English

AIDS CARE (2000), VOL. 12, NO. 2, pp. 149–161

Work-related stress and occupational burnout in AIDS caregivers: test of a coping model with nurses providing AIDS care

V. GUERITAULT-CHALVIN,1 S. C. KALICHMAN,2 A. DEMI1 & J. L. PETERSON1 1Georgia State University & 2Center for AIDS Intervention Research, USA

Abstract Occupational stress and burnout are potential threats to quality of care for people living with HIV/AIDS. A total of 445 nurses who provide care to people living with HIV/AIDS responded to an anonymous postal survey that consisted of: demographic and work history questions, the Maslach Burnout Inventory, Rotter’s Internal–External Locus of Control scale, and internal and external coping styles as measured by the Ways of Coping scale. Path analyses showed that both external (13.6% of the variance) and internal (3.1% of the variance) coping styles signi� cantly predicted levels of burnout among AIDS caregivers, over and above participants’ age, perceived workload and locus of control. Mediation analyses further showed that external coping mediates the effect of locus of control on burnout, but internal coping styles do not exhibit mediation. Results therefore replicate and extend previous research demonstrating the important roles of cognitive and behavioural coping styles in understanding burnout among providers of AIDS care.

Introduction

Occupational burnout is conceptualized as a particular type of stress occurring principally in professional contexts where work demands, especially those of an interpersonal nature, lead to chronic emotional exhaustion, depersonalization and reduced sense of personal accom- plishment (Cordes & Dougherty, 1993). Of particular importance are the consequences of occupational burnout, including staff turnover, absenteeism and reduced productivity; facets of occupational burnout that have serious repercussions for organizations, services and individuals. Occupational burnout has been identi� ed as a signi� cant problem in health care professions, particularly for professionals working in chronic care.

Research in the past few years has shown that occupational burnout is a serious problem in the area of AIDS care (Bennett & Kelaher, 1994; Bennett et al., 1994a; 1994b; 1995). Burnout among AIDS service providers has been discussed more anecdotally than through empirical research. However, some studies reported in recent years shed light on the issues related to AIDS occupational burnout. Whether HIV/AIDS health care workers face unique demands that could make them more prone to experiencing burnout in comparison to

Address for correspondence: Violaine Chalvin, Department of Psychology, College of Arts and Sciences, Georgia State University, Atlanta, Georgia 30303–3083, USA. Tel: 1 1 (770) 641 6341; Fax: 1 1 (770) 641 6322; E-mail: [email protected]

ISSN 0954-0121 print/ISSN 1360-0451 online/00/020149-13 Ó Taylor & Francis Ltd

150 V. GUERITAULT–CHALVIN ET AL.

workers in other health care areas has attracted the most attention of researchers. AIDS care workers face similar stress-producing activities found among cancer care providers and therefore studies have compared workers in those two � elds in order to determine whether AIDS work generates comparable stress. Bennett et al. (1991) compared nurses working in HIV/AIDS with those in oncology units and found that while there were no signi� cant differences in the frequency of burnout, when burnout was observed in HIV/AIDS nurses there was a signi� cantly greater intensity of symptoms. Similarly , Kleiber et al. (1993) found no differences in the levels of burnout reported by HIV/AIDS nurses compared with nurses in oncology and geriatrics. However, Kleiber et al.’s study found that HIV/AIDS nurses had been on the job for shorter periods of time, suggesting that burnout for AIDS care providers may have occurred more rapidly.

It is, however, important to note that there are demands speci� c to AIDS services that may account for the greater intensity and more rapidly developing burnout experienced by AIDS care providers. One major stress-producing concern shared by many AIDS workers is anxiety over safety practices and fears of occupational contagion. These anxieties have been found to be very stressful for AIDS caregivers and rooted in occupational exposures to HIV. Although such instances are rare and the risk of such exposure remains very low (Blumen� eld et al., 1987; Gordin et al., 1987; Pleck et al., 1988; Wallack, 1989), infrequent but serious events are known to produce particularly high levels of fear (Kahneman & Tversky, 1972).

The tendency of many professionals to identify with their patients for whom they feel compassion and empathy can have major implications in terms of emotional stress. Over- identi� cation with patients is often facilitated by the fact that most individuals who have AIDS belong to the same age group as service providers; more than 60% of people with AIDS are under the age of 40. Patients of younger age together with the high mortality rate of AIDS often increases professionals’ fear of death and death-related anxiety. In addition, gay physicians are more likely than heterosexual physicians to report increased fear of death and higher levels of anxiety after working with AIDS patients (Bennett, 1992; Hortsman & McKusick, 1986). Ross and Seeger (1988) also found that professionals for whom the young age of their AIDS patients was a real concern reported higher levels of burnout than professionals who did not show such concerns.

Few studies have investigated the impact of coping styles on frequency and intensity of burnout experienced by HIV/AIDS caregivers. However, results from these studies show that individual coping strategies may be important buffers of perceived stress. For example, Martin (1990) found that coping styles that involve cognitive and behavioural disengagement were associated with higher burnout and greater intentions to discontinue working in the AIDS � eld.

In an in� uential study dealing with the impact of coping styles on reported levels of burnout, Bennett et al. (1993) conducted a longitudinal investigation of nurses, doctors and social workers in three Australian states. This study found that coping style was a major determinant of burnout. External coping strategies such as fatalistic attitudes, negative expectations and reliance on faith, prayers or miracles, as opposed to internal coping strategies such as expression of feelings and emotions, patience and time-out, along with age of the provider and hours per week spent working in an AIDS unit, all predicted burnout. Coping was measured using the Revised Ways of Coping scale (Lazarus & Folkman, 1984). The beginning of the questionnaire was reframed by asking respondents to answer as health professionals in the HIV/AIDS area. While an initial factor analysis produced a number of coping dimensions, a higher-order analysis produced two factors labelled external and internal coping. External coping included fatalistic strategies (passive, fatalistic attitudes), pessimistic strategies (negative expectations), religiosity (faith, prayers and miracles) and

OCCUPATIONAL BURNOUT IN AIDS CAREGIVERS 151

denial (avoidance of dif� cult situations). Respondents who tended to report using external coping with work-related stress did not believe that they could alter the situation they were in and therefore avoided, denied or passively accepted it. Internal coping included positive coping strategies (positive and optimistic attitudes and outcomes), self-expressive strategies (expression of feelings and emotions), vigilant strategies (patience, persistence and waiting) and time-out coping strategies (escapist and self-preserving coping techniques). Thus, internal coping may translate to a greater sense of control over work-related stressors.

External and internal coping styles, as described by Bennett et al. (1993), are reminiscent of Rotter’s (1966) locus of control theory in which individuals describe themselves as being either (1) externally controlled by factors such as luck, chance or fate, or (2) internally controlled, believing that a person has the ability to in� uence their environment and what happens to them. With reference to Bennett et al.’s (1993) study, external control may be similar to external coping, with externally controlled persons being more likely to use external coping strategies because they believe that they cannot alter their environment and must accept the control of others or fate. Internal control is similar to internal coping, perhaps with internally controlled persons being more likely to use internal coping styles because they believe that they can control their environment and stressful events. In their prospective study, Bennett et al. found that the use of external coping mechanisms led to higher levels of burnout among HIV/AIDS caregivers. These results support studies that found external coping styles associated with increased stress and burnout compared to use of internal coping strategies. Indeed, Berkeley Planning Associates (1977) found that people with high internal locus of control reported lower levels of burnout in a study of child abuse service workers. Similarly, Krause and Stryker (1984) reported that individuals who believed they have some control over their own lives coped more effectively with stress.

Considering the importance that different coping strategies seem to have on burnout in the � eld of HIV/AIDS care, the current study was conducted to replicate and extend previous research. We attempted to determine which psychological coping skills are related to burnout in health care professionals working in the HIV/AIDS area. To do so, we used methods similar to those employed by Bennett et al. (1993) and extended their � ndings by expanding the conceptualization of coping strategies. Because the results were based on a single sample and a factor analysis with unknown reliabilit y due to the small size of the sample, these results require replication. Moreover, because Bennett et al.’s sample was Australian, generalization of their � ndings across cultures is an important aspect of continued research. Finally, Bennett tested only a partial model that did not include locus of control or internal coping. Thus, the current study extended Bennett’s earlier work by hypothesizing that: (1) age and perceived workload would predict burnout; (2) HIV/AIDS health professionals’ coping styles as de� ned by Bennett et al. would predict occupational burnout over and above age and workload; and � nally (3), locus of control would predict both coping styles and occupational burnout.

Method

Participants

Nurses specializing in AIDS care (N 5 499) responded to a postal survey. The mean age of the participants was 43.9 years (SD 5 0.41, range 5 23–74); 16% of the respondents were male and 84% were female. Most participants (76%) reported being heterosexual, 19% homosexual and 3% bisexual. In addition, 42% of respondents reported having no children, 39% had one or two children, while 18% had three or more children. The majority of the

152 V. GUERITAULT–CHALVIN ET AL.

Table 1. Descriptive statistics for male and female nurses working in AIDS care

Male Female Variable nurses nurses

N 70 375 Age (years)

M 42.6 44.1 SD 8.3 9.4 Range 25–64 23–74

Sexual orientation (%) Heterosexual 17.1 87.5 Homosexual 78.6 8.8 Bisexual 4.3 2.9

HIV status (%) Negative 78.6 87.2 Positive 11.4 0.5 Unknown 10 12.3

# children (%) No children 84.3 33.6 1–2 children 11.4 45.1 3 or more children 2.9 21.3

Work hours per week M 40.9 38.9 SD 8.8 8.2 Range 4–64 0–60

participants (59%) reported working an average of 40 hours per week (M 5 38.7, SD 5 9.2) (see Table 1).

Instruments

Participants completed an anonymous survey containing demographic information, work history, the Maslach Burnout Inventory (MBI), the Internal–External Locus of Control scale and the Ways of Coping revised scale.

Demographic information. Participants were asked to report their age, sex, number of children, sexual orientation and the highest degree they have obtained.

Work history. Participants provided information about the number of hours per week they worked as a nurse, and in what year they started working as a nurse with HIV/AIDS patients. They reported whether they had taken time off from AIDS care that lasted six months or more and if they responded yes, they were asked how many times they had taken a break. Respondents were asked how many hours per week they provided direct care to people with HIV/AIDS, the number of patients they treat in an average week, the number of patients they have cared for in the last three months and the number of patients who have died of complications of AIDS under their care in the past three months. They were also asked to describe how heavy they perceived their workload to be and whether they were doing any volunteer work in the HIV/AIDS area outside their job.

OCCUPATIONAL BURNOUT IN AIDS CAREGIVERS 153

The Maslach Burnout Inventory (MBI). The Maslach Burnout Inventory is a 22-item questionnaire responded to on a seven-point scale from 0 to 6. The MBI was designed to assess the three main aspects of the burnout syndrome: emotional exhaustion, depersonaliza- tion and lack of personal accomplishment. Each of those aspects is measured by a separate sub-scale. The Emotional Exhaustion (EE) sub-scale measures feelings of being emotionally exhausted by one’s work. The Depersonalization (Dp) sub-scale assesses the negative and detached attitude of a worker toward a client or patient. The Personal Accomplishment (PA) sub-scale measures one’s feelings of competence and successful achievement in one’s work. In the current sample, the MBI was found internally consistent, alpha 5 0.88.

The Internal–External Locus of Control (I–E) scale. The I–E scale is a 29-item, forced-choice test including six � ller items intended to make the purpose of the test more ambiguous. Rotter’s (1966) I–E scale measures the extent to which individuals are either internal or external in their causal attributions for events. Individuals who have an internal locus of control believe they have the ability to in� uence their environment, while individuals who have an external locus of control tend to perceive events as being under the control of luck, chance or fate, i.e. factors external to themselves. In the current sample the I–E scale was found to be internally consistent, alpha 5 0.74.

The Ways of Coping (revised) scale. The Ways of Coping scale is a 67-item inventory responded to on a four-point scale from 0 to 3, intended to assess cognitive and behavioural coping responses to stressful situations. However, in the current study 33-items were selected for various reasons: the items had been demonstrated predictive of burnout in previous research with health care professionals working in AIDS (Bennett et al., 1993), they represented a diversity of coping strategies, and the constraints of the study required abbreviating the original scale.

For the purposes of data reduction to compose two coping scales, a principle compo- nents analysis was conducted selecting two orthogonal factors resulting from a Varimax rotation both of which had Eigenvalues greater than 1. Results showed that the � rst factor (Eigenvalue 5 5.7) consisted of 19 items with factor loadings of 0.30 or greater, accounting for 15.1% of the variance in the analysis. Results indicated that the second factor (Eigen- value 5 3.2) consisted of 14 items with factor loadings of 0.30 or greater, accounting for 11.9% of the variance in the analysis. Inspection of factor 1 showed that it represented problem-focused or internal coping (for example, ‘I changed or grew as a person in a good way’) and factor 2 represented emotion-focused or external coping (for example, ‘I wished that the situation would go away or somehow be over with’). Factor scores, which have a mean of 0 and a standard deviation of 1, were computed for each participant. Factor scores represent weighted combinations of items and are considered, by de� nition, internally consistent.

Procedures

A list of 3,000 nurses, members of the Association of Nurses in AIDS Care (ANAC) was obtained. Names (N 5 1500) were randomly selected from the list by pulling labels in a non-systematic unordered fashion. A random table was not used since labels were sorted geographically (i.e. by states) and it seemed important to get a good representation of all states. Surveys were sent with a coverletter and a postage paid return envelope. A total of 523 surveys were returned, representing a 35% response rate.

154 V. GUERITAULT–CHALVIN ET AL.

Twenty-four of those 523 returned surveys were refusals; seven postage were returns because of an incorrect address, two respondents sent the survey back without completing it because they found it too long, and 15 surveys were sent back with notes explaining that the participants were unable to complete the questionnaire because they were not providing direct care to HIV/AIDS patients. Those 24 refusals were deducted from the total number of returned surveys and the remaining 499 surveys were included in the study.

Data analyses

To control for current work with people living with HIV/AIDS, 54 of the 499 participants who did not report treating at least one patient with HIV/AIDS per week were excluded from the analyses. To test the study hypotheses, a series of hierarchical multiple regression analyses were conducted: (1) MBI burnout scores (outcome variable) were regressed on coping styles and locus of control (internal) controlling for age and workload in the following order: step (a) workload; step (b) age; step (c) internal locus of control; step (d) external coping style; and step (e) internal coping style; (2) locus of control was regressed on age; (3) internal coping style was regressed on internal locus of control after controlling for age and workload; and � nally (4) external coping style was regressed on internal locus of control controlling for age and workload.

Age was selected as the only demographic characteristic to include because no other variable showed an empirical or conceptual relationship to burnout. Mean substitutions were used to replace missing data.

Results

Participants (N 5 445) were included in the analysis on the basis of their treating at least one patient with HIV/AIDS per week. On average, participants reported starting their nursing career in 1979, and the majority worked an average of 40 hours per week (M 5 39, SD 5 8.4). Participants started working with HIV/AIDS patients in the late 1980s and early 1990s (M 5 1989, SD 5 7.2). Most respondents (78%) have not taken a break from AIDS care for six months or more, and among those who did report taking breaks, 51% did so only once. Participants worked approximately 30 hours per week (M 5 28, SD 5 14.1), providing direct care to an average of 20 to 30 AIDS patients per week (M 5 30, SD 5 59.2). Participants reported a mean of 123 AIDS patients in the last three months (SD 5 176). They also reported an average of three patients having died in the last three months when under their care (M 5 3.2, SD 5 6.9). Eleven per cent of respondents perceived their workload to be light , 34% perceived it as medium, 38% heavy and 16% very heavy. When asked whether participants did any volunteer work in the HIV/AIDS area outside of their job, 41% responded that they did volunteer.

Bivariate correlation analyses

Table 2 presents the bivariate correlations for age, internal coping style, external coping style, MBI total scores, the MBI sub-scales of emotional exhaustion, depersonalization, and personal accomplishment, internal locus of control, gender and workload. Results showed the expected association between perceived workload and burnout scores. In addition, external coping was positively related to burnout, greater external coping corresponding to greater burnout and internal coping was inversely related to burnout. Thus, the pattern of bivariat e correlations was as expected.

OCCUPATIONAL BURNOUT IN AIDS CAREGIVERS 155

T a b

le 2 .

B iv

a ri

a te

co rr

el a ti on

s fo

r a ll

v a ri

a bl

es in

cl u de

d in

p a th

a n a ly

se s

a s

w el

l a s

p a rt

ic ip

a n ts

’ ge

n d er

V ar

ia b

le 1

2 3

4 5

6 7

8 9

1 0

1 .

A g e

— 2

. In

te rn

al co

p in

g 0 .0

3 4

— 3

. E

x te

rn al

co p

in g

2 0 .1

1 5 *

0 .0

0 0

— 4

. M

B I

2 0 .1

6 7 * *

2 0

.1 6 9

** 0 .4

2 2 * *

— 5

. E

E S

2 0 .1

5 0 * *

2 0

.0 8 4

0 .3

8 5 * *

0 .8

7 9

** —

6 .

D P

S 2

0 .1

8 9 * *

2 0

.0 7 7

0 .3

9 2 * *

0 .7

3 2

** 0 .5

6 1 * *

— 7

. P

A S

2 0 .0

5 0

2 0

.2 3 9

** 0 .1

8 3 * *

0 .6

2 9

** 0 .2

6 0 * *

0 .2

5 2

** —

8 .

G en

d er

0 .0

6 0

0 .0

0 4

2 0 .1

0 5 *

2 0

.0 6 1

2 0 .0

6 2

2 0 .1

3 0

** 0

.0 3

6 —

9 .

W o rk

lo ad

2 0 .0

7 5

0 .0

1 3

0 .0

8 5

0 .2

4 4

** 0 .3

4 6 * *

0 .0

7 5

0 .0

2 4

0 .0

0 5

— 1

0 .

In te

rn al

lo cu

s o f

co n tr

o l

0 .1

2 0 *

0 .0

7 8

2 0 .2

0 8 * *

2 0

.2 4 0

** 2

0 .1

7 8 * *

2 0 .1

2 9

* 2

0 .2

4 6 * *

0 .0

3 1

0 .0

2 3

—

* C

o rr

el at

io n

is si

gn i�

ca n

t at

th e

0 .0

5 le

ve l ( t

w o -t

ai le

d ) ;

** C

o rr

el at

io n

is si

g n i�

ca n t

at th

e 0

.0 1

le ve

l ( t

w o -t

ai le

d ) .

156 V. GUERITAULT–CHALVIN ET AL.

Table 3. Results of path analysis predicting AIDS-related occupational burnout

Adjusted R square Sig. F Model R square R square change change

1. 0.056 0.053 0.056 0.000 2. 0.084 0.079 0.028 0.001 3. 0.129 0.122 0.045 0.000 4. 0.266 0.258 0.136 0.000 5. 0.296 0.287 0.031 0.000

Note. Dependent variable: MBI. 1. REGR workload; 2. REGR workload, age; 3. REGR workload, age, internal locus of control; 4. REGR workload, age, internal locus of control, external coping style; 5. REGR workload, age, internal locus of control, external coping style, internal coping style.

Path analysis

Workload was � rst tested as a predictor of burnout. Results show that workload signi� cantly predicted burnout, accounting for 5.6% of the variance in burnout scores. Age was then tested as a predictor of burnout and after controlling for workload was found to signi� cantly predict burnout, accounting for 2.8% of the variance over and above workload. Locus of control was also tested as an independent predictor of burnout. After controlling for age and workload, locus of control was found to signi� cantly predict burnout scores, accounting for 4.5% of the variance of burnout. Next, external coping signi� cantly predicted burnout, accounting for 13.6% of the total variance over and above the other variables. Finally, internal coping style was tested as an independent predictor of burnout. Results showed that after controlling for workload, age, locus of control and external coping style, internal coping style signi� cantly predicted burnout, accounting for 3.1% of the variance in burnout over and above the other variables.

Age was also found to predict locus of control, accounting for 2% of its variance and age accounted for 2% of the variance in external coping. Locus of control also predicted external coping style, accounting for 3.7% of its variance. However, locus of control did not signi� cantly predict internal coping style (see Table 3). Figure 1 presents the � nal path model showing only signi� cant paths and their associated adjusted regression weights.

Mediation effects1

Internal coping was tested as a possible mediator in the relation between the predictor variable locus of control and the outcome variable burnout. Results showed that the direct effect (DE) between locus of control and burnout was signi� cant ( b LCB.IC 5 2 0.128). Considering internal coping style as a possible mediator, the indirect effect (IE) calculated as IICLC * b BIC.LC was 2 0.01 and the total effect (TE) was calculated as DEBLC 1 IEBLC was 2 0.14. Given these results, we can say that internal coping style does not have a signi� cant mediating effect on the relation between burnout and locus of control since the direct effect of locus of control on burnout is signi� cantly larger (by 93%) relative to the indirect effect of locus of control on burnout via internal coping.

External coping style was then tested for mediation effects in the relation between locus of control and the dependent variable burnout. To determine whether external coping is a mediator in the relationship between locus of control and burnout, the indirect effect (IE) (IECLC* b BEC.LC) was 2 0.08 and the total effect (TE) (DELCB 1 IELCB) was 2 0.21. These results show that external coping has a signi� cant mediating effect on the relationship

OCCUPATIONAL BURNOUT IN AIDS CAREGIVERS 157

FIG. 1. Path model predicting occupational burnout among nurses working with HIV/AIDS patients (*p , 0.05; **p , 0.01).

between locus of control and burnout because the indirect effect of locus of control on burnout via external coping accounts for 38% of the total effect.

Finally, both internal and external coping taken together were tested for mediating effects in the relationship between locus of control and burnout. Considering internal and external coping as possible mediators, the indirect effect (IE) (IICLC * b ICEC * b ECB) was 0.001 and the total effect (TE) (DELCB 1 IELCB) was 2 0.127. Given these results, internal and external coping taken together do not have a signi� cant mediating effect on the relation between locus of control and burnout, since the indirect effect of locus of control on burnout is signi� cantly larger (by 100%) relative to the indirect effect of locus of control on burnout via internal and external coping styles (see Figure 2).

FIG. 2. Mediation effects of internal and external coping (*p , 0.05; **p , 0.01).

158 V. GUERITAULT–CHALVIN ET AL.

Discussion

The results of the current study replicate and extend previous research on occupational burnout among AIDS care providers, partially supporting the � ndings of Bennett et al. (1993). Bennett et al. found that external coping was positively correlated to burnout, suggesting that those who use a mainly external style of coping may be more prone to burnout. In a like manner, their path analysis suggested that the use of external coping as a primary coping mechanism led to higher levels of burnout. As expected in our hypotheses, we found that both external and internal coping styles signi� cantly predict burnout over and above age and workload with the use of external coping strategies positively correlated with levels of burnout. Internal coping showed a protective effect where people who were using more internal coping strategies reported experiencing lower levels of burnout. These data also support the � ndings of Berkeley Planning Associates (1977) and Martin (1990), which link external coping style to increases in stress and burnout.

Our hypothesis that perceived workload and age would signi� cantly predict burnout was supported. Age signi� cantly predicted burnout, external coping style, as well as locus of control. Older nurses were found to have a more internal locus of control and to use less external coping strategies when dealing with stressful events than did their younger counterparts. Results in this study also indicate that perceived workload is positively correlated to burnout and signi� cantly predicts burnout. This is congruent with the results of Bennett et al. (1992), who likewise found that people who have more concentrated exposure to HIV/AIDS (i.e. hours per week working in the HIV/AIDS � eld) may be more likely to experience burnout. It also supports the � ndings of Hortsman and McKusick (1986), in which more intense exposure to HIV/AIDS was identi� ed as a possible cause of burnout.

In order to extend previous research, the present study measured the extent to which external and internal coping styles derive from individuals’ locus of control (Rotter, 1966). As expected, locus of control signi� cantly predicted external coping where individuals who are more likely to attribute control to internal factors are less likely to use external coping strategies. However, no association was found between locus of control and internal coping, suggesting that people who have an internal locus of control are not more likely to rely on internal coping strategies when dealing with stressful situations. Locus of control itself signi� cantly predicted burnout where people who are internals were less likely to experience high levels of burnout. These � ndings are clari� ed by the mediation analyses, where both external and internal coping were tested for mediation in the relation between locus of control and burnout. While internal coping was not found to be a mediator, external coping was recognized as a signi� cant mediator in the relationship between locus of control and total burnout.

The results of this study must be considered in light of its limitations. The cross- sectional analysis limited our ability to identify cause and effect relations, so it is not possible to ascertain whether the use of external coping causes burnout or whether burnout causes people to use external coping. Then again, a third variable such as perceived workload or a variable that was not measured in this study could be the main cause of both burnout and coping style. Although the sample was diverse, geographically dispersed and represented a wide array of nurses, the 35% response rate cautions against over-generalizing the results to nurses working in AIDS. For example, it is possible that nurses who experience the most burnout in their job would not take time to complete the survey. In a similar fashion, nurses who did not experience burnout may not have felt concerned enough to complete the survey and may have been unable to respond to the type of coping styles they use in response to

OCCUPATIONAL BURNOUT IN AIDS CAREGIVERS 159

job-related stress. This low response rate could also be explained by our sampling proce- dures. Surveys were sent to members of the Association of Nurses in AIDS Care, many of whose members are not nurses taking care of HIV/AIDS patients on a regular basis. Several surveys were sent back with a note explaining that the respondent was not able to complete the survey since he/she was not providing direct care to HIV/AIDS patients either because he/she was now retired or because he/she was a university professor with no direct contact with patients. In any case, a 35% response rate is acceptable for postal surveys (Kerlinger, 1986), but should be considered when interpreting the generalizability of our � ndings.

Keeping these limitations in mind, the results of this study offer implications for preventing occupational burnout in nurses caring for HIVAIDS patients. Coping skills play an important role in the prevention of burnout. Control coping (i.e. internal coping) is cited as being among the resources most critical in � ghting burnout (Maslach & Goldberg, 1998). However, a focus on internal coping skills training may be insuf� cient to develop effective resilience to burnout. What would seem more important is to teach people how to determine the appropriate coping strategies they should use when dealing with speci� c stressors. Research on the prevention of stress and burnout emphasizes the importance of identifying coping strategies to deal effectively with speci� c stressors (Lelord, 1990). For example, interventions may help providers recognize the kind of coping strategies that they commonly use when dealing with stressful events. Once identi� ed, it is necessary to explain the reasons why the use of such strategies is likely to render them more susceptible to burnout. By then explaining how internal coping skills can protect against burnout, internal coping skills training becomes more meaningful. Moreover, as recommended by Lazarus and Folkman (1984) there should be a � t between the coping strategy used by a person and the problem or stress or experienced. According to Lazarus and Folkman’s model, people experience dif� culties when they try to use problem-focused coping (similar to internal coping) when dealing with situations they cannot change. The use of emotion-focused coping (similar to external coping) when faced with situations that can be changed can be just as frustrating, because the person in this case is not addressing the source of stress. In coping effectiveness training, people are trained to use internal coping strategies when faced with situations that can be controlled. When a situation cannot be controlled, external coping can be regarded as more appropriate yet should not be encouraged as a primary form of coping. Bennett et al. (1993) suggested that, ‘unit managers � nd ways to promote the psychosocial achievements of staff by shifting away from the cure-based achievement measures traditionally inherent in health care’ (p. 50). Although there may be no cure for a patient’s condition, internal coping can be encouraged and since previous research has shown that recognition and rewards tend to help AIDS care providers deal with burnout (Bennett et al., 1996), staff members should be rewarded for their attempts to provide palliative care, improve patients’ comfort and help with emotional problems.

Another important implication of the current results is the � nding that individuals who have a low internal locus of control maybe more inclined to seek external coping. Consequently, burnout prevention interventions should take individual differences into consideration by tailoring programmes to individual needs. Because age and per- ceived workload also predict burnout, young nurses who are found to have low internal locus of control and perceive their workload as heavy may be at particular risk for burnout and may require speci� c attention and training before entering an AIDS care unit. Thus, targeted and early intervention that emphasizes effective coping with work-related stress offers several opportunities for preventing and managing occupational burnout in AIDS care providers.

160 V. GUERITAULT–CHALVIN ET AL.

Acknowledgement

This research was supported by grant RO1-MH57624 from the National Institute of Mental Health awarded to Seth. C. Kalichman.

Note

Abbreviations: LC 5 locus of control; IC 5 internal coping; EC 5 external coping; B 5 burnout.

References

BENNETT, L. (1992). The experience of nurses working with hospitalized AIDS patients. Australian Journal of Social Issues, 27 (2), 125–143.

BENNETT, L. & KELAHER , M. (1993). Longitudinal determinants of patient care-related stress in HIV-AIDS health professionals. Presented at the IXth International Conference on AIDS/IVth STD world congress, Berlin, June 6–11.

BENNETT, L. & KELAHER , M. (1994). Longitudinal predictors of burnout in HIV/AIDS health professionals. Australian Journal of Public Health, 18 (3), 334–336.

BENNETT, L., KELAHER , M. & ROSS, M. (1993). Burnout and coping in HIV/AIDS health care professionals. In: H. VAN DIS & E. VAN DONGEN (Eds), Burnout in HIV/AIDS health care and support (pp. 41–51). Amsterdam: Amsterdam University Press.

BENNETT, L., KELAHER , M. & ROSS, M.W. (1994a). The impact of working with HIV/AIDS on health care professionals: development of the AIDS Impact Scale. Psychology and Health: An International Journal, 9 (3), 221–232.

BENNETT, L., KELAHER , M. & ROSS, M. (1994b). Quality of life in health care professionals: burnout and its associated factors in HIV/AIDS related care. Psychology and Health: An International Journal, 9 (4), 273–283.

BENNETT, L., MICHIE, P. & KIPPAX , S. (1991). Quantitative analysis of burnout and its associated factors in AIDS nursing. AIDS Care, 3 (2), 181–192.

BENNETT, L., MILLER, D. & ROSS, M. (1995). Review of the research to date on impact of HIV/AIDS on health workers . In: L. BENNETT, D. MILLER, & M. ROSS (Eds), Health workers and AIDS: research, intervention and current issues in burnout and response (pp. 15–34). London: Harwood Academic Publishers.

BENNETT, L., ROSS, M.W. & SUNDERLAND, R. (1996). The relationship between recognition, rewards and burnout in AIDS caring. AIDS Care, 8 (2), 145–153.

BERKELEY PLANNING ASSOCIATES (1977). Evaluation of child abuse and neglect demonstration projects (1974–1977). Volume IX, Project management and worker burnout. Washington, DC: US Department of Commerce.

BLUMENFIELD, M., SMITH P.J., MILAZZO , J., SEROPIAN , S. et al. (1987). Survey of attitudes of nurses working with AIDS patients. General Hospital Psychiatry, 9, 58–63.

CORDES, C.L. & DOUGHERTY, T.W. (1993). A review and an integration of research on job burnout. Academy of Management Review, 18 (4), 621–656.

GORDIN, F.M., WILLOUGHBY, A.D., LEVINE, L.A., GUREL, L. & NEILL, K.M. (1987). Knowledge of AIDS among hospital workers: behavioural correlates and consequences. AIDS, 1, 1183–1188.

HOMER, J.B. (1985). Worker burnout: a dynamic model with implications for prevention and control. System Dynamics Review, 1, 42–62.

HORTSMAN, W. & MCKUSICK, L. (1986). The impact of AIDS on the physician. In: L. MCKUSICK (Ed.), What to do about AIDS (pp. 63–74). Berkeley: University of California Press.

KAHNEMAN, D. & TVERSKY, A. (1972). Subjective probability: a judgement of representativeness. Cognitive Psychology, 3, 430–454.

KERLINGER, F.N. (1986) Foundations of behavioral research, 3rd edition. New York: Holt, Rinehart and Winston. KLEIBER, D., ENZMAN, D. & GUZY, B. (1993). Stress and burnout among health care personnel in the � eld of AIDS:

causes and prevalence. In: H. VAN DIS & E. VAN DONGEN (Eds), Burnout in HIV/AIDS health care and support: impact for professionals and volunteers (pp. 23–40). Amsterdam: University of Amsterdam Press.

KRAUSE, N. & STRYKER, S. (1984). Stress and well-being: the buffering role of locus of control beliefs. Social Science and Medicine, 18, 783.

LAZARUS, R.S. & FOLKMAN, S. (Eds) (1984). Stress, appraisal and coping. New York: Springer. LELORD, F. (1990). 10 Conseils pour bien vivre avec son stress. Paris: Publi-Union editions. LOWENSTEIN, L. (1991). Teacher stress leading to burnout—its prevention and cure. Education Today, 41, 12–16. LYALL , A. (1989). The prevention and treatment of professional burnout. Loss, Grief, and Care, 3, 27–32.

OCCUPATIONAL BURNOUT IN AIDS CAREGIVERS 161

MARTIN, D.A. (1990). Effects of ethical dilemmas on stress felt by nurses providing care to AIDS patients. Critical Care Nursing Quarterly, 12 (4), 53–62.

MASLACH, C.H. & GOLDBERG, J. (1998). Prevention of burnout: new perspectives. Applied and Preventive Psychology, 7, 63–74.

MASLACH, C.H. & JACKSON, S.E. (1981). Maslach Burnout Inventory manual, 2nd edition. Palo Alto, CA: Consulting Psychologists Press.

PLECK, J.H., O’DONNELL, L., O’DONNELL, C. & SNAREY, J. (1988). AIDS phobia, contact with AIDS and AIDS- related job stress in hospital workers. Journal of Homosexuality, 15 (3/4), 41–55.

ROSS, E. (1993). Preventing burnout among social workers employed in the � eld of AIDS/HIV. Social Work in Health Care, 18, 91–108.

ROSS, M. & SEEGER, V. (1988). Determinants of reported burnout in health professionals associated with the care of patients with AIDS. AIDS, 2, 395–397.

ROTTER, J.B. (1996). Generalized expectancies for internal versus external control of reinforcement . Psychological Monographs, 80 (whole number).

WALLACK , J.J. (1989). AIDS anxiety among health professionals. Hospital and Community Psychiatry, 40, 507–510.