Pneumoperitoneum and PEG Dislodgement Secondary to Noninvasive Ventilation after PEG Tube Placement.

We present a previously undescribed complication after noninvasive ventilation (NIV) for respiratory failure in a patient who required percutaneous endoscopic gastrostomy (PEG) tube placement for long-term nutrition after a complicated coronary bypass operation. A 54-year-old female diagnosed with unilateral vocal cord paralysis after emergent coronary artery bypass grafting (CABG) underwent an uncomplicated PEG tube placement. She was placed on intermittent NIV because of respiratory failure 24 hours after PEG placement, and NIV was continued for several days. Three days later, she was noted to have pneumoperitoneum on an upright chest X-ray. Abdominal CT scan revealed a large amount of pneumoperitoneum with the PEG tube in the correct position and no extravasation of enteric contrast from the stomach. Tube feeds were held and NIV was discontinued. Nonetheless, six days later, the patient was found on CT scan to have partial displacement of the PEG tube with leakage from the gastrotomy requiring operative repair. This case highlights the vulnerability of PEG tubes to management practices in the early postoperative period. Abdominal distention secondary to NIV likely caused pressure necrosis of the PEG site with dislodgement of the tube. This case elicits considerations regarding future management practices of patients receiving NIV in the early postoperative period after PEG placement.

GI bleeding in patients with continuous-flow left ventricular assist devices: a systematic review and meta-analysis.

BACKGROUND: Patients with left ventricular assist devices (LVADs) are at increased risk of GI bleeding (GIB), primarily from GI angiodysplastic lesions (GIAD). OBJECTIVE: To perform meta-analysis of the medical literature in order to determine prevalence and risk factors for GIB. DESIGN: A literature search was performed to identify studies reporting GIB in LVAD patients. We extracted rates of prevalence, rebleeding, and overall mortality from each study. Pooled event rates and odds ratios (ORs) with 95% confidence intervals (CIs) were calculated. SETTING: Meta-analysis of 17 case-control and cohort studies. PATIENTS: A total of 1839 LVAD patients of whom 1697 (92%) had continuous-flow LVADs. RESULTS: The pooled prevalence of GIB in LVAD patients was 23% (95% CI, 20.5%-27%). Subgroup analysis demonstrated that older age (standard difference in means (SDm), 0.69; 95% CI, 0.23-1.15), and elevated creatinine (SDm, 0.65; 95% CI, 0.12-1.18, P = .02) were associated with GIB. Risk factors not associated with GIB included LVAD as destination therapy (OR 1.85; 95% CI, 0.8-4.3), prior history of GIB (OR 2.22; 95% CI, 0.83-5.96), hypertension (OR 1.6; 95% CI, 0.87-2.97), and/or the presence of a continuous-flow LVAD (OR 4.5; 95% CI, 2.1-9.5). Recurrence of GIB occurred in 9.3% (95% CI, 7%-12%), with a GIB mortality rate of 23% (95% CI, 16%-32%). The pooled event rates were 48% (95% CI, 39%-57%) for upper GIB, 22% (95% CI, 16%-31%) for lower GIB, and 15% (95% CI, 8%-25%) for small-bowel bleeding. GIAD in the proximal GI tract were the most common cause of GIB (29%). LIMITATIONS: Lack of information regarding endoscopic therapy and follow-up in most studies. CONCLUSIONS: The prevalence of GIB is increased in patients with continuous-flow LVADs, primarily secondary to the presence of GIAD.

Significantly higher rates of gastrointestinal bleeding and thromboembolic events with left ventricular assist devices.

BACKGROUND & AIMS: The risk of gastrointestinal (GI) bleeding (GIB) and thromboembolic events may increase with continuous-flow left ventricular assist devices (CF-LVADs). We aimed to characterize GIB and thromboembolic events that occurred in patients with CF-LVADs and compare them with patients receiving anticoagulation therapy. METHODS: We performed a retrospective analysis of 159 patients who underwent CF-LVAD placement at 2 large academic medical centers (mean age, 55 ± 13 y). We identified and characterized episodes of GIB and thromboembolic events through chart review; data were collected from a time period of 292 ± 281 days. We compared the rates of GIB and thromboembolic events between patients who underwent CF-LVAD placement and a control group of 159 patients (mean age, 64 ± 15 y) who received a cardiac valve replacement and were discharged with anticoagulation therapy. RESULTS: Bleeding events occurred in 29 patients on CF-LVAD support (18%; 45 events total). Sixteen rebleeding events were identified among 10 patients (range, 1-3 rebleeding episodes/patient). There were 34 thrombotic events among 27 patients (17%). The most common source of bleeding was GI angiodysplastic lesions (n = 20; 44%). GIB and thromboembolic events were more common in patients on CF-LVAD support than controls; these included initial GIB (18% vs 4%, P < .001), rebleeding (6% vs none, P = .001), and thromboembolic events (17% vs 8%, P = .01). CONCLUSIONS: Patients with CF-LVADS receiving anticoagulants have a significantly higher risk of GIB and thromboembolic events than patients receiving anticoagulants after cardiac valve replacement surgery. GI angiodysplastic lesions are the most common source of bleeding.