Background
Healthcare-seeking behaviour is defined as “any activity undertaken by individuals who perceived themselves to have a health problem or to be ill for purpose of finding an appropriate remedy” [
1]. Healthcare-seeking behaviour includes the timing and types of healthcare service utilization and may affect population health outcomes [
2]. Delayed medical attention has been shown to associate with an increased risk of unfavourable outcomes [
3]. For patients with infectious diseases, delay in seeking care may also result in increased transmission risk in the community. Understanding the pattern of healthcare-seeking behaviour could help public health practitioners and policy makers to improve the healthcare system and health promotion strategies.
From a patients’ perspective, healthcare-seeking behaviour tends to be responsive to discomfort or symptoms, rather than to specific diagnosed diseases which were unknown to them before medical consultation. However, many studies examined healthcare-seeking behaviour either focused on a patient care perspective, or restricted to a specific disease related to a few limited symptoms [
4‐
6]. In this study, we focused on healthcare-seeking behaviour specific to symptoms and syndromes, which may more realistically reflect personal responses to sickness in the general population. Such data is still limited in the literature.
A previous study in Denmark showed that for patients with any symptoms, on average < 40% of the patients actually sought healthcare service, though the proportion varied substantially by symptoms [
7]. Here we reported the findings in Hong Kong which also has a well-developed healthcare system composed of both public and private sectors but with very different share in the outpatients and inpatients services: 70% of outpatient services were delivered by private sectors, whereas 90–95% of inpatient services were provided by public sectors [
8]. Also, Hong Kong has its unique mixed culture, which provides and promotes both western and Chinese medicine in the healthcare system. Western medicine has been widely accepted and is the dominant medical system for a long time, but the Hong Kong Government has also actively promoted the development of Chinese medicine.
The objective of this study is to describe the characteristics of healthcare-seeking behaviour due to different symptoms and syndromes related to respiratory and gastrointestinal-related diseases, such as the proportions of patients seeking medical consultation, types of healthcare service utilized, and time from symptom onset to consultation. Data describing healthcare-seeking behaviour could characterize the utilization of the healthcare services, and facilitate risk communication during outbreaks, planning of health care resources, and interpretation of practitioner-based surveillance system.
Discussion
We studied healthcare-seeking behaviour specific to symptoms, which allows interpretation and application of the results in the patient perspective for Hong Kong Chinese population. Our study found that nearly half of the participants reported infectious diseases-related symptoms over a 30-day period, and 41.4% of whom have sought medical consultation (Table
2). Consultation rate varied across symptoms, ranging from 14% due to fatigue, to 78% due to fever, and was usually higher among those with acute/infectious symptoms and lower among those with mild/chronic symptoms (Table
2). The consultation rates were highest in the children and lowest in young adults, suggesting that the working population is least likely to seek medical attention when having infectious disease-related symptoms.
An overall consultation rate of about 40% (Table
2) for symptomatic patients of respiratory and gastrointestinal-related infections suggested that the majority of patients were not captured by the healthcare system, forming the submerged part of the disease iceberg. Understanding the proportions of medically unattended patients may help policy makers for developing health campaigns targeting these individuals or estimating the full burden of disease.
In Hong Kong, the private sector is the major provider of primary care, delivering about 70% of outpatient consultations [
8], and CMP is used as the main alternative and complementary healthcare service in Hong Kong. In our study, we also found that western medicine is the preferred healthcare provider, contributing more than 80% of the consultations (Fig.
2). 16.7% of consultations visited CMP (Fig.
2). A local study showed that 85% of people who have sought medical consultation had consulted western medicine, while 10% had consulted CMP [
8]. Another study found that 8.8% of respondents who reported symptoms during the 30 days before survey had visited a CMP for the discomfort [
15]. In comparison, our finding shows that the preference for CMP may have increased slightly in the last decade with the promotion of Chinese Medicine by the Hong Kong Government. Many patients utilized both systems in parallel, taking western medicine to relieve symptoms and Chinese medicine to restore balance and health. In our study, 3.8% of participants had sought both western and Chinese medicine consultation for the same illness episode (Fig.
2). This could be interpreted as integrative medicine, or was in fact doctor shopping.
Participants had different preference on the type of health service according to their symptoms. Participants with acute symptoms favoured western medicine, whereas participants with gradually developing symptoms prefer to visit CMP. This preference could be explained by the common perception that western medicine is ‘powerful and quick’ comparing to CMP [
16]. Chan et al. found that older, poorer people who have chronic conditions were more sceptical of western physicians [
17]. In our study, we also found that older people having chronic symptoms such as low back pain, myalgia, and fatigue have 10–20% higher utilization of CMP than those of younger age. Considering western medicine only, our study found that patients favoured GP over GOPC regardless of their symptoms, consistent with a study showing that 76% of patients utilized primary care service provided by GPs [
4].
Meng et al. [
18] investigated the difference in healthcare-seeking behaviour of patients with ILI (defined as “at least two of the signs or symptoms [fever ≥37.8 ̊C, cough, sore throat, headache, or myalgia]”, more similar to the definition of ARI in our study) between summer and winter influenza epidemics. Meng et al. [
18] found that 25.0 and 38.6% of respondents reported ILI in summer and winter peak, respectively. Among those with ILI, 42.3 and 48.5% had sought medical care for each peak, respectively. In our study 64.0% of those with ARI sought medical care (Table
2), probably because our surveys were carried out closer to the influenza peak period. In a US study, 40 and 56% of the adults and children respectively who had ILI sought healthcare service during the 2009 H1N1 pandemic [
19], compared to 92 and 84% in our study (Table
2). Patients in Hong Kong were much more likely to seek medical attention when presenting with influenza-associated symptoms.
In our study, 91.7 and 75.4% of the children with ILI and ARI respectively sought medical consultation (Table
2). In Israel, 81.5% of the children under 13 year-old consulted a physician when they had flu-like symptoms [
20]. Both studies showed that children with flu-related symptoms would have a high consultation rate. Age difference in the consultation rate was statistically significant only for ARI (
p-value < 0.001) but not ILI (
p-value = 0.106), with adults having ARI noticeably less likely to seek medical consultation (Table
3). Comparing with ARI, ILI is more specific to influenza infection, and led to high consultation rates irrespective of age (Table
3). The high consultation rates due to ILI may result in school or work absence, which probably reduced influenza transmission risk in schools or workplace. In Hong Kong, medical certificate is required for taking sick leave according to the Employment Ordinance. Though this may not be strictly enforced for short sick leave of 1 or 2 days, the need of medical certificate for the working population cannot explain the lower healthcare-seeking behaviour among adults.
Previous studies showed that some influenza patients did not visit doctors. The proportions vary across countries, for example 55% of ILI patients in the US [
21], and 38% of cases of self-defined influenza in France [
22]. From our study, the proportions were lower in Hong Kong (10 and 35% for patients with ILI and ARI respectively, Table
2). Most of the influenza surveillance systems are established in the clinical settings, which limits its ability to fully capture the burden of ILI/ARI for patients who have mild symptoms or do not seek any medical consultation. Our findings may help to estimate the proportion not being captured in the surveillance system.
Few studies examined the duration between symptom onset to medical consultation for common infectious diseases, in particular with respective to specific symptoms. In our study, more than 60% of participants had sought medical care within 2 days from symptom onset (Fig.
3). A US study showed that among adults with seasonal influenza, 35 and 47% sought medical care within 2 days and within 3–7 days of illness onset respectively [
21], compared to our results for adults with ILI (65 and 35% respectively, combining age groups 16–54 years and ≥ 55 years in Fig.
3). The relatively short duration from illness to medical attention in Hong Kong may be attributed to easy access of medical service in a compact city. Delayed access to healthcare might be associated with longer hospital stays and poorer health outcomes [
23]. Shorter duration between symptom onset and medical consultation may allow patient to have more timely diagnosis and better health outcomes.
There are a few limitations in our study. First, our data had a relatively low response rate and might suffer from under-representation of the older population. We addressed this issue by applying post-stratification weighting methods. Second, some other factors that may affect symptom-specific healthcare-seeking behaviour such as self-medication, and vaccination status were not explored in this descriptive study. Third, there may be recall bias for reporting the illness in the past 30 days. We specifically asked the participants to report the latest illness episode, and provided a list of symptoms to minimize under-reporting. However, very mild and unattended symptoms could still be missed from the survey, especially for symptoms reported by parents of younger children. Fourth, there is seasonal variation in disease activities, the associated symptoms and potentially healthcare-seeking behaviour trigged by these symptoms.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.