Introduction

Frequent attenders, who are also called high users or high utilisers, are common in general practice. They were first described in a paper by Backett, an English general practitioner, in 1954 showing that most of the workload of a doctor was attributed to a small proportion of patients.¹ Since then, many studies have proved the high use of health care by frequent attenders.^2-6 According to these studies, frequent attenders represented 5-15% of the population, and used about 20-40% of health care services.^2-6

What does the term frequent attenders mean? A literature review in 2005 found that there was no generally accepted definition of frequent attendance.⁷ It is difficult to find one single explanation for the phenomenon of frequent attenders. Previous studies have shown that women are generally known to visit physicians more than men, and thus have a greater likelihood of being frequent attenders.⁸ There have also been attempts to identify subgroups of frequent attenders, such as Katon's "distressed high utilizers",⁹ and Karlsson's five patient groups (patients with entirely physical illness, patients with clear psychiatric illnesses, crisis patients, chronically somatising patients, and patients with multiple problems).¹⁰ Overall, the findings from these studies have indicated that frequent attenders comprise a heterogeneous group who may have a genuine need for care, or a troublesome group who simply create an unnecessary workload.¹¹

The first difficulty to define frequent attenders is to define the cut-off point in the number of consultations. The second difficulty is to define the observation period. Most of the studies have used more or less arbitrary numerical definitions of frequent attendance, varying from five to twenty consultations per year.^2-6,10,12,13 One study defined frequent attenders as patients having an annual rate of consultation over twice as high as the practice's sex- and age-related mean.¹⁴

Unnecessary attendances in general practice not only produce excessive workload, but also have such impact on the resources of our healthcare services. Furthermore, there can be significant psychological impact on family physicians since frequent attenders are often perceived as demanding, difficult, and frustrating patients with limited prospects for improvement.^15,16 A full understanding of the characteristics of frequent attenders would be important in planning and improving a management system. As local data are lacking, this study attempted to determine the prevalence of frequent attenders and to analyze the associations between the presence of chronic illness, socioeconomic factors, and the frequency of attendance in the patients with frequent attendances.

Method

This survey was carried out in a general outpatient clinic (GOPC), in a busy part in Hong Kong, which serves about 150,000 patients a year.

Definition of frequent attenders

In this clinic, patients with chronic illnesses are followed up once in every two months, i.e. six times per year. As there is no internationally or locally agreed definition of frequent attendance, a value that is two times the number of follow-ups for patients with chronic illnesses was used as the operational definition. Thus, patients who had attended 12 times or more in a year were considered as frequent attenders.

Study design and patient selection

Data were collected retrospectively from the Clinical Management System (CMS). A list of patients who had attended our clinic twelve or more times between 1^st January and 31^st December 2005 was retrieved from the CMS. For each frequent attender, the following demographic information was collected:

• Age
• Gender
• Payment status: entitled patients (HK$45/consultation), non-entitled patients (HK$215/consultation), comprehensive social security allowance recipients (CSSA) (free), government servants (free), hospital authority staff (free)
• Marital status: married, single, separated/divorced, or widowed
• Occupation: professional, skilled, manual, or nil
• Education level: tertiary or above, secondary or below

In addition, the reason for attendance was collected, in particular whether the encounter was for an acute or chronic problem. Each presenting problem was classified according to the International Classification of Primary Care (ICPC) codes. ICPC coding rates in our clinic are above 99%, accompanied by regular ICPC standardization.

Control group

Data from the Demographic and Statistics Station from the Census and Statistics Department, Hong Kong,¹⁷ was used when comparing data on gender, age, marital status, occupation, and education level. As the Census and Statistics Department does not have any data on payment status, all subjects attending this clinic during the study period, apart from those satisfying the criteria for being frequent attenders, were used as the reference group for purposes of comparsion.

Outcome measures and statistical methods

Data were analysed with the SPSS statistical package. Frequent attenders were categorized according to their demographic differences and presenting illnesses. Differences between groups were analysed by means of T-tests and odds ratios (at 95% confidence intervals).

Prevalence of frequent attenders

Among 35007 attending patients during the study period, 393 (1.1%) patients were classified as frequent attenders. These 393 patients had a total of 5844 consultations, accounting for 4.8% of the total number of consultations.

Subgroups of frequent attenders: episodic, chronic and mixed

Of these 393 patients, 129 patients presented with episodic complaints alone and 264 patients presented with chronic illnesses (Figure 1). Of those patients with chronic illnesses, 230 (85%) also had minor complaints (mixed group) while 34 patients attended for chronic illness only. Patients with chronic illness were more likely to attend frequently [OR: 2.03 (CI 1.65-2.50)] compared to the episodic group.

Frequent attenders with episodic complaints were significantly younger (mean age 53 years) than those with chronic illness, and this was from the comparisons with chronic illness only group (mean age 69 years, p <0.001) and the mixed group (mean age 68 years, p <0.001).

Characteristics of frequent attenders

Characteristics of the episodic group included female, ³ 60 years old, manual worker, an education level of secondary schooling or below, CSSA and staff. A tertiary or above education level was a "protective" factor for frequent attendance (Table 1).

In the chronic illness only group, females, ³ 60 years old, entitled patients and those without occupations were more likely to be frequent attenders, while married and employed patients were less likely to be frequent attenders.

In the mixed group, males, ³ 60 years old, manual workers, no occupation, an education level of secondary or below, CSSA and staff were characteristics of frequent attenders. Married, employed patients, and patients with an education level of tertiary or above were less likely to be frequent attenders.

Number and nature of complaints (Tables 2 & 3)

In the episodic group, each patient on average attended 16 times with 1.2 complaints per consultation. In the chronic illness only group, each patient attended 13 times with 1.9 complaints per consultation. In the mixed group, each patient attended 15 times with 1.6 complaints. Whilst in the control group, each patient attended 3.5 times with 1.5 complaints per consultation.

Among the 129 patients who attended for episodic complaints only, the most common disease group was respiratory system (29%) with upper respiratory tract infection accounting for all cases. The next most common problems included musculoskeletal (28%), skin (14%), and gastro-intestinal (10%). Among the 34 patients who attended for chronic illness only, the most common disease group was cardiovascular (52%) followed by endocrine (27%). In the mixed group, the commonest disease group was cardiovascular (31.7%), followed by endocrine (17.8%), respiratory (17%) and musculoskeletal (9%).

Prevalence

In this study, 1.1% of patients were frequent attenders. Similar studies conducted overseas have reported that the prevalence of frequent attenders varied from 1.7% to 8%.^2-6,10-13 As the criteria used for defining frequent attenders were different in different studies, this variation may well be an artifact. As there were no local data for comparison, further studies with multi-centres involvement are needed to establish the prevalence of frequent attenders in Hong Kong.

Socioeconomic characteristics of frequent attenders

This study concentrated on the associations between the presence of socioeconomic factors and the frequency of attendance among frequent attenders. Results were compatible with the findings from many overseas studies^{5,6,8,14,18,19} in that females, older aged group, those with lower education level, and those without an occupation were more likely to be frequent attenders, whereas being married had a protective effect in preventing patients from becoming frequent attenders.

It has been postulated that people on CSSA might abuse the medical system.²⁰ This study found that CSSA patients were more likely to be frequent attenders, but do they really abuse the system? These patients may simply represent a group that have more social insecurity and therefore develop more psychosocial and physical problems. In order to explore whether abuse of health services characterises their behaviours, further qualitative studies are needed to identify the underlying reasons of their encounters.

In the chronic illness only and mixed groups, the vast majority of those not working were likely to be retired as reflected by their age. However, the lower mean age of the episodic group demonstrated that this was likely a heterogeneous group comprised truly unemployed patients as well as those who had retired. One of the suppositions why patients who were not working attended more than the others is that they might be more likely to have experienced stressful life events. Stressful life events test a person's coping ability. The more distress a person feels, the more likely the person is to see a doctor, especially if their social support is poor.^21,22 On the other hand, a well-prepared retirement may be a protective factor. As retirement and unemployment have different effects on patient's behaviours, the exact impact of these two factors on the frequency of attendance could only be clarified by further studies.

Low education level could imply poorer understanding of one's illnesses and thus, poorer management skills. This in turn would lead to increased attendance.

Figure 2: Gender in episodic vs chronic only vs mixed group

Figure 3: Age distribution in episodic vs chronic only vs mixed groups

Nature of illness of frequent attenders

The main reasons for frequent attenders to consult a doctor are their physical complaints.¹⁰ One overseas study found that respiratory symptoms, gastrointestinal symptoms, and back pain were more common in this group of patients.²³ In this study, upper respiratory tract infections (URTI) contributed to 29% of the total consultations in the episodic group. Over 90% of URTI is self-limiting, also most of it is caused by viruses that do not require any curative treatment.²⁴ The findings in this study reflect a lack of understanding of the nature of illness within at least part of the public. In order to reduce the burden caused by this group of patients, education on this illness should be addressed via primary care physicians as well as the media.

Chronic illness has been known to be a potent risk factor for frequent attendances.¹⁸ In this study patients with cardiovascular or endocrine disease attended more frequently. As Hong Kong's population is ageing, prevalence of patients with chronic illnesses will increase. Local data have revealed that the prevalence of hypertension will likely increase to 26.5% in males and 35.7% in females as age increases from 65 to 74 years.²⁵ It is estimated that cardiovascular disease will become the number one cause of mortality by the year 2020.²⁶ In order to reduce the workload from managing these types of chronic illnesses, primary care physicians have an important role in setting preventive measures and promoting healthy lifestyles.

Overseas studies have also found mental disorders common among frequent attenders (24% to 52% being depressive).^9,10,27 However, in our study, only 3 to 4.4% of complaints in the chronic illness group were identified as psychiatric in origin. One logical question is whether psychiatric illness is under-recognised or if these patients are presenting with somatic complaints that have masked their psychiatric illnesses? If emotional problems are detected and tackled earlier, related frequent attendances could be prevented or at least reduced to more effective consultations.

Limitations

There were two main limitations to this study. Firstly, the sample size of the study was small. Secondly, it was only carried out in one general outpatient clinic, thus, reducing the generalisability. These shortcomings were addressed if the study was carried out in more clinics across Hong Kong Island, Kowloon, and New Territories.

Conclusion

In this study, age, gender, and socioeconomic factors were identified as important characteristics of frequent attenders. Although data from the local Chinese population were lacking, these results are consistent with overseas data. Future qualitative studies are needed in order to explore why these subgroups attend more frequently than other patients.

1. Frequent attenders are common in general practice. They comprise a heterogeneous group who may have a genuine need for care, or a troublesome group who simply create an unnecessary workload.

2. Unnecessary frequent attendance in general practice not only produces excessive workload, it also has a great impact on the resources of our healthcare services.

3. Age, gender, and socioeconomic factors were identified as important characteristics of frequent attenders.

4. Respiratory symptoms, gastrointestinal symptoms, and back pain were more common in frequent attenders. In order to reduce the burden caused by these groups of patients, further education targetting these illnesses would be essential.

5. Future studies with collaboration from other outpatient clinics across Hong Kong Island, Kowloon, and New Territories will help us further understand the profile of frequent attenders in Hong Kong.