Consensus development for producing diagnostic procedure guidelines: SPECT brain perfusion imaging with exametazime

Prior research has shown that informal discussions among experts often lead to inconsistent diagnostic guidelines. No prior work had resolved how to systematically quantify expert agreement on imaging protocols. This gap motivated the need for a structured consensus method. Established knowledge includes the use of Delphi techniques in healthcare, but no prior work had applied semiquantitative scoring to SPECT imaging guidelines. That uncertainty drove the search for a more rigorous approach. It was already known that SPECT cerebral perfusion imaging requires standardized procedures. However, no prior work had demonstrated how to translate expert opinions into actionable guidelines. This paper's contribution lies in its systematic scoring method for consensus building.

The authors aimed to develop a structured consensus method for diagnostic guidelines in SPECT cerebral perfusion imaging. They sought to replace informal discussions with a quantifiable system. Their specific problem was the lack of a systematic approach to rate guideline elements. This motivated the use of a modified Delphi panel with scoring. The goal was to measure changes in expert agreement over time. They wanted to test if semiquantitative scoring could improve consensus. The study also aimed to categorize guideline elements by importance. Their approach allowed for clear documentation of the rationale behind each recommendation.

The team used a modified Delphi panel with three rounds of expert scoring. Panel members generated statements about optimal imaging procedures. Each statement received scores from panelists indicating its importance. The scores were analyzed for average, standard deviation, and variance. They tracked changes in standard deviation across panel rounds. Parametric and nonparametric tests assessed the significance of these changes. Statements were grouped into four categories based on average scores. This categorization formed the basis for a narrative guideline summary. The process took three months and produced a documented rationale.

The average standard deviation decreased by 35% from 2.1 to 1.32 between the first and final panel rounds. This reduction indicated improved consensus among panelists. The change was statistically significant at p < 0.0001. Statements were grouped into four categories based on average scores. Critical elements had the highest average scores. Less important elements had lower scores but remained relevant. Elements of uncertain importance required further discussion. The process was completed in three months at low cost. Clear documentation of the rationale was achieved through this method.

The authors concluded that semiquantitative scoring improved consensus among experts. The method allowed for systematic measurement of agreement changes. Their approach provided a structured way to categorize guideline elements. The results demonstrated that this method could replace informal discussions. The study showed how to rate evidence quality and recommendation strength. It also showed how to assess generalizability to practice conditions. The authors proposed that this method could be adapted for other guideline development. Their findings suggest that structured scoring enhances guideline clarity and consistency.