OBJECTIVE: Differential air-fluid levels are two distinct air-fluid interfaces on horizontal-beam abdominal radiographs that are at different heights but within the same loop of bowel. Differential air-fluid levels have been considered by many to be strong evidence of mechanical bowel obstruction, but others have found this sign unreliable for differentiating mechanical from adynamic obstructions. Neither opinion is supported by evidence from large series of patients. Accordingly, we determined the efficacy of differential air-fluid levels for distinguishing mechanical from adynamic bowel obstruction. MATERIALS AND METHODS: We identified patients who had a total of 62 episodes of proved mechanical bowel obstruction and 38 episodes of adynamic obstruction through a computer search of medical records and radiographic files. On horizontal-beam abdominal radiographs of these patients, the presence and height of intestinal differential air-fluid levels were determined by the consensus of two experienced radiologists. These data were then statistically analyzed to determine the usefulness of differential air-fluid levels for distinguishing between mechanical and adynamic bowel obstructions. RESULTS: Plain films showed differential air-fluid levels in 32 (52%) of the 62 episodes of mechanical obstructions compared with 11 (29%) of the 38 adynamic obstructions, giving a sensitivity for mechanical obstruction of 0.52 and a specificity of 0.71. As the minimum significant height of differential air-fluid levels increased, specificity increased and sensitivity decreased. The positive predictive value also increased as differential air-fluid level heights increased, reaching a level of 0.86 or greater at 20 mm. CONCLUSION: The presence of differential air-fluid levels is an insensitive method of determining if a bowel obstruction is mechanical, because only a small proportion of mechanical obstructions have differential air-fluid levels. In our population of patients, however, a differential air-fluid level of 20 mm or greater was moderately suggestive that a bowel obstruction was mechanical in nature.
OBJECTIVE: Differential air-fluid levels are two distinct air-fluid interfaces on horizontal-beam abdominal radiographs that are at different heights but within the same loop of bowel. Differential air-fluid levels have been considered by many to be strong evidence of mechanical bowel obstruction, but others have found this sign unreliable for differentiating mechanical from adynamic obstructions. Neither opinion is supported by evidence from large series of patients. Accordingly, we determined the efficacy of differential air-fluid levels for distinguishing mechanical from adynamic bowel obstruction. MATERIALS AND METHODS: We identified patients who had a total of 62 episodes of proved mechanical bowel obstruction and 38 episodes of adynamic obstruction through a computer search of medical records and radiographic files. On horizontal-beam abdominal radiographs of these patients, the presence and height of intestinal differential air-fluid levels were determined by the consensus of two experienced radiologists. These data were then statistically analyzed to determine the usefulness of differential air-fluid levels for distinguishing between mechanical and adynamic bowel obstructions. RESULTS: Plain films showed differential air-fluid levels in 32 (52%) of the 62 episodes of mechanical obstructions compared with 11 (29%) of the 38 adynamic obstructions, giving a sensitivity for mechanical obstruction of 0.52 and a specificity of 0.71. As the minimum significant height of differential air-fluid levels increased, specificity increased and sensitivity decreased. The positive predictive value also increased as differential air-fluid level heights increased, reaching a level of 0.86 or greater at 20 mm. CONCLUSION: The presence of differential air-fluid levels is an insensitive method of determining if a bowel obstruction is mechanical, because only a small proportion of mechanical obstructions have differential air-fluid levels. In our population of patients, however, a differential air-fluid level of 20 mm or greater was moderately suggestive that a bowel obstruction was mechanical in nature.
Authors: Carlo Vallicelli; Federico Coccolini; Fausto Catena; Luca Ansaloni; Giulia Montori; Salomone Di Saverio; Antonio D Pinna Journal: World J Emerg Surg Date: 2011-01-07 Impact factor: 5.469