The theory of the government’s test and trace strategy is simple; in order for life to get back to normal we need to be able to detect who is infected, who they have been in contact with and stop them from infecting others.

But the rollout of the scheme is proving difficult. And because a “positive” test result does not confirm whether someone is contagious, there is a risk that test and trace, however well implemented, becomes a blunt instrument.

After a sluggish start, testing capacity in the UK is improving. Around 170,000 tests are now being processed every day, up from 100,000 in mid-June. As the government points out, the 2.4 tests per 1000 population now being carried out per day in the UK is higher than in France, Germany or the United States. In recent weeks the focus of testing has shifted towards local hotspots, like Bolton and Trafford, and care homes, where there is a high level of virus transmission among an at-risk group.

This reallocation has restricted numbers of home testing kits and meant fewer booking slots at local testing centres for people in low-risk areas. Baroness Dido Harding, head of NHS Test and Trace in England, apologised after it emerged that UK laboratories were struggling to keep up with demand. A TV campaign has been urging people to get tested if they show any symptoms but there has been a slew of reports of those wanting a test being told to drive over a hundred miles for an appointment. The backlog of tests has also meant longer turnaround times in recent weeks.

Ministers are hoping that new technology – such as a less invasive spit test that is currently being trialled – will help to boost the numbers. Baroness Harding has said that there will be capacity for 500,000 tests a day by the end of October. Around one third of these tests, however, will be antibody blood tests used to establish whether someone has had Covid-19 in the past.

The second pillar of the strategy – tracing – is proving equally problematic. Latest figures show that the fraction of an infected person’s contacts being reached by tracing teams and asked to stay home is 69.4%, down from 77.1% a week before. The figure has fluctuated between 60% and 80% since June. For the test and trace system to have a chance of successfully extinguishing outbreaks, the rate must be close to one.

In an attempt to improve tracing, 6000 contact tracer jobs have been cut and the roles reallocated to regional teams working with local councils. Public health experts have been arguing that the centrally administered scheme is inefficient and that experienced contact tracers have been employed by councils for decades to fight infections like tuberculosis and sexually-transmitted diseases. The shift will likely mean moving towards a South Korean-style approach with councils sending tracers door to door to alert people who cannot be reached by phone.

Even if the system is improved, there are concerns that the binary nature of test results could undermine the efficacy of test and trace.

Research conducted by Professor Carl Heneghan at the Centre for Evidence-Based Medicine suggests that, due to the sensitivity of a PCR test (the swab test used to diagnose whether a patient is currently infected with the virus), a “positive” result could just mean fragments of the virus from an old infection have been detected.

Even though most people are only infectious for about a week tests can return positive results much later. As Henegan notes in a co-authored paper for the Centre for Evidence-Based Medicine: “If this is not understood, PCR results may lead to restrictions for large groups of people who do not present an infection risk.” The reason people are traced is to prevent them from infecting others. Without being able to distinguish “weak positives” from infectious positives, those testing positive will continue to be contact-traced and quarantined unnecessarily.

The number of confirmed cases in the UK has been rising in recent weeks. Heneghan’s insight could help to explain why there has been no accompanying increase in deaths. A similar pattern was observed when Italy emerged from lockdown; cases rose but hospital admissions and deaths remained low. The proportion of weak positives in the region of Lombardy after lockdown was lifted there was found to be 50%. Heneghan’s paper surveys 14 studies which support unpublished evidence that, in Italy as a whole, only 3% of those testing positive were infectious. As testing has increased, more and more of these weak positives have been detected.

“Purely announcing new ‘positives’ tells us very little” says Heneghan. “A reliable test would inform us of whether a ‘positive’ case is infectious, with a reasonable degree of certainty.” To this end, Public Health England has now published guidance for laboratories instructing them to request a second sample if the first comes back positive. Doing so would help to sift out cases that are at the limits of detection (at the very beginning or end of the virus’s infection cycle) and provide more precise diagnoses. As the PHE guidance notes, laboratories must “strike a balance between the risk of false positive test results and an acceptable level of delay in test turnaround time”.

Despite the increase in confirmed cases, levels of infection in the UK remain low compared to other European countries. But until accurate interpretations of positive test results become the norm the test and trace system will struggle to allocate resources where they are needed and keep outbreaks under control.