At present, few studies have reported the medical institutions choice of hypertensive patients. Dafeng et al. investigated the health-seeking behaviors of 510 patients with hypertension or diabetes in Wuhan in China, and the results showed that 75.35% of patients chose higher-level hospitals for medical treatment . Another community-based study conducted in Guangdong province in China including 154,504 subjects who had a usual source of healthcare demonstrated that 65.3% of patients used outpatient services at primary care regularly for tackling chronic diseases . However, the combined health-seeking behaviors has not been reported in the two studies and it was somewhat different from the result of our study. The population in our study was all from the residents in the central urban area of Chengdu, a central city in western China, with good economic development and community health service construction. We can see that the patients' health-seeking behaviors are completely different due to unbalanced economic development and different maturity levels of community health service centers in China. Further researches need to be done in different regions at different levels of economic development. The results of our study can only represent the health-seeking behaviors of hypertensive patients in the central urban areas.
Health-seeking behaviors is defined as any activity undertaken by individuals who perceive themselves to have a health problem or to be ill for the purpose of finding an appropriate remedy . The desired health-seeking behaviors is responding to an illness by seeking help from a trained allopathic doctor in a recognized health care center . Musinguzi GS et al. found that factors influencing health-seeking behaviors were related to health systems and the patient socioeconomic and structural environment . The main system issues were related to availability and attitudes of staff and shortage of supplies and medicines . The patient factors were related to awareness, perceived severity, perceived effectiveness of therapy, adverse effects, and perceived fears of lifelong dependence on medicines . However, the influencing factors of different health-seeking behaviors are not clear at present. Neither of the two studies we mentioned before have conducted further researches on the factors influencing the different health-seeking behaviors [17, 18]. Our studies demonstrated that education levels and side effects of drugs had impacts on the patients' health-seeking behaviors. The patients choosing higher-level hospitals as the usual places for medical treatment had the largest proportion of middle and advanced education level, and the patients choosing the informal medical facilities as the usual places for medical treatment had the largest proportion of low education level. This is just the same as what Fjaer EL et al. found, that higher socioeconomic position groups are more likely to use health care specialists, compared with lower socioeconomic position groups, and in the context of health care specialist use, education and occupation appear to be particularly important factors . However, the proportion of patients with middle and advanced education was not the highest in the combined health-seeking group. It indicated that education level is not an important factor influencing the choice of combined health-seeking behaviors. Otherwise, the incidence of adverse effects was the highest in the combined group. Although studies have shown that adverse effect is an influential factor for drug adherence [24–26]. However, side effects of drugs may also prompt patients to seek medical treatment from community health service center to higher-level hospitals in our study,so as to receive combined therapy from community health service center and higher-level hospitals. It may also because patients in the combined management group pay more attention to the drug response.
The community-based epidemiological survey of hypertension showed that the overall control rate of hypertension was poor, as we mentioned before, just 5.7% in a population-based screening study  and 38.3% in urban areas, 17.5% in rural areas in another cross-sectional study in China . In the survey conducted in the outpatient, including 5,206 hypertensive patients from 46 hypertension outpatient clinics in 22 provinces, autonomous regions, and municipalities of China, the control rate of hypertension was 44.3%, higher than the control rate obtain in the community . Our study showed that the BP control rate of hypertensive patients who chose higher-level hospitals as the usual places for medical treatment was 43.0%, which was similar to the control rate in previous studies . The control rate of hypertensive patients who chose community health service center as the usual place for medical treatment was 39.4%, which was close to the control rate in group choosing higher-level hospitals. In addition, patients with hypertension who chose the informal medical facilities as the usual places for medical treatment had the lowest BP control rate, which was only 23.8%. The control rate was the highest in the combined management group, reached 54.8%, higher than that in other studies [4, 5, 27], and the average SBP was 136.3/72.6 mmHg in this group. Why was the level of BP the lowest in the combined management group? As we can see, there was no significant difference in the number of antihypertensive drugs among the four groups. And the proportion of combined medication use was relatively higher in the group choosing the higher-level hospitals for medical treatment, but the BP control rate was lower than that in the group with combined health-seeking behaviors. Therefore, it was not the number of drugs that makes the difference in BP. The rationality of prescription and the timeliness of prescription adjustment may play a more important role. As we all know, the drug regimens cannot be more precisely adjusted in patients treated in community health service center when the BP is difficult to control, and the medical problems cannot be solved in time sometimes in patients who simply choose the higher-level hospitals for medical treatment. In the combined management group, patients can not only receive medical services from the community health service center conveniently, but also enjoy better medical services from higher-level hospitals. They would like to visit higher-level hospitals if their BP can’t be controlled to normal in the community health service center and they had the highest medication compliance. However, the loose collaborative management in our study have not significantly improved the BP control rate of hypertensive patients. Yi Q, et al. randomly chose 218 primary hypertensive patients from a community health service center to investigate the effect of the integration pattern of hospital-community on the grade-based management for hypertension in elders. After 6 months of intervention, the control rate in the patients elevated from 22.9–88.1% , higher than the BP control rate of the combined management group in our study. But there was no control group, and the evaluation of health economics and the concrete working mode should be further discussed.
Our study has some limitations. First, the sample size was not large enough to get statistical differences in BP control rate among the 4 groups. But there were statistically significant differences in BP levels among the 4 groups. Although we conducted the survey using a random cluster sampling method, two thirds of the hypertensive patients went voluntarily to the community health service center to receive physical and serological examinations. Those seeking medical help regularly were sequentially enrolled to fill out the questionnaire on health-seeking behaviors. Therefore, there may be some selection bias in our study, which may have an impact on the composition ratio of health-seeking behaviors, but it can also reflect the influence of different health-seeking behaviors on BP. The third, it was a cross-sectional community sampling survey to evaluate the influence of different health-seeking behaviors on BP control rate. As the control rate of BP was also affected by many other factors which could not be corrected one by one, it could only provide a reference for further studies in the future. It is important to further study the effect of hospital-community combined hypertension management on the improvement of the control rate of hypertension and further explore the reasonable combined management mode. In addition, the survey was conducted in the central urban area of Chengdu, and the construction of the community chronic disease management system is relatively complete, it can only represent the characteristics of such type communities. The appropriate management pattern of hypertension in different regions of China still needs to be further studied.