General Health Checks in Adults for Reducing Morbidity and Mortality from Disease

Gustoimages/Science Photo Library Since the latter half of the 20th century, technological advances in modern medicine and a growing emphasis on preventive care have led to increased enthusiasm for screening in apparently healthy people. In primary care practice, the general health check (also termed periodic health evaluation or routine medical examination) is the usual mechanism used to screen asymptomatic people for disease. Although widely practiced, there is no universally accepted definition of what constitutes a general health check; screening may occur at variable frequencies and include a spectrum of diagnostic maneuvers ranging from physical examination to invasive procedures. Despite this variability, all general health checks share a common goal: to reduce morbidity and mortality by detecting disease or modifiable risk factors at an earlier stage—implicitly assuming that this will improve clinical outcomes compared with waiting until symptoms develop. Although general health checks are popular, recent debate has focused on their unintended consequences, including overdiagnosis, which is the detection of disease that (if left undetected) would not have affected the quantity or quality of an individual's life. Abnormal screening results can also lead to further investigations and accompanying risks: anxiety or psychological distress; lost income due to work absences; difficulties securing insurance; and increased healthcare costs. General health checks are not widely recommended by national expert panels. The Canadian Task Force on the Periodic Health Examination recommended against the annual general health check in 1979,1 and similar guidance from the United States Preventative Service Task Force followed in 1989.2 Instead, both bodies recommended focused health checks guided by patient‐specific risk factors. Based on available evidence of efficacy, as well as acceptability to patients and healthcare costs, the Canadian Task Force recommended a limited number of screening tests and at variable intervals.1 Recommendations were made according to the patient's age and sex, and were further qualified with a list of tests that should only be offered in high‐risk patients. Concerns about overdiagnosis notwithstanding (and despite the national guidelines), general health checks are considered by physicians and the public as both necessary and recommended.3, 4, 5, 6 In 2009, the routine general health check was one of the two most common reasons patients visited their doctor in Canada and the United States.7, 8 Japan has a national program that includes mass screening for all adults,9 and the United Kingdom has recently implemented routine screening for several major diseases in adults ages 40 to 74.10 Studies have reported both patient and physician factors as drivers of general health checks.3, 11 In one study, routine screening was viewed by patients as less of a choice and more of an obligation to one's health and family.5 Another survey identified relationship building and the potential benefits of preventive care as the major reasons why primary care physicians continued to offer general health checks.6 The much‐needed Cochrane Review by Krogsbøll and colleagues in the October 2012 issue of The Cochrane Library evaluates the benefits and harms of general health checks in asymptomatic adults.12 The review only included randomized trials studying more than one screening intervention in multiple organ systems, and excluded those enrolling only older populations (people aged over 65 years). All studies evaluated asymptomatic populations that were unselected for disease or risk factors. The interventions studied varied substantially between trials, and used a combination of tests such as health questionnaires, physical examination, blood assays, imaging, stool testing, and assessment of cardiovascular risk factors. All studies compared groups who received general health checks with groups who did not. The review is noteworthy 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. Sixteen randomized studies from primary care or community settings were included; 14 studies (182,880 participants) had outcome data available for analysis. Of the study participants offered screening, a median of 82% (range 50% to 90%) underwent screening at the first round. Nine studies reported on mortality, with a follow‐up of 4 to 22 years. The main conclusion from the review is that general health checks in the community setting do not reduce overall or disease‐specific mortality. The relative risk of death was 0.99 (95% confidence interval 0.95, 1.03), and results were consistent across studies and in a number of sensitivity and subgroup analyses. General health checks either had no effect on other outcomes (patient worry, unscheduled physician visits, hospital admissions, and absences from work), or estimates of effect were unreliable (self‐reported health, disability). Due to missing or unreliable data, the authors could not estimate costs, harms, or the use of follow‐up medications and testing as the result of screening. Screening resulted in a high proportion of abnormal test results and probably increased the number of new diagnoses, although these results were poorly reported in the included studies. Given the lack of overall benefit, this may indicate that general health checks promote overdiagnosis rather than detecting clinically relevant abnormalities. Krogsbøll et al reiterate that the potential for harm is likely to exceed the potential for benefit when screening is implemented in a population where the overall risk of an unfavorable outcome is low. The review's rigorous methods and focus on large randomized trials with long follow‐up periods tend to increase confidence in its findings. However, because the majority of the included studies were unblinded with considerable loss to follow‐up, analyses of outcomes other than death and hospitalization may be subject to bias. In addition, the ability of long‐term studies to correctly estimate the effect of screening on mortality, given improvements in diagnosis and treatment over time, have been questioned.13 This latter point is important, as nine of the 14 studies analyzed by Krogsbøll et al were initiated prior to 1980, and the most recent was initiated in 1992. However, most included studies evaluated basic screening maneuvers that have changed little in the interim, such as height, visual acuity, weight, blood pressure, blood work, physical examination, urinalysis, and questionnaires aimed at establishing risk factors. It is possible that advances in the treatment of diseases or risk factors identified by screening may have improved the benefits associated with general health checks over time. The potential impact of any such advances at the population level may be greatest for those directed at reducing cardiovascular risk: data from the forthcoming Inter99 study (a randomized trial initiated in the late 1990s and addressing this issue in approximately 60,000 patients) should help to confirm or refute this hypothesis. The results of this review are consistent with a previous systematic review by Boulware et al that also evaluated the benefits and harms of general health checks.13 The Boulware review differed in its definition of the general health check, included observational studies, and analyzed fewer randomized trials. Despite these differences, the authors of both reviews reported that general health checks had no effect on mortality, disability and hospitalizations compared with usual care. However, the two reviews drew different conclusions on the overall benefit of screening. Boulware et al did not emphasize the lack of benefit of general health checks on clinical outcomes, instead concluding that such screening is beneficial because it improves the delivery of preventative health care and decreases patient worry. The Krogsbøll review identified two trials that found no benefit on reducing worry whereas Boulware et al identified one randomized trial that found benefit. How should practitioners use the findings of Krogsbøll et al? Although available trials have limitations, there is no convincing evidence that general health checks are beneficial. Since patients who seek or are willing to undergo routine screening are generally healthier than those who are not3 (indicating that general health checks are least likely to reach those who could benefit the most), and because most people do not receive interventions that are known to be beneficial,14 general health checks do not appear to be a wise use of scarce healthcare resources. Heeding the Canadian recommendations (made more than 30 years ago) to abandon routine health checks would save money that could be better used by population‐level interventions supported by effective health policy, such as the campaigns to reduce dietary sodium in Finland and the United Kingdom.15, 16 On the other hand, this important review should not be interpreted as nihilistic; it focuses on the screening of asymptomatic people and does not apply to interventions prompted by clinical judgment or patient concern. For all studies in the review, diagnostic testing and treatments were undoubtedly offered to participants in the control groups for these reasons, and probably improved outcomes while reducing the apparent effect of the intervention. Practitioners should continue to investigate and treat patients with symptoms or clinical clues to underlying disease or its risk factors. When contemplating screening, practitioners should focus on tests that are targeted to the patient's age, sex, and specific risk factors, and that are supported by high‐quality evidence. All screening tests (general health checks or focused screenings based on age, sex, or specific risk factors) have potential for benefits and harms, so consideration of patient preferences is critical, especially for those tests where such preferences vary between individuals or where the overall benefit:harm ratio is less favorable. Future research should determine: the best way to facilitate the sometimes challenging discussions with patients about preferences, benefit, and harm; how to communicate nuanced messages about the advantages and disadvantages of screening to the public; and how to increase the uptake of beneficial targeted screening tests in primary care practice. Of course, better evidence is also needed on the best way to screen for common diseases and their risk factors—and how to optimize the benefit:harm ratio and cost effectiveness.