General health checks in adults for reducing morbidity and mortality from disease: Cochrane systematic review and meta-analysis

BMJ 2012; 345 doi: http://dx.doi.org/10.1136/bmj.e7191 (Published 20 November 2012)
Cite this as: BMJ 2012;345:e7191

Recent rapid responses

Rapid responses are electronic letters to the editor. They enable our users to debate issues raised in articles published on bmj.com. Although a selection of rapid responses will be included as edited readers' letters in the weekly print issue of the BMJ, their first appearance online means that they are published articles. If you need the url (web address) of an individual response, perhaps for citation purposes, simply click on the response headline and copy the url from the browser window.

Displaying 1-10 out of 10 published

The question must be not that health checks reduce mortality and morbidity but whether early detection of conditions so detected like hypertension, diabetes or hyperlipidemia and its management leads to reduction in mortality and morbidity compared to no health checks. That depends on how easily control group accessed health care and how often such checks are done routinely in that population. Perhaps the study should look at how many in control group had the checks routinely done which was offered in health check.

Competing interests: None declared

Kishore SHETTY, General Practitioner

Pallion Health Centre, Sunderland SR4 7XF

Click to like:

As Hozawa et al. have pointed out, estimating the effect of health check-ups on mortality is difficult because participants who attend annual health check-ups are more likely to have healthy lifestyles or health conditions than those who do not. (1) The authors’ study and their reviewed previous studies did not adjust for these lifestyles or health conditions so that the authors concluded that general health checks did not reduce morbidity or mortality, neither overall nor for cardiovascular or cancer causes. However, recent Japanese studies reported the relationship between health check-up and risk of mortality adjusted for possible confounding factors. (1), (2), (3) Khan et al. investigated 3185 men and women for about 15 years and found an inverse relationship between screening and all-cause mortality. (2) Ikeda et al. also investigated 68,825 men and women for 10 years and reported an inverse relationship between participation in screening and all-cause and cause-specific mortality only in women. (3) In fact, neither of the studies adjusted for information on diet, self-rated health, nor physical function, which should also be different between participants who used health check-up (screenees) and those who did not (non-screenees). Hozawa et al. (2010) followed 48,775 Japanese National Health Insurance beneficiaries aged 40–79 years since 1994 for 11 years, and showed that mortality rates are lower among screenees than non-screenees in Japanese health check-ups when propensity matched cohort analyses were used for adjusting confounding factors. (1)

References

1 Hozawa A, Kuriyama S, Watanabe I, Kakizaki M, Ohmori-Matsuda K, Sone T, Nagai M, Sugawara Y, Nitta A, Li Q, Ohkubo T, Murakami Y, Tsuji I. 2010. Participation in health check-ups and mortality using propensity score matched cohort analyses.Prev Med 2010; 51(5):397-402.

2. Khan MM, Goto R, Sonoda T, Sakauchi F, Washio M, Kobayashi K, Mori M. 2004. Impact of health education and screening over all-cause mortality in Japan: evidence from a cohort study during 1984-2002. Prev Med 2004; 38(6):786-92.

3. Ikeda A, Iso H, Toyoshima H, Fujino Y, Mizoue T, Yoshimura T, Inaba Y, Tamakoshi A; JACC Study Group. 2005. The relationships between interest for and participation in health screening and risk of mortality: the Japan Collaborative Cohort Study.Prev Med 2005; 41(3-4):767-71.

Competing interests: None declared

Timon Cheng-Yi Liu, professor

Lab Laser Sports Medicine, South China Normal Univ, University Town, Guangzhou, GD 510006, China

Click to like:

We disagree with Johannes G Scholl on most of the points he raised. He seems to implicitly assume that health checks must necessarily be more effective today than when the trials were performed; that today’s medications must be more effective and less harmful in screened populations than previous treatment options; that changes in surrogate markers such as cardiac risk factors due to health checks must necessarily lead to improved health; that lower thresholds for cholesterol and hypertension today must improve the usefulness of population based health checks; that lifestyle counselling and coaching must be beneficial; and that dietary advice will necessarily lead to benefit. Here are our reasons for disagreeing.

First, general practitioners likely performed less testing and used fewer preventive interventions (medications or lifestyle counselling) as part of their usual care at the time when the large trials were performed that we included in our review. Thus, the potential for finding unrecognised diseases and risk factors was likely greater then. This would have made it more likely to find a benefit in the old trials than if similar trials had been performed today.

Second, new drugs are not always better. An example is rosiglitazone which was on the market for 10 years before being withdrawn because it causes heart disease instead of preventing it [1]. Given the level of secrecy surrounding the industry's trials, and the common manipulations with the results, there are more medications with important unrecognised harms. The field of diabetes drugs is particularly concerning, since drugs have been approved on the basis of improvements in surrogate markers, rather than clinically relevant outcomes [2].

Third, surrogate markers are more susceptible to bias than clinical events and do not capture harmful effects. A recently published trial and meta-analysis of a dietary intervention for reducing cardiovascular disease showed beneficial effects on cholesterol, but an increase in total and cardiovascular mortality[3]. Other examples abound, e.g. torcetrapib increased HDL cholesterol and reduced LDL cholesterol, but also increased total mortality by 58% [4].

Fourth, thresholds for treating risk factors are lower today, but it is not a given that this is beneficial. For example, treatment of mild hypertension has not been documented to be effective, while the risk of harm is certain [5].

Fifth, our review included a pre-specified subgroup analysis of trials that offered lifestyle interventions, but we found no benefit. This is not surprising, as multiple risk factor interventions appear to be ineffective in general populations, in contrast to high-risk hypertension or diabetes populations [6].

Sixth, important parts of dietary advice broadly communicated over the past 50 years may have been wrong [3,7,8], and there is still plenty of uncertainty about what constitutes an optimal diet and BMI [8–10].

Johannes G Schroll represents the company Prevention First, which appears to sell health checks to healthy Germans. Such companies sell a product on the promise that it is beneficial, but the promise is unsupported by the randomised trials.

Harald Lipman raises the question whether risk reduction interventions were used following the health checks, as this would be necessary for health checks to provide benefit. Obviously, the results of the health check were reacted upon, as that is the point of health checks. In the Cochrane review, we described in detail the type of follow-up used in each trial.

Stephan Imsfeld and Ester Ramseier (SE&ER) also highlight the age of the trials and we refer to our reply to Johannes G Scholl and point out that most advances in the field of medicine apply to treatment of the sick, not interventions directed at the healthy.

SE&ER argue that we should not have excluded trials using only one test. However, such trials are irrelevant to our research question and would have increased heterogeneity to the point where conclusions about general health checks could not have been made. Further, we are not aware of a single long-term trial with relevant outcomes investigating the individual components that constitute the core of most health checks, e.g. screening for hypertension, cholesterol, height and weight. We believe that our review makes it clear that systematic offers of bundles of tests in health checks do not seem to be better than usual care, and we concluded that future research should be directed at the individual components of health checks separately, since harmful components could have balanced out beneficial ones in the trials we reviewed.

Regarding heterogeneity, we agree with SE&ER that statistical heterogeneity should not be confused with clinical and methodological heterogeneity, and we described differences between trials in detail for this reason. However, contrary to what SE&ER assert, we do not suffer from the "misconception that statistical figures are more important than common sense." In the article we stated: "Another possible limitation is the clinical and methodological heterogeneity among the included trials, although the results were generally consistent for the frequently reported outcomes." So, although there were variations in the type of health checks used, the lack of effects was a consistent finding. "Common sense" cannot be used to discard our findings.

SE&ER regret that we excluded trials in people over 65. Our reasons were made clear in the article: Geriatric trials usually included several interventions in addition to a health check, e.g. fall prevention, medication review and immunisations. Also, health issues and priorities are sometimes different in older people. For example, a trial of a complex preventive intervention directed against elderly people found that the intervention increased mortality. This was attributed to the fact that the intervention included an autonomy enhancing effort, in part through support for making advance directives[11]. We discussed a review of complex interventions targeting elderly persons, which found small beneficial effects on the risk of not living at home, nursing home admissions, falls, and physical function, i.e. outcomes of particular relevance to elderly. That review did not find an effect on mortality[12].

Regarding the event rates, they are indeed small compared to trials in patients, but given the power in the meta-analyses we would certainly expect to have seen at least a hint of effectiveness if the intervention truly was beneficial. The low event rates, even in trials with long follow-up, show that the absolute benefit of health checks must be very small, if it exists. This increases the need for strong evidence, using reliable outcome measures.

Tomoyuki Kawada also points to the importance of what happens after abnormalities are detected at health checks. We refer to our reply to Harald Lipman, but would also like to add that health checks involve screening healthy people, which brings a particular ethical perspective to it [13]. The medical practice of health checks for healthy adults was historically established in the absence of supporting data. Those who offer this intervention to healthy people have not lifted the burden of proof for effectiveness and safety, as they should. We believe this is unethical, as health checks may cause harm.

References

1. Lehman R, Yudkin JS, Krumholz H. Licensing drugs for diabetes. BMJ 2010;341:c4805.
2. Nissen SE. Cardiovascular effects of diabetes drugs: emerging from the dark ages. Annals of internal medicine 2012;157:671–2.
3. Ramsden CE, Zamora D, Leelarthaepin B, et al. Use of dietary linoleic acid for secondary prevention of coronary heart disease and death: evaluation of recovered data from the Sydney Diet Heart Study and updated meta-analysis. BMJ 2013;346:e8707.
4. Barter PJ, Caulfield M, Eriksson M, et al. Effects of torcetrapib in patients at high risk for coronary events. The New England journal of medicine 2007;357:2109–22.
5. Diao D, Wright JM, Cundiff DK, et al. Pharmacotherapy for mild hypertension. Cochrane database of systematic reviews 2012;8:CD006742.
6. Ebrahim S, Taylor F, Ward K, et al. Multiple risk factor interventions for primary prevention of coronary heart disease. Cochrane database of systematic reviews 2011;1:CD001561.
7. Hoenselaar R. Saturated fat and cardiovascular disease: the discrepancy between the scientific literature and dietary advice. Nutrition 2012;28:118–23.
8. Siri-Tarino PW, Sun Q, Hu FB, et al. Saturated fat, carbohydrate, and cardiovascular disease. The American journal of clinical nutrition 2010;91:502–9.
9. Hooper L, Summerbell CD, Thompson R, et al. Reduced or modified dietary fat for preventing cardiovascular disease. Cochrane database of systematic reviews 2012;5:CD002137.
10. Flegal KM, Kit BK, Orpana H, et al. Association of all-cause mortality with overweight and obesity using standard body mass index categories: a systematic review and meta-analysis. JAMA  2013;309:71–82.
11. Patrick DL, Beresford SA, Ehreth J, et al. Interpreting excess mortality in a prevention trial for older adults. International journal of epidemiology 1995;24 Suppl 1:S27–33.
12. Beswick AD, Rees K, Dieppe P, et al. Complex interventions to improve physical function and maintain independent living in elderly people: a systematic review and meta-analysis. Lancet 2008;371:725–35.
13. Cochrane AL, Holland WW. Validation of screening procedures. British medical bulletin 1971;27:3–8.

Competing interests: None declared

Lasse T Krogsbøll, Doctor

Karsten Juhl Jørgensen, Peter C. Gøtzsche

The Nordic Cochrane Centre, Rigshospitalet, Blegdamsvej 9, 2100 København Ø, Denmark

Click to like:

Recently, Krogsboll and colleagues conducted systematic review and meta-analysis on the efficiency of general health checkups [1]. Their review presented the lack of the effect of health checkups on morbidity and mortality based on randomized controlled trials. The main outcomes were total mortality, cardiovascular mortality and cancer mortality. The morbidity, hospitalization, disability, worry, physician visits, or absence from works were also assessed with no significant advantage of attending general health checkups.

But the uselessness of general health checkups in their paper should be handled with caution. The reviewed studies adopted some different procedures to detect health status in the past four decades, and there have been many changes in medical screening approach in this long period of time [2]. The medical costs of further diagnostic testing and follow-up costs thereafter is important for evaluating the effectiveness of general health checkups, and multi-dimensional approach is indispensable for the cost-effectiveness (benefit) analysis.

There is a gap among perception, awareness and attitude for health outcomes after attending general health checkups. Although the subjects are aware of the link between hypertension and stroke, they do not fully appreciate the consequences of uncontrolled hypertension [3]. Perception of health problems by attending the general health checkups does not always lead to the awareness or attitude for the recovery of health. Desirable healthy activity by the modification of their lifestyles is also required to make lowering of morbidity or mortality. On this point, health education is a very important activity to improve the levels of awareness and attitude for health outcomes. Existence of several steps from general health checkups to morbidity or mortality should be considered before making their conclusion.

REFERENCES
1. Krogsboll LT, Jorgensen KJ, Gronhoj Larsen C, Gotzsche PC. General health checks in adults for reducing morbidity and mortality from disease: Cochrane systematic review and meta-analysis. BMJ 2012;345:e7191.
2. Prochazka AV, Caverly T. General Health Checks in Adults for Reducing Morbidity and Mortality From Disease: Summary Review of Primary Findings and Conclusions. JAMA Intern Med 2013 doi: 10.1001/jamainterned.2013.3187.
3. Volpe M, Dedhiya SD. Physicians, patients, and public knowledge and perception regarding hypertension and stroke: a review of survey studies. Curr Med Res Opin 2006;22(7):1319-1330.

Competing interests: None declared

Tomoyuki Kawada, Professor

Niipon Medical School, Sendagi, Bunkyo-ku, Japan

Click to like:

This is an addendum to my 17 December 2012 rapid response.

Here’s another reason why I believe your online polls deserve scrutiny: the media uses them. In his latest newspaper column, Peter Dawson, a pharmacist, cites the results of a BMJ poll thus: “A British Medical Journal survey revealed 58% of respondents in favour of such changes.” [1] He says nothing more about the “survey”.

The online version of his article has a hyperlink on the word “survey”, which is for a BMJ blog by Tessa Richards. [2] There she describes the results of “a recent BMJ online poll”, without warning of its self-selecting nature or specifying the exact question. A reader of the print version of Dawson’s column is totally blind on the “survey”: he/she only knows its provenance.

It would be unfortunate if public trust in the BMJ brand led the reader to over-interpret and/or misinterpret the poll results Dawson quotes.

1. Dawson P. Better access to patient records is essential to make the NHS a safer place. Guardian 2012 Dec 19. www.guardian.co.uk/society/2012/dec/18/better-access-patient-records-nhs....
2. Richards T. Personal information empowers and its shift to the people makes sense. BMJ blog 2012 Aug 20. blogs.bmj.com/bmj/2012/08/20/tessa-richards-personal-information-empowers-and-its-shift-to-the-people-makes-sense.

Competing interests: None declared

Alex C W May, independent researcher

N/A, Manchester M13 9DP, UK

Click to like:

This is NOT a rapid response to the research paper. Rather, it is a rapid response to the linked poll on bmj.com, “Is there value in conducting periodic health checks?” This is the only way to respond to the poll (I have asked).

The poll question is almost meaningless. The checks certainly have value – importance and money – for those performing them, if they receive payment for the activity. So, ‘value’ for whom, exactly? And what do you mean by ‘value’, anyway? Further, we surely need to consider how healthcare and its financing are organised in the relevant country. What are the priorities of their health system? Etc.

This is not nitpicking: in any poll, the answer you receive depends on the question you ask. There is another reason why I don’t know how to interpret the results. The participants selected themselves. Each week, you run such polls on the home page, reporting the results there and in the following week’s print issue. There is a poll archive. The prominence the BMJ gives the self-selecting polls means they deserve scrutiny. What is their intended purpose?

Competing interests: None declared

Alex C W May, independent researcher

N/A, Manchester M13 9DP, UK

Click to like:

We have read with interest the recent study by Krogsbøll and colleagues (1), who have conducted a review and meta-analysis on morbidity and mortality benefits of general health checks. They conclude that general health checks do not reduce morbidity or mortality, neither overall nor for cardiovascular or cancer causes.

Even though the study was set up as a Cochrane systematic review and meta-analysis and could have helped tackling this important question, it has some drawbacks that we believe seriously threaten the validity of the results and conclusions obtained. First, the average year of the studies used in this review is 1976, i.e. 36 years old, with respective average study start dates around 1968. The medical discipline has undoubtedly made at least some progress since that time.

Second, by excluding single test randomized trials, the authors have deliberately excluded some of the best evidence available, and focused on the combined analysis of test batteries used in the selected studies, irrespective of major differences in design. In addition, adequate screening for diabetes was only performed in three of the 14 trials, mammography in two, and only one trial performed a sigmoidoscopy for colon cancer screening. Omitting such important factors as well as smoking prevention in present day screening trial designs seems unrealistic.

Third, the authors mention that there was no heterogeneity in the analysis of total mortality with no effect of general health checks on total mortality. Though this statement might be statistically correct, it reflects a common misconception that statistical figures are more important than common sense. According to Table 3, the heterogeneity of the chosen studies for this meta-analysis is overwhelming. To name just a few additional sources of heterogeneity, eight of the 14 studies did not include a clinical examination, four did not even include personal history. Only five studies did perform some kind of cancer screening, seven studies did not screen for diabetes at all. Hence the statistical test applied might not indicate heterogeneity, but it should not mislead us to overlook the obvious.

Fourth, the review excluded trials targeting older people and trials which only enrolled people aged > 65. This resulted in a relatively young population of a crudely estimated average age of about 50 years with a median follow up period of 9 years in total and 10.4 years in cause-specific mortality. Accordingly, the event rates were as low as 7% for total and 3.7% for cause-specific mortality, leaving 93% and 96.3% with unknown outcomes. The factors consistently examined across the trials were blood pressure, cholesterol, height and weight, which today are well accepted risk factors for long term, i.e. life term survival, but are not considered as strong predictors for short to mid term survival at these younger ages.

Even though information on performance of health checks is of outmost interest especially in an economical context, we have serious doubts that the study, due to the above mentioned shortcomings, will allow for clarification on this issue. Despite the modern methodology of a cochrane review applied, the authors unfortunately describe fairly historical approaches rather than contemporary medical practice.

Competing interests: None declared

Stephan Imfeld, Vascular Physician

Esther Ramseier

University Hospital Basel, Petersgraben 4, 4031 Basel

Click to like:

27 November 2012

Your editorial contains several relevant caveats in assessing the value of periodic health checks. We all know that optimally those at highest risk should be targeted, rather than undertaking blanket checks. However, that is not really the point.

What are the aims of health checks?
1. To diagnose previously undiagnosed medical problems, such as hypertension & diabetes
2. To assess risk of developing future diseases eg CVD
3. To reduce morbidity & mortality of certain diseases
4. To increase life expectancy and quality of life

Achieving morbidity & mortality reductions and increasing life expectancy & improving life quality can only be long-term aims

Appropriate treatment, of newly diagnosed ailments, when indicated, for example in the case of hypertension & diabetes and initiating measures to reduce risk, such as prophylactic statins will assist with achieving these objectives1

However in all cases measures to modify life-style risk factors, such as excessive tobacco & alcohol, obesity, inappropriate diet and inadequate physical activity must be recommended, implemented and monitored to ensure sustained improvement2

Merely performing health checks solely reveals actual or potential problems and in the absence of subsequent measures to treat or reduce the risk of developing ailments is not going to show any beneficial effect on morbidity or mortality.

Does meta-analysis of trials of periodic health checks assess whether following the health check such measures were initiated?

The value, or lack of value, of health checks can only be assessed if studies include details of measures taken to treat or prevent problems highlighted by the check.

1. Taylor F et al. Statins for the primary prevention of CVD Cochrane summaries Published online May 16 2012
2. European Guideline on CVD Prevention in Clinical Practice (version 2012) European Heart Journal 2012 33, 1635-1701

Dr Harald M Lipman
International Cardiac Healthcare & RiskFactor Modification (ICHARM)

Competing interests: Executive Director ICHARM

Harald M Lipman, Physician

International Cardiac Healthcare & RiskFactor Modification (ICHARM), The Dutch House 77a Fitzjohn's Avenue London NW3 6NY

Click to like:

The authors make several assumptions, that are misleading: They call simple screening procedures a "preventive health check". They analyze very old and old data from times where diagnostic thresholds were very different from today and where a diagnosis of hypertension or hypercholesterolemia was not followed by proper therapy. They analyze studies, in most of which there was no patient education or lifestyle coaching, no nutrition counselling or fitness measurement at all. They then transfer the null-results of such old, unuseful screening procedures onto modern health checks, which is not appropriate.

There are many examples how preventive health checks that include an individual coaching can improve the cardiovascular risk profile of patients.

In our own data we have shown a calculated 25% risk reduction for CVD, improvements in insulin resistance, fitness and BP control.

In my opinion this meta-analysis just shows, that old-fashioned screening procedures or health-checks without appropriate measures (like fitness and metabolic risk evaluation) and without an individual lifestyle coaching are ineffective.

By no means this should be interpreted as ineffectivenes of preventive health care. To the contrary: We must improve our preventive interventions and coaching procedures, because without prevention there will be a very dark future for our health care system: The demographic changes and the growing burden of chronic diseases caused by our "modern" lifestyle will otherwise bring it to its limits.

Dr. med. Johannes Scholl
Prevention First
Ruedesheim, Germany

Scholl J., Schneider M., Lifestyle modification and risk factor management for cardiovascular prevention in a workplace setting: The FIT IN LIFE – FIT ON THE JOB Study; European Heart Journal (2012) 33 (Abstract Supplement) 953

Competing interests: I am working in the field of evidence-based preventive health care.

Johannes G Scholl, Internist

Prevention First, www.preventionfirst.de

Click to like:

Do not trouble until trouble troubles you. This is very true of health checks in adults. We are living in a very materialistic society which includes Nusing Homes and hospitals. When we build nursing homes and corporate hospitals we create the need to generate income to keep the institutions viable and profitable.

One such method (could be beneficial in certain conditions) is master health check up plans which come in one package. Advertisers and medical insurance companies through their influence make the most of the educated elite, and health conscious citizens undergo such check ups. The values in clinical chemistry laboratory reports or the borderline risk elaborated by other diagnostric procedures drive the patient to go for more check ups and end up sometimes in a sojourn whose beneft is less than what has been professed.

The science literate citizen will understand that such check ups are useful in cases where there is a family history of a particular disorder such as diabetes or hypertension or a genetic disorder or stilborn baby.

The average life span of a person has dramatically increased due to more awareness about health, the establishment of Physical Exercise centres or yoga therapy or a desire by the adult to maintain health and physique.

Therefore a health check up is beneficial if taken probably with the advice of a family physician.

Competing interests: None declared

dhastagir s sheriff, Professor

Faculty of Medicine, Benghazi University, Benghazi, Libya

Click to like:

THIS WEEK'S POLL