Practice variation in the use of tests in UK primary care: A retrospective analysis of 16 million tests performed over 3.3 million patient years in 2015/16

Jack O'Sullivan, Rafael Perera-Salazar, Carl Heneghan, Chris Salisbury, Ben Goldacre , Paul Little, Jason Oke, Sarah Stevens, F. D. Richard Hobbs, Clare Bankhead, Jeffrey Aronson

Research output: Contribution to journalArticle (Academic Journal)peer-review

5 Citations (Scopus)
206 Downloads (Pure)


The UK’s National Health Service (NHS) is currently subject to unprecedented financial strain. The identification of unnecessary healthcare resource use has been suggested to reduce spending. However, there is little very research quantifying wasteful test use, despite the £3 billion annual expenditure. Geographical variation has been suggested as one metric in which to quantify inappropriate use. We set out to identify tests ordered from UK primary care that are subject to the greatest between-practice variation in their use.

We used data from 444 general practices within the Clinical Practice Research Datalink to calculate a co-efficient of variation (CoV) for the ordering of 44 specific tests from UK general practices. The co-efficient of variation was calculated after adjusting for differences between practice populations. We also determined the tests that had both a higher-than-average CoV and a higher-than-average rate of use.

In total, 16,496,218 tests were ordered for 4,078,091 patients over 3,311,050 person-years from April 1st 2015 to March 31st 2016. The tests subject to the greatest variation was drug monitoring 158% (95%CI: 153% to 163%), Urine Microalbumin (52% (95%CI: 49.9% to 53.2%)), Pelvic CT (51% (95%CI: 50% to 53%)) and Pap smear (49% (95%CI: 48% to 51%). Seven tests were classified as high variability and high rate (Clotting, Vitamin D, Urine Albumin, Prostate Specific Antigen (PSA), Bone profile, Urine MCS and C-reactive Protein (CRP)).

There are wide variations in the use of common tests, which is unlikely to be explained by clinical indications. Since £3 billion annually are spent on tests this represents considerable variation in the use of resources and inefficient management in the NHS. Our results can be of value to policy makers, researchers, patients, and clinicians as the NHS strives towards identifying overuse and underuse of tests.
Original languageEnglish
Article number229
Number of pages9
JournalBMC Medicine
Publication statusPublished - 20 Dec 2018


  • Overuse
  • Health policy
  • Primary care
  • General practice
  • Test use
  • Imaging


Dive into the research topics of 'Practice variation in the use of tests in UK primary care: A retrospective analysis of 16 million tests performed over 3.3 million patient years in 2015/16'. Together they form a unique fingerprint.

Cite this