Secondary interventions seem to be the most common type of SMIs (Dewe, 1994; Dewe and O’Driscoll, 2001). They aim to reduce the impact stressors exert on employees before they result in any serious health or other well-being problems. In the form of stress management training (SMT) such programmes are “individual-oriented, and seek to educate workers about the nature of stress, and to teach workers specific techniques for reducing physiological and psychological symptoms of stress, and fostering a state of relaxation” (Murphy, 1996/428).
Progressive muscle relaxation, and cognitive-behavioural skills training and to a lesser extent biofeedback and meditation, all of which (apart from meditation) derive from clinical psychology and counselling, are all common forms of SMT. Management skills training may also be included along with some training in time management or interpersonal skills.
Although there is evidence of therapeutic effectiveness associated with the formal use of cognitive behaviour therapy (CBT) in clinical settings to treat various psychiatric conditions and especially depression and anxiety disorders it should not be inferred that SMT techniques based on CBT are necessarily effective. The manner in which SMT is administered differs fundamentally from the ways CBT is clinically applied and, as a number of authors have noted (e. g.
, Murphy, 1996; Reynolds and Briner, 1996), the differences that obtain set SMT apart from other similar forms of intervention and necessitates a different approach to evaluation. Important differences relate to the characteristics of the client groups of CBT and SMT. As Reynolds and Briner (1996) remark, “typical SMT participants are a heterogeneous group of unselected, non-distressed, white-collar employees for whom the potential benefits of SMT can only be assessed in terms of prevention from later disorder” (pp. 147).
Murphy (1996) concludes his review of studies of SMT effectiveness as follows: “Stress management interventions have been generic in nature, not targeting specific work stressors or stress symptoms, and studies comparing the relative effectiveness of different training techniques have produced equivocal results” (p. 437). Additional problems include the absence of appropriate follow-up. In fact, “Where such follow-up has been done, the changes are typically not sustained and there is a regression to the baseline” (Newell, 2002/p.88).
Interestingly, clear evidence of any long-term impact on employee performance is almost totally nonexistent (Heron, McKeown, Tomenson and Teasdale, 1999; Jones and Bright, 2001; Reynolds, 2000; Reynolds and Briner, 1996). As Reynolds and Briner (1996) note: It is… extremely unlikely that the uniformly beneficial results which are promised implicitly and explicitly by occupational stress practitioners and researchers will ever occur in practice” (pp. 153-4).
But even for the shorter-term individual effects demonstrated the evidence is incomplete as it does not reveal the mechanisms that bring them about (Bond and Bunce, 2000). Finally, as many have commented most of the research has been and remains methodologically very weak (Murphy, 1988, Newell, 2002). Other limitations of SMT practice and research would include the lack of any systematic research on the role of individual differences in SMT or the absence of any serious concern with a proper needs assessment before SMT interventions are attempted (Murphy, 1996; Kompier and Kristensen, 2001; Briner and Reynolds, 1996).
In fact, rather than reflecting the nature and intensity of the problems they are supposed to be targeting SMT is offered in a packaged and pre-programmed format. As Kahn and Byosiere (1992) write: “The programs in stress management that are sold to companies show a suspicious pattern of variance; they differ more by practitioner than by company” (p. 623). Attention should also be drawn to another issue: It is not immediately apparent that the skills employees are requested to learn can indeed be learned in the appropriate manner.
In the few studies in the literature that report relevant data, a rather disheartening 30% of participants seem to fail to learn any of the techniques on offer (Murphy, 1984, 1996; see also Kompier and Kristensen, 2001). It is not immediately apparent how the problems identified above with respect to SMT could be resolved. Common proposals on how to proceed with a needs analysis or undertake a risk assessment, for instance, are not without their problems.
Many would suggest a stress audit (Cooper, Dewe and O’Driscoll, 2001). The problem here is that the instruments commonly used in stress audits do not usually reach acceptable levels of demonstrable predictive validity (Rick, Briner, Daniels, Perryman and Guppy, 2001). Stress-audits based on more qualitative methodologies (e. g. , focus groups, interviews), potentially more sensitive to the local context, suffer from problems of their own, especially relating to construct validity and reliability.