[On the overestimation of the benefit of prevention].

BACKGROUND

Both pharmacological and non-pharmacological preventive interventions can do more harm than good. Health checks target a healthy or symptomless population. This is why randomised controlled trials (RCTs) must be conducted to provide high-quality evidence for the benefit of an intervention. The present article presents examples to demonstrate that the benefit of preventive interventions is usually overestimated.

METHODS

Standard screening criteria are used to critically appraise selected preventive interventions. Screening criteria cover the disease, the test, the treatment and the whole programme including evaluation and quality assurance. Type-2 diabetes mellitus is used as an example to discuss specific criteria for preventive interventions. The current state of the evidence is outlined. The article is based primarily on systematic / Cochrane reviews of RCTs.

RESULTS

A recent Cochrane review including 16 RCTs concluded that there is no benefit of general health checks. High-quality evidence on individual components of health checks is frequently missing or inconclusive. Over the last 30 years reference values for normal blood glucose and normal blood pressure as well as treatment targets for patients with type-2 diabetes mellitus and hypertension have been repeatedly decreased though this is not supported by evidence. Recent high-quality RCTs have shown that these "hit hard and early" interventions are detrimental, particularly to those who were the primary target group. Consequently, treatment targets have again been raised and recent guidelines recommend individualisation of treatment goals taking age and comorbidities into account. Important criteria for the implementation of preventive interventions are not currently met. With regard to type-2 diabetes uncertainties remain as to the clinical significance of pre-diabetes, the treatment of pre-diabetes and early treatment of diabetes, the screening tests, and target groups. The ADDITION study was unable to prove the benefit of a diabetes screening. Intensive lifestyle interventions may result in modest reductions of body weight and fewer diabetes diagnoses. However, the clinical relevance of the underlying metabolic changes is doubtful or even negligible. After almost 10 years, the Look AHEAD study has been terminated early due to the ineffectiveness of its intensive lifestyle interventions and the lack of hope that the study will succeed in demonstrating any benefit on the primary cardiovascular endpoints during the originally planned study period of another 3 years.

CONCLUSION

The benefit of prevention is overestimated whereas harm is underestimated. It is most unlikely that medical preventive interventions targeting individual behaviour changes will result in better health for our population.