House of Commons - National Health Screening - Science and Technology Committee

National Health Screening - Science and Technology Committee Contents

3 Reviewing the evidence base

The evidence review process

13. Before a new screening programme can be introduced, it must first go through the UK National Screening Committee's (UK NSC) evidence review process. Figure 1 sets out the process and the four main steps. The purpose, according to Dr Anne Mackie, Director of Programmes, UK NSC, is to decipher whether a screening programme is "likely to do more good than harm at [a] reasonable cost".[34] This chapter examines what may trigger a policy review, and the UK NSC's approach to evaluating the evidence base for screening programmes, through a discussion of each stage of the process.

Figure 1: UK National Screening Committee policy review flow chart[35]

TRIGGERING A POLICY REVIEW

14. Figure 1 identifies two triggers for a policy review. For those screening programmes that have previously been the subject of a review and are already "on the books",[36] witnesses indicated that subsequent reviews will normally be conducted on a three to four year cyclical basis.[37] A scheduled review date, however, may be brought forward: Professor Jane Wardle, Academy of Medical Sciences, drew attention to the more "ad-hoc" nature of some reviews, noting that the "appearance of new evidence, trials or suggestions" may prompt the UK NSC to examine a policy recommendation earlier than originally planned.

15. The vast majority of policy reviews focus on whether a recommendation should be made to implement a proposed screening programme. However, the UK NSC also reviews existing programmes and considers if their delivery requires any amendments (for example altering the frequency at which people are invited to attend for screening) or if there are grounds for the programme to be withdrawn. Despite this regular review process, inconsistencies can emerge. Public Health Wales (PHW) told us that changes to the "age range and frequency" of cervical screening in England were implemented without the UK NSC first reviewing the programme and making such a recommendation.[38] Moreover, PHW noted that the changes in England did not prompt the UK NSC to initiate a review of cervical screening or provide advice to Wales, Northern Ireland and Scotland as to whether they should follow suit.[39] According to PHW, "the UK NSC did not examine the evidence for the age range and frequency of cervical screening until 2012, nine years after the English NHSCSP [NHS Cervical Screening Programme] had changed its policy".[40] As well as producing differences in the delivery of programmes across the UK, PHW indicated that since "Welsh Government Policy is based on UK NSC advice" the absence of that advice "can lead to uncertainty".[41]

16. Dr Anne Mackie, Director of Programmes, UK NSC, clarified that this particular scenario had occurred because:

Whether changes to the delivery of a programme in one country should trigger a UK NSC policy review has, according to Dr Mackie, been resolved: "now, any big change in one of the existing cancer screening programmes or in other programmes […] we would look to bring to the UK NSC".[44]

17. We recognise that the devolved nations have power over public health in their respective territories. However, significant amendments to the delivery of screening programmes by a single nation within the UK (in the absence of a formal recommendation from the UK National Screening Committee (UK NSC)) risk undermining the UK NSC's authority as the body advising all four nations on screening policy. It also generates confusion and uncertainty about current best practice.

18. We welcome the UK National Screening Committee's (UK NSC) decision to ensure that any "big change" to an existing screening programme made by one, or more, of the four nations would now prompt the UK NSC to conduct an evidence review and issue a formal recommendation. We recommend that the UK NSC clarifies in its response to this report what constitutes a "big change" to an existing screening programme that would automatically trigger a UK-wide review and policy recommendation. This information should be made available on the UK NSC's website.

STAGE 1: STAKEHOLDER IDENTIFICATION

19. In some instances an evidence review may be prompted by a request from a stakeholder group. Dr Sian Taylor-Phillips, Warwick Medical School, stated that the UK NSC "say that they will review anything that a significant stakeholder—that is, a stakeholder representing a significant community of people—recommends and can provide a case that it might meet the NSC criteria".[45] According to the UK NSC, the stakeholder identification process is "based on the Single Technology Appraisal process guide" developed by the National Institute for Health and Care Excellence (NICE).[46] Professor David Walker, Chair, UK NSC, told us that the Committee has "a whole range of stakeholders, […] clinicians, patient groups, charities and […] Members of Parliament", any one of which can "ask us to look at any programme at any time".[47]

20. The Government stated that stakeholders were "involved at every stage" of the UK NSC policy review process while Public Health England (PHE) identified stakeholders as having a particular role during the external review and consultation stage when a detailed, draft report on screening for the condition being considered is shared with "expert stakeholders and the public to consult on for a period of three months".[48] According to the UK NSC, the draft report is also made available on its website so that "anyone, including individuals or groups not previously identified as stakeholders" can "provide their feedback".[49] Cancer Research UK stated that, in its experience, this process had "been robust and allowed for external input from a range of different stakeholders".[50]

21. However, patient advocacy groups for newborn screening reported a different experience. Children Living with Inherited Metabolic Diseases (Climb) stated that "patient organisations should be equal and active partners in the health screening decision making processes", but went on to identify "a distinct reluctance on behalf of the UK NSC to incorporate specialist patient group involvement within their structure" and considered the UK NSC's engagement with stakeholders to be "at best lackadaisical in this area".[51] The Save Babies Through Screening Foundation UK stated that the UK NSC's "existing stakeholder list and […] method of consultation with stakeholders [was] quite poor".[52] The Foundation added that while the ability of stakeholders to request that a policy review take place "may appear to be involving and engaging, […] very few can attempt the process without the support and time investment of health professionals".[53]

22. If it is to be effective and trusted, the UK National Screening Committee (UK NSC) must be open to a plurality of perspectives when reviewing the evidence base for its policies. We are satisfied that efforts continue to be made to consult with stakeholders and note that the UK NSC is currently producing updated guidance for stakeholders on "engaging with its policy review process". Engagement, however, should be a two-way process. In addition to being transparent and opening up its policy review process to external input and scrutiny, it is vital that the UK NSC proactively looks beyond traditional, large stakeholder groups and seeks to engage with those smaller—often condition-specific—groups especially where they offer scientific insight. We recommend that the UK National Screening Committee, in its response to this report, details how it will proactively engage with a broader range of stakeholders.

STAGES 2 AND 3: KNOWLEDGE UPDATE / EXTERNAL REVIEW AND CONSULTATION

23. The second stage of the policy review process—the knowledge update—is used to determine whether a full, external review of a particular screening policy is required. According to Dr Sian Taylor-Phillips, Warwick Medical School, it is a "smaller review where they look at, 'Is there any big new evidence in this area? Is this going to be a really interesting and important topic?'"[54] If a further, external review is deemed necessary, Dr Taylor-Phillips suggested that the UK NSC "might farm [it] out to a university" to conduct.[55] This is broadly in line with the information provided on the UK NSC's website which states that external reviews are "carried out by a recognised national expert or academic institution in the field, as identified by the UK NSC Director of Programmes".[56]

24. Group B Strep Support (GBSS) pointed to differences between the theory and practice of conducting evidence reviews of screening programmes. A systematic review of the policymaking processes applied to formulate advice on health screening decisions, conducted by academics at Warwick Medical School, found that the UK "was like most other countries" in using "systematic reviewing to synthesise evidence".[57] This, however, had not been the experience of GBSS: they told us that antenatal screening for Group B Streptococcus "was not carried out as a systematic review" and that this had been "confirmed" to them "in writing by the [UK] NSC".[58] While not commenting on this specific example, Dr Anne Mackie, Director of Programmes, UK NSC, clarified that, in general, the evidence "is brought together by a variety of external organisations" during an external review, adding that "some of that is done in a systematic reviewing way, and some in a literature synthesis way".[59] Literature synthesis differs from a systematic review in several ways. A systematic review typically involves a detailed, replicable search strategy that aims to identify, appraise and summarise all relevant studies (usually primary research) on a particular topic. Literature synthesis, in contrast, tends not to rely on a systematic search of the literature but focuses on a subset of studies on the topic area, usually based on availability or author selection.[60]

25. The methods used to establish the quality of the evidence were a further point of contention. Warwick Medical School noted that "some countries" use "standardised procedures for appraising the quality of evidence for the evidence review"[61]: the US Preventative Services Task Force[62], for example, assesses "the quality of individual studies using objective criteria"[63] while other countries use the methods of GRADE (Grading of Recommendations Assessment, Development and Evaluation).[64] However, Warwick Medical School stated that, in the UK, the steps were "tailored to each review"[65]: according to Dr Anne Mackie, Director of Programmes, UK NSC, "the reviewers are expert at saying what the quality of the evidence is" and "only bring together good quality, peer-reviewed evidence" as part of the external review.[66] HealthWatch[67] told us that "the quality of evidence available should be subject to greater scrutiny" and suggested that "Cochrane Library systematic reviews" represented "the best available source of quality unbiased information".[68]

26. GBSS expressed further concerns about the way that the Group B Streptococcus external review had been reported and noted that it did not meet the "PRISMA checklist requirements".[69] Warwick Medical School acknowledged that it had not evaluated the "usefulness" of making changes to the methods the UK currently uses when "synthesising the evidence for screening programmes" and suggested tools like GRADE and PRISMA would "need to be carefully evaluated for their applicability in the UK context".[70] In her 2013 Annual Report, the Chief Medical Officer for England, Dame Sally Davies, identified the characteristics of "robust reviews", stating that they should:

In the same report, Dame Sally went on to highlight the "PRISMA statement" and the "Cochrane Collaboration" as two of the "commonly accepted methods for the production of unbiased and transparent reviews".[72]

27. We consider the consistent conduct and reporting of systematic reviews to high, well-established standards to be of great importance. We recommend that the UK National Screening Committee (UK NSC) draw on established protocols—such as the "Cochrane Handbook for Systematic Reviews of Interventions"—to standardise the steps within, and the reporting of, each systematic review of a screening programme.

Criteria for appraising the viability, effectiveness and appropriateness of a screening programme

28. When conducting an external review, the UK NSC stipulates that the reviewer(s) must consider the evidence base against its twenty-two criteria for assessing a new programme and, in the resulting report, state whether the proposed programme meets each of those criteria (see Annex).[73] Researchers from Warwick Medical School described this approach as standard, with "most countries assess[ing] the evidence collected against health screening criteria".[74] The criteria cover the condition, the test, the treatment options and the effectiveness and acceptability of the screening programme. According to the UK NSC's first annual report, they are based on Wilson and Jungner's "classic criteria" first published in 1968 to guide the selection of conditions that would be suitable for screening.[75]

29. There were divergent views as to whether the UK NSC's criteria remained fit for purpose. Professor Michael Baum, Advocates for Honesty and Transparency in Breast Screening, noted that the criteria had "stood the test of time" but was open to taking "a fresh look at them"while HealthWatch stated that criteria "remain valid to this day" and urged "their continued use".[76] The PHG Foundation, however, stated that the UK NSC criteria were developed in the context of "the big common conditions such as breast cancer and cervical cancer" and questioned their suitability for appraising new genetic screening programmes.[77] According to the PHG Foundation, proposals for genetic screening programmes "bring complexities to the underlying evidence base" and raise "a breadth of ethical, legal and social" considerations, including questions about "informed consent, informed choice and safeguarding autonomy".[78] The PHG Foundation therefore recommended that there should be "a review of the suitability of the screening criteria for rare inherited and other genetic conditions" and also suggested "setting up a standing group" to advise on the "ethical, legal and social issues" raised by individual proposals.[79] Dr Anne Mackie, Director of Programmes, UK NSC, was open to these suggestions and said that both points were being considered as part of "the current UK NSC consultation".[80]

30. The need for additional clarity regarding how the criteria are evaluated and interpreted was also raised during the inquiry. Cancer Research UK identified the criteria as an "area of concern" on the grounds that they "can be difficult to interpret" and "can lead to controversy".[81] It singled out item 15 from the UK NSC criteria (that the benefit from the screening programme should outweigh the physical and psychological harm) as posing distinct difficulties, stating that "the magnitude of benefits and harms can be quite difficult to define, and compare, in practice".[82] Jessica Kirby, Cancer Research UK, told the Committee that while the criteria were asking the right questions, "it can be difficult sometimes to provide an objective and very clear answer to some of them".[83] For example, there was disagreement as to how the benefits of screening should be measured. Professor Susan Bewley, King's College London, stated that "the proper test of these screening programmes is [reducing] 'all-cause death'" but Professor Jane Wardle, Academy of Medical Sciences, stated that "it would be impossible to argue that our outcome on the positive side should be all-cause mortality, just because of the practical difficulty—nay, impossibility—of asking that kind of question".[84]

31. To improve the overall transparency of the evidence review process Warwick Medical School noted that adding "statements explicitly onto the UK criteria" could clarify the "deliberations and evidence taken into account" by the UK NSC "when judging screening programmes".[85] It went on to suggest that the UK NSC "consider uploading a detailed manual on the UK NSC website, detailing how they collect and assess evidence in the policymaking process".[86] Jessica Kirby, Cancer Research UK, agreed, adding that the UK NSC could "provide a bit of guidance around how evidence will be used and interpreted within the context of some of the criteria".[87]

32. This type of guidance has previously been available: the NSC Handbook of Population Screening Programmes ("the Handbook") was first published in 1998 and outlined the "questions that it [the UK NSC] requires [to be] answered when considering a screening programme".[88] The Handbook stated that the answers to the questions could be compared to the UK NSC's criteria and that this would "enable an objective assessment to be made on the balance of benefit to harm to cost for any particular programme".[89] The expectation, as stated at the front of the Handbook, was that it would "be updated on at least an annual basis".[90] When asked why the Handbook had not been updated and was not available on the UK NSC's website, Dr Anne Mackie, Director of Programmes, UK NSC, replied that "1998 is an awfully long time ago, and things move on"[91], adding that the UK NSC has "a methods process in development" that describes "fairly carefully how we go about our process".[92]

33. We note that the Independent Review of the UK National Screening Committee (UK NSC) is currently examining if the existing criteria for appraising the viability, effectiveness and appropriateness of a screening programme need strengthening or amending to take into account the complexities arising from genetic screening. It is also important that the Independent Panel considers if the evaluation of evidence against these criteria is conducted in a rigorous, transparent and consistent manner. Since the UK NSC does not use the same external reviewer for each review, and given the potential for differences in interpretation, we consider it essential that the UK NSC publishes clear guidance on how it assesses the evidence base against its criteria.

34. We recommend that the UK National Screening Committee publish a revised version of its 1998 Handbook to clarify and add detail to how the UK NSC evaluates the evidence base against its twenty-two criteria. This should be made available on its website no later than March 2015.

STAGE 4: UK NATIONAL SCREENING COMMITTEE DECISION

35. The UK NSC meets three times a year to "review current decisions and make recommendations on screening practices".[93] Though the minutes of each UK NSC meeting are published on its website,[94] it is not clear from these minutes what procedures are used by the UK NSC to ensure that its decisions are robustly and fairly reached. The guidance provided in the Government Office for Science Code of Practice for Scientific Advisory Committees (CoPSAC) suggests that Scientific Advisory Committees "should agree on the mechanisms by which the committee is to reach its final position or advice" and that "open and frank discussion should be encouraged".[95]

36. Dr Sian Taylor-Phillips, Warwick Medical School, told us that her "understanding of how it works in practice is that you would look at all of the NSC criteria, and for it to be implementable as a screening programme it would have to reach a certain level for every single criterion".[96] However, when asked if the UK NSC combines and scores the criteria to reach a decision, Dr Anne Mackie, Director of Programmes, UK NSC, confirmed that it does "not score".[97] According to the PHG Foundation, the "criteria are subjective; there are no clear cut-off points" and there is "frequently a need to trade-off between them".[98] The preamble to a survey that forms part of the Independent Review of the UK National Screening Committee explained that scoring is not used because "not all the criteria can be tested by scientific enquiry and in some cases (comparative research for very rare diseases for example) cannot be fulfilled". Instead, it states that the UK NSC "brings judgement to bear using the scientific literature, expertise, experience and the views of the public in making a recommendation".[99]

37. Jessica Kirby, Cancer Researcher UK, suggested that if there were "considerations" that the UK NSC was "using to make these judgments—then it would be good to have knowledge" of them.[100] When asked if, in the absence of "a scoring process", the UK NSC's decision was based on "subjective assessments", Dr Mackie agreed that "inevitably, some bits of it will be so".[101] She added that this was why involving "as many people as possible" in the policy review process was important.[102] The Minister told us that she had "no reason to think" that she was receiving anything other than consistent advice about screening via the current policy review process but noted "the different contexts in which programmes are offered".[103] The Minister went on to question if it was "possible to achieve perfect consistency because of the very different life stages and differently designed programmes, from the population adult programmes right down to the newborn".[104]

38. Any evidence review process must be flexible enough to accommodate the wide range of screening programmes the UK National Screening Committee (UK NSC) examines and some subjective judgements will be made. However, it is currently unclear what procedures the UK NSC has for reaching decisions about whether to recommend a programme. In line with the guidance outlined in the Code of Practice for Scientific Advisory Committees, we recommend that the UK National Screening Committee formally agree, and make public, the procedural mechanism by which it will reach decisions and recommendations.

Policy before evidence?

39. The UK National Screening Committee (UK NSC) is clear that there "should be evidence from high quality Randomised Controlled Trials (RCTs) that the screening programme is effective in reducing mortality or morbidity" before a systematic, population-based screening programme is introduced.[105] During the course of the inquiry, we were made aware of two programmes where screening policy appears to have been made, and in one instance implemented, in advance of data from RCTs becoming available.

NHS HEALTH CHECK PROGRAMME

40. The NHS Health Check programme was described by Public Health England (PHE) as a "national risk assessment, risk reduction and risk management programme"[106] that aims to help prevent heart disease, stroke, diabetes, kidney disease and certain types of dementia through inviting everyone between the ages of 40 and 74, who has not already been diagnosed with one of those conditions, to have a health check every 5 years to assess their risk.[107] Roll-out of the programme in England formally began in 2009. Under the Local Authorities Regulations 2013,[108] local authorities are mandated to offer a health check to every eligible person in their area, with PHE providing "oversight and implementation support".[109] There is some evidence to indicate that the programme has had an impact on prescribing behaviour. A study examining the relationship between the uptake of the NHS Health Check Programme and the prescription of statins (medicines that can help lower the level of low-density lipoprotein (LDL) cholesterol in the blood) found that 19.4% of patients with a high risk of cardiovascular disease (CVD)[110] were prescribed statins before the introduction of the Health Check, while 43.1% of high risk patients were prescribed statins after the Health Check was introduced.[111] When asked about conflicting reports in the press regarding the safety of statins, Professor David Walker, Chair, UK NSC, pointed to "a scientific consensus that statins are valuable for people who are considered to be high risk" but acknowledged that there is a "serious debate at the moment" about "where that cut-off should be".[112]

41. Several witnesses raised questions about the evidence base for the NHS Health Check programme. Professor Jane Wardle, Academy of Medical Sciences, told us that the programme was not "based on rigorous randomised controlled trial data"[113] while Dr McCartney, a GP from Glasgow, stated that the programme was "launched without an evidence base".[114] The Academy of Medical Sciences also highlighted concerns about the evidence for systematic health checks in general. Pointing to the findings of the Inter99 trial[115] published in June 2014, the Academy noted that the trial found "no reduction in mortality from ischaemic heart disease, stroke or total mortality […] in those who participated in the screening and lifestyle counselling, compared to the control population".[116] Professor Walker was clear that the Health Check programme had "not been through the NSC process" and that, because it was "was implemented through a different route", the UK NSC had not conducted "the rigorous evidence review that [it] would normally do before implementation of this kind of programme".[117]

42. A report published by PHE in July 2013 acknowledged that the NHS Health Check programme was "being implemented in the absence of direct randomised controlled trial evidence" but maintained that "the existing relevant evidence" provided "compelling support for the programme".[118] Speaking to the Health Committee about NHS Health Check in November 2013, Professor Kevin Fenton, Director, Health and Wellbeing, PHE, stated that PHE was "really committed to instilling evidence in the programme" and that "from 1 April [PHE] would ensure that science—evaluation and research—underpinned the evaluation of the programme".[119] When we asked if this meant the NHS Health Check programme had been implemented without conclusive evidence of its effectiveness, Jamie Waterall, National Lead NHS Health Check, stated that while there was "strong evidence for individual risk factor management", there was a need for "better evidence around treating them as a collective".[120]

43. The Minister told us that the Health Check programme was "proceeding on a reasonable evidence base" that would be "built on", noting that results from two evaluations commissioned by the Department of Health would be available in the autumn.[121] The Minister also stated that NHS Health Check "was not strictly a screening programme"; according to Professor David Walker, Chair, UK NSC it was "more of a vascular risk management programme than a screening programme".[122] Nonetheless it was his own "personal view" that "for every programme that looks like a screening programme it would be useful to put it through the [UK] NSC" process.[123] He was hopeful that this approach would be followed in the future.[124]

44. Interventions that display all the hallmarks of being a systematic, population-based screening programme—like NHS Health Check—should not follow a "different route" bypassing the UK National Screening Committee's (UK NSC) evidence review process. To do so risks undermining the UK NSC's authority and, in the absence of the UK NSC's scrutiny, may give rise to serious questions about the quality of the evidence upon which the programme is based. We agree with the UK NSC Chair and recommend that, in the future, any programme that "looks like" a screening programme, regardless of the label it is given, should be subject to the UK NSC's evidence review process.

EXTENDING THE BREAST CANCER SCREENING PROGRAMME

45. All women who are registered with a GP and are aged 50-70 are currently sent an invitation every three years to attend for breast cancer screening. Public Health England told us that the NHS Breast Screening Programme was conducting a randomised control trial (RCT) on the screening of women aged 47-49 and 71-73 to examine whether "screening in the extended age ranges is effective or not".[125] Cancer Research UK described RCTs as the "'gold standard' of clinical evidence" while Breakthrough Breast Cancer and Breast Cancer Campaign stated that "facilitating robust research into the risks and benefits of screening older women [would] ultimately lead to an improved screening programme based on the best possible evidence".[126] Witnesses, however, disagreed whether a decision had already been taken to extend the age range of the breast cancer screening programme prior to this "internationally important" RCT reporting its findings.[127] According to the Cancer Epidemiology Unit, University of Oxford (the "co-investigators" of the age extension trial), Government "policy is to extend the age range for routine screening of all women from 50-70 to 47-73 in 2016"; however, the Unit added that "reliable mortality results" from the trial were "not expected until the early 2020s".[128] HealthWatch echoed these points, stating that the extended age range would "be implemented nationally in advance of [the RCT] being completed or the results being analysed".[129]

46. In contrast, Professor Walker, Chair, UK NSC, told us that:

130

Writing in the British Medical Journal, Professor Susan Bewley, King's College, London, stated that Professor Walker's comments to us represented "a dramatic policy U-turn, as hitherto it has been the Government's stated intention to extend".[131] In subsequent correspondence with Professor Bewley, Professor Walker clarified that his comments "were made on behalf of the UK NSC, not Government", noting that "the UK NSC does not implement policy". In the same letter, he added that, from a "Government perspective, Public Health England is responsible for funding the trial and say no final decision on the extension will be made until the trial results are known".[132] Professor Walker also pointed to the Government's 2011 Strategy for Cancer, which, he said, "made clear that full roll-out to women aged 47-49 and 71-73 was expected to be completed after 2016".[133]

47. We are concerned that there is ambiguity about whether the Government has agreed to the extension of the breast cancer screening programme to cover all women in England aged 47-49 and 71-73. We therefore recommend that, in the Government Response to this report, a clear statement is made about what has, and has not, already been agreed to regarding the extension of the breast cancer screening programme. We ask that this statement also detail the evidential basis for the Government's position.

48. The risk taken in not ensuring a policy is evidence based is poor policy that does not achieve its intended aims. We have heard from witnesses to this inquiry that the NHS Health Check programme may have suffered in this manner. The programme was introduced without an evidence base demonstrating that it could achieve its aims and we are concerned that it could be, as a result, wasting resources. We therefore recommend that the NHS Health Check programme be scrutinised by the UK National Screening Committee, retrospectively, to ascertain its value.

34 Q202; NHS0040 [Public Health England] para 3.2 Back

35 UK National Screening Committee, 'Policy Review Process', accessed 11 July 2014 Back

36 Q6 Back

37 Q6; NHS0017 [British Association of Urological Surgeons] para 4; NHS0026 [Muscular Dystrophy Campaign] para 7 Back

38 NHS0048 [Public Health Wales] Back

39 NHS0048 [Public Health Wales] Back

40 NHS0048 [Public Health Wales] Back

41 NHS0048 [Public Health Wales] Back

42 See paragraph 9 Back

43 Q219 [Dr Mackie] Back

44 Q220 Back

45 Q6 Back