Estimating the prevalence of asthma: QOF v Health Survey for England

Lung and Asthma Information Agency 2006

Prevalence of asthma - background
Asthma is a complex, chronic, episodic disease which is not simply defined and whose nature may vary over the life course. This makes definition of its prevalence similarly complex and open to interpretation and opinion. Common measures for counting those with asthma include: those with a range of symptoms (wheeze, cough, night-waking etc), those diagnosed with asthma, those being treated by GPs for asthma. Counts based on these indicators may relate to different periods of time, ranging from “ever” to “current”. Each of these have different meanings, often different sources, and can to some extent be thought of as part of a spectrum of severity of the disease. Longitudinal studies, which can compare what is currently reported with what was reported earlier in life, have shown that self-reported symptoms suffer from recall problems. Other trend studies have shown that the proportion of wheezers who are doctor-diagnosed with asthma can be low and has varied over time and that some of those suffering symptoms, sometimes even severe symptoms, are not regularly treated by GPs.

Health Survey for England
This survey is a periodic national, representative survey of 40,000 people in England. It has a sound methodology for asking health questions. It asks the same questions on wheezing as the International Survey of Asthma and Allergies in Children (ISAAC) as well as other questions on treatment. It offers several markers of asthma including:

• Have you had any wheezing in the last year?
• Have you ever had a diagnosis of asthma?
• Have you been treated for asthma?
• Have you used an inhaler or other inhaling device in the last 12 months?
• Have you taken any prescribed medication in the last 12 months?

In 2001, a combined variable of recent wheeze, asthma diagnosis ever and treatment for asthma gave an overall prevalence of 8.1% in England and we know from other surveys that this figure is unlikely to have changed much since then. However from the same survey we know that 28% of those with wheeze did not report being treated by a GP or practice nurse in the last year and this was 22% even in those with severe wheeze.

The next Health Survey for England which contains questions on respiratory conditions is being carried out this year and might be available next year.

For more information see the Department of Health web site: http://www.dh.gov.uk/PublicationsAndStatistics/PublishedSurvey/HealthSurveyForEngland/fs/en

QOF
The Quality and Outcomes Framework is a voluntary scheme introduced as part of the new GMS contract for GPs in 2004. It aims to financially reward practices for the provision of quality care for a selection of conditions, of which asthma is one, by assigning points for a series of indicators. In 2004/05 8486 out of a possible 8542 practices made QOF returns, covering 99.5% of all registered patients. A central part of QOF is creating a patient register for each of these selected conditions. The prevalence of asthma in QOF is calculated from the ratio of the practice’s asthma register to the total practice list. In assessing these data we need to consider each element of the definition.

1. The number of patients on the asthma register.
This will depend on patient behaviour (for example in attending the GP), the diagnosis of a patient by the individual doctor involved and the recording system of the practice. Research indicates that often patients are not picked up or diagnosed which leads to lower estimates of prevalence. The register excludes patients with asthma who have been prescribed no asthma-related drugs in the last twelve months. Also, patients can only appear on one disease register. So GPs are instructed that “…patients diagnosed as COPD who were previously on the asthma register should be coded as inactive on the asthma register.” A diagnosis of COPD or asthma will vary by age and can also be very variable between doctors. Hence, the reasons and processes for placing patients on an asthma register may vary between practices and PCTs.

2. The number of patients on the practice list.
This can suffer from “list inflation”, which is where people are still registered and appear on a GP's list, but no longer live in that area and may also be registered elsewhere with another GP. There is evidence that this is often large (particularly in inner cities or other transient populations) and varies greatly between practices. This will also lead to prevalence estimates being lowered.

Initial results from QOF show that prevalence of GP treated asthma from the registers in 2004/05 was estimated at 5.8%, ranging from 3.2% to 7.4% across PCTs in England. As we know that the age-specific prevalence of asthma does not vary much across the country (from ISAAC and other work) most of this variation is likely to be due to other factors. The QOF data also provide the percentage of possible points attained for asthma, which can perhaps be used as a measure of the degree of progress in the implementation of QOF. These varied from 75% to 99% across PCTs. There is some correlation between prevalence and points attained, that is, those PCTs with higher QOF scores tend to report a higher prevalence of asthma. This has also been found in more detailed local analyses. This indicates that currently the QOF prevalence rates may reflect organisational factors, as well as underlying asthma prevalence.

In addition, when comparing rates across PCTs it must be remembered that there are wide variations in the populations of PCTs nationally, such as in age and socio-economic factors. This will make comparisons between PCTs hard to interpret.

For more information see the NHS web site: http://www.ic.nhs.uk/services/qof/data/

Summary of reasons for discrepancies between Health Survey for England prevalence and QOF.

1. They come from different populations: HSE is taken from a representative sample of the general population whereas QOF is taken from patients who see GPs.
2. They are defined differently: HSE measures all self-reported symptoms and treatment in the last year whereas QOF counts those patients seen by a GP and prescribed an asthma drug in the last year.
3. There is reason to think that current QOF prevalence rates are under-estimates due to too few patients on the asthma register and too many on the practice list.
4. QOF aims to clearly distinguish between COPD and asthma whereas HSE does not provide equivalent data on COPD diagnosis.
5. At present no age-specific data are available from QOF and there may be age biases in the reported prevalence

Conclusions and recommendations
QOF is a promising new source of data about the extent of asthma in Britain, specifically about GP consultations for asthma. However, it has not as yet been validated for epidemiological purposes. We therefore do not yet know enough detail about exactly how it is working in practice:

• How are the denominators (patient lists) counted;
• Who exactly will appear on the asthma register;
• How efficient is the recording and reporting system;
• How uniformly is it implemented across practices.
  Also, it is not clear whether or not any age or sex breakdowns will be available.

The Health Survey for England is a well respected survey with reasonably reliable results. It has a wealth of other data available for respondents.

In the end, which data to use will depend to some extent in precisely which aspect of asthma one is most interested. Is it the extent of doctor-diagnosed and treated asthma or is it the wider experience of asthma symptoms in the community?

LAIA recommends that currently the Health Survey for England, and other large well-designed populations surveys in the literature, give the best estimates of asthma prevalence. However QOF offers an alternative source of primary care data and may provide a reliable source of asthma prevalence estimates in the future so its use should be reviewed.