|Developing HIV/AIDS Treatment Guidelines (UNAIDS, 1999, 43 p.)|
|Part 2: Developing valid HIV/AIDS guidelines|
|Steps in developing HIV/AIDS treatment guidelines|
Establishing a consensus among experts helps to identify a common level of understanding on issues that may be controversial. This is usually conducted within a structured environment. The main methods used to achieve this are Delphi techniques, peer groups (nominal group techniques) and consensus conferences.
Four secondary methods of developing a consensus distinguished by the nature of the decision-making can be identified: informal consensus, formal consensus, evidence-linked, and explicit guideline development20. These methods are not mutually exclusive; a combination of them can produce the required agreement. This approach has been used in developing guidelines in several countries. As is common in many issues of a controversial nature, consensus may not always be reached. In such a situation the document produced should indicate this.
Informal consensus development
Commonly used as it is free of the analytical and research methodologies adopted by the others. However, the quality of guidelines developed in this way tends to be poor as meetings may be dominated by an interest group or individual. Many such guidelines lack user confidence owing to uncertainty in their development. Examples of this are the adult and paediatric HIV guidelines developed by The World Health Organization Global Programme on AIDS (GPA)21,22. These guidelines have been adapted for use in Caribbean countries by WHO GPA and the Pan American Health Organization (PAHO)23.
Formal consensus development
Like the Delphi, peer and nominal group methods, this has a clear, identifiable and orderly methodology but may lack the development of explicit linkages between recommendations and quality of evidence. An example of this is the United States National Institutes of Health Consensus Development Programme which takes place over three days. Guidelines are developed in closed session by a panel following a plenary session and open discussion, and are then presented to an audience and press conference on the third day24.
Evidence-linked guideline development
As the name suggests, the guidelines developed have recommendations that are linked to the scientific evidence, enhancing their validity. In the absence of scientific evidence this method cannot be applied and the resulting recommendations tend to be neutral and of little use to clinicians. A methodology that incorporates formal consensus methods has therefore been used to give the results some strength as in the case of the AHCPR. A formal assessment of scientific evidence is conducted followed by a panel meeting of experts and open forum sessions giving more input for practitioners.
Explicit guideline development
In this method guideline developers evaluate the benefits, harms, risks and costs of potential interventions and derive explicit estimates of the probability of each outcome. Several methods of identifying scientific evidence are applied in combination with consensus methods. This method is too complex, costly and time-consuming to be readily applied by the busy guideline developer20.
A methodology for the adaptation of HIV clinical algorithms to local conditions6 using a nominal group process was developed by the World Health Organization Global Programme on AIDS. This is a cheaper method of guideline development but has inherent weaknesses. Examples of these are: domination of a group by forceful personalities, and the lack of an explicit methodological process to reassure the user of its validity. Despite this, most national HIV/AIDS guidelines currently in use in many countries are based on this methodology25.