Intended for healthcare professionals


Reducing hospital admissions

BMJ 2008; 336 doi: (Published 03 January 2008) Cite this as: BMJ 2008;336:4
  1. Sarah Purdy, consultant senior lecturer in primary health care,
  2. Tom Griffin, research associate in primary health care
  1. 1University of Bristol, Bristol BS2 8AA
  1. sarah.purdy{at}

    Guidance should be evidence based and take a holistic view of patient care

    In March 2007, the National Health Service (NHS) Institute for Innovation and Improvement published a Directory of Ambulatory Emergency Care for Adults.1 The directory lists conditions that can be clinically managed outside hospital, with appropriate and prompt access to diagnostic services and specialist advice.

    Admissions to hospital are an increasing source of pressure on health system resources internationally. In the NHS, changes to commissioning arrangements have increased the drive to reduce hospital admissions.2 Unplanned admissions make up 36.7% of hospital admissions in the United Kingdom (4 659 054 emergency admissions in 2005-6).3 The Department of Health has introduced a national target to reduce the number of emergency bed days by 5% by 2008.4

    Most of the evidence on the cost effectiveness of reducing admissions comes from the United States. Ways to reduce admissions include case management, observation units for the evaluation of acute conditions, and the provision of home health care.5 6 Research from other countries, such as Australia, highlights the international relevance of reducing avoidable hospital admissions.7 Therefore, the development and implementation of the directory in the UK is relevant to policy makers and clinicians in other countries.

    The directory lists 49 clinical conditions for which admission to hospital could be avoided in 10-90% of cases. It advises that emergency admission to hospital should be limited to acute illnesses that can be managed only in a hospital bed. We support the principle that admissions should be appropriate, and we welcome evidence based guidance that helps clinicians to reduce inappropriate admission to hospital.

    However, the derivation of the proportion of cases that should be managed in ambulatory care is unclear. One systematic review of the appropriateness of hospital admissions in the UK found that 6-20% of emergency medical admissions were inappropriate, depending on the appropriateness tool used, the sample, and the admitting specialty.8 Similarly, a study in the US found that, depending on the method used, 2.3-12.4% of emergency medical admissions in people over 65 were inappropriate, and these admissions were not related to the overall numbers of medical admissions.9

    The management of urgent medical problems has progressed since these studies were published. None the less, it is hard to reconcile these data with proposed reductions of up to 90% in admissions for some common causes of emergency admission—including chest pain, first epileptic seizure, and abdominal pain—especially when the evidence for these reductions is not referenced.

    The directory refers to the concept of ambulatory emergency care, the implication being that emergency care is provided but the patient is not admitted to hospital. The use of this term is unfortunate, as it may be confused with ambulatory care sensitive conditions—admissioncan be avoided in these conditions if appropriate care is provided in the primary care setting.10 A considerable body of work exists in the US, and a growing amount is being done in the UK, on ambulatory care sensitive admissions.11 Some of the admissions listed in the directory—such as chronic obstructive pulmonary disease, congestive cardiac failure, and urinary tract infection—are ambulatory care sensitive. Initiatives to reduce admissions for these conditions are in place in primary care and community care, the resources needed to manage them without hospital admission are available, and their inclusion in a directory of ambulatory emergency care seems appropriate. However, other conditions—such as pulmonary embolism—for which the directory recommends 60-90% of admissions should be avoided, fit less well with the traditional concept of treatment in ambulatory care. The recommendation to treat patients with pulmonary embolism as outpatients is based on a consensus guideline and case control studies or cohort studies, not randomised controlled trials.12

    Decisions on admission to hospital are usually made with a holistic view of the patient’s current state of health, existing comorbidities, available social support, and the patient’s concerns and expectations. The recommendations in the UK’s directory pertain to an ideal situation, uncomplicated by these factors. The reductions in admissions proposed are based on many assumptions, such as the provision of diagnostic facilities that allow the needs of patients to be assessed reliably at the point of access; a decision on management involving providers of medical and social care; and the availability of services and infrastructure to implement alternatives to hospital admission.

    The directory underplays many challenges encountered in day to day clinical care, including uncertainties in diagnosis and appropriate management; difficulties in effective communication between social services and medical services; the availability of appropriate alternatives to admission; the importance of information and knowledge of services among admitting clinicians; and the effect of incentives (such as time targets in the accident and emergency department) on the admitting clinicians.

    Emergency admissions are expensive, they create difficulties for those responsible for planning and delivering services, and they are distressing for patients and their families. Well thought out evidence based initiatives to reduce inappropriate emergency hospital admission are to be welcomed. Future guidance, in the UK and internationally, should explicitly reference evidence on which recommendations are based, and should incorporate perspectives from social services, primary care, patients, and carers.


    • Competing interests: None declared.

    • Provenance and peer review: Not commissioned; externally peer reviewed.


    View Abstract