“Cost Implications of Varying the Surgical Setting and Anesthesia Type for Dorsal Wrist Ganglion Cyst Excision Surgery.”

Abstract

Background:

Minimizing expenses attributed to dorsal wrist ganglion cyst excisions (DWGE), a common minor surgical procedure, presents potential for healthcare cost savings. Varying the surgical setting [operating room (OR) versus procedure room (PR)] and type of anesthesia [local-only, monitored anesthesia care (MAC), MAC with regional (MAC-R), or general (GA)] may affect total operative costs.

Methods:

Patients who underwent an isolated unilateral DWGE between January 2014 and October 2019 at a single academic medical center were identified by current procedural terminology code. Using information technology value tools provided by our institution, the total direct costs (TDC) for each surgical encounter that met this investigation’s inclusion criteria were calculated. The relative TDC were compared between surgical setting and anesthesia type groups. Univariate and multivariable gamma regression models were utilized to identify factors associated with surgical costs.

Results:

A total of 192 patients were included, of which 14% (26/192) of cases were performed in the PR and 86% (166/192) in the OR. No significant differences in demographic factors were identified between these groups. Univariate analysis demonstrated that use of OR/MAC, OR/MAC-R, and OR/GA groups, as compared to PR/Local-only, yielded significantly greater median costs (1.76-, 2.34- and 2.44-fold greater, respectively). Independent of factors found to be significant in the univariate analysis, multivariable analysis demonstrated 1.80-, 2.10- and 2.31-fold greater costs with use of OR/MAC, OR/MAC-R, and OR/GA relative to PR/Local-only, respectively.

Conclusions:

Performing DWGE in a PR with local-only anesthesia minimizes operative direct costs relative to use of the OR and other anesthetic types.

Introduction:

Dorsal wrist ganglion cysts (DWG) are the most common benign soft-tissue tumor of the hand and wrist.¹ The incidence has been estimated at 43 and 25 cases per 100,000 patients in females and males, respectively.¹ Patients may present for evaluation and treatment of these lesions out of concern for possible malignancy, poor cosmetics, localized pain, or decreased functionality.^1–5 Multiple treatment methods have been utilized for DWG including aspiration, corticosteroid injection, arthroscopic surgical excision, and open surgical excision.^9–13 Recent studies have reported that surgical treatments, such arthroscopic and open DWG excision (DWGE), have demonstrated significantly lower recurrence rates relative to nonoperative interventions.^10,13 As such, many patients affected with DWG elect to proceed with surgical excision.

For appropriately indicated patients, hand surgeons may consider different operative methods when performing a dorsal ganglion cyst excision. Arthroscopic and open surgical techniques have been demonstrated to yield similar recurrence rates^10,13 but with greater costs associated with arthroscopic excision.¹⁴ Different anesthetic types may be used as well, which may have an influence on costs^15,16 without any proven improvement in outcomes associated with one modality over another. Traditionally, DWGE has been performed in the operating room (OR) with monitored anesthesia care (MAC), monitored anesthesia care with regional anesthetic (MAC-R), or general anesthesia (GA).

Value-based payment models have gained increasing popularity in the last decade and have relevant implications for minor hand surgeries, such as DWGE.²¹ Use of local anesthesia without sedation, also known as the wide-awake local-only anesthesia with no tourniquet (WALANT) technique, has also been reported to be effective for these types of surgeries.^17–20 Use of the procedure room (PR) setting with local-only anesthetic administered by a hand surgeon removes the need for pre-and-intraoperative monitoring, as well as the need for anesthesia and PACU teams.¹⁸ The WALANT technique has been shown to reduce direct operative costs in carpal tunnel release procedures for entities such as the National Health Service (NHS),²³ the United States Military Health Care System,²⁴ and also in public health care settings within the United States.^15,25 However, cost implications of using WALANT for DWGE, as compared to use of conventional anesthesia types in the OR, have yet to be elucidated.

With the promising clinical and financial outcomes of WALANT performed in ambulatory- or clinic-based PRs for other simple hand surgical procedures, our study aims to investigate cost implications of using different anesthesia types and surgical settings (PR versus OR) for DWGE. Our null hypothesis is that performing DWGE in a PR with local-only anesthesia (WALANT) will not differ in total direct costs when compared to performing this procedure in an OR utilizing various anesthesia types (MAC, MAC-R, GA).

Methods:

Patient Data Collection and Surgical Procedure

Our IRB approved study consisted of all adult (≥ 18) patients who underwent an isolated, unilateral dorsal wrist ganglion cyst excision between January 2014 and October 2019 by 1 of 5 fellowship-trained orthopedic hand surgeons at a single tertiary academic institution. Current Procedural Terminology (CPT) codes (25110 & 25111) were used to identify potential patients – the revision code (25111) was included in attempt to capture primary excisions that may be miscoded as revisions. Basic demographic data, and surgical/anesthetic times, were extracted electronically from our hospital’s database. Manual chart review of clinic, operative, and anesthesia notes was performed and data pertaining to the surgical encounter, anesthetic type, prior treatments, and follow-up were extracted. In the process, we excluded patients undergoing revision ganglion excision, those with non-dorsal wrist ganglia (e.g. volar or other locations), or those with any simultaneous additional surgical procedures. The decision between PR and OR was based upon surgeon familiarity in addition to a shared decision-making model in which the two operative settings with their advantages and disadvantages were discussed with the patient. No patient was formally ineligible for the PR setting although, subjectively, those with high levels of anxiety commonly preferred the OR. In either surgical setting, DWGE was performed as previously described.²⁶

WALANT Protocol

As previously described,¹⁶ All WALANT surgery is performed in a PR with no preoperative medical workup or anesthesiologist present. A nurse is present as a circulator and for checking patients in, which includes non-continuous assessment of their vital signs. The affected hand is placed on an adjacent mobile hand support, and local anesthesia is administered by the surgeon in a sterile fashion using a formulation of 4.5cc of 0.5% bupivacaine and 4.5cc of 1% lidocaine with 1:100,000 epinephrine buffered with 1cc of 8.4% sodium bicarbonate (10:1 ratio).¹⁹ A total of 10cc of this mixture is injected into the operative site prior to formal preparation and draping to allow proper vasoconstriction mediated by the buffered epinephrine. Although the effect of local anesthetic with epinephrine peaks at around 30 minutes,¹⁹ there is no formal minimum amount of time between injection and the surgery that is regularly followed.

Patients are prepped and draped proximal to the level of the elbow using chlorhexidine soap. Normal operative preparation, such as fasting and cessation of home medications, including anticoagulants, are not necessitated for these procedures. Additionally, no intravenous (IV) access or cardiovascular monitoring is needed during the operation. Involved personnel includes the attending hand surgeon, medical assistant, a registered nurse, and at times a hand surgery fellow or resident. Following the operation, patients are discharged directly to self-care after receiving postoperative instructions.

Operating Room Protocol

For patients undergoing a DWGE in the OR, preoperative preparation consisted of fasting for at least 8 hours prior to their procedure, disrobing and wearing a surgical gown, receiving an IV line, and were subject to a routine evaluation by an anesthesiologist. Once in the OR, the anesthesiologist had a choice of multiple methods of sedation including, monitored anesthesia care (MAC), MAC supplemented with IV regional anesthesia (bier block) using a 1:1 mix of plain lidocaine (1%) and bupivacaine (0.5%), or GA. A tourniquet, and cardiovascular monitoring by an attending anesthesiologist, are both utilized intraoperatively. Following the procedure, the patient is transported to the post-anesthesia care unit (PACU) where they receive further monitoring and nursing services until their discharge to a companion after receiving postoperative care instructions.

Cost Analysis

For each surgical encounter, cost data were extracted from the Value Driven Outcomes (VDO) database. This information technology developed by our institution enables the assessment of detailed item-level total direct costs for a variety of healthcare services including surgical encounters. VDO data have been previously reported, and allows investigators to view the total direct costs associated with various aspects of patient care including: specific supplies and implants, facility utilization (including sterile processing costs), and time-based allocations such as cost of staff (surgical technicians, circulating nurses, etc.).^{15,16,21,28–30} It should be noted that physician reimbursement and indirect costs, such as the costs associated with missed work, are not included in the TDC.

To comply with institutional guidelines, all reported cost data were standardized by dividing the cost of each patient’s encounter by the median total direct cost of the data set, effectively satisfying policies against the reporting of any raw cost data related to the details of non-publicly disclosed contractual agreements. Continuous variables, such as patient age, was summarized as a mean ± SD, where categorical metrics were presented as count and percentage (%). Cost groups were defined by unique combinations of surgical setting (PR, OR) and anesthesia type (local-only, MAC, MAC-R, GA). Specifically, we were interested in comparing costs among PR/Local-only, OR/MAC, OR/MAC-R, and OR/GA. Relative costs were calculated by dividing each distinct group median by the lowest group median such that the group with the lowest cost was assigned a value of 1.0.

Statistical Analysis

A t-test was used to compare continuous baseline patient characteristics between PR and OR patients except length of follow-up which was skewed and compared using a Wilcoxon rank sum test. Fisher’s exact test and chi-squared test were used to compare categorical variables, as applicable. All cost data were adjusted for inflation to 2019 dollars using the Consumer Price Index.³¹ The Kruskal-Wallis test followed by Nemenyi pairwise comparisons were used to compare total direct costs among surgical setting/anesthesia type groups. Cost was approximately normally distributed and was analyzed using univariate generalized linear regression models with a log link to evaluate relationships as relative percent change among surgical setting, age, anesthesia types, surgery time, sex, and surgeon. Using the same model, a multivariable regression was constructed including surgeon and all predictor variables that were significant in the univariate models (p < 0.05). For categorical variables such as race and insurance type, 3 p-values comparing the model with the variable to the null model were used to assess significance. Regression coefficients were exponentiated (eβ) to allow for interpretation of percent-changes over reference groups. Since surgical time, MAC, MAC-R, and GA were unique to OR cases, a supplemental multivariable analysis was performed in the OR cohort to determine the effect of these variables on OR-specific costs. Statistical significance was assessed at the 0.05 level, and all tests were 2-tailed.

A priori power analysis was performed that accounted for the unequal proportions of PR WALANT and OR cases historically seen at our institution. With a two-sided, two-sample t-test, we would need at least a total sample size of 182 (156 OR cases and 26 PR cases) to detect a medium effect size of 0.6 (standard deviation SD units) for the cost ratio between PR and OR groups with 80% of power at 0.05 significance level. All statistical analyses were conducted using STATA MP15. (StataCorp. 2017. Stata Statistical Software: Release 15. College Station, TX: StataCorp LLC.)

Results:

Of the 392 patients identified by CPT coding, the following exclusions were made: 84 had additional simultaneous procedures, 91 presented with volar ganglia, 12 had an incorrect diagnosis, and cost data were missing for 13 patients. Baseline characteristics and surgical details are described in Table 1 for the 192 included patients. Mean age at the time of surgery was 32 ± 12 years, 133 (69%) patients were female, and laterality was equally distributed between right and left wrists (50% for both). With the exception of significant differences in the distribution of surgeons using the PR versus the OR–as expected based upon clinical practice patterns – there were no significant differences in demographic variables between PR and OR groups (p > 0.05 for each comparison). We identified four distinct groups based on surgical setting and anesthesia type, including PR/Local-only, OR/MAC, OR/MAC-R, and OR/GA. Associated sample sizes are summarized in Table 2.

Table 1 -.

Descriptive Summary of Baseline Patient Characteristics and Surgical Details

Variable	Total	Procedure Room	Operating Room	p-value
	N=192	N=26	N=166
Demographic Factors
Age (year)	31.5 ± 11.5	31.7 ± 13.0	31.5 ± 11.3	0.93
Male	59 (30.7%)	7 (26.9%)	52 (31.3%)	0.65
Race				0.20
Black or African American	8 (4.2%)	0 (0.0%)	8 (4.8%)	-
White or Caucasian	144 (75.4%)	17 (65.4%)	127 (77.0%)	-
Insurance ^				0.15
Commercial	125 (65.1%)	19 (73.1%)	106 (63.9%)	-
Government Other	48 (25.0%)	4 (15.4%)	44 (26.5%)	-
Medicaid	9 (4.7%)	2 (7.7%)	7 (4.2%)	-
Medicare	8 (4.2%)	0 (0.0%)	8 (4.8%)	-
Workers Compensation	1 (0.5%)	0 (0.0%)	1 (0.6%)	-
Self Pay	1 (0.5%)	1 (3.8%)	0 (0.0%)	-
Surgical Details
Surgeon ^				<0.05
Surgeon A	28 (14.6%)	0 (0.0%)	28 (16.9%)	-
Surgeon B	24 (12.5%)	1 (3.8%)	23 (13.9%)	-
Surgeon C	14 (7.3%)	0 (0.0%)	14 (8.4%)	-
Surgeon D	73 (38.0%)	23 (88.5%)	50 (30.1%)	-
Surgeon E	53 (27.6%)	2 (7.7%)	51 (30.7%)	-
Laterality				0.40
Left	97 (50.5%)	11 (42.3%)	86 (51.8%)	-
Right	95 (49.5%)	15 (57.7%)	80 (48.2%)	-
Follow-up length (days)	14.0 (13.0 to 20.0)	14.0 (13.0 to 14.0)	13.0 (12.0 to 20.0)	0.93
Prior aspiration (Yes)	59 (30.7%)	12 (46.2%)	47 (28.3%)	0.07
Anesthesia Type ^				<0.05
Local-only	26 (13.5%)	26 (100.0%)	0 (0.0%)	-
MAC with Regional	88 (45.8%)	0 (0.0%)	88 (53.0%)	-
MAC	28 (14.6%)	0 (0.0%)	28 (16.9%)	-
General Anesthesia	50 (26.0%)	0 (0.0%)	50 (30.1%)	-
Surgery time (min)	23.6 ± 14.2	Not reported	23.6 ± 14.2	NA

Summaries are reported as a mean ± standard deviation or count (percent of total) with the exception of follow-up length, which is summarized as a median (interquartile range).

^{^}P value calculated based on a Fisher’s exact test. Bolded P values denote statistical significance.

Table 2 -.

Summary Statistics of Surgical Setting and Anesthesia Type Groups

Surgical setting	Anesthesia type	N
PR	Local-only	26
OR	MAC	28
OR	General Anesthesia	50
OR	MAC with Regional	88

Surgical encounter total direct costs for each of the four groups significantly differed from one another (Figure 1; Table 3). Specifically, univariate analysis demonstrated that OR/MAC, OR/MAC-R, and OR/GA were 76% (estimate 1.76; 95% CI: 1.49 to 2.09; p < 0.05), 134% (estimate 2.34; 95% CI: 1.97 to 2.79; p < 0.05), and 144% (estimate 2.44; 95% CI: 2.06 to 2.88; p < 0.05) more costly than PR/Local-only for DWGE, respectively. In addition, univariate analysis demonstrated that use of the OR, as compared to use of the PR, was associated with a significant increase in direct cost (estimate 2.06-fold greater, or a 106% increase; 95% CI 1.69 to 2.52; p < 0.05). Surgeon was also a significant factor associated with differing costs, in which 3 of 4 surgeons were associated with significantly greater direct costs than the reference surgeon (Surgeon A; p < 0.05 for each of the three comparisons).

Figure 1:

Normalized Total Direct Costs Among Surgical Setting and Anesthesia Type Groups

Table 3 -.

Univariate Regression for Surgical Encounter Total Direct Cost

Variable	Cost Ratio (95% CI)*	P-value
Age (year) +	1.00 (0.99 to 1.00)	0.21
Surgical Setting / Anesthesia Type
PR / Local-only	Reference	-
OR / MAC ^	1.76 (1.49 to 2.09)	<0.05
OR / MAC with Regional ^	2.34 (1.97 to 2.79)	<0.05
OR / General Anesthesia ^	2.44 (2.06 to 2.88)	<0.05
Surgeon
Surgeon A	Reference	-
Surgeon B	1.51 (1.31 to 1.74)	<0.05
Surgeon C	1.66 (1.43 to 1.93)	<0.05
Surgeon D	0.98 (0.85 to 1.12)	0.75
Surgeon E	1.19 (1.04 to 1.36)	<0.05
Insurance
Commercial	Reference	-
Government Other	1.05 (0.95 to 1.17)	0.35
Medicaid	0.79 (0.60 to 1.04)	0.10
Medicare	0.92 (0.72 to 1.18)	0.53
Workers Compensation	1.51 (0.99 to 2.31)	0.06
Self Pay	0.58 (0.19 to 1.74)	0.33
Sex (Male vs. Female)	1.00 (0.91 to 1.11)	0.95
Race
Black or African American	Reference	-
White or Caucasian	0.87 (0.71 to 1.06)	0.17
American Indian and Alaska Native	1.03 (0.67 to 1.57)	0.89
Other	0.79 (0.63 to 0.99)	<0.05
Choose not to disclose	1.06 (0.70, to 1.60)	0.78
Asian	0.65 (0.38 to 1.10)	0.11
Surgical setting
Procedure Room (PR)	Reference	-
Operating Room (OR)	2.06 (1.69 to 2.52)	<0.05

^*Relative cost: For example, a value of 1.4 is interpreted as a 40% increase in total direct costs as compared to the reference category.

⁺Per each additional decade of age.

^{^}Variables that are only available for OR cases.

Bolded P values denote statistical significance.

In the multivariable analysis for surgical encounter total direct costs for the entire cohort (all included patients), OR/MAC, OR/MAC-R, and OR/GA were 80%, 110%, and 131% more costly than PR/Local-only (p < 0.05 for each; Table 4). This was independent of the surgeon, for which significant differences in cost remained between surgeons. To evaluate for association between anesthesia type and cost for OR cases, a separate multivariable analysis of the OR-only sub-cohort revealed that use of MAC or GA were associated with 17% and 29% greater surgical direct costs than use of MAC with regional (p < 0.05 for both; Table 5). Further, each additional 10 minutes of surgical time in the OR increased total costs by 3% (estimate 1.03; 95% CI: 1.02 to 1.04).

Table 4 –

Multivariable Regression for Surgical Encounter Total Direct Cost: Entire Cohort

Variable	Cost Ratio (95% CI)*	P-value
Surgical Setting / Anesthesia Type
PR / Local-only	Reference	-
OR / MAC ^	1.80 (1.55 to 2.09)	<0.05
OR / MAC with Regional ^	2.10 (1.79 to 2.46)	<0.05
OR / General Anesthesia ^	2.31 (1.99 to 2.69)	<0.05
Surgeon
Surgeon A	Reference	-
Surgeon B	1.35 (1.22 to 1.50)	<0.05
Surgeon C	1.47 (1.30 to 1.67)	<0.05
Surgeon D	1.10 (1.00 to 1.22)	<0.05
Surgeon E	1.10 (1.00 to 1.21)	0.06

^*Relative cost: For example, a value of 1.4 is interpreted as a 40% increase in total direct costs as compared to the reference category.

^{^}Variables that are only available for OR cases. Bolded P values denote statistical significance.

Table 5 -.

Multivariable Regression for Surgical Encounter Total Direct Cost: OR-Only Sub-Cohort

Variable	Cost Ratio (95% CI)*	P-value
Anesthesia type
MAC with Regional	Reference	-
MAC	1.17 (1.07 to 1.28)	<0.05
General Anesthesia	1.29 (1.21 to 1.38)	<0.05
Surgeon
Surgeon A	Reference	-
Surgeon B	1.33 (1.20 to 1.48)	<0.05
Surgeon C	1.42 (1.25 to 1.61)	<0.05
Surgeon D	1.08 (0.98 to1.19)	0.12
Surgeon E	1.10 (1.00 to 1.21)	0.06
Surgery time +	1.03 (1.02 to 1.04)	<0.05

^*Relative cost: For example, a value of 1.4 is interpreted as a 40% increase in total direct costs as compared to the reference category.

⁺Per each 10 additional minutes. Bolded P values denote statistical significance.

Discussion:

The main finding of this study is that the surgical encounter total direct costs for dorsal wrist ganglion cyst excision were minimized when performed in the procedure room setting with surgeon-administered local-only anesthesia. This method was significantly less costly than using the OR with various anesthesia types including MAC, MAC with regional, and general anesthesia. Specifically, we found that use of the ambulatory surgery center OR resulted in more than a doubling of surgical costs, as compared to the PR. In light of these findings, we were able to reject our null hypothesis that performing DWGE in a PR with local-only anesthesia will not differ in total direct costs when compared to performing this procedure in an OR utilizing various anesthesia types (MAC, MAC-R, GA).

Our findings are consistent with multiple studies that compare the costs of surgery in the PR with WALANT to the OR with various anesthesia types for different elective hand surgeries such as trigger finger release (TFR) and open carpal tunnel release (OCTR).^{15–18,22,32} For example, it was found that OCTR procedures performed under local-only anesthesia in the OR incurred a 6.3-fold greater total direct procedure cost relative to the same anesthetic type and procedure carried out in the PR. Cost differences between WALANT and use of MAC anesthesia in the OR for OCTR cases were even greater.¹⁵ Similar findings have been reported for TFR in which use of the OR with local-only or MAC anesthesia were found to be 2.2- and 3.2-fold greater, respectively, than WALANT in the PR setting.¹⁶ Additionally, Rhee et al demonstrated that WALANT protocol in a PR yielded cost savings of 70% when compared to OR protocol. This investigation goes on to project savings of up to $400,000 for the Military Health Care System by electing to carry out just 100 procedures in a PR rather than an OR.²⁴ Similar values are also reported by Maliha et al.³²

The reasons for these cost differences between PR with WALANT and OR with various anesthesia types for minor hand surgeries are likely multifactorial. A facility cost difference is likely to substantially contribute, with a prior report noting an 8.4-fold greater cost per minute in the OR versus the PR.¹⁵ Also contributory are the differences in necessary materials and personnel required for each setting. Surgical technicians, anesthesia-related drugs, recovery room nursing staff, and associated recovery room medications and supplies are not required for surgeries done in the PR setting.

These findings are informative in light of previous literature that has shown similar complication rates and patient reported outcomes between the PR with WALANT and OR with various anesthesia types. Garon et al. found that in a study of 189 patients (59 treated in an OR and 130 treated in a PR) who underwent a closed reduction and percutaneous pinning of metacarpal and phalangeal fractures, those treated in a PR had no significantly different rates of post-operative infection, malunion, or nonunion when compared to those treated in an OR.³³ Additional studies have further demonstrated low complication rates for elective hand surgeries in the PR.^34–37 Regarding patient reported outcomes, Rabinowitz et al. demonstrated that patients treated in the PR for TFR were actually significantly more satisfied than patients treated in the OR and had better functional post-operative Disabilities of the Arm, Shoulder and Hand scores.³⁶ Additional studies have also demonstrated high satisfaction with elective hand surgeries performed in a PR- or clinic-basted setting.^24,34,38 Although these studies are limited, which highlights the need for further work in this realm, we found no literature supporting that the PR setting leads to worse outcomes or higher complication rates than the OR. Further, no literature is currently available that reports on the efficacy or postoperative complication rate, including recurrence, of a DWGE performed in a PR relative to an OR.

Our study has some limitations that deserve mention. First, the observed findings are unique to the pricing agreements made between our institution and relevant suppliers, thus, the generalizability of our study must be interpreted within this context and may be different for each healthcare system. Next, all cases that were included in our study were assessments of patients treated at our orthopedic ambulatory care center, and not in a main hospital OR. A greater difference in cost would likely be noted for institutions where DWGE are performed in a main hospital OR.³⁹ Given the retrospective design, our study may be susceptible to selection bias. The cost analysis that was performed is isolated to the surgical encounter itself, and therefore costs related to pre- or postoperative care were not evaluated. However, based on the findings of Lalonde et al. it is likely that the observed cost differences would be greater if these indirect costs were taken into consideration.¹⁸ Likewise, these results reflect only surgical direct costs – we did not evaluate indirect costs such as time off work, nor health states, surgical outcomes, or complications including recurrence. Therefore, a formal cost utility analysis that accounts for these factors would build upon the observations in the current study, which could be informative. Lastly, one surgeon (Surgeon D) performed the majority of PR cases (23/26) of this study which may have affected univariate regression results. As this operative location by surgeon distribution was found to vary significantly among surgeons, it was appropriately controlled for in the multivariable analysis. It is possible that surgeon experience is associated with efficiency and cost, which our study did not account for.

Conclusions:

Dorsal wrist ganglion cyst excision surgical direct costs were minimized, as compared to using the OR with a variety of other anesthetic methods, by performing the surgery in a procedure room setting under surgeon-administered local-only anesthesia. Moving this type of surgery out of the OR and into the PR, for indicated patients, has the potential to lead to substantial health care savings for this common procedure.