Research reportWhich presenteeism measures are more sensitive to depression and anxiety?
Introduction
Presenteeism is broadly understood to reflect the phenomenon of attending work when sick (Aronsson et al., 2000), or “working through illness” (McKevitt et al., 1997). Presenteeism has been estimated to account for a majority of the economic cost of lost productivity from depression (Collins et al., 2005), as much as 86% (Stewart et al., 2003). This hidden cost of mental disorders in the workplace adds an extra dimension to estimating the individual and societal burden of mental disorders, and the potential gains from effective intervention or prevention.
In contrast to absenteeism which is readily validated against administrative records (Ferrie et al., 2005), presenteeism is usually assessed by generic, self-report measures that are applicable to any job. Measures vary in complexity from single items assessing the number of days in a given period in which the person attended work when unwell (Aronsson et al., 2000), to adjusting time at work for perceptions of productivity in relation to self and/or colleagues (Brouwer et al., 1999, Kessler et al., 2003, Stewart et al., 2003, Wang et al., 2003), to domain-based measures that assess health-related limitations in specific job demands (Koopman et al., 2002, Lerner et al., 2004a).
It has recently been suggested that measuring presenteeism per se may be more important than the specific type of presenteeism measure used (Collins et al., 2005). To date, there is insufficient research to support the equivalency of different measures or to help researchers and practitioners choose a measure that is most appropriate for their purposes. Head-to-head comparative studies of more than two presenteeism measures are rare. Most studies have only investigated whether measures are sensitive to differences between those with and without symptoms, so little is known about comparative sensitivity to symptom severity and change over time. Most epidemiologic research on presenteeism has used a non-standard measure of mental health (Sanderson and Andrews, 2006).
The aims of the study were to compare four presenteeism measures on their detection of clinically meaningful differences in depression and anxiety status as measured by the Patient Health Questionnaire (Kroenke et al., 2001). Presenteeism measures included the number of days attending work unwell, the equivalent number of days after adjusting for self-perceived productivity, and two domain-based measures that were developed specifically to include the cognitive/social aspects of work which may be more sensitive to mental health symptoms: the Work Limitations Questionnaire (WLQ, Lerner et al., 2004a) and the Stanford Presenteeism Scale (SPS-6, Koopman et al., 2002). We compared measures on their sensitivity to presence versus absence of depression and anxiety, symptom severity, and change in depression and anxiety over 6 months follow-up. We hypothesized that the more detailed, domain-based measures (WLQ, SPS-6) would show greater sensitivity in all cross-sectional and prospective comparisons than the other presenteeism measures and absenteeism. There has also been one study suggesting the WLQ subscales may be able to discriminate at the level of different types of depressive symptoms (Lerner et al., 2004a). We extended this in a secondary analysis by examining whether there was a differential association between the nine individual symptoms of depression and interference in different types of job demands as measured by the WLQ. We hypothesized that physical symptoms would be associated more strongly with physical work demands and cognitive/affective symptoms with mental work demands. This study builds on previous research by investigating whether more detailed presenteeism measures offer any additional information over briefer alternatives, both for cross-sectional comparisons and a naturalistic observation of change over time, and by using a standardised measure of depression and anxiety in a routine workplace setting.
Section snippets
Design and study setting
The study represented a single-group, prospective design with measurement 6 months apart. The setting was call centres in the South-East region of Queensland, Australia, including the metropolitan centre of Brisbane. Call centres have been a common setting for development and evaluation of presenteeism measures. Data were collected by a self-report questionnaire of approximately 200 items (number of items varied due to internal skips).
Measures
Depression and anxiety: The Patient Health Questionnaire
Sample description
Of the 817 questionnaires distributed at baseline, 436 (53.4%) were returned. The baseline analysis sample (N = 432) included 30 persons (6.9%) with an imputed value on one dependent variable. The sample was predominately female and aged over 30 (see Table 1). A majority of the participants worked full-time and only on weekdays, and most had permanent employment contracts. Sixteen percent (N = 69) of the participants met criteria for any of the depression or anxiety syndromes.
Of the 436 baseline
Overview of study findings
Presenteeism is a relatively new construct used to describe the impact of health problems on productivity when attending work while sick. This study examined the sensitivity of four measures of presenteeism to depression and anxiety in a naturalistic, community-based setting. To our knowledge, this is the one of the first comparative studies of domain-based versus very brief (1 or 2 item) presenteeism measures, and the first to report sensitivity of presenteeism days, inefficiency days, and
Acknowledgements
This research was funded by Australian Rotary Health Research Fund Mental Illness Project Grants (2004, 2005). K. Sanderson was supported by a National Health and Medical Research Council Public Health (Australia) Fellowship (ID 290538), and J. Nicholson was supported by a National Health and Medical Research Council Career Development Award (ID 390136). The study procedures were approved by the Queensland University of Technology Human Ethics Committee.
References (25)
- et al.
Productivity losses without absence: measurement validation and empirical evidence
Health Policy
(1999) - et al.
“Choosing” to work when sick: workplace presenteeism
Soc. Sci. Med.
(2005) - et al.
Anxiety and depression in the workplace: effects on the individual and organisation (a focus group investigation)
J. Affect. Disord.
(2005) - et al.
Detecting panic disorder in medical and psychosomatic outpatients: comparative validation of the Hospital Anxiety and Depression Scale, the Patient Health Questionnaire, a screening question, and physicians' diagnosis
J. Psychosom. Res.
(2003) - et al.
Measuring depression outcome with a brief self-report instrument: sensitivity to change of the Patient Health Questionnaire (PHQ-9)
J. Affect. Disord.
(2004) - et al.
Comparative validity of three screening questionnaires for DSM-IV depressive disorders and physicians' diagnoses
J. Affect. Disord.
(2004) - et al.
Sickness presenteeism: prevalence, attendance-pressure factors, and an outline of a model for research
J. Occup. Environ. Med.
(2005) - et al.
Sick but yet at work. An empirical study of sickness presenteeism
J. Epidemiol. Community Health
(2000) - et al.
The assessment of chronic health conditions on work performance, absence, and total economic impact for employers
J. Occup. Environ. Med.
(2005) - et al.
A comparison of self-reported sickness absence with absences recorded in employers' registers: evidence from the Whitehall II study
Occup. Environ. Med.
(2005)