The publication of a meta-analysis earlier this year showing annual health checks are useful has revived a long-running debate about the worthiness of the practice.
General health checks started in the middle of the last century in North America, and spread to other places, as “executive health checks”. As the name suggests, they were usually undertaken for high-ranking people in large companies to check there was no remedial, but undiagnosed, problem.
Soon many physicians were offering such checks to all their patients, and it became standard practice to undergo an annual check-up. The check-ups rather mimicked the type of examination we undergo for life insurance, to check whether someone is a good risk, or whether premiums need to be adjusted upwards.
Indeed, general health checks seem to offer an analogy with servicing cars – an essential thing to do to avoid breakdown in the near future. But do they actually do any good?
These taskforces were huge endeavours, costing millions of dollars, and they were one of the first forays into what we now call evidence-based care.
The results, when they started filtering in, found a lot of what was being offered (listening to the heart and lungs with a stethoscope, for instance, or feeling the abdomen, and routine ECGs) didn’t have evidence of benefit. Although few things, such as taking blood pressure, and of course childhood vaccinations, did.
Still, most of the rituals of annual health checks looked as if they might be worthless. A controversy was starting to brew.
In the meantime, the notion of opportunistic preventive care was being developed across the Atlantic. A pair of academic general practitioners in the United Kingdom had started talking about the potential to do good during every consultation by focusing on at least one preventive activity.
The choice of what that should be was influenced by a randomised controlled trial published at around the same time. It showed general practitioners telling people to stop smoking, together with giving them a leaflet (pretty unsubtle behavioural science by today’s standards) resulted in a reduction of smoking by about 4%.
This sounds pretty modest, but when multiplied up by the vast numbers of general practice consultations, the high rates of smoking at that time, and the multiple forms of harm tobacco causes, the health benefits of such a simple intervention were astounding.
Knowing your patients’ smoking status in the medical record became (and still is, actually) a marker of good quality general practice.
Thus were the battle lines drawn up between opportunistic preventive care protagonists, and the annual health check ones. And battles are still fought over what preventive activities are effective, meaning “is there trial evidence the intervention saves lives?”
Controversies rage over individual activities. Currently the most bitter are over prostate cancer screening, breast cancer screening in the age group between 40 and 50, and vaccinating against influenza.
A few activities have such good non-trial evidence that they’re accepted without it, such as screening women for cervical cancer.
And the battle lines move with the arrival of new nuggets of information. The new anti-human-papilloma-virus vaccine against cervical cancer, for instance, is likely to render the smear test obsolete in years to come, in many western countries.
A 2012 Cochrane review of the evidence for annual (or even longer) health checks seemed to put a nail in the coffin of the practice. It found no benefits in terms of reduced deaths (“mortality”) or even reduced health outcomes (“morbidity”).
It did find evidence of harms, especially overdiagnosis of conditions that would never have caused problems in life, and overtreatment of those conditions.
But the annual health check was given a reprieve earlier this year with the publication of another meta-analysis. This one was undertaken slightly differently, in particular with a focus on surrogate endpoints. In other words, rather than mortality or morbidity, measures such as blood pressure, body mass index, and cholesterol, were used to assess effectiveness.
Those of us who practice evidence-based medicine are somewhat suspicious of surrogate endpoints. All too often, these are easy to change but with no change in what really matters – how well people are rendered, or if fewer die earlier.
And indeed, in this meta-analysis, there was no reduction in deaths, or even deaths from cardiovascular disease (rather, embarrassingly it was the opposite, a slight statistically significant increase in cardiovascular deaths). But, we shouldn’t lose sight of individual preventive health activities that are worthwhile.
General health checks are hard to justify, and may even be harmful, while a limited (but nonetheless extensive) number of individual preventive activities are beneficial. These can be undertaken opportunistically during normal general practice problem solving.
Chris Del Mar, Professor of Public Health