The value of conducting periodic health checksBMJ 2012; 345 doi: http://dx.doi.org/10.1136/bmj.e7775 (Published 20 November 2012) Cite this as: BMJ 2012;345:e7775
Routine health checks or periodic health examinations, a term used in North America, are seductive. They seem sensible. If prevention is better than cure then they seem a socially responsible approach to caring for patients. Doctors, politicians, and the public buy into the idea that a systematic routine check can identify health problems at an early stage and put them right. A body maintenance programme—like a vehicle service to ensure we are roadworthy—sounds like a good idea. But not every good idea stands up to critical appraisal.
In a linked systematic review and meta-analysis (doi:10.1136/bmj.e7191), Krogsbøll and colleagues comprehensively searched for randomised controlled trials that examined the effectiveness of health checks in adults in reducing morbidity and mortality.1 Carried out according to the exacting methodological standards of the Cochrane Collaboration, the review analysed data from 14 trials with a variety of interventions and outcomes in different settings.2 The authors found no evidence that general health checks made any difference to any of the studied outcomes. They also concluded that such health checks might increase overdiagnosis.
These findings are important for general practice in many countries. Furthermore, health checks are also promoted beyond primary care services, by private clinics, and by industry, where some companies support initiatives to improve employees’ health. The current study carries a useful message for those who create national policy—that such initiatives are ineffective and probably a waste of resources. The original research included in the review originates from many countries with different health systems. Even in countries where health insurance funds care the findings support a clear message to insurers.
What are the limitations of this review? Some of the studies are old and perhaps not as immediately relevant as they once were. There is also considerable heterogeneity between the studies. The interventions took place in several countries, involved different healthcare professionals, and had different outcomes. Some were based in primary care but others were in different settings. Critics might argue that it was inappropriate to gather all these trials together in one meta-analysis. In some trials participation rates were less than ideal, which may reflect the reality of practice. Implementation of interventions within the trials may have been suboptimal, but we cannot tell because details of individual trials cannot be scrutinised. Could it be that follow-up wasn’t long enough to show whether the health check interventions made a difference? This is possible, but the mean duration of the studies suggests that it is unlikely. We should maintain a degree of circumspection in view of the wide confidence intervals, however.
The most interesting question is whether health checks do harm. People may gain inappropriate reassurance from a verdict of a “clean bill of health,” which may lead to continued risky behaviour. A false positive test result may cause considerable worry and upset, not to mention inappropriate treatment. A false negative result also provides inappropriate reassurance. The availability of the routine health check may also divert patients from presenting appropriately with symptoms, signs, or complaints of concern, leaving it up to the doctor or health check to discover the problem. Furthermore, resources may be diverted from diagnosis and treatment to ineffective anticipatory care. Krogsbøll and colleagues found, for example, that health checks increased the diagnosis and treatment of hypertension but with no improvement in outcomes, which, they concluded, suggested overdiagnosis and overtreatment.
When the Oxford and Collaboration Health Check (OXCHECK) and British Family Health studies, which evaluated regular health checks delivered by nurses, were published in the 1990s they stimulated considerable debate.3 4 They showed that scheduled medical examinations led to small changes in cardiovascular risk factors, but there was no consensus on whether it was worth the considerable effort and expense in running these nurse led clinics.
In their review of the cost effectiveness of these studies, Wanderling and colleagues pointed out that the mean number of life years gained per person screened from the British Family Heart Study ranged between 0.0062 (assuming a one year effect) and 0.2035 (assuming a 20 year effect) for men and between 0.0011 and 0.0626 for women.5 The mean number of life years gained from the OXCHECK study ranged between 0.0034 (assuming a one year effect) and 0.1093 (assuming a 20 year effect) for men and between 0.0018 and 0.1065 for women. In a BMJ editorial,6 Nigel Stott discussed the findings of these two studies, pointing out that the health check approach through primary care alone would not produce large reductions in the risk of cardiovascular disease. He called for more effective legislation on controlling the use of tobacco and promoting the consumption of healthy food instead, messages that remain important today.
The current study finds that regular health checks are ineffective. It robustly shows evidence of little effect. It remains possible that targeted health checks might offer some benefit. This study looks at health checks in well people only, and initiatives that are focused on particular population groups with identifiable risk factors and conditions could possibly be effective, but evidence of this is needed. The history of health promotion through routine health checks has been one of glorious failure, but generations of well meaning clinicians and public health physicians struggle to allow themselves to believe it. We need to reinforce the message lest some enthusiast reinvent the health check in another guise. Policy should be based on evidence of wellbeing, rather than on well meant good intentions.
Cite this as: BMJ 2012;345:e7775
Competing interests: The author has completed the ICMJE uniform disclosure form at www.icmje.org/coi_disclosure.pdf (available on request from the corresponding author) and declares: no support from any organisation for the submitted work; no financial relationships with any organisations that might have an interest in the submitted work in the previous three years; no other relationships or activities that could appear to have influenced the submitted work.
Provenance and peer review: Commissioned; not externally peer reviewed.