No Need, No Bleed: A Quality Improvement Initiative to Rationalise Blood Testing on an Older People's Medicine Ward
Abstract
Background and aim
Unnecessary blood testing increases patient discomfort, workload and costs. We aimed to reduce unnecessary blood tests by standardising decisions about testing frequency and increasing senior clinical oversight.
Methods and interventions
Monthly blood-test activity was collected from September 2024 - June 2025. In July 2025 we used a Plan–Do–Study–Act approach.
Cycle 1 involved local education, discussion at safety huddles and consultants setting expectations for frequency of blood tests during ward rounds. Business managers supported with messaging and data collection.
The primary outcome measure was the monthly number of blood tests performed.
Cycle 1
During the 5 months following the intervention, this fell to 824 per month, from a baseline of 1099. representing a reduction of 25.4% with 6 consecutive points below the median on our run chart.
Cycle 2
After doctor changeover, and a busy winter period with historically high admissions, blood tests went up again to over 1000 tests for 2 months. Renewed messaging was accompanied by a "No need, No bleed" poster, highlighting benefits to both resident doctors and patients. The ward registrar and PA helped resident doctor level engagement.
Weekly data was utilised to provide real-time feedback to teams when testing numbers changed.
February – April data showed blood testing came back down to 846 tests per month. Official monthly data for May and June are outstanding but weekly results suggest sustained improvement.
Over a 10 month period, 2212 fewer blood tests were performed than the previous 10 months representing a saving well over £18000
Conclusions
Clear guidance, senior decision-making, education and repeated feedback reduced blood testing by approximately one quarter. The intervention was embedded within routine practice, even as doctors rotated through the ward. Length of stay, mortality and discharge rates were not adversely affected.