Q Is a new guideline that does not include an implementation strategy likely to prove of lasting benefit?
A Developing a guideline without an implementation strategy is like making a 'to do' list instead of doing what needs to be done, or having a policy without an operational plan.
Research and experience consistently show that making recommendations and distributing them achieve very little: only a proportion of recipients open the document; only a proportion of these read the information thoroughly; only a proportion of these fully understand the implications of what is written; only a proportion of these put their understanding into practice; and only a proportion of these maintain changes in practice. If the proportion is 1 in 5 at each stage, only 1 in 3125 people wil end up taking action!
Although there is no question that the implementation of every guideline should be managed – rather than left to chance and the good intentions of those concerned – it may not be necessary for all guidelines to include an implementation strategy.
First, implementation has to be tailored to the local situation: recommendations that fit all are likely to suit no-one very well.
Second, an implementation strategy has to be lived more than it has to be written. The process of considering the local implementation of a guideline is itself part of the implementation strategy. A written strategy, handed down by experts, is likely to have no more effect on behaviour than does sending out the guideline without an implementation strategy.
However, every guideline should emphasise in general terms the need for an organised local process of implementation, and should also make suggestions, particularly metho-dological ones, that will help participants implement the guideline better. For example, it may be helpful to suggest definitions, offer model proformas, propose ways of identifying or following up patients, or highlight key organisational issues that are likely to be encountered.
Guideline producers should play their part in ensuring that the follow-through is as effective and efficient as possible.
Q What level of evidence is necessary before research findings can be applied more generally, in the form of a guideline, which may, after all, greatly influence practice, costs and outcomes?
A This is a difficult question to answer in the abstract. It depends on a number of factors, some of which are subjective:
- The amount and type of evidence available
- The evidence that is likely to become available in the short to medium term
- Whether the need for a guideline is pressing (e.g. whether there is concern about wide variation in practice)
- The degree to which a consensus can be reached on the basis of the evidence available
- The potential risks or benefits from changing – or not changing – clinical practice
- The potential costs or savings from changing clinical practice, or leaving it as it is.
There is a general expectation that guidelines should indicate the strength of the evidence on which each recommendation is based, and this helps potential guideline users to decide for themselves whether the evidence is strong enough.
This strength has to be measured on two dimensions: the confidence one can have that one can rely on the results, and the strength of the conclusion if the results are reliable. Standard classifications of evidence are now used to make life easier for authors and readers.
Despite all the talk of evidence, many things remain matters of opinion because scientific evidence is absent, equivocal, or irrelevant. Nevertheless, we have to make decisions when patients come into the surgery, and have to do this as best we can and use the information we have to the patient's best advantage.
Q We all use devices in diagnosis and treatment. I recently heard of an instance where the manufacturer's instructions on the use of a device said one thing, but the nurses using the device had read more recent research that gave different advice. The MDA advises us to follow the manufacturer's instructions. How should we approach a situation like this?
A Devices are approved for specific purposes on the basis of the manufacturer's instructions. Use of the device in any other way is the responsibility of the professional concerned, and should problems arise, the consequences will fall on that person rather than the manufacturer.
As with pharmaceuticals, you may feel confident in using devices outside their licensed indications or in a wayithat differs from that recommended by the manufacturer, but you should be very cautious in doing this.
The research report read by the nurses may have been incorrect in some way: the data or analysis may have been flawed, or the authors may have drawn the wrong conclusions. The nurses may have misinterpreted the paper, e.g. the suggestions may apply to a slightly different device or in different circumstances.
In this situation, you should contact the manufacturer or the MDA, and discuss the apparent clash of information. The manufacturer will want to alter the instructions if new evidence has come to light, otherwise it may be liable for damages. The MDA will want to ensure that approval of the device is current and appropriate.