Tubercle Tuber& (1991) T&261-264 @LmgmmGroupUKLtd1991
Factors associated with poor patient compliance with antituberculosis therapy in Northwest Perak, Malaysia S. Y. CHUAH Department of Medicine, Leicester General Hospital, Leicester, UK
smmary- A retrospectivestudy of factors associated with poor patient compfiancewith antitubercUosistherapy was conducted in Taiping, Perak. 219 patients were studied. Male patients and hospital referrals were significantly more likely tp default. Patients with tubemubus lymphadenitisalone had a greater rate of default, but this just failed to reach signific;ande(0.05 < p e 0.10). Six of 7 mate hospital referrals with tubemubus lymphadenitisatone defautted. Patients treated as outpatients from the start were more compliant. Housewiveswere also highly compliant. It was noticed that patients who defaulted tended to do so during early stages of treatment.
Introduction
Patientswho had startedtreatmentin 1982 were studied in August 1984. All 227 were expected to Worldwidepatient compliance with antituberculous tegitherapyhasbeenpoorandremainstheprincipalcause have completedtheii 14month antituberculosis menbymid-1984.Dataoneachpatientwerereuxded of treatmentfailure[l, 23. This studywas undertaken todetenninesomeofthefactorsassociatedwithpoor onastandardformandincludedname,age,ethnic patient compliancewith antituberculosistherapyin origin,sex, place of residence,detailsof familyhistoryofdiseaseanditstmatment NorthwestPerakin Malaysia.The areais servedby a districtgeneralhospitalin Taiping.The countryhas SnlrceofreferraL Ttkesources to this informationincluded the pahada NationalTIrberculosis ControlRogmmmesince dents’ clinical pogress cards, drug ucntmemcards 1961 and its objectivesare to fbui infectious cases andtheTBCla(Rev72)formsoftheContmtTuberand renderat least 95%of thempermanentlynon-infectious with adequatetreatmentand to maintainat culosis Registrywhich are completedfor every new least 75% herd immunityat all times throughBCG patientregisteredin the country. Patientswere consideredto have defaultedif they vaccination. neededadditionallettersorhomevisitstorecallthem for tEatmen regardlessof the outcome of these Patients and methods measures.InformatkmaboutthedistanceofaparThe catchmentareaof Taipmgdistrictgeneralhospititular village or town from Taiping tal includes: hoi@alwasobmiuedfkomtheclinic ofhista&swastodrivetheclinicnuaset0peticnts’ 1. Larut,
[email protected]= bomcswheneWcftheneedalosc. 268 369). 2. KrianQqulation = 165 525). Datawekanalysedby means of the Minitabstatisticalcomputerpackage. 3. Bemas and Dinding (population = 153 361). conupmdsnce to:DrS.Y. C&u& bpmtnmt
of Medicine. Leiccatcr Gcneml Hospital. Leiancr LES 4IW UK.
261
262 Results Of the 227 patients who started treatment in 1982,146 completed treatment without default, 73 defaulted at least once, and 8 died as inpatients during their initial intensive therapy. These 8 deaths were excluded from the study. The default rate was consequently 33%. There were no significant differences between defaulters and regular attenders in terms of age, ethnic origin, area of residence, distance from hospital, month of presentation, personal history of tuberculosis, treatment regimen, sputum status or radiology at presentation. Variables which were significantly different included sex, occupation, source of referral and whether initial treatment was as inpatient or outpatient (Table). In 1982, 148 new tuberculosis patients were males. Based on equal numbers of males and females in the population, the incidence of tuberculosis was significantly higher in men (~2 = 27.792, p < 0.01). 91% of new cases had pulmonary tuberculosis, 6% tuber-
culous lymphadenitis, and 3% had other forms of extrapuhnonary tuberculosis. More than a third of new tuberculosis patients in 1982 were unemployed. The second largest group was personal service workers, semi-skilled and unskilled manual workers. Housewives formed 15% of the case load. Group 1 which consisted of professionals, nonmanuaI workers and manual workers in a supervisory role were most compliant with no defaulters. The compliance rate for housewives was 91%. More than 90% of new cases were hospital referrals. 60% of new cases were treated as inpatients for some part of their 2-month initial intensive therapy. Just over 10% of new cases were transferred from other chest clinics. The remainder were treated entirely on an outpatient basis. Of the 73 patients who defaulted, 18 (expected 10) did so during the 2-month initial intensive phase, and 55 (expected 63) in the subsequent 12-month continuation phase (~2 = 7.42, p < 0.01).
Table The features of patients defaulting from treatment for tuberculosis in Northwest Pen& Malaysia in 1982
Feature Sex Mak!S Females Occupational groups 1. Professionals, non-mambsl workers; manual workers in supervisory role 2. Skilled mamtal workers 3. Own-account workers Farmers :: Personal service, semi-skilled and unskilled manual workers Agricultural workers $: Armed forces 8. Unemployed 9. School and pre-schoal children 10. Hatsewives 11. Pensioners 12. Prisoners. imnatcs of institutions Source of referral Hospital Others Initial treatment Inpatient Outpatient Outside the region Site of tuberculosis Pulmonary Tuberculous lymphadenitis Extrapulmonary Pulmonary and extrapulmonary
Default Yes
No
x2adp
61 12
87 59
0 3 4 0
11 2 5 2
18 6 1 30 4 3 1 3
23
71 2
131 15
x2 = 3.86 p < 0.05
49 14 10
83 49 14
x2 = 5.10
62 8 1 2
137 5 3. 1
~2 = 12.77 p < 0.01
x2 = 25.25 p < 0.01
9 46 I 30 6 8
p < 0.05 x2 = 6.70 p < 0.10
PATIENT CO MPLIANCEWIl’HTHERAW
Diussion
263
Women patients in Northwest Perak were usually of childbearing age and were significantly more comPatient characteristics such as age, sex, sociciizo- pliant with treatment than men. The reasons for this nomic status, ethnic&y and area of residence cannot are unknown but may reflect concern about the spread be controlled or significantly altered to improve treat- of disease to children or simply a greater health ment compliance. However, a knowledge of features awareness as in the case for occupational group 1. In Malaysia, if a hospital patient is found to have associated with poor compliance can help identify groups at risk of defaulting and lead to improved tuberculosis, the responsible clinician has a legal oblipatient education. Patient compliance with therapy is gation to refer him or her to the chest clinic, i.e. the affected by issues of belief, health motivation, per- patient is not the party taking the initiative. However, ceived susceptibility to disease and its severity, views other forms of referral involve the patient taking the on the benefit of professional intervention and knowl- initiative to consult his or her family doctor, the local edge of the condition [3,4]. These latter factors can health centre or going directly to the chest clinic. The be moditied by improving the patient-doctor relation- concern of these patients about their health probably leads to their greater compliance with treatment. ship and through health education [5]. Traditional treatment regimens begin with a short In this study, the distance from Taiping district genperiod of daily treatment in hospital which allows eral hospital and place of residence had no sign&ant effect on defaulting behaviour. This is probably due close observation of therapeutic responses and drug to the network of government health centres scattered toxicity, and gives an opportunity for health education. In this study, such an approach gave no advanthroughout the country. In Northwest Perak, age and ethnic origin did not tage in terms of compliance with treatment but may affect compliance. This was surprising as the three reflect a clinical decision to offer outpatient treatment major races in Malaysia have different religious be- only to selected patients. Having identitled patients who are likely to default, liefs, and the attitudes of older and younger generations also differ. In contrast in rural Pakistan 141, health education could then be efficiently directed to inbred fears and supernatural beliefs were two major maximise patient compliance. factors affecting patients’ compliance with treatment. Occupations, such as paddy planting and sea fish- Acknowledgement ing, are seasonal and could be expected to affect the defaulting behaviour. However, a detailed analysis of I thnnkDr J.F. Mayberry, Cats&ant Physician, Leicester General the Krian data where the predominant occupation is Hospital, for his invaluable advice. paddy planting, showed no such seasonal variation References (x2 = 7.73, p > 0.70). Although having a relative with tuberculosis might 1. Fox W. The problem of se.&administration of drugs; with parinstil fear into a new patient, it had no effect on deticular reference to pulmonary tuberculosis. Tuber& 1958; faulting behaviour which is consistent with the find39: 269-274. 2. Anderscm S. Bane@ D. A socidogical enquiry into an urban ings of Blackwell [6] that fear may not be a positive tuberculosis control programme in India. Bull WHO 1%3; 29: motivational factor in health behaviour. 685-700. When medication is directly administered under su3. Rosmstodr IM. Patient compliance with health regimens. pervision, the prescribed regimen is adhered to. Also, JAMA 1975; 234: w. 4. Sloan JP, Sloan MC. An assessment of default and nanama number of studies confirm good long-term compliance in tuberculosis camd in Pakistan. Trans R Sot Trap pliance when parenteral treatment is used [7-131. In Mad Hyg 1981; 75: 717-718. Malaysia, most patients am on supervised intermit5. Francis V, Korsch BM. Morris MJ. Gaps in doctor-patient tent chemotherapy which consisted of streptomycin canmunicati~, patient’s response to medical advice. N Engl and isoniazid with or without rifampicin or pyraziJ Med 1969; 280: 535-540. 6. BlackweB B. Trutment adherence. Br J Psych 1976; 129: namide. (In April 1984, a new 6-month regimen of 513-531. quadruple chemotherapy in the initial intensive phase 7. Feinstein AR, Wood HF. Epstein JA d al. A controlled study of 2 months and triple chemotherapy in the continuaof three methods of pmphyIaxis against w infection phase of 4 months was introduced to replace the tion in a poplath of rheumatic children. II. Residtr of the first thme Yeats d study, including methods for evaluating old 14-month regimen). A patient may request oral the maintenance of oral pmphylaxis. N Engf J Mcd 1959;260: treatment but the final decision rests with a clinician. 697-702 In Northwest Perak, patients on oral regimens were 8. Feinstein AR, Spagnuolo M. Jonas S ez al. Pqhylaxir of teless compliant (42.9% defaulted), although this did current rheumatic fever. menpeutic - m oral paticillin VI manthly injections. JAMA 1968; 206: 565-568. not reach statistical significance.
264 9. Johnson DAW, Freeman H. Long-acting tranquilizers. Pracrilionei 1972; 208: 395-400. 10. Stradling P, Poole GW. Twice weekly streptomycin plus isoniazid for tuberculosis. Tubercle 1970; 51: 44-I7. 11. Strong EJ. Intermittent chemotherapy in the northern territory of Australia. Med J AIM 1970; 2: 948-950.
CHUAH
12. Chaulet P, Larbaoui D, Grosset J et al. Intermittent chemotherapy with isoniazid and streptomycin in Algiers. Tuber& 1967; 48: 128-136. 13. Albert RK, Sbarbaro JA, Hudson LD et al. High dose ethambutol: its role in intermittent chemotherapy, a six-year study. Am Rev Respir Dir 1976; 114: 699-704.