Skip to NavigationSkip to content

Look beyond the screen

Published on 03/03/14 at 08:20am
Blood pressure image
Blood press screening is an 'old favourite' which is crucially underpinned by strong evidence

It makes sense, doesn’t it? You have a drug that’s effective for a certain disease, but no patients. What do you do? Go out and look for them, of course! But how do you do that legally and ethically (assuming that there’s anyone out there who bothers with such niceties)? 

This is exactly the challenge that faced a company of my acquaintance a few decades ago. The potential was seen as huge - existing prevalence data were limited, but suggested that there were patients at risk in vast numbers.

So a project was designed that looked guaranteed to be self-financing. Patients in primary care would be invited for health and lifestyle screening. Data would be collected on the prevalence of risk factors, which would be genuinely useful to clinicians, especially on the predictive value of certain markers.

But at the same time patients at high risk, and eligible for treatment with the company’s drug, would be identified. All the figures pointed towards a very nice impact on sales. What could go wrong? 

This is what went wrong: to make the protocol ethical, high-risk patients had to be offered a non-drug option that was at virtually zero cost. All the previous data suggested that this was a poorly effective measure, so most patients would go onto drug treatment. Experience proved otherwise. 

Administered under close supervision by GPs, the non-drug option was much more effective than expected, so far fewer patients went onto drug treatment. How do you think the company responded to this bad news? Sometimes the best response can be counter-intuitive, and that’s what the company did. It set up another screening programme, 10 times the size.

But it was the wrong decision - the programme ran out of money, and had to be closed, amid much wailing and gnashing of teeth. The original project was completed and published, so at least science gained something out of it. But we are left with a load of questions about screening programmes, and not just whether they make business sense.

The truth of course, is that health screening is big business. Just look at the ads from BUPA and other private providers. Again, it makes intuitive sense - prevention must be better than cure. However, the extent to which that holds depends on how good the test is.

Note what BUPA says about the prostate-specific antigen test: “This is not a reliable test for prostate cancer and may miss and/or suggest the presence of a cancer where there is none.” What underlies that statement is the fact that the PSA test was never developed as a diagnostic, but was for the monitoring of established disease. It also begs the question of why customers are invited to pay for it. 

The reason intuition doesn’t work in this situation is that we tend to think dichotomously - i.e., something is either good or it isn’t. Science doesn’t work like that, which is why we have statistics; science deals with probabilities, and that makes health screening a little hard to understand.

I have seen no better explanation than that of Dr Margaret McCartney in her book The Patient Paradox. To paraphrase, suppose there is a disease that affects 1% of the population, for which there is a diagnostic test that is 90% accurate. This means that of 100 tests, 90 will be correct and 10 will either be false positives or false negatives.

If 1,000 people are screened using this test, 10 will actually have the disease. But the test will say that about 100 people have the disease when they don’t. The reason is that as 99.9% of people are not affected, most errors will be false positives.

So, the screening process turns up far more false positives than real positives - and a 90% accurate test is usually considered quite good. 

When this scenario was presented to a cohort of GPs, they mostly thought that 90% of people with positive tests would have the disease, when the truth is that the figure is less than 10 per cent. 

We next need to consider what we do with all of these patients. We know that some of them need treatment, and we also know that most of them don’t. Let’s suppose that patients identified using this process are deemed eligible for treatment with a medicine.

No drug is equally effective in all patients, and the cold hard truth is that most drugs are ineffective in a substantial proportion of patients. This is why we do controlled clinical trials, to measure the proportions of responders in active drug groups compared to control groups (placebo or another drug).

Here we have a population in which most patients with a positive test result don’t have the disease, and a drug which, naturally, won’t work in any of them, and also won’t in quite a lot of the minority that do have the disease. So out of the whole exercise, do we have a triumph for the pharmaceutical companies? I don’t think so.

Yes, there are rudimentary safeguards built into some screening programmes. In breast cancer screening, for example, secondary screening seeks to confirm the first result, as well as take biopsies. But the same rules apply to these measures - nothing is 100% accurate, meaning the problem of overdiagnosis can be reduced, but never eliminated.

Much of the uncertainty surrounding routine mammography stems from the Cochrane review of 2011, which concluded: “For every 2,000 women invited for screening throughout 10 years, one will have her life prolonged.

“In addition, 10 healthy women, who would not have been diagnosed if there had not been screening in the first place, will be diagnosed as breast cancer patients and will be treated unnecessarily. Furthermore, more than 200 women will experience significant psychological distress for many months because of false positive findings”.

You’ll find these figures reflect Margaret McCartney’s hypothetical scenario rather well. Yet mortality from breast cancer has improved dramatically. Ten-year survival is now 77% compared with 54% 20 years ago (figures from Cancer Research UK). Is this not a triumph for mammography screening?

Not necessarily. It is far more likely to be related to much more effective treatment - of which the pharmaceutical industry can be justly proud. So what of screening for other cancers? There doesn’t seem to be very robust evidence for any, although bowel cancer screening is in progress in the UK - not without some criticism.

The US Preventive Services Task Force does not recommend routine screening for lung cancer, although this might be worthwhile for high-risk groups, especially smokers.

Of course, many diseases other than cancer can kill you. Alongside ‘prevention is better than cure’ lies the mantra ‘early rather than late’. Again, while this may sound like good sense, young people tend to be healthier, so you will have to screen a lot more to find those who do need treatment.

This was tested in the US by Wang et al., who found that screening blood pressure in adolescents was only marginally cost-effective. Population-based interventions such as dietary salt reduction and physical exercise are likely to be more cost-effective.

Conversely, our old favourite, blood pressure screening in the adult general population, still enjoys a widepread positive reputation. The US Preventive Services Task Force also looked at this in 1996, and again in 2003, and concluded that “strong indirect evidence supports screening adults for hypertension”.

However, this was qualified by several questions; for example where and how to measure blood pressure (office, or ambulatory monitoring), the role of so-called ‘white coat’ hypertension (where stressed patients’ blood pressure rises in their doctors’ clinics), and the negative effect of stigmatising affected patients.

So the message is that all forms of health screening must be underpinned by robust evidence. Sometimes the evidence is good, often it’s not. Yet something weird is going on with breast cancer screening in the UK. The government has committed to extending the age range for routine mammography by three years at either end.

But the Department of Health agrees that there is no robust evidence that screening is effective in the extended age range, so it is conducting a randomised controlled trial to test the effect of it. Yes, you read that right, they are rolling out the extra screening and testing whether it works at the same time! 

What happens if at the end of the 10-year trial they find it doesn’t work, and thousands of women were screened and treated unnecessarily? I have not found anyone who wants to answer that question. 

So what should industry do about getting involved in screening? On balance, it appears not to be a good idea. As we have seen with breast cancer, where industry shines is in treatment, and if companies are short of cash, then surely they should put what they do have into developing even better drugs? 

Perhaps it no longer makes commercial sense to spend twice as much on sales and marketing as we do on R&D. The difficulty lies in balancing short-term gain in sales against revenues from the next blockbuster drug, which could be 10 years away.

Les Rose is a freelance clinical research consultant and medical writer. 

Mission Statement is a leading portal for the pharmaceutical industry, providing industry professionals with pharma news, pharma events, pharma service company listings and pharma jobs,
Site content is produced by our editorial team exclusively for and our industry newspaper Pharmafocus. Service company profiles and listings are taken from our pharmaceutical industry directory, Pharmafile, and presented in a unique Find and Compare format to ensure the most relevant matches