Targeting asthma care in general practice using a morbidity index

Clinical implications of the Royal College of Physicians three questions in routine asthma care: a real-life validation study

BACKGROUND

Annual recording of the Royal College of Physicians three questions (RCP3Q) morbidity score is rewarded within the UK 'pay-for-performance' Quality and Outcomes Framework.

AIMS

To investigate the performance of the RCP3Qs for assessing control in real-life practice compared with the validated Asthma Control Questionnaire (ACQ) administered by self-completed questionnaire.

METHODS

We compared the RCP3Q score extracted from a patient's computerised medical record with the ACQ self-completed after the consultation. The anonymous data were paired by practice, age, sex, and dates of completion. We calculated the sensitivity and specificity of the RCP3Q scale compared with the threshold for good/poor asthma control (ACQ ≥1).

RESULTS

Of 291 ACQ questionnaires returned from 12 participating practices, 129 could be paired with complete RCP3Q data. Twenty-five of 27 patients who scored zero on the RCP3Q were well controlled (ACQ <1). An RCP3Q score ≥1 predicted inadequate control (ACQ ≥1) with a sensitivity of 0.96 and specificity of 0.34. Comparable values for RCP3Q≥2 were sensitivity 0.50 and specificity 0.94. The intraclass correlation coefficient of 0.13 indicated substantial variability between practices. Exacerbations and use of reliever inhalers were moderately correlated with ACQ (Spearman's rho 0.3 and 0.35) and may reflect different aspects of control.

CONCLUSIONS

In routine practice, an RCP3Q score of zero indicates good asthma control and a score of 2 or 3 indicates poor control. An RCP3Q score of 1 has good sensitivity but poor specificity for suboptimal control and should provoke further enquiry and consideration of other aspects of control such as exacerbations and use of reliever inhalers.

Asthma management in primary care: does increasing patient medication improve symptoms?

AIMS

We set out to observe the General Practitioner (GP) management of a cohort of asthma patients over a 2-year period by comparing asthma health status, spirometry, British Thoracic Society treatment step, inhaled medication uptake and psychological status. Changes in these parameters were assessed over the 2-year period.

METHODS

One hundred fourteen subjects were recruited from four GP practices, two in the inner city and two in suburbia. Subjects were assessed at baseline and at 2 years using the Juniper asthma quality of life questionnaire, the locally devised Q score (a simple patient-focused morbidity index) and the hospital anxiety and depression (HAD) scale. Spirometry (forced expiratory volume in 1 s, forced vital capacity and peak expiratory flow) was recorded using a Micro Medical portable spirometer (Micro Medical Ltd, Chatham, UK); the best values from three acceptable attempts were recorded.

RESULTS

Data for 90 subjects reviewed at baseline and at 24 months were compared to the original cohort of 114 subjects. Patients who had the treatment increased showed no apparent benefit over 2 years. They had similar physiology and symptom scores at baseline but had higher HAD scores (particularly depression element P < 0.05) initially. The picture remained constant over the 2 years.

CONCLUSION

Asthma guidelines define the aim of treatment to minimise or abolish symptoms. We suggest that if the symptoms are not measured, they remain unrecognised. This is something that could and should be incorporated routinely into clinical practice because this is morbidity that is largely treatable.

Using an asthma control questionnaire and administrative data to predict health-care utilization

OBJECTIVE

To examine the merits of the Asthma Therapy Assessment Questionnaire (ATAQ) control index together with prior asthma health-care utilization from administrative data in predicting future acute asthma health-care utilization.

DESIGN

Prospective cohort study.

POPULATION

A total of 4,788 adult asthma patients aged 17 to 93 years who completed a baseline evaluation and had at least 6 months of follow-up data.

STATISTICAL METHODS

Classification and regression tree methodology to predict future risk of acute health-care utilization events.

RESULTS

These results show that the ATAQ control index and administrative data are jointly useful for predicting future health-care utilization. The utility of the ATAQ control index in the presence of information about prior health-care utilization is to further stratify risk among the subset of younger individuals who did not have any prior acute health-care utilization. While administrative health-care utilization data served as the strongest predictor of future health-care utilization, the ATAQ control index helped to identify 1% of individuals without recent acute care that had approximately a sixfold elevated risk (95% confidence interval, 4.2 to 8.4) of future acute health-care utilization. This is an important result since only a small fraction of individuals with acute events in a given year will have had acute events in the previous year.

CONCLUSION

These findings should assist the practicing clinician and organizations interested in population-based asthma disease management.

Targeted routine asthma care in general practice using telephone triage

BACKGROUND

There is a high non-attendance rate for traditional clinic-based routine asthma care in general practice. Alternative methods of providing routine asthma care need to be examined.

AIM

To examine the cost and effectiveness of targeted routine asthma care in general practice using telephone triage, compared to usual clinic care.

DESIGN OF STUDY

An open randomised controlled trial.

SETTING

A single semi-rural practice in the southwest of England.

METHOD

Adult patients with asthma were randomised to receive either their routine asthma care in the surgery or care by telephone triage. Asthma control parameters, health status and NHS resource utilisation were measured over the 12-month study period.

RESULTS

One hundred and ninety-four patients were randomised and 35% per cent more patients (n = 84 versus n = 62) received more than one consultation in the telephone group. Asthma control as measured by the asthma control questionnaire (ACQ) was similar in the clinic and telephone groups: mean change in ACQ = -0.11 (95% CI = -0.32 to 0.11) versus -0.18 (95% CI = -0.38 to 0.02). Mean NHS costs were 210 pounds sterling per patient per year in the telephone group compared to 334 pounds sterling in the clinic group (P-value of bootstrapped difference = 0.071).

CONCLUSION

Targeted routine asthma care by telephone triage of adult asthmatics can lead to more asthma patients being reviewed, at less cost per patient and without loss of asthma control compared to usual routine care in the surgery.

Measuring outcomes with tools of proven feasibility and utility: the example of a patient-focused asthma measure

Best quality care is clearly desirable and instituting quality assurance should be logical and simple. However, the practicality of setting standards for a product or service, and designing systems to measure against the standards, is more difficult. In the health service it is only likely to be feasible if data can be generated from efficient and reliable information systems. The ideal measure of quality is an outcome measure that evaluates whether or not the quality of care has achieved the desired standard of outcome. Direct measures of outcome are not easy to construct and the information systems required to provide data are not widely available. The National Centre for Health Outcomes Development (NCHOD) has produced a series of indicators in 10 areas of health care, where an indicator is a pointer to, rather than a direct measure of, a desired outcome. Feasibility studies measuring their sensitivity and reliability have drawn attention to their possible utility within different health care settings. This paper reports on an investigation into a patient-focused outcome indicator for asthma. There is broad agreement about the need to measure the outcome of disease. However, when outcome indicators are defined there are major obstacles to their successful uptake. A key challenge for outcomes measurement is to ensure that the cost of collecting the data and ensuring completeness, accuracy and standardization are justified by the benefits derived. Health outcome indicators should not be treated as a panacea, but as a part of the clinical and health care tool kit.

Symptomatic asthma: attendance and prescribing in general practice

Under-prescribing and low attendance continue to be cited as reasons for ongoing asthma symptoms in primary care despite marked increases in prescribing and structured care for asthma over the past 10 years. The objective of this study was to determine the relationship between continuing asthma morbidity and the attendance of and prescribing for symptomatic asthmatic patients in primary care. A random sample of 402 subjects from 801 who reported at least one of six symptoms in the previous month on most or every day were identified from responses to a validated morbidity questionnaire. An analysis of their care over a 2-year period (1 year before and 1 year after the questionnaire) was carried out from their general practice case-notes. Data on 308 patients was available for analysis. Ninety-four per cent of these symptomatic asthma patients attended over the 2-year period, with 77% attending for an asthma related consultation. Most patients were managed exclusively in primary care. Inhaled steroids were prescribed for 78% of patients and high dose inhaled steroids (> or = 800 mcg of beclomethasone or equivalent per day) were prescribed for 38%. Patients with most symptoms were more likely to be prescribed inhaled steroids. Rescue courses of oral steroids were prescribed for 29% of patients. Changes in asthma medications were recorded for 31% during the study period. Metered dose inhalers (MDI) were prescribed for 86% with more than half prescribed MDIs combined with some other delivery device. Elements of structured care were more frequently recorded in patients who reported most symptoms. In conclusion the asthma management of the majority of patients in this study was active with high levels of steroid prescribing. There appeared to be room to increase prescribing and to improve the structure of care. While patients who were 'symptomatic on steroids' should have had their medications, delivery devices and structured care reviewed regularly many were already on maximal treatment and were therefore likely to remain symptomatic. It is unclear how practitioners could improve morbidity in many of these patients as under-treatment and low attendance seem unlikely to be the principal causes of continuing symptoms.

Association of asthma control with health care utilization: a prospective evaluation

Population-based disease management should be enhanced by good risk assessment models and instruments. We prospectively evaluated the ability of a simple measure of short-term asthma control (scored 0 to 4) to predict asthma 12-mo health care utilization (HCU). A total of 5,172 adult asthma patients completed a brief questionnaire in fall 1997 to assess current level of asthma control. We then evaluated HCU for calendar year 1998. Ninety-three percent had health plan eligibility in 1998 and were included in this analysis. Both acute and routine asthma utilization increased with increasing numbers of asthma control problems. Rates of acute care episodes were 3.5 (95% confidence interval [CI] = 2.9, 4.3) times more likely for those with 3 to 4 control problems versus those with no control problems. Lesser, but statistically significant, increases were seen for those with two (relative risk [RR] = 1.7, 95% CI = 1.4, 2.2) or one (RR = 1.4, 95% CI = 1.1, 1.8) control problems. These patterns were similar for men and women, and diminished with increasing age. The asthma control index contributed significantly to prospective prediction models even after adjusting for administrative data such as medication use and prior HCU. These data reinforce the usefulness of measures of short-term asthma control both for the individual clinician and for those interested in population-based asthma management.

Using a revised asthma morbidity index to identify varying patterns of morbidity in U.K. general practices

Both in terms of morbidity and mortality, the therapeutic and organizational management of asthma pose a considerable and continuing challenge to healthcare delivery. One element in attempts to meet this challenge is the recognition of appropriate outcome measures to assess progress in tackling the burden of this disease. This study therefore aimed to assess pragmatically the effectiveness of a revised asthma morbidity index in identifying varying patterns of morbidity in U.K. general practices. A postal survey was conducted of 2,762 patients believed to have or have had asthma from the lists of 12 general practices within the Eden Valley in Cumbria, using a questionnaire which combined the revised Jones morbidity index with questions on age, medication and perception of current asthma. Prescribing data were also recovered for 11 of the practices for the quarter within which the postal survey was conducted. Responses were obtained from 2,123 subjects (77%), of whom 1,474 (70%) believed themselves to be currently asthmatic. In this group, 18% reported low morbidity 34% medium morbidity and 48% high morbidity Age and inhaled steroid use were both positively and significantly associated with high morbidity. Those taking inhaled steroids were 1.4 times more likely to report high morbidity than those nottaking steroids. The prescribed corticosteroid/bronchodilator ratio for cost was both negatively and significantly associated with high morbidity. The revised morbidity index is a simple tool of use in the surveillance of asthma in primary care. It identifies spectra of morbidity which vary between practices, which may be of use in assessing the quality of asthma care provided in the community.

Randomised controlled trial of general practice based asthma clinics

OBJECTIVE

To compare the effects on asthma morbidity of asthma clinics based in general practice with standard general practice care.

DESIGN AND SETTING

A randomised controlled trial in eight general practices. Patients, general practitioners and outcomes assessors were not blinded to treatment allocation.

PARTICIPANTS

195 patients with asthma aged 5-64 years; 191 completed the trial.

INTERVENTION

Three asthma clinic sessions over six months involving nurse counselling, education about asthma management, spirometry and consultation with the general practitioner.

MAIN OUTCOME MEASURES

Patients reporting days lost from work or school, number of days lost, the presence of morning or nocturnal asthma symptoms, use of an action plan, medication use, current smoking, hospitalisation, and emergency visits.

RESULTS

Asthma clinics were associated with a greater reduction in nocturnal symptoms, an increase in the ownership of peak flow meters and an increase in the number of patients commencing or resuming smoking. Both control and intervention groups showed similar improvement in days lost from work or school, the presence of symptoms, use of an action plan and taking reliever medication.

CONCLUSION

Our study does not show that asthma clinics are more effective than standard general practice care in reducing asthma morbidity. It is uncertain how much of the improvement in outcomes was due to the asthma clinic, the influence of the study itself upon patients and practitioners, or other factors, such as the tendency for a patient's asthma management to improve over time.

Associations between an asthma morbidity index and ideas of fright and bother in a community population

There is a need for simple asthma outcome measures for primary care which are not only valid in terms of their relationship with lung function but also in terms of pragmatic psychological constructs. This study assesses the usefulness of adding items on the degree of 'bother' and 'fright' caused by the condition to a previously validated simple asthma morbidity index. A postal questionnaire survey comprising a simple asthma morbidity index and questions on 'fright' and 'bother' was conducted in one general practice in the north-east of England. Responses were obtained from 570 individuals. Of these, 184 (32%) reported low, 133 (23%) medium and 253 (44%) high morbidity. Twenty-nine per cent of respondents had felt frightened by their asthma in the previous 4 weeks. Both the 'fright' and 'bother' items were significantly associated with the morbidity index. The addition of 'bother' and/or 'fright' questions may improve both the content, construct and predictive validity of the morbidity index, but this needs to be established prospectively.

Predictive value of a simple asthma morbidity index in a general practice population

BACKGROUND

There is a need in primary care for simple asthma outcome measures that are valid in terms of their relationship with lung function and capable of predicting those patients for whom additional management is indicated.

AIM

To assess the predictive validity of a revised asthma morbidity index in United Kingdom (UK) general practice.

METHOD

Morbidity index and peak flow rate data were gathered from nine general practices over a three-month period. Two postal questionnaire surveys, one year apart, were conducted in one Tyneside general practice. Morbidity index data from 570 asthmatic patients were gathered in the first survey and used to predict morbidity over the next year.

RESULTS

For 120 responders with low morbidity, mean peak flow as a percentage of the predicted value was 91% (SD = 21%); for 91 responders with medium morbidity, the percentage was 77% (SD = 21%); and for 90 responders with high morbidity, it was 63% (SD = 29%). Fifty-seven per cent of the morbidity index categories remained unchanged after 12 months. The relative risks of high morbidity for having any acute asthma attacks, more than four attacks, and needing oral steroids during a one year period were 2.88 (CI = 1.87 to 4.43), 2.52 (CI = 1.84 to 3.44) and 2.38 (CI = 1.70 to 3.33) respectively.

CONCLUSION

The revised morbidity index is a simple and valid tool for the opportunistic surveillance of asthma in primary care.

How to avoid underdiagnosed asthma/chronic obstructive pulmonary disease?

The prevalence of asthma and chronic obstructive pulmonary disease (COPD) is rising in most countries, including The Netherlands. It has been suggested that a majority of these cases of (self-reported) symptoms related to asthma/COPD are not diagnosed in general practice. We compared a population screening for underdiagnosed asthma/COPD with a high-risk approach by a questionnaire form with specified questions about asthma/COPD-related symptoms. A case-controlled study including a record review was performed of cases and controls. The results of a population screening were used to classify patients as (a) asthma/COPD, (b) at risk for asthma/COPD, or (c) no asthma/COPD. Eleven hundred fifty-five patients were screened. One hundred fifty-five patients reported previous asthma/COPD-related care (cases). The difference between number of cases and controls in asthma/COPD diagnosis was chosen as main outcome measure. The population screening revealed 85 subjects with a diagnosis of asthma/COPD and 154 subjects with an increased risk. Nineteen diagnoses could be made in cases, and eight diagnoses in controls. The chart review showed that only seven cases and two controls were known to the general practitioner. From this study it can be concluded that in order to reduce the number of un- and underdiagnosed patients, all listed patients in general practice should be screened. However, if screening of all patients is not feasible, active case finding by asking a few questions about shortness of breath or wheezing to all patients in the group of listed individuals is recommended.

Asthma management. The case for aiming at control rather than merely relief

Asthma is a common and eminently treatable disease. Most asthma patients are initially seen by primary care physicians, who are likely to have a significant part in management. Research has provided physicians with a better understanding of the pathophysiologic basis of asthma and new antiasthma agents. Guidelines have been published to disseminate asthma management information. On the basis of these advances, there appears to be no reason that informed physicians cannot play an important role in reversing recent trends of rising asthma morbidity and mortality.

Reducing asthma morbidity in the community: the effect of a targeted nurse-run asthma clinic in an English general practice

Although most primary health care teams in the U.K. now offer proactive care for patients with asthma, there is relatively little published evidence showing the effectiveness of such innovations. This may be due in part to lack of targeting of extra care towards those most in need. Therefore, to demonstrate the benefits of targeted nurse-run asthma clinic care in a seven-partner general practice in a mixed urban and rural area of North Lincolnshire in the east of England, a cohort of 173 patients, with asthma selected predominantly by having high morbidity in a postal survey, completed 12 months follow-up in a nurse-run asthma clinic. A longitudinal comparison was conducted in terms of: changes in morbidity index category, inhaler technique score, knowledge score, use of inhaled steroids, use of salmeterol, method of administration of beta(2)-agonist medication and frequent use of peak flow meters. The number with high morbidity fell from 123 (71.1%) at the initial consultation to 14 (8.1%) at the 12-month review. Those with full marks on inhaler technique rose from 28 (16.2%) to 142 (82.1%), and with full marks on asthma knowledge rose from 7 (4.0%) to 98 (56.6%). The numbers of patients using inhaled steroids and salmeterol rose from 127 (73.4%) to 171 (98.9%) and from 5 (2.9%) to 35 (20.2%), respectively. The preferred inhaler device for beta(2)-bronchodilator medication changed from metered dose aerosol to dry powder. Regular use of peak flow meters in 157 subjects aged 5 years and over rose from 43 (27.4%) to 116 (73.9%). These data clearly demonstrate the benefits of targeted proactive nurse-run asthma care in terms of reduced morbidity for patients. The authors recommend the morbidity index targeting concept to other primary health care teams.

Definitions of severity and outcome measures

Outcome measurement is still a difficult area in general, and in asthma in particular, with further research needed. (Attributable) outcomes of health care are the only sort of outcome measure which are of direct use as a contracting tool. However, less well-researched and understood outcomes are useful as quality improvement tools, and within more open-ended discussions involving purchasers and providers. In terms of hospital care of acute asthma, there is no well-defined outcome measure which reflects the quality of hospital care; re-admission rates show promise as an outcome measure which relate to the quality of discharge planning and merit further study. In terms of ambulatory care, there is an urgent need to develop and evaluate a symptom-based outcome measure which would be usable in routine practice and could be recommended for widespread use. As a physiological outcome measure, percentage of best function is one which corrects for the degree of irreversible air flow obstruction and is independent of treatment step. It is valuable for individual patients by providing a realistic gold standard and if best function is assessed in a standard manner, it also allows results of groups of patients to be compared in a meaningful manner. Severity scores, which might allow categorization of patients on the basis of characteristics other than current symptoms or therapy, are currently being evaluated.

Do self-management plans reduce morbidity in patients with asthma?

BACKGROUND

Self-management plans may help patients with asthma intervene when symptoms deteriorate, thus preventing asthma attacks.

AIM

A study set out to test whether a self-management plan tailored to the circumstances of the individual reduces morbidity from asthma.

METHOD

General practitioners who had participated in a national audit of asthma attacks were randomized into intervention and control groups. Six months after the intervention group had issued self-management plans to patients with asthma, both groups of practitioners completed morbidity questionnaires on patients. Morbidity outcomes were compared for the 6-month periods before and after the issue of the plans.

RESULTS

In the 6 months before the study, the 376 patients enrolled by the intervention group experienced higher levels of morbidity than the 530 patients for whom details were recorded by the control group. In the 6 months after the issue of the plans, control group patients showed little change in levels of morbidity, but intervention group patients showed significant reductions in hospital admissions, consultations for asthma symptoms, asthma review consultations, courses of oral steroids and use of emergency nebulized bronchodilators.

CONCLUSION

General practitioners appeared to operate enthusiast bias' and issued more self-management plans to patients with uncontrolled asthma. The reduction in morbidity in this group is probably a result of the use of the plans, but the verdict on whether plans reduce morbidity must be deemed 'not proven'.

Chronic disease or physical disability? The role of the general practitioner

There has been a lack of discussion and consensus as to what the role of the general practitioner should be in the care of patients with chronic diseases. Should general practitioners concentrate on the disease or should their remit include the resultant disability and handicap? General practitioners have tended to concentrate on the disease, but this may be inappropriate. For many disabled people, their general practitioner is their only source of health care and is the gatekeeper to other services. Greater knowledge among doctors of the functional and social aspects of disease would therefore improve the quality of care for patients, and should be assessed through clinical audit. Ways are described in which general practitioners, working together with their patients with chronic diseases and with other health professionals, can improve aspects of the care of these patients.

Monitoring outcomes in routine practice: defining appropriate measurement criteria

With the development of an internal market for health care, 'purchasing for outcomes' has become an important if somewhat rhetorical catchphrase. While there is emerging understanding about how it can be pursued, doubts are being expressed over an outcomes rather that a process emphasis. This debate has been confused by a failure to differentiate the role and importance of monitoring outcomes at an individual patient care level from those at an aggregate population/purchaser level. The clinical need to collect outcomes data on individual patient care within routine care settings places additional requirements on measurement development and selection. Traditional measurement criteria, stressing reliability, validity and responsiveness to change, must be supplemented by criteria of feasibility of use, clinical utility and acceptability. One option is to select domains or items of interest from longer instruments initially designed for research, carefully selected in relation to the purposes of measurement. Further measurement criteria must be addressed which stress the relevance of the proposed instrument to the condition and to the participants in the clinical interaction: in particular, patient-centredness and sensitivity to the setting. Monitoring the outcomes of individual patient care within routine clinical practice poses considerable challenges to researchers who are developing instruments and to clinicians who collect and use the data. A shift in emphasis is required towards more context-specific tests, addressing relevance to lay perceptions, to clinical use and to the condition and setting under review. The content validity, the responsiveness to patient-relevant and clinically relevant change and, of course, reliability must have greater primacy. In this way, outcome data which measure the quality of clinical practice and which provide appropriate criteria for research into effectiveness can be generated.

Peak flow based asthma self-management: a randomised controlled study in general practice. British Thoracic Society Research Committee

BACKGROUND

Peak flow based asthma self-management plans have been strongly advocated in consensus statements, but convincing evidence for the effectiveness of this approach has been largely lacking.

METHODS

A randomised controlled trial was conducted in 25 general practices comparing an asthma self-management programme based on home peak flow monitoring and surgery review by a general practitioner or practice nurse with a programme of planned visits for surgery review only over a six month period.

RESULTS

Seventy two subjects (33 in the self-management group and 39 in the planned visit group) completed the study protocol, but diary card data for at least three months were available on a total of 84 (39 in the self-management group and 45 in the planned visit group). Teaching self-management took longer than the planned visit review. In the self-management group home peak flow monitoring was felt to be useful by doctors and patients in 28 (85%) and 27 (82%) cases, respectively. There were no between group differences during the study period in terms of lung function, symptoms, quality of life, and prescribing costs. Only within the self-management group were improvements noted in disturbance of daily activities and quality of life. Possible explanations for these negative results include small numbers of subjects, the mild nature of their asthma, and inappropriate self-management strategies for such patients.

CONCLUSIONS

Rigid adherence to long term daily peak flow measurement in the management of mild asthma in general practice does not appear to produce large changes in outcomes. Self-management and the use of prescribed peak flow meters need to be tailored to individual circumstances.

Comparison of asthmatic patients admitted to hospital from health districts experiencing high and low asthma mortality rates

Geographical variation in asthma mortality rates within the United Kingdom could be a reflection of variability in effectiveness of medical care services, or epidemiological variation. In order to ascertain whether differing hospital admission processes could contribute to this variation, asthmatic patients admitted from two districts, experiencing above and below average mortality rates were compared. The present study was part of a cohort study of 1,200 consecutive acute adult admissions in 1987/88. In the main study, social data and information on referral were collected by interview for all patients. The admitting doctors' perception of the patient's severity was assessed on the basis of the severity of symptoms, and likelihood of morbidity or mortality if the patient was not admitted. Further information on asthmatic patients (treatment and physiological measurements) was retrieved from the notes. Sixty-six asthmatic patients resident in Wandsworth (a district with high asthma mortality rates) were admitted to St George's Hospital or St James' Hospital (WW) and 31 patients resident in East Surrey (ES) (a district with low asthma mortality rates) were admitted to the East Surrey Hospital (ESH). Notes were obtained on 55 (83%) and 27 (87%) of patients in the two districts, respectively. WW received significantly more patients by self-referral: 68% of patients called an ambulance or came directly to casualty compared with 30% in ES (chi-squared = 13.7, d.f. = 2, P = < 0.001). There was a tendency for more admissions to ESH to be taking oral steroids (chi-squared = 3.2, d.f. = 1, P = 0.07). Patients admitted in WW tended to have more severe disease: 39 (85%) of patients admitted to WW had peak expiratory flow less than 200 1/minute on admission compared to 14 (58%) in ES (chi-squared = 6, d.f. = 1, P = 0.01). In WW the mean first recorded peak expiratory flow on admission was 154 1/minute compared to 172 1/minute in ES; their mean peak flow on discharge was 318 1/minute compared with 377 1/minute in ES. Twenty-one (38%) of admissions in WW were considered to be very urgent by the admitting hospital doctor compared to four (15%) in ESH (chi-squared = 4.67, d.f. = 1, P = 0.03). This opportunistic study found that, in an area experiencing high mortality rates, more patients with severe disease were admitted to hospital compared to a low mortality area. This does not appear to be due to differing hospital practices but rather to increased levels of morbidity in the community. As patients with more severe asthma are at a greater risk of dying, these finding reinforce the need to standardize asthma treatment in the community.