LAUSR

Background In April 2014 the UK government launched the ‘NHS Visitor and Migrant Cost Recovery Programme Implementation Plan’ which set out a series of policy changes to recoup costs from ‘chargeable’ (largely non-UK born) patients. In England, approximately 75% of tuberculosis (TB) cases occur in people born abroad. Delays in TB treatment increase risk of morbidity, mortality and transmission in the community. We investigated whether diagnostic delay has increased since the Cost Recovery Programme (CRP) was introduced. Methods There were 3342 adult TB cases notified on the London TB Register across Barts Health NHS Trust between 1st January 2011 and 31st December 2016. Cases with missing relevant information were excluded. The median time between symptom onset and treatment initiation before and after the CRP was calculated according to birthplace and compared using the Mann Whitney test. Delayed diagnosis was considered greater or equal to median time to treatment for all patients (79 days). Univariable logistic regression was used to manually select exposure variables for inclusion in a multivariable model to test the association between diagnostic delay and the implementation of the CRP. Results We included 2237 TB cases. Among non-UK born patients, median time-to-treatment increased from 69 days to 89 days following introduction of CRP ( p  < 0.001). Median time-to-treatment also increased for the UK-born population from 75.5 days to 89.5 days ( p  = 0.307). The multivariable logistic regression model showed non-UK born patients were more likely to have a delay in diagnosis after the CRP (adjOR 1.37, 95% CI 1.13–1.66, p value 0.001). Conclusion Since the introduction of the CRP there has been a significant delay for TB treatment among non-UK born patients. Further research exploring the effect of policies restricting access to healthcare for migrants is urgently needed if we wish to eliminate TB nationally.