What do hospital admission rates say about primary care?

Their limitations suggest the need for more appropriate measures

General practice pp 94, 98

In its white paper on the English NHS the government has emphasised the need for an accountability framework against which to measure its objectives. A consultation document has suggested that health authorities should use performance indicators such as hospital admission rates as a measure of the quality of primary care.1 Yet two papers in this week's BMJ cast further doubt on hospital admission rates as a good measure of general practice performance.

In the United States high hospital admission rates for chronic diseases like asthma, hypertension, congestive cardiac failure, chronic obstructive airways disease, and diabetes have been associated with lack of access to a primary care physician.2 In the United Kingdom substantial variation exists in admission rates among both health authorities 3 and general practices.4 Variation in general practitioner referral rates is correlated with subsequent variation in elective admissions.5 Given the continuing rise of both elective and emergency admissions in the United Kingdom, a major drive for the use of this indicator is to increase the accountability of general practitioners both for the quality of their services and for financial control. In addition to the direct financial costs, an unnecessary admission may expose the patient to iatrogenic harm and waste resources.

The two studies in this issue have analysed routine data on hospital admission rates and suggest that general practitioners have a limited influence over such rates.3 4 In one study a twofold variation in emergency and elective hospital admissions between general practices in a London health authority was mostly explained by sociodemographic differences, by patient morbidity, and by hospital factors (p 98).3 In addition, general practices which scored highly on other quality indicators, such as high rates of uptake of cervical smear examinations and the provision of minor surgery and child health surveillance, were associated with higher emergency and elective admission rates. The second study, which examined admission rates for chronic diseases such as epilepsy, asthma, and diabetes (which might be expected to be reduced by good management in general practice), also concluded that variation was mainly explained by factors outside primary care (p 94).4 Both studies indicate that admission rates, despite being easily measurable, are probably poor indicators of quality in primary care. Perhaps this is not surprising, given that it is mainly hospital doctors who decide which patients should be admitted.

More influence from general practitioners might be expected over the fourfold variation observed in outpatient referrals for elective treatment.6 When judged by consultants or against locally agreed guidelines, 10% and 15% of referrals respectively are considered inappropriate.7 There is also evidence, however, of underreferral, though the subsequent use of referral guidelines, feedback on referral rates, and audit of referrals has produced only limited change in overall rates.7 8 The limited impact of referral criteria may be because they have failed to take account of important factors such as ease of access, the interest and skills of the individual clinician, and past experience of or future threat of litigation or complaint.8

A performance framework can be sustainable only if it motivates the practitioners whose performance it seeks to appraise. A balance is needed between what is measurable (such as admission rates) and what is important (such as improved health care and health for patients). In addition, the practitioners must have the ability to effect change. Many general practitioners see the validity of immunisation and cervical smear rates,9 10 which are clearly linked to health benefits, as indicators of quality in primary care. Although general practitioners recognise the importance of chronic disease management in primary care,10 they are less convinced about referral rates 9 10 and would presumably be even more sceptical about hospital admission rates.

Good quality primary care provides a comprehensive, coordinated, continuous programme of prevention, treatment, and care, at first contact. An important aspect is effective communication where the patient and the primary care professional explore a number of possible options. There is empirical evidence that better patient information about the possible benefits and risks of various management options can reduce hospital use.6 The general practice assessment survey (GPAS) attempts to capture the patient's view of communication during the consultation.11 The patient enablement instrument tries to capture what contribution primary care makes to the patient's health and ability to cope with life and patients' knowledge of and ability to cope with their illness.12 The use of these instruments, which are currently under evaluation, have yet to be shown to provide health benefits for patients. Such measures would be relevant to the vast majority of consultations in general practice that do not result in a referral to hospital. These approaches, if extended to other disciplines, may prove to be better for assessing the quality of 90% of patient contacts in the NHS.