Long-Term Oral Steroid Use: A Unique Risk Factor in 4-Corner Fusion Compared With Other Wrist Salvage Operations

Abstract

Introduction:

Salvage procedures such as proximal row carpectomy, 4-corner fusion, total wrist arthroplasty, and total wrist arthrodesis are commonly used at the end stages of wrist arthritis. These operations have high complication rates, and significant controversy exists regarding the selection of procedure. Long-term oral glucocorticoid therapy has previously been identified as a risk factor for complications in multiple orthopedic procedures. The purpose of this study is to investigate the effect long-term oral preoperative corticosteroid use has on complications after different salvage operations for wrist arthritis.

Methods:

The National Surgical Quality Improvement Program database was queried to identify patients who underwent proximal row carpectomy, 4-corner fusion, total wrist arthroplasty, or total wrist fusion between 2005 and 2020. Patients were classified by steroid use. Univariate analysis and multivariate logistic regression were used to assess the risk of complications.

Results:

A total of 1298 patients were identified. Overall, steroid use was found to be independently associated with a higher complication rate. On multivariate analysis of patients who underwent 4-corner fusion, steroid use was found to be associated with higher complication rate and surgical site infection rate. Steroid use was not associated with increased complications in patients who underwent proximal row carpectomy, total wrist arthroplasty, or total wrist fusion when examined individually.

Conclusion:

Long-term oral corticosteroid therapy was associated with an increased risk of postoperative infections in patients who underwent 4-corner fusion, which was not found in other wrist salvage operations.

Introduction

Osteoarthritis (OA) is among the most prevalent musculoskeletal conditions, affecting nearly 10% of the world’s population above the age of 60 years.^1,2 Wrist arthritis has multiple etiologies, including post-traumatic, idiopathic, and secondary to inflammatory arthropathies such as gout or rheumatoid arthritis. ³ Depending on the specific insult, degeneration may progress in a predictable fashion. The most recognized of these patterns is scapholunate advanced collapse (SLAC), which occurs as a sequela to a scapholunate ligament injury. ⁴ In addition, a nonunion of a scaphoid fracture can result in a chronic degenerative pattern known as scaphoid nonunion advanced collapse (SNAC).^3,5 Other common etiologies of wrist arthritis include rheumatoid disease, gout, neuropathic conditions, and amyloid. ⁶ In most cases, degeneration develops insidiously over years and is often subclinical until the arthritis becomes more advanced. ⁷

Treatment for wrist arthritis is guided by the patient’s presenting complaints more so than radiographic findings, as patients with SLAC or SNAC wrist may be asymptomatic. ⁸ Symptomatic management begins conservatively with wrist splinting, oral anti-inflammatory medications, and targeted intra-articular steroid injections. ⁹ After failing conservative management, numerous surgical options are available, with difficult treatment decisions made based on patient-specific factors, surgeons’ experience, radiographic findings, and specific joints involved. The least destructive surgical option is the partial or total wrist denervation procedure, which aims to remove some or all of the major nerves associated with pain perception within the wrist joint. This procedure involves an anterior interosseous nerve (AIN) and posterior interosseous nerve (PIN) neurectomy, and in some descriptions other sensory nerves in the periphery of the joint.^10,11 Excision procedures including radial styloidectomy or distal scaphoid pole excision can be used in symptomatic early-stage SLAC or SNAC wrist cases.^12,13 In the absence of midcarpal or radiolunate arthritis, a proximal row carpectomy (PRC) may be used. A PRC involves removal of the scaphoid, lunate, and triquetrum, relying on a new articulation between the lunate fossa and the proximal capitate. Historically, this procedure has not been recommended for younger active patients due to concerns for decreased grip strength and early wear of the radiocapitate articulation.^14,15 For this reason, there are a variety of fusion procedures that have been described. Limited carpal fusions are an option for patients with midcarpal arthritis without significant radiolunate arthritis. Four-corner fusion (4CF) and capitolunate arthrodesis have both been used for this purpose. These both allow the removal of the scaphoid from an arthritic radioscaphoid articulation while preserving the length of the carpus and stabilizing the link between the proximal and distal row, a role formerly filled by the now removed scaphoid. Capitolunate fusion has had a recent resurgence in popularity and seems to have results similar to the classic 4CF with less surfaces requiring bony healing. ¹⁶ Total wrist arthroplasty (TWA) is an option for patients with pancarpal arthritis and has become an increasingly popular surgical option. Total wrist arthroplasty preserves wrist motion but necessitates permanent weightbearing limitations and is classically used for lower demand patients due to known risks of dislocation and loosening.^17,18 Total wrist fusion (TWF) is reserved for patients with pancarpal arthritis, those who have failed previous limited carpal fusions, or those who are not candidates for TWA.^{19 -21}

Long-term oral glucocorticoids are prescribed for a variety of conditions, including inflammatory arthropathies, and are prevalent in slightly more than 1% of the US population. ²² Long-term steroid use has previously been identified as a substantial risk factor for surgical complications, most notably wound-healing complications. The risk of long-term steroid use on surgical complications has been demonstrated in multiple orthopedic disciplines including spine and total joint arthroplasty.^23,24 Within the hand surgery literature, multiple studies have demonstrated that injected steroids are a risk factor for postoperative complications.^{25 -27} However, the data regarding long-term oral steroid use on postoperative outcomes in hand and wrist surgery have remained limited. ²⁸ The purpose of this study is to investigate the effect preoperative long-term oral steroid use has on postoperative complications following various salvage operations for wrist arthritis.

Materials and Methods

A retrospective review using the American College of Surgeon’s National Surgical Quality Improvement Program (NSQIP) database was the source of patient information in this study. It is a validated database used in nearly every surgical field to assess surgical outcomes. The NSQIP database was chosen because of its proven ability to better capture complication data, due to data collection via chart abstraction, rather than the claims data that are present in other similar databases. ²⁹ It also offers national representation and large patient sample size, making it optimal for maximizing power in more rare surgeries. The database includes deidentified demographic information, medical comorbidities, and complications within 30 days of surgery. This study was granted exemption by the Medical University of South Carolina Institutional Review Board as a nonhuman subject study with deidentified data. All patients under the age of 18 years are excluded from this database study. The NSQIP database from 2005 to 2020 was searched by Current Procedural Terminology (CPT) codes for all patients who underwent PRC (25215), 4CF (25820, 25825), TWF (25800, 25805, 25810), and TWA (25446). Patients with missing complication data and open/infected wounds were excluded to minimize variability.

Demographic information collected included age, sex, race, height, weight, and smoking status. Body mass index (BMI) was calculated using height and weight as (weight in kg/height in m²).

Medical comorbidities included diabetes, hypertension, heart failure, chronic obstructive pulmonary disease, dialysis dependency, long-term oral steroid use, and American Society of Anesthesiology (ASA) class. Perioperative factors included operative time and length of hospitalization. Complications assessed include surgical site infections (SSIs), wound dehiscence, deep vein thrombosis, sepsis, pneumonia, and any complication (defined as any one of the previous complications). Rates of unplanned hospital readmission or reoperation within 30 days from the index operation were also queried. Long-term oral steroid use is defined by NSQIP as regular use of oral steroid medications within 30 days prior to surgery for a chronic medical condition. Patients receiving 1-time doses, short courses, or topical or inhaled steroids are not included.

The cases that met inclusion criteria were separated by preoperative long-term steroid use and assessed for differences in demographics and complication rates with univariate analysis. To better understand whether these trends were consistent in the entire cohort or driven by a specific subset(s) of cases, similar analysis was conducted for each of the 4 subgroups (PRC, 4CF, TWF, and TWA). Univariate analysis consisted of χ² test, Fisher exact test, and independent paired t test as appropriate. Binary logistic regression was used in all cases with significance found in multiple variables to account for confounding variables. In this and all calculations, a P value of <.05 was deemed significant.

NSQIP Data Use Statement

American College of Surgeons-NSQIP and the contributing hospitals are the source of data used herein; they have not verified and are not responsible for the statistical validity of the data analysis or the conclusions derived by the authors.

Results

A total of 1298 patients who met inclusion and exclusion criteria were identified from the NSQIP database. The overall cohort demonstrated a mean age of 56.2 years, 58.3% men, and 84.7% white, and 22.4% were smokers within 1 year of surgery. Of this group, 93 (7.16%) were on long-term oral steroids prior to surgery. In all, 87.3% of the procedures were performed as an outpatient, with same-day discharge. There were significant demographic differences between the steroid and nonsteroid groups (Table 1). The steroid group was more likely to be women (P < .001), have a lower BMI (P = .014), be nonsmokers (P = .002), and have a higher ASA score (P < .001). There was also an increase in operative time in the steroid group. There was no statistically significant difference among race (P = .414) or age (P = .450) between groups.

Table 1.

Demographics and Perioperative Details by Steroid Use.

Demographics and comorbidities	No steroid use (N = 1205)	Steroid use (N = 93)	P value
Sex			<.001
Male	60.0%	35.5%
Female	40.0%	64.5%
Race			.414
White	84.4%	88.3%
Nonwhite	15.6%	11.7%
Age (mean)	56.07	57.26	.450
BMI (mean)	30.48	28.59	.014
Current smoking	23.4%	9.7%	.002
Diabetes mellitus	14.1%	11.8%	.663
Dialysis	0.2%	0.0%	1.000
Heart failure	0.5%	2.2%	.107
COPD	4.9%	7.5%	.266
Hypertension	44.2%	39.8%	.405
ASA classification			<.001
1	8.4%	0.0%
2	56.8%	45.2%
≥3	34.9%	54.8%
Operative time, min (mean)	110.16	141.41	<.001
Length of hospitalization, d (mean)	0.54	0.86	.368

Note. All statistically significant values (P < .05) are in bold. BMI = body mass index; COPD = chronic obstructive pulmonary disease; ASA = American Society of Anesthesiologists.

There were 802 patients in the PRC group, 235 in the 4CF group, 100 in the TWA group, and 193 in the TWF group. The different demographic and perioperative details for each group can be seen in Table 2. Significant findings from this analysis revealed the steroid cohort was female-predominant in the 4CF (P = .003) and TWA (P = .008) groups, associated with decreased BMI in the 4CF group (P = 0.027) and increased ASA score in the PRC group (P = .004).

Table 2.

Demographics by Surgery and Steroid Use.

	Proximal row carpectomy (N = 802)			Four-corner fusion (N = 235)			Total wrist arthroplasty (N = 100)			Total wrist arthrodesis (N = 193)
Demographics	No steroid use (N = 769)	Steroid use (N = 33)	P value	No steroid use (N = 212)	Steroid use (N = 23)	P value	No steroid use (N = 87)	Steroid use (N = 13)	P value	No steroid use (N = 164)	Steroid use (N = 29)	P value
Sex			.279			.003			.070			.008
Male	61.2%	48.5%		62.1%	30.4%		43.7%	15.4%		58.3%	31.0%
Female	38.8%	51.5%		37.9%	69.6%		56.3%	84.6%		41.7%	69.0%
Race			.802			.705			.586			1.000
White	82.2%	85.7%		87.1%	82.4%		88.6%	100%		89.1%	88.9%
Nonwhite	17.8%	14.3%		12.9%	17.6%		11.4%	0.0%		10.9%	11.1%
Age, mean	55.58	56.79	.630	55.79	58.91	.337	63.39	58.31	.213	55.04	55.79	.817
BMI, mean	31.09	29.87	.356	29.90	26.97	.027	29.74	29.11	.734	28.43	27.84	.647
Current smoking	25.8%	12.1%	.100	17.5%	4.3%	.139	20.7%	15.4%	.736	23.8%	13.8%	.333
Diabetes mellitus	15.1%	12.1%	.539	10.8%	8.7%	.412	17.2%	23.1%	.058	13.4%	6.9%	0.519
On dialysis	0.1%	0.0%	1.000	0.0%	0.0%	—	0.0%	0.0%	—	0.6%	0.0%	1.000
Heart failure	0.5%	3.0%	.188	0.5%	4.3%	.187	0.0%	0.0%	—	0.6%	0.0%	1.000
COPD	5.5%	6.1%	1.000	2.8%	8.7%	.179	8.0%	7.7%	1.000	5.5%	10.3%	.395
Hypertension	43.9%	48.5%	.391	41.5%	34.8%	.533	63.2%	46.2%	.239	39.0%	27.6%	.240
ASA classification			.004			.051			.712			.051
1	7.5%	0.0%		12.3%	0.0%		3.4%	0.0%		9.1%	0.0%
2	56.3%	36.4%		60.4%	52.2%		54.0%	46.2%		56.1%	44.8%
≥3	36.2%	63.6%		27.4%	47.8%		42.5%	53.8%		34.8%	55.2%
Operative time, mean	89.50	120.88	.050	134.82	139.04	.743	130.01	139.08	.594	174.18	183.00	.701
Length of hospitalization, mean	0.47	0.27	.775	0.33	1.83	.166	0.67	1.23	.100	1.06	1.36	.569

Note. All statistically significant values (P < .05) are in bold. BMI = body mass index; COPD = chronic obstructive pulmonary disease; ASA = American Society of Anesthesiologists.

The differences in complications between the steroid and nonsteroid groups can be seen in Table 3. When comparing complications among all procedures between the steroid and nonsteroid groups, patients on preoperative steroids were at increased risk of any complication (P = .009; odds ratio [OR], 4.83; confidence interval [CI], 1.702-13.728) and pneumonia (P = .015; OR, 26.462; CI, 2.377-294.591). The rate of SSI and wound dehiscence approached but did not meet statistical significance (P = .157 and P = .138, respectively). Logistic regression of all significant demographic, medical, or perioperative factors determined that steroid use was independently associated with the presence of any complication (P = .013; OR, 6.212; CI, 1.356-12.732) (Table 4).

Table 3.

Complications by Steroid Use.

Complications	No steroid use (N = 1205), %	Steroid use (N = 93), %	P value
Any complication	1.2	5.4	.009
Surgical site infection	0.7	2.2	.157
Wound dehiscence	0.1	1.1	.138
Deep vein thrombosis	0.1	0.0	1.000
Sepsis	0.2	0.0	1.000
Pneumonia	0.1	2.2	.015
Reoperation	0.7	1.1	1.000
Readmission	1.7	2.2	.655

Note. All statistically significant values (P < .05) are in bold.

Table 4.

Logistic Regression for Any Complication by Steroid Use in All Cases.

Demographics	OR	95% CI	P value
Sex	1.627	0.618-4.286	.324
BMI	0.972	0.905-1.044	.431
Current smoking	1.400	0.488-4.021	.531
ASA status	—	—	.554
Steroid use	6.212	1.356-12.732	.013
Operative time	1.002	0.997-1.007	.463

Note. All statistically significant values (P < .05) are in bold. OR = odds ratio; CI = confidence interval; BMI = body mass index; ASA = American Society of Anesthesiologists.

Univariate analysis of the frequency of complications by steroid status in each of these groups can be seen in Table 5. This analysis revealed a relationship between steroid use and the occurrence of SSIs (P = .026) and any complication (P = .003) in 4CF, which was not found in the other 3 wrist salvage operations. Binary logistic regression was performed for 4CF (Table 6), which showed steroid use as an independent predictor for any complication (P = .017; OR, 19.463; CI, 1.719-220.401) and SSIs (P = .048, OR, 14.742; CI, 1.028-211.414).

Table 5.

Complications by Surgery and Steroid Use.

	Proximal row carpectomy (N = 802)			Four-corner fusion (N = 235)			Total wrist arthroplasty (N = 100)			Total wrist arthrodesis (N = 193)
Complications	No steroid use (N = 769), %	Steroid use (N = 33), %	P value	No steroid use (N = 212), %	Steroid use (N = 23), %	P value	No steroid use (N = 87), %	Steroid use (N = 13)	P value	No steroid use (N = 164), %	Steroid use (N = 29), %	P value
Any complication	0.8	0.0	1.000	0.5	13.0	.003	0.0	0.0	—	4.3	6.9	.626
Surgical site infection	0.4	0.0	1.000	0.5	8.7	.026	0.0	0.0	—	2.4	0.0	.626
Wound dehiscence	0.1	0.0	1.000	0.0	0.0	—	0.0	0.0	—	0.0	3.4	.150
Other wound complication	0.3	0.0	1.000	0.0	0.0	—	0.0	0.0	—	1.2	0.0	1.000
Deep vein thrombosis	0.1	0.0	1.000	0.0	0.0	—	0.0	0.0	—	0.0	0.0	—
Sepsis	0.3	0.0	1.000	0.0	0.0	—	0.0	0.0	—	0.0	0.0	—
Pneumonia	0.0	0.0	—	0.0	4.3	.098	0.0	0.0	—	0.6	3.4	.279
Reoperation	0.4	0.0	1.000	0.5	0.0	1.000	2.3	0.0	1.000	2.4	3.4	1.000
Readmission	1.0	0.0	1.000	2.1	9.1	.117	3.7	0.0	1.000	4.1	0.0	.591

Note. All statistically significant values (P < .05) are in bold.

Table 6.

Logistic Regression in 4-Corner Fusion Cases.

Demographics	Any complication			Surgical site infection
	OR	95% CI	P value	OR	95% CI	P value
Sex	—	—	.996	—	—	.996
BMI	1.026	0.875-1.203	.752	1.070	.909-1.259	.417
Steroid Use	19.463	1.719-220.401	.017	14.742	1.028-211.414	.048

Note. All statistically significant values (P < .05) are in bold. OR = odds ratio; CI = confidence interval; BMI = body mass index.

Power analysis conducted post hoc demonstrated that this study was well powered regarding the relationship between steroid use and occurrence of any complication in patients undergoing 4CF (P = .84) and slightly underpowered regarding SSIs (P = .73).

Discussion

There are numerous surgical options available for treating wrist arthritis; however, it is the role of the surgeon to determine the best option for each individual patient.^{10 -15,17 -21} To optimize outcomes for each patient, surgeons must understand the unique risk factors associated with surgical options available. Previous studies have indicated that long-term steroid use is a risk factor for postoperative surgical complications; however, in hand and upper extremity surgery, these trends have not always held true.^23,24,28 This study uses a nationally validated database, with particular strength in complication identification, to demonstrate the effect preoperative steroids have on complications following surgical treatment options of wrist arthritis, including PRC, 4CF, TWF, and TWA.

When analyzing all types of salvage operations together, long-term preoperative steroid use was found to be independently associated with any complication and postoperative pneumonia. Although the rate of SSI was nearly 4 times as high in the steroid cohort (0.6% vs 2.2%), statistical significance was not reached (Table 3). Upon subgroup analysis of the 4 different groups, patients undergoing limited carpal fusion such as 4CF were at increased risk of surgical site complication and any complication (Table 5), which was not present in any of the other salvage operations studied. Our study did demonstrate an increased operative time in the steroid group, potentially resulting from challenges in fixation and more complicated deformities in these patients on chronic steroids. This increased time and inherent soft tissue insult may contribute to infection risk, but when controlling for cofounders, these trends were present (Table 6). Preoperative steroid use was not found to be significantly predictive of complications in any of the other surgical subgroups studied, thus implying that the significance of the entire cohort is a result of the 4CF group.

Long-term steroid use has been identified as a surgical risk factor in multiple orthopedic subspecialties. A retrospective review including more than 360 000 patients conducted by Singla et al, ²³ published in 2019, found that patients undergoing lumbar spine fusion are at increased risk of mortality and SSIs. This increased risk has also been demonstrated in other orthopedic disciplines. A study by Cordero-Ampuero and de Dios ²⁴ found that patients on long-term oral glucocorticoids were at increased risk of infection following total hip arthroplasty or hemi-arthroplasty. Furthermore, there is evidence for increased risk of localized infection in time periods closely following corticosteroid injections within the upper extremity.^{25 -27} Although long-term steroid use in lower extremity orthopedics and localized injections in the upper extremity do seem to have specific increases in complication profile, short-term perioperative and postoperative corticosteroid use has demonstrated beneficial effects on recovery in upper extremity surgery without an apparent increase in risk. ³⁰

To our knowledge, there has been no study investigating the effect of long-term steroids on wrist salvage procedure outcomes. In studies of other upper extremity surgeries, steroids were not identified as risk factors. A study by Jain et al ²⁸ of patients with rheumatoid arthritis undergoing hand surgery found no significant relationship between steroids or methotrexate and surgical wound complications. A database study by Jiang et al ³¹ of more than 3000 patients who underwent surgical fixation of distal radius fractures found that patients on steroids were at increased risk of any complication within 30 days following surgery; however, they determined that steroids were not independently associated with complications after performing multivariate analysis. Risk factors for complications or reoperation following 4CF have previously been studied. Work by van Hernen et al ³² and Li et al ³³ identified preoperative narcotic use, history of manual labor, and the absence of AIN/PIN neurectomy to be risk factors for complications or reoperation following 4CF. Neither of these studies addressed the role of oral corticosteroids. Vitamin A has been shown to have the unique ability to counteract the inhibitory effects of steroids on wound healing.^34,35 In light of this, future studies investigating the possible ability of vitamin A supplementation to improve wound outcomes in orthopedic patients on chronic oral steroids may be beneficial.

There are limitations that warrant discussion. Because of the nature of this project as a database study, it is limited by coding accuracy and data input errors. In addition, no standardized patient-reported outcome metrics are included in the NSQIP database. The reporting of complications is limited to 30 days following surgery, and thus, complications such as reoperation due to continued pain, hardware complications, or nonunion are not included in the NSQIP data. Although this study attempts to control by medical comorbidities, patients are taking chronic oral steroids for a medical condition, and therefore, this study is not fully able to differentiate whether the oral steroids or the underlying disease for which the steroids are being taken is the risk factor. This study includes only procedures performed in the hospital setting, both as an inpatient and as an outpatient. It is possible that patients in the chronic corticosteroid therapy cohort were more likely to have surgery performed in such a setting, rather than a free-standing surgery center, introducing a confounding factor. Finally, although this study included a large cohort of patients, a larger sample of patients would be helpful in evaluating the rare complications of less commonly performed procedures such as TWA or TWF. Future studies would benefit from having longer patient follow-up, standardized outcome metrics to assess quality of life, and an increased study population size.

Total wrist arthrodesis and wrist arthroplasty have specific indications that do not typically overlap with those of 4CF and PRC. On the contrary, it is common that treating surgeons are faced with choosing between PRC and limited carpal fusion, classically reserving PRC for lower demand patients with less reliance on grip strength which may be decreased in PRC. While grip strength has been shown to be decreased slightly following PRC, there are increasing data demonstrating that outcomes following PRC and 4CF are likely equivalent even in younger patients.^36,37 In addition, there is an increasing trend toward utilization of PRC so as to eliminate the risk of nonunion present with midcarpal fusions. Furthermore, it seems the rates of progression of symptomatic radiocarpal arthritis following PRC and 4CF are equivalent as both alter the normal kinematics of the wrist. ³⁸

Conclusion

Long-term oral corticosteroid therapy is associated with an increased risk of postoperative infections in patients who underwent 4CF, which was not found in the other wrist salvage operations. The results of this study may help to provide further information to guide treatment decisions in this often-difficult clinical scenario, potentially favoring PRC for patients on long-term steroid therapy.