Burnout and salutogenic functioning of nurses

Opsomming The aim of this research was to determine the nature of the relationship between burnout and salutogenic function­ ing, specifically sense of coherence, hardiness and learned resourcefulness. The measurement was done with the Maslach Burnout Inventory and the Antonovsky, Kobasa and Rosenbaum questionnaires, administered on a sample of 105 registered general nurses. Correlations, exploratory and confirmatory factor analysis are reported. The results indicate a significant negative correlation and a model of good fit, confirming a structural relationship between burn­ out and salutogenic functioning in its conceptualisation and its measurement. The nature of these relationships are discussed and recommendations are formulated towards more effective individual, group and organisational coping with and preventing of burnout. Die doel van hierdie navorsing was om die aard van die verband tussen uitbranding en salutogeniese funksionering te bepaal, meer spesifiek die konstrukte sin vir koherensie, gehardheid en aangeleerde vindingrykheid. Die meting is gedoen met die M aslach uitbrandingskaal, en die Antonovsky, K obasa en Rosenbaum vraelyste, geadministreer op ‘n steekproef van 105 geregistreerde algemene verpleegkundiges. Korrelasies, ondersoekende en bevestigende faktoranalise word gerapporteer. Die resultate dui op ‘n beduidende negatiewe korrelasie en ‘n goeie passingsmodel, wat ‘n strukturele verband bevestig tussen uitbranding en salutogeniese funksionering in die konseptualisering sowel as die meting daarvan. Die aard van hierdie verbande word bespreek en aanbevelings word geformuleer ten opsigte van meer effektiewe individuele, groeps en organisasie-hantering en voorkoming van uitbranding.


62
Curatlonis May 2003 Vines (1991) recommended that research in nursing should not only focus on coping and coping methods to control burnout.Additionally, it should search for mediating variables such as self-esteem, motivation and personality.Since the 1980's the focus in the social sciences has in fact moved away from study ing stress and general coping behaviour (Badenhorst, 1997;Ngwezi, 1998) from an abnormal behavioural paradigm, towards studying specific personality coping constructs form the posi tive psychology (F red erick so n , 2001;S eligm an & C sikszentm ihalyi, 2000;Sheldon & King, 2001) and salutogenesis paradigms (Antonovsky, 1979;Breed, 1997), in cluding psycho-fortology as a field of study (Coetzee & Cilliers, 2001).Recently burnout is studied in terms of various so-called salutogenic coping constructs such as sense of coherence, internationally (Palsson, Hallberg, Norberg & Bjorvell, 1996) as well as locally (Levert, Lucas & Ortlepp, 2000).The most popular construct used in studying burnout in nursing, is har diness (see Boyle, Grap, Younger & Thomby, 1991;Collins, 1996;De Pew, Gordon, Yoder & Goodwin, 1999;Marsh, Beard & Adams, 1999;Simoni & Paterson, 1997;Sims, 2000;Topf, 1989).A limiting factor in all of the above studies is that only one salutogenic construct is used to explain coping with burn out.According to Antonovsky (1979) andStrumpfer (1990;1995) there are many such behavioural constructs acting as mediating variables explaining coping behaviour (such as selfefficacy, locus of control, resilience and happiness).Only one research project (De Wet, 1999 -a qualitative study, N=23) could be traced using a combination of salutogenic constructs, namely sense of coherence, hardiness and learned resource fulness.This choice was based on the suggestion by Rich (1991) and Sullivan (1989) that these three constructs may be the most relevant in the nursing field.From the above it seems that individual salutogenic constructs are relevant in understanding coping with burnout amongst nurses.What is not clear is how a combination of these con structs relate statistically to burnout and what the underlying factor structure of these measures are.

Burnout
Burnout (BO) is described as a persistent, negative, work-re lated state of mind (or syndrome) developing gradually over time in originally highly motivated, striving, achieving and non compromising individuals with good intentions and high ex pectations (which are sometimes out of touch with reality), who stretch themselves beyond the normal work boundaries for a long period of time in their quest for meaning.The indi vidual then develops an array of physical, psychological and attitudinal symptoms, primarily emotional exhaustion, accom panied by distress, depersonalisation, a sense of reduced ef fectiveness, decreased motivation and the development of dys functional personal and societal attitudes and behaviours at work (Chemiss, 1995;Golembiewski & Munzenridder, 1998;Jackson, 1982;Maslach, 1976;1982;1993;Maslach& Jackson, 1982b;1984;Pines & Aronson, 1981;1988;Schaufeli & Enzmann, 1998).BO is not the same as depression (Maslach & Schaufeli, 1993) or stress (Pines, 1993;Schaufeli & Enzmann, 1998).
The symptoms of BO can be categorised as follows (Chemiss, 1980;1995;Golembiewski & Munzenridder, 1998;Jackson, 1982;Maslach, 1976;1982;Maslach & Jackson, 1982b;1984;Pines & Aronson, 1981;1988;Schaufeli & Enzmann, 1998)  emotional exhaustion; males higher on depersonalisation pos sibly because of gender role stereotypes), marital status (higher amongst unmarried men), less education, less work experience and work load (due to reality shock, an identity crises due to unsuccessful occupational socialisation or a selection or sur vival bias).3. BO correlates with personality constructs such as type-A behaviour, neuroticism, high and unrealistic expectations and external control, and work related constructs such as over load, role conflict, role ambiguity, poor collegial support, lack of feedback in decision making and autonomy.
2. BO correlates with young age, gender (females higher on 4. BO has consequences for the individual nurse, the patient as well as for the larger institution.It leads to deterioration in the quality of service rendered, personal dysfunction and leads to personnel issues such as staff turnover, absenteeism and low morale.Dubrin (1990) even suggested that the bumt-out manager spreads it to subordinates.5.In American hospitals, up to 70 percent of staff nurses re sign from their jobs during a typical year because of BO symp toms.
6. BO can be countered by a confronting coping style, high self-esteem and extroversion.7. Coping with BO lies in the individual's predisposition (Semmer, 1996) described as having positive beliefs about the world, realising that he/she has possibilities in dealing with it, to perceive events and circumstances as stressful and to have ways of coping with them and to deal with failure in coping.
BO is discussed in the literature as impacting on the individual, interpersonal and organisational levels.
1.The individual has the responsibility to recognise the signs and symptoms of BO (Muldary, 1983;Pines, 1993).Individual coping is described as an intrapersonal and action-oriented effort to manage the environmental and internal demands and conflicts, through awareness, understanding and taking re sponsibility for action (Lazarus, 1974;Lazarus&Launier, 1978;Pines & Aronson, 1981).2. Interpersonal coping strategies refer to having and using social support systems defined as networks of occupational relationships, which could comprise one or more of the follow ing: emotional support (admiration, respect, liking), affirmation or appraisal (acknowledgment of the appropriate behaviour of another), and aid (direct giving of materials, information or serv ice) (Morano, 1993;Pines & Aronson, 1988).

Salutogenic functioning
The salutogenic paradigm (Antonovsky, 1979) focusses on the origins of health and well-ness (Latin salus = health / Greek genesis = origins), the location and development of personal and social resources and adaptive tendencies which relate to the individual's disposition, allowing him/her to select appro priate strategies to deal with confronting stressors.For the purpose of this research, the salutogenic constructs sense of coherence, hardiness and learned resourcefulness were cho sen.
1. Sense of coherence (SOC).Antonovsky (1984;1987) defines SOC as a global orientation that expresses the extent to which the individual has a pervasive, enduring, though dynamic feel ing of coherence, that the stimuU deriving from his/her internal and external environments in the course of living are struc tured, predictable, and explicable, that the resources are avail- (2) control (a proclivity to make the individual feel and act as if he/she is influential in the face of the varied contingencies of self-control skills by which the individual self-regulates inter nal responses that interfere with the smooth execution of an ongoing behaviour.It consists of specific behavioural skills namely the ability to choose and implement effective problem solving skills, the ability to use cognitive skills such as self talk to control internal processes, the ability to delay the gratifica tion of needs and the tendency to evaluate the self as efficient and effective in situations (Rosenbaum, 1988).
The salutogenic personality profile incorporates the following behaviour (Viviers & Cilliers, 1999); On the cognitive level, the individual is able to view stimuli from the environment in a positive and constructive manner, and to use the information towards effective decision making.On the affective level, the individual functions with self-awareness, is confident, self-fulfilled, views stimuli as meaningful and feels committed towards life in a mature manner.On the motivational level, the indi vidual is driven from within, perceives stimuli as a challenge which directs his/her energy to cope, solve problems and achieve results.The interpersonal characteristics entail the capacity to form meaningful relationships with others within a support system at work and in society.

66
Curationis Curationis 2003 HAR (Kobasa, 1982;Lambert & Lambert, 1987) is seen as an inherent health promoting factor with a direct relevance to nurs ing practice where HAR can be taught to help nurses increase their tolerance to stress, to screen nurses who might be ex posed to high stress in the work environment, and to aid in preventing stress-related illnesses.This finding has been con firmed for ICU nurses (Consolvo, Brownewell & Distefano, 1989;Manning, Williams & Wolfe, 1988;Rummel, 1991;Taylor & Cooper, 1989).According to Rosenberg (1990) HAR pre dicts 34% of the variance in nurses' lifestyle.Commitment as a

Method
The sample An sample of convenience (Anastasi, 1990) was used, consist ing of 105 registered general nurses from various large hospi tals in Gauteng Province.Each had a three year nursing di ploma and at least 5 years nursing experience.Only females were included with ages ranging between 28 and 57 years.There were 73 white and 32 black nurses.All were involved in general nursing and worked full-time.dimension and years employed have the largest beta weights and are the most predictive of a healthy lifestyle (Boyle et al, 1991;Dermatis, 1989;Gillmore, 1990).

Theoretical statement and research question
The central theoretical statement of this research can be for mulated as follows: The individual nurse who functions on high levels of BO (with its mentioned symptoms) will function on low levels of salutogenic functioning (SOC, HAR and LR) and vice versa.The research question being investigated is whether a meaningful negative relationship between BO and salutogenic functioning as measured by these three constructs, exist.This knowledge could help in formulating future indi vidual, group and organisational coping strategies for nurses.

Aim and research design
The aim of this research was to determine whether a meaning ful relationship exist between BO and salutogenic functioning, and to ascertain the nature thereof.A survey design with quan titative measurement of and statistical analyses on the four constructs (SOC, HAR and LR) was used.

iVleasurement instruments
The following four measuring instruments were chosen be cause of their (1) conceptual congruence to the above defini tions of the constructs and (2) acceptable psychometric char acteristics provided in the literature.
1.The Maslach Burnout Inventory (MBl) (Maslach & Jackson, 1981;1982a;1986), measures (BO) in three sub-scales namely emotional exhaustion (a reduction in emotional resources, feel ing drained, used up and physically fatigued), depersonalisa tion (an increase in negative, cynical and insensitive attitudes towards colleagues, clients and/or patients) and personal ac complishment (a feeling of being unable to meet the other's needs and to satisfy essential elements of job performance).Maslach and Jackson (1986) as well as Schaufeli and Janczur (1994) offer factor-analytical support for the usage of the sub scales separately.Maslach and Jackson (1981;1986) as well as Lahoz and Mason (1989) Dana, Hoffman, Amstrong and Wilson (1985) on concurrent validity and Payne (1982) on its construct validity.3. The Personal Views Survey (Kobasa, 1979) measures HAR in a total score as well as in three sub scores namely commit ment, control and challenge, as defined above.Parkes and Randall (1988) report a reliability coefficient of 0.78, Manning et al (1988) report a range between 0.75 and 0.90 and Funk (1992) a test-retest coefficient of 0.60 after two weeks.Internal consistency ranged from 0.68 to 0.89 in several studies reported by Maddi and Khoshaba (1994).Concurrent validity is reported by Bartone (1989), Campbell, Amerikaner, Swank andVincent (1989) and Parkes and Randall (1988).Maddi and Khoshaba (1994) report that factor analysis confirmed the three constructs of HAR as clearly identifiable factors.4. The Self-control Schedule (Rosenbaum, 1980) measures LR in a single score as defined above.Rosenbaum (1980), Leon and Rosenthal (1984) report test-retest reliability of 0.86 after four weeks and 0.77 after eleven months.Redden, Tucker and Young (1983) report the internal consistency to be 0. 82 and Rosenbaum (1988) between 0.78 and 0.91.Leon and Rosenthal (1984) and Rosenbaum and Ben Ari (1985) report on the fa vourable convergent and discriminant validity.Rosenbaum (1988) reports that factor analysis confirmed three factors namely problem-focussed coping, mood / pain control and ex ternality.
model in this research, the data representing the variables were factor analysed.The orthogonal transformation matrix rotated factor pattern method was used to determine the factor struc ture of the variables.The retention of the factors is based on certain rule of thumb principles.For principal-components analysis, it can be argued that the Kaiser criterion of retaining factors, with eigenvalues greater than one, appears to be the most appropriate (Ford, MacCallum & Tait, 1986).Because not all statisticians agree (for example Floyd & Widaman, 1995), a commonly used rule for specifying factors was used, namely that only variables with loadings greater than 0.40 on a factor should be considered significant and used in defining a factor (Comrey, 1978).

Confirmatory factor analysis.
Because exploratory factor analysis has limited value for the specification and testing of an hypothesis relating to model structure, confirmatory factor analysis was done, allowing the researcher to specify the hypotheses and providing informa tion to determine whether the observed data confirm the hy pothesised model structure.SEPATH in Statistica (1999) was used to specify and analyse such models and thereby validate the data.It measures the fit of the hypothetical model to the data (goodness-of-fit statistics), measures and tests specific elements of the model, such as structural parameters (Hughes, Price & Marrs, 1986;MacCallum, 1998).

Data gathering
The staff offices of the various hospitals were contacted and informed about the research and its objective, and asked to have access to interested staff members.After appointments were made, the sample attended the psychometric sessions in groups of up to 15, lasting about 90 minutes, with the four measuring instruments administered in sequence.In terms of ethical considerations, the following can be stated: all nurses participated voluntarily, their results were treated confiden tially and individual feedback was promised and given to those individuals who were interested in receiving it.This was done by the author (a psychologist) who took special care to ensure that no-one was hurt or left the research situation with unfin ished emotional business.

Data processing
The data was analysed by means of the Statistica (1999) StatSoft package and the following statistics (Nunnally & Bernstein, 1994) are reported:

Descriptive statistics.
Reliability of the instruments -Cronbach alpha coeffi -cients.Clark and Watson (1995) suggested that a Cronbach alpha of between 0.5 and 0.6 is satisfactory for research purposes.Inter-correlations -Pearson-product moment correla -tion coefficients.

Expioratory factor analysis.
The research measured 12 observed variables through the four measuring instruments.As a basis for establishing (close up) a The following two statistical hypotheses are tested: H 1 A meaningful relationship exist between BO (negative for emotional exhaustion and depersonahsation, and positive for personal accomplishment) and salutogenic functioning with its nine dimensions H2 A good fit exists between the theoretical structure of the two constructs and the empirical data

Descriptive statistics
The descriptive statistics are reported in Table 1.

Reliability of instruments
The Cronbach alphas and total dimensions of the four instru ments were as follows, suggesting that the data were rehable in that respondents tended to answer in a consistent manner: BO: 0.75 / 3 dimensions; SOC: 0.85 / 3 dimensions and the total score; HAR: 0.85 / 3 dimensions and the total score; LR: 0.80 one dimension.

inter-correlations
An overall significant negative correlation exists between BO and salutogenic functioning (r = -0.35,p < 0.001).The signifi cant correlations between the various dimensions of BO and salutogenic functioning are shown in Table 2. Emotional ex haustion and depersonalisation both correlate negatively with all the salutogenic dimensions of SOC, HAR and LR.Personal accomplishment correlates positively to all the salutogenic di mensions.

Exploratory factor analysis
Table 3 shows the 2-factor exploratory factor analysis model (in bold).The model explains the relationship between the di mensions.It separates the BO construct form the salutogenic functioning construct.Thus, a 2-factor model is established namely Factor 1 -Salutogenic functioning and Factor 2 -Burn out.

Confirmatory factor analysis
Table 4 shows the parameter estimates and structural relation ships for the 2-factor model.All of the parameter estimates are significant.The model shows a reasonable fit with a Steiger-Lind RMSEA index of 0.15.According to MacCallum (1998), a good fit = <0.10.
Table 5 shows the Steiger-Lind RMSEA index of fit for each of the measuring instruments.The findings indicate that the in struments are valid for the purpose of which they are being used.Thus, the empirical data fit the theoretical model.

Discussion
The results are discussed in terms of the two hypothesis.

Hypothesis 1
This hypothesis could not be rejected because of the overall and dimensional negatively significant relationships between the constructs BO and salutogenic functioning.This supports previous research findings in terms of SOC (Basson & Rothmann, 2001 with pharmacists;Levert, Lucas & Ortlepp, 2000 with nurses), and HAR (Bonalumi & Fisher, 1999;Boyle et al, 1991;Constantini, Solano, Di-Napoli & Bosco, 1997;Kennedy, 1999;Marsh et al, 1999;Sims, 2(XX)).Similar results are reported for HAR when BO was measured by the Pines Burnout Scale (Collins, 1996;Simoni & Paterson, 1997).BO as a behavioural phenomenon can be seen as the opposite of or in contrast to salutogenic functioning in the following manner: 1.On the physical level, the BO symptoms of distress, illness, extreme physical and psychosomatic reactions, are contrasted by the salutogenic functioning individual's strong resistance resources, including a healthy immune system to fight off ill ness.2. On the cognitive level, the poor performance, concentration, decision making and the making of mistakes, are contrasted by the salutogenic functioning individual's strength in under standing of and making sense out of the demands of his/her environment in a positive, realistic, constructive and truthful way.This leads to effective decision making, problem solving and achievement of results.3. On the affective level, the emotional exhaustion and lack of energy is contrasted by an emotional identification with mean ingful stimuli and life events.The negative self-concept and feelings of helpless-/ hopeless-/ powerlessness, depersonali sation and the lack of individual distinctiveness, are contrasted by a mature and realistic sense of self, characterised by self and emotional awareness, fulfilment and confidence, feeling in charge, influential and optimistic with a strong belief in own worth.4. On the motivational level, the external locus of control is contrasted by an internal locus.The lack of initiative, enthusi asm and interest is contrasted by a commitment towards de mands and challenges which are experienced as manageable.The low frustration tolerance and impulse control are con trasted by the capacity to delay own need gratification -the individual experiences his/her own sphere of influence vividly and feels in control of inner responses, experiences and choices. 5. On the behavioural level, the impulsiveness, procrastina tion, doubt and lack of focus is contrasted by stress resistance resources.The individual experiences life as coherent ("to gether"), predictable, explicable and see change as a challenge and a growth enriching opportunity.The dependancy on stimu lants from outside is contrasted by internal balance and ten sion management.6.On the interpersonal level, the disinterest, isolation, with drawal, indifference, hostility and suspicion are contrasted by establishing, working hard on and being committed to mean ingful relationships with significant others, within a support system. 7.In terms of work, the resistance to get involved, the reduced performance, effectiveness, productivity and job satisfaction, is contrasted by an experience of strength, trying to make the job and its tasks predictable and manageable, the individual is willing to participate fully and a has a sense of accomplish ment.

Hypothesis 2
This hypothesis could not be rejected because of the estab lished model linking the constructs through confirmatory fac tor analysis.This means that the operationalisation of the con structs in the four used instruments, corresponds with the conceptuahsation of the constructs in the hterature of BO and salutogenic functioning.This result confirms that these in struments can be used effectively in nursing research to meas ure BO, SOC, HAR and LR as well as to address the effects of BO from a salutogenic perspective.

Conclusion and recommendations
Firstly, this study confirmed that the measurement of the con structs fits the theoretical models of BO and salutogenic func tioning.
Secondly, this study confirmed that BO represents the oppo site of salutogenic functioning.The individual nurse who func tions on high levels of BO (with its mentioned symptoms) will function on low levels of salutogenic functioning (SOC, HAR and LR) and the individual nurse functioning on high levels of the salutogenic constructs, will be able to counter the effects ofBO.It is recommended that BO symptoms are addressed on three levels of the individual, the group and the organization: Muldary (1983) and Pines (1993) suggest that the in dividual takes responsibility to recognise the signs and symp toms of BO through awareness, understanding and action.According to Rosenbaum (1988) this refers to the implementa 69 Curationis May 2003 tion of effective problem solving skills, using cognitive skills, such as self talk, to control internal processes, to delay the gratification of needs and to evaluate the self as efficient and effective.This may be difficult for the individual who's general resistance resources are already lessened by prolonged BO and it may even be impossible for the individual who denies these symptoms.It is hypothesised here that the individual will only succeed with support from others giving feedback and support, providing opportunities to become aware of own issues.

•
The above activities must rather be implemented by the individual in interaction with significant others, acting as a social support system."Self talk with others" (Hawkins & Welsh, 1999) or a facilitated growth group experience (Cilliers & Terblanche, 2000) is suggested consisting of emotional sup port, affirmation and behavioural feedback.This climate of unconditional acceptance seems to be the best way to combat BO.

•
Hospital administration officers in co-operation with Industrial Psychologists should regularly identify BO amongst its staff -quantitatively by means of questionnaires and quali tatively by means of focus groups, being observant and show ing interest.This data should be used to plan and present organisational development inputs designed specifically for countering BO from a salutogenic perspective.
Lastly, it is recommended that future research on BO from a salutogenic perspective, includes more constructs in combi nation, as described in the positive psychology paradigm (see Snyder & Lopez, 2002).Examples of these are resilience, emo tional intelligence and happiness.

Table 3 : The 2-factor exploratory factor analysis model
BOLD indicates significanceThe literature suggests that coping with BO is the responsibil ity of the individual nurse.Although institutional inputs could be helpful, his/her own inner strength forms the basis to over come the symptoms effectively.This strength is next concep tualised from the salutogenic paradigm.abletomeetthe demands posed by these stimuli, and that these demands are challenges worthy of investment and en gagement, The SOC predicts the extent to which the individual feels that there is a probability that things will work out well(Antonovsky, 1979).It consists of three core personality char

Table 4 Parameter estimates and structural relationships for the two-factor model
esting incentives to growth rather than threats to security) versus threat to security.The hardy personality uses an opti mistic and cognitive appraisal of events which will determine the subsequent actions directed towards those events (Man ning,Williams & Wolfe, 1988).3.Leamed resourcefuhiess (LR).Rosenbaum (1983; Rosenbaum  & Ben-Ari, 1985)defines LR as an acquired personality reper toire existing of a set of behaviours and mostly cognitive