Volume 166 Issue 6

Hepatitis C transmission on the north coast of New South Wales: explaining the unexplained Tim J Sladden, Alan R Hickey, Therese M Dunn and John R Beard MJA 1997; 166: 290 For comment see Wodak Readers may print a single copy for personal use. No further reproduction or distribution of the articles should proceed without the permission of the publisher. For permission, contact the Australasian Medical Publishing Company Journalists are welcome to write news stories based on what they read here, but should acknowledge their source as "an article published on the Internet by The Medical Journal of Australia <http://www.mja.com.au/>". Abstract - Introduction - Methods - Results - Exposures - Transmission to sexual partners and offspring - Discussion - Acknowledgements - References - Authors' details - - ©MJA1997 Abstract Objective: To determine the routes of hepatitis C virus (HCV) transmission in an Australian community. Design: Questionnaire-based, cross-sectional survey of notified HCV cases. Subjects and setting: All cases notified to the New South Wales North Coast Public Health Unit between 1 January 1993 and 30 September 1994. Outcome measures: Frequency of potential transmission exposures (parenteral and sexual); most likely primary exposure; HCV infection rates in sexual partners and offspring. Results: 467 subjects responded (47% of resident cases). Of these, all but one reported actual or potential blood exposures (injecting drug user [IDU], 85%; IDU with sharing of injection equipment, 76%; pre-1990 blood transfusions, 6%; other blood exposures, 8%). Most subjects reported multiple exposures and none reported sexual contact as the only potential exposure. Of 233 sexual partners tested for HCV, 83 were positive; 54 of these were questioned and all had other parenteral exposures. Only three children out of 91 children tested were positive for HCV (two expressing maternal antibodies). Conclusions: In contrast with previous studies, possible HCV transmission modes were identified for almost all respondents. Most respondents in this community were IDUs. Non-parenteral transmission appeared minimal. Novel approaches to preventing HCV transmission in IDUs are needed. MJA 1997: 166; 290-293 Introduction Few studies have investigated transmission of hepatitis C virus (HCV) in representative, population-based samples.1-3 Most previous studies have examined patient series,4-6 those at risk (e.g., transfusion recipients and injecting drug users [IDUs])7-9 or other particular groups (e.g., blood donors, pregnant women and prisoners).10-12 Previous studies have also failed to identify exposures in significant numbers of cases (up to 45% of respondents).1-5,8 HCV transmission is predominantly parenteral13 (via shared drug injection equipment, infected blood products [almost entirely before screening was introduced in February 1990], unsterile skin penetration practices [e.g., tattooing, ear/skin piercing, acupuncture], needlestick and "sharps" injuries and shared personal items, such as toothbrushes and razors14 ). However, patient-to-patient transmission (via contaminated anaesthetic circuitry)15 and surgeon-to-patient transmission (via percutaneous injury)16 have both been demonstrated. Sexual transmission without blood contact appears rare,5,6,8,13 but the risk of blood exposure may be increased by sexual contact during menstruation and anal intercourse. Vertical transmission also appears rare.8,13,17 Both sexual and vertical transmission appear viraemia-dependent,8,9,13,17 and may be facilitated by genital lesions.6,17 Nipple trauma may enable postnatal transmission.18 Household transmission is probably restricted to infected personal items.19 Arthropod vectors have not been identified. Despite evidence that there is no risk of transmission via casual contact, community concern about hepatitis C transmission remains evident in the discrimination reported anecdotally by people with HCV. Clarification of how HCV is transmitted is needed both to allay this concern and to allow the development of new prevention strategies.20 One focus of the current Australian hepatitis C epidemic is the north coast of New South Wales (NSW), where the notification rate (201/100 000 residents) is double the NSW rate (103/100 000 residents),21 and nearly three times the Australian average (74/100 000 residents).22 We investigated the mode of transmission in notified cases in residents on the NSW north coast over a 21-month period. Methods All people diagnosed with HCV infection (through duplicate second generation anti-HCV antibody tests) who were notified to the North Coast Public Health Unit between 1 January 1993 and 30 September 1994 were invited to participate. Subject name, diagnosis and contact address were verified with attending doctors. Questionnaires were mailed to subjects with covering letters, consent statements, and reply-paid return envelopes. Non-respondents were recontacted by mail six weeks later. Data, with identifying codes removed, were entered into a restricted-access database. We developed a questionnaire that included demographic questions, a checklist of possible ways the subject may have contracted hepatitis C (see Box 1) and questions about the HCV status of the subject's current sexual partner and, for women, of their children. The questionnaire was pilot-tested on clients of a local sexual health service. The study and questionnaire were approved by the North Coast Region Health Service Ethics Committee. For subjects with multiple potential exposures, exposures were ranked according to expected risk, and the exposure with the highest risk was considered the most likely primary exposure. Parenteral exposures were assumed to be higher risk than sexual exposure and ranked as: High risk -- (in descending order of risk) injecting drug use (IDU) (with and without reported sharing of injection equipment), receipt of pre-1990 blood products, dialysis and transplants; Medium risk -- needlestick injuries, tattooing, or blood splashes into a cut, eye or mouth (i.e., during an accident or fight); and Potential or low risk -- origin from countries where medical services have used inadequate infection control procedures in the past (such as reuse of needles) or where unsterile tattooing or scarification is common (including the Mediterranean area, Eastern Europe, Asia, South America, Africa and the Middle East),23 post-1989 blood transfusion, ear or skin piercing and other clinical procedures. Transmission between study subjects and their current sexual partners was investigated; a modified questionnaire which included only the checklist of potential transmission routes was mailed to all HCV-positive partners. Household and social contact were examined when no other exposure was reported. Subjects with only low-risk or no parenteral exposures were offered testing for HCV RNA by polymerase chain reaction (PCR) to confirm their hepatitis C status. Demographic characteristics of respondents and non-respondents were compared with chi-squared tests and t tests, and sharing of injection equipment by current and former IDUs was compared with chi-squared tests. Results Of 1487 notified cases, 487 were excluded (395, temporary residents or non-residents; 67, with address unknown; seven, aged under 18 years; 18, other reasons), leaving a study population of 1000. Questionnaires were returned by 467 (46.7%). Respondents differed significantly from non-respondents in sex and age distributions: respondents included a significantly higher proportion of women than non-respondents (respondents: 219/467 [47%]; non-respondents: 212/533 [40%]; chi-squared = 4.42, P = 0.035) and were significantly older (respondents: mean age, 37.4 years; 95% confidence interval [CI], 36.6-38.3; range, 18-86; non-respondents: mean age, 34.9 years; 95% CI, 34.4-35.5; range, 18-62); t [unequal variances] = 4.70; df = 852, P < 0.001). Exposures Potential HCV exposures are shown in Box 1. Almost all 467 respondents reported at least one potential transmission exposure and most reported multiple exposures. The potential exposure with the highest expected risk was classified as the most likely primary exposure. Most respondents (398 [85%]) were IDUs (222 men, 176 women), including 287 (72%) former IDUs (although 45 [11%] of these reported stopping during the survey period). None of the 19 subjects aged over 60 years were IDUs. Of the IDUs, 354 (89%) reported having shared injection equipment. Frequency of sharing injection equipment by IDU status is shown in Box 2 (below). Current IDUs reported sharing significantly less than former IDUs ( chi-squared = 5.82; df = 1; P= 0.016). Hepatitis B infection was reported by 184 subjects (39%); 175 of these (95%) were IDUs. All but 11 sharers of injection equipment and all non-sharer IDUs reported other parenteral exposures (Box 1). Among subjects who were not IDUs, the most likely primary exposure (Box 1) was pre-1990 blood transfusion for 30 (6%), dialysis for two (0.4%), needle-stick injuries for six (1%) (including three health care workers, one with a known exposure to hepatitis C virus, a garbage collector and the partner of an IDU), tattooing for four (0.9%), blood splashes for four (0.9%), origin in a "high-risk" country for three (0.6%), post-1989 blood transfusion for two (0.4%), skin piercing for 13 (3%) and medical procedures for four (0.8%). About a third of subjects (151 [ 32%]) reported past or current HCV-positive sexual partners. However, 147 (97%) also had potential blood exposures (141, IDU; three, blood transfusion; and one each, tattoos, needlestick injury and blood splash). The remaining four had low-risk potential blood exposures (pierced ears or skin for three and a clinical procedure for one). Thus, sexual contact did not occur without concurrent or potential blood exposure(s). Sharing of personal items (e.g., toothbrushes) also could not be excluded for subjects reporting sexual contact. For 15 subjects, no sexual exposure and only low-risk potential parenteral exposures were reported. Another subject reported no risk factors. Polymerase chain reaction (PCR) was offered to these 16 to confirm their anti-HCV antibody test results. Ten were lost to follow-up (including the subject with no risk factors), three were confirmed HCV-positive (all had had clinical procedures, two had pierced ears and the third had social contact with an HCV-positive person) and three were HCV-negative, indicating either resolved infections or false positive initial antibody results. Transmission to sexual partners and offspring Three-hundred-and-twenty subjects (69%) had current sexual partners. Of 233 partners who had been tested for HCV, 83 (36%) were positive, 138 negative and 12 had unknown results. Of the positive partners, 80 (96%) were partners of IDUs, with independent parenteral exposures determined for 54 (68%) and unknown for the rest. Of the negative partners, 70 (51%) practised "unsafe" sex (defined as unprotected oral, anal or vaginal sex) with study subjects. Among the 219 women subjects, 173 had had children; 56 had had one child tested for HCV and 35, a second child. Only three children were anti-HCV positive -- two infants expressing maternal anti-HCV antibodies, and a three-year old. Discussion Many studies have investigated HCV transmission in limited patient series or specific groups,4-12 and full identification of exposures has been difficult.1-5,8 Our study was a population-based survey, and we were able to identify potential blood exposures for 99% of respondents. However, the response rate was low (47%), reflecting the difficulties of community-based surveys and, possibly, community sensitivities about bloodborne viruses. In addition, there were small but statistically significant age and sex differences between respondents and non-respondents, with a higher proportion of women among respondents and most older subjects responding. However, none of the 19 subjects aged over 60 years were IDUs. We suggest that non-IDUs would be more likely to respond than IDUs, and that a higher response rate in non-IDUs accounted for the age difference between respondents and non-respondents. Questionnaire comprehension was not thought to be a problem as the north coast population is predominantly English-speaking (96%).24 The high proportion of subjects declaring use of illicit drugs in the 60 years and under age group (89%) implies they were reporting truthfully. Therefore, we postulate that the respondents were representative of those infected with HCV in this community. If more respondents had denied injecting drugs, selection bias due to non-response of IDU subjects might have been a legitimate concern. Our results suggest that, on the NSW north coast, injecting drug use with sharing of injection equipment accounted for transmission in 76% of all people with hepatitis C. Transmission during injecting drug use remained possible in a further 9% who denied sharing injection equipment, as unrecognised contamination of such equipment (e.g., spoons, filters, water or swabs), poor recall or denial of sharing may have occurred. Apart from IDU, other high- or medium-risk blood exposures appeared to be responsible for a further 10% of respondents, and all but one of the remaining 5% had low-risk or potential blood exposures, with 1% of these also having sexual contact. While rates of IDU (past or current) may be relatively high on the NSW north coast, we suggest that almost all HCV transmission is via blood exposure, with varying proportions of different types of blood exposure in different populations. Thus, there appeared to be minimal HCV transmission via sexual, perinatal, household, occupational or social contact, provided blood exposures were avoided. Supplemental testing identified some false positives among the few subjects without obvious blood exposures. While a third of subjects had had sexual contact with HCV-positive partners, all of these had additional blood exposures (high risk for 95%, medium risk for 2% and low risk for 3%). Our results should help allay community concern about HCV transmission via casual contact. Sexual transmission appears to be minimal, but the risk may increase with menstruation, anal sex, concurrent STDs that involve scratching, sores or blisters, and increasing viraemia (often observed in early, acute stages of HCV infection). Uninfected partners should also avoid oro-facial abrasions of infected sexual partners (e.g., from toothbrushing or razor cuts). People with hepatitis C who are contemplating having children would be advised to seek medical advice regarding their HCV-RNA PCR status, hepatic enzyme function, and clinical symptoms as markers of viral activity. The low frequency of unexplained transmission in this study (< 1%) was almost entirely attributable to the high rate of reporting of injecting drug use. Other studies have found much lower rates of injecting drug use in HCV-positive subjects, but have been unable to explain transmission in a much higher proportion (27%-45%),1,3-5 possibly because of reluctance to admit illicit drug use. In our study, the privacy afforded by the self-administered questionnaire may have encouraged more truthful reporting than may occur in an interview or clinic situation. Many IDUs were former users, suggesting not only reluctance of current IDUs to participate, but also that much HCV infection is due to past drug use. Current IDUs shared injection equipment significantly less often than former IDUs, indicating increased awareness of the dangers and the effectiveness of needle and syringe exchange programs. However, sharing of injection equipment remains the commonest route of transmission of HCV, responsible for an estimated 10 000 new infections each year in Australia.25 The proportion of cases due to IDU will increase with blood product screening. This emphasises the need for transmission prevention and harm-minimisation programs, especially targeting adolescents before any experimental drug-taking. Community development and peer education of IDUs to promote safer injecting practices should be strengthened.26 The impacts of improved access to needle and syringe exchange and methadone programs, campaigns to encourage non-injecting routes of drug administration, development of non-reusable syringes,25 supply of heroin to registered users27 and provision of "safe-house" injecting venues on HCV transmission all need to be investigated. Acknowledgements Kieran Mutimer and staff at the Lismore Sexual Health and AIDS Service assisted with piloting and circulating questionnaires. References Alter MJ, Hadler SC, Judson FN, et al. Risk factors for acute non-A, non-B hepatitis in the United States and association with hepatitis C virus infection. JAMA 1990; 264: 2231-2235. Mistry SA. Hepatitis C notifications in the Australian Capital Territory, January to June 1993. Comm Dis Intell 1995; 19: 183-188. Curran M. Acute hepatitis C notifications and associated risk factors in Australia, 1995 first quarter report. Comm Dis Intell 1995; 19: 615-617. Strasser SI, Watson KJR, Lee CS, et al. Risk factors and predictors of outcome in an Australian cohort with hepatitis C virus infection. Med J Aust 1995; 162: 355-358. Weinstock HS, Bolan B, Reingold AL, Polish LB. Hepatitis C virus infection among patients attending a clinic for sexually transmitted diseases. JAMA 1993; 269: 392-394. Tor J, Llibre JM, Carbonell M, et al. Sexual transmission of hepatitis C virus and its relation with hepatitis B and HIV. BMJ 1990; 301: 1130-1133. Crofts N, Hopper JL, Bowden DS, et al. Hepatitis C virus infection among a cohort of Victorian injecting drug users. Med J Aust 1993; 159: 237-241. Meisel H, Reip A, Faltus B, et al. Transmission of hepatitis C virus to children and husbands by women infected with contaminated anti-D immunoglobulin. Lancet 1995; 345: 1209-1211. Bresters D, Mauser-Brunschoten EP, Reesink HW, et al. Sexual transmission of hepatitis C virus. Lancet 1993; 342: 210-211. Kaldor JM, Archer GT, Buring ML, et al. Risk factors for hepatitis C virus infection in blood donors: a case-control study. Med J Aust 1992; 157: 227-230. Fairley CK, Leslie DE, Nicholson S, Gust ID. Epidemiology and hepatitis C in Victoria. Med J Aust 1990; 153: 271-273. Anand CM, Fonseca K, Walle RP, et al. Antibody to hepatitis C virus in selected groups of a Canadian urban population. Int J Epidemiol 1992; 21: 142-145. van der Poel CL, Cuypers HT, Reesink HW. Hepatitis C virus six years on. Lancet 1994; 344: 1475-1479. Davis AR. Contaminated razor blades as a possible source of hepatitis C virus infection [letter]. Med J Aust 1995; 163: 275. Chant K, Kociuba K, Munro R, et al. Investigation of possible patient-to-patient transmission of hepatitis C in a hospital. NSW Public Health Bull 1994; 5: 47-51. Esteban JI, Gomez J, Martell M, et al. Transmission of hepatitis C virus by a cardiac surgeon. N Engl J Med 1996; 334: 555-560. Zanetti AR, Tanzi E, Paccagnini S, et al. Mother-to-infant transmission of hepatitis C virus. Lancet 1995; 345: 289-291. Grayson ML, Braniff KM, Bowden DS, Turnidge JD. Breastfeeding and the risk of vertical transmission of hepatitis C virus [letter]. Med J Aust 1995; 163: 107. Davis AR, Kowalik AM. Hepatitis C virus transmission to heterosexual partner: bedroom or bathroom hazard? [letter] Med J Aust 1996; 164: 126. Strasser SI. Hepatitis C: questions still to be answered [editorial]. Med J Aust 1996; 164: 132-133. Sladden TJ, Hickey AR, Beard JR. Hepatitis C on the North Coast of NSW [letter]. Med J Aust 1995; 162: 166. Longbottom H, Evans D, Myint H, Hargreaves J. Annual Report of the National Notifiable Diseases Surveillance System. Comm Dis Intell 1994; 18: 521, 533. National Health and Medical Research Council Hepatitis C Working Party. Draft report on a strategy for the detection and management of hepatitis C virus in Australia. Canberra: NHMRC, 1996. Census Applications. Socio-economic profile of the north coast of New South Wales. Lismore: North Coast Regional Co-ordination Pilot Program, NSW Premier's Department, 1996. Wodak A, Crofts N. HIV revisited: preventing the spread of blood-borne viruses among injecting drug users. Aust J Public Health 1994; 18: 239-240. Feachem RGA. Valuing the past. . .investing in the future. Evaluation of the National HIV/AIDS Strategy 1993-94 to 1995-96. Canberra: Commonwealth Department of Human Services and Health, 1995 (ISBN 0644356944). Bammer G, editor. Feasibility research into the controlled availability of opioids. Canberra: NCEPH/AIC, 1991 (ISBN 0731512359). (Received 8 Feb, accepted 20 Nov, 1996) Authors' details North Coast Public Health Unit, NSW Health Department, Lismore, NSW. Tim J Sladden, MSc, MPH, Epidemiologist; Alan R Hickey, RN, Research Assistant; Therese M Dunn, BAppSc(Comp), Research Assistant; John R Beard, MB BS, FAFPHM, Director. Reprints: Mr T J Sladden, North Coast Public Health Unit, NSW Health Department, PO Box 498, Lismore, NSW 2480. E-mail: tslad AT doh.health.nsw.gov.au - - To top of article - ©MJA 1997 <URL: http://www.mja.com.au/> © 1997 Medical Journal of Australia.

Tim J Sladden · Alan R Hickey · Therese M Dunn · John R Beard

Reported use of asthma management plans in South Australia Justin J Beilby, Melanie A Wakefield and Richard E Ruffin MJA 1997; 166: 298 For comment see Jenkins & Bauman Readers may print a single copy for personal use. No further reproduction or distribution of the articles should proceed without the permission of the publisher. For permission, contact the Australasian Medical Publishing Company Journalists are welcome to write news stories based on what they read here, but should acknowledge their source as "an article published on the Internet by The Medical Journal of Australia <http://www.mja.com.au/>". Abstract - Introduction - Methods - Results - Asthma management plans - Good management - Discussion - References - Authors' details - - ©MJA1997 Abstract Objective: To examine the use of asthma management plans (AMPs) and specific elements of AMPs in a South Australian community. Design: Cross-sectional questionnaire survey by experienced interviewers. Participants and Setting: 307 adults with current asthma, confirmed by a doctor, who were among 3001 respondents of a representative sample of 4065 South Australian adults recruited by multistage systematic clustered area sampling for a larger study in March 1995. Main Outcome Measures: Prevalence of reported use of AMPs and of specific elements of AMPs, such as knowledge of trigger factors and having an action plan, and the proportion of adults with asthma receiving what we defined as "good management". Results: 46% reported having an AMP; 63% had discussed trigger factors with their doctors and 39.5% had action plans. 82% had been told the severity of their asthma, and 23.3% were receiving good management. Conclusion: Less than half of people with asthma use AMPs, although certain elements within the AMPs are reportedly being used more frequently. More research is required on why AMPs are not more generally used and, more importantly, on what elements within the AMP model are useful and effective for people with asthma and likely to be used by their general practitioners (GPs). MJA 1997; 166: 298-301 Introduction Asthma causes substantial morbidity1 and mortality,2 and costs -- in the range of $585-$720 million in Australia in 1991.3 Approximately 9% of Australians (over 1.4 million people) have reported asthma as "either a recent or long term condition in the last 12 months".4 There is evidence that the prevalence of the disease is increasing,5 and that it is both undertreated and underdiagnosed.6 The variability of the condition in both adults and children further complicates these concerns. In an attempt to create a unified approach to the management of asthma, a group of "accepted experts" developed the Asthma Management Plan (AMP), a set of consensus guidelines for health professionals.7 The Australian AMP comprises six steps: Assessing the severity of asthma; Achieving best lung function; Maintaining best lung function by identifying and avoiding triggers; Maintaining best lung function with optimal medication; Developing an action plan; and Educating the patient and reviewing regularly. These steps are aimed at doctors as an aid to achieving good asthma management. Hence, any reported use of elements of the AMP will, in part, be dependent on the use of these steps by doctors as well as on patient factors. In this study we examine the current use of AMPs and determine the specific elements of the plan that have been adopted by people with asthma and their doctors. Such information is important in planning ongoing asthma management campaigns. Methods In March 1995, a representative population sample of adults in South Australia was obtained for the Health Omnibus Survey,8 as follows. A multistage systematic clustered area sample of households was used to give a sample comprising 75% selected from the Adelaide metropolitan area and the rem ainder from country centres with populations of 1000 or more. At each household, one person aged 15 years or over (the person whose birthday was last) was selected for interview. Interviews were conducted in the respondent's home by trained interviewers, with up to six call-back visits if the selected respondent was not at home. For the purposes of our study, we included a number of questions in the Health Omnibus Survey. Respondents were asked whether they had asthma, and if they answered "yes" they were then asked to respond to the questions listed in Box 1. For the purposes of the interviews, we defined: Asthma management plans as "a coordinated method of management that covers aspects of your asthma -- medication, trigger factors, lung function measurement, etc."; Trigger factors as "incidents that can cause you to have an asthma attack"; and An asthma action plan as "written or verbal instructions of what to do if your asthma is out of control, i.e., having a bad attack". We initially examined questionnaire responses against sociodemographic variables, including age, sex, place of birth, educational attainment, and household income. Conventional chi-squared tests were used for comparisons at the univariate level, and unconditional logistic re gression analysis to adjust for potential confounders. In further analyses, we adjusted for confounding and/or interaction between variables such as place of residence, severity, age, sex and whether respondents had been told about their level of asthma severity. Separate logistic regression analyses were run for the subgroup who had been told about their asthma severity to determine the independent contribution of severity level on the dependent variable of interest. Similarly, the responses of those who had a regular doctor were analysed separately to determine differences between sources of care. Finally, we considered that patients who reported having an AMP, being told about trigger factors and having an action plan were receiving "good management", and examined factors related to good management. We excluded the optimisation of pharmacological intervention from our definition. SPSS was used for all statistical analyses.9 Results 3001 of a possible 4065 people responded (response rate, 73.8%). Of these 3001 respondents, 558 (18.6%) indicated that they had ever had asthma, and 521 (93.4%) of these had had their asthma confirmed by a doctor. Of these 521, 307 (58.9%) indicated that they still had asthma, representing 10.2% of the total sample and comprising 120 men and 186 women (one respondent did not indicate sex), with a mean age of 39.4 years (SD, 18.3; range, 15-84 years). Box 2 (below) shows that current medically confirmed asthma was more common among women than men, and declined in prevalence with increasing age. It was also more common among rural than metropolitan respondents, and among Australian-born respondents than those born elsewhere; this difference remained significant after adjusting for age and sex in logistic regression analyses. There was no relationship between current confirmed asthma prevalence and post-secondary education or annual gross household income, even after adjusting for age and sex. Asthma management plans Box 3 (above) summarises our findings on asthma management among respondents with currently confirmed asthma. Further analyses showed that having an AMP was not related to age, sex, area of residence, or whether people had been told their level of severity. However, those who had a regular doctor were more likely to have an AMP (P = 0.03). Those who saw both a general practitioner (GP) and specialist were significantly more likely to have an AMP (81.7%) than those who saw a GP only (47.6%), specialist only (41.1%) or no one (22.6%; P = 0.02). Among respondents who had been told their level of severity, those with moderate or severe asthma were more likely to report having an AMP than those with mild asthma (P < 0.01), after adjustment for other covariates. Women were more likely than men to have been told about their asthma severity (86.3% v. 75.0%; chi-squared = 6.1; P = 0.01), and those who had been told tended to be younger than those who had not (t = 3.1; P = 0.07). Among those who had been told, reported level of severity increased with advancing age (t = 5.3; P < 0.01). Severity did not vary by sex, but was greater among rural respondents than metropolitan respondents (chi-squared = 8.6; P = 0.01). The reported prevalence of discussing trigger factors and having an action plan increased with increasing asthma severity (see Box 4). It is notable that only 14 (25.4%) of those who had not been told their asthma severity had an action plan. Men were more likely than women to have no regular doctor for asthma care (22.7% v. 12.4%; chi-squared = 5.7; P = 0.02); there was no difference by age. Overall, those who had discussed trigger factors did not differ by sex, area of residence or whether they had been told their level of severity. After adjustment for covariates, discussion of trigger factors was more likely among younger respondents (P < 0.05) and those who had a regular doctor (P < 0.01). Among those who had been told their level of severity, logistic regression analysis showed that those with severe asthma were more likely than those with moderate or mild asthma to have discussed trigger factors (P < 0.05). Of those with a regular doctor, source of care was unrelated to discussion of trigger factors. Respondents who reported having an action plan did not vary by age, area of residence or whether they had been told their level of severity. Action plans were more common among women (P < 0.05) and those who had a regular doctor (P < 0.05), and among those who had been told they had moderate or severe asthma than mild asthma (P < 0.01). For respondents with a regular doctor, source of care was unrelated to having an action plan. Good management Seventy-one respondents with currently confirmed asthma (23.3%) were receiving "good management". Good management was unrelated to age, sex, area of residence or whether respondents had been told their level of severity, but was more likely among those who had a regular doctor (P < 0.05). Among those who had been told their level of severity, those with moderate or severe asthma were more likely to be receiving good management than those with mild asthma (P < 0.001). Of those with a regular doctor, source of care was unrelated to receiving good management. Discussion Given that the AMP model was initiated in 1989, our finding of a prevalence of reported use of AMPs of 46%, with much higher prevalences for certain elements of AMPs, is reassuring. It is noteworthy that we found AMPs to be more common among people who had been told their asthma was severe and those with a regular doctor. Our finding that 39.5% of respondents had an action plan is a somewhat higher result than those of previous reports,10,11 possibly indicating an increase in their use. As we found that reported use of action plans did not vary by age, the different age profile of our study compared with that of Comino and colleagues10 (mean, 39.4 years; range, 15-84 years v. mean, 37.4 years; range, 18-50 years) is unlikely to have contributed to the apparent doubling in prevalence of action plan use that we observed. However, unlike the other study, we included verbal as well as written action plans, and this may have been responsible for the difference. Our study was based on self-report, and responses were not validated. However, our method of assessing self-reported asthma has been used in a previous validated epidemiological study.12 There may have been under-reporting in the responses to the questions, even though we used trained interviewers. However, despite these methodological problems, the key measure of the acceptability of the AMP and its elements is their reported use in the community. We are aware that the term "asthma management plan" may have introduced some problems in that our def inition of the AMP encompassed as - sessment of severity, trigger-factor discussions and the development of an action plan. Hence, it is possible that people who reported using an AMP may have been including the other three elements. However, as a main objective was to compare the uptake of the "complete" AMP with that of its individual elements, we believed it was important to include all these questions in the survey. We found that certain elements of the AMP are being adopted more frequently than others, which may indicate that patients find these steps more appropriate to their specific needs, or that doctors (GPs and specialists) find certain elements of the AMP easier to use or implement. GPs are the principal providers of asthma care in the community and it is probable that they have a considerable effect on the uptake of the AMP model. Developing a "complete" management plan may be too onerous a task for a busy GP. Conversely, GPs may find the AMP model does not satisfy their individual needs and, as a consequence, choose specific elements from the AMP model for their everyday practice.13 Over 80% of our respondents, most of whom had mild to moderate asthma, saw a GP for all or part of their asthma care. As there is evidence that GPs' management of asthma is very variable,14,15 it is important to concentrate on providing useful and focused education for GPs. Urgent clarification is needed of the key steps within the AMP model that will reduce mortality and morbidity in people with asthma and improve their quality of life. Interestingly, we found that only 41% of the respondents attending a specialist only had an AMP. As specialists spend more time with their patients and deal with people with more severe asthma, this finding requires further investigation. It has been suggested that some steps of the AMP may be inappropriate for general application. For example, peak flow meters are supposed to be used in both steps 2 and 3, but Ruffin and Pierce have proposed that aiming to have all asthmatic patients record peak expiratory flow over the long term may not be worthwhile,16 and that focusing on, for example, those at risk of dying, those in uncertain clinical situations (e.g., those with exercise-induced asthma) and those who are "particularly interested" in monitoring their asthma would be more cost effective. They emphasised the need for a more rigorous scientific approach to examining "the basis for and benefit of" peak expiratory flow.16 In a more recent paper Gibson et al. state that "action plans have received little controlled evaluation".17 Among 35 adult asthma patients, they found that the current increase in medication recommended in the AMP action plan had a high failure rate. On the other hand, there is evidence that regular review (step 6 of the AMP) can reduce days lost from work or school18 and home visits by GPs,19 and that patient education (step 5) improves symptom control20 and reduces visits to accident and emergency departments.21 The guidelines in the AMP model have come to be accepted as the "gold standard" of asthma management in most countries.22,23 However, what is needed now is a careful re-evaluation 24 to ensure that "unsound practice" does not become formalised.25 A new perspective may be required that embraces the variability in management and clinical encounters.26 Not surprisingly, we found that having a regular doctor is the main determinant of "good management" -- regular review is likely to allow time for the introduction of more steps in the AMP. The challenge in continuing to evaluate an introduced management program in clinical practice is substantial. If over half the population of asthmatic patients have already been exposed to the asthma management plan in part or in full, how can this be done? Small intensive clinical trials with objective outcomes such as peak flow measurements, or case-control studies comparing asthmatic patients with and without good quality of life, could be performed, but would have the disadvantage of not being generalisable to the population at large. Large population studies with specific outcome measures would rely on the outcome measures being validated against accepted criteria of asthma control. A national project that used specific health outcomes, such as prescriptions of reliever or preventer agents, and monitored hospital admissions for asthma could be used to evaluate interventions introduced on a State or regional basis. To coordinate and cooperate in such a project would present a challenge to organisations involved in asthma management. There is current interest in coordinated care programs for chronic diseases and the time is right to introduce an evaluation strategy for different interventions in asthma. References Campbell D, Ruffin R, McEvoy RD, Crockett AJ. South Australian asthma symptoms prevalence survey [abstract]. Paper presented at the Annual Scientific Meeting of The Thoracic Society of Australia and New Zealand. 1991 7-11 Apr; Lorne, Vic. Aust N Z J Med 1991; 21: 658. Jenkins WA, Hurley SF, et al. Trends in Australian mortality of asthma 1979-1985. Med J Aust 1988; 149: 620-624. National Asthma Campaign. Report on the cost of asthma in Australia 1994. Melbourne: National Asthma Campaign, 1994. National Health Survey 1989/90. Canberra: Australian Bureau of Statistics, 1991. Robertson CF, Heycock E, Bishop J, et al. Prevalence of asthma in Melbourne school children: change over 26 years. BMJ 1991; 302: 1116-1118. Bauman A, Young L, Peat J, et al. Asthma under-recognition and under-treatment in an Australian Community. Aust N Z J Med 1992; 22: 36-40. Woolcock A, Rubinfield A, Seale P, et al. Asthma management plan, 1989. Med J Aust 1989; 151: 650-652. Wilson D, Wakefield M, Taylor A. The South Australian Health Omnibus Survey. Health Promotion J Aust 1992; 2: 47-49. Nornsis MJ. SPSS for Windows. Advanced statistics release 6.0. Chicago: SPSS Inc, 1993. Comino E, Mitchell C, Bauman A, et al. Asthma management in eastern Australia, 1990 and 1993. Med J Aust 1996; 164: 403-406. Tse M, Bridges-Webb C, Bauman A. The impact of a mass communication campaign on the reported management of asthma by general practitioners. Fam Pract 1993; 10: 263-267. Abramson M, Kutin J, Bowes G. The prevalence of asthma in Victorian adults. Aust N Z J Med 1992; 22: 358-363. Armstrong D, Fry J, Armstrong P. General practitioners' views of clinical guidelines for the management of asthma. Int J Qual Health Care 1994; 6: 199-202. Bauman A, McKenzie D, Young L, Yoon R. Asthma education: the perception of family physicians. J Asthma 1990; 27: 385-392. Coates J, Steven I, Beilby J, et al. Knowledge and reported asthma management among South Australian General Practitioners. Br J Gen Pract 1994; 44: 123-126. Ruffin R, Pierce R. Peak flow monitoring -- which asthmatics, when and how? Aust N Z J Med 1994; 24: 519-520. Gibson P, Wlodarczyk J, Hensley M, et al. Using quality-control analysis of peak expiratory flow recordings to guide therapy for asthma. Ann Intern Med 1995; 123: 488-492. Charlton I, Charlton G, Broomfield J, Rullee M. Audit of the effect of a nurse run asthma clinic on workload and patient morbidity in general practice. Br J Gen Pract 1991; 41: 227-231. Charlton I, Charlton G, Broomfield J, Campbell M. An evaluation of a nurse-run asthma-clinic in general practice using an attitudes and morbidity questionnaire. Fam Pract 1992; 9: 154-160. Wilson S, German D, Sulochina L, et al. A controlled trial of two forms of self-management education for adults with asthma. Am J Med 1993; 94: 564-576. Bolton M, Tilley B, Kuder J, et al. The cost and effectiveness of an education program for adults who have asthma. J Gen Intern Med 1991; 6: 401-407. Guidelines for the diagnosis and management of asthma. Bethesda, Md.: United States Department of Health and Human Services, August 1991. (Publication No. 91-3042.) International consensus report on diagnosis and managment of Asthma. Bethesda, Md.: United States Department of Health and Human Services. June 1992. (Publication No. 92-3091.) Thompson P, Lavender M, Madhok R. How to ensure that guidelines are effective. BMJ 1995; 311: 237-242. Delamothe T. Wanted: guidelines that doctors will follow. Implementation is the problem. BMJ 1993; 307: 218. Parmley W. Clinical practice guidelines. Does the cookbook have enough recipes? JAMA 1994; 272: 1374-1375. (Received 17 Apr, accepted 11 Nov, 1996) Authors' details Research & Health Promotion Unit, Royal Australian College of General Practitioners, Adelaide, SA. Justin J Beilby, MPH, FRACGP, Director. Behavioural Epidemiology Unit, South Australian Health Commission, Adelaide, SA. Melanie A Wakefield, MA, PhD, Senior Behavioural Scientist. Department of Thoracic Medicine, The Queen Elizabeth Hospital, Adelaide, SA. Richard E Ruffin, MD, FRACP, Director. No reprints will be available. Correspondence: Dr J J Beilby, Department of General Practice, University of Adelaide, Adelaide, SA 5005. - - To top of article - ©MJA 1997 <URL: http://www.mja.com.au/> © 1997 Medical Journal of Australia.

Justin J Beilby · Melanie A Wakefield · Richard E Ruffin