Guidelines Development

Guideline Selection and Approval Process
New guideline topics and revisions to current guidelines are prioritized according to relevance to the Society’s mission and to patients and their families, in addition to budgetary considerations.

The Guideline Topic Proposal Form is due March 31 annually. All guidelines proposals go through the ACCM BOR approval process. The selection process for new guidelines and focused updates uses specific criteria to rank the highest-priority topics as outlined in the SOP Manual. Submitted proposals are evaluated, scored, and ranked by the BOR using the Guidelines Proposal Priority Scoring Tool. At the May BOR meeting, the results are discussed, and BOR members vote on whether a guidelines proposal should be amended, approved, or rejected. The guidelines authors are then notified of the decision via email by May 30. If the guidelines proposal is approved, the work begins on October 1.

Once a guidelines proposal is selected for development or revision, the BOR initiates the process of selecting guidelines leadership, comprised of cochairs and co-vice-chairs. Leaders are selected based on such factors as strong leadership skills, subject matter expertise, and multiprofessionalism. The guidelines leadership and panel should be chosen in accordance with principles of diversity, equity, and inclusion (DEI). SCCM uses the DEI survey to ensure inclusion on the panel of a wide range of professionals from diverse healthcare settings and cultural backgrounds. Potential conflicts of interest (COI), including intellectual and financial influences, are also weighed.
 

Guidelines Development Process and Milestones
Every guideline or practice statement has a timeline that the panel develops within the first three months of its formation. By using this timeline as reference, the cochairs and co-vice-chairs can ensure that the activity stays on target. If the guidelines are complex, the timeline may need to be extended; however, generally guidelines should be completed and submitted to the journal within 18 months after the panel’s first full meeting.
 

Steps: PICO to Voting

Step 1: Forming PICO Questions
SCCM has adopted the PICO format, which defines clinical questions in terms of a specific problem and aids the panel in narrowing clinically relevant evidence. Sometimes refinement of PICO questions is required based on the literature review. PICO questions are first formed by the guidelines leadership and group heads,  if applicable, and then are vetted by the panel. SCCM’s template for development of PICO is as follows:
 

PICO Step	Example
Patient	Critically ill adults at risk for stress-related upper gastrointestinal bleeding
Intervention	Pharmacologic stress ulcer prophylaxis
Comparison	No pharmacologic stress ulcer prophylaxis
Outcome	Reduced incidence of stress-related upper gastrointestinal bleeding

Step 2: Systematic Review and Meta-Analysis
SCCM has contracted with the Guidelines in Intensive Care Medicine, Development and Evaluation (GUIDE) group of McMaster University to perform a systematic review of the available evidence to address each PICO question. First, a medical librarian performs literature searches, and the group narrows the large body of evidence by applying inclusion/exclusion criteria to identify studies specifically relevant to the PICO questions. Then, a GUIDE methodologist extracts information from the included studies and performs meta-analyses to determine the magnitude of effect of an intervention on specific outcomes to assess efficacy. Studies that cannot be meta-analyzed may be included as supporting evidence.

Step 3: Grading of Evidence
The methodologist evaluates the quality and strength of the evidence using the grading of recommendations, assessment, development, and evaluation (GRADE) methodology. GRADE has four main components:

• Quality of evidence reflects the confidence that the estimate of effect from meta-analyses or single studies for a given intervention represents the true effect for patients. Estimates of effect are clinically significant if the estimate exceeds a predetermined threshold set by the guidelines panel.
• Balance of benefits and harms is assessed to identify the direction and magnitude of the effect of a given intervention.
• Patient values and preferences are considered to determine the likelihood that most well-informed patients would choose to use the intervention over the comparator based on the evidence.
• Resource use considers the availability and cost of the intervention to the target patient population.

Step 4: Recommendation Formulation and Voting
After completing the systematic review and grading the evidence, the methodologist meets with the guidelines panel to formulate recommendations. The Evidence-to-Decision Framework within the GRADE Pro software is used to facilitate the transition from evidence to recommendation. Recommendations are worded as actionable statements rather than statements of fact. Recommendations are also worded as either “strong” or “conditional.” A “strong” recommendation indicates something that a clinician should almost always do, while “weak” or “conditional” recommendations indicate that the clinician and patient must determine the most appropriate option based on the balance of the quality of evidence, benefits and harms, values and preferences, and resource implications.

SCCM’s template for writing recommendations:

We recommend/suggest that clinicians [insert intervention] compared with [insert comparator] be [insert action word, e.g., used, implemented, adopted, etc.] for the treatment/management/diagnosis of [insert patient/population description].

All panelists without COI can provide input on the final recommendations. For each recommendation, an online survey will be sent to panel members without COI, asking whether they agree or disagree with the strength and direction of the recommendation. Recommendations that fail to achieve at least 80% consensus will be revised by the group head, panel members, and methodologist, then sent electronically for another round of voting, up to a maximum of three rounds of voting. If no consensus is reached after the third round of voting, the panel will not issue a recommendation.

Journal Submission and Publication
All guidelines, except those governed by memoranda of understanding (MOUs) that specify otherwise, are submitted to one of the SCCM journals. Guidelines related to pediatrics are submitted to Pediatric Critical Care Medicine. Adult guidelines are submitted to Critical Care Medicine. There is no provision for joint copyright for SCCM guidelines with other organizations except in the case of MOUs with approvals from the editors-in-chief that specify joint guidelines. Specific instructions for the preparation of guidelines manuscripts include length, tables, ease of reading, and single voice and are described in Appendix D of the SOP Manual.

Guidelines Dissemination and Uptake
Toolkits are extremely helpful and are considered essential to facilitate the application of guidelines at the bedside. Toolkits can include sample protocols, teaching slides, infographics, instructional videos, gap analysis tools, pocket cards, checklists, and other visually pleasing implementation resources. Toolkit materials are typically submitted to the BOR for testing and input before publication. SCCM communications staff are responsible for disseminating information about new and revised guidelines after publication. SCCM allows translations of guidelines and executive summaries following Council policy to use professional translation services or via MOU with approved professional medical societies. Translated guidelines will be posted and links provided where possible on the journals’ websites.

Submit a Guideline Topic
New guideline topics and revisions to current guidelines are prioritized according to relevance to the Society’s mission and to patients and their families, in addition to budgetary considerations. Guideline topic submitters need not be SCCM members. The Guideline Topic Proposal Form is due March 31 annually. Email notification of review outcomes will be sent to the indicated contact by May 30. Submitted proposals will be reviewed by the ACCM. Learn more about the selection process and criteria.