Re: Government’s plans for universal health checks for people aged 40-75
The editorial by Goodyear-Smith on health checks nicely summarises the uncertainties about their effects . She cites the recent report from Public Health England, "NHS Health Check: our approach to the evidence," for its criticism of our Cochrane review. This report refers to a commentary published on the website of the NHS Health Check. We respond here to the criticisms raised in both places and describe our unsuccessful attempt to have our response published.
Last autumn, we published a Cochrane review on general health checks in adults for reducing morbidity and mortality from disease [2,3]. Our review was large. It included 182,880 randomised participants and 11,940 deaths, and the median follow-up time in the meta-analyses was nine years. There was no reduction in morbidity or mortality whereas health checks may cause overdiagnosis and overtreatment. Subsequently, the NHS Diabetes and Kidney Care, in conjunction with the Department of Health, published an eBulletin, "Response to the Cochrane review," on the website of the NHS Health Check programme . It is 800 words long and contains what looks like serious criticism of our work, but as we demonstrate below, it is misleading.
We were made aware of the eBulletin when a UK general practitioner copied us on a query she made to the NHS Diabetes and Kidney Care about whether she should continue doing NHS Health Checks in light of the results of our review, and what she should tell her patients about the benefits and harms. She was referred to the eBulletin.
We sent a detailed response to the criticism to the National Director of NHS Diabetes and Kidney Care on 7th January 2013 requesting that it be published on the website alongside the criticism. Two weeks later, we asked whether our email had been received, which was confirmed. On 18th February, we asked again for a reply to our request, and on 25th February we inquired where we could file a complaint about the lack of reply. The next day, we were informed that our response would not be published.
The letter stated that the decision had already been taken by Government that NHS Health Checks will be carried out as a national priority; that “the website is not a forum for debate or discussion on the merits of conducting NHS Health Checks;” and that “there are other more appropriate places to discuss Government policy.” If that was really the case, we wonder why the Programme chose to publish its criticism of our work on its website, and not in a scientific journal, and why it did exactly what it advised against: discussed the merits of health checks on its website . As the Programme furthermore refers health professionals with questions about health checks to this information, it suggests to us that the NHS has sacrificed its own principles about evidence-based health care.
It is interesting to contrast the reactions to our review in the UK with the reactions in Denmark. In the UK, an anonymous Department of Health representative told BBC News that “The NHS Health Check programme is based on expert guidance” and the eBulletin cites a simulation study . These are hardly sources of evidence that can compete with a comprehensive systematic review of the randomised trials, which showed a clear negative result. In Denmark, systematic health checks had not yet been implemented, but they were high on the then new government's agenda. Even so, the Danish Minister of Health stated: “The analysis from the Nordic Cochrane Centre does not come as a surprise... I have put our old suggestion of systematic health checks on ice because they will not have the desired effect” .
THE MISLEADING CRITICISM
The eBulletin states that our review does not specify what constitutes a general health check, its content, or its objectives . This is not correct. It is a requirement that Cochrane reviews clearly define the interventions they examine, and we detailed this both in the peer-reviewed protocol, in our Cochrane review, and in its BMJ version [2,3].
Other comments are similarly misleading, e.g. “some interventions included relevant measures such as blood pressure and cholesterol, but not all”  gives the impression that these measures were not general features of the included studies. In actual fact, blood pressure was measured in 13 of the 14 trials (the last trial was unclear about this), and cholesterol was measured in 11 trials (unclear in one trial, and in the remaining two trials it was very likely measured).
It is also stated that there was no specification of the follow-up actions to identified abnormalities . However, we mentioned in both of our papers that some trials used follow-up by specialists or by using treatment algorithms, which likely bias results towards greater effect [2,3]. In most trials, follow-up of identified abnormalities was done by the participant's regular physician, which gives a more realistic and generalisable picture of the effect.
Our inclusion of unpublished mortality data from the OXCHECK trial is described as an “unusual approach,” which is “unlikely to be considered appropriate in other circumstances” . This reflects a disturbingly poor understanding of the fundamental principles for systematic reviews. Searching for and including unpublished outcome data is very important since negative results are less often published than positive ones and our approach is standard Cochrane methodology.
Another argument is that none of the included trials precisely match the NHS Health Check. This argument can be used to denigrate any systematic review results that are unwelcome. The overlap between the tests used in the trials in our review and those used in the NHS Health Check is large and include cornerstones of the intervention (e.g. blood pressure, cholesterol and weight). Screening for diabetes was done in seven of our trials, and these trials did not show positive effects either. A recent trial of screening people at high risk for diabetes also failed to find beneficial effects .
The eBulletin states that trials with long follow-up are necessarily older than those with short follow-up, and that this reduces the relevance of trials with long follow-up . This is essentially a peculiar argument against using trials with meaningful outcomes and long follow-up in general. Surely, old trials with clinically relevant outcomes are more important than recent trials with surrogate markers such as blood pressure and cholesterol. Furthermore, there was less opportunistic screening when the old trials were performed, and it would therefore have been easier to find an effect of health checks if there was one.
Several other points raised were largely identical to those raised in rapid responses to the BMJ version of our review. We therefore refer readers to our reply .
The NHS Health Check programme operates in clear conflict with the best available evidence and in violation of the criteria of the UK National Screening Committee: “There should be evidence from high quality randomised controlled trials that the screening programme is effective in reducing mortality or morbidity“ .
An administration's automatic defence of an existing screening programme can be viewed as a defence of its own interests and of the public funds and personal prestige that have been invested, but it threatens the very idea of evidence-based public health care, particularly when it involves censorship of an open scientific debate.
We urge the NHS Health Check Programme to start a discussion with the Government about closing the programme. The programme consumes vast resources, £332m per year , that could be used for a better purpose, e.g. on interventions with documented benefits.
Lasse T. Krogsbøll
Karsten Juhl Jørgensen
Peter C Gøtzsche
1 Goodyear-Smith F. Government's plans for universal health checks for people aged 40-75. BMJ 2013;347:f4788.
2 Krogsbøll LT, Jørgensen KJ, Grønhøj Larsen C, Gøtzsche PC. General health checks in adults for reducing morbidity and mortality from disease. Cochrane database of systematic reviews 2012;10:CD009009.
3 Krogsbøll LT, Jørgensen KJ, Grønhøj Larsen C, Gøtzsche PC. General health checks in adults for reducing morbidity and mortality from disease: Cochrane systematic review and meta-analysis. BMJ 2012;345:e7191.
4 NHS Health Check eBulletin. Response to the Cochrane Review. www.nhshealthcheck.nhs.uk/?iid=11 (accessed 9 April 2013).
5 www.bbc.co.uk/news/health-19964600 (Accessed 9 April 2013)
6 Minister: Vi har lagt helbredstjek på is. Ugeskr Læger 24 October 2012. www.ugeskriftet.dk/portal/page/portal/LAEGERDK/UGESKRIFT_FOR_LAEGER?publ... (Accessed 9 April 2013)
7 Simmons RK, Echouffo-Tcheugui JB, Sharp SJ et al. Screening for type 2 diabetes and population mortality over 10 years (ADDITION-Cambridge): a cluster-randomised controlled trial. Lancet 2012;380:1741–8.
8 Krogsbøll L, Jørgensen K, Gøtzsche P. Re: General health checks in adults for reducing morbidity and mortality from disease: Cochrane systematic review and meta-analysis. BMJ 2013;14 March. www.bmj.com/content/345/bmj.e7191/rr/636040 (accessed 9 April 2013).
9 UK national screening committee. Programme appraisal criteria. www.screening.nhs.uk/criteria (accessed 9 April 2013).
Competing interests: No competing interests