The Institute of Medicine Standards for Trustworthy Clinical Practice Guidelines: 4 Years Later

Has the development of clinical practice guidelines (CPGs) made meaningful progress since the IOM report was released in 2011?  Not an easy question to answer, but several promising signposts point to at least some progress to date for many of the recommendations made by that report.
The evidence base for the comparative effectiveness of medications, procedures and imaging – the underpinning for CPGs – has been greatly enhanced in focus, direction and public attention/scrutiny with the contributions made by the Patient Centered Outcomes Research Institute (PCORI) both methodologically and in the transparency and dissemination associated with increased stakeholder involvement.  The measurement of quality of care, led by organizations such as the National Quality Forum and stimulated by provisions of the ACA, has also received considerable public and professional scrutiny leading to enhancement of the evidence base for quality assessment and improvement.
While the scientific rigor for clinical effectiveness and quality of care research has progressed, CPGs continue to lie in a scientific and political no-man’s land.  The good work of the Guidelines International Network (G-I-N), of many professional societies and of the AHRQ National Guideline Clearinghouse (NGC) has led to some striking examples of excellence in the development of CPGs.  However, the lack of public scrutiny for the processes leading to the development of trustworthy guidelines that would lead to measurable improvements in clinical practice remains a barrier.  Even with the inclusion of some guidelines or guideline elements into the EHR (e.g. into a reminder system), there is little standardization of the criteria needed to establish what would represent a “good housekeeping seal” to represent high quality guidelines that should be broadly implemented.
The development of high quality CPGs is an arduous process involving the synthesis of the body of evidence supporting the multiple management decisions that collectively represent appropriate care for a specific condition, testing or clinical circumstance.  That synthesis requires the careful scrutiny, integration and collective interpretation of results of individual comparative effectiveness studies.  Further, when the scientific evidence is either missing or has gaps, the guideline development progress then requires the careful consideration by and consensus of groups of experts involved in the care of those patients.  The assessment of performance should be subsequently based on and derive from unbiased CPGs of high quality.  Until there is recognition of the relationship between and order of development of effectiveness studies, CPGs and quality measurement, guidelines may continue to occupy a scientific and public no-man’s land with no standardized criteria for excellence and no established development principles.
Yet, some progress has been made, particularly with respect to the quality of systematic reviews, the basis for CPGs and performance measures.  The efforts of AHRQ NGC have facilitated the ability to derive quantitative estimates of the effectiveness of medications, devices and tests/images.  Other guideline development groups also appear to be similarly adopting the criteria set out for systematic reviews in the IOM report.  However, it has to be clear that the improvements in systematic reviews will not necessarily be reflected in an increase in the number of independent, external, reviews of the type supported by AHRQ. Instead, and this has been heartening to the indefatigable stalwarts in G-I-N and to those in the AHRQ NGC program, there has been incremental improvement in criteria for quality reviews, even while being performed by the organizations themselves.  Independent reviews remain the aspiration; however, if the improvement is measureable and meaningful and not just window dressing, improvements could go a considerable way toward the goal of acceptable, trustworthy CPGs.  These improvements should be monitored closely by AHRQ so that the public can be informed as to whether the thresholds for high quality systematic reviews are being exceeded.
Another area of progress, with considerable work remaining to be done, has been the attention to the role of conflict of interest in the development of CPGs.  In many ways this area remains the biggest challenge and most formidable barrier to the creation of credible and useful practice guidelines.  As noted in a recent article in Health Affairs[1], conflict of interest among physicians – with respect to testing, medications, referrals, etc. – may be on the rise.  The implications for the construction of high quality CPGs are not trivial and are highly controversial.  The aspirational goal reflected in the IOM report was that each CPG development group is chaired by an individual with no conflict of interest in the management area under consideration, and that at least 50% of the members of the committee be from a discipline other than the primary specialty associated with the aspect of care addressed by the guideline.
A key aspect of the definition of conflict of interest is that a specific person may be conflicted if he/she earns a “substantial part of his/her income from services pertinent to the CPG”.  In practice, this definition is difficult to implement since expertise in the content area addressed is key to the creation of a high quality CPG.  The goal of development of conflict-free CPGs is still aspirational, but there is evidence of increased awareness of at least the need to manage conflicts.  The recurrent debate over mammograms that resurfaced within the past several weeks underscores the intractability of the conflict of interest dilemma and the deepening need to confront it.  Overcoming the COI problem cannot be accomplished by organizations such as AHRQ, professional societies or G-I-N alone.  Concerted pressure from the leadership of organized medicine in conjunction with regulating bodies and payers is needed to make progress in this area.
Two other challenges, both reflected in the issues related to systematic reviews and conflict of interest remain.  As discussed in Chapter 5 of the IOM report, to be broadly applied to clinical practice, CPGs must address heterogeneity of treatment effects[2].  In practice, one size fits none.  The average treatment effect combines the sometimes contradictory effects of treatments for key subgroups of patients.  Most CER, and most systematic reviews have focused on the synthesis of main effects of treatments, but the targets of CPGs are all or at least the majority of patients with a given condition or treatment need.  To build this body of evidence, representing a more diverse group of patients, more pressure on the sponsor of clinical research, such as NIH and PCORI should be brought to assure that studies are designed a priori to address heterogeneity vs. relying on post-hoc analyses.
Closely related to the challenge of conflict of interest is the need to address, integrate or “harmonize” multiple competing guidelines.  As noted by Richard Shiffman in his excellent summary[3] of the issues of conflict of interest in guideline development, there are multiple factors that contribute to the proliferation of competing guidelines.  As of now, there are no disincentives for the multiple stakeholders seeing patients addressed by guidelines to generate guidelines that reflect their interests.  So why bother?  What is wrong with ‘dueling guidelines’?  First, if CPGs are to serve as the basis for performance assessment, and potentially as the basis in turn for Value Based Purchasing, integration of guidelines is critical.
Second, if CPGs serve to inform reimbursement policies by payers, again, synthesis of guidelines is essential.  Finally, to avoid confusion in the public, as CPGs and performance measures are publicly available and reported, harmonizing competing guidelines is crucial.
Progress is indeed being made.  The stakes are high and much remains to be done.  The public is counting on us all.

[1] Rickert, J. Health Affairs Blog. April 10, 2015. What Can Patients Do In The Face Of Physician Conflict Of Interest? (accessed May 11, 2015)
[2] Greenfield S, Kravitz R, Duan N, Kaplan SH. Heterogeneity of Treatment Effects: Implications for Guidelines, Payment, and Quality Assessment. Am J Med. 2007;120(4A):S3-S9.
[3] Shiffman, RN. E-GAPPS II Conference Proceedings and Materials. March 2, 2015. Understanding Guideline Conflict.  (accessed May 11, 2015)