Cavalea Alexander C, Heidel Robert E, Daley Brian J, Lawson Christy M, Benton Darrell A, McLoughlin James M
Am Surg. 2017 Aug 1;83(8):825-831.
Pneumatosis intestinalis (PI) identified on computed tomography (CT) suggests an underlying pathology including bowel ischemia. Patients receiving tube feeds can develop PI, potentially requiring surgical intervention. We identify clinical factors in PI to predict those that may be safe to observe versus those that need immediate intervention. We retrospectively reviewed patients from a single institution from 2008 to 2016 with CT findings of PI and an enteric feeding tube. Patients who had not received tube feeds within one week of the CT were excluded. We analyzed clinical, operative, and outcome data to differentiate benign from pathologic outcomes. P values < 0.05 were set as significant. Forty patients were identified. We classified 24 as benign (no intervention) and 16 as pathologic (requiring intervention). A pathologic outcome was demonstrated for free fluid on CT [odds ratio (OR) = 5.00, confidence interval (CI) 1.23-20.30, P = 0.03)], blood urea nitrogen (BUN) elevation (OR = 8.27, CI 1.53-44.62, P = 0.01), creatinine (Cr) elevation (OR = 5.00, CI 1.27-19.62, P = 0.02), BUN/Cr ratio >30 (OR = 8.57, CI 1.79-40.98, P = 0.006), and vomiting/feeding intolerance (OR = 9.38, CI 1.64-53.62, P = 0.01). Bowel function within 24 hours of the CT, bowel dilatation (small ≥ 3 cm; large ≥6 cm), and lactic acidemia were not significant. Peritonitis was only seen in pathologic states, but this did not reach statistical significance (P = 0.06). This represents the largest single-center retrospective analysis of tube feeding-induced PI to date. The presence of free fluid on CT, BUN and Cr elevation, BUN/Cr >30, vomiting/feeding intolerance and peritonitis were predictive of a pathologic etiology of PI.
计算机断层扫描（CT）发现的肠壁积气（PI）提示存在包括肠缺血在内的潜在病理状况。接受管饲的患者可能会发生PI，可能需要手术干预。我们确定PI的临床因素，以预测哪些患者可以安全观察，哪些患者需要立即干预。我们回顾性分析了2008年至2016年来自单一机构的患者，这些患者的CT检查发现有PI且带有肠内喂养管。CT检查前一周内未接受管饲的患者被排除。我们分析了临床、手术和结局数据，以区分良性结局和病理性结局。P值<0.05被设定为具有统计学意义。共确定了40例患者。我们将24例归类为良性（无需干预），16例归类为病理性（需要干预）。CT显示有游离液体（比值比[OR]=5.00，置信区间[CI]1.23 - 20.30，P = 0.03）、血尿素氮（BUN）升高（OR = 8.27，CI 1.53 - 44.62，P = 0.01）、肌酐（Cr）升高（OR = 5.00，CI 1.27 - 19.62，P = 0.02）、BUN/Cr比值>30（OR = 8.57，CI 1.79 - 40.98，P = 0.006）以及呕吐/喂养不耐受（OR = 9.38，CI 1.64 - 53.62，P = 0.01）均提示病理性结局。CT检查后24小时内的肠道功能、肠扩张（小肠≥3 cm；大肠≥6 cm）和乳酸血症无统计学意义。腹膜炎仅见于病理性状态，但未达到统计学意义（P = 0.06）。这是迄今为止关于管饲引起的PI的最大规模单中心回顾性分析。CT显示有游离液体、BUN和Cr升高、BUN/Cr>30、呕吐/喂养不耐受以及腹膜炎可预测PI的病理性病因。
