Infection rates dropped dramatically on the Isle of Wight following the launch of an NHS contact tracing phone app, which was later scrapped in an embarrassing government U-turn, according to an analysis. After having initially suggested the app would be rolled out nationally by the end of May, the government ditched the project and decided to focus on manual contact tracing, with no fixed date for when an app might be available. Now new research shows that the region went from among the worst R rates in the country – 147th out of 150 local authorities – to having one of the lowest R rates (10th out of 150) in the country after the app was launched. “It really seems that something different happened on Isle of Wight from elsewhere in the country,” said Michelle Kendall, from Oxford University’s Nuffield Department of Medicine and first author of the analysis. “There were no other areas that had the same trajectory or even close.” The decline was steeper for the Isle of Wight than any other region and the scientists behind the work said it was “disappointing” that an app had not yet been deployed given the potential benefits for containing new outbreaks. David Bonsall, a clinician and senior researcher at Oxford’s Nuffield Department of Medicine, who has been advising the NHS team working on the app, said: “I’m disappointed the app hasn’t been released in the UK by now. I do strongly feel that the app could have a real benefit to the individual in providing them with that personalised information on their risk that allows them to take appropriate decisions on the spread of the virus further.” In April, the Isle of Wight was amongs the worst affected regions. Following the introduction of the test and trace programme, including the use of an NHS app, in May both incidence and the R-value rapidly declined. By mid-June the epidemic appeared to be under control with an estimated reproduction number of around 0.5 and an average of less than 1 new hospital-based infection per week from late May to mid-June. Community testing data confirmed the same positive trends with estimates valuing the reproduction number on the Isle of Wight 20-25% lower than the national level during May and June. The trial on the Isle of Wight began, to much fanfare, at the start of May and the contact-tracing app initially seemed poised to be rolled out to the rest of England within weeks. Then the timetable was pushed back to June, then winter, amid disagreements over whether a centralised model for the app was appropriate. Officials eventually decided to ditch the app that was trialled entirely in favour of the de-centralised model preferred by Apple and Google. The analysis shows that the app reached a far higher number of potential cases and contacts than manual contact tracing. Between the 6 May and 26 May, 160 cases on the Isle of Wight were reported to manual contact tracing, resulting in 163 individuals receiving a notification and request to self-isolate. During the same period 1,524 people reported symptoms to the app resulting in 1,188 receiving an exposure notification. “You can see straight away that you have the potential for a digital platform to work much faster and at scale,” said Kendall. “What you’d like to know is how many actually developed symptoms or were infections that were taken out of the epidemic.” The analysis was not able to prove that the drop in infections on the Isle of Wight was due to the deployment of the app, but Kendall said that this seemed the likely explanation for the trend. A similar drop was not seen in the rest of England when manual test and trace was rolled out, but that could be due to lockdown restrictions being eased simultaneously. The findings, which are not yet peer reviewed, are published on the website MedRxiv.