541–549 DOI 10.1007/s00127-008-0480-4 ORIGINAL PAPER Tanya Nelson Æ Sonia Johnson Æ Paul Bebbington Satisfaction and burnout among staff of crisis resolution, assertive outreach and community mental health teams A multicentre cross sectional survey Received: 12 June 2008 / Accepted: 14 November 2008 / Published online: 12 December 2008 j Abstract Background The NHS Plan required extensive changes in the configuration of mental health services in the UK, including introduction of crisis resolution teams, CRTs. Little is known about the effects of these changes on mental health staff and their recruitment and retention. Aims To assess levels of burnout and sources of satisfaction and stress in CRT staff and compare them with assertive outreach team (AOT) and community mental health team (CMHT) staff. Method Cross sectional survey using questionnaires, including the Maslach Burnout Inventory, the Minnesota Satisfaction Scale and global job satisfaction item from the Job Diagnostic Survey. All staff in 11 CRTs in 7 London boroughs were included. Results One hundred and sixty-nine questionnaires were received (response rate 78%). CRT staff were moderately satisfied with their jobs and scores for the three components of burnout were low or average. Their sense of personal accomplishment was greater than in the other types of team. Conclusion Our results suggest that CRTs may be sustainable from a workforce morale perspective, but longer term effects will need to be assessed. j Key words crisis resolution – staff satisfaction – burnout – cross-sectional survey Introduction Implementation of the NHS Plan [18], involving a move from institutional care to community care, has required extensive changes throughout the UK mental health system. Many staff have moved into new forms of specialist team, including crisis resolution teams (CRTs), assertive outreach teams (AOTs) and Early Intervention Services. Specialist CRTs are dedicated to providing short term intensive home treatment for people presenting in acute psychiatric emergencies, and have been shown to decrease overall acute adult inpatient admissions by 21%, reduce occupied bed days by 11% [17] and possibly increase user satisfaction [10, 11]. AOTs provide intensive treatment and support for service users in the community with a history of disengagement from mainstream mental health services, one of their aims being to prevent repeated hospital admissions [14]. National service mapping data indicate that 262 CRTs and 261 AOTs were operating in England in March 2005 [8]. A crucial issue is how these changes affect the morale of the mental health workforce: unless these new types of team can recruit and retain high quality and committed staff they are unlikely to be maintain these models of care and replicate the good outcomes produced by experimental teams [1, 3, 6, 16, 23] and improve upon results achieved. CRTs and AOTs require very intensive community work with disturbed patients, usually with shift working, and there have been concerns that this may lead to low morale and burn out among staff. On the other hand, staff may find short term working with an innovative team approach satisfying. ‘Burnout’ is a term coined by Freudenburger [7] to describe demoralisation, disillusionment and Harvey Whiteford, served as a guest editor for the manuscript and was responsible for all editorial decisions, including the selection of reviewers. The policy applies to all manuscript with authors from the editor’s institution. T. Nelson, MBBS, MRCPsych, MSc (&) Crisis Resolution Team Devon Partnership NHS Trust Wonford House Hospital Dryden Rd Exeter (Devon) EX2 5AF, UK E-Mail: tanya.nelson@nhs.net S. Johnson, BA, BM, BCh, MSc, MRCPsych, DM Æ Prof. P. Bebbington, MA, MPhil, PhD, FRCP, FRCPsych Dept. of Mental Health Sciences University College London London, UK SPPE 480 542 exhaustion he observed in human service workers. It is conceptualised as a process that occurs over time, rather than a state. Over the last quarter century, the job strain model [13] has become the predominant theory of work stress. This identifies job characteristics as the principal source of psychological strain in the workplace, proposing that this results from the combination of high job demands and low job control (also referred to as decision latitude). Thus workers experience distress when there is a mismatch between the challenges of their job and their ability to formulate and implement effective coping responses. The model has been extended to include lack of support as a third source of job strain: this has become known as the Demand-Control-Support model. Many of the studies that have tested this model in its various forms support it [27], particularly in its additive version which suggests that high demands, low control and low social support can each result in psychological strain, but that strain is greatest when all are present. In investigations with nurses, high control has been found to be particularly associated with good job satisfaction and motivation, while high job demands and low social support have been found to be especially associated with health complaints and absenteeism [26]. Most of the factors thus far identified in research on mental health staff as potentially important influences on morale fit into this well-established model. The studies of the CMHT workers that took place in the nineties [4, 5, 20, 22, 29] found high levels of emotional exhaustion and burnout. Commonly reported sources of stress were to do with fears over future job security, not having enough facilities in the community to refer on to and long waiting lists to access these services. This may be explained by the changes in configuration and delivery of mental health services at that time, which involved uncertainty about the new role required by mental health professionals, the sustainability of working in this way and the lack of provision of adequate services in the community. These concerns do not seem to feature so highly in more recent studies [2, 21] which have tended to show slightly lower levels of burnout in the community than in the early studies, though still remaining in the high emotional exhaustion category for some groups. Patient centred characteristics were cited as important sources of stress, such as ‘‘having to work with suicidal patients on my own’’, patients with a known history of violence, dual diagnosis and complex clinical and social problems. There are no published data regarding the morale of staff working in CRTs, and a Cochrane Review suggested that staff satisfaction and burnout are key indicators relevant to the sustainability of these teams, and must be considered in future research [12]. The aim of the current study was to address this important gap in the evidence underpinning policy and service planning. Our primary research hypothesis was that crisis team workers would have high scores on the emotional exhaustion component of the Maslach burnout inventory owing to the demands of mainly working with acutely ill patients, but that job satisfaction and personal accomplishment would be good owing to good levels of support within the team and clearly defined objectives. The secondary hypothesis was that staff satisfaction would be greater in crisis team workers than in those people working in community mental health teams or assertive outreach teams. Method Multicentre research ethics committee approval was obtained for this descriptive cross-sectional study. j Sample Managers of eleven CRTs in 7 London boroughs (Camden, Islington, Southwark, Lambeth, Waltham Forest, Kensington and Chelsea) agreed that their staff might be approached and their consent sought for participation. These were all the established CRTs in London when the study started in 2003. No CRTs that were approached declined to take part. Comparison data on Assertive Outreach and CMHT staff were drawn from the Pan-London assertive outreach (PLAO) Study, a collaboration between the Sainsbury Centre for Mental Health and the departments of psychiatry at Royal Free and University College, Queen Mary and Westfield, and St George’s Hospital Medical Schools [2]. In the course of the PLAO study, data were collected in 2001 on staff in all London assertive outreach teams and 8 randomly selected community mental health teams using a set of measures which overlap extensively with those used in the present study. j Measures Staff were asked to complete a questionnaire which included: • Background information consisting of brief demographic and professional details, including age, sex, profession, level of seniority, hours of work, shift work, age and site of team. • The Maslach Burn-out Inventory MBI [15]. This defines burnout operationally through the following three dimensions: emotional exhaustion—the depletion of emotional resources as a direct result of the excessive psychological and emotional demands of working intensively with people: depersonalisation—the tendency for people to develop negative and cynical attitudes and feelings towards others: lack of personal accomplishment—negative self-evaluation, particularly with regard to one’s work with clients. The MBI categorised each dimension of burnout into low, medium and high levels, producing standardised norms for various professions. According to their norms for mental health staff, high burnout is characterised by a score of 21 or more in emotional exhaustion, 8 or more on depersonalisation and 28 or less on personal accomplishment. Average burnout is indicated by 14–20 emotional exhaustion, 5–7 on depersonalisation and 33–29 on personal accomplishment. Low burnout is indicated by scores of 13 or less on emotional exhaustion, 4 or less on depersonalisation and 34 or more on personal accomplishment. Total scores are calculated for each subscale. The MBI has been widely used and well validated [24]. • The Minnesota Satisfaction Scale, MSS [28] .This consists of 18 items of which 12 items can be ascribed to determine intrinsic and 6 can be ascribed to determine extrinsic job satisfaction. Intrinsic subscale items are such as ‘the chance to make use of 543 my abilities’, whereas extrinsic items include factors like satisfaction with pay, company policy’s, supervision and chances for advancement. The different subscale scores may be added up to a summary score. • The general job satisfaction item from the Job Diagnostic Survey [9]. This consists of 5 items, each measured on a 7 point Likert scale. A global score can be computed from the 5 items in this sub-scale, a higher score representing greater general job satisfaction. Details of the psychometric properties of the scale are explored by Taber and Taber [25]. • The PLAO questionnaire on sources of stress and satisfaction was also administered [2], augmented by some items specifically focusing on CRT work. These included free text spaces for more detailed answers regarding sources of stress and satisfaction, questions regarding leadership, whether members of staff wished to continue working in CRTs in the long and short term, reasons to leave or stay, future plans and suggestions to improve the working environment. between the three types of team so as to be able to identify differences in staff characteristics that might account for any differences found in burnout and satisfaction. Compared with CMHTs, CRT staff were younger (v2 = 4.1, P = 0.0006), and had worked for a shorter time in their current team (adjusted mean difference 1.67 years, P = 0.002), current type of team (adjusted mean difference 3.35 years, P < 0.0005) and in mental health services (adjusted mean difference 2.75 years P = 0.03). Mix of disciplines differed significantly between the 3 team types, with CRTs having the highest proportion of nurses and psychiatrists. The CRTs had been in existence at the time of the study for a mean of 3.5 years (minimum 8 months, maximum 5.6 years). j Data collection The principal investigator visited each CRT to explain the nature and purpose of the study and distribute questionnaires together with stamped addressed envelopes for return. In order to match results, codes were used to identify individuals. Questionnaires were left for any staff not in attendance at the team meeting. To ensure a maximal response rate, staff were reminded by telephone and a further visit to the team. j Satisfaction and burnout Tables 2 and 3 show the mean job satisfaction and burnout scores comparing CRTs with AOTs, and then with CMHTs. Table 2 shows that both CRT and AOT staff were similarly moderately satisfied on the JDS global job satisfaction item.1 CRT and AOT staff scored very similarly on the Minnesota Satisfaction Scale.2 The MBI3 shows that the CRT and AOT mean score for emotional exhaustion were each in the average burnout category. Regarding depersonalisation, both CRT and AOT staff scored in the low burnout category, the AOT score being lower with the adjusted mean difference of )1.23 just reaching statistical significance (P = 0.05). Both CRT and AOT staff scored in the low burnout category on the personal accomplishment dimension, but CRT staff experienced greater sense of personal accomplishment than AOT staff. The adjusted mean difference was 3.4 points, which was highly significant (P = 0.003). Table 3 shows that CRT staff appeared to experience greater job satisfaction than CMHT staff, but this did not remain significant after adjusting for potential confounders. CRT and CMHT staff appeared to 1 j Statistical analysis We used SPSS for Windows Release 12.1 and STATA (release 8.0; Stata Corporation, 2003) for data analysis. Results for staff working in the same team are not independent, so all significance tests and confidence intervals were computed using robust standard errors that allow for clustering by team. The survey estimation commands in STATA allow the computation of means, confidence intervals and v2 tests adjusted for clustering, and were used for descriptive statistics and tests of association between categorical variables. The principal analyses comprised the following stages. (a) The socio-demographic and occupational characteristics of crisis team staff were described and comparisons made with AOT and CMHT staff. (b) Comparisons were made between scores for CRT staff and AOT and CMHT staff for each of the main satisfaction and burnout indicators. To identify confounders and explore factors associated with burnout and job satisfaction, the main sociodemographic and occupational variables were added as additional independent variables in regression analyses, with burnout and satisfaction scores as the dependent variables. Results The response rate from the CRTs was 78% (132 of a total of 169 staff). In the PLAO study responses were obtained from 89% of AOT staff (187 of a total of 211 staff), while the lowest response rate came from CMHT staff, at 75% (114 of a total of 152 staff). j Demographic characteristics of staff Table 1 shows the demographic characteristics and job details for the respondents from the CRTs. We made exploratory comparisons on all these variables For the Job Diagnostic Survey, a mean score is calculated with 1 indicating severe dissatisfaction, 7 indicating a very high level of satisfaction and 4 indicating neither overall satisfaction nor dissatisfaction with work. 2 For the Minnesota score, a neutral attitude is indicated by scores of 60 for overall satisfaction, 18 for extrinsic satisfaction and 36 for intrinsic satisfaction. 3 The Maslach Burnout Inventory, MBI, categorised each dimension of burnout into low, medium and high levels, producing standardised norms for various professions. According to their norms for mental health staff, high burnout is characterised by a score of 21 or more in emotional exhaustion, 8 or more on depersonalisation and 28 or less on personal accomplishment. Average burnout is indicated by 14–20 emotional exhaustion, 5–7 on depersonalisation and 33–29 on personal accomplishment. Low burnout is indicated by scores of 13 or less on emotional exhaustion, 4 or less on depersonalisation and 34 or more on personal accomplishment. 544 Table 1 Shows the demographic characteristics and job details for the respondents of all three types of team Characteristic Crisis resolution team staff (n = 132) n (%) Assertive outreach team staff (n = 187) n (%) Community mental health team staff (n = 114) n (%) 45 (39.5) 69 (60.5) 4 30 38 23 17 76 5 14 11 4 4 53 31 5 12 8 4 1 (3.6) (26.8) (33.9) (20.5) (15.2) (66.7) (4.4) (12.3) (9.6) (3.5) (3.5) (46.5) (27.2) (4.4 0 (10.5) (7.0) (3.5) (0.9) Significance testa Gender (%) Male Female Age (%) 18–25 26–35 36–45 46–55 55+ Ethnicity (%) White Mixed race Asian Black Chinese Other Occupation (%) Nursing Social work OT Psychiatry Psychology Supp work Other Experience (mean, CI) Mean no. of years in current teamb Mean no. of years in current type of team Mean no. of years worked in mental health service Position in team (%) Manager/deputy/consultant Other mental health worker Pattern of work (%) Within office hours (Monday–Friday, 9–5) only Shifts, inc. some evenings and/or weekends but no overnight work Shifts inc. some evenings and weekends; on call overnight by telephone only Shifts inc. evenings, weekends and overnight Other a 66 (50) 66 (50) 6 50 49 20 7 96 4 11 17 0 3 78 19 1 21 1 11 1 (4.5) (37.9) (37.1) (15.1) (5.3) (73.3) (3.1) (8.4) (13.0) (2.3) (59.1) (14.3) (0.8) (16.0) (0.8) (8.3) (0.8) 94 (50.3) 93 (49.7) 1 80 74 30 2 116 7 14 42 3 4 65 32 17 15 5 32 21 (0.5) (42.8) (39.6) (16.0) (1.1) (62.4) (3.8) (7.5) (22.6) (1.6) (2.1) (34.8) (17.1) (9.1) (8.0) (2.7) (17.0) (11.2) v2 = 2.9 P = 0.06 v2 = 4.1 P = 0.0006 v2 = 1.7 P = 0.09 v2 = 6.5 P = 0.0005 1.6 (1.3–2.0) 2.0 (1.5–2.5) 10.9 (10.0–11.8) 2.0 (1.4–2.6) 2.1 (1.5–2.7) 10.4 (0.7–9.1) 3.3 (2.3–4.3) 5.4 (4.2–6.6) 13.6 (11.3–16) Crt vs. cmht P = 0.002 Crt vs. aot P = 0.29 cmht vs. aot P = 0.03 Crt vs. cmht P = 0.0005 Crt vs. aot P = 0.8 cmht vs. aot P = 0.0005 Crt vs. cmht P = 0.03 Crt vs. aot P = 0.5 cmht vs. aot P = 0.02 v2 = 0.11 P = 0.87 v2 = 14.8 P = 0.0005 25 (18.9) 107 (81.1) 27 (20.5) 14 (10.6) 28 (21.2) 63 (47.7) 0 36 (19.4) 150 (80.7) 93 (50.3) 65 (35.1) 21 (11.4) 1 (0.5) 5 (2.7) 20 (17.9) 92 (82.1) 93 (83.0) 13 (11.6) 1 (0.9) 4 (3.6) 1 (0.9) All values for v2 tests and associated P values in this table have been adjusted to take account of clustering by team using the survey estimation commands in STATA b For these variables the P values are derived by carrying out a regression analysis with the variable listed as the dependent variable and type of team as the sole independent variable, adjusting for clustering by team have very similar mean scores on the Minnesota Satisfaction Scale, both before and after adjusting for confounding factors, indicating mildly positive satisfaction. On the emotional exhaustion component of the MBI, both team types scored in the average burnout category. CRT staff appeared to experience less depersonalisation (adjusted mean diff is )3.2 P = 0.01), and also had a significantly greater sense of personal accomplishment (adjusted mean difference 4.4, CI 1.4–7.3, P = 0.006) than CMHT staff. j Burnout variation among the teams Figure 1 shows the distribution of team means on the three components of burnout. Regarding the emotional exhaustion component of the MBI, the AOTs were fairly evenly distributed between high, medium and low burnout categories. The majority of the CRTs, 8 of the 11(73%), had team means in the average burnout category, with 2 and 1 teams scoring low and high burnout respectively. The majority of the CMHTs; 7 of the 9 teams (78%) also scored in 545 Table 2 Means, confidence intervals and adjusted mean difference in satisfaction and burnout between CRT and AOT Measure Crisis resolution team staff (n = 132) mean (95% CI)a 5.3 (5.0–5.6) 71.2 (68.7–73.8) 42.8 (40.9–44.7) 21.0 (20.4–21.6) 17.7 (15.9–19.6) 4.8 (3.8–5.9) 36.7 (35.4–38.0) Assertive outreach team staff (n = 187) mean (95% CI) 5.1 (4.7–5.5) 70.8 (68.1–73.5) 42.2 (41.5–42.9) 20.8 (19.9–21.7) 17.4 (15.2–19.6) 4.4 (3.7–5.2) 34.8 (33.8–35.8) Unadjusted mean diff (CI)b P Adjusted mean diff (CI)c P JDS Global satisfaction Minnesota satisfaction scale General job satisfaction Intrinsic satisfaction Extrinsic satisfaction MBI Emotional exhaustion Depersonalisation Personal accomplishment a b 0.2 ()0.3 to 0.8) 0.4 ()3.4 to 4.2) 0.3 ()2.2 to 2.8) 0.2 ()0.9 to 1.3) 0.3 ()2.6 to 3.3) 0.4 ()0.9 to 1.7) 1.9 (0.2–3.5) 0.4 0.8 0.8 0.7 0.8 0.5 0.03 0.2 ()0.4 to 0.9) 0.2 ()2.6 to 3.0) )0.2 ()2.1 to 1.5) 0.4 ()0.5 to 1.2) )0.73 ()4.4 to 2.8) )1.23 ()2.6 to 0.03) 3.4 (1.3–5.6) 0.5 0.9 0.7 0.4 0.7 0.05 0.003 Means and confidence intervals adjusted to take clustering by team into account using the survey estimation commands in STATA Release 8 Teams compared using regression analysis with the burnout or satisfaction variable as the dependent variable and team type as the independent variable, adjusting for clustering by team c Teams compared using regression analysis with independent variables including the main baseline variables that were potential confounders Table 3 Means and confidence intervals for satisfaction and burnout: comparison between crisis resolution and community mental health teams Measure Crisis resolution team staff (n = 132) mean (95% CI)a 5.3 (5.0–5.6) 71.2 (68.7–73.8) 42.8 (40.9–44.7) 21.0 (20.4–21.6) 17.7 (15.9–19.6 4.8 (3.8–5.9) 36.7 (35.4–38.0) Community mental health team staff (n = 114) mean (95% CI) 4.7 (4.4–5.0) 70.9 (69.6–72.2) 42.2 (41.5–42.9) 21.2 (20.7–21.8) 19.0 (16.4–21.6) 5.7 (4.3–7.0) 32.7 (31.7–33.7) Unadjusted mean diff (CI)b P Adjusted mean diff (CI)c P JDS Global satisfaction Minnesota satisfaction scale General job satisfaction Intrinsic satisfaction Extrinsic satisfaction MBI Emotional exhaustion Depersonalisation Personal accomplishment a b 0.6 (0.2–1.1) 0.3 ()2.1 to 2.7) 0.6 ()1.0 to 2.1) )0.2 ()1.0 to 0.5) )1.26 ()4.6 to 2.1) )0.8 ()2.6 to 1.0) 4.0 (2.2–5.7) 0.008 0.8 0.5 0.6 0.4 0.3 0.0005 0.8 (0.1–1.5) 0.9 ()2.6 to 4.4) 0.5 ()1.8 to 2.8) )0.1 ()1.3 to 1.0) )2.2 ()7.3 to 2.9) )3.2 ()5.6 to 8.1) 4.4 (1.4–7.3) 0.3 0.6 0.6 0.8 0.4 0.01 0.006 Means and confidence intervals adjusted to take clustering by team into account using the survey estimation commands in STATA release 8 Teams compared using regression analysis with the burnout or satisfaction variable as the dependent variable and team type as the independent variable, adjusting for clustering by team c Teams compared using regression analysis with independent variables including the main baseline variables that were potential confounders the average burnout category, 2 teams incurring high burnout and no teams in the low burnout margin. Regarding depersonalisation, the majority of the CRTs (7of 11) and CMHTs (5 of 9) scored in the average burnout category. Half of the AOTs scored in Fig. 1 Burnout scores across the teams. NB: Staff are categorised as ‘low burnout’ for the ‘personal accomplishment’ component of burnout if they have high scores on this component 18 16 14 No. of teams 12 10 8 6 4 2 0 the low burnout category for depersonalisation, with only 2 of the 24 teams scoring high burnout on the depersonalisation dimension. The greatest proportion of CRTs (10 of 11) and AOTs (17 of 24) scored low burnout on the personal AOT CMHT CRT low average high low average depersonalisation Burnout high low average high emotional exhaustion personal accomplishment 546 accomplishment dimension of the MBI, whereas 6 of the 9 CMHTs scored in the average burnout category. member is associated with a greater sense of personal accomplishment. j Multivariate analysis Table 4 shows the results from an exploratory regression analysis of scores for all three types of team, aiming to identify factors that may be associated with job satisfaction and burnout. Multiple regression analyses were carried out with burnout and satisfaction scores as the dependent variables and the main socio-demographic and occupational factors as independent variables. They suggest that being Black or of mixed race is associated with poorer job satisfaction, particularly on the Minnesota Job Satisfaction Scale, but that being of Asian ethnicity is associated with less emotional exhaustion, less depersonalisation and a greater sense of personal satisfaction. It appears that being a psychiatrist is associated with greater intrinsic job satisfaction and psychologists experience less depersonalisation but that social workers have less extrinsic job satisfaction. A longer career in mental health and being older also seem to be associated with experiencing less emotional exhaustion and less depersonalisation. The analysis also indicates that CMHT members have poorer general job satisfaction, a greater sense of depersonalisation and less personal accomplishment. However being an AOT j Sources of satisfaction and stress Figure 2 shows the mean ratings for the sources of satisfaction for the three teams, with 0 indicating that this factor was not considered a source of satisfaction at work, and 4 being a very important source. The profiles were in general very similar for each type of team. CRT workers had significantly higher satisfaction from; ‘the company of the staff they worked with’ (P = 0.022), and ‘contributing to the team’s service’ (P < 0.0005) compared to CMHT workers; ‘managing and supervising others’ compared to AOT (P = 0.007) and CMHT (P = 0.002) staff and ‘the salary they got paid’ compared to AOT (P = 0.011) and CMHT workers (P < 0.0005). These were, however, exploratory comparisons without adjustment for multiple testing. The areas of potential satisfaction specific to CRTs and AOTs, such as the ‘‘team approach to patients’’ and ‘‘working in an innovative type of team’’, attracted high ratings. Similarly, factors that were specific to CRTs alone, such as ‘‘keeping clients out of hospital’’, ‘‘the benefits of team working’’, ‘‘seeing change more quickly’’ and ‘‘greater autonomy’’ emerged as important sources of satisfaction, apart from ‘‘the on call aspects of the job’’, which were not considered a source of satisfaction. Table 4 Variables associated with satisfaction and burnout scores in crisis resolution, assertive outreach and community mental health teams Characteristics associated with measure at least P = 0.05 level Job diagnostic survey Global satisfaction Minnesota satisfaction scale General satisfaction Intrinsic satisfaction Extrinsic satisfaction Maslach burnout inventory Emotional exhaustion Depersonalisation CMHT membera (poorer satisfaction) Longer career in mental health (greater satisfaction) Mixed ethnic group (poorer satisfaction) Black ethnic group (poorer satisfaction) Office hours, incl. evenings & weekends, no overnight (less satisfaction) Psychiatrist (greater satisfaction) Mixed ethnic group (poorer satisfaction) Black ethnicity Mixed ethnic group (poorer satisfaction) Social Worker (poorer satisfaction) Asian ethnic group (less exhaustion) Time as mental health worker (less exhaustion with longer career) CMHT member (greater depersonalisation) Female (less depersonalisation) Psychologist (less depersonalisation) Asian ethnic group (less depersonalisation) Age 46–54 (less depersonalisation) Aged over 55 (less depersonalisation) Time as mental health worker (less depersonalisation with longer career) Black ethnic group (less depersonalisation) Office hours, incl. evenings and weekends, no overnight (greater depersonalisation) Shifts, telephone on call only (greater depersonalisation) Shifts, plus on call overnight (greater depersonalisation) AOT member (greater personal accomplishment) CMHT member (less personal accomplishment) Asian ethnic group (greater personal accomplishment) Time in type of team (greater personal accomplishment) Regression coefficient (95% CI) )0.5 ()0.9 to )0.1) 0.03 (0.003–0.05) )7.1 )2.9 )2.8 2.5 )4.5 )2.1 )2.2 )1.0 )5.5 )0.2 1.8 )1.3 )2.6 )1.9 )3.3 )4.7 )0.1 )2.1 1.6 1.9 1.9 3.0 )2.1 3.4 0.5 ()12.0 to )2.3) ()5.6 to )0.2) ()5.4 to 0.3) (0.4–4.7) ()7.6 to )1.4) ()3.8 to )0.3) ()3.8 to )0.6) ()1.9 to )0.1) ()9.0 to )2.0) ()0.4 to )0.01) (0.6–3.0) ()2.2 to )0.4) ()5.0 to )0.2) ()3.4 to )0.4) ()6.3 to )0.4) ()8.3 to )1.2) ()0.2 to )0.01)) ()3.3 to )0.9) (0.5–2.8) (0.4–3.4) (0.3–3.5) (0.8–5.2) ()4.1 to )0.1) (0.8–6.0) (0.2–0.9) P 0.008 0.03 0.004 0.04 0.04 0.02 0.005 0.02 0.006 0.03 0.002 0.03 0.003 0.003 0.04 0.002 0.03 0.008 0.03 0.001 0.005 0.02 0.02 0.008 0.04 0.01 0.005 Personal accomplishment a CRTs were the reference category with which other team types were compared 547 Fig. 2 Mean ratings for sources of staff satisfaction for the three teams, with 0 indicating that this factor is not considered a source of satisfaction at work and 4 being a very important source CMHT on call aspects keeping clients out of hospital benefits of teamworking seeing change more quickly greater autonomy working in a new type of team salary I get paid contributing to the teams' service managing and supervising others helping clients company of staff 0 0.5 1 1.5 2 2.5 3 3.5 AOT CRT Figure 3 shows the sources of job stress for the three teams. All the teams rated ‘a lack of people/resources in the community to refer on to’ as the highest source of stress followed by ‘too much administrative work’. Areas rated as more important sources of stress for CRT staff than for others were; ‘Lack of support from seniors’ compared to CMHT (P = 0.004) and ‘poor communication within the team’ compared to both AOT (P = 0.013) and CMHT (P < 0.0005). j Setting There are very clear guidelines set out in the MHPIG for the form and function that a CRT should have. We attempted to assess whether some basic aspects of ‘model fidelity’ existed in the teams surveyed by asking for the different team configurations and characteristics. This study incorporated CRTs that had been in existence for and average of 3 years and 5 months, ranging form 8 months to 5 years and half years duration. The teams were fairly uniform in terms of composition and service provision. The CRTs operated in a similar way with all teams acting as gatekeepers to admission and most teams operating over the 24-h period. Only two teams were not providing a 24 h service, and one of these planned to when they had recruited the full complement of staff. It would be worth surveying this again after a period of time had elapsed to see whether any adaptations had occurred. CMHT AOT CRT Discussion A high response rate was achieved giving good representation of staff views of urban community mental health teams and newer specialist community based services. Findings will be less representative of rural areas, but we felt that this was an acceptable limitation given that 10% of CRTs are located in rural areas [19]. Fig. 3 Mean ratings for sources of stress for AOT CRT and CMHT staff with 0 indicating that this factor is not considered a source of stress at work and 4 being a very important source having to persuade professionals about home treatment visiting people who live in dangerous areas working with suicidal people in their homes working with violent people in their homes taking a team approach being called out at night poor communication within the team too much admin work lack of support from seniors lack of people/resources in the community to refer to 0 0.5 1 1.5 2 2.5 548 The national survey of crisis resolution teams in England in 2005 [19] (team managers entered data directly via a web site with telephone interviewer support) found that only 40% of teams were set up as per MHPIG recommendations and that teams were at around 88% of their recommended staffing capacity. Lack of resources and particularly staff were the most frequently cited obstacle to effective implementation, although with resources in place CRTs were not perceived as difficult to recruit to. Working collaboratively with other parts of the local system was perceived as a significant challenge to proper implementation of an effective CRT function. The survey did not cite dissatisfaction or burnout with the nature of the work as factors that impeded the work of CRTs. j Socio-demographic characteristics and job details The composition of the CRTs was notably medical, with a high proportion of community psychiatric nurses and doctors. This was a more restricted disciplinary mix than suggested in the Mental Health Policy Implementation Guidelines and other expert opinion, and may create a relatively narrow focus on symptom management rather than the intended social systems approach. The relatively large number of doctors probably reflects the fact that most of the CRTs surveyed covered a number of community sectors in their catchment area and had to liaise with several sector consultants rather than having a dedicated CRT consultant. The majority of staff working in CRTs (79%) worked a shift pattern including being on call during the evenings and weekends, whereas half of the AOT and most of the CMHT staff (83%) worked office hours. This may be explained by the possibility of office hours being more attractive to an older age group who may have family commitments, whereas younger staff may be attracted to the smaller specialist teams that work in an innovative way. variables. Results suggest that having a longer career in mental health confers less emotional exhaustion and greater sense of personal accomplishment. This might be explained by people who aren’t happy in the job tending to leave, or that the confidence and coping skills of long experience reduce depersonalisation. Our analysis also found that being of Asian ethnicity is associated with less emotional exhaustion, depersonalisation and a greater sense of personal accomplishment and being Black or mixed race was associated with greater burnout. This contrasts with the PLAO study and other studies [2, 22] which actually note lower burnout for black staff. The basis for this is unclear. Regarding the burnout scores across the teams, in contrast to the AOTs where the teams were quite widely distributed in terms of burnout scores with some teams looking quite burnt out, the CRT pattern looks more uniform, with only one team scoring in the high category on only one indicator. So despite the variations in team age, composition, working practices, the model seems to be reasonably agreeable in terms of staff experiences. Thus overall the new configuration of mental health services appears so far to be relatively benign in terms of effects on staff morale, and CRTs in particular appear to offer a satisfying work environment, though it is possible that they have attracted the most motivated and satisfied staff. j Factors influencing CRT morale According to the Demand-Control-Support Model of job stress [27], the demands of working in CRTs are likely to be high, but these may be mitigated by the sense of autonomy staff report, likely to indicate relatively high job control, and the support resulting from working in a cohesive team. This study tentatively supports this notion, as factors specific to CRTs are satisfying aspects of the work, these being: greater autonomy; seeing change more quickly; the benefits of team working; and keeping clients out of hospital. A further factor that may explain staff’s relatively positive views of this model is that the clearly defined specialist role of the team may mean that lack of professional role clarity is less of a problem in the newer specialist teams than among the relatively demoralised CMHTs surveyed in some of the 1990s studies. In order to assess this fully we would need to see what happens to staff satisfaction and burnout in the long term. j Main findings CRT and AOT staff showed similar levels of satisfaction with their work, both groups reporting fairly good levels of satisfaction and sense of personal accomplishment. CMHT staff were less satisfied, and had lower levels of personal accomplishment. No type of team showed high burnout. AOT and CRT staff had a higher sense of personal accomplishment than those working in CMHTs, CRT staff having the highest level of personal accomplishment overall. These findings regarding CMHTs are consistent with other more recent studies, though less with the studies of CMHTs in the 1990s that tended to indicate high levels of emotional exhaustion [4, 5, 22, 29]. The multivariate analysis used satisfaction and burnout scores as dependent variables to explore the effect of the socio-demographic and occupational j Limitations These findings will be less representative of rural and smaller town settings. There were differing modes of operation and configurations of teams within each type of team, therefore teams may be too heteroge- 549 neous to aggregate, although we attempted to control for this using the clustering technique in the regression analyses. Staff may have been wary of disclosing their true feelings about work, although no more so than in other self completed questionnaire surveys. j Declaration of interest j Funding None. None. References 1. Audini B, Marks I, Lawrence R, Connolly J, Watts V (1994) Home based versus out-patient/inpatient care for people with serious mental illness: Phase II of a controlled study. Br J Psychiatry 165:204–210 2. Billings J, Johnson S, Bebbington P, Greaves A, Preibe S, Muijen M, Ryrie I, Watts J, White I, Wright C (2003) Assertive outreach teams in London: staff experiences and perceptions. PanLondon assertive outreach study part 2. Br J Psychiatry 183:139–148 3. Dedman P (1993) Home treatment for acute psychiatric disorder. Br Med J 306:1359–1360 4. Edwards D, Burnard P, Coyle D, Fothergill A, Hannigan B (2000) Stressors, moderators and stress outcomes: findings from the all wales community mental health nurse study. J Psychiatr Ment Health Nurs 7:529–537 5. Fagin L, Brown D, Bartlett H, Leary J, Carson (1995) The Claybury community psychiatric nurse study: is it more stressful to work in hospital or the community' J Adv Nurs 22:347–358 6. Ford R, Minghella E, Ryrie I (2001) Assertive outreach and crisis resolution: moving forward the research and development agenda. The Sainsbury Centre for Mental Health. http://www.scmh.org.uk 7. Freudenburger H (1974) Staff burn-out. J Soc Issues 30:159–166 8. Glover G, Arts G, Babu-Kannan S (2006) Crisis resolution/home treatment and psychiatric admission rates in England. Br J Psychiatry 189:441–445 9. Hackman JR, Oldham GR (1975) Development of the job diagnostic survey. J Appl Psychol 60:159–170 10. Johnson S, Nolan F, Hoult J, White IR, Bebbington P, Sandor A, Mckenzie N, Patel S, Pilling S (2005) The outcomes of crises before and after the introduction of a crisis resolution team. Br J Psychiatry 187:68–75 11. Johnson S, Nolan F, Pilling S, Sandor A, Hoult J, Mckenzie N, White IR, Thompson M, Bebbington P (2005) Randomised controlled trial of mental health care by a crisis resolution team: the North Islington crisis study. Br Med J 331:599 12. Joy CB, Adams CE, Rice K. (last updated 2006) Crisis intervention for people with severe mental illnesses. Cochrane Database Syst Rev 1998, Issue 4. Art. No.: CD001087. doi:10.1002/14651858.CD001087.pub3 13. Karasek R Jr (1979) Job demands, job decision latitude and mental strain: implications for job redesign. Adm Sci Q 24:285– 308 14. Killaspy H, Bebbington P, Blizard R, Johnson S, Nolan F, Pilling S, King M (2006) The REACT study: randomised evaluation of assertive community treatment in north London. Br Med J 332:815–820 15. Maslach C, Jackson S (1986) Maslach burnout inventory, 2nd edn. Oxford Psychological Press, Oxford 16. Mental Health National Service Framework (and the NHS Plan) Workforce Planning, Education and Training. Underpinning Programme: Adult Mental Health Services http://www.dh.gov.uk 17. National Audit Office (2007) Helping people through mental health crisis: the role of the crisis resolution and home treatment services. Report by the Comptroller and Auditor General. HC 5 Session 2007–2008 18. NHS Plan (2000) a plan for investment, a plan for reform. Department of Health http://www.dh.gov.uk 19. Onyett S, Linde K, Glover G, Floyd F, Bradley S, Middleton H (2006) A national survey of crisis resolution teams in England. http://eprints.nottingham.ac.uk/789/1/national-crisisteam-survey. Currently in press 20. Onyett S, Pillinger S, Muijen M (1997) Job satisfaction and burnout among members of community mental health teams. J Ment Health 6(1):55–66 21. Priebe S, Fakhoury W, Hoffman K, Powell R (2005) Morale and perception of community mental health professionals in Berlin and London. Soc Psychiatry Psychiatr Epidemiol 40:223–232 22. Prosser D, Johnson S, Kuipers E, Dunn G, Szmuckler G, Reid Y, Bebbington P, Thornicroft G (1999) Mental health, ‘‘burnout and job satisfaction in a longitudinal study of mental health staff. Soc Psychiatry Psychiatr Epidemiol 34:295–300 23. Sainsbury Centre for Mental Health (2001) Finding and keeping: a review of recruitment and retention in the mental health workforce. The Sainsbury Centre, London 24. Schaufeli W, Enzman D, Girault N (1993) Measurement of burnout: a review. In: Schaufeli W, Maslach C, Marek C (eds) Professional burnout: recent developments in theory and practise. Taylor & Francis, Washington, DC 25. Taber T, Taber E (1990) A review and evaluation of the psychometric properties of the job diagnostic survey. Pers Psychol 43(3):467–500 26. Tummers GER, Janssen PPM, Landeweerd A, Houkes I (2001) A comparative study of work characteristics and reactions between general and mental health nurses: a multi-sample analysis. J Adv Nurs 36(12):151–162 27. Van der Doef M, Maes S (1999) The job demand-control (-support) model and psychological well being: a review of 20 years empirical research. Work Stress 13:87–114 28. Weiss DJ, Davis RV, England GW, Lofquist CH (1967) Manual for the Minnesota Satisfaction Questionnaire. Industrial Relations Center, University of Minnesota, Minneapolis 29. Wykes T, Sevens W, Everitt B (1997) Stress in community care teams: will it affect the sustainability of community care' Soc Psychiatr Epidemiol 32:398–407