Factors influencing compliance with antiepileptic drug regimes

Seizure 1997; 6: 87-93

Factors influencing drug regimes

compliance

DEBORAH BUCK*, ANN JACOBY*,

GUS A. BAKEFIt & DAVID W. CHADWICK?

‘Centre for Health Services Research, University Neurosciences, Walton Hospital, Liverpool, UK Correspondence Newcastle-upon-Tyne,

to: MS Deborah Buck, Newcastle-upon-Tyne,

Centre

with antiepileptic

of Newcastle-upon

for Health Services NE2 4AA, UK

Research,

Tyne, UK and fDepartment

21 Claremont

Place,

University

of

of

Failure to comply with drug regimesis prevalent amongst patients with epilepsy and the consequenceof this is often an increasedrisk of further seizures.This paper describesthe level of, and influencesupon, non-compliance with antiepileptic drug (AED) treatment. A postal questionnaire was sent to an unselected,community-based population of patients with epilepsy. This instrument included questions about patients’ AED treatment, any related side-effects,and AED-taking behaviour. Univariate analysisshowedthat factors associatedwith compliance were patient age, how important patients felt it wasto take drugs asprescribed,whether patients reported feelings of stigma, whether on mono- or polytherapy, whether they were experiencing any side-effectsbecauseof AEDs, whether patients had a regular arrangementto seetheir GP about epilepsy and how easy they found their GP to talk to. Multivariate analysisshowedthat the strongestpredictors of non-compliancewere feeling it wasnot very or not at all important to take AEDs asprescribed,being a teenager, being aged under 60 and being on monotherapy. Further implementation of educational programmesfor people with epilepsy would help to improve levels of compliancethereby reducing the risk of unnecessaryseizures. Key words: epilepsy: compliance:antiepileptic drugs; health care.

INTRODUCTION The sole treatment available for the majority of patients with epilepsy is antiepileptic drug (AED) therapy’. Research has shown that missing or altering antiepileptic drug dosages can have adverse reactions, that is, increase the chance of seizure recurrence. For instance, Mattson er af* report that over one quarter of seizures occurred at or before reports of inadequate medication levels, and Stanaway et al3 found that over one-third (38%) were due to either missed medication or inadequate drug levels. Failure to follow prescribed drug regimes is known to be fairly widespread amongst patients with epilepsy: research has shown that between 25 and 75% fail to adhere”‘, in particular very young or very old patients”, and teenagers”. Given that failing to comply may increase the likelihood of hospital admission”, there could be a potential saving to the National Health Service if compliance could be improved. The use of the term ‘compliance’ in relation to 1059-131

l/97/020087

+ 07

$12.00/O

health care became commonplace amongst physicians during the late 197Os,and it was then that it came to be recognized as a vital factor in the assessment of treatments”. The definition of compliance/non-compliance is not straightforward, however. Ley ‘* lists among its aspects not taking the correct dosage (too much or too little), failing to leave the recommended length of time between doses, not taking medication for the duration specified, taking other drugs not prescribed. Some commentators believe that any patient who fails to adhere on one occasion or more is non-compliant whereas others feel that only those who fail to comply more than occasionally, say at least 25% of the time, should be classed as non-compliant’*. Just as the definition of compliance is problematic, so too is its measurement. Various methods used include patients’ reports, pill counts, blood tests, outcome (i.e. shifts in behaviour or changes in the condition) and doctors’ estimations-the latter is commonly 0

1997

British

Epilepsy

Association

88

believed to be the least valid method’*. The method of counting pills in order to verify whether or not the correct number have been used up has also been widely criticized-patients have been known to empty out pills without actually taking them. A recently-developed electronic device which records the date and time on each opening of the bottle13 has been heralded as a significant advance in the measurement of compliance-but even with such sophisticated methods there can be no guarantee that pills are actually taken unless a physician or researcher stands guard over each individual patient every time their doses are due. The term ‘compliance’ has been frowned upon because it has connotations of subservience. Stockwell Morris and Schulz14 argue that the term non-compliance implies deviance and so is inappropriate, since patients’ decisions result from a reasoned attempt to balance not only clinical but also social, psychological and other factors. Although some have suggested ‘nonadherence’ as an alternative, less negative term others argue that there appears to be no other single word more suitable”. We have opted to refer to ‘compliance’ and ‘non-compliance’ throughout this paper, simply because they seem to be the most commonly-used expressions. Although the bulk of the literature on compliance grew out of a concern with clinical outcomes, some studies with a wider focus have examined personal and other non-clinical outcomes’“. According to Stockwell Morris and Schulz’4, non-compliance is usually intentional and rational, in that patients assess the effects of their medication on their everyday life generally and the side-effects on their social functioning specifically. Pryse-Phillips, et a/l5 argue that to regard compliance as merely a patient’s ability to comply with a drug regime exactly as instructed is too simplistic and they define it instead as ‘an inclination or readiness to yield to the demands of others’. They also suggest that there are two types of compliance: firstly, compliance is simply the ‘regular ingestion of medication as and when prescribed’; secondly, compliant (or insightful) behaviour is that where patients have an active and informed part to play in a therapeutic situation. This sort of insightful behaviour (e.g. reducing one’s dosage because of a particular side-effect and subsequently improving outcome) was displayed by 20% of the patients in Pryse-Phillips et al’s study. Schneider and ConradI also identified patients who exhibited ‘compliant behaviour’ and defined these as

D. Buck

et

al

‘self-regulators’. According to Pryse-Phillips er al”, this sort of behaviour may be just as vital to condition-management as the traditionallydefined form of compliance. Research has shown that there are several factors which hinder compliance. Although forgetfulness plays a large part, together with confusing advice and fear of dependence or side-effects4, factors such as doubting the diagnosis, uncertainty about the necessity for drugs and anxiety over the complexity of the drug regimen also contribute’. Another influential factor is patients’ perceptions of their physician: Freeman et al” found that patients were more likely to take their medication as required when they perceived their physicians as being easy to talk to. Ley and Morris” found providing patients with written information resulted in better adherence rates; though numerous studies have shown that informing patients has no effect14.“. The present paper reports on the degree to which a sample of adult patients with epilepsy missed taking their AEDs and examines various factors associated with this non-compliance.

STUDY

DESIGN

AND

METHODS

The data reported here are derived from a community-based study of epilepsy undertaken in 1993 in one UK health region. Individuals with a history of seizures during the previous two years, or seizure-free in that time but taking AED medication, were defined as eligible for the study. They were identified from the records of 31 general practices, selection of the practices being stratified by health district and practice size. Study methods have been described in detail elsewhere”. Nine hundred and seventy-five subjects aged 16 or over were sent a postal questionnaire and 769 were returned of which 40 were rejected as unusable (more than 10% of relevant questions had been unanswered) and a further 33 because they had been completed by a proxy informant. Thus, 696 usable questionnaires were available for analysis (a response rate of 71%). The postal questionnaire included questions about the clinical nature and history of subjects’ epilepsy, demographic details and medical care (general practitioner and hospital clinic). It also included a previously validated seizure severity scale2’, comprising two sub-scales related to patients’ perception of control (such as the timing of the seizures, whether or not the patient could predict them) and ictal and post-ictal effects (such

Factors influencing

compliance

wtth AED regimes

as loss of consciousness or post-ictal confusion). Measures of psychosocial functioning included: the hospital anxiety and depression scale (HAD)“; the impact of epilepsy scale**; the life fulfilment scale23; and the stigma of epilepsy scale”. Information was sought about current AED therapy, how well patients felt the drugs controlled their attacks, whether or not they believed it was important to take drugs as prescribed, and side-effects of drugs using a recently developed adverse events profile=. Drug-taking behaviour was examined by asking patients whether and how regularly they missed taking their AEDs. This was the only measure of compliance used in the present study, and is the method used most often in research’*. The SPSSx statistical package for social sciences26 was used to analyse the data. Univariate analysis consisted of contingency table analysis, with differences using chi squared tests reported at the 5% and 1% level: differences in proportion and 95% confidence intervals (CIs)*’ are also reported. Multivariate analysis employed logistic regression, where the dependent variable was whether or not patients complied with their AED therapy regimes: odds ratios are also reported.

RESULTS

Out of 696 patients, 95% were taking AEDs (68% were on one type of drug only, i.e. monotherapy, and 27% were on polytherapy). Of these 95%, almost three-quarters (72%) of patients said they never missed taking their AEDs, 15% missed less than once a month, 9% missed more than once a month but less than weekly and 4% missed at least once a week. Patients were asked how well they felt AEDs controlled their attacks: 61% said they were very well controlled, 32% said fairly well, 6% said not very well and 1% said not at all. Fifty per cent of patients taking AEDs reported experiencing AED side-effects, the most commonly reported being related to the central nervous system: tiredness (80% of those on AEDs), memory problems (71%) concentration (63%) sleepiness (63%) depression (60%) and headaches (58%).

89

variable was collapsed into three categories: never missed, missed less than once a month and missed once a month or more, for the purpose of univariate analyses.

Patient characteristics

Table 1 shows that patients’ age was significantly associated with whether or not they missed taking their tablets, with younger patients (i.e. those under 60) being less likely than older ones to always comply (66% and 86%, respectively, difference in proportions 20, 95% CIs 13-27). Furthermore, teenagers were less likely than others to always comply (52% compared with 72% of those aged over 19, difference in proportions 20, 95% CIs l-39), and were more likely to miss infrequently (though there were no differences between teenagers and others in the percentages missing often). Social class (x2 = 4.85, P = 0.09) and gender of patient k* = 1.74, P = 0.04) were not significantly related to likelihood of missing medication. Not surprisingly, patients who felt that it was very important to take their tablets as prescribed were the group most likely to comply-76% always complied compared with 29% of those who felt it was only fairly or not at all important (difference in proportions 49, 95% CIs 38-60). There was a significant relationship between whether or not patients reported feeling stigmatized by their epilepsy and likelihood of compliance, with only 66% of those who felt stigmatized saying that they always complied compared with 74% of those who did not feel stigmatized (difference in proportions 8,95% CIs 1-15).

Clinical characteristics

Factors influencing compliance to drug regime

There was no relationship between whether or not patients missed taking their tablets and seizure frequency (x2 = 2.79, P = 0.6), seizure type h’= 4.68, P = 0.3) whether patients felt their attacks were well-controlled (x2 = 3.19, P = 0.2) seizure severity (ICTAL P = 0.05, odds ratio 0.1; PERCEPT P = 0.8, odds ratio O.l)*, length of time seizure-free (P = 0.5, odds ratio 0.2) or duration of epilepsy (P = 0.07, odds ratio

Given that the number of patients who said they missed taking their AEDs at least once a week was so low (n = 31), the ‘frequency missed’

‘The relationship between compliance and continuous variables such as seizure severity and duration of epilepsy were each examined using logistic regression techniques.

D. Buck

90 Table

1: Factors

affecting

compliance

with AED regimes Frequency

miss taking


Age: Under 60 (n = 460) 60 or over (n = 180)

66 86

18 8

17 6

x2 = 26.14 P < O.ooool

Teenager (n = 25) Over 20 (n = 615)

52 72

32 14

16 14

x2 = 6.66 P < 0.05

76

15

9

29

17

53

66 74

19 I3

15 13

x2 = 6.82 P < 0.05

(11 = 467) 01 = 190)

68 82

I7 9

15 9

x2 = 13.61 P
Side-effects due to AEDs: Yes (n = 326) No (/I = 328)

67 77

I8 I2

I6 11

x2 = 9.6 PCO.01

How perceive general practilioner: Easy to talk 10 (n = 394) Not easy (n = 63)

73 57

I4 24

14 19

x2 = 6.58 P < 0.05

Have IO see Yes No

81 68

19* 32

Reported feelings Yes (n = 245) No (n = 394) No. of drugs Monotherapy Polytherapy

a month

AEDs:

Never %

How important to take drugs as prescribed: Very important (n = 597) Fairly/not at all important ()I= 64)

et al

at least once a month %

x2 = 100.50 P
of stigma:

regular arrangement GP about epilepsy: (n = 69) (n = 382)

x2 = 4.61 P < 0.05

* Although there was no significant difference between never missing, missing less than once a month month and having a regular arrangement to see GP, the difference was significant when the ‘frequency collapsed into two categories: whether missed at all or never missed.

0.2). Patients on monotherapy were less likely than those on polytherapy to always comply with their medication (68% and 82%, respectively, difference in proportions 14, 95% CIs 7-21). Patients who reported experiencing side-effects because of their AEDs were somewhat less likely to comply than those who did not (67% and 77%, difference in proportions 10, 95% CIs 3-17). Although compliance levels were not directly related to seizure frequency, patients on monotherapy had less frequent seizures than others, with 22% of those on just one type of drug having seizures at least once a month compared with 33% of patients on more than one type (x2 = 9.70; P ~0.01). Also, patients on monotherapy reported a significantly shorter duration of epilepsy than those on polytherapy (49% had had epilepsy for at least ten years compared with 77% of those on polytherapy, x2 = 49.45; P < O.OOOOl), were less likely to report experiencing side-effects (46% compared with 59%, x2 = 8.11; P < 0.01)

or missing at least once a missed’ variable was

and were more likely to consider their attacks well-controlled (95% compared with 87%, x2 = 10.92; P < 0.001). Physician

characteristics

Despite some evidence indicating that information provision increases compliance*“, no relationship was found between amount of information received from the GP (x2 = 2.54, P = 0.3) or clinic doctor 01’ = 1.72, P = 0.4) and levels of compliance. However, feeling that the doctor was easy to talk to was important, at least in relation to GPs; of those who found their GP easy to talk to, 73% always complied compared with only 57% of patients who did not find their GP easy to talk to (Table 1, difference in proportions 16, 95% CIs 3-29). There was no association between how easy the hospital clinic doctors were to talk to and level of compliance &* = 1.15,

P = 0.6).

Factors influencing

compliance

with AED regimes

Having a regular arrangement to see the GP about epilepsy was significantly related to whether patients missed taking drugs or not: those who did have such an arrangement were less likely to miss than those who did not (19% compared with 32%, difference in proportions 13, 95% CIs 3-23). There was no significant relationship between compliance and the number of times patients had seen their GP (P = 0.2, odds ratio 0.2) or clinic doctor (P = 0.05, odds ratio 0.1) in last year. A logistic regression whereby all clinical and other patient characteristics were included in the model (Table 2) showed that the best predictors of missing AEDs were: feeling it is not very or not at all important to take AEDs as prescribed, being a teenager (i.e. 16-19), being aged under 60, being on monotherapy.

DISCUSSION

The nature of the present study and the methods of data collection employed precluded the possibility of any pill counting procedures or blood level testing, and information about compliance was therefore limited to patients’ own accounts. However, Ley” notes that studies of correlations between patients’ reports and mechanical methods indicate a mean correlation of +0.80, and results from patients’ reports will generally be akin to those of other methods. Despite this Table 2: Explanatory model and variables Included

clinical variables included predicting missing AEDs

in the

in the model:

Whether on monotherapy or polytherapy Seizure frequency Seizure type Duration of epilepsy Seizure severity Whether or not have any side-effects due to AEDs Number of side-effects Whether or not feel attacks are well-controlled by AEDs How important feel it is to take AEDs as prescribed Whether feel stigmatised because of epilepsy Gender Social class Whether aged under 60 or 60+ Whether a teenager (i.e. 16-19 years old) Predicting

variables:

Feel it is not very or unimportant to take AEDs as prescribed Teenager Aged under 60 Beine on monotheraov

P<

odds ratio

0.0001 0.05 0.02 0.03

15.6 5.4 3.2 2.2

91

possible limitation, our findings verify much previous research into the issue of compliance, and contribute to our knowledge of the drugtaking behaviour of adults with epilepsy. Previous studies indicate that non-compliance with AED therapy is fairly evenly spread throughout social classes and between gender’ and in this study also, no significant differences were found between either social class or gender and levels of compliance. Age, however, has been shown to be important. In particular, failure to comply with AED therapy is common amongst young people29; Takaki et af9 found that teenagers were the most erratic in terms of complying. In the present study, too, although the numbers of teenagers was small, they were significantly more likely than others to miss taking their AEDs. Clearly, on top of the general problems faced by any teenager, the effects of being in some way ‘different’, because of having a condition which peers do not comprehend, can lead to a denial of the presence of the condition manifested by non-compliance with drug regimes; equally, for some teenagers non-compliance may be attributable to the lack of a clear understanding of why it is important both to take AEDs and to take them precisely as prescribed2’. There were no significant differences between whether or not patients missed taking drugs and some basic clinical characteristics, namely seizure type, seizure severity, duration of epilepsy, length of time seizure-free and, perhaps surprisingly, seizure frequency. We did find that patients on polytherapy were more likely to comply than those on monotherapy and other studies have had similar results? this could be due to the fact that patients on polytherapy have a history of more frequent seizures whatever their current seizure activity, and so come to feel it is more important to comply with their medication regime. It was also the case that subjects who reported experiencing any AED side-effects were more likely to be non-compliant, a finding which corresponds to that reported in the Veterans’ Administration Study2. It may be that the nature of side-effects associated with AEDs-often central nervous system-related-increases the likelihood of noncompliance because such side-effects are more obtrusive on psychosocial functioning than those of other drugs. A review of studies on the importance of written information’8 suggests that information provision leads to greater levels of compliance. It has been argued, however, that ability to comply with a medication regime depends not only on understanding and following doctors’ orders but

92

also, and perhaps more importantly, on how these fit into patients’ lives3’-a sound understanding of information is no guarantee of compliance. In the present study, univariate analysis showed that satisfaction with the amount of information given by GP or clinic doctor had no effect on the likelihood of missing taking AEDs, but that patients who thought it was unimportant to take medication as prescribed were less likely to comply: this could reflect the possibility that they were making a decision based on their lifestyle and personal experience, rather than simply following doctor’s orders blindly. Patients may well have been displaying the sort of ‘insightful behaviour’ which Pryse-Phillips et al I5 described and which is reported in a number of studies of patients with epilepsy. Scambler and Hopkins3’, for example, found that some patients took it upon themselves to alter their drug-taking behaviour because of adverse side-effects, fear of addiction or a general dissatisfaction with treatment and Conrad32 found that patients with epilepsy would test whether or not they still had the condition by decreasing or ceasing to take prescribed drugs. However, an alternative explanation is that information, particularly regarding the issue of the importance of compliance, may not have been sufficiently targeted. Another potential contributing factor in noncompliance may be the patient’s perception of the doctor’4: if the doctor is viewed as concerned, and prescribing of medication as a reflection of that concern, then compliance may be higher than if the prescription is seen as an indication that the doctor has no time for or interest in the patient. In the present study a significant relationship was found between how easy patients found their GPs to talk to and levels of reported compliance, with those feeling they were not easy to talk to being less likely to comply than those who felt they were. Similar findings are reported in earlier research”. The amount of contact that patients have with their physicians may also be an influential factor. Wannamaker et aP3 discovered that patients who saw their doctors most often were most likely to comply. Although in the present study there was no significant relationship between compliance and the number of times patients saw their GP or clinic doctor during the previous year, those patients who had a regular arrangement to see their GP specifically about epilepsy were more likely to comply; this may well reflect the possibility that regular contact with a clinician offered an opportunity to discuss any problems arising over drug-taking and to emphasize the

D. Buck

et a/

importance of good compliance in the control of seizures. Scambler and Hopkins3’ emphasize that many patients are not passive, but play an active part in managing their own condition, and at times this will be in conflict with the advice of their doctor. Of particular concern to some patients is the stigma associated with having epilepsy: if continued therapy is seen as implying continued epilepsy then, it has been argued, by reducing or missing their AEDs some patients are seeking to renegotiate an undesirable labe13’.32. In the present study, patients who reported feelings of stigma associated with their condition were more likely than others to miss taking their AEDs. This study has highlighted a number of reasons why patients may not comply with their medication regimes, and a key predictor of noncompliance was feeling that it was not very or not at all important to take AEDs as prescribed. One possible way to encourage compliance is through specific educational programmes such as those developed by Helgeson and Mittan” for epilepsy and by Bailey et aP4 for asthma. Implementation of such programmes for people with epilepsy, which acknowledge the rational basis for non-compliance and the factors influencing it, would perhaps help to improve levels of compliance and so reduce the risk of unnecessary seizures, as would specific programmes for teenagers which would need to take into account the particular impact that epilepsy has on their self-perception and behaviour and equip them with sufficient knowledge about their condition and the treatment they receive for it. ACKNOWLEDGEMENTS

The authors would like to thank the Wellcome Trust for their support and financial backing in this study, the patients who took the time and trouble to participate, and Pauline Potts at the Centre for Health Services Research (CHSR) for her assistance with the data analysis. Also, thanks to Professor John Bond and Dr Lois Thomas at the CHSR for their comments on an earlier draft of this paper. REFERENCES 1. Richens. A. Drug Trea~menr of Epilepsy. Chicago, Year Book Medical Publishers. 1976. 2. Mattson, R.H., Cramer, J.A. and Collins, J.F. VA Epilepsy Cooperative Study Group. Aspects of compliance in epilepsy: taking drugs and keeping clinic

Factors influencing

3. 4. 5. 6. 7.

8. 9.

10. 11. 12.

13. 14. 15. 16. 17. 18.

19.

compliance

with AED regimes

appointments. In: Compliance in Epilepsy (Eds D. Schmidt and I.E. Leppik). Amsterdam, Elsevier, 1988: pp. 111-117. Stanaway, L., Lambie, D.G. and Johnson, R.H. Noncompliance with anticonvulsant therapy as a cause of seizures. New Zealand Medical Journal 1985: 98: 150-152. Eisler. J. and Mattson, R.H. Compliance in anticonvulsant drug therapy. Epilepsia 1975; 16: 203. Christensen, D.B. Drug-taking compliance, a review and synthesis. Health Services Research 1988: 13: 171-187. Scambler, G. Epilepsy. London, Routledge, 1989. Cramer, J.A. and Mattson, R.H. Monitoring compliance with antiepileptic drug therapy. In: Patient Compliance in Medical Practice and Clinical Trials (Eds J.A. Cramer and B. Spilker). New York, Raven Press Ltd, 1991: pp. 123-137. Blackwell, B. Patient compliance. New England Journal of Medicine 1973: 289: 249-252. Takaki, S., Kurokawa, T. and Aoyama, T. Monitoring drug noncompliance in epileptic patients: assessing phenobarbital plasma levels. Therapeuric Drug Moniforing 1985: 7: 87-91. Ausburn, L. Patient compliance with medication regimes. In: Advances in Behavioural Medicine (Ed J.L. Sheppard). Sydney. Cumberland College, 1981: 1: pp. 227-240. Feinstein, A.R. On white-coat effects and the electronic monitoring of compliance. Editorial. Archives of Internal Medicine 1990; 150: 1377-1378. Ley, P. Improving patients’ understanding, recall, satisfaction and compliance. In: Health Psychology-Processes and Applications (Ed A.K. Broome). London, Chapman & Hall, 1987: pp. 74-102. Cramer, J.A., Scheyer, R.D. and Mattson, R.H. Compliance declines between clinic visits. Archives of lnrernal Medicine 1990: 150: 1509-1510. Stockwell, Morris, L. and Schulz, R.M. Medication compliance: the patient’s perspective. Clinical Therapeufits 1993: 15: 593-606. Pryse-Phillips. W.. Jardine, F. and Bursey, F. Compliance with drug therapy by epileptic patients. Epilepsia 1982; U: 269-274. Schneider, J. and Conrad, P. Having Epilepsy: rhe Experience and Control of illness. Philadelphia, Temple University Press, 1983. Freeman, B.. Negrete. V.. Davis, M. and Korsch, B. Gaps in doctor-patient communication: doctor-patient interaction analysis. Paediatric Research 1971; 5: 298-311. Ley. P. and Morris, L.A. Psychological aspects of written information for patients. In: Conrriburions to Medical Psychology 3 (Ed S. Rachman, S). Oxford, Pergamon Press. 1984: pp. 117-149. Jacoby, A., Baker, GA., Steen, N. and Chadwick, D. The clinical course of epilepsy and its psychosocial correlates: findings from a UK community study. Epilepsiu 1996: 37:

93 148-161. 20. Baker, G.A., Smith, D.F., Dewey, M., Morrow, J., Crawford, P.M. and Chadwick, D. The development of a seizure severity scale as an outcome measure in epilepsy. Epilepsy Research 1991: 8: 245-251. 21. Zigmond. A.S. and Snaith, R.P. The Hospital Anxiety and Depression Scale. Acta Psychiatrica Scandinavica 1983; 67: 361-370. 22. Jacoby, A., Baker, G.A., Smith, D.F.. Dewey, M. and Chadwick, D.W. Measuring the impact of epilepsy: the development of a novel scale. Epilepsy Research 1993; 16: 83-88. 23. Baker, G.A., Jacoby. A., Smith, D.F., Dewey, M.E. and Chadwick, D.W. The development of a novel scale to assesslife fulfilment as part of the further refinement of a quality-of-life model for epilepsy. Epilepsia 1994; 35: 591-596. 24. Jacoby, A. Felt versus enacted stigma: a concept revisited. Evidence from a study of people with epilepsy in remission. Social Science and Medicine 1994: 38: 269-274. 25. Baker, G.A., Jacoby, A. and Chadwick D.W. The associations of psychopathology in epilepsy: a community study. Epilepsy Research 1996: 25: 29-39. 26. SPSS Inc. SPSS base system user’s guide for SPSS 4.0. New Jersey, Prentice Hall, 1993. 27. Gardner, M.J. and Altman, D.G. Confidence intervals rather than P values: estimation rather than hypothesis testing. British Medical Journal 1986: 292: 746-750. 28. Helgeson, D.C. and Mittan, R. Efficacy of a psychoeducation treatment program in treating medical and psycho-social aspects of epilspsy. Epdepsio 1990: 31: pp. 35-82. 29. McMenamin, J. and O’Connor Bird, M. Epilepsy. A Parent’s Guide. Dublin, Brainwave, The Irish Epilepsy Association, 1993. 30. Hunt, L.M.. Jordan, B., Irwin, S. and Browner, C.H. Compliance and the patient’s perspective: controlling symptoms in everyday life. Culfure, Medicine ond Psychiutry 1989: 13: 315-334. 31. Scambler, G. and Hopkins, A. Accommodating epilepsy in families. In: Living wirh Chronic Illness: The Experience of Patients and their Families (Eds R. Anderson and M. Bury). London, Allen & Unwin, 1988: pp. 156-176. 32. Conrad, P. The meaning of medications: another look at compliance. Social Science and Medicine 1985: 24%2937. 33. Wannamaker, B.B., Morton, W.A.. Gross, A.J. and Saunders, S. Improvement in antiepileptic drug levels following reduction of intervals between clinic visits. Epilepsia 1980: 21: 155-162. 34. Bailey, WC., Richards, J.M., Brooks, C.M., Soong, S.J., Windsor, R.A. and Manzell, B.A. A randomized trial to improve self-management practices of adults with asthma. Archives of internal Medicine 1990; 150: 1654.