Public health interventions tend to be complex and need a lot of planning and preparation before they can be rolled out successfully. But how much should we be willing to spend in order to complete the roll-out sooner?
The astonishing answer from our analysis of HPV primary testing in cervical screening is that, based on standard willingness to pay thresholds, it would have been worth £5 million to introduce HPV testing just 2 months earlier in England.
Economists are used to thinking about the value of information, but how often do health planners think about the value of time? We argue that as soon as an intervention has been shown to be efficacious (i.e. to work in ideal circumstances), planning should begin to ensure that the necessary additional evidence and piloting is undertaken in a timely manner.
Insufficient quality assurance can lead to an ineffectual intervention. Insufficient piloting can lead to unforeseen knock-on effects that negatively impact on other aspects of the health service. To avoid such disasters, we tend to play it safe and take our time before rolling out a new service or a modifying an existing programme. But what is the cost of taking longer than is strictly necessary? Consideration should be given to the value of completing such evaluation and subsequent roll-out sooner. Preparing a project plan with a timeline and budget early on could optimise the cost-effectiveness of timely implementation.
Our paper in the Journal of Medical Screening used primary HPV testing as an example of what is at stake. It is 11 years since HPV testing was first shown in a randomised controlled trial to be better at detecting cervical cancer precursors than traditional cytology-based screening. And yet, few countries have fully switched over from cytology to HPV testing in their cervical screening programmes. Co-testing (with both HPV testing and cervical cytology) was introduced by Kaiser Permanente Northern California in 2003, but there is no comprehensive national screening programme in the USA and screening is quite heterogeneous. Sweden and The Netherlands introduced HPV testing in 2017 and Wales in September 2018, but England will not switch over until the end of 2019.
In the paper, we estimated the number of new cases of cervical cancer that would be diagnosed in England between now and 2030 under two scenarios:
- HPV testing is brought in for cervical screening in December 2019 (as planned)
- HPV testing is brought in for cervical screening in December 2020 (one year later than planned).
We also estimated at what stage those cancers would be diagnosed, how much it would cost to treat them and how many would lead to premature death. These could then be converted into Quality Adjusted Life Years (QALYs), a standard measure of the benefits of health care in terms of duration and quality of life. 1 QALY represents 1 year of life in perfect health. There are agreed thresholds as to what the NHS should be willing to pay to provide an extra QALY. Using such a threshold, we estimated how much we (as a society) should be willing to pay to avoid a further delay in the roll-out of HPV primary screening.
The result took us by surprise. We estimated that we should be willing to pay (at least) £32 million to avoid a 12-month delay in the roll-out of HPV primary screening. Had everyone realised that it would have been worth spending £20 million to roll-out HPV primary screening in April 2019 rather than December 2019, might things have been done differently? It is too late for HPV testing, but let’s make sure that cost (as well as the advantages) of moving slowly is factored in to the timelines for introducing future public health interventions.
Some people would like to lay blame for the extremely slow-roll-out of HPV testing (the first clinical trial of HPV testing in primary cervical screening was published, 23 years ago, in 1995), but no organised national screening programme had adopted primary HPV testing by 2010. It is not possible to blame a political party (we have had Labour, Conservative and coalition governments since 1995) or Public Health England (it was not established until April 2013). Should academics have done more? What about charities? The problem is that no one (and no organisation) has the job to ensure that a promising public health intervention is evaluated, piloted and rolled-out in a timely manner.
In our paper we concluded that there “is a measurable loss” that “should be considered in prioritising decision-making in screening”. This is perhaps a timely message with the very recent announcement of the positive results from a large European trial of low-dose CT scanning for lung cancer screening. There seems little doubt that lung cancer screening will be introduced across Europe, but will it take 5, 10, 15 or even 20 years to roll-out? As we wrote in our paper “While careful planning is essential, sometimes there is a heavy price to pay for being overcautious”.
The views expressed are those of the author. Posting of the blog does not signify that the Cancer Prevention Group endorse those views or opinions.