Defining and classifying peri-anaesthetic mortality in dogs and cats using the Delphi method: the Definition Research for Establishing Anaesthetic Mortality (D.R.E.A.M.) consensus study

Peri-anaesthetic mortality is an important indicator of anaesthetic safety in dogs and cats, but comparisons between studies are limited by inconsistent definitions, temporal windows and attribution criteria. This three-round online Delphi study sought expert consensus on principles for defining and conceptually classifying peri-anaesthetic mortality in dogs and cats. Spanish-speaking experts in veterinary anaesthesia and analgesia completed an exploratory open-ended round followed by two structured consensus rounds. Round 2 evaluated 38 statements on a 1–9 agreement scale, and Round 3 evaluated 18 refinement statements on a 0–10 scale transformed to a continuous 1–9 analytical scale. Positive consensus was defined as median ≥ 7 and interquartile range ≤ 2, and negative consensus as median ≤ 3 and interquartile range ≤ 2. Of 57 experts invited, 47 participated in Round 1, and all completed Rounds 2 and 3. In Round 2, 23 statements reached positive consensus, one reached negative consensus and 14 did not reach consensus. In Round 3, nine statements reached positive consensus, one reached negative consensus and eight did not reach consensus. The panel supported separating temporal eligibility from causal attribution, assigning one primary cause when sufficient information was available and recording contributing factors separately. Supported causal domains included anaesthesia, surgery, diagnostic procedures, pre-existing disease and sedation. Fixed 24-hour, 48-hour and 7-day windows did not reach consensus as standalone criteria. A flexible model using a 24-hour or 48-hour base window, with extension up to 7 days only when supported by evidence of an anaesthetic complication, reached positive consensus. Unresolved areas included the base window, single-count reporting, anaesthesia as a contributing but non-primary factor, treatment limitation or economic constraints, an independent sedation-specific definition and an indeterminate or unclassifiable category. This study supports a preliminary hierarchical causal framework rather than classification based on timing alone or a binary anaesthetic versus non-anaesthetic distinction. The findings provide a conceptual basis for developing and validating a practical classification tool for research, morbidity and mortality review and clinical audit, but the framework should not yet be interpreted as a validated adjudication tool.