Online testing for sexually transmitted infections: A whole systems approach to predicting value

Abstract

Background: Online testing for sexually transmitted infections has a lower unit cost than testing in clinical services and economic analysis has focused on the cost per test and cost per diagnosis in clinics and online. However, online services generate new demand for testing and shift activity between services, requiring system-level analysis to effectively predict cost-effectiveness.

Methods and findings: Routinely collected, anonymised, retrospective data on sexual health service activity from all specialist services (clinic and online) within an inner London sexual health economy were collated and harmonised to generate a complete dataset of individual level clinic attendances. Clinic activity and diagnoses were coded using nationally standardised codes assigned by clinicians. Costs were taken from locally or regionally agreed sexual health tariffs. The introduction of online services changed patterns of testing. In an inner London sexual health economy, online STI testing increased total number of tests, the total cost of testing and total diagnoses while slightly reducing the average cost per diagnosis. Two years after the introduction of online services 37% of tests in the were provided online and total diagnoses increased. The positivity of online services is generally lower than that in clinics but varies between contexts. Where the positivity ratio between clinic and online is less than the cost ratio, online services will reduce cost per diagnosis. In this analysis, areas with different classifications as urban and rural had different clinic/online positivity ratios changing the cost effectiveness between areas. Even after the introduction of online services, simple STI testing activity continues in clinics and providers should consider online-first options where clinically appropriate.

Conclusions: Online services for STI testing are not 'stand alone'. They change STI testing behaviour with impacts on all elements of the sexual health economy. Planning, development and monitoring of such services should reference the dynamic nature of these systems and the role of online services within them.

Fig 1. Overall volume of STI testing by service provider and quarter from Q1 2014-Q3 2016.

Fig 2. Change in average cost per test (combined online and clinic tests) as the proportion of tests done online increases, under different assumptions of online test costs.

Assumptions: Clinic test costs £66. Scenarios: Vary online test cost as a proportion of clinic cost (0.7, 0.6, 0.5, 04).

Fig 3. Change in average cost per diagnosis as the ratio of test positivity between clinic and online tests is varied.

Assumptions (baseline): 10% of clinic tests are positive, 30% of tests take place online, the cost of online test is 50% of the clinic test cost, the total number of tests is constant. Scenarios: a) Increase the proportion of tests online to 40%, b) Increase the cost of online test to 70% of clinic cost. Red vertical line indicates above online testing decreases the average cost of testing at baseline (positivity ratio 50%).

Fig 4. Effect of return rate on average cost per completed test for online test.

Assumptions: Online test cost £33, sending out test kit costs 15% of total test cost. Scenarios a) Sending test kit costs 10% of total test cost b) Sending test kit costs 30% of total test cost. Red vertical line: Return rates for the online service in this study were above 70%.