Skip to main content Skip to navigation

Short Project Descriptions

Show all news items

Improving Quality of Anaemia Diagnosis (iQUAD) in Primary and Secondary Care

Leads: Prof Dan Lasserson (Acute Care Interfaces), Dr Nick Parsons (Meths)

Dates: January 2020 - December 2022

Background:

Anaemia is one of the most common laboratory abnormalities in both community and hospital settings, across high- and low-income countries (Seitz, et al. J Am Geriatr Soc. 2018; Melku, et al. Curr Gerontol Geriatr Res. 2018). Even after adjusting for underlying causes, anaemia is associated with a range of poorer health outcomes including poorer quality of life, hospital admission, progression to institutional long term care, and death (Michalak, et al. Ann Hematol. 2018; Zilinksi, et al. Ann Hematol. 2014). In spite of the prevalence and importance of anaemia, its management is not subject to any national guidelines in the general healthcare setting. Although there are guidelines that apply to specific differentiated anaemia syndromes, e.g. iron deficiency anaemia (Goddard, et al. Gut. 2011), there is no mandated guideline to deliver a standardised approach to anaemia detection and initial investigations in primary care or in acute medical settings. Our previous research at the level of a 600,000 population found evidence of a highly variable approach in primary care to the investigations of patients with anaemia, particularly affecting older people (McCartney, et al. BJGP Open. 2017). The prevalence of anaemia was 30% among patients 65 years and older, and of those with anaemia, only around half had subsequent investigations, often incomplete, to fully differentiate the cause of the anaemia. Where ferritin was measured, it was low in around 30% of patients, indicating the significant need for appropriate iron replacement (McCartney, et al. BJGP Open. 2017). Similar marked inequalities have been found in a cohort of >13,000 patients treated in the acute medical unit at the Queen Elizabeth Hospital, Birmingham. Anaemia was common, with over 50% of patients aged 80 years and above meeting World Health Organization (WHO) criteria for anaemia but investigation was very limited with only 11% of anaemic patients getting any further tests to investigate the cause. There is an urgent need to deliver a standardised approach to anaemia management in both community and acute hospital settings. This will ensure that at every healthcare contact, the patient with anaemia will get the right investigations at the right time leading to the right treatment.

Policy and Practice Partners:

Sandwell and West Birmingham Hospitals NHS Trust, Royal Wolverhampton NHS Trust, Dudley Group NHS Foundation Trust, Walsall Healthcare NHS Trust.

Co-Funding Partners:

Sandwell and West Birmingham Hospitals NHS Trust, Royal Wolverhampton NHS Trust, Dudley Group NHS Foundation Trust, Walsall Healthcare NHS Trust.

Aims and Objectives:

  1. Develop and embed a computer algorithm within the laboratory information management system (LIMS) of Sandwell and West Birmingham NHS Trust to automatically identify patients with anaemia from blood tests undertaken in primary or secondary care. The algorithm will determine the need for additional testing and suggest appropriate investigation and/or treatment which will be communicated electronically to the requesting clinician.
  2. Determine the impact of the decision support algorithm on the proportion of patients with fully differentiated anaemia on laboratory testing.
  3. Assess the impact of the clinical advice on improvements in Hb levels and frequency of repeat testing for patients in acute and community settings.

Methods:

Development and embedding of the algorithm at SWBH and the wider Black Country Pathology Partnership.

Consensus will be determined across clinical specialties to construct the decision support algorithm to reduce the proportion of patients with inadequate investigation for anaemia and to reduce over-investigation where clear answers to clinical questions are already contained within existing results in the laboratory database. The algorithm will therefore support rational diagnostic testing for anaemia, using existing resources tailored to the local population.

We will use case scenarios to reflect the range of laboratory and clinical features of anaemia and use an appropriateness ratings methodology to clarify where there is disagreement and promote consensus building (Blakeman, et al. BMJ Open. 2016).

This will primarily identify patients with a full blood count indicating anaemia on WHO criteria, and, through linked haematology and clinical biochemistry datasets, will determine which tests are required to provide a first line differentiation of the cause of anaemia (as some tests may have been completed within a relevant timeframe and will not need to be repeated). Relevant additional tests potentially include blood samples e.g. haematinic deficiencies, inflammatory status markers, immune conditions, endoscopy as well as radiological tests (CT colonography). The output from this decision support algorithm will be displayed within the laboratory report, which will be sent directly to General Practices (for bloods taken in community settings) or appear within the LIMS report to the electronic hospital record for bloods taken in hospital. Advice on treatment strategy if needed e.g. haematinic supplementation or cessation will be included, with links to current guidance. The decision support algorithm guidance will be developed in conjunction with acute medical team representatives and general practitioners so that the outputs are delivered in a format that is acceptable to requesting clinicians. Prof Lasserson’s previous role in the NHS England Acute Kidney Injury Programme Board (co-chair of Measurement Workstream) provides additional expertise in barriers and facilitators to embedding algorithms within the LIMS as well as understanding the optimisation of messages from alerts to the clinical community (Blakeman, et al. BMJ Open. 2016).

Main Results:

Awaited.

Conclusions:

Awaited.

Implications for Implementation:

If our results confirm that the computer decision support is feasible and likely to be beneficial, then it has immediate implications for primary care services, laboratory services, and commissioning organisations throughout the country and beyond.