Pathways to care for alcohol use disorders

Transcription

1 Journal of Public Health Medicine Vol. 21, No. 1, pp Printed in Great Britain Pathways to care for alcohol use disorders M. J. Commander, S. O. Odell, K. J. Williams, S. P. Sashidharan and P. G. Surtees Abstract Background The aim of the present study was to examine access to care for people with alcohol use disorders. Method An alcohol screening questionnaire was completed by 444 respondents in a community survey. During a designated week, 1009 patients presenting in primary care were assessed by their doctor and 773 of these completed the same questionnaire. Over a six month period 223 people with alcohol use disorders were identified using specialist addiction and psychiatric services, of whom 58 were admitted to hospital. One month prevalence rates of alcohol morbidity were determined for people aged between 16 and 64 years at all five levels in the pathways to care model. Results Around half the people with alcohol morbidity in the community never consulted their general practitioner and of those who did only half had their problem identified. Case recognition was particularly poor for women, young people and Asians. The main filter to people accessing specialist services came at the point of referral from primary care. This was especially marked for young people and for ethnic minorities. Conclusions Strategies are required to improve the identification and treatment of alcohol morbidity in primary care. Deficits in access to specialist services for women, young people and ethnic minorities need to be addressed. Keywords: access, pathways, alcohol, primary care Introduction Along with primary prevention strategies, such as increasing the cost of alcohol, 1 the early identification and management of alcohol use disorders in primary care offers one of the main opportunities for reducing alcohol related problems 2 and meeting the Health of the Nation target with respect to alcohol consumption (namely, to reduce the proportion of men drinking more than 21 units by 10 per cent and the proportion of women drinking more than 14 units by 4 per cent by the year ). However, despite evidence pointing to the effectiveness of simple interventions, 4 general practitioners (GPs) often fail to detect alcohol use disorders and many lack enthusiasm for treating these conditions. 5 Furthermore, the impact of efforts to re-focus specialist addiction services outside of hospital settings and to improve their integration with primary care and nonstatutory agencies remains unclear. Epidemiological data are required to identify barriers to care and to guide the planning of alcohol services. Goldberg and Huxley s pathways to care model 6 has been used as a framework for examining access to mental health care. Five levels of morbidity are represented (from the community through primary care to specialist services) with each separated by one of four filters (see Table 1). We aimed to examine differential access to services for selected sub-groups in the population with alcohol morbidity using this approach. Method The study was undertaken in a deprived inner-city health district (the old West Birmingham Health District). The district has a substantial ethnic minority population (14 per cent of the adult population being black and 23 per cent Asian) and is consistently ranked in the top ten most deprived districts in England. 7 Epidemiological surveys of adults (16 64 years) living in the district were undertaken in three settings: specialist addiction and psychiatric services, primary care and the community. Survey of specialist services All residents using specialist services on a specified day (23 March 1992) and over the following six months were identified. This included people using addiction services (the Regional Addiction Unit being located on the site of one of the two 1 Academic Unit, Northern Birmingham Mental Health Trust, Trust Headquarters, 71 Fentham Road, Erdington, Birmingham B23 6AL. 2 Addiction Treatment Unit, 44 London Road, Gloucester GL1 3N7. 3 MRC Biostatistics Unit, Institute of Public Health, University Forvie Site, Robinson Way, Cambridge CB2 2SR. M. J. Commander, 1 Senior Research Fellow, University of Birmingham and Hon. Consultant Psychiatrist, Northern Birmingham Mental Health Trust S. M. Odell, 1 Research Associate, University of Birmingham. K. J. Williams, 2 Consultant Psychiatrist with special responsibility for substance misuse, Severn NHS Trust. S. P. Sashidharan, 1 Professor of Community Psychiatry, University of Birmingham and Hon. Consultant Psychiatrist, Northern Birmingham Mental Health Trust P. G. Surtees, 3 Member of Scientific Staff MRC BSU and Hon. Clin. Psychol. to the Addenbrookes NHS Trust Address correspondence to M. J. Commander. Faculty of Public Health Medicine 1999

2 66 JOURNAL OF PUBLIC HEALTH MEDICINE Table 1 Pathways to care model Level LEVEL 5 In-patient care LEVEL 4 Specialist service use LEVEL 3 Primary care: conspicuous morbidity LEVEL 2 Primary care: total morbidity LEVEL 1 Community Filter FILTER 4 admission FILTER 3 referral FILTER 2 case recognition FILTER 1 decision to consult Source: Ref. 6. psychiatric hospitals included in the survey) as well as people in contact with psychiatric services receiving a primary diagnosis of an alcohol use disorder. Data were also obtained on people admitted to hospital during the same period. All points of contact with these services were ascertained. A range of demographic and clinical data were collected from staff using a specially designed proforma. The draft version of ICD-10 8 was adopted for the diagnostic classification. Primary care survey In addition to data on their practice lists held by the Family Health Services Authority, we used information obtained from a postal survey of GPs (n = 125) working in practices located within the district (39 single-handed and 27 group practices) to select five representative practices (three single-handed and two group practices). 9 All residents who attended non-specialized surgeries at these practices during one week were asked to complete a questionnaire as they waited to see the doctor. This included demographic items corresponding to those collected in the associated surveys as well as an alcohol screening questionnaire, the CAGE. 10 The latter consists of the following four questions (answered either yes or no): During the last month (1) Have you felt that you ought to cut down on your drinking? (2) Have people annoyed you by criticizing your drinking? (3) Have you felt bad or guilty about your drinking? (4) Have you had a drink first thing in the morning to steady your nerves or get rid of a hangover? The conventional threshold of two or more affirmative replies was used to distinguish case from non-case. After the consultation, and unaware of the patient self-report ratings, GPs completed a modified version of the World Health Organization triaxial classification of problems presenting in primary care. 11 This included a tick box item on the presence or absence of symptoms of alcohol misuse and a section to record a diagnosis of an alcohol use disorder. The practices were surveyed between December 1992 and March Community survey A computer-generated random sample of residents was drawn from the Family Health Services Authority database. After seeking the GPs permission, all eligible subjects were contacted by letter and subsequently approached. There was no replacement for people who either refused or could not be contacted. A poor response rate in the pilot study prompted us to offer 10 to each participant. This was paid on the return of a questionnaire incorporating items used in the primary care survey, including the CAGE. The study was completed between December 1994 and May In the three surveys, ethnic group codes were based on the OPCS 1991 Census classifications. The term Asian is used here to include people identified as Indian, Pakistani or Bangladeshi and does not include Chinese or Vietnamese. Self-reported ethnicity was used for the primary care and community surveys but in the survey of specialist services ethnicity was sometimes assigned by staff. The local Ethics Committee gave approval to each of the surveys. Analysis Data were analysed using the Statistical Package for the Social Sciences 12 and the Confidence Interval Analysis program. 13 Levels of significance were tested using x 2 where appropriate. Confidence intervals were estimated at the 95 per cent level. To allow comparison of morbidity rates across all levels in the pathways model estimates of the one month period prevalence of alcohol morbidity were unified across all the studies to coincide with the assessment time frame used for the CAGE. The ratios of the rates at successive levels were determined to assess the likelihood of people passing through the intervening filters. Results Morbidity rates at all levels in the pathways to care model The overall results of the three surveys are presented in Table 2 alongside corresponding data for access to care for people with mental health problems in the same district. 9

3 CARE FOR ALCOHOL USE DISORDERS 67 Table 2 Pathways to care for alcohol use disorders* Alcohol Mental health Survey Method Rates Ratio Rates Ratio LEVEL 5 Specialist services Case register 1 11 FILTER LEVEL 4 Specialist services Case register FILTER LEVEL 3 Primary care GP rating FILTER LEVEL 2 Primary care CAGE FILTER LEVEL 1 Community CAGE *One month rates/ Levels 1 and 2 based on GHQ-30 scores. 9,15 Ratio of level n to level n +1. Survey of specialist services The numerator for the one month prevalence rate was based upon people using specialist services on a specified day (n = 156) and an adjusted monthly number determined from those who contacted services during the following six months (n = 67/ 6). The number using in-patient care was calculated in the same way (i.e. nine in-patients on the specified day plus one-sixth of the 49 patients admitted during the following six months). Data from the 1991 Census, corrected for under-enumeration, 14 provided the denominator (n = ). Primary care survey Total morbidity. Of the 1123 eligible patients, 84 were unable to complete the questionnaire in English. Of those remaining, 266 (26 per cent) either refused to participate or did not complete sufficient responses on the CAGE to be designated a case or non-case. When those who completed the CAGE were compared with those who did not a similar response rate was found for men and women. However, people aged years were significantly more likely to respond than those aged (77 per cent vs 70 per cent; x 2 = 7.4, df = 1, p = 0.007) and whites (89 per cent) more so than either blacks (69 per cent) or Asians (52 per cent; x 2 = 85.3, df = 2, p < ). Of the 773 patients who completed the CAGE, 63 were cases (8 per cent). Conspicuous morbidity. Data were collected from the 19 participating GPs on 1009 patients, of whom 34 were identified as having symptoms of alcohol misuse or an alcohol use disorder (3 per cent). To compare morbidity rates across different levels of care, rates are expressed as a percentage of the population at risk. Estimates of the population rates for levels 2 and 3 were calculated by multiplying the percentage of cases by the proportion of the population consulting their doctor in the previous month; 36 per cent of those interviewed in the community survey. Community survey To achieve a target number of 500 interviews, 1500 people were sampled. Of these, 95 could not be contacted and 585 were excluded, most often because either they did not live at the given address (n = 362) or their GP refused permission (n = 135). A further 91 people were excluded because they were unable to complete interviews in English. Of the remaining 729 eligible subjects, 444 (61 per cent) answered sufficient questions on the CAGE to be coded as a case or non-case. The likelihood of completing the CAGE did not vary significantly according to age, sex or ethnicity. Of those who completed the CAGE, 25 were cases (6 per cent). Overview Only half the people identified as cases using the CAGE consulted their GP and of those that did only half had their condition recognized (Table 2). However, the main filter to accessing specialist care came at the point of referral, only around one in ten people identified as having symptoms of alcohol misuse or an alcohol use disorder in primary care going on to use specialist services. Demographic factors Men and women were found to be equally likely to consult their GP and to be referred to specialist services but women less often had their alcohol problem recognized by their doctor (see Table 3). Young people were least likely to consult, to have their problems detected and to be referred to specialist services. Asians were highly likely to consult their GP whereas blacks were least likely to do so. However, case identification in primary care was poor for Asians, and both Asians and blacks

4 68 JOURNAL OF PUBLIC HEALTH MEDICINE Table 3 Pathways to specialist alcohol services by selected demographic factors* Sex Age (years) Ethnicity Men Women Asian Black White LEVEL FILTER LEVEL FILTER LEVEL FILTER LEVEL Ratio level *One month rates per were far less likely than their white counterparts to be referred to specialist services. When the intervening filters between people with alcohol morbidity in the community and specialist services were examined simultaneously (see Table 3; Ratio level 1 4) access to care was by far the worst for women, young people and ethnic minorities. Discussion It is necessary to be cautious when drawing conclusions from prevalence surveys and in particular when making comparisons across surveys. Nevertheless, the method adopted here uniquely offers a comprehensive description of access to care for alcohol problems. The composite study carries the specific limitations of the three separate surveys (covered in greater detail elsewhere 9 ). The aims of the study were wide ranging. 9 Consequently, the numbers in some sub-groups are small and the opportunities for addressing potential confounding factors limited. The response rate between sub-groups varied and was especially poor for Asians in primary care. Although the same time frame was utilized, methods of case identification varied across the surveys. This inevitably restricts the opportunity to examine the influence of clinical factors on the permeability of the filters and potentially confounds comparisons made across groups. Despite being validated for use in primary care, 16 the CAGE does not provide a gold standard assessment of alcohol use disorders and the methods used to determine morbidity at levels 3 and above also fall short of ideal. Our results are remarkably similar to those reported more than 20 years ago by Edwards et al., who discovered that only per cent of people suffering from alcohol abuse were in contact with appropriate services. 17 As with the findings on access to care for people with mental health problems in the community, 9 around half the people with problems never see their GP and of those that do only half have their condition recognized. Also comparable is the fact that the main filter to reaching specialist services occurs at the point of referral from primary care (although the impact of clinical factors is likely to be particularly important here). Access to in-patient care is especially stringent, yet still almost one in ten people using services is admitted to hospital. Our data ignore the increasingly important influence of nonstatutory agencies and could potentially paint an unduly gloomy picture. Yet only 3 per cent of patients with identified morbidity in primary care were referred to such services by their GP, suggesting that they are not currently making a substantial impact on the model described. It would appear then that little has changed in the past two decades. 17 The community morbidity rates are largely consistent with existing findings 1,18 as are the low levels of access to specialist care for women and young people. 19 It is evident from our results that the latter has as much to do with problems surrounding case recognition and referral practices in primary care as with the lack of acceptable specialist provision for these groups. Although at odds with the Office of Population Censuses and Surveys (now Office for National Statistics) national survey of psychiatric morbidity, 18 the higher morbidity rate found in Asians is in accord with evidence that Sikhs are more likely to be regular drinkers than whites and that few Muslims drink at all; 20 none of our Asian cases were Muslim. Access to care was impeded for blacks by their failure to consult their GP and for Asians by poor case recognition in primary care. Also, compared with whites both Asians and blacks were under-served by specialist services. The reasons for this are unclear but are more likely to stem from deficiencies in existing health provision than because of a greater uptake of services on offer from non-statutory agencies. 21 The UK Government has recently called for a strategy on alcohol to set out a practical framework for a responsible approach. 22 The current study highlights some areas of weakness in existing health provision. It remains the case that most dedicated alcohol interventions are undertaken in specialist settings, and steps need to be taken to extend services

5 CARE FOR ALCOHOL USE DISORDERS 69 away from the hospital to the community if they are to make a significant population-wide impact. 4 GPs clearly occupy a central role in this process and need to be better equipped both to detect and to manage alcohol use disorders in primary care. 23 New community based services must be closely linked with primary care and should be especially alert to shortcomings in provision for women, Asians and blacks and young people. References 1 Anderson P. Alcohol as a key area. Br Med J 1991; 303: Edwards G, Unnithal S. Alcohol misuse. In: Stevens A, Raftery J, eds. Health care needs assessment. Volume 2. Oxford: Radcliffe Medical Press, 1994: Department of Health. The health of the nation. Technical supplement. London: HMSO, Edwards G. Alcohol policy and the public good. Addiction 1997; Supplement 1: S73 S79. 5 Unnithan S, Ritson B, Strang J. Organising treatment services for drug and alcohol misusers. In: Chick J, Cantwell P, eds. Seminars in alcohol and drug misuse. London: Royal College of Psychiatrists, 1994: Goldberg DP, Huxley P. Mental illness in the community: the pathway to psychiatric care. London: Tavistock, Smith P, Sheldon TA, Martin S. An index of need for psychiatric services based on in-patient utilisation. Br J Psychiat 1996; 169: World Health Organization. ICD-10. Chapter V, mental and behavioural disorders. Diagnostic criteria for research. Draft for field trials. Geneva: WHO, Commander MJ, Sashidharan SP, Odell S, Surtees PG. Access to mental health care in an inner city health district. 1. Pathways into and within psychiatric services. Br J Psychiat 1997; 170: Mayfield D, McCleod G, Hall P. The CAGE questionnaire: validation of new alcoholism screening questionnaire. Am J Psychiat 1974; 131: Clare A, Gulbinat W, Sartorius N. A triaxial classification of health care problems presenting in primary health care. Social Psychiat Psychiat Epidemiol 1992; 27: SPSS. SPSS for Windows. Release 6.0. Chicago, IL: SPSS Inc., Gardner MJ, Gardner SB, Winter PD. Confidence interval analysis. Version 1.2. London: BMJ Publishing, Office of Population and Census Surveys Census user guide no. 58: Undercoverage in Great Britain. London: OPCS, Goldberg DP, Williams P. A user s guide to the general health questionnaire. Windsor: NFER-Nelson, King M. At risk drinking among general practice attenders: validation of CAGE questionnaire. Psychol Med 1985; 16: Edwards G, Hawker A, Hensman C, Petro J, Williamson V. Alcoholics known/unknown to agencies: epidemiological studies in a London suburb. Br J Psychiat 1973; 123: Meltzer H, Gill B, Petticrew M, Hinds K. OPCS surveys of psychiatric morbidity in Great Britain. Report 1. The prevalence of psychiatric morbidity among adults living in private households. London: HMSO, Thom B. Social factors and alcohol abuse. In: Bhugra D, Leff J, eds. Principles of social psychiatry. Oxford: Blackwell, 1992: Cochrane R, Bal S. The drinking habits of Sikh, Hindu, Muslim and white men in the West Midlands: a community survey. Br J Addiction 1990; 85: Cochrane R, Sashidharan SP. The mental health needs of ethnic minorities. In: Ahmad W, Sheldon TA, eds. The health needs of ethnic minorities, CRD report no. 6. York: University of York, 1996: Department of Health. Our healthier nation. A contract for health. a consultation paper. London: HMSO, Anderson P. Effectiveness of general practice interventions for patients with harmful alcohol consumption. Br J Gen Pract 1993; 43: Accepted on 22 September 1998