Making sense of health data, research and public health
ICD-10 codes in the R00–R99 chapter are often read as a sign of medical mystery. In NHS hospital statistics and ONS mortality data, they more often record a familiar administrative and clinical reality: a patient needed a code before a firm diagnosis was available.
Death certificates become public health statistics through a chain of clinical judgement, legal process and coding. This guide explains how cause-of-death data is produced in the UK, what underlying and contributing causes mean, and why mortality figures deserve careful reading rather than immediate conclusions.
As winter 2026 approaches, NHS bed occupancy and discharge delays return to the headlines. But flow data are operational instruments, not a verdict on care. This analysis explains what the numbers record, where they mislead and how to read them with winter in mind.
The MHRA Yellow Card scheme turns individual suspicions about medicines into a national safety signal. But a report is not a proven harm, and a rise in entries is not automatically a rise in disease. This guide explains how to read the data with the caution the system itself requires.
The national data opt-out is a right, and a small but consequential one. This deep dive examines who withholds their GP data, why missing records are rarely missing at random, and how researchers are trying to keep UK disease estimates honest without eroding the trust the system depends on.
Most alarming UK health statistics are arithmetically correct and analytically shaky. This explainer unpicks incidence, prevalence, testing rates and population bases, and offers a five-question checklist for reading the next scare story with a cooler head.
Open prescribing data offer one of the most granular views of long-term illness in the UK, but they record treatment rather than diagnosis. This guide explains how to read repeat prescriptions for statins, inhalers, antidepressants and diabetes drugs without mistaking the pharmacy ledger for a disease register.
Rare disease registries are among the UK's most ambitious attempts to make small patient populations visible to clinicians and researchers. They also expose hard questions about consent, standardisation, linkage and statistical disclosure—questions that determine whether the data can be trusted and used well.
A routine GP consultation produces more than a narrative note. It generates structured codes, quality indicators and free text that can be extracted for research—if the data are good enough. This explainer traces how a ten-minute appointment becomes secondary-use health data, and where patients should look for the gaps.
Wastewater sampling can detect influenza, RSV, norovirus and SARS-CoV-2 before clinical testing climbs—but a rising signal is not an outbreak. As the UK enters the 2026 winter respiratory season, we explain how sewage surveillance works, where interpretation falters, and how public-health teams decide when to act.
Smartwatches now generate heart rhythms, oxygen levels, sleep scores and readiness indices by the hour. We separate the handful of measurements with genuine clinical traction from the ones that are, for now, consumer decoration — and explain how to raise a wearable finding without derailing a ten-minute appointment.
Interactive maps of neighbourhood-level disease, deprivation and service use are now published in abundance. This sceptical field guide explains how to read them properly — why small numbers wobble, why registers are not populations, and which questions to ask before concluding anything about a postcode's health in 2026.
Digitised tissue samples and AI tools are moving from research laboratories into NHS pathology networks. The promise is faster, more consistent diagnosis; the reality is a work in progress, with evidence gaps, governance questions and a continuing need for human judgement.
Anonymised NHS records are used to spot diabetes hotspots, track mental-health prescribing and plan services. This explainer sets out seven signals worth watching in 2026—and the safeguards and limits that shape what the numbers can reasonably tell us.