The Rationale for the Rules
Why are the guideline standards so important?
That’s the question you need answered – what is the rationale behind the rules?
If you understand the why behind the rules, you’ll be better equipped to not just use the standards that exist today but to be able to adopt and adapt to new standards as the methods for guideline development improve.
There are many ways to approach understanding the reasons for the standards. Remember your school training when you worked hard for “straight As” to show you mastered new material? Let’s use “straight As” now to master the rules for guideline development.
What are the essential elements of a clinical practice guideline? Guidelines are tools used by providers and patients who want to make the best decisions about appropriate, individualized, care. In order to meet those goals, a guideline should be:
The guideline you develop or choose to adopt must provide the information needed by the care provider and the patient to make informed decisions about the management of the health of that patient. This means the guideline and its recommendations must be relevant and specific to patient and provider needs for information. How can you be sure the guideline will be relevant to clinicians and patients?
This is the rationale behind the guideline development standards that recommend establishing a panel with appropriate clinical expertise (2.1 IOM and Composition GIN) and patient input (2.3 IOM and Composition GIN). Ideally, the guideline development panel will include all the types of clinicians who treat patients with the condition of interest, and will also include patients who have the condition of interest. At a minimum you’ll want to include patients, clinicians, and other stakeholders during the scope and question development and during the review of recommendations.
Points to consider include:
- Which providers will use the guideline?
- Which patients will use the guideline?
- What care decisions will the guideline address?
To be trustworthy, the guideline must be developed following the best existing methods for ensuring accuracy and reliability. The guideline development must use sound science. Current best standards for accessing data buried in published and unpublished studies and file drawers requires:
o Extensive and unbiased search for the studies and accounting for missing studies
o Assessment of both the relevance and accuracy of the studies and content retrieved
o Appropriate methods to combine and evaluate the study findings
o Translation into actionable evidence using a system that rates the strength
This process is what occurs in systematic review. It requires you to do your best at finding all the evidence, not just studies that are easy to find or that you agree with, but to look for all the studies to access all the data. Because of the need for extensive searches for studies, not all studies retrieved will end up being relevant to the guideline questions that patients and providers want answered ie. applicable. And not all studies, even studies published in top journals, are valid. Studies that are poorly designed and conducted will not provide you with usable data. Would you build a house using faulty tape measures? After extensive search, followed by selection using pre-defined criteria, then evaluating the studies for quality, the data must be combined to be useful to clinicians and patients. They may be combined numerically through meta-analysis, if the data are suitable, or descriptively. Finally, the authors of the systematic review should label the strength of the synthesized evidence, so providers and patients know how confident they can be in the results.
These standards are so detailed, that when the IOM references them in their standards for guidelines (4.1 and 4.2), they refer developers to another IOM document on systematic review standards. The G-I-N standards also require the use of systematic review to identify and evaluate evidence but allow developers to follow either IOM standards or use G-I-N templates.
Gone are the days when “trust me, I’m the doctor” was sufficient rational for treatment decisions. Now, thank goodness, trust must be earned. To earn trust, the guideline must not only follow good scientific methods, it must demonstrate that it did so in a way that is clear to the users. Guideline developers are held accountable to follow the current standards and they must describe every step in the process with enough detail to demonstrate they followed the current standards. To make it easy for the user to evaluate, the guideline should clearly state the plan and goals and how they will be achieved. Another way to describe this is to say, guideline developers should create and share their protocol. Then the guideline developers should carefully follow the protocol, or provide detailed explanation of why the protocol had to be changed. In shorthand the guideline panel should:
DEVELOP A PROTOCOL
FOLLOW THE PROTOCOL
This is the rationale behind many of the guideline development standards that recommend descriptions and disclosures of all process methods and how decisions were handled. It’s also behind the requirements for disclosure of all potential sources of bias in the development of the guideline and how they were managed, including potential or perceived intellectual and financial conflicts of interest. All descriptions should be clear enough to allow the user to determine if the user agrees or disagrees with the decisions made throughout the process and the potential impact on the final recommendations.
An important part of the reasoning for making process decisions ahead of time is to minimize making choices during the process that could be influenced by prior belief or expert opinions. The goal of evidence-based guideline development is to use science to guide the process to make it as objective as possible.
So for accountability and also for accuracy, there should be no financial or intellectual conflicts of interest, especially in chairs who will lead the guideline development. The panel members should not be composed of people with answers, but instead of people with good questions and with skills to look for unbiased answers to those questions.
The IOM standards 2.1, 2.2, 2.3, and 2.4, address disclosure and management of conflicts of interest (COI) for these reasons, and G-I-N standards do likewise.
To be useable, the guideline must be:
- easy to obtain by the providers and patients that want to use it
- easy to understand
- easy to implement
Guideline developers can make the guideline available by offering it free on their website. They can also share it through the National Guideline Clearinghouse and the Guidelines International Network. Of course it must meet the standards of those organizations to be shared. To be easy to understand and implement, the guideline and recommendations should clearly state:
Who should do What, When, How, and Why
The recommendations should be actionable and unambiguous. They should make clear the actions for the provider and for the patient to achieve the best outcome. The recommendations must take into consideration the totality of the evidence for beneficial outcomes and the evidence of potential harms, making clear where evidence stops and conjecture begins. Recommendations based on strong evidence can be clear in their recommendations and use strong language to recommend for or against the action. As Yoda says in Star Wars to young Luke Skywalker, strong recommendations will say
“Do. Or do not”
(“There is no try”)
But weak recommendations, which will be those based on limited evidence, may suggest try.
Strong recommendations are warranted when you are confident that the evidence supports a net benefit from the action and when the patient prefers or accepts the action and outcome. Weak recommendations are warranted when the evidence is limited or the net benefit is small. In case you are not familiar with the term net benefit, it means when the overall benefit exceeds the harm for that particular patient. For example, chemotherapy as a therapy is risky, and has potential for harm because of the toxicity to cells other than cancer cells. But it can have net benefit for some cancer patients.
These principles are clearly described in the standards of both the IOM and the G-I-N.
Additionally, strong recommendations may be adopted by regulatory or payer organizations to use to measure care performance, since strong recommendations should only be developed when the evidence is strong enough to warrant them.
So the standards for developing and labeling recommendations as either strong or weak are important, because they may lead to widespread adoption of care practices labeled by strong recommendation and avoidance of care practices labeled as weak. It is important that the labels be correct. If strong recommendations are not based on strong evidence, they may institutionalize substandard or harmful care practices.
The IOM standard 6.2 says strong recommendations should be worded so compliance can be evaluated.
Adaptable means the guideline offers options for individualized care. The guideline needs to be applicable to individual patients seen in clinical settings. For a guideline to be adaptable requires sufficient evidence to characterize both the effectiveness and the safety of the clinical action. Users need to be able to assess the risk of all important outcomes by important individual risk factors and be able to see the change in risk with all potential treatments. This requires enough detail to assess the impact of the alternative clinical actions in specific patient subgroups so that the potential risks and benefits can be assessed and discussed for shared decisions. This feature, at present, is often lacking in clinical practice guideline development, and is not precisely defined in any of the standards.
But it should be the goal we seek as we develop and select guidelines to optimize the right care for the right patient and the right time.
Applicable - Accurate - Accountable - Accessible - Adaptable