Evaluating the Safety of the Hand Surgery Procedure Room: A Single-Center Cohort of 1404 Surgical Encounters

Abstract

Purpose:

Performing hand surgeries in the procedure room (PR) setting instead of the operating room (OR) effectively reduces surgical costs. Understanding the safety or complication rates associated with the PR is important in determining the value of its utilization. Our purpose was to describe the incidence of medical and surgical complications among patients undergoing minor hand surgeries in the PR.

Methods:

We retrospectively reviewed all adult patients who underwent an operation in the procedure room setting between December 2013 to May 2019 at a single tertiary academic medical center by 1 of 5 fellowship-trained orthopaedic hand surgeons. Baseline patient characteristics were described. Complication rates were obtained via chart review.

Results:

For 1404 PR surgical encounters, 1796 procedures were performed. Mean patient age was 59 ±15, 809 (57.6%) were female, and average follow-up was 104 days. The most common surgeries were carpal tunnel release (39.9%), trigger finger release (35.9%), and finger mass/cyst excision (9.6%). Most surgeries (58%) were performed using a non-pneumatic wrist tourniquet, whereas 42% used no tourniquet. No patient experienced a significant medical complication. No procedure was aborted due to intolerance. No patient required admission. No intraoperative surgical or medical complications occurred. Observed complications included delayed capillary refill requiring phentolamine administration following a trigger thumb release performed using epinephrine without a tourniquet (n=1; 0.1%), complex regional pain syndrome (CRPS; n=3; 0.2%), infection requiring surgical debridement (n=2; 0.2%), and recurrent symptoms requiring reoperation (n=8; 0.7%).

Conclusion:

In this cohort of patients in whom surgery was performed in a procedure room, there were no major intra-operative surgical or medical complications. There was a low rate of post-operative infection, development of CRPS and a low need for revision surgery. These observations do not support the concern for safety as a barrier to performing minor hand surgery in the procedure room setting.

Level of Evidence:

IV, Therapeutic

Introduction:

The emphasis on improving the value of healthcare delivery has increased in importance in recent decades, in part due to policies that allow for value-based payment models.¹ The value of care may be thought of as the outcome or level of improvement per unit cost.^2–5 One means of improving the value of care is through decreasing surgical costs while maintaining comparable clinical, functional, and safety outcomes.

There is no question that performing minor hand surgery in the procedure room (PR) setting leads to reduced surgical costs, as compared to utilizing the OR. Performing hand surgery in the PR with local-only anesthesia has been shown to decrease the need for costly preoperative medical testing.⁶ Recent literature has also demonstrated that moving minor hand surgeries out of the operating room (OR) and into the procedure room (PR) is an effective means of directly decreasing surgical encounter costs.^7–11 Direct surgical costs for carpal tunnel release (CTR) in the OR range from 6- to 29-fold higher than the cost of performing open CTR in the PR.⁷ A similar trend was reported for trigger finger release (TFR) by which performing surgery in the OR was 2.2 to 3.2-fold more costly than in the PR.⁸ Other studies with similar observations also exist ^{9, 11–14}

Less is known about the second component of value in relation to the PR: safety and outcomes. Several published reports exist in which carpal tunnel release, trigger finger release, hardware removal, nail ablation, tendon repair, De Quervain release, and Dupuytren open palmar fasciectomy were successfully performed under WALANT (wide-awake, local-only anesthesia, no tourniquet) in the PR setting.^{7, 8, 11, 15–17} Multiple studies have reported complications associated with performing hand surgery in the PR.^{14, 18–20} Some of these studies are also limited in that they evaluate only specific procedures such as CTR^{19, 21} or TFR.¹⁴ Therefore, information regarding the safety of additional hand procedures performed in the PR could be informative.

Given the absence of detailed information regarding complication rates associated with performing minor hand surgeries in the PR, the purpose of this study was to describe the rate of major and minor complications for a large cohort of patients treated by fellowship-trained hand surgeons in the procedure room setting.

Methods:

Our study was approved by our institution’s review board. We retrospectively reviewed all patients treated in the procedure room setting by 1 of 5 fellowship-trained orthopaedic hand surgeons at our tertiary academic medical center between December 2013 and May 2019. Any patient meeting the indications for minor hand surgery, during our study period was potentially a candidate for the PR instead of the OR. Surgical setting selection was determined using shared decision-making. Although patients with decompensated medical comorbidities were considered to be ineligible for elective surgery in general, ASA scores and comorbidities were not used as exclusion criteria for eligibility for the PR. Therefore, patients with significant but stable medical comorbidities were available for inclusion in our study. Manual chart review of all clinic, emergency room, and hospital visits and procedural notes by the first author (ARS) identified the procedures performed, associated diagnoses, usage of epinephrine or tourniquets, intraoperative medical and surgical complications, and postoperative complications. All post-operative follow-up visits were conducted in person. Through review of the medication reconciliation in the electronic medical record, we also recorded whether or not diazepam was ordered preoperatively, and whether narcotics were prescribed post-operatively.

WALANT Protocol

Preoperative medical or anesthesia evaluation was not required for patients regardless of their comorbidities. Our WALANT surgeries were performed in a PR located adjacent to our ORs in an ambulatory surgical center. From the waiting room, patients were brought directly to the PR and placed supine. The patient’s arm(s) were placed on an adjacent mobile hand table and a 5cc injection of local anesthesia was administered in a sterile fashion at the appropriate operative site(s) prior to formal draping, using a formulation of 4.5 cc of 1% lidocaine and 4.5 cc of 0.5% bupivacaine buffered with 8.4% sodium bicarbonate. The addition of epinephrine was surgeon- and procedure-dependent and the use of phentolamine was available in case critical digital ischemia occurred.^{15, 22, 23} Tourniquet usage was likewise surgeon-dependent.

Patients were not required to disrobe, and their arm was prepped and sterilely draped. Fasting and/or holding of home medications, including anticoagulation agents, was not required in preparation for the operation. Cardiovascular monitoring was not used, and an intravenous (IV) line was not placed. Intraoperative antibiotics (IV or oral) were not administered. PR staffing included the attending hand surgeon, medical assistant whose primary role was to help maintain sterility, hand surgery fellow or resident, and a registered nurse. After completion of the procedure, patients received postoperative care instructions and were discharged with self-care directly to home.

Variables of interest were descriptively summarized at the visit level (N=1404). For continuous variables, mean, standard deviation (SD), median, interquartile range (IQR), and range were summarized. For categorical variables, frequency and percentage were calculated. Statistical significance was assessed at the 0.05 level using two-tailed tests.

Based on the overall complication rate of 2.5% as reported in the study by Lipira et al.,²⁵ with a sample size of 1400 we expected a 95% confidence interval ranging from 1.7% to 3.5% (a width of 1.8%) to describe the complication rate of the current study. Given that our infection rate was found to be 0.6%, the 95% CI had greater precision, at 0.3% to 1.2% (a width of 0.9%).

Results:

A total of 1796 procedures were performed for 1263 patients within 1404 unique procedural encounters. Taking into account that 127 patients had multiple procedure encounters over the study period, our cohort was 57.6% female with an average age of 59 (SD ± 15.4). Additional demographic data are presented in Table 1. The maximum number of procedures performed at a single encounter was 6, and a single surgical procedure was performed for 1083 of the encounters (Table 2). The median follow-up was 14 days (IQR: 13, 63). Local anesthetic with epinephrine was used in 35.8% of encounters, and plain local anesthetic was used in the remainder of encounters. A non-pneumatic tourniquet (Hemaclear; Grandville, MI) was used in 58% of encounters, and no tourniquet was used in the remainder. The top 3 procedures were CTR (717), TFR (645), and digital mass excision (173). The number of each procedure performed is shown in Figure 1. Postoperative narcotics were prescribed to 559 (39.8%) patients. Of the 1404 procedural encounters, no patients required admission to the hospital, and none of the procedures had to be aborted due to patient intolerance or surgeon inability to complete the procedure for any reason. No intraoperative medical complications were observed.

Table 1.

Descriptive Summary of Demographics at a Visit Level

Variable	Levels	Summary (N=1404)	#Missing
Age	Mean (SD)	59.36 (15.40)	0
	Median (IQR)	61.00 (49.00, 71.00)	-
	Range	(17.00, 96.00)	-
Gender	Female	809 (57.6%)	0
	Male	595 (42.4%)	-
Race	White or Caucasian	1202 (85.6%)	0
	Hispanic/Latino	103 (7.3%)	-
	Other	30 (2.1%)	-
	Asian	29 (2.1%)	-
	Black or African American	12 (0.9%)	-
	Patient Refused	11 (0.8%)	-
	American Indian and Alaska Native	7 (0.5%)	-
	Native Hawaiian and Other Pacific Islander	6 (0.4%)	-
	Choose not to disclose	4 (0.3%)	-
BMI	Mean (SD)	29.64 (7.56)	29
	Median (IQR)	28.20 (24.60, 32.90)	-
	Range	(14.90, 77.80)	-
Smoking	Yes	70 (5%)	9
	No	1325 (95%)	-
Diabetes	Yes	161 (11.5%)	0
	No	1243 (88.5%)	-
Opioids	Yes	169 (12%)	0
	No	1235 (88%)	-
Days of follow-up	Mean (SD)	104.93 (222.44)	0
	Median (IQR)	14.00 (13.00, 62.78)	-
	Range	(0.00, 1616.68)	-
Death	Yes	14 (1%)	0
	No	1390 (99%)	-
Site	1	1124 (80.1%)	0
	2	254 (18.1%)	-
	3	26 (1.9%)	-
Attending provider	A	12 (0.9%)	0
	B	354 (25.2%)	-
	C	156 (11.1%)	-
	D	617 (43.9%)	-
	E	265 (18.9%)	-
Insurance	Commercial	784 (59%)	76
	Medicare	439 (33.1%)	-
	Medicaid	54 (4.1%)	-
	Self Pay	18 (1.4%)	-
	Workers Compensation	17 (1.3%)	-
	Other Government	13 (1%)	-
	Other	3 (0.2%)	-
ASA	0	506 (49.7%)	385
	1	54 (5.3%)	-
	1E	1 (0.1%)	-
	2	289 (28.4%)	-
	2E	2 (0.2%)	-
	3	157 (15.4%)	-
	3E	2 (0.2%)	-
	4	7 (0.7%)	-
	5	1 (0.1%)	-

Table 2.

Descriptive Summary of Procedure Characteristics at the Surgical Encounter Level

Variable	Levels	Summary (N=1404)	#Missing
Number of procedures in a single visit (continuous)	Mean (SD)	1.28 (0.59)	0
	Median (IQR)	1.00 (1.00, 1.00)	-
	Range	(1.00, 6.00)	-
Total number of procedures in a single visit (categorical)	1	1083 (77.1%)	0
	2	265 (18.9%)	-
	3	41 (2.9%)	-
	4	13 (0.9%)	-
	6	2 (0.1%)	-
Unilateral vs. bilateral surgeries	Unilateral	1226 (87.3%)	0
	Bilateral	178 (12.7%)	-
Tourniquet use	Yes	817 (58.2%)	0
	No	587 (41.8%)	-
Epinephrine use	Yes	503 (35.8%)	0
	No	901 (64.2%)	-
Preop valium ordered	Yes	35 (2.5%)	0
	No	1369 (97.5%)	-
Postop narcotics ordered	Yes	559 (39.8%)	0
	No	683 (48.6%)	-
	Already taking	162 (11.5%)	-

Figure 1:

The number of each procedure performed

No evidence of iatrogenic nerve, artery, or tendon laceration was observed intraoperatively or at subsequent follow-up visits. Only one surgical complication was identified: delayed capillary refill requiring phentolamine administration following a trigger thumb release. During the procedure, a tourniquet and local anesthetic with epinephrine was used, and this patient had normal capillary refill after the procedure. About 6 hours after surgery, the patient noticed that the finger appeared blue. She was instructed to come back to clinic immediately. On exam, the tip of her operative digit had a blueish discoloration, decreased turgor and absent Doppler arterial signals. After subcutaneous administration of phentolamine, perfusion normalized within minutes and was normal at a 2-week postoperative visit.

During the study period, 14 of the 1263 patients died: 10 deaths occurred > 1 year postoperatively, 1 death was between 6–12 months postoperatively, 1 occurred between 3–6 months postoperatively, and 2 patients died < 3 months postoperatively (1 at 3 weeks, 1 at 4 weeks). The patient who died 4 weeks postoperatively after trigger finger release had known coronary artery disease with ischemic cardiomyopathy and was doing well at the 2-week postoperative visit with their hand surgeon. Although this patient had a pre-scheduled coronary bypass grafting surgery performed 3 weeks after their PR visit, the cause of death is unknown. The patient who died 3 weeks postoperatively following trigger finger release had known diabetes, chronic obstructive pulmonary disease, and heart failure with reduced ejection fraction. Documented oxygen saturations were within normal limits within the week prior to hand surgery, no complications were noted during surgery, and the patient reported they were “doing well” when contacted via phone on postoperative day 1. The cause of death, or whether the procedure was contributory, are both unknown for this individual.

Of the 1207 visits for 1070 patients who had recorded follow-up encounters, 7 (0.6%) patients had superficial skin infections requiring administration of oral antibiotics, 2 (0.2%) of which went on to require surgical debridement in the OR. There was one CTR patient with an infection (0.2%), and five TFR patients (1.0%). Recurrence of symptoms occurred in 14 patients: 6 were observed after a needle aponeurotomy, 3 after TFR, 4 after mass excisions, and 1 after a nail ablation. Five required reoperations in the OR (3 needle aponeurotomies and 2 mass excisions) and 3 had reoperation in the PR (1 nail ablation, 1 mass excision, and 1 TFR). Post-operative development of CRPS was noted in 3 (0.2%) patients at follow-up. All had undergone CTR. One of these patient’s symptoms resolved by 4 months post-operatively and one patient was doing well after a stellate block 6 months after their CTR. The final patient continued to have symptoms at 12-month post-operatively despite multiple stellate injections. A comprehensive summary of all major and minor complications is shown in Table 3.

Table 3 -.

Descriptive Summary of Major and Minor Complications^*

Major Complications	N (%)
Had to abort procedure due to patient intolerance?	0 (0%)
Phentolamine use	1 (0.1%)
Intraoperative surgical complication	1 (0.1%)
Intraoperative medical issue	0 (0%)
Admitted on Date of Surgery	0 (0%)
Nerve/artery/tendon transection complication in follow-up	0 (0%)
Recurrence and Reoperation in OR	5 (0.4%)
Recurrence and Repeat Surgery in Procedure Room	3 (0.3%)
Infection Requiring Surgical Debridement in OR	2 (0.2%)
Complex Regional Pain Syndrome	3 (0.2%)
Minor Complications
Flexion Contracture 0–10 Degrees	61 (5.0%)
Flexion Contracture 11–20 Degrees	11 (0.9)
Flexion Contracture 21–30 Degrees	3 (0.2%)
Flexion Contracture >30 Degrees	1 (0.1%)
Infection Total	7 (0.6%)
Infection Necessitation Administration of Antibiotics	5 (0.4%)
Lacks 5 Degrees of Full Flexion of PIP	1 (0.1%)
Persistent Numbness	9 (0.7)
Persistent Pain	3 (0.2%)
Persistent Catching	2 (0.2%)
Persistent Symptoms (Pronator Release)	1 (0.1%)
Recurrence no Reoperation	6 (0.5%)
Stiffness	1 (0.1%)
Persistent Numbness, Steroid Injection Given	1 (0.1%)

^*No patient had more than one Major Complication. For Minor Complications, it was possible for a given patient to have one or more, therefore, the sum of the complications does not represent the sum of the patients with complications.

Discussion:

The main finding of our study was that surgical procedures done in the procedure room setting had a low rate of overall complications. No major intraoperative medical or iatrogenic surgical complications were observed in 1404 procedural encounters and 1796 individual surgical procedures. Observed rates of postoperative complications, including those requiring unplanned reoperation, were low. Taken together, these finding suggest that performing minor hand surgeries in the procedure room setting is safe. Although the procedure room at our institution is located in an office-like room within an ambulatory surgery center, we believe our findings are likely generalizable to any procedure room setting: the common denominator for the procedure room, whether in-office or associated with a surgery center, or within a tertiary or community hospital, is use of surgeon-administered local anesthesia outside of the OR and a lack of anesthesia support.

Our results with respect to a low rate of complications, are consistent with prior literature for minor hand surgeries. Bismil et al. also demonstrated no major intra-operative medical or surgical complications for 1,000 procedure performed in the procedure room.²⁰ LeBlanc et al. and Halvorson et al. both demonstrated low infection rates (0.4% and 2.2% respectively) after CTR performed in the PR.^{19, 21} The use of WALANT for larger surgeries, such as distal radius fracture fixation and trapeziectomies, has also been shown to be safe.^26–28 Our findings are notable in that this is the first large cohort study to evaluate complication rates of all elective hand surgeries in addition to CTR and TFR (De Quervain’s, mass excision, needle aponeurotomy etc.) in the United States. Our results demonstrate that many small hand procedures can be safely performed in the PR setting. The results of our study are also consistent with rates previously demonstrated for similar procedures performed in the OR. The risk of major medical and surgical complications in our cohort is similar to that seen for hand surgery performed in the OR. Lipira et al. demonstrated that the risk of myocardial infarction, pulmonary embolism, shock, stroke, hemorrhage, or nerve injury for outpatient hand surgery to all be less than 0.1%.²⁵ The minor complication rate for our cohort was similar and often qualitatively lower than published OR values. In our study we observed an overall rate of infection requiring operative debridement of 0.5%. This is similar to the findings of Tosti et al. and Lipira et al. who both demonstrated overall infection rates after elective hand surgeries to be approximately 0.7%.^{25, 29}, but contrary to the findings of Platt et al. who found the rate of surgical site infections for elective hand surgeries to be around 11%, which is questionably high.³⁰ Previous literature has indicated that the infection rate is between 1–11%^{31, 32} for CTR and 5–6%.^{33, 34} for TFR, values qualitatively higher than our rates of 0.2 and 1% respectively. Although limitations exist regarding diagnosis, the incidence of CRPS after CTR has been previously demonstrated to be between 2.1 and 8.3%,^{35, 36} which is qualitatively greater than the 0.3% rate observed in the current cohort. Recurrence of CTS requiring reoperation has been shown to be between 2% and 8%,^{37, 38} which is qualitatively greater than the 0.2% in the current study. Finally, a Cochrane review of 14 randomized controlled trials found the recurrence of trigger finger after surgery to be approximately 7%,³⁹ in contrast to the 0.2% rate observed in the current study. Rhee et al. who found the complication rate of clinic-based WALANT for a spectrum of minor hand surgeries to be 3%. In their study of 100 consecutive patients, the post-operative infection rate was 1%. Phentolamine use was not required and their patients did not experience any major medical complications, require hospital admission, or monitoring post-operatively.¹¹ However, the strength of this finding is limited by the low sample size of the study (100 patients). In total, our results indicated that the use of the PR for hand surgery is comparably safe in terms of low complication rates in relation to those described in the literature for minor hand surgeries.

Our study has several limitations that warrant discussion. The retrospective nature of our study makes it susceptible to selection bias. Our study did not evaluate patient-level factors that might have an impact on a surgeon’s likelihood to offer the PR or the patient’s preference in selecting the PR versus OR. It is likely that differences exist between patients who opted to be treated in the PR and those who declined. These differences could bias our results, which highlight the benefits of a randomized study design. Our study design and results do not allow us to comment on the potential for litigation or liability related to potential major intra-operative complications that occur in the PR as compared to that in the OR – these issues likely vary by country, state, and the specific medico-legal milieu of each hospital system. Although these concerns may still pose a barrier for some surgeons or health care administrators to implement the PR, our results and those of other published studies on PR complications do not support a level of complications that exceeds that of the OR. Although we conclude that the PR is safe with a low complication rate that is comparable to that of historical complication rates among OR patients, a formal statistical comparison between our observed rates and internal (e.g. a control group) or historic complication rates was not performed. A subjective comparison of our results with historical published complications is also limited given possible discrepancies in defining and reporting complications in addition to differences in the perioperative and operative management and follow-up course. A large sample would be needed to adequately evaluate differences in safety between procedures performed in the PR versus the OR. Future work possibly using an administrative data is needed to further clarify differences in safety and complications between surgical settings. Given that the median follow-up duration of our study was 14 days, it is likely that our study may not have captured all possible complications such as recurrences and need for reoperation. Additionally, 193 patients had no follow-up visits. It is unknown if these patients had major complications and decided to pursue their care elsewhere. Our short mean follow-up duration of 14 days limits the strength of our analysis of reoperation rates for recurrent or persistent symptoms. Finally, our study did not evaluate the clinical or functional outcomes of patients treated in the PR.

Given that the complication rates seen in our study are low and comparable or lower than those reported for similar procedures done in the OR, we conclude that concern for major medical and surgical complications should not be a barrier to performing minor hand surgeries in the procedure room setting. In fact, in light of these findings and literature supporting patient satisfaction and overall lower costs associated with the PR, we recommend the utilization of the PR for appropriately qualified patients when performing minor hand surgeries.