PHA ups the ante in CPG Development

One of the objectives of the Philippine Heart Association is to improve the quality of cardiovascular care in the Philippines. Thus, the development of relevant clinical practice guidelines (CPGs) that are appropriate for the local Philippine setting has been one of the main thrusts of the PHA.

To fast-tract the development of these CPGs, the PHA in partnership with the Asia Pacific Center for Evidence-based Medicine (APCEBM) conducted a CPG Development Workshop at the Crown Regency Hotel, Ortigas Center, Pasig last August 22-23, 2009. Over 50 members representing the 16 councils, chapters and board members of the PHA attended the workshop.

Dr. Antonio Dans of APCEBM highlighted the importance of developing CPGs in the local setting. Practicing medicine backed by valid studies (evidence) is crucial in qualifying and quantifying the benefits and risks of interventions, whether they are diagnostic, therapeutic, or rehabilitative.

The �explosion of evidence� makes it difficult for the clinician to keep up and integrate evidence into clinical practice. �Less than half of the knowledge and technology [published] is being used,� said Dr. Dans. �Instead of looking for new knowledge, why not focus more on using what we already know?� This is called �knowledge translation��translating currently available knowledge into useable knowledge. Like other integrative articles (e.g., systematic reviews, meta-analyses), CPGs are a way of translating knowledge, and helping physicians utilize evidence-based medicine.

But most studies are conducted on a Western population, which limits the applicability of these studies and the CPGs developed from these. Hence, local CPGs need to be developed in many (but not all) situations. �Globalize the evidence, localize the decision.�

Preparing for CPGs

Developing CPGs requires a 4-step cycle: 1) Preparation phase; 2) Data synthesis; 3) Consensus development; and 4) Monitoring of impact. The CPGs may then be updated, going through the same cycle, after the impact of the CPGs have been evaluated.

The preparation phase starts by identifying the problem: �Is there a need to do CPGs in the first place?� CPGs are appropriate when any of the following are present:

Significant variations in practice
Poor quality of care
Rising costs
General confusion.

Many times, there are already published CPGs available from other countries or organizations. These CPGs have to be evaluated, as part of data synthesis. First, are these CPGs valid (Are all groups considered in the recommendations? Are the interventions and outcomes relevant? Are there processes to consider like patient values and preferences? Do the CPGs indicate strengths of recommendations?) Secondly, are the CPGs applicable in the local setting? Are there differences (biological, technical, etc) that may affect the applicability of the guidelines? A local panel for applicability should evaluate both the recommendations and the evidence used in developing the recommendations.

The nitty gritty of CPG development

The panel is also in charge of reviewing the evidence and developing the recommendations (consensus development), specifically:

Judging the level of evidence (Is the quality of evidence high, medium, low, or very low?)
Evaluating the balance sheet of benefit (Does the evidence demonstrate net benefit?)
Developing and grading the recommendations (Should the clinician perform the intervention?)

A panel consists of a non-conflicted editor (a methods expert with no conflicts of interest), a deputy editor (a content expert), and panel members that should ideally include frontline physicians.

Considering the significant impact of panel�s decisions on the CPG recommendations, panel members should disclose conflicts of interest (COI). There are two classifications of COI: primary and secondary.

� Primary COI includes financial or intellectual (primary article authorship) relationships, including those of spouses or partner, with possible interested institutions within 36 months of the guidelines development (voting). Those with primary COI cannot participate in the drafting, deliberation, and voting of recommendations.

� Secondary COI includes financial relationships with interested institutions, but regarding an unrelated product; or authorship in secondary papers (e.g., reviews, other CPGs). Those with secondary COI may perform all roles, but should declare the said COI explicitly.

Making decisions

The International Clinical Epidemiology Network (INCLEN) advocates the use of the Knowledge Management Plus (KM+) CPG development process. This process attempts to standardize the way evidence is graded, benefits are qualified, and recommendations are set and equity is addressed.

Previous CPGs have used an over-simplistic approach to grading of evidence, based solely on the study designs (observational studies are of poor quality, while randomized controlled trials are of high quality).

But this approach fails to consider several factors, such as the specific intricacies of each individual study, as well as the quality of how each study was conducted. It is also possible to make strong recommendations in the absence of RCTs, if the only available evidence are observational studies (e.g., there are no RCTs demonstrating harm caused by smoking, but recommendations on smoking cessation are strong).

More importantly, this approach seldom considers essential factors in actual practice, such as drug availability or costs.

Hence, the KM+ method looks beyond the study designs, and looks at the overall climate of the evidence on the topic. The KM+ grading system specifically qualifies the extent to which one can be confident that an estimate of the effect or association is correct:

A high grading means further research is unlikely to change the confidence in the estimate of the effect.
A moderate grading means further research is likely to have important impact and may change the estimate.
A low grading means further research is very likely to change the estimate.
A very low grading means any estimate is very uncertain.

The panel also decides on the benefit and risks of the interventions (balance sheet). The panel first determines which outcomes the intervention needs to affect to be able to be considered beneficial. Only outcomes deemed critical will be taken into consideration. The panel then reviews the evidence, and judges whether the intervention has the following effect on these critical outcomes:

Net benefit (the intervention clearly does more good than harm)
Trade-offs (there are some tradeoffs between good and harm)
Uncertain benefit (it is unclear whether the intervention does more good than harm)
No net benefit (the intervention clearly does more harm than good).

Lastly, the panel decides on appropriate recommendations, and the strength of recommendations (or the degree of confidence that desirable effects outweigh undesirable effect). The recommendation will be:

Strongly do it.
Uncertain.
Don�t do it.
Strongly don�t do it.

This step considers all aspects, including the quality of the evidence, the seriousness of the outcome, the magnitude and precision of the benefits and risks, as well as the cost of the intervention.

A strong recommendation means all physicians should perform the recommendation. A weak (uncertain) recommendation means that patient choices and preferences can dictate whether or not the intervention should be performed, and that the aversion of harm becomes the primary objective.

Equity through guidelines

In the INCLEN KM+ CPG development process, the KM part includes an evaluation of efficacy, effectiveness and efficiency. Meanwhile, the �Plus� evaluates appropriateness and more importantly, equity.

According to Margaret Whitefield, inequities are differences that are unnecessary, avoidable, and more importantly, unfair and unjust. These inequities may occur between Places of residence, Race, Occupation, Gender, Religion, Education, Socioeconomic status, and Social capital/resources (PROGRESS).

The KM+ Equity Lens utilizes the following scrutinizing questions:

Is the health problem a priority for potentially disadvantaged groups?
Did the guidelines look into the possibility of differential effects in potentially disadvantaged populations?
Is the interest of potentially disadvantaged populations represented in the expert panel?
Is the interest of potentially disadvantaged populations represented in the feedback process?
Were feasible knowledge transfer strategies laid out to address barriers to the implementation of the guidelines in the disadvantaged population?
Did the impact assessment include evaluation of health gains across disadvantaged populations?

These 6 questions should be in mind during the entire process of CPG development. Only when CPGs address health inequities can these recommendations truly be faithful to the PHA Creed: We believe � In Tempering the science with the art of Medicine, practicing a holistic approach � In Health as vital in the development and progress of our nation � In Excellence, and the pursuit of it � In Alliances, constructive and synergistic, with all who share our vision � In Respect for human life.

Finally, the PHA Board members, Council chairs and chapter officers decided to adapt this model of Guideline development. The PHA Board might decide to create a Health Policy Committee who will oversee all guidelines and ensure implementation through legislation.