General health checks, such as the “annual physical”, are common elements of health care in many countries, including the United States. Their purpose is to reduce ill health and prevent premature death by improving early detection and treatment of diseases. Now a systematic review for the October issue of The Cochrane Library questions whether they have any benefit, especially as it finds they do not reduce deaths, either overall or from serious diseases like cancer and heart disease.
Lead researcher Lasse Krogsbøll, of The Nordic Cochrane Centre in Copenhagen, Denmark, and colleagues, say general health checks should not form part of a public health programme.
“What we’re not saying is that doctors should stop carrying out tests or offering treatment when they suspect there may be a problem. But we do think that public healthcare initiatives that are systematically offering general health checks should be resisted,” says Krogsbøll in a statement.
Writing in an accompanying editorial, Stephanie Thompson and Marcello Tonelli, of the Department of Medicine at the University of Alberta in Canada, support this view (which they note is also shared by national expert panels), and urge doctors contemplating screening to make routine health screening more patient-specific, such as taking into account risk factors, supported by high quality evidence.
They add that the review is important because “it focuses on patient-relevant outcomes such as all-cause and disease-specific mortality, as well as morbidity, hospitalization, patient worry, self-reported health, and cost.”
One of the disadvantages raised by the review is that general health checks can potentially lead to diagnosis and treatment of conditions that may never produce symptoms of disease or shorten life.
For their study, Krogsbøll and colleagues reviewed 14 randomized trials from primary care or community settings involving a total of 182,880 people. They did not include trials that enrolled only people aged 65 and over.
Each trial had divided the participants into at least two groups. As a minimum, they all compared a group of people who had been invited to have general health checks, with a group who had not.
The people invited for health checks had undergone at least one type of screening or evaluation, assessing more than one part of the body (“multiple organ systems”). None had been invited because of a particular disease or risk factor, ie all of the studies had evaluated “asymptomatic populations”.
The health check interventions varied considerably across the trials, and included health questionnaires, stool tests, imaging, blood tests, physical exams, and assessment of cardiovascular risk.
Of the participants in groups invited for checks, the percentage of those who underwent the first round ranged from 50% to 90% (median 82%).
9 of the trials reported on mortality, and followed the participants for 4 to 22 years. Analysis of these found no difference in the number of deaths between those invited for health checks and those not invited: either overall or specifically due to cancer or heart disease.
The relative risk of death was 0.99 (with a 95% confidence interval ranging from 0.95 to 1.03), and results were consistent across studies and in a number of sensitivity and subgroup analyses.
The researchers note that other outcomes were generally poorly studied, but suggest offering general health checks has no impact on hospital admissions, disability, worry, specialist referrals, additional visits to doctors or time off work.
There was one trial where health checks had led to more diagnoses of all kinds. And in another trial, participants in the group invited for health checks were more likely to be diagnosed with high blood pressure or high cholesterol, as one might expect.
In three of the trials, large numbers of abnormalities were identified in the groups invited for health checks.
Krogsbøll says from the evidence they examined, “inviting patients to general health checks is unlikely to be beneficial”.
“One reason for this might be that doctors identify additional problems and take action when they see patients for other reasons,” he suggests.
The researchers suggest new studies should concentrate on the individual components of health checks, and targeting of conditions, such as kidney disease and diabetes. They should look in more detail at the potential harm that health checks can do, as these are often ignored, leading to a false impression of the balance between benefits and harms.
Another problem they raise is that, of people invited to attend health checks, those who actually show up may be different to those who choose not to participate. There is a good chance that people who are at high risk for a serious illness are less likely to attend.
Written by Catharine Paddock PhD