- Email: [email protected]
General practice or drug clinic for methadone maintenance? A controlled comparison of treatment outcomes
International Journal of Drug Policy 12 (2001) 81 – 89 www.elsevier.com/locate/drugpo
General practice or drug clinic for methadone maintenance? A controlled comparison of treatment outcomes David Lewis a,*, Mark Bellis b b
a Vauxhall Primary Health Care, Limekiln Lane,Li6erpool L5 8XR, UK Public Health Sector, School of Health, Li6erpool John Moores Uni6ersity, 70 Great Crosshall Street, Li6erpool L3 3AB, UK
Abstract The model for management of opiate dependence in the United Kingdom includes long term methadone maintenance. A consequence is either long waiting lists for treatment or that treatment capacity is expanded. General practitioners are encouraged to prescribe methadone for opiate dependent patients, but little is known about the differences between patients or outcomes in primary or secondary care settings. This paper compares patients’ characteristics and treatment outcomes in a specialist drug clinic and a general practice operating a shared care policy (with the specialist clinic). We undertook a retrospective review of patient records. All patients prescribed methadone maintenance during a 2 year period in one general practice were compared with one in three patients treated at a drug clinic during the same period. Outcome was determined at the end of a treatment episode or on 30 June 1997 (whichever was sooner). Eighty-nine drug clinic and 36 general practice patients were followed up for a minimum of nine months each. Patient characteristics were similar at the start of treatment. A ‘good’ outcome (remaining in treatment or becoming drug free) was equally likely in either setting. Patients treated in the general practice setting were significantly more likely to be immunised (or have known natural immunity) against hepatitis B (adjusted odds ratio 6.0). Our findings suggest that with similar patient groups this model of care in general practice can produce results at least as good as those of a drug clinic. © 2001 Elsevier Science B.V. All rights reserved. Keywords: Methadone; Opiate dependence; Primary care; Secondary care; Outcome; General practice
Introduction Methadone maintenance has been shown in controlled trials to reduce criminal activity, mortality, and illicit drug use (Dole et al., 1969; Dole and Joseph, 1978; Newman and Whitehill, 1979; Gronbladh et al., 1990), as well as time in prison (Keen et al., 2000), and * Corresponding author.
risk taking behaviour (Farrell et al., 1994). Previously, successful treatment has been described in primary care, either in routine general practice (Hindler et al., 1996) or in special methadone clinics (Wilson et al., 1994), sometimes with substantial support from secondary care (Greenwood, 1992). Surveys of drug users have shown that they may have a low opinion of their general practitioners (survey of patients in a hospital clinic
0955-3959/01/$ - see front matter © 2001 Elsevier Science B.V. All rights reserved. PII: S 0 9 5 5 - 3 9 5 9 ( 0 0 ) 0 0 0 7 7 - 3
82
D. Lewis, M. Bellis / International Journal of Drug Policy 12 (2001) 81–89
— Telfer and Clulow, 1990) or conversely that they may value care in general practice (survey of patients receiving prescriptions in general practice — Leaver et al., 1992), but often seek out a doctor known to be sympathetic. Potentially primary care can allow easy access, holistic care for all medical problems, building of long term relationships, and avoidance of the stigma attached to attendance at drug clinics. However, the advantages of secondary care for drug users include the greater experience of doctors, who often have more time available, and the fact that many general practitioners (GPs) do not wish to prescribe methadone, and are unlikely to (Glanz, 1986; McKeganey and Boddy, 1988). One British survey indicated that most GPs felt inadequately trained to prescribe methadone, but would be encouraged to be involved in treatment if there were better support from specialist services existed (Davies and Huxley, 1997). A national review of drug services encouraged methadone maintenance in primary care (Department of Health, 1996), as have recent UK Government guidelines (Department of Health, 1999). Furthermore, the National Treatment Outcome Research Study (NTORS) recently compared six month treatment outcomes for patients receiving methadone maintenance either in a specialist clinic or general practice (Gossop et al., 1999). Results were as good in primary as secondary care, and patients were similar in age, sex, length of dependence, injecting habit, other drug use, and previous treatment attempts. Despite this, many GPs rarely (knowingly) see drug misusers, most are not confident in their ability to manage drug dependence, and only a small minority view primary care as an appropriate setting in which to treat drug dependence (Deehan et al., 1997). Like NTORS, this study compares characteristics and outcomes of patients treated in a
drug clinic and a general practice (Gossop et al., 1999). However, since length of time in treatment has been shown to be strongly associated with good long term outcome (less severe drug use, fewer drug related problems and less criminal activity) (French et al., 1993), this study makes comparisons over a longer period and in addition examines harm prevention measures such as hepatitis B immunisation.
Methods Liverpool, UK, has a population of about 460 000, and some of the worst indices for unemployment and poverty in the UK (Department of the Environment, Transport and the Regions, 2000). Vauxhall Primary Health Care (VPHC) is an inner city general practice in Liverpool, whose list size increased from 1300 to 3200 over the period of the study, with initially two, later 2.5 whole-time doctor equivalents. All doctors have an interest in drug dependence, but no specific training other than experience and some informal sessions in a clinic with a specialist in drug dependency. The practice operates a ‘shared care’ policy where all patients are initially assessed and followed up by a Community Psychiatric Nurse (CPN) from the local drug dependency clinic, with whom the doctors work closely. Following the initial assessment most care, such as review of progress and methadone prescriptions, is provided by the GPs, and stable patients often may not be seen by CPNs. In general all patients registered with the practice who request it are considered for methadone maintenance. ‘Priority’ patients, such as sex workers and pregnant women, may be but not necessarily referred to the specialist clinic but not necessarily. All 39 patients prescribed methadone maintenance at VPHC during 1995 or 1996
D. Lewis, M. Bellis / International Journal of Drug Policy 12 (2001) 81–89
were considered acceptable for inclusion in the study (all patients had entered treatment by October 1996). For comparison, a second set of drug users in treatment were selected from the local specialist drug clinic. This clinic is where the CPNs are based and is run by a consultant psychiatrist. For the duration of the study there was no other local specialist provider of methadone maintenance, and private general practice is extremely rare in Liverpool. Patients may refer themselves to the clinic, so include those registered with GPs who do not wish to treat opiate addiction. There is a waiting list for care, usually several months, except for ‘priority’ groups — sex workers, pregnant women or people with acute hepatitis. The clinic keeps a chronological list of all new patients seen. Every third patient from their list for the same period (January 1995 to October 1996) was included, with the exception of those who were not prescribed methadone, resulting in 96 patients. Data collection included retrospective examination of case notes from the time of starting treatment, with follow up ending on 30th June 1997, or when the patient was last seen prior to that date. All patients were followed up for at least nine months, and patients in primary and secondary care did not differ in their mean length of follow up. All the notes were included and data entered on to an extraction sheet. Data were collected on drug and alcohol use before and during treatment (supported by urine testing where possible); experience of prison; previous treatment history; hepatitis B and C testing and immunisation for hepatitis B. All patients’ GPs were contacted by letter or phone to assess from records attendance at Accident and Emergency and hospital admissions during the treatment period. This assumed that the GPs would have been informed of the patients’ attendance at hospi-
83
tal, which would be routine practice in local hospitals. Not all GPs still had the relevant records (usually because the patient had changed doctor), and one GP refused to give the information (see altered sample sizes in Table 2). At the end of follow up patients were classified as still in treatment, drug free, using illicit opiates (i.e. not in treatment), in prison, moved away, transferred to another service, or status unknown. Remaining in treatment or becoming drug free were classified as ‘good’ outcomes; moving away or being transferred elsewhere were classified ‘OK’; and using illicit opiates (not in treatment but known to be using according to information recorded in the notes), being in prison or outcome unknown (i.e. ceasing attendance for treatment with no reason known) were classified as ‘bad’ outcomes. It is possible that some patients were imprisoned for offences committed before they started treatment, but we could not verify this. We assumed that those with an unknown outcome were either imprisoned or using illicit street drugs, but it is feasible that some were drug free.
Statistical analysis Data were entered into Excel 5.0 and analysed with STATA 5.0. (Stata Corporation, Texas 77840, USA). Data were compared using chi square or Fisher’s exact test for categorical variables or Student’s t-test for continuous variables. Adjusted odds ratios were calculated using logistic regression analysis. Kaplan –Meier survival analysis was performed for remaining in treatment for at least 9 months. For simplicity, this analysis assumed that the one patient who became drug free after four months was still ‘in treatment’ at 9 months.
84
D. Lewis, M. Bellis / International Journal of Drug Policy 12 (2001) 81–89
Results Results are presented for the 89 drug clinic patients and 36 general practice patients. Table 1 shows pre-treatment characteristics for the two groups. All patients used heroin at the start of treatment. The two groups did not differ significantly with respect to age; sex; opiate-using and injecting history; previous prison sentences; and previous courses of methadone treatment. Although not statistically significant, there was a tendency towards women at the drug clinic being more likely to be pregnant at the start of treatment (7 of 32 compared with 0 of 9), and more likely to be commercial sex workers (13 of 41 compared with 1 of 9); both pregnancy and
commercial sex work are criteria for priority treatment at the drug clinic. In addition, four clinic patients had acute hepatitis at the start of treatment-another criterion for priority treatment. However, across all assessed criteria there were no statistically significant differences between the groups at the start of treatment. Table 2 shows outcomes for the two groups. The likelihood of a good (in treatment or drug free), ‘OK’ (moved away or transferred to another treatment facility) or ‘bad’ (in prison, using heroin on the street, or unknown) outcome did not differ significantly between groups. Fig. 1 shows KaplanMeier analysis for survival in treatment up to nine months (the minimum time for which all
Table 1 Characteristics of patients treated in the drug clinic and general practice at start of treatment Characteristic
Drug Clinic, % (n)a or mean (n)
Male Mean age, years Mean age first use opiates, years Mean length of opiate use, months Injected in last month Shared needles ever Acute hepatitis at start Pregnant at start (% of women) Commercial sex worker (% of women)
64 (89) 28.4 (89) 20.1 (87) 101 (87) 52 (84) 44 (62) 4 (89) 22 (32) 41 (32)
Drug use at start of treatment (within pre6ious month) Heroin 100 (89) Amphetamine 5 (86) Cocaine 37 (84) Benzodiazepines 22 (86) Excessive alcohol 8 (88) Opiates other than heroin 66 (64) Prison ever 55 (71) Prison within 1 year 23 (60) Previous methadone treatment 71 (85)
General practice, % (n) or mean (n) 75 (36) 27.6 (36) 20.2 (31) 90 (31) 42 (31) 37 (27) 0 (36) 0 (9) 11 (9)
100 3 32 21 6 84 57 17 58
(36) (31) (31) (33) (34) (19) (30) (29) (33)
P for difference 0.24 0.42 0.90 0.31 0.32 0.57 0.32 0.31 0.13
– 0.74 0.64 0.92 0.70 0.12 0.87 0.51 0.18
a n is sample size (maximum 89 in drug clinic and 36 in general practice). Differences reflect missing information from case notes, except for pregnancy and sex workers where sample size is the number of women (no missing data).
D. Lewis, M. Bellis / International Journal of Drug Policy 12 (2001) 81–89
85
Table 2 Outcomes for patients in drug clinic or general practicea Outcome
Drug clinic % (n) or mean (n)
General practice % (n) or mean (n)
P for difference
Methadone dose at 1 month, mls daily, mean Methadone dose at end (mls daily), mean Other illicit drug use at end,% Months in treament, mean Follow up (months from start to 30/6/99), mean
46 46 75 11 19
41 39 56 15 21
(SE 1.5) (SE 2.2) (32) (36) (36)
0.11 B0.05 0.054 B0.5 0.2
(SE 1.6) (SE 1.6) (83) (87) (89)
Situation on 30 /6 /97 (or date last seen if left treatment) Still in treatment, % 50 (89) Drug free, % 2 (89) In prison, % 9 (89) Unknown, % 20 (89) Known to be using illicit opiates (not in 11 (89) treatment)% Transferred to drug clinic,% 6 (89) Transferred to primary care,% 1 (89) Moved away,% 1 (89) ‘Good’ outcome,% 52 (89) ‘OK’ outcome,% 8 (89) ‘Bad’ outcome,% 40 (89)
53 3 25 3 6
(36) (36) (36) (36) (36)
0.74 1 B0.05 0.01 0.5
6 0 6 56 11 33
(36) (36) (36) (36) (36) (36)
1 1 0.2 0.7 0.56 0.46
Hepatitis 6ariables Tested for Hepatitis B, %
39 (89)
67 (36)
Hepatitis B positive (of tested), % Hepatitis B positive (of all), % Full hepatitis B immunisation or natural immunity, % Tested for hepatitis C, %
49 (35) 19 (89) 24 (89)
8 (24) 6 (36) 44 (36)
29 (89)
63 (36)
Hepatitis C positive (of tested), % Hepatitis C positive (of all), % Hospital admission (whilst on methadone), % Attendance at A&E (whilst on methadone), %
79 21 16 31
52 31 11 33
a
(24) (89) (49) (48)
(21) (36) (36) (36)
B 0.01, OR= 3.1 B0.01 0.056 B0.05, OR = 2.6 B0.001, OR= 4.2 0.057 0.275 0.5 0.84
OR, odds ratio if in General Practice; n, sample size (maximum 89 in drug clinic and 36 in general practice).
patients were followed up). Again, results are very similar for each group. Differences between groups however included patients in general practice being much more likely to have been tested for hepatitis B (odds ratio 3.1), while clinic patients were more likely (odds ratio 2.6) to be positive for hepatitis B, both as a proportion of the whole sample and
of those tested. Patients in general practice were significantly more likely (odds ratio 2.6) to have been fully immunised against hepatitis B (three injections with seroconversion confirmed) or to have known natural immunity (positive hepatitis B surface antibody). In order to adjust for potential confounding factors, adjusted odds ratios (AORs) were
86
D. Lewis, M. Bellis / International Journal of Drug Policy 12 (2001) 81–89
calculated correcting for age, sex, injecting habit, length of opiate use, and whether or not patients had previously been on methadone maintenance. The difference between samples in immunisation levels remained significant (PB 0.01) with an AOR of 6.0. The same technique confirmed that general practice patients were also more likely to have been tested for hepatitis C (PB 0.001, AOR 7.2). There was no significant difference in the number of patients attending Accident and Emergency or being admitted to hospital, and the mean number of attendances or admissions was the same. However, data for hospital attendances were only available for just over half the clinic patients. Finally, for those patients followed up for longer than nine months the chance of survival in treatment was calculated at 15, 20 and 24 months (excluding those not followed up for the respective time). Table 3 shows odds ratios for remaining in treatment, and the adjusted odds ratios corrected for age, sex, length of
opiate use, injecting habit, and previous courses of methadone maintenance. However, here the number of patients is quite small, and therefore the potential for confounding bias larger
Conclusions Long term methadone maintenance results in large numbers of patients being treated and, unless specialist drugs services expand substantially, primary care practitioners will have to be involved. This study compares baseline data and outcome at nine months for opiate dependent patients given methadone maintenance at a drug dependency clinic and a general practice. Most patients were not regular injectors and this reflects the pattern of drug use locally (The Drug Misuse Research Unit, 1998). As in the NTORS study, the patients were similar at baseline, and outcomes for general practice patients were at least as good as for those in the
Fig. 1. Kaplan–Meier survival estimates for the proportion of patients remaining in contact with the drug clinic or general practice.
D. Lewis, M. Bellis / International Journal of Drug Policy 12 (2001) 81–89
87
Table 3 Patients still in treatment after longer follow up (n is number of patients who entered treatment at least 15/20/24 months before 30 June 1997) Follow up time
Drug clinic % patients still in treatment (n)
General practice% patients still in treatment (n)
15 months 20 months 24 months
43 (67) 37 (46) 27 (26)
52 (23) 69 (16) 71 (14)
specialist clinic. Patients in general practice were more likely to be tested for and immunised against hepatitis B, an important aspect of harm minimisation (Department of Health, 1999). One possible explanation for this is the proximity of a practice nurse, and therefore an increased potential for opportunistic venepuncture and immunisation in primary care and commensurate increase in uptake of vaccination. Testing for hepatitis C infection is recommended (Department of Health, 1999), and patients treated in general practice were more likely to have this done. However, in our experience those who are positive rarely attend Infectious Diseases clinic for hepatitis C if referred, and very few would be motivated enough to receive treatment with interferon, should this be indicated. The effects of the test on behaviour remain unstudied, although a recent study suggests an association between having a test and reduction in sharing behaviour (Cook et al., 2000). Although in the short term duration in treatment was similar in both groups, when they were followed up for longer periods it appears that general practice patients were more likely to remain in treatment after 20 or 24 months. A longer prospective study would be needed to confirm or refute this. It may be that general practice provides a stable, more flexible and convenient setting, with the integration of other aspects of health care. One general practice treating patients without any secondary care support has described an aver-
Odds ratio (P for difference x2) 1.4 (0.46) 3.8 (0.03) 6.8 (B0.01)
Adjusted odds ratio (P) 2.2 (0.17) 5.6 (0.03) 19.6 (0.02)
age treatment time of over 5 years (Martin et al., 1998). Rates of retention in treatment in the present study are similar to those published elsewhere, for example about 60% at 9 months in practices throughout Glasgow (Gruer et al., 1997), and 66 and 60% for patients treated in general practice or specialist clinic respectively in the NTORS study (Gossop et al., 1999). As in these other studies, both groups tended to have doses of methadone lower than those now recommended (Department of Health, 1999), and were often not receiving treatment under direct supervision. This study examined one method of ‘shared care’ in general practice where GPs work closely with CPNs from the drug clinic and have easy access to advice from CPNs or a consultant psychiatrist at the drug clinic; all the doctors involved are interested in drug dependence and the practice was supported by the Health Authority specifically to treat this group of patients. Attitudes towards treatment of drug users remain split. In one area of London, most GPs did not consider primary care an appropriate setting in which to treat opiate dependence and would not wish to do so even if given additional training and financial incentives (Deehan et al., 1997). Despite this, a scheme in Glasgow, Scotland found that many GPs are willing to prescribe methadone to some patients if given appropriate support (Gruer et al., 1997), including payment.
88
D. Lewis, M. Bellis / International Journal of Drug Policy 12 (2001) 81–89
Given that many GPs will never want to treat drug dependence, that many opiate misusers wish treatment but do not receive it, and that this model depends on an effective local specialist service, the role of the drug clinic is unlikely to diminish in the near future. General practice may not be an appropriate setting for treatment of some chaotic drug users. No economic analysis has compared the costs of treatment in primary with secondary care. This was a retrospective study, so not as robust as a long (more costly) prospective one; but consideration of sometimes lengthy waiting lists for drug clinics, combined with a slow turnover of clients, means that alternative settings and approaches for the provision of drugs services need to be urgently examined. This study represents one such examination and identifies that this model of care can produce results at least as good as those of a drug clinic in patients with similar drug using backgrounds.
Acknowledgements Thanks to all staff at Vauxhall Primary Health Care and Liverpool Drug Dependency Clinic who helped find patient records; to Lisa Cummins and staff at RADICAL (Research Audit and Development In Communities Across Liverpool) who entered data on to the database; and to Dr Chris Whitty for statistical advice. D.L. was employed by Liverpool Health Authority through its Primary Care Initiative.
References Cook P., McVeigh J., Patel A., Syed Q., Mutton K., Bellis M. Hepatitis C in injecting drug users in the North West. Liverpool John Moores University, 2000.
Davies A, Huxley P. Survey of general practitioners’ opinions on treatment of opiate users. British Medical Journal 1997;314:1173– 4. Deehan A, Taylor C, Strang J. The general practitioner, the drug misuser, and the alcohol misuser: major differences in general practitioner activity, therapeutic committment, and ‘shared care’ proposals. British Journal of General Practice 1997;47:705– 9. Department of the Environment, Transport and the Regions. Indices of Deprivation 2000. London: Department of the Environment, Transport and the Regions, 2000. Department of Health. Task force to review services for drug misusers: report of an independent review of drug treatment sevices in England. London: Department of Health, 1996. (Chairman: The Reverend Dr John Polkinghorne) Department of Health. Drug misuse and dependence – Guidelines on clinical management. London: Department of Health, 1999. (Chairman Prof J Strang). Dole VP, Joseph H. Long term outcome of patients treated with methadone maintenance. Annals of the New York Academy of Science 1978;311:181– 9. Dole VP, Robinson WR, Orraca J, Towns E, Searcy P, Caine E. Methadone treatment of randomly selected criminal addicts. New England Journal of Medicine 1969;280:1372– 5. The Drug Misuse Research Unit (University of Manchester), Drug Monitoring Unit (Liverpool John Moores University). Drug Misuse in the North West of England 1997, University of Manchester and Liverpool John Moores University, 1998. Farrell M, Ward J, Mattick R, Hall W, Stimson G, des Jarlais D, Gossop M, Strang J. Methadone maintenance treatment in opiate dependence: a review. British Medical Journal 1994;309:997– 1001. French MT, Zarkin GA, Hubbard RL, Rachal JV. The effects of time in drug abuse treatment and and employment on posttreatment drug use and criminal activity. American Journal of Drug and Alcohol Abuse 1993;19(1):19– 33. Glanz A. Findings of a national survey of the role of general practitioners in the treatment of opiate misuse: views on treatment. British Medical Journal 1986;293:543– 5. Gossop M, Marsden J, Stewart D, Lehmann P, Strang J. Methadone treatment practices and outcome among opiate addicts treated in drug clinics and in general practice: results from the National Treat-
D. Lewis, M. Bellis / International Journal of Drug Policy 12 (2001) 81–89
ment Outcome Research Study. British Journal of General Practice 1999;49:31–4. Greenwood J. Persuading general practitioners to prescribe-good husbandry or a recipe for chaos? British Journal of Addiction 1992;87:567–75. Gronbladh L, Ohlund LS, Gunne LM. Mortality in heroin addiction: impact of methadone treatment. Acta Psychiatry Scandanavia 1990;82:223–7. Gruer L, Wilson P, Scott R, Elliott L, Macleod J, Harden K, Forrester E, Hinshelwood S, McNulty H, Silk P. General practitioner centred scheme for treatment of opiate dependent drug injectors in Glasgow. British Medical Journal 1997;314:1730– 5. Hindler C, King M, Nazareth I, Cohen J, Farmer R, Gerada C. Characteristics of drug misusers and their perceptions of general practitioner care. British Journal of General Practice 1996;46:149– 52. Keen J, Rowse G, Mathers N, Campbell M, Seivewright N. Can methadone maintenance for heroin-dependent paients retained in general practice reduce criminal conviction rates and time
.
89
spent in prison? British Journal of General Practice 2000;50:48– 9. Leaver EJ, Elford J, Morris JK, Cohen J. Use of general practice by intrvenous heroin users on a methadone programme. British Journal of General Practice 1992;42:465– 8. Martin E, Canavan A, Butler R. A decade of caring for drug users entirely within general practice. British Journal of General Practice 1998;48:1679– 82. McKeganey NP, Boddy FA. General Practitioners and opiate-abusing patients. J R Coll Gen Pract 1988;38:73– 5. Newman R, Whitehill WB. Double-blind comparison of methadone and placebo maintenance treatments of narcotic addicts in Hong Kong. Lancet 1979;ii:485– 8. Telfer I, Clulow C. Heroin misusers: what they think of their general practitioners. British Journal of Addiction 1990;85:137– 40. Wilson P, Watson R, Ralston G. Methadone maintenance in general practice: patients, workload and outcomes. British Medical Journal 1994;309:641– 4.
General practice or drug clinic for methadone maintenance? A controlled comparison of treatment outcomes David Lewis a,*, Mark Bellis b b
a Vauxhall Primary Health Care, Limekiln Lane,Li6erpool L5 8XR, UK Public Health Sector, School of Health, Li6erpool John Moores Uni6ersity, 70 Great Crosshall Street, Li6erpool L3 3AB, UK
Abstract The model for management of opiate dependence in the United Kingdom includes long term methadone maintenance. A consequence is either long waiting lists for treatment or that treatment capacity is expanded. General practitioners are encouraged to prescribe methadone for opiate dependent patients, but little is known about the differences between patients or outcomes in primary or secondary care settings. This paper compares patients’ characteristics and treatment outcomes in a specialist drug clinic and a general practice operating a shared care policy (with the specialist clinic). We undertook a retrospective review of patient records. All patients prescribed methadone maintenance during a 2 year period in one general practice were compared with one in three patients treated at a drug clinic during the same period. Outcome was determined at the end of a treatment episode or on 30 June 1997 (whichever was sooner). Eighty-nine drug clinic and 36 general practice patients were followed up for a minimum of nine months each. Patient characteristics were similar at the start of treatment. A ‘good’ outcome (remaining in treatment or becoming drug free) was equally likely in either setting. Patients treated in the general practice setting were significantly more likely to be immunised (or have known natural immunity) against hepatitis B (adjusted odds ratio 6.0). Our findings suggest that with similar patient groups this model of care in general practice can produce results at least as good as those of a drug clinic. © 2001 Elsevier Science B.V. All rights reserved. Keywords: Methadone; Opiate dependence; Primary care; Secondary care; Outcome; General practice
Introduction Methadone maintenance has been shown in controlled trials to reduce criminal activity, mortality, and illicit drug use (Dole et al., 1969; Dole and Joseph, 1978; Newman and Whitehill, 1979; Gronbladh et al., 1990), as well as time in prison (Keen et al., 2000), and * Corresponding author.
risk taking behaviour (Farrell et al., 1994). Previously, successful treatment has been described in primary care, either in routine general practice (Hindler et al., 1996) or in special methadone clinics (Wilson et al., 1994), sometimes with substantial support from secondary care (Greenwood, 1992). Surveys of drug users have shown that they may have a low opinion of their general practitioners (survey of patients in a hospital clinic
0955-3959/01/$ - see front matter © 2001 Elsevier Science B.V. All rights reserved. PII: S 0 9 5 5 - 3 9 5 9 ( 0 0 ) 0 0 0 7 7 - 3
82
D. Lewis, M. Bellis / International Journal of Drug Policy 12 (2001) 81–89
— Telfer and Clulow, 1990) or conversely that they may value care in general practice (survey of patients receiving prescriptions in general practice — Leaver et al., 1992), but often seek out a doctor known to be sympathetic. Potentially primary care can allow easy access, holistic care for all medical problems, building of long term relationships, and avoidance of the stigma attached to attendance at drug clinics. However, the advantages of secondary care for drug users include the greater experience of doctors, who often have more time available, and the fact that many general practitioners (GPs) do not wish to prescribe methadone, and are unlikely to (Glanz, 1986; McKeganey and Boddy, 1988). One British survey indicated that most GPs felt inadequately trained to prescribe methadone, but would be encouraged to be involved in treatment if there were better support from specialist services existed (Davies and Huxley, 1997). A national review of drug services encouraged methadone maintenance in primary care (Department of Health, 1996), as have recent UK Government guidelines (Department of Health, 1999). Furthermore, the National Treatment Outcome Research Study (NTORS) recently compared six month treatment outcomes for patients receiving methadone maintenance either in a specialist clinic or general practice (Gossop et al., 1999). Results were as good in primary as secondary care, and patients were similar in age, sex, length of dependence, injecting habit, other drug use, and previous treatment attempts. Despite this, many GPs rarely (knowingly) see drug misusers, most are not confident in their ability to manage drug dependence, and only a small minority view primary care as an appropriate setting in which to treat drug dependence (Deehan et al., 1997). Like NTORS, this study compares characteristics and outcomes of patients treated in a
drug clinic and a general practice (Gossop et al., 1999). However, since length of time in treatment has been shown to be strongly associated with good long term outcome (less severe drug use, fewer drug related problems and less criminal activity) (French et al., 1993), this study makes comparisons over a longer period and in addition examines harm prevention measures such as hepatitis B immunisation.
Methods Liverpool, UK, has a population of about 460 000, and some of the worst indices for unemployment and poverty in the UK (Department of the Environment, Transport and the Regions, 2000). Vauxhall Primary Health Care (VPHC) is an inner city general practice in Liverpool, whose list size increased from 1300 to 3200 over the period of the study, with initially two, later 2.5 whole-time doctor equivalents. All doctors have an interest in drug dependence, but no specific training other than experience and some informal sessions in a clinic with a specialist in drug dependency. The practice operates a ‘shared care’ policy where all patients are initially assessed and followed up by a Community Psychiatric Nurse (CPN) from the local drug dependency clinic, with whom the doctors work closely. Following the initial assessment most care, such as review of progress and methadone prescriptions, is provided by the GPs, and stable patients often may not be seen by CPNs. In general all patients registered with the practice who request it are considered for methadone maintenance. ‘Priority’ patients, such as sex workers and pregnant women, may be but not necessarily referred to the specialist clinic but not necessarily. All 39 patients prescribed methadone maintenance at VPHC during 1995 or 1996
D. Lewis, M. Bellis / International Journal of Drug Policy 12 (2001) 81–89
were considered acceptable for inclusion in the study (all patients had entered treatment by October 1996). For comparison, a second set of drug users in treatment were selected from the local specialist drug clinic. This clinic is where the CPNs are based and is run by a consultant psychiatrist. For the duration of the study there was no other local specialist provider of methadone maintenance, and private general practice is extremely rare in Liverpool. Patients may refer themselves to the clinic, so include those registered with GPs who do not wish to treat opiate addiction. There is a waiting list for care, usually several months, except for ‘priority’ groups — sex workers, pregnant women or people with acute hepatitis. The clinic keeps a chronological list of all new patients seen. Every third patient from their list for the same period (January 1995 to October 1996) was included, with the exception of those who were not prescribed methadone, resulting in 96 patients. Data collection included retrospective examination of case notes from the time of starting treatment, with follow up ending on 30th June 1997, or when the patient was last seen prior to that date. All patients were followed up for at least nine months, and patients in primary and secondary care did not differ in their mean length of follow up. All the notes were included and data entered on to an extraction sheet. Data were collected on drug and alcohol use before and during treatment (supported by urine testing where possible); experience of prison; previous treatment history; hepatitis B and C testing and immunisation for hepatitis B. All patients’ GPs were contacted by letter or phone to assess from records attendance at Accident and Emergency and hospital admissions during the treatment period. This assumed that the GPs would have been informed of the patients’ attendance at hospi-
83
tal, which would be routine practice in local hospitals. Not all GPs still had the relevant records (usually because the patient had changed doctor), and one GP refused to give the information (see altered sample sizes in Table 2). At the end of follow up patients were classified as still in treatment, drug free, using illicit opiates (i.e. not in treatment), in prison, moved away, transferred to another service, or status unknown. Remaining in treatment or becoming drug free were classified as ‘good’ outcomes; moving away or being transferred elsewhere were classified ‘OK’; and using illicit opiates (not in treatment but known to be using according to information recorded in the notes), being in prison or outcome unknown (i.e. ceasing attendance for treatment with no reason known) were classified as ‘bad’ outcomes. It is possible that some patients were imprisoned for offences committed before they started treatment, but we could not verify this. We assumed that those with an unknown outcome were either imprisoned or using illicit street drugs, but it is feasible that some were drug free.
Statistical analysis Data were entered into Excel 5.0 and analysed with STATA 5.0. (Stata Corporation, Texas 77840, USA). Data were compared using chi square or Fisher’s exact test for categorical variables or Student’s t-test for continuous variables. Adjusted odds ratios were calculated using logistic regression analysis. Kaplan –Meier survival analysis was performed for remaining in treatment for at least 9 months. For simplicity, this analysis assumed that the one patient who became drug free after four months was still ‘in treatment’ at 9 months.
84
D. Lewis, M. Bellis / International Journal of Drug Policy 12 (2001) 81–89
Results Results are presented for the 89 drug clinic patients and 36 general practice patients. Table 1 shows pre-treatment characteristics for the two groups. All patients used heroin at the start of treatment. The two groups did not differ significantly with respect to age; sex; opiate-using and injecting history; previous prison sentences; and previous courses of methadone treatment. Although not statistically significant, there was a tendency towards women at the drug clinic being more likely to be pregnant at the start of treatment (7 of 32 compared with 0 of 9), and more likely to be commercial sex workers (13 of 41 compared with 1 of 9); both pregnancy and
commercial sex work are criteria for priority treatment at the drug clinic. In addition, four clinic patients had acute hepatitis at the start of treatment-another criterion for priority treatment. However, across all assessed criteria there were no statistically significant differences between the groups at the start of treatment. Table 2 shows outcomes for the two groups. The likelihood of a good (in treatment or drug free), ‘OK’ (moved away or transferred to another treatment facility) or ‘bad’ (in prison, using heroin on the street, or unknown) outcome did not differ significantly between groups. Fig. 1 shows KaplanMeier analysis for survival in treatment up to nine months (the minimum time for which all
Table 1 Characteristics of patients treated in the drug clinic and general practice at start of treatment Characteristic
Drug Clinic, % (n)a or mean (n)
Male Mean age, years Mean age first use opiates, years Mean length of opiate use, months Injected in last month Shared needles ever Acute hepatitis at start Pregnant at start (% of women) Commercial sex worker (% of women)
64 (89) 28.4 (89) 20.1 (87) 101 (87) 52 (84) 44 (62) 4 (89) 22 (32) 41 (32)
Drug use at start of treatment (within pre6ious month) Heroin 100 (89) Amphetamine 5 (86) Cocaine 37 (84) Benzodiazepines 22 (86) Excessive alcohol 8 (88) Opiates other than heroin 66 (64) Prison ever 55 (71) Prison within 1 year 23 (60) Previous methadone treatment 71 (85)
General practice, % (n) or mean (n) 75 (36) 27.6 (36) 20.2 (31) 90 (31) 42 (31) 37 (27) 0 (36) 0 (9) 11 (9)
100 3 32 21 6 84 57 17 58
(36) (31) (31) (33) (34) (19) (30) (29) (33)
P for difference 0.24 0.42 0.90 0.31 0.32 0.57 0.32 0.31 0.13
– 0.74 0.64 0.92 0.70 0.12 0.87 0.51 0.18
a n is sample size (maximum 89 in drug clinic and 36 in general practice). Differences reflect missing information from case notes, except for pregnancy and sex workers where sample size is the number of women (no missing data).
D. Lewis, M. Bellis / International Journal of Drug Policy 12 (2001) 81–89
85
Table 2 Outcomes for patients in drug clinic or general practicea Outcome
Drug clinic % (n) or mean (n)
General practice % (n) or mean (n)
P for difference
Methadone dose at 1 month, mls daily, mean Methadone dose at end (mls daily), mean Other illicit drug use at end,% Months in treament, mean Follow up (months from start to 30/6/99), mean
46 46 75 11 19
41 39 56 15 21
(SE 1.5) (SE 2.2) (32) (36) (36)
0.11 B0.05 0.054 B0.5 0.2
(SE 1.6) (SE 1.6) (83) (87) (89)
Situation on 30 /6 /97 (or date last seen if left treatment) Still in treatment, % 50 (89) Drug free, % 2 (89) In prison, % 9 (89) Unknown, % 20 (89) Known to be using illicit opiates (not in 11 (89) treatment)% Transferred to drug clinic,% 6 (89) Transferred to primary care,% 1 (89) Moved away,% 1 (89) ‘Good’ outcome,% 52 (89) ‘OK’ outcome,% 8 (89) ‘Bad’ outcome,% 40 (89)
53 3 25 3 6
(36) (36) (36) (36) (36)
0.74 1 B0.05 0.01 0.5
6 0 6 56 11 33
(36) (36) (36) (36) (36) (36)
1 1 0.2 0.7 0.56 0.46
Hepatitis 6ariables Tested for Hepatitis B, %
39 (89)
67 (36)
Hepatitis B positive (of tested), % Hepatitis B positive (of all), % Full hepatitis B immunisation or natural immunity, % Tested for hepatitis C, %
49 (35) 19 (89) 24 (89)
8 (24) 6 (36) 44 (36)
29 (89)
63 (36)
Hepatitis C positive (of tested), % Hepatitis C positive (of all), % Hospital admission (whilst on methadone), % Attendance at A&E (whilst on methadone), %
79 21 16 31
52 31 11 33
a
(24) (89) (49) (48)
(21) (36) (36) (36)
B 0.01, OR= 3.1 B0.01 0.056 B0.05, OR = 2.6 B0.001, OR= 4.2 0.057 0.275 0.5 0.84
OR, odds ratio if in General Practice; n, sample size (maximum 89 in drug clinic and 36 in general practice).
patients were followed up). Again, results are very similar for each group. Differences between groups however included patients in general practice being much more likely to have been tested for hepatitis B (odds ratio 3.1), while clinic patients were more likely (odds ratio 2.6) to be positive for hepatitis B, both as a proportion of the whole sample and
of those tested. Patients in general practice were significantly more likely (odds ratio 2.6) to have been fully immunised against hepatitis B (three injections with seroconversion confirmed) or to have known natural immunity (positive hepatitis B surface antibody). In order to adjust for potential confounding factors, adjusted odds ratios (AORs) were
86
D. Lewis, M. Bellis / International Journal of Drug Policy 12 (2001) 81–89
calculated correcting for age, sex, injecting habit, length of opiate use, and whether or not patients had previously been on methadone maintenance. The difference between samples in immunisation levels remained significant (PB 0.01) with an AOR of 6.0. The same technique confirmed that general practice patients were also more likely to have been tested for hepatitis C (PB 0.001, AOR 7.2). There was no significant difference in the number of patients attending Accident and Emergency or being admitted to hospital, and the mean number of attendances or admissions was the same. However, data for hospital attendances were only available for just over half the clinic patients. Finally, for those patients followed up for longer than nine months the chance of survival in treatment was calculated at 15, 20 and 24 months (excluding those not followed up for the respective time). Table 3 shows odds ratios for remaining in treatment, and the adjusted odds ratios corrected for age, sex, length of
opiate use, injecting habit, and previous courses of methadone maintenance. However, here the number of patients is quite small, and therefore the potential for confounding bias larger
Conclusions Long term methadone maintenance results in large numbers of patients being treated and, unless specialist drugs services expand substantially, primary care practitioners will have to be involved. This study compares baseline data and outcome at nine months for opiate dependent patients given methadone maintenance at a drug dependency clinic and a general practice. Most patients were not regular injectors and this reflects the pattern of drug use locally (The Drug Misuse Research Unit, 1998). As in the NTORS study, the patients were similar at baseline, and outcomes for general practice patients were at least as good as for those in the
Fig. 1. Kaplan–Meier survival estimates for the proportion of patients remaining in contact with the drug clinic or general practice.
D. Lewis, M. Bellis / International Journal of Drug Policy 12 (2001) 81–89
87
Table 3 Patients still in treatment after longer follow up (n is number of patients who entered treatment at least 15/20/24 months before 30 June 1997) Follow up time
Drug clinic % patients still in treatment (n)
General practice% patients still in treatment (n)
15 months 20 months 24 months
43 (67) 37 (46) 27 (26)
52 (23) 69 (16) 71 (14)
specialist clinic. Patients in general practice were more likely to be tested for and immunised against hepatitis B, an important aspect of harm minimisation (Department of Health, 1999). One possible explanation for this is the proximity of a practice nurse, and therefore an increased potential for opportunistic venepuncture and immunisation in primary care and commensurate increase in uptake of vaccination. Testing for hepatitis C infection is recommended (Department of Health, 1999), and patients treated in general practice were more likely to have this done. However, in our experience those who are positive rarely attend Infectious Diseases clinic for hepatitis C if referred, and very few would be motivated enough to receive treatment with interferon, should this be indicated. The effects of the test on behaviour remain unstudied, although a recent study suggests an association between having a test and reduction in sharing behaviour (Cook et al., 2000). Although in the short term duration in treatment was similar in both groups, when they were followed up for longer periods it appears that general practice patients were more likely to remain in treatment after 20 or 24 months. A longer prospective study would be needed to confirm or refute this. It may be that general practice provides a stable, more flexible and convenient setting, with the integration of other aspects of health care. One general practice treating patients without any secondary care support has described an aver-
Odds ratio (P for difference x2) 1.4 (0.46) 3.8 (0.03) 6.8 (B0.01)
Adjusted odds ratio (P) 2.2 (0.17) 5.6 (0.03) 19.6 (0.02)
age treatment time of over 5 years (Martin et al., 1998). Rates of retention in treatment in the present study are similar to those published elsewhere, for example about 60% at 9 months in practices throughout Glasgow (Gruer et al., 1997), and 66 and 60% for patients treated in general practice or specialist clinic respectively in the NTORS study (Gossop et al., 1999). As in these other studies, both groups tended to have doses of methadone lower than those now recommended (Department of Health, 1999), and were often not receiving treatment under direct supervision. This study examined one method of ‘shared care’ in general practice where GPs work closely with CPNs from the drug clinic and have easy access to advice from CPNs or a consultant psychiatrist at the drug clinic; all the doctors involved are interested in drug dependence and the practice was supported by the Health Authority specifically to treat this group of patients. Attitudes towards treatment of drug users remain split. In one area of London, most GPs did not consider primary care an appropriate setting in which to treat opiate dependence and would not wish to do so even if given additional training and financial incentives (Deehan et al., 1997). Despite this, a scheme in Glasgow, Scotland found that many GPs are willing to prescribe methadone to some patients if given appropriate support (Gruer et al., 1997), including payment.
88
D. Lewis, M. Bellis / International Journal of Drug Policy 12 (2001) 81–89
Given that many GPs will never want to treat drug dependence, that many opiate misusers wish treatment but do not receive it, and that this model depends on an effective local specialist service, the role of the drug clinic is unlikely to diminish in the near future. General practice may not be an appropriate setting for treatment of some chaotic drug users. No economic analysis has compared the costs of treatment in primary with secondary care. This was a retrospective study, so not as robust as a long (more costly) prospective one; but consideration of sometimes lengthy waiting lists for drug clinics, combined with a slow turnover of clients, means that alternative settings and approaches for the provision of drugs services need to be urgently examined. This study represents one such examination and identifies that this model of care can produce results at least as good as those of a drug clinic in patients with similar drug using backgrounds.
Acknowledgements Thanks to all staff at Vauxhall Primary Health Care and Liverpool Drug Dependency Clinic who helped find patient records; to Lisa Cummins and staff at RADICAL (Research Audit and Development In Communities Across Liverpool) who entered data on to the database; and to Dr Chris Whitty for statistical advice. D.L. was employed by Liverpool Health Authority through its Primary Care Initiative.
References Cook P., McVeigh J., Patel A., Syed Q., Mutton K., Bellis M. Hepatitis C in injecting drug users in the North West. Liverpool John Moores University, 2000.
Davies A, Huxley P. Survey of general practitioners’ opinions on treatment of opiate users. British Medical Journal 1997;314:1173– 4. Deehan A, Taylor C, Strang J. The general practitioner, the drug misuser, and the alcohol misuser: major differences in general practitioner activity, therapeutic committment, and ‘shared care’ proposals. British Journal of General Practice 1997;47:705– 9. Department of the Environment, Transport and the Regions. Indices of Deprivation 2000. London: Department of the Environment, Transport and the Regions, 2000. Department of Health. Task force to review services for drug misusers: report of an independent review of drug treatment sevices in England. London: Department of Health, 1996. (Chairman: The Reverend Dr John Polkinghorne) Department of Health. Drug misuse and dependence – Guidelines on clinical management. London: Department of Health, 1999. (Chairman Prof J Strang). Dole VP, Joseph H. Long term outcome of patients treated with methadone maintenance. Annals of the New York Academy of Science 1978;311:181– 9. Dole VP, Robinson WR, Orraca J, Towns E, Searcy P, Caine E. Methadone treatment of randomly selected criminal addicts. New England Journal of Medicine 1969;280:1372– 5. The Drug Misuse Research Unit (University of Manchester), Drug Monitoring Unit (Liverpool John Moores University). Drug Misuse in the North West of England 1997, University of Manchester and Liverpool John Moores University, 1998. Farrell M, Ward J, Mattick R, Hall W, Stimson G, des Jarlais D, Gossop M, Strang J. Methadone maintenance treatment in opiate dependence: a review. British Medical Journal 1994;309:997– 1001. French MT, Zarkin GA, Hubbard RL, Rachal JV. The effects of time in drug abuse treatment and and employment on posttreatment drug use and criminal activity. American Journal of Drug and Alcohol Abuse 1993;19(1):19– 33. Glanz A. Findings of a national survey of the role of general practitioners in the treatment of opiate misuse: views on treatment. British Medical Journal 1986;293:543– 5. Gossop M, Marsden J, Stewart D, Lehmann P, Strang J. Methadone treatment practices and outcome among opiate addicts treated in drug clinics and in general practice: results from the National Treat-
D. Lewis, M. Bellis / International Journal of Drug Policy 12 (2001) 81–89
ment Outcome Research Study. British Journal of General Practice 1999;49:31–4. Greenwood J. Persuading general practitioners to prescribe-good husbandry or a recipe for chaos? British Journal of Addiction 1992;87:567–75. Gronbladh L, Ohlund LS, Gunne LM. Mortality in heroin addiction: impact of methadone treatment. Acta Psychiatry Scandanavia 1990;82:223–7. Gruer L, Wilson P, Scott R, Elliott L, Macleod J, Harden K, Forrester E, Hinshelwood S, McNulty H, Silk P. General practitioner centred scheme for treatment of opiate dependent drug injectors in Glasgow. British Medical Journal 1997;314:1730– 5. Hindler C, King M, Nazareth I, Cohen J, Farmer R, Gerada C. Characteristics of drug misusers and their perceptions of general practitioner care. British Journal of General Practice 1996;46:149– 52. Keen J, Rowse G, Mathers N, Campbell M, Seivewright N. Can methadone maintenance for heroin-dependent paients retained in general practice reduce criminal conviction rates and time
.
89
spent in prison? British Journal of General Practice 2000;50:48– 9. Leaver EJ, Elford J, Morris JK, Cohen J. Use of general practice by intrvenous heroin users on a methadone programme. British Journal of General Practice 1992;42:465– 8. Martin E, Canavan A, Butler R. A decade of caring for drug users entirely within general practice. British Journal of General Practice 1998;48:1679– 82. McKeganey NP, Boddy FA. General Practitioners and opiate-abusing patients. J R Coll Gen Pract 1988;38:73– 5. Newman R, Whitehill WB. Double-blind comparison of methadone and placebo maintenance treatments of narcotic addicts in Hong Kong. Lancet 1979;ii:485– 8. Telfer I, Clulow C. Heroin misusers: what they think of their general practitioners. British Journal of Addiction 1990;85:137– 40. Wilson P, Watson R, Ralston G. Methadone maintenance in general practice: patients, workload and outcomes. British Medical Journal 1994;309:641– 4.