When introducing programmes to support employees changing to a more healthy lifestyle, how do you assess whether they are having an impact? The occupational health service of the Health Service Executive Dublin North East carried out a lifestyles screening pilot programme with its ambulance service and used data analysis tools to produce statistics showing how the intervention affected employees. This article explains the approach.
The World Health Organization describes health promotion as “the process of enabling people to increase control over, and improve their health”1. It advocates the workplace as an ideal setting for health promotion as it provides access to a defined population, where employees can gain support from their peers, and where established channels of communication already exist. The benefits of promoting workplace health are twofold: for the individual in terms of promoting and maintaining their physical and mental health and for the organisation in terms of productivity and morale.
Specialists in occupational health are acutely aware that many lifestyle factors can ultimately lead to ill health at work. Prolonged hours in a sedentary setting can result in a wide range of musculoskeletal disorders. A study in Finland reported the annual incidence of neck pain among display screen equipment users to be 34%2.
Research also suggested that an increased prevalence of upper extremity musculoskeletal symptoms may be associated with increased computer mouse use3, 4, 5. Specific exercises can help to reduce the onset of repetitive strain injuries such as carpel tunnel syndrome for employees who carry out continuous unvaried work.
Exercise helps prevent injuries by strengthening muscles and joints. It maintains flexibility, which aids posture and range of motion, and may help improve balance and co-ordination for the individual. One way to highlight risk factors is through an individual lifestyle screening programme in the workplace. Early education and intervention will, therefore, ultimately promote and maintain the highest degree of health at work and thus benefit the employee and the employer.
The OH service of the Health Service Executive Dublin North East chose its ambulance service because, at 130 employees, the size of the service reduced the risk of selection bias.
Baseline assessments of those who chose to participate were carried out by the investigating OH nurse adviser from October-December 2004, using the software package Preventive Healthcare Services from Warwick International Systems. Participants were reassessed six months later to establish if any lifestyle changes had occurred and also to reinforce any positive lifestyle behaviour changes.
An OH nurse adviser travelled to each ambulance base and completed individual assessments on site. Participants who did not attend for reassessment were notified by post and asked to contact the OH service to reschedule their appointment. Those who failed to contact the OHS for a follow-up appointment were sent a general lifestyles appraisal by post together with a stamped addressed envelope.
A printout of results and an advice sheet were then given to all participants. The final data set was collected in September 2005 and analysed using the Statistical Package for the Social Sciences.
Forty-two per cent of the sample population responded to the initial invitation to participate (54 out of 130 employees). Eighty per cent of those who participated in phase 1 attended for reassessment in phase 2. In phase 2 all participants on duty on the day of assessments in each base attended.
The programme
A general lifestyles appraisal was completed on all participants to obtain a baseline, which highlighted possible risk factors for cardiovascular disease or musculoskeletal disorder. A subsequent cardiovascular risk appraisal or musculoskeletal assessment was then completed by the investigating OH nurse adviser where indicated.
The musculoskeletal assessment was unpopular with participants due to its perceived limited utility, and was discontinued following a trial period. Participants found the detailed interactive back pain computer-based questionnaire very time consuming, and did not find the information and reports produced on their musculoskeletal disorders useful.
General lifestyle appraisal: Data fields recorded in the general lifestyles appraisal included: grade, gender, age, body mass index, past medical history of heart disease, cerebro-vascular accident, diabetes, smoking status, alcohol consumption, level of physical activity and dietary habits. Participants’ present level of stress was interpreted using a doctor-rated screening test developed using latent trait analysis by Goldberg et al7. This test interpreted the participant’s present level of stress as either “normal, moderate or high”.
Each participant was given a personalised printout of results, which interpreted each answer in terms of the normative range. Health advice was given on how to maintain or achieve a result within the acceptable limits. A healthy lifestyle action plan was given, based on the information given in the appraisal, which indicated the lifestyle changes (if any) required to maintain good health.
Cardiovascular risk assessment: The cardiovascular risk appraisal followed a similar format to the general lifestyle appraisal. Data collected specifically related to cardiovascular risk factors. Participants’ 10-year coronary heart disease risk was then calculated and was measured against the predicted risk for each participant.
Musculoskeletal assessment: The musculoskeletal assessment consisted of a detailed interactive back pain computer-based questionnaire. In this questionnaire the participant was prompted to complete anatomical diagrams and multiple-choice answers in an endeavour to identify and analyse detailed information about their musculoskeletal symptoms.
The participant was given a printout of their results.
Results
Demographics: More than three-quarters of the sample population were male and 17% of the sample population was female. The majority of the sample population were emergency medical technicians 4% were members of patient transport service 6% were clerical officers. The age of participants ranged from 27- to 60 years old. Almost two-thirds of the sample population were more than 35 years of age, and this age profile was consistent with the stated intention of the pilot programme to target the over 35s. The mean and median was 40 years of age. The mode was 33 years, while the range was 37.
Fibre intake per week: Participants were asked to estimate how much fibre they consumed per week, by choosing options from a dropdown menu which included every meal, daily, several times a week or seldom. A Wilcoxon Signed Rank Test confirmed there was a statistically significant difference in the mean rank amount of fibre consumed per week from phase 1 to phase 2.
Saturated fat intake per week: Examples of saturated fats were discussed with participants. They were then asked to estimate how much saturated fat they consumed per week, by choosing options from a dropdown menu which included every meal, daily, several times a week or seldom. A Wilcoxon Signed Rank Test confirmed there was a statistically significant difference in the mean rank amount of saturated fat consumed per week from phase 1 to phase 2.
Physical activity per week: Participants were asked how many times they were physically active for 30 minutes or more per week. A Paired Samples T-Test confirmed there was a statistically significant difference in the mean amount of exercise per week from phase 1 to phase 2.
Stress: A screening test for anxiety and depression was incorporated into the general lifestyles questionnaire. Participants were asked a number of questions relating to anxiety and depression. Each question scored one point for “yes”. The scores for the anxiety scores were totalled, as were the scores for the depression scale. Participants perceived present level of stress was calculated (rating as normal, moderate or high) using a doctor-rated screening test, developed using latent trait analysis by Goldberg et al7. Doctor-rated screening tests are a structured way of interviewing and rating the severity of an illness.
A Wilcoxon Signed Rank Test confirmed there was a borderline statistically significant difference in the mean rank amount of perceived level of stress experienced per week from phase 1 to phase 2.
High-density lipoprotein cholesterol (HDL-C): Participants’ HDL-C was measured. A Paired Samples T-Test confirmed there was a borderline difference in mean HDL-C from phase 1 to phase 2.
Ratio of total cholesterol to HDL-C: This ratio was recorded. A Paired Sample T-Test confirmed a statistically significant difference in the mean ratio from phase 1 to phase 2.
Overall, one can conclude that the statistical findings were not due to chance variation, and may be attributed to an improvement in lifestyle behaviour.
Discussion and conclusions
This pilot programme set out to examine the characteristics of those participating in a lifestyles screening programme, and the degree to which their lifestyle behaviours changed following intervention.
The programme began with a general lifestyles appraisal to obtain a baseline, and also to identify the presence of risk factors for cardiovascular disease or musculoskeletal disorder. Eighty-two per cent of participants were found to have three or more risk factors for cardiovascular disease, triggering the criteria for an additional cardiovascular risk appraisal. The musculoskeletal assessment was unpopular with participants due to its perceived limited utility, and was discontinued following a trial period.
A disadvantage of this type of intervention is that healthy individuals usually volunteer thus introducing volunteer bias, which adds difficulty in targeting interventions to the population that would probably benefit the most. In an attempt to overcome this problem the positive benefits of the programme were promoted at each stage of the programme in a supportive and proactive way.
Depending on the data, a Paired Samples T-Test or Wilcoxon Signed Rank Test was performed to evaluate the impact of the intervention on participants’ lifestyle factors.
Statistically significant lifestyle improvements were noted in four variables studied (level of physical activity per week, fibre intake per week, saturated fat intake per week and the ratio of total cholesterol to HDL-C). Borderline statistically significant improvements in lifestyle behaviours were noted in a further two variables (present level of stress and level of HDL-C).
In a written evaluation on completion of the programme by participants, 97% stated unhealthy lifestyle behaviours were identified, of which 74% stated they had made positive lifestyle changes in relation to the maladaptive life behaviours previously pursued. Some participants were sceptical of the limitations of a computer-screening package regarding this type of intervention, although 90% of participants found it very beneficial and requested that it be available on an annual basis.
The authors are Mary Devaney, OH nurse adviser and Dr Peter Noone, consultant occupational physician, of the occupational health service, Health Service Executive, Dublin. The authors are sincerely grateful to the Health Service Executive, Dublin North East’s Ambulance Service for their participation and help with this pilot programme. They also wish to thank the executive’s Department of Health Promotion
References
1 The Ottawa Charter for Health Promotion. Geneva: World Health Organization 1986.
2 Jens Wahlstrom. Ergonomics, musculoskeletal disorders and computer work. Occupational Medicine 2005 55:168-
3 Karlqvist L, Hagberg M, Koster M, et al. Musculoskeletal symptoms among computer-assigned design (CAD) operators and evaluation of a self-assessment questionnaire. International Journal of Occupational Environmental Health 19962:185-194.
4 Fogleman M, Brogmus G. Computer mouse use and cumulative trauma disorders of the upper extremities. Ergonomics 199538:2465-2475.
5 Jensen C, Borg V, Fines L, et al. Job demands, muscle activity and musculoskeletal symptoms in relation to work with the computer mouse. Scandinavian Journal of Work Environmental Health 1998 24:418-424.
6 Statistical Package for the Social Sciences Version 14.0.1
7 Goldberg D, Bridges K, Duncan-Jones P, Grayson D. Detecting anxiety and depression in general medical settings. British Medical Journal 1988 297: 897-899.
Statistical analysis
The data was analysed using the Statistical Package for the Social Sciences.
Depending on the participant’s answers, a Paired Samples T-Test or Wilcoxon Signed Rank Test was performed to evaluate the impact of the intervention on participant’s lifestyle factors. The T-Test procedure compares the means of two variables for a single group.
The procedure computes the differences between values of the two variables for each case and tests whether the average differs from 0. Like the T-Test, the Wilcoxon Signed Rank Test applies to two-sample designs involving repeated measures, matched pairs, or “before” and “after” measures.
These tests were used because the Paired Sample T-Test assumes that the observations for each pair should be made under the same conditions. The mean differences should be normally distributed. Variances of each variable can be equal or unequal. The Wilcoxon Signed Rank Test does not require assumptions about the form of the distribution of the measurements. It should therefore be used whenever the distributional assumptions that underlie the T-Test cannot be satisfied.
A result is considered statistically significant if the significance value, also known as the probability (p) value is less than 0.05. When the p value is less than 0.05, the reader can conclude that this factor was not due to chance variation, and can be attributed to the improvement of the participant’s lifestyle.
Statistically significant improvements in lifestyle behaviour were found in four of the variables studied. These included the level of physical activity per week (p=0.001), fibre intake per week (p=0.000). saturated fat intake per week (p=0.002) and ratio of total cholesterol to high-density lipoprotein cholesterol (p=0.013). A borderline significant improvement in lifestyle behaviour was found in present stress level (p=0.054), and level of HDL-C (p=0.054).
Laerning for life: CPD quiz
1 What is the process of ‘enabling people to increase control over, and improve their health’ a definition of?
a) Health
b) Occupational health
c) Health education
d) Health promotion
2 Which group of employees were chosen for the study?
a) Accident and emergency staff
b) Laboratory workers
c) Ambulance service workers
d) Sedentary workers
3 How long was left between the first assessment and the second?
a) One month
b) Six months
c) One year
d) Six years
4 Why didn’t participants like the musculoskeletal assessment?
a) The questionnaire was time consuming to complete and the printout information was not useful
b) The questionnaire was time consuming but the printout information was very useful
c) The questionnaire was simple but the printout information was not useful
d) The questionnaire was a waste of time and there was no printout information.
5 What screening test was used to interpret participants’ stress levels?
a) Nurse-rated test
b) Doctor-rated test
c) Patient-rated test
d) Stress-rated test
6 What test was used to assess the cardiovascular risk?
a) Ten-year coronary heart disease risk
b) Five-year coronary heart disease risk
c) Ten-year cardiovascular disease risk
d) Five-year cardiovascular disease risk
7 Which of the following was NOT one of the improvements found at the end of the study?
a) Physical activity per week
b) Fibre intake per week
c) Sugar intake per week
d) Saturated fat per week
8 What type of cholesterol was measured?
a) HDL only
b) HDL and total cholesterol to HDLC
c) LDL only
d) Neither
9 What percentage of participants found this study beneficial?
a) 10%
b) 25%
c) 50%
d) 90%
Sign up to our weekly round-up of HR news and guidance
Receive the Personnel Today Direct e-newsletter every Wednesday
10 What statistical package was used for this study?
a) Stata
b) SAS
c) SPSS
d) Statistica