Family Practice Vol. 17, No. 4, 298-304
© Oxford University Press 2000
Socio-economic characteristics of adult frequent attenders in general practice: secondary analysis of data
a Sub-Unit for Medical Statistics, Nuffield Institute for Health and
b Centre for Research in Primary Care, University of Leeds, Leeds and
c Department of Primary Care and General Practice, University of Birmingham, Birmingham, UK.
PS Gill, Senior Lecturer, Department of Primary Care and General Practice, University of Birmingham, Birmingham B15 2TT, UK.
Scaife B, Gill PS, Heywood PL and Neal RD. Socio-economic characteristics of adult frequent attenders in general practice: secondary analysis of data. Family Practice 2000; 17: 298–304.
Received 17 September 1999; Accepted 13 March 2000.
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionReferences |
|---|
Objective. This study was carried out to determine the effect of a range of socio-economic features on frequent attendance in general practice from a large database of general practice consultations using two definitions of frequent attendance.
Methods. Secondary analyses were carried out of data from the Fourth National Survey of Morbidity in General Practice covering 60 general practices in England and Wales. A total of 283 842 adult patients and their consultations between September 1991 and August 1992 were examined. The main outcome measure was the odds ratio of being a frequent attender (95% confidence intervals).
Results. Using a definition of 12+ consultations/year, men were less likely to be frequent attenders (OR 0.14, 95% CI 0.13–0.17); however, the difference between men and women lessens with age. Patients who were more likely to be frequent attenders included those who were divorced or widowed (1.41, 1.31–1.51); from social classes IIIM (1.23, 1.17–1.29) and IV/V (1.33, 1.26–1.41); South Asian people (1.38, 1.16–1.65); or unemployed (1.61, 1.46–1.77). Other factors signifying isolation or poverty were also linked to frequent attendance. Using the definition of ‘6+ consultations for minor problems’ produced broadly similar results although the relative weight of the factors showed some differences.
Conclusions. Socio-economic factors were important indicators of frequent attendance in general practice. Results were very similar using either definition, suggesting that both are valid for further work. Furthermore, frequent attendance is a complex process associated with many factors outside the control of the GP.
Keywords. Frequent attenders, general practice.
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Whilst it has been established that higher rates of consultation are associated with a range of socio-economic factors,1,2 the precise relationship between these factors and frequent attendance is unknown. Establishing this relationship is important because a minority of patients do consult very frequently in general practice.3–6 This group of patients accounts for a disproportionate number of consultations; our analysis of data from the Fourth National Survey of Morbidity in General Practice (MSGP4)7 showed that the most frequently consulting 4.7% of patients used 21% consultations over 1 year; these findings are in keeping with those of smaller studies from the UK3 and abroad.8 Frequent attendance may be regarded as a problem in terms of this excessive workload as well as the feelings of ‘heartsink’ that some frequent attenders generate in their doctors,9 and the poor clinical or pastoral care that they may receive, despite their numerous consultations.
It has been well established that female gender, increasing age, physical and psychiatric morbidity, and somatization are all important factors in determining frequent attendance.2,10–13 However, there are only limited data about socio-economic factors which come from small studies in discrete geographical locations, whose results may be hard to extrapolate to a wider community.5,14
Defining frequent attendance, for research or clinical purposes, is problematic.4,15 Whilst most of the published work uses a definition based on a number of consultations in a fixed time period (ranging from 516 to 2017,18 in one calendar year), this is not ideal. It only represents a ‘snapshot’ of their consulting during this time period; some frequent attenders will not have attended frequently prior to the study period, and some will cease to attend frequently afterwards.19–21 Furthermore, there are considerable variations in the distribution of the numbers of consultations, and therefore of frequent attendance, between different cultural and geographical settings, and patients of different age, gender and levels of morbidity. The most widely used definition is between 10 and 13 consultations over a 12 month period;5,14,22–24 in most practices, this is likely to equate to the most frequent 3% of attenders.3 It represents, on average, about one consultation a month, a definition with which most GPs would concur as frequent. Alternative definitions include using a percentile of the most frequently attending patients in a given time period,12,25 or by age and sex;13 by a combination of criteria;20 or by frequent consultations for minor illness.17 The issue of how different definitions compare with each other using the same samples needs to be addressed.
The aim of this study is to establish the socio-economic indicators of frequent attendance from a large database of general practice consultations in order to inform the development of appropriate intervention strategies. In doing this, different definitions of frequent attendance will be tested and compared, enabling the best definition to be used in further work.
MSGP4 collected data from 60 volunteer practices between September 1991 and August 1992 on consultations with doctors and practice nurses, which were defined as "... a contact with a doctor or practice nurse whether at the surgery or at home".7 In this study, we considered multiple consultations on one day as a single attendance, as it was impossible to ascertain from the data whether they represented several problems at the same consultation or several consultations on one day. Socio-economic data were collected by interview from 83% of the patients, whether or not they made a consultation in the year of the study. This present study looks at data from the 283 824 adult patients who were in the study for the whole year for whom socio-economic data were available and the 1 013 560 attendances they made.
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Definitions of frequent attendance
In this study, a variety of definitions of frequent attendance were examined in detail; these produced a broadly similar pattern in the results. The two definitions for which results are presented were selected because they exhibit similarities to the various definitions and also draw out some of the more significant differences. The two definitions used were those with 12 or more consultations over the year (13 271 patients, 4.7%) and those who attended for >6 consultations for minor illnesses (11 291 patients, 4.0%) (Table 1
). Only 4772 patients were common to both definitions. The choice of these two definitions was based on the assumptions that patients attending 12 or more times a year would be seen as frequent attenders and that patients attending six or more times a year for minor illnesses would be perceived more readily as a frequent attender than patients attending as often for more serious complaints. The second definition takes into account, to some extent, the difference between a patient who attends frequently and a patient who is seen as a frequent attender. Twelve and six attendances were used for the cut-off points because they allowed comparison with the literature and they were felt to be of the ‘right order’.
|
Statistical methods
The data were analysed using logistic regression, using Stata v6.0. All socio-economic variables available were included in the analysis plus those interactions which it seemed clinically reasonable to consider. The coding of the variables for the analysis is given in Table 2
. Analyses of frequent and non-frequent attenders' attendances, using the ‘12+’ definition were undertaken for three sets of variables also recorded in the MSGP4 data set: location (home or surgery), severity of illness (severe, intermediate or minor; Table 1
) and the type of consultation (‘illness’ or ‘investigation’).
|
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Analysis
The coding of the data (Table 2
) is a result mainly of the data collection process used for MSGP4 and the anonymizing of those data prior to their release to us. Age and distance from practice, however, were split into quintiles for the analysis as it seemed unlikely that any effect from these variables would be linear. (The age data originally were given in 5-year age groups, hence the quintiles were not exactly equal.) The interactions used took into account: that age would have a different effect in men and women; that men and women may exhibit different behaviours due to their co-habitation status; that there may be an effect due to being a single parent; and that distance from the practice may have a different effect on patients living in urban and rural environments.
The results of the analysis by logistic regression for the ‘12+’ and ‘6+ minor’ definitions of frequent attender are shown in Table 3.
|
Interpretation of the 12+ attendances/year results
Whilst, overall, men were much less likely than women to be frequent attenders, the effect was much reduced for older men. Although increasing age reduced the odds ratio (OR) associated with being a frequent attender, the ORs associated with the interaction term ‘sex x age’ show how men's likelihood of being a frequent attender increase with age. This can be seen clearly in Figure 1
, which shows overall median attendance rates by sex and age. Being married did not significantly affect the OR of being a frequent attender; divorced or widowed patients, however, were more likely to be frequent attenders, with an OR of 1.41 [95% confidence interval (CI) from 1.31 to 1.51] compared with those who had always been single. Patients with children over 5 years of age were less likely to be frequent attenders, with an OR of 0.79 (0.75–0.84). Patients with children under 5 years, however, were more likely to be frequent attenders. When considering single parents, we had to take into account the variables ‘Parent of children under/over 5’ and ‘Sole adult in household’ and their interaction, added to the fact that they were not co-habiting. This gave an overall OR of 0.85 for single parents of children under 5 years and 0.88 for single parents of older children. Yet when marital status was taken into account (OR of 1.41 for ‘divorced or widowed’), we obtained a more interesting result: the OR was 1.20 for those with younger children and 1.24 for those with older children. Single parents who once had a partner were more likely to be frequent attenders than those who had always been single.
|
Other important findings were that patients from social classes IIIM, IV and V were more likely to be frequent attenders than those from social classes I, II and IIIN; South Asian and Afro-Caribbean patients were more likely to be frequent attenders than white patients, although the CI for Afro-Caribbean patients was very wide. Patients living in rented accommodation were more likely to be frequent attenders, as well as patients not in full-time work. The permanently sick had, unsurprisingly, the highest OR for frequent attendance (5.25, 95% CI 4.86–5.67). The unemployed and ‘other economically inactive’ (mostly retired or housewives) had similar and high ORs (1.61 and 1.67). Part-time workers' OR was between that of full-time workers and the economically inactive. Students were less likely to be frequent attenders. Patients living very close to the practice were more likely to be frequent attenders; the relationship of distance to likelihood of being a frequent attender, however, was clearly not linear. Frequent attendance was increased amongst urban patients a moderate distance from the practice (1.1–2.9 km) compared with rural patients the same distance away.
Differences in the 6+ minor attendances/year results
The OR associated with being male was lower than for the ‘12+’ definition. However, the interaction term that looked at being both male and older had higher ORs than for the ‘12+’ definition. There was weak evidence for a change in the direction of the OR for the ‘other ethnic’ group (mostly Chinese). It appears that the different ages of children had a more extreme effect on the odds of being a frequent attender under this definition, with the OR for children over 5 years being lower and that for children under 5 years being higher. The ORs for the economic status variables were generally lower; this was particularly marked for the permanently sick. Under this definition, distance from the practice had much less effect.
Differences in the consultations of frequent attenders
Frequent attenders defined using the ‘12+’ definition attended significantly more for illness rather than for investigation; the OR for a frequent attender attending for illness was 1.53 (95% CI 1.50–1.55) compared with a non-frequent attender (Table 4). Frequent attenders were also about twice as likely to be seen at home than other patients, with an OR of 2.07 (2.04–2.10) (Table 5). Frequent attenders, compared with non-frequent attenders, were less likely to attend for minor illnesses; the OR was 0.61 (0.60–0.61) (Table 6).
|
|
|
|
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
The results show that a variety of socio-economic factors strongly influence the chance of a patient being a frequent attender; these results are broadly in line with those of Carr-Hill et al.1 and Campbell and Roland,2 who looked at factors affecting the rates of consultation in the general population. The overall pattern of characteristics, apart from age and sex, which make a patient more likely to be a frequent attender indicates that social isolation and increasing poverty are the most important factors. Hence there are increased chances of being a frequent attender amongst the unemployed, housewives and pensioners. This effect is seen strongly amongst the secondarily single, particularly when there were dependent children. Social class, economic status and housing tenure all lead to increasing likelihood of frequent attendance as one moves towards the poorer end of these scales. The importance of the data presented here is that they demonstrate the significance of these factors in frequent attendance, rather than looking at increased consultation rates across the board. The results are very similar for both definitions of frequent attender, suggesting that for future work the use of either is valid.
There is an association between poor health and social deprivation, and in turn with frequent attendance, which many of the factors above indicate, although the nature of this link is unclear. This study suggests that further work is needed to establish the nature of this link. It is clear, however, that the likelihood of a patient being a frequent attender is strongly dependent on a set of social factors, without reference to physical or psychological status.
The data demonstrate that the ‘12+’ group have fewer consultations for minor illnesses; this may explain why they have more consultations in their own homes and more consultations for illnesses rather than investigations.
Whilst it is established that physical and psychiatric morbidity, and somatization are important determinants of frequent attendance, and therefore addressing these may form part of an intervention strategy, the ‘problem’ of frequent attendance is more complex than this, and is due, at least in part, to a manifestation of wider social problems. Many of the factors which increase the likelihood of frequent attendance are outside the control of the GP. This may go some way to explain the ‘heartsink’ felt by GPs when faced with frequent attenders:9 they may be unlikely to alter the attendance patterns whatever management they employ. It also suggests that wider actions suggested by the recent independent inquiry into inequalities in health25 may be more effective than interventions within primary care. This work therefore begins to answer some of the ‘how’ questions regarding intervention for frequent attendance, increasing calls to determine the why?
|
Acknowledgments |
|---|
We thank John Charlton for supplying the data set. This project was funded by a grant from Northern and Yorkshire R & D.
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
1 Carr-Hill RA, Rice N, Roland M. Socioeconomic determinants of rates of consultation in general practice based on the fourth national morbidity survey of general practices. Br Med J 1996; 312: 1008–1013.
2
Campbell SM, Roland MO. Why do people consult the doctor? Fam Pract 1996; 13: 75–83.
3 Neal RD, Heywood PL, Morley S, Clayden AD, Dowell AC. Frequency of patients' consulting in general practice and workload generated by frequent attenders: comparison between practices. Br J Gen Pract 1998; 48: 895–898.[ISI][Medline]
4 Neal RD, Dowell AC, Heywood PL, Morley S. Frequent attenders—who needs treatment? Br J Gen Pract 1996; 46: 131–132.
5
Heywood PL, Cameron Blackie G, Cameron IH, Dowell AC. An assessment of the attributes of frequent attenders to general practice. Fam Pract 1998; 15: 198–204.
6
Schrire S. Frequent attenders—a review. Fam Pract 1986; 3: 272–275.
7 Royal College of General Practitioners, Office of Population Censuses and Surveys, Department of Health. Morbidity Statistics from General Practice. Fourth National Study 1991–1992. London: HMSO, 1995 (Series MB5 No. 3.).
8 Browne GB, Humphrey B, Pallister R, Browne JA, Shetzer L. Prevalence and characteristics of frequent attenders in a prepaid Canadian family practice. J Fam Pract 1982; 14: 63–71.[ISI][Medline]
9 O'Dowd TC. Five years of heartsink patients in general practice. Br Med J 1988; 297: 528–530.
10 Sensky T, MacLeod AK, Rigby MF. Causal attributions about common somatic sensations among frequent general practice attenders. Psychol Med 1996; 26: 641–646.[ISI][Medline]
11 Portegijs PJM, van der Horst, Proot IM, Kraan HF, Gunther NCHF, Knottnerus JA. Somatization in frequent attenders of general practice. Soc Psychiatry Psychiatr Epidemiol 1996; 31: 29–37.[ISI][Medline]
12 Baez K, Aiarzaguena JM, Grandes G, Pedrodo E, Aranguren J, Retolaza A. Understanding patient-initiated frequent attendance in primary care: case control study. Br J Gen Pract 1998; 48: 1824–1827.[ISI][Medline]
13 Svab I, Zaletel-Kragelj L. Frequent attenders in general practice: a study from Slovenia. Scand J Primary Health Care 1993; 11: 424–430.
14
Karlsson H, Lehtinen V, Joukamaa M. Frequent attenders of Finnish public primary health care: sociodemographic characteristics and physical morbidity. Fam Pract 1994; 11: 424–430.
15
Neal RD. Frequent attenders. Fam Pract 1995; 12: 370.
16 Andersson SO, Mattson B, Lynoe N. Patients frequently consulting general practitioners at a general practice in Sweden—a comparative study. Scand J Soc Med 1995; 23: 251–257.[ISI][Medline]
17 Slater BCS. A study of morbidity in general practice with particular reference to adult frequent attenders. MD thesis. University of Edinburgh, 1964.
18 Robinson JO, Granfield AJ. The frequent consulter in primary medical care. J Psychosom Res 1986; 30: 589–600.[ISI][Medline]
19 Neal RD, Heywood PL, Morley S. Freight trains and supernovas: the use of a sorting task to determine patterns within long term frequent attendance to general practice. Primary Health Care Res Dev 2000; in press.
20 Gill D, Dawes M, Sharpe M, Mayou R. GP frequent consulters: their prevalence, natural history, and contribution to rising workload. Br J Gen Pract 1998; 48: 1856–1857.[ISI][Medline]
21 Karlsson H, Joukamaa M, Lahti I, Lehtinen V, Kokki-Saarinen T. Frequent attender profiles: different clinical subgroups among frequent attender patients in primary care. J Psychosom Res 1997; 42: 157–166.[ISI][Medline]
22 Semmence A. Chronic high users in a general practice: a preliminary study. J R Coll Gen Pract 1969; 17: 304–310.[Medline]
23 Wamosher Z. The returning patient—a survey of patients with a high attendance rate. J R Coll Gen Pract 1966; 11: 166–173.
24 Westhead JN. Frequent attenders in general practice: medical, psychological and social characteristics. J R Coll Gen Pract 1985; 35: 337–340.[ISI][Medline]
25 Independent Inquiry into Inequalities in Health. London: Stationery Office, 1998.
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||