Association of health checkups with health-related quality of life among public servants: a nationwide survey in Taiwan

This study is the first nationwide survey of public servant workplace health in Taiwan and employs multistage stratified random cluster sampling according to proportional probabilistic sampling (PPS). A two-stage systematic stratified probabilistic sampling combined with a cluster sampling approach were applied to select the study sample. In the first stage, the selected unit is the chief administrative institutions, and each selected unit may have different weights according to the probability proportional to size sampling principle. In the second stage, the selected unit is the subordinate institution, and all staff members of the selected unit were treated as our study subjects. The survey was launched by Taiwan Health Promotion Administration (HPA) to guide the implementation of an intervention program to improve public servant’s health. The results served as a baseline to evaluate the efficacy of the intervention program for public servants. This study obtained ethical approval from the Institutional Review Board of China Medical University Hospital (CMUH105-REC3-091). Participants were informed that their data would be handled confidentially. Information regarding this study was sent to government institutions to encourage public servants to participate in our survey. After agreeing to participate, public servants filled out informed consent forms and completed an online questionnaire. Our study enrolled 21,583 participants, resulting in an overall response rate of 35.8%. Reasons for non-response included unwillingness, vacation, time off, or having insufficient time to fill out the questionnaire, but did not affect the validity of the data. A group of 11,454 public servants aged 40 to 65 years was selected after excluding cases with missing BMI (Body Mass Index) information. Our sample of 11,454 participants had more than 99% power to detect effects size of 0.1 for the significance of the multiple regression analysis with a type I error rate of 0.05. The power was calculated using the statistical package G*Power 3.1.9.2

The online questionnaire included questions related to demographics (gender, age, educational level, monthly income, and marital status), job characteristics (sector type, shift work, and managerial position), health behavior (regular physical exercise and consumption of 5 portions of vegetables or fruit a day), and health status (chronic conditions and obesity). Sectors were categorized as administration, public enterprise, medical services, or public school administration. Job characteristics including shift work and managerial position were crucial considerations. The health behavior section enquired whether respondents exercised more than 3 times a week and consumed 5 portions of vegetable or fruit a day. Obesity was defined using BMI, which is expressed as a weight and height ratio (kg/m²). Based on Taiwan HPA guidelines (2012) for Taiwanese adult men and women, BMI was categorized as follows: 18.5 to 24, healthy; 24.1 to 27.0, overweight; and ≥ 27, obese (27–30, mildly obese; 30–35, moderately obese; and ≥ 36, severely obese). Participants were categorized as having a chronic condition if they confirmed ever receiving a diagnosis of one or more of the following diseases: diabetes, hyperlipidemia, hypertension, heart disease, stroke, kidney disease, osteoarthritis, gout, peptic or duodenal ulcer, chronic obstructive pulmonary disease, liver disease, mental disorder, and cancer. Three types of health checkups during the preceding 3 years were classified: GPHC (government-paid health checkup), SPHC (self-paid health checkup), and NOHC (no health checkup). The group with GPHC was defined if they underwent HPA paid health checkups in the previous 3 years or received a subsidy from the government for public servants over the previous 2 years, but they have not received any self-paid health checkups. The participants who underwent any self-paid health checkups were defined as SPHC, and those with no health checkups in the preceding 3 years were defined as NOHC.

The 36-item questionnaire (SF-36) was used to measure HRQoL for each public servant, which included eight scales: physical functioning, role limitations due to physical problems, bodily pain, general health, vitality, social functioning, role limitations due to emotional problems and general mental health [11, 12]. In Taiwan, this measure has been translated into Chinese, and its validity and reliability have been verified [13]. Scores of physical component summary (PCS), and mental component summary (MCS) were calculated from raw data according to the SF-36 manual. For comparison, the domain scores were calculated to fit a scale of 0–100. The z-scores and factor loadings from the US population were used for the summaries [14, 15]. The average scores of MCS and PCS were used as two primary outcomes of receiving different types of health checkups for cross-sectional HRQoL.

The reliability and validity tests were conducted before the initiation of the survey. The content validity index (CVI) was 81% based on the assessment of content validity by five experts. Two types of interviews also yielded findings that were highly consistent with those of the online questionnaire, which indicates the reliability of the results. In addition, we found consistent results from both high- and low-response groups, and no significant differences were observed in the findings of the two response groups. Our results validated the reliability of the questionnaire with 20,046 participants, which included both a face to face and online questionnaire.

Our study population consisted of 11,454 participants. Table 1 lists information of all the variables included in this study, with 46.6% male and 53.4% female participants. The independent variables for different types of health checkups were mutually exclusive and categorized into 3 groups depending on health checkup use in the 3 preceding years: GPHC, SPHC, and NOHC, representing 30.1%, 41.4%, and 28.5%, respectively, of the study population. The mean age was 49.4 ± 6.02 years and most participants had a bachelor’s degree. The public servants who didn’t attend any health checkups were statistically more likely to be younger, male, have lower economic status, work in administration sectors, have non-managerial positions, shift-work, unhealthy lifestyles and no history of chronic conditions when compared with those that attended health checkups. This complies with the results of other studies.

Table 2 lists the different types of health checkups, the mean score of the PCS and MCS scores, and pair comparisons of mean in the crude model and the three other models. In crude Model, scores of PCS in the GPHC group were significantly higher than those who were in SPHC groups while scores of MCS in the GPHC group were significantly higher than those who were in other groups. In Model 1 and Model 2, scores of PCS in the GPHC groups were significantly higher than those who were in other groups while MCS in both GPHC and SPHC groups were significantly higher than those who were in NOHC groups. In Model 3, after adjustments for covariates, GPHC group had significantly higher PCS scores than those in the SPHC group (Mean difference = 0.62, P = 0.0001); however, no significant difference was noted for MCS scores.

Table 3 shows the difference between PCS and MCS in three groups after stratification for the number of chronic diseases (no chronic disease, having only 1 chronic disease, having 2 or more chronic diseases) using Model 3. In the no chronic disease group, the GPHC group had significantly higher PCS scores than those in the NOHC group (Mean difference = 0.55, P = 0.008). Furthermore, those in the GPHC group (Mean difference = 0.65, P = 0.02) or SPHC group (Mean difference = 0.81, P = 0.003) were significantly higher MCS scores than those in the NOHC group. Only the SPHC group was significantly higher in MCS score than those in the NOHC group (Mean difference = 1.4, P = 0.008) for the group with 1 chronic disease. Participants with 2 or more chronic diseases in the GPHC (Mean difference = 1.79, P = 0.002) or SPHC groups (Mean difference = 1.39, P = 0.009) had significantly higher PCS scores than those in the NOHC group, but no significant difference was noted for MCS score.

Figures 1 and 2 illustrate the PCS and MCS scores across different health checkups groups. The trend is the same in both PCS and MCS, where scores show the GPHC group is greater than the SPHC group, and the SPHC group is greater than the NOHC group. This may indicate that public servants undergoing GPHCs are more likely to report higher PCS and MCS scores than are those undergoing SPHCs.

We identified that 71.5% of public servants received health checkups, including GPHC and SPHC, this is greater than the report of 64.5% of citizens aged 40–64 who received health checkups from the Ministry of Health and Welfare [16]. The difference may be accounted for by the fact that only public servants received subsidiaries. In terms of middle or higher socio-economic status, literature has indicated a positive relationship between higher income / social class and utilization rates of preventive services [17, 18].

The findings also showed that non-attendees, NOHC, are younger, male, have a lower educational level, unhealthy lifestyle and multiple chronic conditions compared to attendees; this complies with other study results [19‐21]. Related factors like job characteristics, those who work in the administration sector, those in non-managerial positions, and shift work will be less likely to receive health checkups. Since the subsidiary for public servants is in accordance with their job grade, this may result in those with managerial positions attending health checkups and those with shift work failing to schedule checkups with health institutes. The reasons for not attending might include lack of awareness or knowledge, misunderstanding the purpose of the health checkup program, unwillingness to use preventive medicine, time constraints, and difficulties with access to general practices, and doubts regarding clinics as appropriate settings [22, 23].

With respect to health checkup type were associated with a significant difference in PCS and MCS scores among the public servants suggesting that those undergoing GPHC had higher PCS scores than are those undergoing SPHC. The mean scores of PCS (51.2) and MCS (43.4) in the present study were varied across different countries, US study (48.6, 53.1), and the Whitehall II cohort study (51.2, 51.1), respectively [24]. The score of PCS tended to similar to the Whitehall II cohort study while MCS was lower in our study, perhaps reflecting intrinsic cultural differences compared to the other two studies. Another reason is that public servants have been exposed to a high workload and job stress due to government organizations downsizing, manpower decreasing and the public demand increasing.

Our study shows that the mean scores of PCS and MCS of different types of health checkup decreased from Model 1 to Model 3 adjusted for demographics, job characteristics and life style as well as health status sequentially, and these factors are all related to HRQoL and health checkups. In model 1 and model 2, perhaps items included in GPHC are related to basic physical examination and laboratory tests, the score of PCS in GPHC are higher than the other groups, while the scores of MCS were lower in the NOHC group. This suggests that people with high health consciousness are more likely to have a desire for extensive health check-ups [25]. In addition, self-paid physical checkup programs may complement government-sponsored health screening programs and add value to free or even mild disease screening for health maintenance and help provide good post-checkup care [26‐28]. For the SPHC group, the fact that they may already have health problems means they tend to be willing to pay higher out-of-pocket expenses, especially if they have complementary private health insurance [29]. That implies having diseases or good health awareness and perception have an influence to lead people to seek preventive health checkups.

That people with multiple chronic conditions correspondingly had worse HRQoL than those with 1 or no chronic condition and that frequent physical distress was more common than frequent mental distress was consistent with previous studies [5, 30]. We found that public servants with more than 1 chronic disease condition who underwent GPHC are significantly more likely to report higher PCS scores than those undergoing SPHC. The pattern was GPHC was higher than SPHC and NOHC in score of PCS across 3 chronic conditions groups. Different subgroups, including those with no chronic disease, those with one chronic disease, and those with 2 or more chronic diseases, might exhibit different health-seeking behavior and health awareness. On the other hand, both PCS and MCS scores have a significant pair difference for each type of health checkup as the number of chronic conditions increases the score decreases and the two comparisons are statistically significant. Based on health utilization theory, periodic physical health checkup was an enabling factor of awareness of some disease and for disease literacy covering the knowledge of disease screening guidelines and risk factors [31]. A study suggests that low health literacy may affect behaviors necessary for the development of self-management skills [32]. Another consideration is that the specific disease and the diagnosis date were lacking in our data, we may extend our study to investigate the issue. Due to certain chronic diseases having more of an effect on up-to-date screening status than others for different cancers [33]. In some countries, health was an instrumental value exploited as an economic resource not only during periods of well-being but also during illness, by individuals not seeking preventive or timely health care because of the fear of losing their jobs [34].

This was the first nationwide study of public servants in Taiwan to investigate the association of preventive health checkup profiles with HRQoL, with both mean scores of PCS and MCS. The representative population was selected using PPS and consistent results were validated using both online- and paper-based questionnaires. Our results can aid in the implementation of an intervention program for high-risk groups and consequently be used for follow-up evaluations on the efficacy and effectiveness of the program. However, our study also has several limitations. First, establishing causality was difficult because of the cross-sectional design. Second, complete theoretical factors were not collected for each participant, such as components in the health belief model, self-rated physical factors, and mental outcome in the health utilization model. Third, chronic conditions play a critical role to motivate the seeking of health services as can be noted in Table 3. But the specific diagnosis time and disease need further investigation in future studies. Fourth, we cannot predict if some participants may receive a health checkup program that was covered by an insurance company not an out-of-pocket payment and resulted in the overuse of medical resources. Finally, the differences in preventive health checkup profiles between public servants and the general population encumber the comparison of PCS and MCS scores.