Behavioral Health Research Center of the Southwest, 612 Encino Place NE, Albuquerque, NM 87102, USA
Received 8 April 2002; in revised form 15 October 2002; accepted 4 November 2002
![]() |
ABSTRACT |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
![]() |
INTRODUCTION |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Despite this demonstrated success, behavioural health issues, such as substance misuse, have received very little attention as a component of health promotion programmes, in part because prevention programmes lack empirical evidence. In a review of published studies of health promotion programmes at the worksite, Heaney and Goetzel (1997) found only two studies that addressed alcohol, both of which were judged inconclusive in support of the intervention. Another review of worksite interventions for alcohol (Roman and Blum, 1996
) found support for employee assistance programme (EAP)-based rehabilitation of employees with severe alcohol problems, but little evidence of any research focused on prevention or early intervention. Subsequently, two additional reports have contributed more uncertainty about the effectiveness of reducing excessive drinking through a worksite wellness programme. Heirich and Sieck (2000)
demonstrated the effectiveness of a proactive cardiovascular worksite programme in reducing risky drinking of auto workers, whereas Richmond et al.(2000)
found little effect of a similar lifestyle campaign in changing alcohol habits of postal workers in Australia. Both programmes were innovative in that they placed alcohol use messages within broader concepts of wellness and health.
Substance misuse problems cost companies billions of dollars in absenteeism (Lehman and Simpson, 1992), turnover (Lehman and Simpson, 1990
), injury rates (Alleyne et al., 1991
), and increased medical care costs (Sheridan and Winkler, 1989
). For example, it has been estimated that for each employee who misuses alcohol or drugs, companies lose $7261 per year (U.S. Bureau of Labor Statistics, 1989
). This underscores a need to further explore and test interventions in the workplace. To address this gap in knowledge regarding the impact of substance misuse prevention on workplace health, the Substance Abuse and Mental Health Services Administrations Center for Substance Abuse Prevention funded nine partnerships between managed care providers and employer subscribers, with a goal of improving existing substance misuse programmes and assessing their impact.
This report is the first of two publications presenting results of research supported by one of these partnerships a prevention programme for health care workers. The project objective was to determine whether enhancements to the existing employee wellness programme (EWP) and EAP reduced harmful or hazardous alcohol use and improved key indicators of employee health. The programme employed features suggested by earlier research to improve outcomes: health risk appraisals (HRAs) in combination with health promotion campaigns, targeting higher-risk individuals, and integrating alcohol messages within the broader context of overall wellness.
![]() |
METHODS |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
To examine similarities between the groups, the intervention and comparison sites were compared on gender, age and job classifications (Table 1). The intervention site employed a slightly higher number of men than the comparison site and consisted of a somewhat younger population. The comparison site was also relatively over-represented in the other job class (Table 1
), which consisted of sales, production/repair, machine operators/assemblers, transportation/moving, unskilled work and service. Thus, the other category generally refers to manual or blue-collar labour professions. The intervention site had greater representation of administrative, administrative support and professional/technical staff.
|
The EWP, administered by a health educator, offered services designed to increase awareness of potential health problems and to improve mental and physical well-being. Programme components included completion of the HRA, developed by HealthFirst of Albuquerque, NM, health promotion classes, subsidized health club memberships and participation incentives. Substance misuse prevention consisted of a self-care manual, with a chapter on alcohol and drug problems, which was distributed to all new employees. Yearly health fairs also provided information on multiple health and wellness issues and solicited EWP enrolment and HRA completion.
Project WISE was introduced to the intervention site on 1 September 1998, and it augmented services provided by EAP and EWP. New elements added by Project WISE included substance misuse awareness training for management staff, mailed substance misuse prevention information to those completing the HRA, brief motivational counselling for employees and family members, and a campaign designed to increase employee awareness of substance misuse as a health problem. The brief counselling component was under-utilized, with fewer than 50 employees receiving personalized counselling. Details of the Project WISE intervention have been published (Lapham et al., 2000).
The EAP counsellor and the hospitals medical review officer conducted single 2-h training sessions for supervisors and managers. Topics centred on substance misuse awareness and procedures for intercession and referral of employees. Sixty per cent (n = 69) of supervisors at the intervention site were trained. Those who did not attend were made aware of a self-instructional packet that was maintained by the staff education department. During the 2-year intervention period, 189 employees at the intervention site, identified as moderate- or high-risk drinkers by their HRA scores, received a free self-help manual, Patient Workbook for Quitting or Cutting Down (Sanchez-Craig, 1994). Low risk drinking was defined as drinking <7 days/week and no more than 12 drinks per session (see individual items below). Borderline to moderate risk was defined as drinking up to 34 drinks per occasion or drinking every day. High-risk drinking was defined as usual or occasional drinking of 5 or more drinks per occasion. Binge drinking was defined as drinking 5 or more drinks on at least one occasion in the past 30 days. This definition is from the Substance Abuse and Mental Health Services Administration: National Household Survey on Drug Abuse: Main Findings 1991 (SAMHSA, 1993
). Binge drinkers, along with employees who had risks associated with poor nutrition, inactivity or stress, were also mailed videotapes designed to educate them about healthy alternatives to alcohol and drug use for stress reduction and coping (Cook and Back, 1990
; Connection Series, 1996
), and were invited to attend a free health coaching session to discuss lifestyle modification.
In addition, Project WISE provided educational information about substance misuse to all employees. One-page informational flyers covering eight topics about hazardous alcohol use and drinking were developed and dispersed to all workers through the employee mail system. Furthermore, promotional events were held to raise awareness, including prize lotteries for completing an HRA. A week-long media campaign was held for Alcohol Awareness Month, and there were displays at the annual employee health fair. Videotapes and literature were placed in the 31 staff lounges and freely given by request at the intervention site. At the conclusion of the intervention, approximately 450 videos, 16 000 flyers and 450 workbooks had been distributed.
The HRA was used throughout the study at both sites. The self-report instrument contained sections on heart disease, stress, depression, diabetes, nutrition, cancer, safety and preventive health care risks. Enrolees were also weighed, their heights measured, and blood drawn for measuring cholesterol levels. Data were entered by a member of staff, who generated and mailed a feedback report to the employee that itemized personal risks and recommendations for improving risk. Participants received points, which could be exchanged for prizes and other incentives. The HRA used in the year prior to the intervention start date did not include questions on alcohol use, so the programme was customized to incorporate four questions on alcohol use (Lapham et al., 2000). Alcohol use questions from the HRA included the following: (1) on average, how often do you drink beer, wine, liquor, or other beverage containing alcohol?; (2) on days when you drink, how many drinks do you usually have? (1 drink = 12 oz. beer, 4 oz. glass of wine, or shot of liquor); (3) during the past 30 days, on how many days did you have 5 or more drinks on the same occasion? (for example, if a person had 2 beers with lunch and 3 glasses of wine with dinner, that would be considered 2 drinks on one occasion and 3 drinks on another; this would not be counted as 5 or more drinks per occasion, even though that individual had 5 drinks throughout the day); (4) in the next 6 months or so, do you want to reduce the amount of alcohol that you drink? The HRA did not contain questions on illicit drug use.
Study design
The impact of Project WISE was assessed by comparing HRA changes over time (Pre and Post) and between groups (Intervention and Comparison). Pre-intervention data were compiled from an 8-month period (January 1998 to August 1998) prior to the start date, when the revised HRA included alcohol items. Post-intervention data were gathered from a 22-month period (September 1998 to July 2000).
One goal of Project WISE was to reduce and prevent binge drinking, measured as the number of days a respondent reported 5 or more drinks per occasion in the past 30 days. Motivation to reduce drinking was captured by question (4) above. Since alcohol use may be related to level of stress and depression, scores on these measures were considered in the analysis. Composite stress scores were derived from responses to 19 questions addressing agreement with feeling calm, security, problem solving ability, sense of happiness and joy, sense of fear and worry, and the frequency of somatic conditions such as headaches, light-headedness, stomach pains and sweatiness. Depression was similarly measured by 20 questions.
The Human Resources database (worksite, age, gender and job classification) and the EWP database, which housed HRA results, were the two data sources for the analysis. One thousand and sixty employees completed 1391 revised HRAs (which included the questions on alcohol use) at both sites. Each of these employees completed between one and four individual HRAs (mean HRAs completed per employee = 1.3) (Table 2).
|
The HRA was considered the unit of analysis, and since some employees completed more than one HRA, modelling required special statistical techniques. Each repeated HRA outcome was unlikely to be independent and identically distributed from the binomial error distribution assumed in the logistic regression model. Omitting all but one HRA result for each respondent can mitigate problems of applying standard statistical procedures to dependent data. However, we wished to take advantage of the large data set provided in this study without applying arbitrary exclusion criteria for multiple HRAs taken by each respondent. Therefore, generalized estimating equations (GEEs) were used for fitting repeated measures models to each outcome measure in this analysis (Liang and Zeger, 1986). These estimation procedures provided unbiased parameter estimates and confidence intervals with repeated measures data and allowed us to use all available data for estimating the intervention effects on drinking behaviour. SAS procedure GENMOD was utilized for these analyses (Release 8.02, SAS Institute Inc., Cary, NC). Both analyses were evaluated at an alpha level of P
0.05.
A high level of job stress, as well as feelings of sadness and depression, can increase alcohol consumption (Graham and Schmidt, 1999; Sexton et al., 1999
; McCreary and Sadava, 2000
; San Jose et al., 2000
). Therefore, it was considered necessary to control for stress and depression scores in the analysis of alcohol outcomes. However, preliminary analysis revealed that stress and depression were highly linearly correlated (Pearsons r = 0.81, P < 0.0001), and because putting both in the statistical model would inflate the standard error estimates, only one of the two variables was allowed to remain. A coarse analysis of each outcome measure fit independently to stress and depression showed consistently higher regression coefficients for stress; consequently, stress was selected and depression omitted.
HRAs were generated from four theoretical groups: comparison site pre- and post-intervention (ComparisonPre, ComparisonPost) and the intervention site pre- and post-intervention (InterventionPre, InterventionPost). Changes in attitudes and behaviours were anticipated after the intervention start date among employees at the intervention site; these changes represent the InterventionPost group. The effect of the intervention was evaluated by comparing HRA outcomes for the InterventionPost group to the three other groups, while adjusting for the factors of age, gender, stress and job class. A logistic regression model was used to predict the daily probability of binge drinking during the past 30 days and the probability of increasing the desire to reduce drinking. The mean of each outcome measure was modelled as a function of the following predictors: Group (Intervention, Comparison); Time (Pre-, Post-); the interaction between Group and Time; and demographic factors including age (30, 3140,
41 years), gender (female, male), and job class (professional/technical, executive administrative, administrative support, other). Regression coefficients for each predictor were adjusted for all other terms in the model. Adjusted odds ratios estimated by the analysis were interpreted as approximate relative risks, averaged over all members of the study population.
![]() |
RESULTS |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Binge drinking
The mean number of days spent binge drinking was negatively associated with age (Table 3). Furthermore, on average, men reported more than twice the binge drinking frequency of women (P < 0.001). Administrative support personnel reported more than double the number of days binge drinking as did the executive/administrative employees. Stress scores were slightly higher for younger subjects and those in the administrative support job classes. The desire to cut down on drinking was predictably higher for those with episodes of binging, but was also higher among older subjects, women, and those in the executive/administrative job class. Employees in the other job class, representing blue-collar workers, were least likely to express a desire to cut down on alcohol use. These mean differences were not tested inferentially, due to the repeated-measures nature of these data.
|
|
|
Desire to reduce drinking
Logistic regression was used to model the mean probability of expressing a desire to reduce drinking as a function of intervention status, controlling for stress, age, gender and job class. Among those reporting zero binge days, 552 individuals provided 696 HRAs used in this analysis; among individuals admitting to one or more days of binging, 224 individuals provided 261 HRAs. The results of the logistic regression analysis are also given in Table 5, which shows no significant effect of any predictor on the desire to reduce drinking among those employees who reported zero days of binge drinking in the past 30 days. These results contrast with the analysis of employees who reported one or more days of binging in the past 30 days. Individuals who participated in the intervention were nearly 2.6 times as likely as all other employees (P < 0.05) to express a desire to reduce their drinking behaviour in the post-intervention time period. Employees in the other job classes reported a significantly lower probability of willingness to reduce drinking (P < 0.05), almost half that expressed by employees in the professional/technical group.
![]() |
DISCUSSION |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Our finding that over 35% of the binge drinkers reported a desire to reduce their alcohol use, is encouraging. Furthermore, self-reported binge drinkers in the intervention site had significantly higher odds of reporting a desire to reduce their alcohol use after exposure to the Project WISE intervention. Although the one question on motivation provided only a limited scope for interpretation as regards motivation for change, that employees exposed to Project WISE were 2.59 times more likely to express a desire to cut down, suggests the programme made an important contribution toward stimulating employees critical inspection of their drinking habits. The finding that motivation to change was affected, while binging behaviour was not, was somewhat expected within the short 2-year period of this study. Because alcohol habits are interconnected with cultural norms and reinforced by social customs, altering drinking behaviour in a group that was not already predisposed toward change would be difficult and would require time to progress through the stages of change (Prochaska, 1996).
The findings of this investigation support the strategy of addressing binge drinking through interventions to reduce stress levels. While stress scores on the HRA were associated with increased binging, they did not appear to be associated with the desire to reduce drinking. One may hypothesize that employees who feel their stress levels peak also perceive that the use of alcohol is justified, either as self-medication or reward. It may also be that employees who were feeling higher levels of stress did not consider their emotional reserves sufficient to take on more change. In any case, these employees did not respond to the intervention, which indicates, for programmes such as Project WISE, that stress reduction would be a valuable component of targeting these resistant employees. It should be noted that health care work is accompanied by high stress levels, making this population particularly vulnerable (Hardy et al., 1997; Firth-Cozens, 1998
).
This study had several limitations. First was the fairly low (23.4%) participation in completing the EWP and HRA. Typical participation rates in HRAs have been established at 2030% of workers (Edington, 2001), which means that the majority of workers are generally not reached through wellness programmes. Our finding that males and also the highest paid employees were less likely to participate is not surprising, but the gender discrepancy is particularly disheartening as males are more likely than females to engage in binge drinking (SAMHSA, 1996
). For programmes like Project WISE, HRAs are necessary to identify and target high-risk individuals. Therefore, to succeed, employers must either create incentive packages that strongly encourage participation or possibly even mandate membership. Secondly, the study was limited by the low number of subjects who actually engaged in binge drinking, with a mean of only 0.54 days of binge drinking in the past 30 days for the intervention, and 0.66 for the comparison sites. This left little room for improvement, and the need for change was not great for most drinkers. Binge drinking was defined as 5+ drinks per occasion at least once in the past 30 days for both male and female employees. However, some researchers recommend using a cut-off of 4 or more drinks per occasion for women (Wechsler et al., 1994
). This could account for some of the discrepancy between binge drinking rates of men and women in this report.
The low rates for this population of health care workers may reflect a tendency for employees drawn to a voluntary wellness programme to be more health conscious than their non-participant counterparts. Those who choose to engage in health promotion are more likely to consider fitness a high priority (Eakin et al., 1988), have greater concern for health (Morgan et al., 1984
), and have increased feelings of self-efficacy (Alexy, 1991
; Steinhardt and Young, 1992
). The 22-month follow-up in this investigation was sufficient to capture early changes in risk, but inadequate to determine longer-term outcomes. Outcomes measurement ideally should continue for up to 5 years following intervention to resolve whether the impact is sustainable or enhanced by repetitive exposure to the intervention, and at what point in time the maximum benefit has been achieved.
![]() |
ACKNOWLEDGEMENTS |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
![]() |
FOOTNOTES |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
![]() |
REFERENCES |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Alleyne, B. C., Stuart, P. and Copes, R. (1991) Alcohol and other drug use in occupational facilities. Journal of Occupational Medicine 3, 496500.
Bamberg, R., Acton, R., Goodson, L., Go, R., Struempler, B. and Roseman, J. (1989) The effect of risk assessment in conjunction with health promotion education on compliance with preventive behaviors. Journal of Allied Health 18, 271280.[Medline]
Brown University (1998) Social marketing program slows binge drinking on college campuses. Brown University Digest Addiction Theory Application 17, 18.
Connection Series (1996) 201 North Union Street, Suite 330, Alexandria, VA 22314, ISA Associates.
Cook, R. F. and Back, A. (1990) SAY YES! Healthy Choices for Feeling Good. ISA Associates, 201 North Union Street, Suite 330, Alexandria, VA.
Eakin, J. M., Gotay, C. C., Rademaker, A. W. and Cowell, J. W. (1988) Factors associated with enrollment in an employee fitness center. Journal of Occupational Medicine 30, 633637.[ISI][Medline]
Edington, D. (2001) Emerging research: a view from one research center. American Journal of Health Promotions, 15, 341349.[ISI][Medline]
Edington, D., Yen, L. and Witting, P. (1997) The financial impact of changes in personal health practices. Journal of Occupational and Environmental Medicine 39, 10371046.[ISI][Medline]
Firth-Cozens, J. (1998) Individual and organizational predictors of depression in general practitioners. British Journal of General Practice 48, 16471651.[ISI][Medline]
Gebhardt, D. L. and Crump, C. (1990) Employee fitness and wellness programs in the workplace. American Psychologist 45, 262272.[CrossRef][ISI][Medline]
Goetzel, R., Jacobson, B., Aldana, S. et al. (1998) Health care costs of worksite health promotion participants and non-participants. Journal of Occupational and Environmental Medicine 40, 341346.[ISI][Medline]
Graham, K. and Schmidt, G. (1999) Alcohol use and psychosocial well-being among older adults. Journal of Studies on Alcohol 60, 345351.[ISI][Medline]
Hardy, G. E., Shapiro, D. A. and Borrill, C. S. (1997) Fatigue in the workforce of the National Health Service Trusts: levels of symptomatology and links with minor psychiatric disorder, demographic, occupational and work role factors. Journal of Psychosomatic Research 43, 8392.[CrossRef][ISI][Medline]
Heaney, C. A. and Goetzel, R. Z. (1997) A review of health-related outcomes of multi-component worksite health promotion programs. American Journal of Health Promotion 11, 290308.[ISI][Medline]
Heirich, M. A. and Sieck, C. J. (2000) Worksite cardiovascular wellness programs as a route substance abuse prevention. Journal of Occupational and Environmental Medicine 42, 4756.[CrossRef][ISI][Medline]
Lapham, S., Chang, I. and Gregory, C. (2000) Substance abuse intervention for health care workers: a preliminary report. Journal of Behavioral Health Services and Research 27, 131143.
Lehman, W. E. and Simpson, D. D. (1990) Patterns of drug use in a large metropolitan workforce. NIDA Research Monograph 100, 4562.[Medline]
Lehman, W. and Simpson, D. (1992) Employee substance use and on-the-job behaviors. Journal of Applied Psychology 77, 309321.[CrossRef][ISI][Medline]
Liang, K. and Zeger, S. L. (1986) Longitudinal data analysis using generalized linear models. Biometrika 73, 1322.[ISI]
McCreary, D. R. and Sadava, S. W. (2000) Stress, alcohol use and alcohol-related problems: the influence of negative and positive affect in two cohorts of young adults [in process citation]. Journal of Studies on Alcohol 61, 466474.[ISI][Medline]
Morgan, P. P., Shephard, R. J., Finucane, R., Schimmelfing, L. and Jazmaji, V. (1984) Health beliefs and exercise habits in an employee fitness programme. Canadian Journal of Applied Sport Sciences 9, 8793.
Pelletier, K. R. (1999) A review and analysis of the clinical and cost-effectiveness studies of comprehensive health promotion and disease management programs at the worksite: 19951998 update. Science of Health Promotion 13, 131143.
Prochaska, J. O. (1996) A stage paradigm for integrating clinical and public health approaches to smoking cessation. Addictive Behaviors 21, 721732.[CrossRef][ISI][Medline]
Richmond, R., Kehoe, L., Heather, N. and Wodak, A. (2000) Evaluation of a workplace brief intervention for excess alcohol consumption: the workscreen project. Preventive Medicine 30, 5163.[CrossRef][ISI][Medline]
Roman, P. M. and Blum, T. C. (1996) Alcohol: a review of the impact of worksite intervention on health and behavioral outcomes. American Journal of Health Promotion 11, 136149.[ISI][Medline]
SAMHSA (1993) Substance Abuse and Mental Health Services Administration: National Household Survey on Drug Abuse: Main Findings 1991. DHHS Pub. No. (SMA)93-1980. SAMHSA, Rockville, MD.
SAMHSA (1996) Preliminary Estimates from the 1995 National Household Survey on Drug Abuse. (18), U.S. Department of Health and Human Services. SAMHSA, Rockville, MD.
Sanchez-Craig, M. (1994) Low-Risk Drinking: Patient Workbook for Quitting or Cutting Down. The College of Family Physicians of Canada, Missauga, Ontario, Canada.
San Jose, B., van de Mheen, H., van Oers, J. A., Mackenbach, J. P. and Garretsen, H. F. (2000) Adverse working conditions and alcohol use in men and women. Alcoholism: Clinical and Experimental Research 24, 12071213.[CrossRef][ISI][Medline]
Serxner, S., Gold, D., Anderson, D. and Williams, D. (2001) The impact of a worksite health promotion program on short-term disability usage. Journal of Occupational and Environmental Medicine 43, 2529.[CrossRef][ISI][Medline]
Sexton, H., Lipton, R. I. and Nilssen, O. (1999) Relating alcohol use and mood: results from the Tromsø study. Journal of Studies on Alcohol 60, 111119.[ISI][Medline]
Sheridan, J. and Winkler, H. (1989) An evaluation of drug testing in the workplace. NIDA Research Monograph 91, 195216.[Medline]
Sorensen, G., Stoddard, A., Hunt, M. K., Hebert, J. R., Ockene, J. K., Avrunin, J. S., Himmelstein, J. and Hammond, S. K. (1998) The effects of a health promotion-health protection intervention on behavior change: the wellworks study. American Journal of Public Health 88, 16851690.[Abstract]
Stein, A., Shakour, S. and Zuidema, R. (2000) Financial incentives, participation in employer-sponsored health promotion, and changes in employee health and productivity: HealthPlus Health Quotient Program. Journal of Occupational and Environmental Medicine 42, 11481155.[CrossRef][ISI][Medline]
Steinhardt, M. A. and Young, D. R. (1992) Psychological attributes of participants and nonparticipants in a worksite health and fitness center. Behavioral Medicine 18, 4046.[ISI][Medline]
Stunkard, A. J., Cohen, R. Y. and Felix, M. R. (1989) Weight loss competitions at the worksite: how they work and how well. Preventive Medicine 18, 460474.[CrossRef][ISI][Medline]
U.S. Bureau of Labor Statistics (1989) Survey of Employer Antidrug Programs. (760), U.S. Department of Labor, Washington, DC.
Wechsler, H., Davenport, A., Dowdall, G., Moeykens, B. and Castillo, S. (1994) Health and behavioral consequences of binge drinking in college: a national survey of students at 140 campuses. Journal of the American Medical Association 272, 16721677.[Abstract]