Gastrostomy Versus Non-gastrostomy Enteral Access for Severe Traumatic Brain Injury

Traumatic brain injuries (TBIs) can be devastating injuries that result in significant morbidity and mortality. As part of optimal care for these patients, it is important to establish a plan for nutrition. These patients can have impairment with oral intake and may require naso-gastric access for a period to ensure adequate enteral nutrition. Patients that require prolonged naso-gastric nutrition due to neurologic dysfunction will have gastrostomy tubes placed to facilitate longer term enteral access while the patient is being rehabilitated. Establishing early enteral access to facilitate enteral nutrition has been shown to have a mortality benefit in TBI patients. ¹ As with most critically ill patients, administering enteral nutrition will stimulate the GI tract and preserve immunological function, decreasing the risk of infection and septic complications. ² Previous studies looking at oropharyngeal dysphagia in the TBI population have identified low admission Glasgow Coma Scale (GCS) score, low Ranchos Los Amigos (RLA) score, CT findings including evidence of midline shift or brain stem involvement, prolonged ventilatory support, and need for tracheostomy tube as predictors of prolonged oropharyngeal dysphagia. ³ As technology and treatment plans in the trauma population have advanced, low RLA score alone has been shown to be an independent predictor of long-term dysphagia and not necessarily an association with GCS score or length of ventilator use. ⁴ Deciding on enteral access in TBI patients is not without risk. There are complications associated with the different routes of enteral access. Naso-/oro-gastric (NG/OG) access can be associated with malposition, bleeding, dislodgement, aspiration, local skin irritation/necrosis, and discomfort. Percutaneous endoscopic gastrostomy (PEG) tubes have procedural risks as well as post-procedural risk that can include PEG site infection, inadvertent removal, hemorrhage, perforated viscus, and misplacement of the PEG. These varying complications can significantly prolong hospitalization and may require additional procedure/operative intervention to address.

This study was performed at a level 1 trauma center to evaluate which TBI pattern is more likely to have prolonged dysphagia and receive benefit from PEG tube placement. After review by our institutional review board (IRB) committee, this study was approved. A retrospective chart review of patients admitted to the trauma service at a level 1 trauma center with severe TBI (admission GCS of 8 or less) between January 2016 through December 2020 was performed. Patients were excluded if younger than 18 or older than 89, underwent a laparotomy, thoracotomy for any reason other than to address PEG complication, or underwent a surgical (open) gastrostomy tube placement. The pattern of TBI was determined by the official radiology report of the computed tomography (CT) Head from the initial trauma evaluation. TBI patterns were documented based on the presence of the various types of injury patterns documented in the CT Head report. Patients were determined to have multicompartment TBI if they had 2 or more differing types of TBI. Time until NG/OG removal, PEG removal, and initiation of oral intake was recorded in days. Patients that underwent swallowing evaluations with speech therapy or a per os (PO) challenge with documentation of failure was also recorded. Length of stay (LOS), intensive care unit (ICU) length of stay, and ventilator requirement were also documented in days. Statistical analysis was performed utilizing Google Sheets (Google, Mountainview, Ca.). Statistical tests include Welch’s t-test for unequal variances and Fisher’s test.

A total of 332 patients met the inclusion criteria. 268 patients were included in the non-PEG group and 64 in the PEG group. Average age of patients that did not undergo PEG was 48.14. Average age of patients that did undergo PEG was 36.7 (P<.05). Average admission GCS of patients that did not undergo PEG was 4.1, and 4.23 for PEG patients. The most common TBI documented for non-PEG patients was subdural hematoma (SDH) with 67.4% of patients. In PEG patients, SDH was the most common TBI with 76.5% patients. (Table 1). For the non-PEG group, 26% required operative/procedural intervention to address their TBI. The PEG group had 59% of their patients undergoing operative/procedural intervention (P<.05). The common intervention for TBI was craniectomy which was required in 41.9% of PEG patients and only 10.4% of non-PEG patients (P<.05). In our study population, 19.5% of patients underwent PEG placement with the average time interval of placement being 9.42 days. Percutaneous endoscopic gastrostomy patients remained on the ventilator for 15.5 days and non-PEG patients for 3.9 days (P<.05). 93.8% of PEG patients required tracheostomy placement compared to 3.7% of non-PEG patients. ICU length of stay was 5.0 days for non-PEG patients and 14.56 days for PEG patients (P<.05). Total LOS was 10.27 days on average for non-PEG patients and 44.7 days for PEG patients (P<.05). Non-PEG patients tolerated a diet at 6.9 days from trauma, and PEG patients tolerated PO 49.48 days from trauma (P<.05). 17.8% of non-PEG patients had at least 1 failed swallow evaluation/PO challenge. 78.1% of PEG patients had at least 1 failed swallow evaluation/PO challenge (P<.05). 78% of PEG patients were discharged with their PEG in place. 31.25% of PEG patients had documentation of PEG removal at an average of 98.65 days from placement. The complication rate for patients that underwent PEG placement was 18.75%. Inadvertent removal of the PEG tube occurred in 30% of the documented complications. The remaining complications in order of frequency included hemorrhage (16.6%), colonic placement (8.3%), and catheter malfunction (8.3%). Patients were discharged home in 16.8% of non-PEG patients and 7.8% of PEG patients. Percutaneous endoscopic gastrostomy patients were discharged to a skilled nursing facility in 21.8% of patients and 7.8% of non-PEG patients (P<.05). 45.3% of PEG patients were discharged to an inpatient rehab facility compared to 13.8% of non-PEG patients (P<.05). Mortality of non-PEG patients was 51.5% and mortality for PEG patients was 4.6% (P<.05).

OPEN IN VIEWER

Table 1.

TBI Patterns in PEG vs Non-PEG Patients. SDH = subdural hematoma; SAH = subarachnoid hematoma; EDH = epidural hematoma; IPH = intraparenchymal hematoma; IVH = intraventricular hematoma; x2 = 2 compartment TBI; x3 = 3 compartment TBI; x4 = 4 compartment TBI.

	SDH	SAH	EDH	IPH	IVH	x2	x3	x4
Non-PEG	67.4%	63.2%	7.1%	40.1%	14.4%	34.1%	26.1%	3.8%
PEG	76.5%	54.6%	12.5%	46.8%	10.9%	35.9%	23.4%	7.8%
P value	.15	.22	.16	.38	.47	.77	.65	.16

Predicting recovery in patients with severe TBI remains challenging as patients can have differing patterns of injury and the evolution of these TBIs can be variable. Percutaneous endoscopic gastrostomy placement is an excellent way to provide enteral nutrition in patients with prolonged dysphagia in the setting of severe TBI. In our study, patients most likely to have prolonged dysphagia requiring PEG placement due to severe TBI were younger, had prolonged ventilatory requirements, were more likely to undergo tracheostomy placement, and had at least one failed swallow evaluation/PO challenge. The PEG group was more likely to require operative intervention to address their TBI in comparison to the non-PEG patients (Table 2). We did not identify a statistical difference in the pattern of TBI or the presence of multicompartment TBI as a factor that led to prolonged dysphagia. Admission GCS did not demonstrate any difference between the two groups which may demonstrate severe TBIs are not all equal. This is further evidenced by our study demonstrating that patients that underwent PEG placement required more operative interventions to address their TBI. Mandaville et al identified older patients as one of the predictors of PEG placement and discharge from the hospital with PEG in place in patients with severe TBI; this differed from our study in that our population of PEG patients were younger than the non-PEG patients. There was a similarity in that tracheostomy requirement, and failed swallow evaluation was predictor of prolonged dysphagia. ⁴ Our study demonstrated that severe TBI patients that required PEGs tended to have a prolonged hospitalization as well as prolonged ICU level of care. With a vast majority of PEG patients also undergoing tracheostomy placement, it is reasonable to strongly consider placing PEG tubes in patients with anticipated prolonged ventilatory requirements. Given that our study is retrospective, no one clinical factor should be the sole determinator of placing a PEG tube; however if a patient has many of the factors identified in our study, strong consideration should be made in placing PEG tubes in the severe TBI population.

OPEN IN VIEWER

Table 2.

Comparison of Age, GCS, Ventilator Days, Length of Stay, ICU Length of Stay, Days Until PO Intake, Failed Swallow/PO Challenge, and Operative Intervention for TBI in Non-PEG vs PEG patients. Asterisk denotes statistical significance (P<.05).

	Non-PEG	PEG
Age*	48.1	36.7
GCS	4.1	4.2
Ventilator, days*	3.9	15.5
Tracheostomy placement*	3.7%	93.8%
Length of stay, days*	10.3	44.7
ICU length of stay, days*	5.0	14.6
Days until PO intake*	6.9	49.5
Failed swallow/PO challenge*	17.8%	78.1%
Operative intervention for TBI*	26.0%	59.0%