Potential for health care cost savings with preoperative gastrostomy tube placement in the head and neck cancer population

Abstract

Background:

The purpose of this study was to examine the cost differences between preoperative and postoperative placement of gastrostomy tubes (G-tubes) in patients with head and neck cancer.

Methods:

We conducted a retrospective chart review of patients with aerodigestive tract cancers from 2010 to 2015. Data included inpatient and postdischarge costs, demographics, tumor characteristics, surgical treatment, length of stay (LOS), time spent in the intensive care unit (ICU), and readmissions.

Results:

Five hundred ninety patients were included in this study. There was a $7624 inpatient cost savings (P = .002) for those G-tubes placed preoperatively ($26 060) versus postoperatively ($33 754). Postdischarge costs did not differ significantly between groups (P = .60). There was a $9248 total costs savings (P = .009) for those patients with G-tubes placed preoperatively ($39 751) versus postoperatively ($48 999), despite patients with preoperative G-tubes having lower body mass index (BMI; P = .009), higher Association of Anesthesiologist (ASA) class (P = .02), more preoperative radiation (P < .001), and more free tissue transfer reconstruction (P = .007).

Conclusion:

There is potential for savings by placing G-tubes preoperatively, possibly driven by decreased LOS, despite data suggesting that patients with G-tubes placed preoperatively are higher risk.

1 |. INTRODUCTION

Head and neck cancer is one of the most expensive cancers to treat, with studies showing the highest expenditure within the first year after diagnosis.^1–3 The high costs across the healthcare system can be attributed to high incremental costs of care and increasing prevalence.^4–8 Cancer care has been found to consume nearly 50% of all healthcare expenditures.⁹ Given these demands, institutions are looking more frequently at measures to increase cost efficiency due to growing concerns about healthcare sustainability.¹⁰

In the patients with head and neck cancer population, gastrostomy tubes (G-tubes) are an important consideration when disease burden or treatment impairs effective swallowing and maintenance of nutrition. It is important that the placement of G-tubes not only be appropriately timed for patients’ nutrition needs but also in a cost-efficient manner to facilitate resourceful healthcare delivery. The purpose of this study was to determine costs of care differences between placement of G-tubes before surgical resection versus postoperatively in an effort to identify an area for potential cost savings.

2 |. MATERIALS AND METHODS

A retrospective review of patient charts from the Wake Forest Baptist Health Otolaryngology - Head and Neck Oncology clinic was performed. Patients were identified based on a comprehensive database of all surgical procedures performed by the 3 Head and Neck Oncology faculty between the dates January 1, 2010, and May 31, 2015, with the International Classification of Disease-9 codes 140.0–149.9 and 160.0–162.0. This database was compiled and released by the medical records department after institutional review board approval was obtained. Each patient chart from this database was screened for participation in this study.

The eligibility criteria included: all patients who underwent surgical resection for head and neck upper aerodigestive tract squamous cell cancers that required G-tubes preoperatively or postoperatively. We excluded those patients who had G-tubes placed > 3 months after the resection or placed prophylactically due to anticipated effects of adjuvant therapy; these G-tubes were considered to have been placed due to factors other than the disease or effects of surgery. Our goal was to include only those patients who had G tubes placed secondary to detrimental nutritional decline preoperatively and who required persistent enteral feeding needs related to surgery (ie, prolonged NPO time for wound healing, wound issues etc). The G-tubes that were placed after 3 months of surgery were deemed unlikely to have been placed due to the direct effects of surgery and, therefore, were excluded. Patients who solely underwent resection of neck nodal disease without primary site resection or those patients whose primary tumor site was not in the upper aerodigestive tract (eg, skin, parotid, or thyroid) and those with insufficient clinical data were also excluded. Patients who died within 6 months of the resection, as their total cost of care was unable to be assessed during our defined postoperative time period, were also excluded.

Using the electronic medical records, patient charts were screened for demographic characteristics, including age, sex, and body mass index (BMI). Tobacco use (oral or inhaled), heavy alcohol use (>2 drinks per day), medical comorbidities, American Society of Anesthesiology (ASA) class, and history of preoperative radiation to the tumor site were included. The ASA class is a subjective assessment of a patient’s overall health that is based on 5 classes: (1) the patient is a completely healthy and fit patient; (2) the patient has mild systemic disease; (3) the patient has severe systemic disease; (4) the patient has severe systemic disease that is a constant threat to life; and (5) the patient is not expected to survive. At our institution, the ASA class is documented on every patient preoperatively during the medical clearance visit by our anesthesia colleagues and, thus, was used in this study as a surrogate for overall health.

Tumor characteristics, such as TNM classifications were also recorded. Surgery information, such as the type of reconstruction, and timing of the G-tube placement (or lack thereof) were included. The indications for placement of a G-tube in the postoperative period were based on the combined assessments of the surgeon and speech-language pathologist as to whether they predicted a prolonged recovery of swallowing. Although not all patients had postoperative swallowing evaluations (ie, modified barium swallow or functional endoscopic evaluation of swallowing), evidence of aspiration on these studies certainly assisted the team in determining whether a G-tube was necessary. In general, the surgeons and speech pathologists recommend G-tubes in the setting of gross aspiration with poor adaptation and management of secretions and in settings in which adequate PO cannot be obtained. Most patients with preoperative G-tubes had them placed before presentation at our institution but those who did receive them under our guidance had them placed due to severe malnutrition preoperatively when it was felt that preoperative optimization was necessary. The length of stay (LOS) and time in the intensive care unit (ICU) during the primary resection hospitalization were also recorded. Complications during the inpatient stay and postdischarge were recorded and defined as either wound or nonwound related. Wound-related complications were defined as those that directly involved the surgical site, such as flap failure, fistula, and donor site breakdown. The nonwound-related complications were defined as those that did not directly involve the surgical site, such as myocardial infarction or colitis. Hospital readmissions in the 6-month postdischarge period were also recorded.

True inpatient costs included all costs associated from the time of admission for surgery to the time of discharge. This included G-tube placement costs for those patients who received postoperative G-tubes during their inpatient hospital stay. Charges billed from our institution served as the basis for our cost analysis. For the purpose of this study, the costs of the primary surgical resection inpatient hospital stay minus the cost of G-tube placement (when appropriate) will henceforth be termed “inpatient costs.”

In order to perform a comprehensive 6-month postdischarge cost analysis, several factors were included. Hospital readmissions and emergency department visits within 6 months after discharge for indications deemed directly related to the surgery or the disease process were included. For example, an emergency department visit for neck swelling related to a fistula would be included, whereas an emergency department visit for chest pain would not. Routine outpatient care for 6 month’s postdischarge, including clinic visits, scans, home health services, or other ancillary services, such as swallow evaluations, were included. The costs of the 6-month postdischarge period minus the cost of G-tube placement (when appropriate) will henceforth be termed “postdischarge costs.”

In order to control for G-tube costs between groups, the average G-tube placement cost was calculated using institutional data for surgical and perioperative costs and incorporated in the costs for those patients whose G-tubes were performed at external institutions in which specific cost data were not available. The Current Procedural Terminology codes 49440, 49441, 49442, 43653, 43246, and 43830 were used to isolate outpatient G-tubes over a 5-year time period, from 2010 to 2015. Only perioperative and procedural costs were included. Perioperative costs included operating room time, operating room supplies, anesthesia labor and supplies, and postanesthesia recovery. Procedural costs included procedure time and supplies. The average G-tube cost at our institution was found to be $2772. The term “G-tube costs” will be used to reference costs specific only to G-tube placement (Table 1). At our institution, patients who undergo G-tube placement in the outpatient setting are admitted to our observation day hospital postprocedure for education and initiation of feeding. This is not considered a full inpatient admission; instead, it is a 23-hour observation admission.

TABLE 1.

Charge breakdown for each primary cost category

Inpatient costs	6-mo postdischarge costs	Gastrostomy-tube placement costs
Perioperative costs of surgical resectiona	Emergency department visits	Perioperative costs of tube placementa
Procedural costs of surgical resectionb	Hospital readmissions	Procedural costs of tube placementb
Bed unit	Routine expected outpatient care (clinic visits, imaging, home health services, or other ancillary services such as swallow evaluations)	Bed unit
Imaging		Imaging
Laboratory results		Laboratory results
Pharmacy		Pharmacy
Therapy		Therapy

^aPerioperative costs include operating room time, operating supplies, anesthesia labor and supplies, and postanesthesia recovery.

^bProcedural costs include procedure time and supplies.

The term “total costs” will be used to reference the inpatient costs plus the postdischarge costs, excluding the G-tube placement costs. All cost encounters were screened to ensure they were appropriate for inclusion in the analysis.

2.1 |. Statistical methods

Descriptive statistics for all patients were generated for all measures, including means, SDs, and 95% confidence intervals (CIs) for continuous measures and frequencies and proportions for categorical variables. Next, 2-sample t tests and the Fisher’s exact tests were performed to examine the relationship between each of the individual patient measures and the 2-level variable for G-tube (preoperative G-tube and postoperative G-tube). Finally, a stepwise linear regression model was fit to identify risk factors to predict total costs separately for preoperative and postoperative G-tubes. All analyses were performed using SAS 9.3 (Cary, NC).

2.2 |. Sample size and power

The primary comparisons in these analyses focus on comparing the 184 participants across the 2 G-tube groups (preoperative = 73 patients, and postoperative = 111 patients); thus, we can estimate the power to detect differences between these groups for mean values of total cost. Based on these data, there would be 80% power to detect a difference between groups equivalent to 0.426 SDs (ie, an effect size of 43%), assuming a 2-sided 2-sample t test with alpha = 0.05.

3 |. RESULTS

Five hundred ninety patients underwent surgical resections during the study period. Sixty-nine percent of the patients did not require a G-tube, whereas 13% required preoperative G-tubes and 18% required postoperative G-tubes. Of those patients who received preoperative G-tubes (n = 73), 11 received them in the inpatient setting before resection and 62 had them at a prior admission or at a institution before resection. Of those patients who received postoperative G-tubes (n = 111), 55 had them placed in the inpatient setting and 56 had them placed in the outpatient setting.

Those patients who required a preoperative G-tube as an inpatient resection had an LOS of 3.2 days (P < .001). This group also had lower BMIs (P = .009) and higher ASA class (P = .02) and were more likely to have had preoperative radiation (P < .001). They were also more likely to require free flaps or pedicled rotation flaps for reconstruction versus primary closure or skin graft closure (P = .007). All other patient demographics, disease characteristics, comorbidities, and treatments were comparable in both the preoperative and postoperative G-tube groups (Tables 2 and 3)

TABLE 2.

Analysis of pretreatment and treatment-related variables between preoperative and postoperative gastrostomy-tube groups

Variables	Preoperative G-tubes	Postoperative G-tubes	P value	No G-tube
Age, years	60	61	.66	63
BMI	23	26	.009	27
Sex	F: 12 (17%)	F: 29 (26%)	.15	F: 119 (29%)
	M: 60 (83%)	M: 82 (74%)		M: 287 (71%)
Tobacco use	No: 15 (21%)	No: 15 (14%)	.23	No: 102 (25%)
	Yes: 58 (79%)	Yes: 96 (86%)		Yes: 304 (75%)
Heavy alcohol use	No: 57 (79%)	No: 89 (80%)	.69	No: 337 (83%)
	Yes: 15 (21%)	Yes: 22 (20%)		Yes: 69 (17%)
ASA class	1: 0	1: 0	.02	1: 4 (1%)
	2: 1 (1%)	2: 11 (10%)		2: 92 (23%)
	3: 60 (82%)	3: 91 (82%)		3: 275 (68%)
	4: 12 (17%)	4: 9 (8%)		4: 35 (8%)
T classification	T1: 13 (17%)	T1: 14 (12%)	.20	T1: 158 (39%)
	T2: 15 (21%)	T2: 31 (28%)		T2: 106 (26%)
	T3: 22 (30%)	T3: 24 (22%)		T3: 64 (16%)
	T4: 23 (32%)	T4: 42 (38%)		T4: 78 (19%)
N classification	0: 45 (62%)	0: 59 (54%)	.25	0: 281 (69%)
	1: 3 (4%)	1: 13 (12%)		1: 39 (10%)
	2: 25 (34%)	2: 37 (33%)		2: 85 (21%)
	3: 0	3: 1 (1%)		3: 0
Preoperative radiation	No: 29 (40%)	No: 83 (75%)	< .001	No: 343 (84%)
	Yes: 44 (60%)	Yes: 27 (25%)		Yes: 63 (16%)
Reconstruction type	Primary closure: 14 (19%)	Primary closure: 35 (32%)	.007	Primary closure: 235 (58%)
	STSG: 0	STSG: 2 (2%)		STSG: 19 (5%)
	Free flap: 32 (44%)	Free flap: 56 (50%)		Free flap: 93 (23%)
	Pedicled rotation flap: 27 (37%)	Pedicled rotation flap: 18 (16%)		Pedicled rotation flap: 58 (14%)

Abbreviations: ASA, American Society of Anesthesiology; BMI, body mass index (calculated as weight in kilograms divided by height in meters squared); G-tubes, gastrostomy tubes; STSG, split-thickness skin graft.

Note: For comparison, data for patients without G-tubes was also included. The P value represents comparison between preoperative and postoperative G-tube placement.

TABLE 3.

Analysis of posttreatment variables between preoperative and postoperative gastrostomy-tube groups

Variables	Preoperative G-tubes	Postoperative G-tubes	P value	No G-tube
ICU time	2.2 d	3.0 d	.09	1.2 d
LOS	9.4 d	12.6 d	< .001	5.9 d
Any complication (wound + nonwound)	No: 18 (25%)	No: 28 (25%)	.93	No: 205 (50%)
	Yes: 55 (75%)	Yes: 83 (75%)		Yes: 201 (50%)
Wound complications	No: 40 (55%)	No: 53 (48%)	.37	No: 277 (68%)
	Yes: 33 (45%)	Yes: 58 (52%)		Yes: 128 (32%)
Nonwound-related complications	No: 37 (51%)	No: 45 (41%)	.23	Yes: 141 (35%)
	Yes: 36 (49%)	Yes: 66 (59%)		No: 265 (65%)
Hospital readmission	No: 56 (77%)	No: 83 (75%)	.86	No: 367 (90%)
	Yes: 17 (23%)	Yes: 28 (25%)		Yes: 39 (10%)

Abbreviations: G-tubes, gastrostomy-tubes; ICU, intensive care unit; LOS, length of stay.

Note: For comparison, data for patients without G-tubes were also included. The P value represents comparison between preoperative and postoperative G-tube placement.

The average inpatient costs between both groups was $30 685 ± $2251 (95% CI $27 401-$33 969). The average postdischarge costs between both groups was $14 819 ± $23 871 (95% CI $11 346-$18 290]. The average total costs between both groups was $45 329 ± $29 943 (95% CI $40 974-$49 685).

There was an average savings of $9248 in total costs (P = .009) for those patients who had G-tubes placed preoperatively ($39 751) versus postoperatively ($48 999). There was an average inpatient cost savings of $7694 (P = .002) for those patients who had G-tubes placed preoperatively ($26 060) versus postoperatively ($33 754). Postdischarge costs did not differ significantly between the preoperative ($13 690) and postoperative ($15 561) G-tube groups (P = .60; Table 4, Figure 1).

TABLE 4.

Cost differences between preoperative and postoperative gastrostomy-tube groups

Study period	Preoperative G-tube costs (95% CI)	Postoperative G-tube costs (95% CI)	Cost difference	Difference	P value
Inpatient costs	$26 060 ± $21 155 ($21 124–$30 996)	$33 754 ± $22 954 ($29 416–$38 091)	$7694 savings	23%	.002
Postdischarge costs	$13 690 ± $24 301 ($8020–$19 359)	$15 561 ± $23 666 ($11 109–$20 012)	$1871 savings	12%	.60
Total costs	$39 751 ± $29 168 ($32 945–$46 556)	$48 999 ± $30 007 ($43 354–$54 643)	$9248 savings	19%	.009

Abbreviations: 95% CI, 95% confidence interval; G-tubes, gastrostomy-tubes.

FIGURE 1.

Cost comparison between preoperative and postoperative gastrostomy tube (G-tube) placement groups. The figure highlights the increased costs associated with patients receiving G-tubes in the postoperative period, across the study periods. Differences in inpatient costs and total costs were statistically significant

Models were created for both groups to predict total costs. For the preoperative G-tube group, the LOS was noted to be the strongest predictor (P < .001; Table 5). The LOS was also the strongest predictor for the postoperative G-tube group (P < .001); followed by ICU time, ASA class, sex, and reconstruction type (Table 6).

TABLE 5.

Predictive model for total costs of care in the preoperative gastrostomy-tube group

Predictor	Coefficient	P value	Interpretation
LOS	2251	< .001	$2251 for each additional daya
Hospital readmission	15 012	.04	$15 012 for each readmission

Abbreviation: LOS, length of stay.

^aThe intercept for comparison is $12 094. To interpret an intercept, those patients with LOS of 1 day and no hospital readmission would be expected to cost $12 094. Changes in cost then vary based on longer LOS and presence of hospital readmission.

TABLE 6.

Predictive model for total costs of care in the postoperative gastrostomy-tube group

Predictor	Coefficient	P value	Interpretation
LOS	1442	< .001	$1442 for each additional daya
ICU time	2308	< .001	$2308 for each additional day
ASA class	ASA 3: −15490 ASA 4: −24013	.03	From the intercept, subtract $15 490 for ASA 3 patients, and $24 013 for ASA 4 patients
Sex	−11 846	.03	Females cost $11 846 more than males
Reconstruction type	STSG: −28 270, Free flap: −10 345, Pedicled rotation flap: −11 353	.05	From the intercept, subtract $28 270 for STSG closure patients, $10 345 for free flap patients, and $11 353 for pedicled rotation flap patients

Abbreviations: ASA, American Society of Anesthesiology; ICU, intensive care unit; LOS, length of stay; STSG, split-thickness skin graft.

^aThe intercept for comparison is $42 959 with the reference for ASA class being ASA 2 and the reference for reconstruction type being primary closure. To interpret an intercept, those patients with LOS of 1 day, ICU time of 0 days, ASA class 2, female sex, and primary closure would be expected to cost $42 959. Changes in cost then vary based on differences in these variables.

4 |. DISCUSSION

The development of improved therapies for cancer treatment has led to increasing cancer survival with estimates of approximately 14.5 million cancer survivors in the United States in 2014, with this number expected to reach 19 million as of 2024.¹¹ The annual medical cost of cancer care was estimated to be $124.6 billion in 2010 and is projected to increase to $157.8 billion by 2020.¹² To account for this increase and to account for patients whose complex care requirements constitute the bulk of healthcare costs, our healthcare system has changed the payment structure in the form of bundling and pay-for-performance systems.^13,14 In turn, many third party payers have shifted these costs onto the patients in the form of heightened premiums and copayments.¹⁵ Therefore, a move toward reduction of costs of care where possible without compromising overall outcomes is an important process to ensure a successful oncologic practice.

Healthcare cost analyses, much like this study, have become a more prevalent part of our oncologic literature because of the recent economic and financial pressures put on the healthcare system. Some studies have shown that the treatment of hypopharyngeal and esophageal cancers has some of the highest costs of care among cancers.^2,6 The study by de Oliveira et al⁶ (2013) found that esophageal cancers have the highest postdiagnosis costs ($50 620), the highest cost for hospital admissions among all cancers ($27 506), and the highest follow-up costs evidenced by the high costs for physician services ($4757) and home care ($4058). They attributed this cost to the expense of multiple procedures, such as esophageal dilation or repeated biopsies.⁵ Efforts to reduce costs wherever possible are now all the more imperative, thus, the relevance of healthcare cost analyses, such as our own.

In this study, we attempted to define any potential healthcare cost savings by placing G-tubes preoperatively versus postoperatively. To our knowledge, no cost analysis of this type has previously been performed in the literature. Our findings suggest that G-tube placement preoperatively can result in considerable cost savings (19% savings) and seems to be driven most strongly by a reduction in the costs associated with the inpatient stay. There was a 3.2-day LOS difference between groups, which drove an inpatient cost savings of 23%, a finding also evidenced by our predictive model. Multiple other studies have recognized a similar relationship between increased LOS and increased healthcare costs.^16–19 At our institution, we routinely find that coordinating placement of a G-tube during the inpatient hospital stay creates considerable delays. Specific causes of those delays have ranged from coordinating placement with other services to teaching the patient and caregiver about G-tube care. If delays could be avoided by preoperative placement, this may expedite discharge and facilitate cost savings. Efforts to reduce LOS and facilitate efficient discharge may be a key initiative to reducing healthcare costs in our changing care climate.

These cost figures represent the experience of a single institution’s practice. We recognize that the practice setting and practice patterns of other institutions may differ, making generalization of our findings with other practices somewhat difficult. For example, important drivers of cost in our sample were LOS, readmissions, and ICU stay; however, other practice models may not have such high prevalence of these factors. It is also important to recognize that the reasoning behind placement of G-tubes is multifactorial. In situations that will require prolonged NPO time, such as in the setting of a free tissue transfer in previously radiated patients, wound breakdown, and fistula, a G-tube may be placed to reduce the discomfort and poor cosmesis associated with a long-term nasal feeding tube. In situations in which patients exhibit gross aspiration or inadequate oral intake postoperatively, a G-tube may be placed to allow for full swallow rehabilitation and/or airway protection. Most, if not all, of our patients are seen by our speech-language pathologists postoperatively and concerns that are raised based on their objective swallow measures are communicated to the attending surgeon. It is our group’s practice to discharge patients with nasal feeding tubes and planned swallow rehabilitation versus prolonging inpatient stay to ensure full PO diet before discharge. We argue that recognition of the complex nature of swallowing and nutrition management in the patients with head and neck cancer is important to reduce inevitable delays that can come from delaying enteral nutrition in certain patients.

When thinking about frequent causes of increased costs of care, the reader likely would expect that patients with more comorbidities or larger disease burden may have higher costs secondary to increased propensity to poor outcomes or complications. Interestingly, our findings showed that patients who had preoperative G-tubes had lower costs of care despite having lower BMI, higher ASA class, and a history of treatment (to include, at a minimum, preoperative radiation). In addition, this group required more free flap or pedicled rotation flap reconstructions, as are often necessary to repair larger surgical defects, when compared with patients with postoperative G-tubes. As one might expect, we found that those patients with higher ASA class and free flap/pedicled rotation flap reconstruction actually had longer ICU time and LOS, potentially accounting for the expected cost increases with these variables. However, even with this longer LOS and ICU time in those patients with preoperative G tubes, these seemingly “sicker” patients had lower costs of care than those that were seemingly “healthier” with postoperative G-tubes. We believe that this finding is one of our most profound in this study and that it highlights the impact that obtaining G-tubes preoperatively may have on cost savings because of the fact that seemingly higher risk and sicker patients had lower costs with preoperative G-tube placement.

Despite our data that suggests the potential for considerable cost savings with preoperative G-tubes, predicting which patients would benefit from preoperative nutritional optimization and those who predictably may require them postoperatively as a result of surgery can sometimes be difficult. Several predictive algorithms^20,21 have been developed to identify high-risk characteristics for G-tube needs. The model created by Mays et al²² is the most comprehensive predictive model to date and listed higher TNM classifications, free flap or pedicled rotation flap reconstruction, tracheotomy placement, preoperative radiation, dysphagia/ preoperative weight loss, and supracricoid laryngectomy as the most predictive characteristics. Although more work in stratifying high-risk and low-risk groups needs to be done, this model may serve as a reasonable risk assessment for the clinician during the preoperative evaluation. The decision for G-tube placement at any point is multifactorial and requires careful assessment by the clinician to ensure prudent use of this nutrition aid.

The strengths of this study include that it was performed on a large patient population cared for in a multisurgeon practice at a large tertiary care facility. Our 2 groups of interest were similar, suggesting routine utilization of both preoperative and postoperative G-tubes. Finally, to control for patients who were either lost to follow-up or obtained routine care at an outside hospital or facility, patients who did not maintain follow-up with our clinic during the study period were excluded from this study. By implementing this exclusion criterion, we aimed to use our routine surveillance visits to keep record of any events, readmissions, or swallowing-related issues that may prompt G-tube placement.

There are several limitations of this study. Our data was somewhat reliant on the accuracy and completeness of clinic notes. With numerous providers involved in preoperative clinical evaluation, there was certainly variability in the standard patient preoperative evaluation. Further, we relied on the patient report or outside records to alert us to complications, readmissions, or emergency department visits that occurred outside our facility. If patients did not report these at their routine follow-up appointments, or we did not receive these outside records, these would not have been included. We were unable to obtain specific cost data for G-tube placement on those patients who had G-tubes placed at an outside facility, which theoretically could have skewed our data in those patients. However, by pooling our own institution’s G-tube cost data over a 5-year period and capturing changes in healthcare trends and inflation, we felt the cost estimates that we utilized were reasonable and accurate.

5 |. CONCLUSION

There may be potential for healthcare cost savings by placement of G-tubes preoperatively when the concern for long-term enteral access is high. We found that this cost savings was driven most strongly by a decrease in LOS at the time of resection. Regardless of the practice pattern in head and neck surgical patients, our data suggest that an awareness of the often inevitable delays associated with postoperative swallowing rehabilitation and obtaining enteral access is imperative to reduce LOS. Further, we found that even those patients with high-risk characteristics requiring preoperative G-tubes, that intuitively may have increased costs based on a propensity toward more complications, were found to have lower costs than seemingly lower-risk patients. Further research is warranted to better delineate this relationship.