Article Text
PDF
Abstract
Background There is contradictory evidence on the association between health check-ups and future morbidity. Among the general population, those with high socioeconomic position participate more often in health check-ups. The main aims of this study were to analyse if attendance to health check-ups are socioeconomically patterned and affect sickness absence over a 10-year follow-up.
Methods This register-based follow-up study included municipal employees of the City of Helsinki. 13 037 employees were invited to age-based health check-up during 2000–2002, with a 62% attendance rate. Education, occupational class and individual income were used to measure socioeconomic position. Medically certified sickness absence of 4 days or more was measured and controlled for at the baseline and used as an outcome over follow-up. The mean follow-up time was 7.5 years. Poisson regression was used.
Results Men and employees with lower socioeconomic position participated more actively in health check-ups. Among women, non-attendance to health check-up predicted higher sickness absence during follow-up (relative risk =1.26, 95% CI 1.17 to 1.37) in the fully adjusted model. Health check-ups were not effective in reducing socioeconomic differences in sickness absence.
Conclusions Age-based health check-ups reduced subsequent sickness absence and should be promoted. Attendance to health check-ups should be as high as possible. Contextual factors need to be taken into account when applying the results in interventions in other settings.
- HEALTH SERVICES
- INEQUALITIES
- OCCUPATIONAL HEALTH
- SICKNESS ABSENCE
Statistics from Altmetric.com
Introduction
There is evidence that health check-ups may contribute positively to subsequent health, as measured with different health indicators, such as sickness absence1 and mortality.2–4 Common health check-ups for populations are often suggested as a means to improve the overall health of the populations and to reduce socioeconomic differences. However, the results vary between countries and time of the studies.
A comprehensive analysis was carried out in a systematic Cochrane review on the outcomes of randomised controlled studies on the effects of health check-ups. The results were disappointing regarding the effectiveness of health check-ups. Participation in non-focused health check-ups increased the number of diagnoses on, for example, hypertension and hypercholesterolaemia, but did not manage to reduce subsequent morbidity and mortality. According to the authors, possible explanations for the finding were that participation in health check-ups might be selective and that most of the included studies were from the 1960s and 1970s, when preventive treatments were not as developed as they are now.5 A French study with a large data set from the 1980s showed that high participation in recurrent health examinations predicted lower mortality in a 25-year follow-up in women and men.4 Japanese studies using data from the 1990s indicated that participation in general health examinations was an independent predictor of lower mortality among women and men,2 also when a wide set of health behaviour characteristics were taken into account.3 The varying findings indicate that the topic needs more attention using up-to-date data from different countries and contexts.
Previous studies on health check-ups have used morbidity and mortality as outcome measures, but only few have used sickness absence. There are several reasons for the use of sickness absence. It is a reliable health indicator,6–8 causes high costs,9 ,10 is often used to justify health interventions, but is rarely taken into account in health intervention studies.11 ,12 In a Swedish study, health check-ups were used to identify risk factors for sickness absence. Both narrow and wide examination protocols identified employees with higher risk of subsequent sickness absence,1 but the effectiveness of health check-ups in reducing sickness absence was not studied.
It is known that participation in health check-up plans varies by socioeconomic position. Usually, people with a higher socioeconomic position,13 ,14 more positive attitude towards health promotion and better health behaviour, such as non-smoking, less overweight and lower alcohol consumption,14 participate more often in health check-ups. However, it is not known how participation in health check-ups affects subsequent socioeconomic differences in health. The logic may go in both directions. First, it can be assumed that if preventive actions can be initiated among populations with lower socioeconomic position, this may reduce socioeconomic differences. Second, if participation in health check-ups is substantially higher among higher socioeconomic groups, non-focused health check-ups may even increase socioeconomic differences in health, as preventive actions are disproportionately carried out among the more privileged. The underlying assumption in both possibilities is that health check-ups are an effective way to promote health.
The main objectives of this study were to examine (1) how attending health check-ups is divided between socioeconomic groups, (2) how attending health check-ups will predict medically certified sickness absence and (3) whether health check-ups will narrow socioeconomic differences in medically certified sickness absence.
Methods
Data
This register-based follow-up study is part of the Helsinki Health Study, focusing on health and well-being of the employees of the City of Helsinki.15 The City of Helsinki is the largest employer in Finland and has ca. 40 000 employees.16 The City of Helsinki produces various services for the citizens ranging from healthcare, social welfare services and primary education to technical services. As a large employer, the City of Helsinki provides the employees free occupational healthcare, ranging from traditional preventive care to primary healthcare services. From the perspective of employment security, the early 2000s was a relatively stable period, but from 2009 onwards, an economic downturn resulted in a decrease in the number of staff. However, the reductions were carried out without lay-offs simply by not recruiting as much new staff as retired.
The participants of this study included full-time and permanently employed municipal employees of the City of Helsinki aged 40, 45, 50, 55 and 60 years in each year of 2000–2002, who were invited to age-based health check-up carried out by occupational health nurse for all employees with respective age. The health check-ups included answering a short health questionnaire including questions on health behaviour and work ability, 60 min counselling of the occupational health nurse with physical measurements including height, weight, calculating body mass index and blood pressure, and taking a limited set of laboratory examinations such as complete blood count, serum cholesterol, fasting blood sugar and urine test. During the health check-up, some health advice was given based on individual needs, but the general purpose was to find employees with a need for further examinations or an appointment to occupational physician. Health check-ups were free for the employee and employees could participate in them during working hours. The occupational healthcare is organised in-house and provides primary healthcare services for all employees of the City of Helsinki. In all, 13 037 employees were invited to health check-ups of whom 8071 attended, leaving a non-participation rate of 38%. A general non-participation analysis showed that low income was moderately associated with non-participation in health check-ups among women and men, whereas low occupational class was associated only among men.17 Long sickness absence in the study year was not associated with the participation in health check-ups,17 but inclusion of employees with temporary working contract may have affected these results, as it is known that this group has lower sickness absence rates than the permanently employed. Employees, who attended health check-ups and gave consent to data linkage for research purpose, were classified as attendees and the others as non-attendees.
Information on education, occupational class and individual income from the same year the employee was invited to attend health check-up was obtained from the employer's personnel register.
Education level was classified as high, if the employee had graduated from the upper secondary school. Intermediate and comprehensive school were classified as intermediate-education group. Employees with only compulsory education were classified as low-education group.
Occupational class was classified into four categories according to the job title of each employee at baseline: (1) managers and professionals, such as teachers and physicians, (2) semiprofessionals, such as nurses and foremen, (3) routine non-manuals, such as clerical employees, child minders, home assistants and practical nurses and (4) manual workers, such as technical and cleaning workers.18
The employees were divided into income quartiles according to their monthly individual salary from the City of Helsinki. This measure did not include income from other sources, such as investment income or transfer payments. The individual income represents more closely individual socioeconomic position than, for example, household disposable income.19
The number of sickness absence spells of 4 days or longer per 100 person years was measured at baseline and during the follow-up period. The latter was used as an outcome variable. The employer requires a certification from an occupational health nurse or a physician for such absences. Information on each individual sickness absence period was gathered from the employer's personnel register with an accuracy of 1 day. Consecutive and overlapping sickness absence periods were combined.
The employees were followed until end of 2012 using the employer's personnel register. Owing to turnover and retirement, the mean follow-up time was 7.5 years.
Methods
Besides descriptive distributions, presented in table 1, age-adjusted numbers of sickness absence spells were calculated. Poisson regression analysis was used to estimate age-adjusted relative risks for sickness absence during follow-up and their 95% CIs in accordance with previous sickness absence studies.7 ,20 ,21 For the results presented in table 2, models were fitted to examine the effect of health check-up attendance, and by adding socioeconomic position measures one at a time to the models. In addition, baseline sickness absence was included in the final model to control for differences in health status and context related to individual work ability at baseline. For the results presented in table 3, models were fitted to examine socioeconomic differences in sickness absence. Other socioeconomic position measures and baseline sickness absence were added to the models to examine their modifying effects as in previous analyses. Health check-up attendance was included in the final model to analyse its effectiveness in reducing socioeconomic differences.
Distributions of socioeconomic position measures, health check-up attendance and age-adjusted average number of sickness absence spells per 100 person years among the staff of the City of Helsinki
Sickness absence spells per 100 person years among the staff of the City of Helsinki by health check-up attendance and socioeconomic position
Sickness absence spells per 100 person years among the staff of the City of Helsinki by socioeconomic position measures and health check-up attendance
Results
In table 1, basic distributions and age-adjusted levels of sickness absence are presented. Attendance to health check-ups was quite evenly distributed or so that groups with lower socioeconomic position had slightly higher attendance. There was also difference between genders so that men attend more frequently to health check-ups than women. Sickness absence rates were unequally distributed so that employees with less education, lower occupational class and lower income had higher rates of sickness absence.
Relative risks for long sickness absence over the follow-up by attending health check-ups are presented in table 2. From the results, it is evident that non-attendance constantly predicted a higher risk of further long sickness absence. The relative risk of the non-attendees was 1.30 (95% CI 1.20 to 1.40) among women and 1.17 (95% CI 0.97 to 1.42) among men. Taking all socioeconomic position measures into account simultaneously reduced the difference by attendance by 9% among women and 52% among men. It can thus be noticed that the predictive effect of attendance was independent from socioeconomic position and also baseline sickness absence. However, these results were statistically significant only among women.
Sickness absence spells by all three socioeconomic position measures are presented in table 3. The results were consistent showing that low socioeconomic position predicted higher subsequent sickness absence. When other socioeconomic position measures and baseline absence were included in the model, education and occupational class remained independent determinants among women. The same applied for occupational class and income among men. Health check-up attendance slightly reduced the occupational differences among women and practically not at all among men.
Discussion
In this register-based 10-year follow-up study, the main aims were to analyse how attending age-based health check-up was divided between socioeconomic groups, predicted sickness absence and affected socioeconomic differences. The main findings were that low socioeconomic position groups and men participated to some extent more often in health check-ups, attending health check-ups predicted lower sickness absence among women and that participation in health check-ups did not practically affect socioeconomic differences in sickness absence. There are several explanations for the findings and are discussed next.
Contrary to previous studies, low socioeconomic position groups and men participated more actively in health check-ups.13 ,14 Several potential explanations stand out. First, it is possible that upper socioeconomic groups have given less often consent to give information for research purposes, but this is not supported by a previous study carried out in Helsinki Health Study where giving consent to research purposes did not differ by socioeconomic position or sickness absence.17 Second, in several manual occupations, it is already mandatory to attend regular health check-ups and this may reduce participation in voluntary check-ups. These occupations include, for example, fire fighters and bus drivers, in which there are special health criteria to be allowed to work. It seems that this does not reduce participation in age-based health check-ups but can make them socially more accepted. As men dominate these occupations, this may also explain partly the fact that men and especially men in lower socioeconomic groups attended more often in health check-ups. Age-based health check-ups have been carried out for years and they have evolved to be a generally accepted and even awaited means to maintain health and well-being among employees, and might also explain higher participation rate among men. We assume that this cultural and contextual explanation is the most likely, but its verification is not possible using only register-based data.
Attending health check-ups predicted lower sickness absence in the follow-up. This is a novel finding. Previous studies using other health indicators have been ambiguous,4 ,5 but more recent Japanese studies on the effects of health check-ups on general population indicate that they are effective in maintaining health.2 ,3 Partially, this study supports these newer findings on other health indicators. Several potential explanations come up. First, it is possible that employees who have generally a more positive attitude towards health promotion and have better health behaviour are more likely to participate in health check-ups as found before.13 ,14 Such employees might just want to strengthen their perception on their own health and health behaviour. Second, employees with long-standing illnesses or sicknesses may have a regular contact with health services and therefore may not see an extra benefit in attending age-based health check-ups. Finally, it is also possible that health check-ups are effective in what they aim to do, that is, in finding risk factors and sickness in early phases and initiate preventive measures and treatment for them before they cause work disability and sickness absence. The effect of attendance was, however, statistically significant only among women. The total attendance rate was highest among men in routine non-manual and manual occupations and men with lowest income. These groups tend to have high sickness absence rates22 so this selection is likely to explain some of the gender differences in the effect of health check-up attendance. The positive predictive effect of attendance to health check-ups on future sickness absence was independent from socioeconomic position. So, it can be thought that all socioeconomic groups benefit from health check-ups.
It was found that low socioeconomic position predicted high sickness absence rates. This finding was in line with previous research.6 ,21 ,22 Non-existent or small differences between income groups among women may result from narrow income differences in the public sector and also between occupational groups. One possible explanation for this is that there are large groups of social and healthcare employees who receive shift work compensation and thus a higher salary, which narrows income differences. Health check-ups were not effective in reducing socioeconomic differences in sickness absence, as they benefit all socioeconomic groups in relatively equal way.
The strengths of this study were that the data set is comprehensive and follow-up time was 10 years. The register data can also been regarded as accurate and up to date, as the employer uses the data also in financial reporting and payroll accounting. Using register-based sickness absence data is a strength compared with self-reported sickness absence.23 A relative weakness in the study design is that information on those who attended health check-up but did not give a written consent to allow the information to be used for research purposes are mixed with those who did not attend health check-ups at all. However, the number of these is likely small, and the selection would result in smaller differences, and therefore would affect the results towards conservative direction. This study was carried out in large and comprehensive public sector workplace in Finland, so one must be careful when generalising the results to other organisations or social security settings.
Conclusions
It can be concluded that age-based health check-ups are likely to be beneficial and may contribute to lower subsequent sickness absence. This contribution remained when baseline sickness absence was taken into account. However, general age-based health check-ups are not sufficient in reducing socioeconomic health differences at least in a context, where participation in health check-ups is generally accepted and is relatively equally divided between socioeconomic groups.
The practical implication is that the coverage of health check-ups should be as high as possible, as all groups benefited from attending. Health service producers should consider ways to promote higher attendance to health check-ups regardless of initial state of health, but especially among population who currently does not participate in health check-ups, so that preventive actions or treatments can be carried out with them. These employees might have poorer health behaviour, such as high alcohol consumption, higher rates of smoking and poorer dietary habits, and higher subsequent sickness absence, as previously shown.24 Planners of services and interventions might learn from marketing of, for example, other services or consumer goods consumed and bought by these groups.
Age-based health check-ups require large amounts of resources. There is a tendency to select participants to extended health check-ups more precisely, using digital self-administered health surveys or light health check-ups. It is unexplored how such screening methods affect the participation in health check-ups or subsequent health.
Further research should be targeted to study more precisely what kind of health check-ups and protocols are most effective. There are many important research questions, for example, whether health check-ups should (1) use wide or limited protocols, (2) target general or vulnerable populations besides employees, (3) be age-specific or need-based and (4) what is the optimal interval of health check-ups. In addition, it would be important to study the effect of regular participation in health check-ups in subsequent health. The methods should include randomised controlled trials to find potential effects, observational studies to find out real-life effects and cost–benefit analysis to further justify actions to decision makers.
What is already known on this subject?
Studies from 1980s and 1990s indicate that health check-ups are effective in reducing future morbidity, but the overall results are ambiguous.
People with a higher socioeconomic position attend health check-ups more often.
What this study adds?
Men and employees with a lower socioeconomic position participated in health check-ups more often than women and those with a high socioeconomic position.
Attending health check-ups was an independent determinant of lower sickness absence among women.
Health check-ups were not effective in reducing socioeconomic differences in sickness absence.
References
Footnotes
Contributors All the authors developed the research questions and planned the analyses. KP, EL and OR acquired the data. KP conducted the analyses. KP drafted the manuscript, and all the authors revised it critically. KP is the guarantor.
Competing interests None declared.
Ethics approval This study was conducted with the approval of the ethics committee of the Department of Public Health at the University of Helsinki and the City of Helsinki Health Authorities.
Provenance and peer review Not commissioned; externally peer reviewed.