Practice parameters developed by the American Society of Anesthesiologists (ASA) have been an important resource for physicians and other health-care workers for more than 20 years. The intention of the ASA evidence-based practice parameter is to enhance and promote safe medical practice as well as offer guidance for diagnosing, managing, or treating a variety of clinical conditions. ASA evidence-based practice parameters consist of a “broad body of documents developed on the basis of a systematic and standardized approach to the collection, assessment, analysis and reporting of: scientific literature, expert opinion, ASA member opinion, feasibility data and open forum commentary.” Evidence-based practice parameters may take the form of guidelines or advisories.
Before the development of a policy for evidence-based practice parameters in 1991, ASA practice parameters were primarily consensus-based documents, and the majority of these documents were practice standards. Practice standards were typically declarative statements focusing on simple aspects of patient care applicable to virtually all relevant anesthetic situations. The standards were well received within both the anesthesia community and allied medical professions and positioned the ASA and the Anesthesia Patient Safety Foundation of the ASA at the forefront of medical practice by demonstrating the benefits of a proactive approach to patient safety.
Many aspects of practice, however, could not be adequately covered by the relatively limited and prescriptive recommendations of practice standards. When broader and more flexible recommendations for practice were needed, the ASA broadened its scope to encompass practice guidelines. The practice guidelines were initially formulated on the basis of evidence generated by the same consensus-based methodology used in the development of standards. To effectively evaluate the increasing breadth and complexity of issues considered by practice guidelines, the ASA Committee on Standards and Practice Parameters (Committee) determined that a systematic evaluation of scientific evidence was necessary to fully support recommendations driven by expert opinion. Using a method that systematically combined a synthesis of the literature with opinions from experts and other sources, the ASA produced the first two evidence-based practice guidelines in 1993. In developing these guidelines, the Committee recognized the unique properties of both the anesthesia literature and the practice of anesthesiology and realized that further methodologic changes were needed. Over the next few years, a more elaborate multidimensional method to guideline development evolved. It contained four critical components: (1) a rigorous review and evaluation of all available published scientific evidence, (2) meta-analytic assessments of controlled clinical studies when appropriate, (3) a statistical assessment of expert and practitioner opinions obtained by formally developed surveys, and (4) the informal evaluation of opinions obtained from invited and public commentary.
Process of Parameter Development
The process used by the ASA to develop evidence-based practice parameters normally begins when the Committee identifies an issue or clinical problem. The Committee then appoints a task force of 8 to 12 anesthesiologists who are recognized experts on the issue or clinical problem to advise the Committee on the need for a practice parameter. Task force members are carefully chosen to not only provide representation from both private practice and academia but also ensure representation across major geographic areas of the United States. Occasionally, nonanesthesiologists may also be appointed to a task force if the Committee determines that their appointment would add specific subspecialty expertise (e.g., the appointment of a radiologist to the magnetic resonance imaging task force). Conflict of interest issues are fully evaluated before individuals are selected to serve on a task force, and such information is fully transparent to the reader.
If the task force determines that sufficient evidence is available, the process of defining goals and objectives within the mandate established by the Committee begins. During this conceptualization phase, approximately 75 to 150 peer-review consultants are identified as secondary external sources of opinion, practical knowledge, and expertise. Consultants typically are recognized experts in the subject matter and, like the task force members, represent a balance of practice settings and geographic locations. Depending on the clinical topic, individuals from nonanesthesia medical specialties or organizations may be selected as consultants.
An initial step in the development of an evidence-based practice parameter is to survey the task force members to identify target conditions, patient or clinical presentations, providers, interventions, practice settings, and other characteristics that help define or clarify the parameter. On the basis of the survey responses, members of the task force collectively develop a list of clinical interventions and expected outcomes. The list, typically referred to as “evidence linkages” between interventions and outcomes, forms the foundation on which evidence is collected and organized and provides structure for formulation of recommendations. When possible and appropriate, evidence linkages are designed to describe comparative relationships between interventions and outcomes. For example, the linkage statement “spinal opioids versus parenteral opioids improve maternal analgesia for labor” identifies a specific intervention (spinal opioids), a comparison intervention (parenteral opioids), and a specific clinical outcome (maternal analgesia) thought to be affected by the intervention. Once all evidence linkages for the parameter are specified, the task force then begins the process of collecting evidence.
Sources of Evidence
The multiple sources of information used by a task force in developing an evidence-based practice parameter are displayed in Table 1-1 . During the search for evidence, the task force considers two major sources: literature-based evidence and opinion-based evidence. Within the domain of literature-based evidence, meta-analytic findings are reported when sufficient numbers of randomized controlled trials (RCTs) are available, and descriptive outcome data summaries (e.g., means, ranges, and sensitivity/specificity values) are reported for interventions not subject to evaluation by RCTs. For opinion-based evidence, consensus-based information obtained from formal surveys as well as informal sources (e.g., open forum commentary and Internet comments) is considered. The final determination of whether the document is a guideline or an advisory is based on the totality of evidence accumulated.
| Source of Evidence | Type of Evidence |
|---|---|
| Literature-Based Evidence | |
| Randomized controlled trials | Comparative statistics |
| Nonrandomized prospective studies | Comparative statistics |
| Controlled observational studies | Correlation/regression |
| Retrospective comparative studies | Comparative statistics |
| Uncontrolled observational studies | Correlation/regression/descriptive statistics |
| Case reports | No statistical data |
| Opinion-Based Evidence | |
| Consultants | Survey findings/expert opinion |
| ASA members | Survey findings/opinion |
| Invited sources | Expert opinion |
| Open forum commentary | Public opinion |
| Internet commentary | Public opinion |
The Literature Search
The initial literature search includes a computerized search of PubMed and other large reference sources/databases and usually yields 2000 to 5000 citations for each practice parameter. Manual searches are also conducted when supplemental references are supplied by the consultants and members of the task force.
In the selection of published studies, three conditions must be met. First, the study must assess one or more of the interventions being considered. Second, the study must report an anesthetic or clinical outcome or set of findings that can be tallied or quantified, thereby eliminating reports that contain only opinion (e.g., editorials and news reports). Third, the study must be an original investigation or report. Review articles, books or book chapters, and manuscripts that report findings from previous publications are not used as sources of evidence. After the initial electronic review, letters, editorials, commentaries, and other literature with no original data are removed from consideration. Typically, only 1000 to 2500 articles prove suitable for retrieval and further review.
Evaluating and Summarizing the Literature
The literature review process focuses on studies that report outcomes relevant to an identified intervention. A standard classification system separates findings by strength and quality of research design, statistical findings, and type of data. RCTs offer the strongest evidence; findings from studies using other research designs are separately categorized as observational. Observational studies contain critical information not necessarily found in RCTs. For example, a nonrandomized comparative study may provide evidence for the differential benefits or risks of select interventions. Observational studies may report frequency or incidence data revealing the scope of a problem, event, or condition or may report correlations that associate clinical interventions and outcomes. In addition, when case reports describe adverse events that are not normally reported in controlled studies, they can be a source of important cautionary notations within a recommendation or advisory. Case reports also may be the first indication that a new drug or new technique is associated with a previously unrecognized benefit or unwanted side effect.
One of the strengths of the ASA protocol for developing evidence-based practice parameters is that the primary search and evaluation of the literature are jointly conducted by the clinicians and methodologists of the task force. Consequently, the clinical and practical significance of a study, as well as its research design and statistical aspects, are appropriately and thoroughly evaluated. The protocol is evaluated with the use of formal reliability testing by task force members and methodologists. Interobserver agreement values for research design, type of analysis, linkage assignment, and study inclusion are calculated with both two-rater agreement pairs (kappa) and multirater chance-corrected agreement (Sav) calculations. These values are reported in the final published document.
Evaluating and Summarizing Consensus Opinion
Although literature-based scientific evidence is a critical part of the process of developing an evidence-based practice parameter, the literature is never used as the sole source of evidence. Scientific findings are always supplemented by the practical knowledge and opinions of expert consultants. The consultants participate in formal surveys regarding conceptualization, application, and feasibility, and they review and comment on the initial draft by the task force. Opinion surveys of the ASA membership also are conducted to obtain additional consensus-based information used in the final development of an evidence-based practice parameter. The evidence obtained from surveys of consultants and ASA members represents a valuable and quantifiable source, critical to the formulation of effective and useful practice parameters.
In addition to survey information and commentary obtained from consultants and practitioners, the task force continually attempts to maximize the amount of consensus-based information by obtaining opinions from a broader range of sources. These sources include comments made by readers of a draft of the practice parameter posted on the ASA website ( www.asahq.org ) and comments from attendees of public forum presentations of the practice parameters scheduled during major national meetings. After collection and analysis of all scientific and consensus-based information, the draft document is further revised, and additional commentary or opinion is solicited from invited sources, such as the ASA Board of Directors and presidents of ASA component societies.
Meta-Analytic Evidence
When sufficient numbers of controlled studies are found addressing a particular evidence linkage, a formal meta-analysis for each specific outcome is conducted. For studies containing continuous data, either general variance-based methods or combined probability tests are used. When studies report dichotomous outcomes, an odds-ratio procedure is applied. In summarizing findings, an acceptable significance level typically is set at p < 0.01 (one-tailed) and effect size estimates are determined.
Reported findings in the anesthesia literature often use common outcome measures, thereby enhancing the likelihood that aggregated (i.e., pooled) studies will be homogeneous. Because homogeneity is generally expected, a fixed-effects meta-analytic model is used for the initial analysis. If the pooled studies for an evidence linkage are subsequently found to be heterogeneous, a random-effects analysis is performed, and possible reasons for the heterogeneous findings are explored. The heterogeneous findings are reported and discussed as part of the literature summary for an evidence linkage.
Whenever possible, more than one test is used so that a better statistical profile of the evidence linkage can be evaluated. For example, when a set of studies allows for more than one meta-analysis (e.g., using both continuous and dichotomous findings), separate meta-analyses are conducted. To be conclusive, the separate findings for the results of the analysis must agree. Additionally, the results should be in agreement with the directional evaluation of the literature and with consensus opinion before an unequivocal supportive recommendation is offered. If the results do not agree, the disparity is fully reported in the summary of evidence and acknowledged in caveats or notations to the recommendation.
Related posts:
Who Should Have a Preoperative 12-Lead Electrocardiogram?
Should Patients with Stable Coronary Artery Disease Undergo Prophylactic Revascularization before Noncardiac Surgery?
Is There a Best Approach for Patients with Difficult Airways: Regional versus General Anesthesia?
Does the Choice of Fluid Matter in Major Surgery?
Are There Special Techniques in Obese Patients?
Is Propofol Safe If Given by Nonanesthesia Providers?
Full access? Get Clinical Tree
