2017-18 HSC Section 3 Green Book

whereas the use of endovascular embolization appears to be remaining flat. Table II lists hospital characteristics between the two groups. Compared to the embolization group, the ligation group was older (64.1 years vs. 62.4 years; P 5 0.027) and had less of a male gender predominance (55.6% vs. 62.5%; P 5 0.003). No difference in racial dis- tribution was seen between the two groups. Nonelective admissions were more common in the embolization group (11.2% vs. 6.3%; P < 0.001). On average, patients in the ligation group stayed in the hospital for fewer days (3.6 days vs. 4.0 days; P 5 0.014) and incurred low- er hospital charges ($33,030 vs. $69,304; P < 0.001). Comorbidity frequencies are listed in Table III. The three most common comorbidities were hypertension, diabetes mellitus, and chronic pulmonary disease. The only comorbidity that was significantly different between the groups was congestive heart failure, which was more common among patients receiving surgical ligation (15.1% vs. 9.8%; P < 0.001). In-hospital complications are listed in Table IV. The only complication to vary significantly between the groups was intubation/tracheostomy, which was signifi- cantly less common in the ligation group (2.8% vs. 5.3%; P 5 0.009). Transfusion was the most common complica- tion that we analyzed and was not significantly different between the two groups (24.3% in ligation vs. 22.8% in embolization; P 5 0.450). Rates of stroke and blindness were low and not significantly different between the two cohorts. The in-hospital mortality rate was also low and not significantly different between the two groups (1.1% in ligation vs. 0.6% in embolization; P 5 0.450). DISCUSSION Intractable epistaxis is a frequently encountered condition that can be life-threatening as well as costly, often requiring aggressive management with surgical arterial ligation or arterial endovascular embolization for treatment. The purpose of the present analysis was to compare these two treatment modalities and offer updated information about their costs and benefits. One of the major advantages to endovascular embo- lization versus arterial ligation is that it can be per- formed under local anesthesia, thus avoiding the risks of general anesthesia in patients unfit for surgery due to

TABLE I. ICD-9 Codes for Variable Derivation

Variable

ICD-9 Code

Epistaxis

478.29

Surgical ligation*

21.04, 21.05, 21.06, 21.09 39.72, 38.82, 38.80, 38.79

Arterial embolization*

Hereditary hemorrhagic telangiectasia

448.0

Stroke*

39.74, 436, 997.02

Blindness*

369.00, 369.60, 369.67, 950.09

Blood transfusion

99.04

Intubation/tracheostomy

96.01, 96.02, 96.03, 96.04, 96.05, 31.1, 31.21, 31.29

*Adapted from Villwock and Goyal. 9 ICD-9 5 International Classification of Diseases, 9th Revision, Clini- cal Modification.

the Healthcare Cost and Utilization Project (HCUP), sponsored by the Agency for Healthcare Research and Quality. The data- base represents a sample of approximately 20% of all hospital discharges in the United States, excluding rehabilitation and long-term acute care hospitals. Prior to 2012, the NIS data sam- pling included 100% of discharges from 20% of HCUP hospitals. Since 2012, however, the sampling method was reconfigured to include 20% of discharges from 100% of HCUP hospitals to facilitate more accurate national estimates. Each case in the NIS represents one hospitalization, from admission to discharge. Henceforth, we will refer to an NIS case as a patient. Inclusion criteria for the present analysis included all patients with a primary International Classification of Diseases, 9 th Revision, Clinical Modification (ICD-9) diagno- sis code for epistaxis and an associated ICD-9 procedure code for arterial ligation or embolization. Table I lists relevant ICD-9 diagnosis and procedure codes. Patient characteristics and in- hospital outcomes were compared between patients receiving either surgical ligation or arterial embolization. Data collection and analysis were completed in June 2016. Unweighted data were reported and used for all statistical analyses, unless other- wise specified. The Rutgers New Jersey Medical School Institutional Review Board (IRB) does not consider the use of de-identified data from the NIS to meet the regulatory definition of human subject research provided in 45 CFR 45.102; therefore, IRB approval was not required for the present study. Statistical Analysis Two tailed t tests, Pearson’s chi square, and Fisher’s exact test were used as appropriate. SPSS version 22 (IBM, Armonk, NY) was used for all statistical analyses, with threshold for sig- nificance set at P < 0.05. RESULTS A total of 1,813 patients with epistaxis, who were treated with either surgical ligation or arterial emboliza- tion, were identified in the NIS between 2008 and 2013. Of those, 1,035 (57.1%) patients underwent surgical liga- tion, and 778 (42.9%) of these patients underwent arteri- al embolization. Figure 1 shows the trend in number of cases per year using weighted data. The national use of surgical ligations appears to be trending downward,

Fig. 1. National cases of ligation versus embolization per year.

Laryngoscope 127: May 2017

Sylvester et al.: Ligation Versus Embolization in Epistaxis

203