Burnout and Social Support

in Rural Mental Health Counselors

J. Alan Kee, Ph.D., Don Johnson, Ph.D., & Portia Hunt, Ph.D.

In recent years, burnout has become a well known problem in the human service professions. A number of researchers have indicated that burnout is mentally and physically debilitating for workers, costly to institutions and agencies, and harmful to clients (Cherniss, 1980; Edelwich & Brodsky, 1980; Farber, 1983; Maslach, 1982; Maslach & Schaufeli, 1993; Pines, Aronson, & Kafry, 1981; Rogers, 1987). Even though a large body of literature on burnout has been developed over the past two decades, research on burnout in rural mental health providers has remained virtually nonexistent. Due to the unique practice conditions of the rural mental health provider, it may be that these practitioners are particularly at risk for burnout. Consequently, more research is needed in order to better understand the effects of the rural work environment on the mental health provider.

Herbert Freudenberger (1975) was the first clinician to use the term "burnout" in the literature to describe a state of physical and emotional depletion experienced by himself and his colleagues while working intensively in the free clinic movement of the late 1960s and early 1970s. During the late 1970s and early 1980s, the burnout literature was primarily anecdotal, and interest in the subject was mostly among practitioners in various human service fields rather than researchers (Daley, 1979; Forney, Wallace-Schutzman, & Wiggens, 1982; Kremer & Owen, 1979; Warnath & Shelton, 1976). The literature offered no consistency in the definition of the term "burnout." For example, Freudenberger and Richelson (1980) defined burnout as a "state of fatigue or frustration brought about by devotion to a way of life, or relationship, that has failed to produce the expected reward" (p. 13), whereas Edelwich and Brodsky (1980) defined burnout as a "progressive loss of idealism, energy, purpose, and concern as a result of conditions of work" (p. 14). The concept had been expanded to the point that it did not provide a specific and meaningful construct (Maslach, 1993).

A landmark event in the study of burnout that addressed this issue of a lack of a consistent definition was the development of the Maslach Burnout Inventory (Maslach & Jackson, 1981). This standardized measure, based on extensive empirical research, provided researchers with the methodological specificity necessary for studying the syndrome of burnout. From their analysis, three dimensions of burnout emerged: emotional exhaustion, depersonalization of clients, and lack of feelings of personal accomplishment. Maslach and Jackson describe emotional exhaustion as a depletion in emotional resources, feeling that one is no longer able to give of himself or herself. They characterize depersonalization as the development of negative, cynical attitudes and feelings toward one's clients. A lack of feelings of personal accomplishment is the tendency to view one's work with client's in a negative light, feeling a dissatisfaction with their accomplishments on the job.

From the late 1980s to the present, research on burnout has expanded into the development of more comprehensive models. Maslach (1993) indicates that several researchers have written about burnout from multiple perspectives (e.g., existentialism, social comparison theory, action theory, ego psychology, general stress theory, self-efficacy, organizational theory, and the psychology of creativity), and more current models of burnout are based on the outcomes of empirical research (e.g., Golembiewski, Scherb, & Boudreau, 1993; Maslach, 1993; Leiter & Maslach, 1988; Leiter, 1993). Maslach (1993) points out other current trends in research in burnout to include cross-national studies and expansion of research with other human service occupations (e.g., police, correctional officers, prison guards, librarians) as well as other types of occupations and other non-occupational areas of life (e.g., business world, sports, and family life).

Research clearly indicates that psychotherapists working in public agencies such as community mental health centers are among a group of employees categorized in a high-stress work environment (Dunbar, McKelvey, & Armstrong, 1980). Studies show that those employed in public settings are more dissatisfied and prone to occupational stress and burnout than those in private practice (Ackerley et al., 1988; Cherniss & Egnatios, 1978; Raquepaw & Miller, 1989). Rural mental health counselors are inclined to be subject not only to the stress of working in a public setting, but they also face an unusual array of potentially stressful conditions endemic to the rural community. These include ethical issues related to limits of competence and dual relationships (Berry & Davis, 1978; Flax, Wagonfeld, Ivens, & Weiss, 1979; Hargrove, 1982; Hargrove, 1986; Horst, 1989; Jennings, 1992; Schank & Skovholt, 1997), impact of a deteriorating economic base on funding for mental health services (Dyer, 1997; Human & Wasem, 1991; Murray & Keller, 1991; Paulsen, 1988), geographic barriers to the delivery of social services (Cohen, 1992), and professional isolation (Elkin & Boyer, 1987; Richards & Gottfredson, 1978; Sladen & Mozdzierz, 1989; Wagonfeld & Buffum, 1983).

The potential for burnout in counseling practice has been well documented. For example, Farber and Heifetz (1982) investigated prevalence of burnout in mental health professionals. These researchers found burnout in 71% of the psychologists, 43% of the psychiatrists, and 73% of the social workers. Another study by Farber (1985) discovered that thirty-six percent of their sample of mental health professionals reported moderate levels of burnout, and only 6.3% indicated a high degree of burnout. A study by Ackerley, Burnell, Holder, & Kurdek (1988) examined the level of burnout in a national sample of licensed psychologists. The results indicated that more than a third of the psychologists reported experiencing high levels of both emotional exhaustion and depersonalization. Also, 21% of the psychologists indicated that they would choose a career other than psychology if given the opportunity. Furthermore, Raquepaw & Miller (1989) completed a study of 68 Texas psychotherapists utilizing the Maslach Burnout Inventory. The psychotherapists in this study reported only low to moderate levels of burnout.

The literature regarding prevalence of burnout in many of the human service professions has been extensive (e.g., Maslach, 1982). However, the research on prevalence of burnout in particular groups has been limited and somewhat inconsistent. As seen in the research cited above, there is little consistency in reports of prevalence of burnout in psychotherapists. Also, at this time, no empirical study of burnout has been completed on rural mental health counselors. Maslach (1993) has suggested that studies on burnout should focus on specific work settings. This study specifies masters level rural mental health counselors as the work setting under investigation. Furthermore, this study is not only examining the prevalence of burnout in rural mental health counselors, but also it is examining the relationship between burnout and social support. Social support has been well documented as a highly effective intervention for coping with burnout (Pines, Aronson, & Kafry, 1981).

Hagihara, Tarumi, and Miller (1998) define social support as “the provision and receipt of tangible and intangible goods, services and benefits (such as encouragement and reassurance) in the context of informal relationships (e.g., family, friends, coworkers, and boss)” (p.75). Because current researchers recognize that social support is too complex to be limited to a single theoretical concept (Vaux, 1988), comprehensive models that incorporate the major elements of most current conceptualizations of social support have been developed. The model utilized in this study was proposed by Cutrona and Russell (1987). These researchers emphasize that in order to comprehensively understand social support, the following six dimensions of social support have to be taken into account: attachment - provided by relationships where a person receives a sense of security and safety; social integration - provided by relationships in which persons share interests and concerns; reassurance of worth - provided by relationships in which the person's skills and abilities are acknowledged; guidance - provided by relationships with trustworthy and authoritative persons who provided advice; reliable alliance - derived from relationships in which persons can count on others for assistance under any circumstances; and opportunity for nurturance - derived from relationships in which a person is responsible for the well-being of another.

The current study had three objectives. The first objective was to examine the relationship between the three dimensions of burnout and social support. The second objective involved predicting types of burnout (i.e., emotional exhaustion, depersonalization, and reduced personal accomplishment) from measures of social support. Finally, the third objective involved examining the relationship between level of risk for burnout and types of social support.

METHOD

Procedures

A list of potential participants was generated by purchasing the mailing list of all master’s level practitioners from the State of Kansas Behavioral Sciences Regulatory Board. Each potential sample member was set a packet containing a cover letter introducing the project, an informed consent form, demographic questionnaire, and all the instruments used in the current study. Packets were coded so that once the informed consent form was removed, the investigators could maintain client confidentiality. Upon receipt of the returned packets, investigators inspected demographic information to determine that the participants met the inclusion criteria. For those who did not, the data was discarded. All valid data was entered into a spread sheet format and analyses were performed using SPSS, version 8.0.

Instruments

The investigators selected the Maslach Burnout Inventory (Maslach & Jackson, 1981) and the Social Provisions Scale (Cutrona & Russell, 1987) for use in the current study.

Maslach Burnout Inventory (MBI)

The MBI (Maslach & Jackson 1981) is a 22-item self-report instrument for assessing burnout among human service professionals. It yields scores on three subscales: Emotional Exhaustion (EE), 9 items; Depersonalization (DP), 5 items; and Personal Accomplishment (PA), 8 items. The EE scale measures the inability of the worker to give of himself or herself at a psychological level. The DP scale measures the development of negative or cynical feeling towards one’s clients. The PA scale measures the worker’s perception of the importance of his or her work with clients. All 22 items use a seven-point Likert scale indicating the frequency of a feeling or perception. Item scores range from 0-6 (0 = Never, 6 = Every Day). The first two scales (EE & DP) are scored so that higher scores indicate greater problems with burnout. In its typical form, the PA scale is scored in the direction of lower scores indicating greater levels of burnout. Because the current investigators wanted to be able to develop a total score for burnout, the PA scale scoring procedure needed to be altered. Consequently, all items were reversed scored on the PA scale and cutoff scores for this scale were determined by subtracting the manual’s norm scores from the total score possible. Thus, the direction of the PA score and its interpretation were reversed. In this altered scoring format, higher PA scores indicate higher levels of burnout. Higher scores are then interpreted as indicating low levels of personal

accomplishment, associated with higher levels of burnout. Further, the scores from the three scales may then be added together to obtain a total burnout score.

Maslach conceives of burnout as a continuous variable and each scale is treated as an interval scale of measurement. Cutoffs have been determined to separate high, moderate and low risk levels for each of the three types of burnout. The three ranges for EE are: Low (0-16), Moderate (17-26), and High (27-54). For DP, the ranges are: Low (0-6), Moderate (7-12), and High (13-30). Using the reversed scored norms, the ranges for PA are: Low (0-8), Moderate (9-15), and High (16-48).

According to data from the test manual, the MBI scales have acceptable levels of reliability (Maslach et al., 1996). The three scales of EE, DP and PA have coefficient alphas of .90, .79, and .71 respectively. Test-retest at one year for the three scales was .60, .80, and .82 respectively.

Investigators established criterion validity of the MBI by comparing burnout scores to ratings of personal experience (Maslach & Jackson, 1981; and Maslach et al., 1996), dimensions of job experience (Maslach & Jackson, 1984; and Stout & Williams, 1983), or personal outcomes (Corcoran, 1986). In addition, researchers (Maslach et al., 1996; McCann & Pearlman, 1990; and Cox, Kuk & Leiter, 1993) have found evidence of construct validity. MBI scale scores have been shown to be related to measures of job satisfaction, depression, compassion fatigue, and occupational stress.

Social Provisions Scale (SPS)

The SPS (Cutrona & Russell, 1987) is a 24-item instrument measuring six different aspects of social support. Thus, it differs from other global measures (House, 1981; Holahan & Moos, 1982; and Wilcox, 1981) of social support. Participants respond on a four point rating scale indicating how true a statement is of their personal experience. The scale is from “not at all true” to “very true”. Four items (two positively stated and two negatively stated) measure each of six areas. These areas are: a) Attachment, when people feel a sense of safety and security; b) Social Integration, the perception of shared interests and concerns; Reassurance of Worth, when one’s skills and abilities are acknowledged; Guidance, the perception of the providing of trustworthy and authoritative advice; Reliable Alliance, being able to count on others for assistance under any circumstances; and Opportunity for Nurturance, from relationships in which a person is responsible for the well-being of another.

Cutrona and Russell (1987), in obtaining initial data on the instrument, found coefficient alpha reliabilities ranging from .65 to .76 across the six scales and .915 for the total score.

Russell, Cutrona, Rose, and Yurko (1984) found construct validity between attachment and social loneliness (Weiss, 1973; 1974). Further, they found a relationship between satisfaction with friendships, kin relationships, and work relationships and reliable alliance, reassurance of worth, and social integration. These relationships suggested good concurrent validity (Cutrona & Russell, 1987).

RESULTS

Participants were all volunteer, full-time, practicing, master’s level Licensed Professional Counselors (LPC) or Licensed Masters Level Psychologists (RMLP) in a midwestern state. Investigators obtained a list of all the LPC (316) and RMLP (502) practitioners from the state’s Behavioral Sciences Regulatory Board. Of these, 513 met the inclusion criteria for the present study, 177 LPC and 336 RMLP practitioners. The inclusion criteria included the following considerations. First, practitioners had to work in a “rural” area. That is, they had to work in an area smaller than a Standard Metropolitan Statistical Area (SMSA). An SMSA is defined as an area with one city larger than 50,000 inhabitants, or with a city larger than 25,000, which together with contiguous populated areas, totals more than 50,000 inhabitants. Second, the individual must be a provider of mental health services. Finally, third, the individual must work full-time as a deliverer of mental health care services.

All 513 sample practitioners criteria were mailed survey packets. Of that number, 220 individuals returned sample packets. Upon review of the demographic data, 28 of these individuals did not currently meet the inclusion criteria, and the investigators discarded their data. This left a total of 192 individuals in the final sample for a return rate of 37.4%. The investigators had previously run a power analysis with the following parameters: alpha = .05; effect size = medium (.30); and power = .80. This analysis indicated that investigators would need a minimum sample of 95. Therefore, the sample in the current study provided ample power.

The demographic makeup of the final sample was as follows. There were 124 females

(64.6%) and 68 males (35.4%). Participants ages ranged from 25-67 years with a mean of 44.84 and a median of 46.00. The ethnic breakdown of the sample indicated there were: 0 (0%) African Americans; 1 (.5%) Hispanic; 11 (5.7%) Native Americans, 2 (1%) designated “Other”; and 177 (92.2%) were White. They reported working in the following settings: 6 (3.1%) in private practice; 85 (44.3%) in community mental health centers; 9 (4.2%) in university counseling centers; 13 (6.8%) in hospitals; 51 (26.6%) in school settings; 8 (4.2%) in correctional settings; and 6 (3.1%) in “other” settings. Twenty-two (11.5%) reported having a secondary private practice to their primary work setting, but most (139, 72.4%) had no secondary practice setting. Their current length of employment in their primary work setting ranged from 1 month to 34 years with a mean of 9 years, 2 months; a standard deviation of 7 years, 10 months; and a median of 7 years. Their length of employment in the mental health care provider field ranged from 10 months to 46 years with a mean of 12 years, 10 months; a standard deviation of 8 years, 7 months; and a median of 10 years. Client/patient contact hours per week ranged from 0-60 with a mean of 24.12, a standard deviation of 10.66, and a median of 25.00.

The current study asked three questions. First, is there a relationship between social support and burnout in rural mental health counselors? Second, can the three different types of burnout be predicted from measures of social support? Finally, the investigators divided rural mental health counselors into three groups which were identified as being at either low, moderate or high risk of burnout, based on normative sample cutoff scores. They then sought to determine if there were group mean differences in measures of social support among these three groups?

The investigators used the Maslach Burnout Inventory (Maslach, Jackson, & Leiter, 1996) as a measure of burnout. The total sample means and standard deviations for the three scales were: Emotional Exhaustion, Mean = 23.32, SD = 11.31; Depersonalization, Mean = 7.41, SD = 5.45; and Low Sense of Personal Accomplishment, Mean = 7.35, SD = 4.99. The sample was broken down to determine risk levels using the normative sample cutoff scores. The results are presented in Table 1. Note that the current sample appeared to be at greater risk for emotional exhaustion, 37.5% in the high risk category, than either depersonalization, 17.7%, or having a low sense of personal accomplishment, 6.8%.

Table 1.

Prevalence of Risk Levels of Types of Burnout

Type of Burnout               Risk Level            Frequency       Percent            Cum. %

Emotional Exhaustion

Low                     59                    30.7                 30.7

                                        Moderate             61                    31.8                 62.5

                                        High                     72                    37.5                 100.0

Depersonalization

                                        Low                    97                    50.5                 50.5    

                                        Moderate            61                    31.8                 82.3

                                        High                   34                    17.7                 100.0

Low Sense of Personal Accomplishment

                                        Low                 119                  62.0                 62.0    

                                        Moderate          60                    31.3                 93.3    

                                        High                 13                    6.8                  100.0

The investigators used the Social Provisions Scales (Cutrona & Russell, 1987) to measure social support. Sample means, standard deviations, and prevalence data are presented in Table 2.

Table 2.

Sample and Normative Group Means and Standard Deviations, and Comparative Levels of Social Support

                                                        Current                          Normative

Scales                                      Mean            SD              Mean            SD            % > Mean *            % < Mean *

Guidance                                  13.07           2.15            14.18            2.23            72 (37.5%)            120 (67.5%)

Reassurance of  Worth              13.78           1.93            13.29            2.02            131 (68.2%)           61 (31.8%)

Social Integration                       12.80          1.99            14.01            1.09            47 (24.5%)            145 (75.5%)

Attachment                               12.50          2.13            13.72            2.42            53 (27.6%)            139 (72.4%)

Nurturance                               11.31           1.87            12.82            2.28            37 (19.3%)            155 (80.7%)

Reliable Alliance                        13.77          1.92            14.43            1.91             85 (44.3%)           107 (55.7%)

Total Social Support                  76.81          9.87            82.45            9.89            54 (28.1%)            138 (71.9%)

* “% > Mean” is the percent of the current sample that falls above the mean of the normative sample, “% < Mean” is the percent of the current sample that falls below the mean of the normative sample.

Note that the current sample appeared to be low in social support in the areas of Guidance, Social Integration, Attachment, Opportunity for Nurturance, and Total Social Support. The current sample appeared similar to the normative sample in Reliable Alliance, and to be high in social support in the area of Reassurance of Worth.

Pearson r statistics indicated a negative relationship between total social support and the three burnout measures. The correlations between the Total Social Support score and the burnout measures were as follows: Emotional Exhaustion, r = -.361, p< .001; Depersonalization, r = -.218, p< .002; and Low Sense of Personal Accomplishment, r = -.313, p< .001. The relationships between the burnout measures and the individual social support scales is presented in Table 3.

Table 3.

Intercorrelations Among Burnout and Social Support Scales

Scales

EE

  DP

LPA

GD

RW

SI

AT

ON

RA

EE

1.000

 

 

 

 

 

 

 

 

DP

540**  

1.000

 

 

 

 

 

 

 

LPA

.487**

.456**

1.000

 

 

 

 

 

 

GD

 -316**

-.166*

-.256**

1.000

 

 

 

 

 

RW

 -.321**

-.184* 

-.331**

.730**

1.000

 

 

 

 

SI

 -.392**

-.219**

-.344**

.810**

.715**

1.000

 

 

 

AT 

-.385** 

-.200**

-.327**

.830**

.711**

.796**

1.000

 

 

ON

.049

  .010 

-.050

.335**

.304**

.403**

.318**

1.000

 

RA 

-.354**

-.225**

-.286**

.830**

.704**

.755**

.777**

.289**

1.000

Note: EE = Emotional Exhaustion; DP = Depersonalization; LPA = Low Sense of Personal Achievement; GD = Guidance; RW = Reassurance of Worth; SI = Social Integration; AT = Attachment; ON = Opportunity for Nurturance; and RA = Reliable Alliance. * p< .05, ** p<.01.

The measures of burnout were all significantly positively correlated to each other. Also, the social support scales were all significantly positively correlated to each other. With the exception of Opportunity for Nurturance, which was uncorrelated to burnout, all five of the other social support scales were significantly negatively correlated with the three burnout measures. So, results indicated that, with the lone exception of Opportunity for Nurturance, as social support increased, burnout decreased and as social support decreased, burnout increased.

Can burnout then be predicted by social support? Obviously, it can. The investigators used a stepwise multiple regression analysis to test this hypothesis. The criterion for entry of a variable was set at p< .05 and for removal at p>.10. Results indicated that a two variable model was the best predictor of the Emotional Exhaustion variable, R2 = .201. The two variables were Social Integration and Opportunity for Nurturance. The F test to determine that the R2 of the two variable model was a significant increase over that of the one variable model yielded the following: F(1,190) = 11.840, p< .05. For Depersonalization a one variable model, using Reliable Alliance, was the best predictor, R2 = .151. The F test to determine that the R2 of the one variable model was a significant increase over that of the intercept model yielded the following: F(1,190) = 10.170, p< .05. Finally, for Low Sense of Personal Accomplishment, only Social Integration was a significant predictor, R2 = .121. The F test to determine that the R2 of the one variable model was a significant increase over that of the intercept model yielded the following: F(1,190) = 26.209, p< .05.

The three burnout variables were added together to form a total score. Investigators used this total score to divide participants into three groups according to their risk level of burnout. Investigators used group membership as the independent variable in a one-way analysis of variance (ANOVA) for each of the six social support variables. The researchers wanted to determine if these burnout risk level groups differed in their mean social support. The one-way ANOVA was followed by a Tukey’s HSD multiple comparison procedure to determine how the groups differed, when they did differ. An ANOVA summary table is presented in Table 4.

Table 4.

Summary of Analysis of Variance for Burnout Risk Groups for the Six Types of Social Support

Scale                                        SS                    df            MS                  F

Guidance

Between Group                        595.471           2         297.735          7.641 *

Within Group                            7364.196        189      38.964

Reassurance of Worth

Between Group                        609.362           2          304.681          9.992 *

Within Group                            763.304          189      30.494

Social Integration

Between Group                        1019.690        2          509.845          16.201 *

Within Group                            5947.789        189      31.470

Attachment  

Between Group                        609.616           2         304.808          6.193 *

Within Group                            9301.304        189      49.215

Opportunity for Nurturance

Between Group                          2.391             2         1.195              0.038

Within Group                            5980.276        189      31.642

Reliable Alliance

Between Group                        647.503           2         323.751       10.774 *

Within Group                           5679.367        189      30.050

              * p< .05

The three burnout risk groups differed in mean social support for five of the six social support variables. Only Opportunity for Nurturance was non-significant. The Tukey’s HSD results are presented in Table 5.

Table 5.

Means and Standard Deviations for the Six Social Support Measures Across Three Burnout Risk Levels and the Tukey’s HSD Multiple Comparison Analyses

Risk Levels                                       Low                                    Moderate                        High

SS Variables                            Mean              SD                Mean            SD            Mean           SD

Guidance                                  41.10 a          5.16            39.68 a           6.21         36.25 b        7.82

Reassurance of Worth              43.03 a          5.05            41.49 a           5.18          38.13 b        6.85

Social Integration                      40.71 a          5.05            38.57 a           4.99          34.35 b        7.21

Attachment                               39.62 a          5.57            37.90 a           5.85          34.70 b        7.32

Opportunity for Nurturance       33.97 a          6.13            33.80 a           5.37          34.08 a        5.24

Reliable Alliance                       43.44 a          4.63            40.94 b            5.39         38.45 c        6.85

Note: Means sharing the same letter were not significantly different, and means with different letters were significantly different in the Tukey’s HSD analysis.

For four of the social support variables: Guidance, Reassurance of Worth, Social Integration, and Attachment; results indicated a significant difference in mean social support between both the low and moderate risk groups and the high risk group. The low risk and moderate risk groups did not significantly differ. For Opportunity for Nurturance, there were no significant group mean differences. For Reliable Alliance, all three groups significantly differed from each other in their means.

DISCUSSION

Based on the findings of the current study, it is now evident that rural mental health counselors are at great risk for burnout and inadequate social support on the job. In fact, approximately 65% of the sample of rural mental health counselors scored at the moderate level or greater in burnout, and approximately 72% of the sample scored below the normative sample means for social support. Specifically, the present study was designed to investigate the relationships between social support systems and burnout among rural mental health counselors. The first objective of this study was to examine the relationship between the three dimensions of burnout and social support. As predicted, there is a statistically significant negative relationship between each of the three dimensions of burnout (i.e., emotional exhaustion, depersonalization, low sense of personal accomplishment) and total social support. Consequently, greater burnout was associated with diminishing social support.

The mean levels of burnout indicated that the current sample reported a greater degree of burnout than the normative sample of the MBI, except in relation to low sense of personal accomplishment. The present sample did not significantly differ from MBI norms in regard to personal accomplishment. The current sample was divided (according to the cut-off points listed on the scoring key of the MBI) into the three groups based on degree of risk for overall burnout, the results showed that 20.8% of this sample of rural mental health counselors reported a high level of burnout while 43.8% of the sample reported a moderate level of burnout. Also, in comparison to studies by Raquepaw and Miller (1989), Farber (1985), and Maslach et al., (1996) the present sample reported greater prevalence of emotional exhaustion, depersonalization, and less feelings of personal accomplishment.

In terms of prevalence of social support, an important finding in this study was that 71.9% of the sample of rural mental health counselors obtained scores below the normative sample means provided for the SPS (Cutrona & Russell, 1987). Even though this sample of rural mental health counselors reported an above average level of reassurance of worth, the current sample reported a below average level of guidance, social integration, attachment, opportunity for nurturance, and reliable alliance. These below average levels of social support may speak to the geographic and professional isolation commonly experienced by rural mental health counselors (Cohen, 1992; Murray & Keller, 1991; Richards & Gottfredson, 1978; Sladen & Mozdzierz, 1989; Wagonfeld & Buffum, 1983).            

The second objective of this study involved predicting types of burnout from measures of social support. The relation between the types of social support and burnout varied depending on the dimension of burnout. For emotional exhaustion, only social integration and opportunity for nurturance were statistically significant predictors. Therefore, one might conclude that rural mental health counselors, who may be isolated and do not have colleagues in which to share interests and concerns and who do not have mutually nurturing relationships, are at greater risk for emotional exhaustion.

Likewise, for depersonalization, only reliable alliance was a statistically significant predictor. That is, feelings of depersonalization were associated with lower levels of relationships in which persons can count on others for assistance. This suggests that rural mental health counselors who do not have others on whom they can rely on for assistance when needed are at greater risk for feelings of depersonalization.

Finally, for low sense of personal accomplishment, only social integration was a statistically significant predictor. Thus, rural mental health counselors with deficits in social support provided by a network of social relationships in which individuals share interests and concerns were at greater risk for experiencing a low sense of personal accomplishment.

The third objective involved the relationship between level of risk for burnout and types of social support. Lack of sufficient guidance, reassurance of worth, social integration, and attachment were associated with the rural mental health counselors at high risk for burnout. As one can see from the shortages in this sample of the various provisions of social support, it appears likely that these clinicians are not receiving adequate support on the job.

This study has made several contributions with regard to research and practice. The present study has been the first to test relationships between types of social support and burnout in rural mental health counselors. Additionally, Maslach (1993) has indicated that more theory-driven studies on burnout are needed that utilize a multidimensional model of burnout in contrast to the many of the early studies on burnout utilizing a unitary construct model.

In addition to contributing to theory-driven research, this study has addressed another problem related to research: investigators have acknowledged the paucity of data on burnout for specific work settings (Golembiewske, 1989; Leiter & Maslach, 1988; Ross et al., 1989). The present study has helped correct the deficit of information regarding specific work settings by documenting base rates of burnout and social support within a particular professional group, rural mental health counselors.

This study has made a contribution to the research on social support by addressing a major problem in the social support literature: failing to utilize instruments that assess most or all of the functions and behavioral categories of social support reviewed by Cohen and Wills (1985). This study has used a measure of social support that can differentiate the specific functions or types of social support.

Furthermore, this study has applications for clinical practice in the rural areas. Given the general findings for this sample of rural mental health counselors (i.e., 20.8% high level of burnout, 43.8% moderate level of burnout, and 71.9% below average level of social support), the increasing mental health needs of rural people, and the shortage of mental health providers practicing in the rural areas, it is imperative to know how to better train, recruit and retain clinicians who are considering practicing in the underserved rural areas. Maslach (1986), and Maslach and Leiter (1997) emphasize that social support is a key intervention to prevent as well as treat burnout.

It appears that the rural mental health counselors in this study are receiving adequate social support outside of work in terms of their family and friends, but they are experiencing a deficiency of coworker support. Each of the forms of social support provided by coworkers was below average except for reassurance of worth. Training programs and administrators in rural mental health need to be aware of the potential for limited support from coworkers, and they need to focus on strategies that address the development of the forms of social support reported below average in this study: guidance, social integration, attachment, opportunity for nurturance, and reliable alliance.

Maslach (1986) has recommended not only informal meetings over lunch and socializing after work but also more formal strategies such as staff meetings, conferences, retreats, and formal support groups led by a trained leader to facilitate the improvement of coworker relations. While some of these suggestions could be put into action to improve coworker relations, others may not be feasible for the rural mental health counselor. For example, the rural mental health counselor may not have access to the funding for such options as retreats or professionally led support groups. Similar to Maslach’s suggestions, Cherniss and Dantzig (1986) provide detailed information on the mutual aid group as another strategy for expanding social support. Mutual aid groups can foster social support among coworkers by providing information on how to deal with problems, material assistance, corrective feedback, reassurance and confirmation.

Future studies need to continue to provide theory-driven research. Even though it has been consistently demonstrated that social support and burnout are critical factors related to the well-being of the providers in the human services professions, there are still more details to uncover. A helpful direction for this end would be to continue to research Maslach’s multidimensional conceptualization of burnout, as well as multidimensional conceptualizations of social support. This approach will continue to advance the development of theory and application by providing a consistent frame of reference for comparison.

Furthermore, future research should consider longitudinal research designs. Schaufeli, Maslach, and Marek (1993) emphasize the importance of longitudinal research designs because it is only through this design that “causal relationships can be established between burnout and both its precipitating factors and its outcomes ... in addition to providing better evidence for causal relationships, longitudinal studies would yield valuable information about the development and successive phases of burnout” (p. 258). In addition to utilizing longitudinal designs, the use of path analysis in place of correlation may further reduce ambiguity of results by providing information on causal relationships. Moreover, the methodologies used to measure burnout need to be expanded to include measures other than self-report, such as job turnover, absenteeism, work performance, ratings by others (i.e., coworkers, supervisors, clients, family members, and friends), and physiological measures of health.

Future research also needs to consider replicating this study with a national sample of rural mental health counselors. A research project of this magnitude would not only help validate the results of this study, but also would ensure representation of the entire rural mental health counseling profession. Also, research on a national sample would provide information on base rates of burnout and social support for different rural regions of the country. Furthermore, with a national sample, it is likely that one would have enough subjects to complete statistics on questions related to differences in ethnicity and rural counselors in private practice vs. rural counselors in community agencies.

Finally, a primary challenge for counseling programs is to train effective rural mental health providers. This study has expanded the scarce empirical research on rural mental health issues by examining the relations between burnout and social support in rural mental health counselors. The results of this study, as seen above, do have important implications for intervention programs directed at rural mental health counselors by clarifying information related to base rates of burnout and social support, and the forms of social support predictive of burnout. The application of these findings will foster the prevention (or at least minimization) of burnout and promote the social support needed by rural mental health counselors. This improvement may, in turn, enhance both the recruitment and retention of the rural mental health counselor. These changes will hopefully have a beneficial effect on the quality of life of the rural citizen by improving availability and quality of mental health care to those underserved areas.

REFERENCES

Ackerley, G. K., Burnell, J., Holder, D.C., & Kurdek, L.A. (1988). Burnout among licensed psychologists. Professional Psychology: Research and Practice, 19(6), 624-631.

Berry, B., & Davis, A. E. (1978). Community mental health ideology: A problematic model for rural areas. American Journal of Orthopsychiatry, 48, 673-679.

Bodden, J. L. (1985). Review of Maslach Burnout Inventory. In J. V. Mitchell, Jr. (Ed.), Ninth Mental Measurements Yearbook: Vol. 1. Lincoln, NE: University of Nebraska Press.

Cherniss, C. (1980). Professional burnout in human service organizations. New York: Praeger.

Cherniss, C., & Dantzig, S. (1986). Preventing and managing job-related stress. In R. R. Kilburg, P. E. Nathan, R. W. Thoreson (Eds.), Professionals in distress: Issues, syndromes, and solutions in psychology. Washington D.C.: American Psychological Association

Cherniss, C., & Egnatios, E. (1978). Is there job satisfaction in community mental health. Community Mental Health Journal, 14(4), 309- 318.

Cohen, D. (1992). Occupational hazards of the rural psychotherapist. Psychotherapy in Private Practice, 10(3), 13-36.

Cohen, S., & Wills, T.A. (1985). Stress, social support, and the buffering hypothesis. Psychological Bulletin, 98(2), 310-357.

Corcoran, K. J. (1986). Measuring burnout: A reliability and convergent validity study. Journal of Social Behavior and Personality, 1(1), 107-112.

Cox, T., Kuk, G., & Leiter, M.P. (1993). Burnout, health, workstress, and organizational healthiness. In W. B. Schaufeli, C. Maslach, & T. Marek (Eds.), Professional burnout: Recent developments in theory and research (pp. 177-193). Washington D.C.: Taylor & Francis

Cutrona, C.E., & Russell, D. (1987). The provisions of social relationships and adaptation to stress. In W.H. Jones & D. Perlman (Eds.), Advances in personal relationships (Vol. 1, pp. 37-67). Greenwich, CT: JAI Press.

Daley, M. R. (1979). Burnout: Smoldering problems in protective services. Social Work, 24, 375-379.

Dunbar, E., McKelvey, J., & Armstrong, L. (1980). Characteristics of workplace and workers in rural mental health settings. Regional Human Resource Development Task Force, Washington D.C.: Department of Health, Education, and Welfare.

Dyer, J. (1997). Harvest of Rage: Why Oklahoma City Is Only the Beginning. Boulder, CO: Westview Press.

Edelwich, J., & Brodsky, A. (1980). Burnout: Stages of disillusionment in the helping professions. New York: Human Services Press.

Elkin, B., & Boyer, P.A. (1987). Practice skills and personal characteristics that facilitate practitioner retention in rural mental health settings. Journal of Rural Community Psychology, 8(2), 30-39.

Farber, B.A. (1983). Introduction: A critical perspective on burnout. In B. Farber (Ed.). Stress and burnout in the human service professions (pp. 1-22). Elmsford, NY: Pergamon Press.

Farber, B.A. (1985). Clinical psychologists' perceptions of psychotherapeutic work. The Clinical Psychologist, 38, 10-13.

Farber, B.A., & Heifetz, L.J. (1982). The process and dimensions of burnout in psychotherapists. Professional Psychology, 13(2), 293-301.

Flax, J. W., Wagonfeld, M. O., Ivens, R. E., & Weiss, R. J. (1979). Mental health and rural America: An overview and annotated bibliography. Washington, DC: U. S. Government Printing Office.

Forney, D.S., Wallace-Schutzman, F., & Wiggens, T. T. (1982). Burnout among career development professionals: Preliminary findings and implications. Personnel and Guidance Journal, 61, 435-439.

Freudenberger, H.J. (1975). The staff burnout syndrome in alternative institutions. Psychotherapy: Theory, Research and Practice, 12(1), 73-82.

Golembiewski, R. T. (1989). A note on Leiter’s study: Highlighting two models of burnout. Group and Organizational Studies, 14 (1), 5-13.

Golembiewski, R.T., Scherb, K., & Boudreau, A. (1993). Burnout in cross-national settings: Generic and model-specific perspectives. In W.B. Schaufeli, C. Maslach, & T. Marek (Eds.), Professional burnout: recent developments in theory and research (pp.217-234). Washington, DC: Taylor & Francis.

Hagihara, A., Tarumi, K., & Miller, A. S. (1998). Social support at work as a buffer of work stress-strain relationship: A signal detection approach. Stress Medicine, 14, 75-81.

Hargrove, D. S. (1982). The rural psychologist as generalist: A challenge for professional identity. Professional Psychology, 13(2), 302-308.

Hargrove, D. S. (1986). Ethical issues in rural mental health practice. Professional Psychology: Research and Practice, 17(1), 20-23.

Holahan, C. J., & Moos, R. H. (1982). Social support and adjustment: Predictive benefits of social climate indices. American Journal of Community Psychology, 10, 403-415.

Horst, E. A. (1989). Dual relationships between psychologists and clients in rural and urban areas. Journal of Rural Community Psychology, 10(2), 15-24.

House, J. S. (1981). Work stress and social support. Reading, MA: Addison-Wesley.

Human, J., & Wasem, C. (1991). Rural mental health in America. American Psychologist, 46(3), 232-239.

Jennings, F. L. (1992). Ethics of rural practice. Psychotherapy in Private Practice, 10(3), 85-104.

Kremer, B.J. & Owen, W.A. (1979). Stress in the life of the counselor. School Counselor, 26, 40-45.

Leiter, M. P. (1993). Burnout as a developmental process: considerations of models. In W.B. Schaufeli, C. Maslach, & T. Marek (Eds.), Professional burnout: recent developments in theory and research (pp.237-249). Washington, DC: Taylor & Francis.

Leiter, M. P., & Maslach, C. (1988). The impact of interpersonal environment on burnout and organizational commitment. Journal of Organizational Behavior, 9, 297-308.

Maslach, C. (1982). Burnout: The cost of caring. New York: Prentice Hall.

Maslach, C. (1986). Stress, burnout, and workaholism. In R. R. Kilburg, P. E. Nathan, R. & W. Thoreson (Eds.), Professionals in distress: Issues, syndromes, and solutions in psychology. Washington D.C.: American Psychological Association

Maslach, C. (1993). Burnout: A multidimensional perspective. In W.B. Schaufeli, C. Maslach, & T. Marek (Eds.), Professional burnout: recent developments in theory and research (pp.19-32). Washington, DC: Taylor & Francis.

Maslach, C., & Jackson, S. E. (1981). The measurement of experienced burnout. Journal of Occupational Behavior, 2, 99-113.

Maslach, C., Jackson, S. E., & Leiter, M. P. (1996). Maslach Burnout Inventory Manual Third Edition. Palo Alto, CA: Consulting Psychologists Press, Inc.

Maslach, C. & Leiter, M. P. (1997). The truth about burnout: how organizations cause personal stress and what to do about it. San Francisco, CA. Jossey-Bass Inc.

Maslach, C., & Schaufeli, W. B. (1993). Historical and conceptual development of burnout. In W. B. Schaufeli, C. Maslach, & T. Marek (Eds.), Professional burnout: recent developments in theory and research (pp.1-16). Washington, DC: Taylor & Francis.

McCann, I., & Pearlman, L. (1990). Vicarious traumatization: a framework for understanding the psychological effects of working with victims. Journal of Traumatic Stress, 3(1), 131-149.

Murray, J. D., & Keller, P. A. (1991). Psychology and rural America. American Psychologist, 46(3), 220-231.

Paulsen, J. (1988). Crisis, culture, and response. Journal of Rural Community Psychology, 9(1), 5-11.

Pines, A., Aronson, E., & Kafry, D. (1981). Burnout: From tedium to personal growth. New York: The Free Press.

Raquepaw, J. W., & Miller, R. S. (1989). Psycho-therapist burnout: A componential analysis. Professional Psychology: Research and Practice, 20, (1), 32-36

Richards, J. M., & Gottfredson, G. D. (1978). Geographic distribution of U.S. psychologists: A human ecological analysis. American Psychologist, 33, 1-9.

Rogers, E. R. (1987). Professional burnout: A review of a concept. The Clinical Supervisor, 5 (3), Fall, 91-106.

Ross, R.R., Altmaier, E.M., & Russell, D.W. (1989). Job stress, social support, and burnout among counseling center staff. Journal of Counseling Psychology, 36(4), 464-470.

Russell, D., Cutrona, C.E., Rose, J., & Yurko, K. (1984). Social and emotional loneliness: An examination of Weiss's typology of loneliness. Journal of Personality and Social Psychology, 6, 1313-1321.

Schank, J. A., & Skovholt, T. M. (1997). Dual-relationship dilemmas of rural and small-community psychologists. Professional Psychology: Research and Practice, 28(1), 44-49.

Warnath, C. F., & Shelton, J. L. (1976). The ultimate disappointment: The burned-out counselor. Personnel and Guidance Journal, 55, 172-175.

Schufeli, W. B., Maslach, C., & Marek, T. (1993). The future of burnout. In W. B. Schaufeli, C. Maslach, & T. Marek (Eds.), Professional burnout: recent developments in theory and research (pp.1-16). Washington, DC: Taylor & Francis.

Sladen, B. J., & Mozdzierz, G. J. (1989). Distribution of psychologists in underserved areas: Changes over time. Professional Psychology, 20, 244-247.

Vaux, A. (1988). Social support: theory, research, and intervention. New York: Prager.

Wagonfeld, M. O., & Buffum, E. W. (1983). Problems in, and prospects for, rural mental health services in the United States. International Journal of Mental Health, 12(1-2), 89-107.

Weiss, R. (1973). Loneliness: The experience of emotional and social isolation. Cambridge, MA: MIT Press.

Weiss, R. (1974). The provisions of social relationships. In Z. Rubin (Ed.), Doing unto others (pp. 17-26). Englewood Cliffs, NJ: Prentice-Hall.

Wilcox, B. (1981). Social support, life stress, and psychological adjustment: A test of the buffering hypothesis. American Journal of Community Psychology, 9, 371-386.

Contact the Author:

Alan Kee, Ph.D. Department of Psychology Amarillo College P. O. Box 447 Amarillo, TX 79178-0001 (806) 371-5183 (work) (806) 345-5589 (FAX) kee-ja@actx.edu

Don Johnson, Ph.D. Department of Psychology West Texas A&M University Canyon, TX

Portia Hunt, Ph.D. Department of Counseling Psychology Temple University Philadelphia, PA