13 Ensuring that published guidelines are current and accurate

This chapter describes the process and methods for checking that published guidelines are current and deciding whether updates are needed (surveillance).

13.1 Aims of surveillance

The aim of surveillance is to check that guidelines are up to date. This is done by exploring if there is any new evidence to contradict, reinforce or clarify guideline recommendations. Surveillance also identifies new interventions that may need to be considered within the guideline. Finally, it explores changes in context that may mean modifications are needed, for example, changes in policy, infrastructure, legislation or costs.

NICE has more than 270 published guidelines so the number of checks needed is considerable. A proactive approach is taken that includes reacting to events at any time after guideline publication (for example, publication of a key study) and a standard check every 5 years. Where possible, surveillance of guidelines covering similar populations or settings is undertaken at the same time.

13.2 Reacting to events

Some topic areas are fast moving and this increases the risk of guidelines having out-of-date recommendations. Therefore, NICE maintains an event tracker containing information on key events, such as ongoing studies, that are judged to be relevant to the guideline content. Ongoing studies are identified for the event tracker through the standard check and also through NICE's engagement with the National Institute for Health Research. The event tracker means that NICE can react quickly to changes in the evidence base, by initiating a check of the guideline as soon as the event has occurred. A check does not necessarily mean that the guideline will be updated.

An event that could affect the guideline could include:

  • publication of a study that is directly relevant to NICE guidance and has the potential to affect recommendations

  • substantial changes in policy or legislation (an example includes changes to the UK physical activity guidelines by the Chief Medical Office)

  • development of a related piece of NICE guidance that contradicts recommendations in another NICE guideline

  • withdrawal of a drug from the market or a clinically significant drug safety update from the Medicines and Healthcare products Regulatory Authority (MHRA)/Commission on Human Medicines.

This list is not exhaustive and individual events will be considered on a case-by-case basis. Events are identified through constant intelligence gathering, for example, the standard check, the guideline development process and stakeholder correspondence and enquiries.

Process for reacting to events

The NICE surveillance team considers how an event could affect a guideline. If an event is likely to affect guideline recommendations a check is performed before the next scheduled standard check. This involves considering the impact of the event on the guideline recommendations and incorporating feedback from topic experts in the area. The check may include intelligence gathering and literature searches, if needed, involving the same approach as for the standard check.

Checks in response to events do not undergo stakeholder consultation because they focus only on an important event and potentially a small section of a guideline. However, the decisions are communicated on the NICE website.

If NICE's Guidance Executive decides that an update of the guideline is needed after this type of check, registered stakeholders are informed of the planned approach.

13.3 The standard check

The median lifespan of a clinical guideline is 60 months (Alderson et al. 2014). More recent work within NICE incorporating data for public health guidelines supports this conclusion. Therefore, all NICE guidelines will be checked every 5 years using the approach described below.

Topic expert engagement

Topic experts including members of NICE's Expert Advisers Panel are invited to participate in surveillance and provide views about the continued relevance of recommendations. But if their response is limited or further specialist input is needed, we may seek input from other experts, such as government bodies or representatives from a Quality Standards Advisory Committee.

Intelligence gathering

Topic experts are surveyed for their views on the continued relevance of the published guideline and recommendations, and their knowledge of recent developments in the topic area and any important new evidence since publication of the guideline. Feedback is also sought from internal teams within NICE who have expertise in the topic area under surveillance (for example, where there is a social care or medicines focus in the guideline). We may also ask stakeholders for their views, including organisations representing the interests of patients, people using services, carers and the public.

Additional intelligence might include:

  • external queries and comments received since publication of the guideline (these are collated in an issues log for consideration during surveillance)

  • related NICE guidance and quality standards (including placeholder statements in NICE quality standards) developed since the guideline was published

  • information about guideline implementation, including evidence derived from analysis of primary data on the uptake of recommendations

  • information about important ongoing studies in the area covered by the guideline (identified through searches of trial databases)

  • changes in licensing status of medicines

  • updated or new national policy.

Literature searching

Published evidence is identified through searching a range of bibliographic databases relevant to the topic, which are generally based on those searched for the published guideline. Sources searched may vary depending on the topic. In general, MEDLINE, MEDLINE in Process, Embase, Cochrane Database of Systematic Reviews and Cochrane Central Register of Controlled Trials (CENTRAL) would be considered.

Search approaches

The search approach will vary between topics and may depend on priority areas highlighted through topic expert engagement and intelligence gathering. The following search approaches can be used:

  • population or population/intervention search as needed for the guideline scope with:

  • focused search(es) for a specific question or a new question, meaning that the study type searched for (RCTs or observational studies) should reflect the type expected to address the question.

  • citation search forward/back (this option would be supplemented with either a restrictive full scope search or focused searches).

Other considerations

It may be appropriate to consider setting limits for the searches, which could include, but are not limited to:

  • study design using appropriate search filter(s)

  • date

  • location

  • population(s)/subpopulation(s)

  • intervention

  • service delivery aspect

  • prognostic factors.

Search period

The search period will start at the:

  • end of the search for the last update of the guideline

  • end of the search for the last standard check.

The search date ends on the date the search is conducted.

Decision-making

Proposals on the need to update a guideline include an element of judgement and are based on an assessment of the relevant evidence published since guideline publication (abstracts of primary or secondary evidence), information obtained through intelligence gathering and feedback from stakeholder consultation.

The update proposal will be based on the following options:

  • no update (check again in 5 years)

  • no update at present but date of next check should be brought forward or pushed back (this decision would be made exceptionally, for example where it is clear that new evidence critical to this decision is due to be published)

  • full update (develop replacement guideline)

  • partial update (update defined sections of the guideline)

  • transferring the guideline to the static list (see section 13.4)

  • refreshing the guideline (see section 14.4)

  • withdrawing some recommendations or the whole guideline.

When a guideline is being updated, the original scope may be used (unchanged), the original scope may be modified (for example, where new areas have been identified that require an extension to the scope) or a new scope may be developed. See chapters 2 and 14 for further details.

Stakeholder consultation

For all proposals other than a full update, we consult with stakeholders for 2 weeks. There is no stakeholder consultation on the proposal for a full update.

Signing off the final decision

All surveillance proposals go through a validation process at NICE (including sign‑off by the associate director and centre director) before approval is sought from NICE's Guidance Executive.

13.4 Static list check

Guidelines are considered static when the recommendations are still current and should continue to be implemented, but are unlikely to change in the foreseeable future (because the evidence base or practice is unlikely to change). Guidelines are only considered static after consultation with stakeholders, and providing the following criteria are met:

  • there is a decision not to update following a standard check and no major ongoing research expected to publish before the next standard check or

  • the guideline is not intervention-based (for example, it focuses on commissioning or implementation) and no major changes to commissioning or service configurations have occurred since guideline publication, or are expected.

Following stakeholder consultation, a proposal to add the guideline to the static list may no longer be appropriate if stakeholders have made us aware of:

  • relevant research or

  • pertinent issues that need to be monitored or

  • information that would impact on the 'no update' proposal.

Any ongoing research would be added to the event tracker to feed into a guideline check process.

Process

Static guidelines are looked at 5 years after they go on the static list and then every 5 years to determine whether they should undergo a standard check. This preliminary check is similar to the standard check but no literature searches are done. Topic experts are asked to supply information on any new published evidence that could affect the recommendations. Information is also gathered from the event tracker. This is likely to highlight the main events that could trigger a standard check and a possible update of the guideline. This process is applied consistently across static list guidelines with key decisions recorded as part of an audit trail. Guidelines are removed from the static list when the preliminary check suggests new evidence may affect the recommendation. They then undergo the standard check as described in section 13.3.

13.5 References and further reading

Alderson LJ, Alderson P, Tan T (2014) Median life span of a cohort of National Institute for Health and Care Excellence clinical guidelines was about 60 months. Journal of Clinical Epidemiology 67: 52–5

Shekelle P, Eccles MP, Grimshaw JM et al. (2001) When should clinical guidelines be updated? British Medical Journal 323: 155–7

Shekelle PG, Ortiz E, Rhodes S et al. (2001) Validity of the Agency for Healthcare Research and Quality clinical practice Alderson LJ, Alderson P, Tan T (2014) guidelines: how quickly do guidelines become outdated? JAMA 286: 1461–7