Undetected oesophageal perforation and feeding-tube misplacement

41 downloads 0 Views 759KB Size Report
and feeding-tube misplacement. Early enteral nutrition in critical illness is associated with reduced complications and mortality (Heyland,. 2013). Delayed gastric ...
Undetected oesophageal perforation and feeding-tube misplacement Stephen J Taylor, Catherine Ross and Timothy Hooper

Abstract

This is a case report of an electromagnetically (EM)-guided Cortrak feeding tube that perforated the lower oesophagus and was neither detected by the EM trace nor plain X-ray. Misplacement was diagnosed from computed tomography (CT) following injection of radio-contrast down the tube. Recommendations are offered for use of the EM trace in patients at high risk of oesophageal perforation. Key words: Cortrak ■ Misplacement ■ Nasointestinal tube ■ Nasogastric tube

E

arly enteral nutrition in critical illness is associated with reduced complications and mortality (Heyland, 2013). Delayed gastric emptying (DGE) is common, occurring in 30–70% patients (Columb et al, 1992; Nguyen et al, 2007), and may be overcome with nasointestinal (NI) feeding, unless ileus supervenes (Taylor et al, 2010a). EM-guidance is used in preference to blind placement because success rates are higher; confirmation is in real-time, thereby avoiding lung damage (Taylor, 2014); and it is as accurate as plain X-ray (Powers et al, 2013; Taylor et al, 2014), thus enabling rapid NI tube placement and a higher percentage of the feed goal being retained (Taylor et al, 2010a).

Medical care Trust policy is that any patient with DGE, defined as a gastric residual volume >250  mL or vomiting, is started on 10 mg metoclopramide IV three times daily. If this fails, after 24  hours an NI tube is placed when possible. The Cortrak EM-guided NI tube system is a conventional Corflo feeding tube, but with a guide-wire incorporating an EM transmitter and a receiver unit placed over the xiphisternum, both connected to a computer display. During tube insertion, the receiver picks up the EM signal. The signal is graphically displayed on a computer screen showing anterior-posterior (AP) and depth (cross-section or lateral) paths taken by the feeding tube as it proceeds down the oesophagus, through the

stomach and into the small intestine. A 56-year-old woman was admitted to a gastroenterology ward with jaundice and abdominal pain. She reported drinking one and a half bottles of vodka per week and subsequent investigation diagnosed hepatic cirrhosis, oesophageal varices, acute kidney injury, coagulopathy and possible spontaneous bacterial peritonitis. She was started on prednisolone and supplementary nasogastric tube feeding. After 5  days, she became hypotensive and required terlipressin, underwent ascitic drainage, and was started on piperacillin-tazobactam to treat the presumed spontaneous bacterial peritonitis. Later that day, she developed haematemesis, became haemodynamically unstable, and was transferred to the intensive care unit (ICU). On admission to ICU, three of four oesophageal varices were endoscopically banded. On ICU day 5, an NGT was placed and feeding started. However, the woman had an asystolic cardiac arrest requiring cardiopulmonary resuscitation, and upper GI bleeding necessitated placement of a Sengstaken-Blakemore tube. Endoscopy showed no active bleeding and the patient underwent tracheostomy on day 6 and endoscopic NGT placement on day 7. DGE developed despite metoclopramide treatment and an NI tube was placed day 9. On day 10, haematemesis again required endoscopic intervention. This caused loss of the NI tube. Another Sengstaken-Blakemore tube was placed and, on day 12, endoscopy showed persistent oesophageal ulcers and a bleeding ulcer on the greater gastric curvature. The latter was injected with adrenaline, an IV proton-pump inhibitor started, and Cortrak NI and NG tubes were placed. The NI tube was found to be misplaced and was therefore removed and nutrition given via total parenteral nutrition. Despite supportive treatment and antimicrobials for confirmed vancomycin-resistant enterococcal bacterial peritonitis, the woman acutely deteriorated on day 26 with reduced consciousness level, worsening metabolic acidosis and increasing oxygen requirements. After discussion with her family, a decision was made to focus on end-of-life care and the woman died later that day.

Accepted for publication: June 2014

16

Two NG and two NI tubes were placed using the EM trace in addition to two blindly placed NG tubes and one endoscopically placed NG tube. The first EM-guided NG tube placement was aborted because of resistance at the level of the gastro-oesophageal junction on day 1 of the ICU admission.The first NI tube placement (day 3) was uneventful (Figure 1). However, during the second NI tube placement,

British Journal of Nursing, 2014, Vol 23, No 19

© 2014 MA Healthcare Ltd

Tube placements Stephen J Taylor is Research Dietitian, Department of Nutrition and Dietetics; Catherine Ross is Core Trainee in Anaesthesia and Intensive Care; and Timothy Hooper is Consultant in Anaesthesia and Intensive Care, Department of Anaesthetics, Southmead Hospital, Bristol

??????????? there was some resistance to advancement beyond the lower oesophagus. Tube advancement was mostly done with the guide-wire partially withdrawn, creating a flexible tip to reduce trauma risk. The EM trace ultimately described the expected path from the oesophagus (AP: vertical; depth: deep), through the stomach (AP: clockwise circuit; depth: becoming shallow) and was consistent with entry into duodenum part-1 (AP: crossed midline, began to move up then down in an anti-clockwise path; depth: deeper). However, further advancement was impossible. Within 24 hours (day 13) The woman developed abdominal distension and an EM-guided NGT was placed, without resistance, to the superior duodenal flexure, and was then withdrawn into the gastric body. A chest X-ray was requested to assess the distension and showed free air under the diaphragm; the NIT was interpreted as being in duodenum part-2 (Figure 1). Surgical review suggested gastric ulcer perforation that, in light of the woman’s comorbidities and ongoing haemodynamic instability, required conservative treatment with TPN and free drainage of both feeding tubes. On ICU day 20, an abdominal computed tomography (CT) scan following injection of radio-contrast down the NI tube showed this tube exiting the lower oesophagus and passing behind the stomach within the peritoneal cavity, as shown on an oblique X-ray (Figure 2). The path followed by the misplaced NI tube approximated a path following the greater gastric curvature on both EM-trace and plain X-ray, suggesting that the tube tip was situated at the superior duodenal flexure and duodenum part-2, respectively. On review of previous images, two discrepancies were noted: ■■ Cortrak: although the AP EM trace crossed the centreline, the ‘false’ superior duodenal flexure was far less pronounced than on the NI_1 or NG_2 ■■ X-ray: while the tube appeared to follow the greater gastric curve, it turned down, as if into the duodenum, left, rather than right, of centreline. However, interpretation of these images may have been complicated by patient rotation and the presence of ascitic fluid.

© 2014 MA Healthcare Ltd

Discussion and recommendations Neither an EM trace nor plain X-ray can necessarily detect a tube misplaced in the peritoneum when it has perforated the oesophagus and follows the greater gastric curvature in both the AP and depth paths. At least one other intra-peritoneal placement has occurred after oesophageal perforation with a Cortrak tube (Food and Drug Administration (FDA), 2007), but observer error is likely in this case as the EM trace should have been able to warn that the tube was in the pelvis. Intraperitoneal misplacement is rare and the EM trace is likely to be detected as abnormal when interpreted by an expert. However, we recommend extra precautions when patients have a friable oesophagus and these include: ■■ Determine whether risk-benefit makes the procedure worthwhile compared with endoscopic or fluoroscopic tube placement. ■■ If the deepest point of placement is less than duodenum part-3, confirm position with an X-ray after injecting radio-contrast down the feeding tube. ■■ When tubes have reached duodenum part-3 or beyond, an

British Journal of Nursing, 2014, Vol 23, No 19

Figure 1. Most distal GI position of placement: a) NI_1, b) NI_2, c) NG_2

expert should review the EM trace. It should be extremely difficult for a tube actually within the peritoneum to mimic an EM trace that has been interpreted as being in duodenum part-3 or beyond. Where an EM trace exists from a tube previously confirmed to be correctly placed, an expert should look for discrepancies that might warn of misplacement, taking into account changes in the patient or their position. Putting this incident into context, 1.5% of blindly placed tubes are misplaced, 0.5% result in pneumothorax or pneumonia and 0.27% of patients die following tube placement (Taylor, 2014). Most of these complications occur during placement, not because misplacement remained

17

Nguyen NQ, Ng MP, Chapman M, Fraser RJ, Holloway RH (2007) The impact of admission diagnosis on gastric emptying in critically ill patients. Crit Care 11(1): R16 Powers J, Fischer MH, Ziemba-Davis M, Brown J, Phillips DM (2013) Elimination of radiographic confirmation for small-bowel feeding tubes in critical care. Am J Crit Care 22(6): 521–7. doi: 10.4037/ajcc2013755 Taylor SJ (2014) Cortrak tube placement: Advanced training. http://www. nutritionsupport.info. Silhouette Publications, UK. ISBN 978-0-95745583-2 Taylor S, Manara A, Brown J (2010a) Treating delayed gastric emptying in critical illness: Metoclopramide, erythromycin and bedside (cortrak) nasointestinal tube placement. JPEN J Parenter Enteral Nutr 34(3): 289–94. doi: 10.1177/0148607110362533 Taylor, S, Manara A, Brown J (2010b) Nasointestinal placement versus prokinetic use when treating delayed gastric emptying in ICU patients. British Journal of Intensive Care 20: 38–44 Taylor SJ, Allan K, McWilliam H, Manara A, Brown J, Toher D, Rayner W (2014) Confirming nasogastric tube position: Electromagnetic tracking versus pH or X-ray and tube radio-opacity. Br J Nurs 23(7): 352, 354–8

Figure 2. Plain X-ray (a) and oblique X-ray (b)

undetected. For this reason, expert use of an EM trace or another bedside-guided technique offer an opportunity to reduce complication risk (Taylor et al, 2014). In addition, NI feeding appears to be more effective than prokinetic drugs in achieving feed tolerance in patients with DGE (Taylor et al, 2010a) and, for NI placement, the EM trace achieves similar success rates to endoscopic and fluoroscopic placement (Foote, 2004; Holzinger et al, 2011) but will reduce the cost and delay to placement (Taylor et al, 2010b). These precautions are additional safeguards for special circumstances, not contraindications to use of the EM trace. BJN  Conflict of interest: Stephen Taylor worked on a Corpak consultation committee once in 2007. Corpak funded the time and equipment for an unrelated audit, but played no part in study design, execution, analysis or reporting of results. Columb MO, Shah MV, Sproat LJ, Sherratt MJ, Inglis TJ (1992) Assessment of gastric dysfunction: current techniques for the measurement of gastric emptying. British Journal of Intensive Care 2: 75–6, 78, 80 Food and Drug Administration (2007) MAUDE Adverse Event Report. http:// tinyurl.com/qjrjdd5 (accessed 6 October 2014) Foote JA, Kemmeter PR, Prichard PA, Baker RS, Paauw JD, Gawel JC, Davis AT (2004) A randomized trial of endoscopic and fluoroscopic placement of postpyloric feeding tubes in critically ill patients. JPEN J Parenter Enteral Nutr 28(3):154–7 Heyland D (2013) Canadian Clinical Practice Guidelines 2013. www. criticalcarenutrition.com (accessed 6 October 2014) Holzinger U, Brunner R, Miehsler, W, Herkner H, Kitzberger R, Fuhrmann V, Metnitz PGH, Kamolz L-P, Madl C (2011) Jejunal tube placement in critically ill patients: a prospective, randomized trial comparing the endoscopic technique with the electromagnetically visualized method. Crit Care Med 39(1): 73–7. doi: 10.1097/CCM.0b013e3181fb7b5f

KEY POINTS n Cortrak

electromagnetic trace is effective in guiding nasointestinal tube placement. However, in patients with a friable oesophagus, extra precautions are needed to confirm tube position

n Detecting

intraperitoneal misplacement may necessitate injection of radio-contrast before X-ray bedside nasointestinal tube placement is advantageous, but where risk of oesophageal perforation is high, the risk vs benefit of endoscopic or fluoroscopic placement should be considered

© 2014 MA Healthcare Ltd

n Rapid

n Expert

review and comparison with previous electromagnetic traces can help spot misplacements

18

British Journal of Nursing, 2014, Vol 23, No 19