Covid cases rising, a guide to life’s risks, and racing jelly-fish

Episode Summary

Executive Summary: The episode examines rising UK Covid-19 case numbers, showing they appear to reflect a real increase rather than a data anomaly, while cautioning that testing patterns and local outbreaks affect interpretation. It also explains how to think about individual Covid risk, critiques government test-and-trace performance and testing turnaround targets, and debunks a jellyfish speed claim.

Main Topics: Rising UK Covid-19 cases and what they mean (Priority: 5/5): Tim Harford and Josephine Cassidy assess whether the late-summer rise in reported cases is real. They rule out a backlog explanation and conclude infections are genuinely rising, though testing patterns and regional hotspots complicate interpretation. How to understand personal Covid risk (Priority: 5/5): Tim Harford discusses a headline mistake comparing Covid risk to taking a bath, then speaks with Sir David Spiegelhalter about separating the risk of catching the virus from the risk of severe illness if infected. Limitations of case counts and ONS survey data (Priority: 4/5): The show contrasts daily reported cases with the Office for National Statistics infection survey, explaining that ONS is more systematic but slower, less able to detect local outbreaks, and not yet reflecting the newest rise. Test and trace targets and government performance (Priority: 5/5): Simon Mabin analyses whether NHS Test and Trace is meeting official targets. The programme argues the government is using a misleading denominator and is still well short of SAGE’s 80% contact-reach advice. Testing turnaround promises (Priority: 4/5): The episode checks Boris Johnson’s promise that all non-postal tests would be turned around within 24 hours and finds the system remains far from that target across different test channels. Debunking jellyfish swim-speed claims (Priority: 2/5): Jordan Dunbar investigates whether jellyfish can swim as fast as Olympic swimmers. Expert evidence shows the claim is false in absolute terms, though some jellyfish may be fast relative to their body size.

Key Arguments: The recent rise in UK Covid-19 cases is not just a one-day reporting glitch or backlog; experts believe the increase is real. Testing volume and location can distort daily case numbers, but they do not explain the overall upward trend. The ONS infection survey is a better population measure than daily testing data, but it has a lag and can miss localized outbreaks. Risk should be separated into two questions: chance of catching the virus and chance of severe illness if infected. For an older, healthy person, age is a major driver of Covid severity; personal risk is materially lower than in March-April but not negligible. NHS Test and Trace looks better only when using a narrower metric; the more relevant close-contact figure is below the 80% threshold. The government’s 24-hour test-turnaround promise has not been met across the system, despite better performance in some settings. The jellyfish claim is false if interpreted literally; the speed comparison only becomes plausible when normalized by body size.

Data Points: UK daily reported Covid cases: 1,813 to 2,988 - Reported cases jumped from Saturday to Sunday, then stayed near 3,000 on Monday before falling back on Tuesday. Share of positive cases from recent specimens: 95% - Evidence suggests most positive cases come from specimens taken in the last five days. Latest ONS survey cutoff: 25 August - The ONS infection survey lag meant it did not capture the newest rise in cases. Birmingham case count: 212 - Named as the largest English local authority with the highest number of new cases. Hospital Covid patients in England: 425 to 537 - Hospitalizations rose within a week, indicating a worrying upward trend. Possible March-April infection level: 150,000 per day - Estimates suggested infections were vastly higher during the first wave than current levels. Contacts reached where details provided: 84.3% - Government figure cited in Parliament for contacts reached and asked to self-isolate when contact details were provided. True close-contact reach rate: 69% - More relevant measure for actual close contacts reached by NHS Test and Trace. Positive cases reached by Test and Trace: 80% - General coverage of referred positive cases; used in calculating overall effectiveness. SAGE advice threshold: 80% - Advisors say at least 80% of contacts of all positive cases should be reached. Trace improvement needed: One extra contact for every two currently traced - Estimated gap between current performance and SAGE’s recommendation. Prime Minister’s promised turnaround time: 24 hours - Boris Johnson’s June commitment for non-postal test results. Mobile unit tests within 24 hours: 5% - Government data contradicted the PM’s statement that all such tests were within 24 hours. Hospital/outbreak tests within 24 hours: 90% - Best-performing test category, accounting for about a third of tests. Regional and mobile test centres within 24 hours: 50% to 60% - Performance for another third of tests. Satellite centre tests within 24 hours: 1% - Very poor turnaround for urgent satellite tests. Daily risk of Covid death for an early-60s person: 1 in 2 million - Tim Harford’s corrected estimate used to compare individual risk levels. Annual risk of dying in the bath: 1 in 3 million - The incorrect comparison that led to the headline mistake. Chance of dying if infected for a healthy person in early 60s at epidemic peak: about 1% - Sir David Spiegelhalter’s estimate, noting it has likely fallen since then. Bolton case rate example: 98 per 100,000 per week - Used as an example of local risk being far above the national average. Southampton case rate example: 3 per 100,000 per week - Used to show how misleading national averages can be. Average chance of catching Covid in ONS-style data: about 1 in 25,000 per day - Spiegelhalter’s estimate of a more realistic average daily infection risk. Healthy 75-year-old chance of dying if infected: 3% to 4% - Spiegelhalter’s estimate for an older but otherwise healthy person. Annual chance of cancer/heart attack/stroke for a healthy 75-year-old: 5% to 6% - Used to compare Covid risk with other common health risks. Long-distance train study threshold: less than 10 hours - German-Austrian rail study found no infections among passengers with stays shorter than this. Austria proven cases in contact tracing example: 4,700 - No public-transport infections were found among these proven cases. Jellyfish speed claim: 2.1 meters per second - A 1960 report claimed a box jellyfish swam at Olympic-sprinter speed, but current effects were not measured. Typical lab-measured jellyfish speed: about 0.05 meters per second - Modern measurements show much slower swimming speeds. Olympic swimmer comparison: 25 meters in 12 seconds - Contrast used to show how far slower jellyfish are in absolute terms. Jellyfish proportional speed: 1 to 5 body lengths per second - Relative-size comparison suggesting some jellyfish may be fast for their size.

Pivotal Quotes: "the consensus is that amid all the noise, this is real." — Josephine Cassidy: On whether the rising Covid case numbers reflect a genuine increase or a reporting artifact "Claims which go viral because they're surprising and interesting are often surprising because they're untrue." — Tim Harford: Discussing the newspaper headline comparing Covid risk to taking a bath "pay enormous attention to what is the actual event we're talking about? And in particular, what time period are we measuring it over?" — Sir David Spiegelhalter: Explaining how to correctly assess risk and avoid misleading comparisons

Implications: Listeners should treat Covid statistics cautiously: daily case totals are informative but incomplete, and personal risk depends on both exposure and vulnerability. The episode also warns that official targets can be made to look better through selective wording, so underlying definitions matter.