Intended for healthcare professionals


Rising emergency admissions

BMJ 1995; 310 doi: (Published 28 January 1995) Cite this as: BMJ 1995;310:207
  1. Richard Hobbs
  1. Professor Department of General Practice, The University of Birmingham, Birmingham B15 2TT

    No good evidence yet that this trend is inappropriate

    The National Association of Health Authorities and Trusts reports that providers of secondary care are under pressure from ever rising numbers of acute medical admissions, which were up between 7% and 13% from average rates during 1993–4.1 This is not a recent concern: the Audit Commission, extrapolating from routinely collected hospital data, predicted a doubling of acute admissions to Birmingham hospitals between 1987 and 1996.2

    The National Association of Health Authorities and Trusts suggests that the probable causes of the increase include more detected illness, raised expectations of general practitioners (especially fundholders), premature discharge, and worries over litigation. Others point accusingly to variations in general practitioners' referral rates; Acheson was irritated in 1985 that “a phenomenon so gross can continue to defy analysis.”3 Fundholders are accused of protecting budgets, although they generally do not have to pay for emergency care.

    Unfortunately, evidence to illuminate this issue is scarce. Although a substantial body of research exists on total referrals to hospital, remarkably few data exist on emergency admissions. Overall, general practitioners' referral rates vary threefold4 to fourfold,5 but this variation is less than that which occurs in other clinical activities, such as prescribing. Variables relating to patients and providers explain surprisingly little of this variation—perhaps as little as 7%.2 Roland suggests that this is mainly due to the lack of precision in measurement and the inability of many studies to distinguish between random variation and systematic differences.6

    Even the true scale of this rise in admissions is unknown because of hospitals' reclassification or misclassification of the category of admission. No uniform definition exists of what constitutes an emergency admission: if, during an outpatient appointment booked weeks before, the consultant decides to admit the patient that day, is this always classified as an emergency? If a patient seen after a minor overdose is kept in hospital until a daytime assessment by a psychiatrist, does this constitute a true urgent admission (important since drug overdose may be the single most common cause of emergency admission7)?

    The proportion of referrals from general practitioners classified as immediate or urgent is relatively low in Britain compared with much of Europe (37% in Britain compared with 44% in France and 89% in Germany) and equivalent to that in countries not providing out of hours general practitioner services such as Denmark (37%).8 A stark message for the new NHS market lies in these data, as countries with consumer led systems, such as Germany, exhibit the highest ratios of acute to routine admissions, thus “achieving consumer objectives regardless of medical priorities.”9

    The growth in acute admission rates is a medical rather than a surgical issue. Permanently reallocating “surgical” beds to medical firms would largely solve the shortages of beds, according to the Audit Commission.2 Presumably, the continued shift to day surgery will make this more feasible. There are probably just enough hospital beds in the system, but they may be in the wrong place (in towns rather than cities) or allocated to the wrong specialty.

    The referring behaviour of general practitioners does not seem of paramount importance as hospital doctors—not general practitioners—decide on admission. Furthermore, one study found that the general practitioner had been involved in only half of emergency medical admissions to one hospital.7

    Clinical practice over discharge is considered to be a further influence. The increased throughput of patients despite the existence of fewer beds has been achieved by falls in average hospital stays10 and increased rates of day case surgery (which rose by 16.5% between 1992–3 and 1993–4, to 2.1 million cases in England10). However, it remains a perverse index of NHS efficiency that measures early discharge of a postoperative patient who needs readmission 48 hours later because of a wound infection (perhaps under a different consultant or at a different hospital) as desirable activity—two short finished consultant episodes rather than one longer one. Readmission rates may be a more useful indicator of quality than league tables of mortality.

    Clearly, unexpected admissions are more difficult to cost accurately than planned activity. Some trends may, however, be predicted. The availability of fibrinolytic treatment in acute myocardial infarction should be driving up emergency admissions. Indeed, the NHS should perhaps be expecting a larger proportion of admissions to become immediate in a service with more outpatient diagnostic services available to more comprehensively skilled general practitioners. Should patients who do not require urgent care or planned invasive procedures be in high dependency institutions?

    Commissioning research on some basic interventions seems worth while. One example might be to give general practices budgets to purchase nursing home care for acute social or low dependency medical crises, with cover provided by general practitioners. We need more data, which should be collected against explicit definitions, routinely validated, and systematically analysed; more debate, as opposed to argument, between purchasers and providers over the appropriateness of acute admissions; and more precise research and evaluated interventions. Meanwhile, whether the rising trend of emergency admissions is unjustified is debatable.


    1. 1.
    2. 2.
    3. 3.
    4. 4.
    5. 5.
    6. 6.
    7. 7.
    8. 8.
    9. 9.
    10. 10.