The four pillars of medical ethics are autonomy, justice, beneficence, and nonmaleficence. The developers of clinical practice guidelines (CPGs) must ensure that each of these principles is respected in creating a trustworthy clinical guideline. Historically, guidelines were created by assembling experts and reaching consensus on how to manage a particular condition. The literature was selected to support the consensus, without any attempts at systematic review. Unfortunately, these recommendations were not always free of bias or ethically sound. Although the methodology has evolved for evaluating the literature and creating well-grounded guidelines, problems persist. In the last decade, guidelines on breast and prostate cancer screening -- as well as on the management of Barrett esophagitis and Lyme disease -- have come under scrutiny for excess bias. Despite the fact that these guidelines were touted as “evidence based,” critics argued the recommendations were biased and self-serving for the medical societies. In some cases, two organizations looking at exactly the same information created guidelines with contrasting recommendations. These concerns led the Institute of Medicine (IOM) to evaluate the current state of CPGs and their validity.
In 2011, the IOM produced two documents in response to these concerns: Clinical Practice Guidelines We Can Trust and Finding What Works in Healthcare: Standards for Systematic Reviews.(1,2) The first document establishes a set of standards for developing CPGs while the second outlines standards for performing systematic reviews of the literature. The Society of Critical Care Medicine (SCCM) collaborated with the Council of Medical Specialty Societies to develop realistic policies consistent with the IOM’s recommendations.(3) This was done to ensure that CPGs produced by SCCM are consistent with the four pillars of medical ethical principles and based on sound scientific evidence.
The first pillar, autonomy, is the patient’s right to make informed decisions regarding treatment options. To do this, the patient should understand the risks, benefits and alternate treatments. The obligation is on the developer of the CPG to clearly quantify these risks and benefits. Guidelines that only provide recommendations for care and omit a balanced discussion of the risks are incomplete and place the clinical provider in the position of explaining these risks. Realistically, providers will not have the same level of expertise as the CPG developer; any decision based solely on the clinical provider’s knowledge will be less informed than it would have been had the information been provided in the CPG.
The second pillar of medical ethics is the need to distribute any burden or benefit of a treatment in a fair and just manner. CPG developers must consider this when assessing increasingly expensive and complex therapies. Treatments that marginally improve care but significantly increase costs will not always be available in an era of cost constraints. For instance, a drug that increases cancer survival by three or four months at a cost of hundreds of thousands of dollars may not be available to every patient. Also, consideration should be given to a successful therapy that requires sophisticated or patented manufacturing processes. For example, antiretroviral therapies have demonstrated significant impact on treating human immunodeficiency virus (HIV) infections, yet making these drugs available to infected individuals in sub-Saharan Africa has posed significant challenges to the world health community. Use of CPGs is not confined to the United States or even to wealthy countries. An ethical dilemma arises when trying to create a CPG that can be used universally. The Surviving Sepsis Campaign, led by SCCM and the European Society of Intensive Care Medicine, has attempted to balance recommendation of therapies available to most countries against those limited to developed countries.
The final two pillars -- beneficence and nonmaleficence -- are interrelated and have generated the most controversy in the ethics of CPG development. Taken broadly, beneficence requires that recommendations are made with the intent of doing good for the patient. Nonmaleficence requires that the recommendations do not cause harm. Together, these requirements are the crux of an objective and systematic review of the literature, followed by unbiased recommendations coming from that review. Unfortunately, the majority of clinical questions addressed by CPGs are not associated with a body of multiple, congruent, prospective clinical trials. In fact, most of the evidence is limited in one way or another. This is why an entire science related to assessing the quality and strength of evidence has evolved. Regardless of the methodology employed, human bias enters into the process.
Bias can negatively impact attempts to do good and prevent harm. SCCM has adopted the Council of Medical Specialty Societies’ policies to mitigate bias in CPG development. The two most significant types of bias in guideline development come from industry and self-serving specialty involvement. The perceived industry bias occurs when a CPG developer has an established financial relationship with industry. Reviewers who have engaged in industry-sponsored trials are subject to this form of bias, and IOM has recommended eliminating all such individuals. When the clinical question under study involves a drug or device, these relationships are suspect. However, those who participate in such trials are often the very physicians who have the most knowledge and experience in the matter. These potential conflicts of interest (COI) need to be appropriately managed so that the resulting CPG is trustworthy. The first step in managing COI is establishing and maintaining strict policies on disclosure of potential competing interests and transparency. SCCM has adopted such policies and strives to ensure that, whenever possible, CPG groups are free of members with conflicts. When it is necessary to include members with prior industry relationships (for their expertise in the subject matter), additional safeguards are necessary. The potential COI and how it was managed must be disclosed in the guideline. Also, individuals with potential COI should be in the minority and should never act as chair of the committee or subcommittee. Additionally, they should not be directly involved in any decisions related to their potential COI, but may participate in the more general discussions where their expertise may be necessary. Finally, individuals with potential COI should not be permitted to vote on recommendations related to that conflict.
Specialty bias can be more difficult to mitigate because the evidence is rarely so overwhelming that clear and convincing recommendations can be made. This is why the systematic review of the literature should be done by individuals free of potential COI. Many societies contract out the systematic reviews to avoid any apparent conflicts. Unfortunately, these services -- provided by organizations such as the Agency for Healthcare Research and Quality -- are costly and financially prohibitive for small organizations. Once the systematic review is completed and the evidence graded, recommendations must be formulated. These recommendations should be differentiated by the strength of the related evidence. How these decisions are to be made should be determined in advance and made explicit in the CPG. The results of voting on the recommendations (in totals, not individuals’ votes) should be available as an appendix to the CPG. This level of transparency decreases potential biases and allows readers to determine the trustworthiness of the CPG. Additionally, any similar existing guidelines should be acknowledged, with differences and contrasts explained. This effort to harmonize CPGs is an important component in the development of a trustworthy guideline. Specialty/society bias can be further mitigated by allowing a public comment period, so concerns can be addressed before the final document is circulated for peer review and publication.
Finally, the CPG must be clear on what we know, what we think we know, and what we don’t know. Standardized processes, transparency and peer review all help in the development of a scientifically and ethically sound CPG. The American College of Critical Care Medicine and SCCM are dedicated to producing high-quality, relevant and up-to-date guidelines.
1. Institute of Medicine. Clinical Practice Guidelines We Can Trust. Committee on Standards for Developing Trustworthy Guidelines. Graham R, Mancher M, Wolman DM, Greenfield S, Steinberg E, eds. Washington, DC: The National Academies Press, 2011. Available at http://www.nap.edu/catalog.php?record_id=13058
. Accessed August 11, 2014.
2. Institute of Medicine. Finding What Works in Health Care: Standards for Systematic Reviews. Committee on Standards for Systematic Reviews of Comparative Effectiveness Research. Eden, J, Levit L, Berg A, Morton S, eds. Washington, DC: The National Academies Press, 2011. Available at http://www.nap.edu/catalog.php?record_id=13059
. Accessed August 11, 2014.