ideasIdea
Use the trend in routine blood counts, not a single threshold, to spot cancer
A platelet count that is normal but rising year on year, or haemoglobin drifting down, can signal cancer. Records already hold these trends; software could use them.
Thrombocytosis and unexplained anaemia are known predictors; longitudinal within-person trends are more informative than population thresholds. Propose trend-based alerting in primary care (e.g. a 20% platelet rise over two years within the normal range), evaluated in a linked-data cohort and then an RCT.
Hypothesis
Within-person trend flags identify at least 20% more incident cancers 6-12 months before diagnosis than threshold-based flags at the same alert rate.
Rationale
Analogous to using individual baselines rather than population cut-offs in other fields; the data is already collected.
What would test it
Retrospective validation in CPRD or similar linked datasets (small cost); then a prospective pragmatic trial.
Maturity
speculative
Who has to act
data
Cost to try
Small (under $1M)
Years to first evidence
3
Bottlenecks it attacks
- Most lethal cancers are found late · Screening exists for only a few cancers. Pancreatic, ovarian, liver, oesophageal and most lung cancers are found when cure is unlikely.