The Influence of Corticosteroid Injections on Postoperative Outcomes of Carpal Tunnel Release: A Systematic Review

Background  Carpal tunnel syndrome can be treated with corticosteroid injections (CIs) and surgery. In this systematic review, the influence of previous CI on different postoperative outcomes after carpal tunnel release is evaluated. Methods  A systematic literature search using several databases was performed to include studies that examined patients diagnosed with carpal tunnel syndrome who received preoperative or intraoperative CIs. Results  Of 2,459 articles, 9 were eligible for inclusion. Four papers reported outcomes of preoperative and four outcomes of intraoperative CIs. One study evaluated patients who received both intraoperative and preoperative corticosteroids. Conclusion  Intraoperative CIs are associated with reduced postoperative pain after carpal tunnel release and support earlier recovery of the hand function that can be objectified in a faster median nerve conduction speed recovery and lower Boston Carpal Tunnel Questionnaire (BCTQ) scores. Using preoperative CIs did not lead to enhanced recovery after carpal tunnel release, and both preoperative and intraoperative CIs might be predisposing factors for infections.


Introduction
Carpal tunnel syndrome (CTS) is a commonly occurring nerve entrapment disorder with a reported incidence rate of up to 2.3 cases per 100 person-years. 1Symptoms mostly consist of tingling, numbness, or burning sensation following the median nerve path distally from the carpal tunnel.Of all conservative treatments, splinting and corticosteroid injection (CI) are the most notable ones.For some patients, the injection of a corticosteroid provides enough relief to CTS symptoms that they no longer need any further treatment.However, when failure of response to CI and for that matter other conservative treatments occurs, patients will likely eventually undergo surgical carpal tunnel release (CTR). 1,2ven though multiple studies investigated the influence of CIs on CTR, the clear influence of pre-and postoperative CIs remains unclear.
Therefore, a systematic review was conducted to summarize the benefits and disadvantages of pre-and intraoperative CIs in patients with CTR.

Methods
The PRISMA guidelines were followed during all stages of this systematic review. 3

Search Methods
In collaboration with a professional independent librarian, multiple databases, MEDLINE, Embase, Web of Science, and PubMed, were systematically searched to find eligible articles on CTR and CIs (►Appendix A). References lists of included studies were hand searched for relevant studies.There were no geographic or language restrictions.The last update on the search results was performed in September 2022.No contact was made with authors for unpublished data or results.The review of the titles, abstracts, and full-text articles were done independently by two researchers.

Eligibility Criteria
Studies were included if they contained at least 10 clinically diagnosed patients with atraumatic CTS who underwent open CTR and underwent a pre-or intraoperative CI.
Studies were excluded based upon the following criteria: not acquirable in full text, or no clear description of postoperative results.
Endnote X5 was used during the processing of the search results. 4Duplicate articles were deleted.
For the first round of reviewing, articles were included based upon their titles and abstracts.A third individual researcher acted as the third observer, who was consulted when a consensus was out of reach.Following this, all articles were under full text review to be checked for inclusion and exclusion criteria by both researchers.

Quality Assessment
Depending on study type, quality assessment and risk-ofbias evaluation were done by two reviewers.A modified Cochrane risk-of-bias tool (RoB 2) was used for the randomized trials. 5After scoring the RoB 2, an overall risk-of-bias judgment of either low, some concerns, or high was made (►Table 1).Quality assessment of the cohort studies was done using the Newcastle-Ottawa Scale (NOS). 6The finalized result of the NOS is a score ranging from 1 to 9 stars, with 7 to 9 stars being regarded as high quality, 4 to 6 moderate quality, and 1 to 3 low quality (►Table 2).

Data Extraction
The data of the studies were independently extracted by two researchers.From the studies, the following data were extracted: patient population, demographics, number of injections for each patient, duration between injection and surgery, type and dosage of injection, number of infections, functional outcome report, nerve conduction measurements, Boston Carpal Tunnel Questionnaire (BCTQ) scores, and eventually other relevant outcomes (►Table 3).

Outcomes
The primary outcomes of this study are infection and nerve conduction rates for patients receiving either pre-or intraoperative CI and for control patients.Secondary outcomes are any patient-reported outcomes with well-described outcomes before and after treatment and other CTS-related complaints.

Results
The database search identified 2,470 papers, and 1 other paper was added through scanning the reference lists of included articles. 7A total of 1,372 papers were left after removing duplicates and these papers were screened for eligibility on title and abstract.][9][10][11][12][13][14][15] One study used both intraoperative and preoperative CIs. 14 The reviewing process is shown in ►Fig. 1.
Four of five studies using intraoperative CIs were randomized controlled trials, and one study was a retrospective cohort study.This study also included a preoperative CI cohort. 14A summary of a risk of bias is given in ►Table 1.All five studies using preoperative CIs were retrospective cohort studies.A summary of the quality assessment of these articles is given in ►Table 2.

Intraoperative Corticosteroid Injections
The five studies included in the intraoperative CI consisted of four randomized controlled trials and one retrospective trial,

Infections
8][9] Hanssen et al reported 13 (22%) postoperative infections in the intraoperative CI group and 2 (4.5%) in the control group (►Table 4). 14The intraoperative CI group had a significant (p < 0.002) higher risk of having an infection.There was a difference between incidence of infections in male (0.87%) and female (0.25%) patients, which was also significant (p < 0.02).In addition, a total of 19 patients received prophylactic antibiotics, which did not change the statistical analysis; the difference in postoperative wound infections between the intraoperative CI and control group was still significant.

Postoperative Pain
Only one of the included studies reported on postoperative pain after CTR. 9 Naji evaluated postoperative pain after CTR in 40 patients.In total, 20 patients received an intraoperative CI and none of them (0%) reported postoperative pain complaints.In the control group of 20 patients, 11 patients (55%) had persisting agonizing pain postoperatively at the ulnar side of the wrist (incision site), which was not responding to analgesic treatment.The other nine patients (45%) did not have any postoperative pain complaints.This difference was found to be significant (p < 0.05).

Symptom Relief
One study described postoperative symptom relief of CTS. 10 Stepić et al registered the disappearance of symptoms in 40 patients after CTR, of which 20 patients received an intraoperative CI.After 3 months, only one patient (5%) receiving intraoperative CI still experienced symptoms, compared with two patients (10%) in the control group.Intraoperative CI did not lead to an increased proportion of patients without symptoms when compared with patients receiving no CI (►Fig.2).

Median Nerve Conduction Speed
Two studies performed electrophysiological measurements in a total of 82 patients before and after CTR, comparing 40 patients receiving an intraoperative CI with 42 patients in a control group without CI. 7,10The median (sensitive) nerve conduction speed and the distal motor and sensory latencies were measured.In 20 patients receiving intraoperative CI, a faster and bigger recovery in electrophysiological measurements was established, compared with the control group.
In Stepić et al, values of the median nerve conduction speed and sensitive conduction speed were measured both preoperatively and postoperatively after 7, 30, and 90 days. 10wenty patients receiving intraoperative CI were compared with 20 patients in the control group (►Fig.3a, b).The median nerve conduction speed measurements 7 days postoperatively were not significantly different between these groups.After 30 days, the median nerve conduction speed improved in both groups, but this improvement was significantly greater in the CI group (p ¼ 0.025).After 90 days, both groups showed improvement of the median nerve   Corticosteroid Injections and Carpal Tunnel Release Kumaş et al.
conduction speed in comparison to the preoperative situation, which was significantly greater in the CI group (p ¼ 0.043).
In Mottaghi et al, the median sensitive conduction speed after 30 days also improved significantly in the CI group compared with the control group (p ¼ 0.018). 7The measurements after 90 days did not show any significant difference between the CI group and the control group.
Furthermore, the nerve conduction speed in the distal motor and sensory nerves showed a greater improvement in the intraoperative CI group compared with the control group (►Fig.4). 7The distal motor latency improved by 30.0% in the steroid group and by 25.3% in the control group, compared with the preoperative situation.The distal sensor latency improved by 37.7% in the steroid group and by 26.3% in the control group.However, the data were not significantly different (preoperative p ¼ 0.671, postoperative 7 days p ¼ 0.404, postoperative 30 days p ¼ 0.16, postoperative 90 days p ¼ 0.977).

Boston Carpal Tunnel Questionnaire
In two studies, pre-and postoperative BCTQ scores were compared in a total of 62 patients: 30 intraoperative CI patients and 32 control patients. 7,8The improvement in the BCTQ scores of the CI groups was shown in both studies.The mean BCTQ score improved by 63.3% in 30 patients with intraoperative CI and by 48.7% in 32 patients without CI.
The first study, Mottaghi et al, used the sum of the symptom and functional scales to determine a BCTQ index (►Fig. 5). 7Both groups showed improvement of symptoms (p < 0.001) after CTR, but the CI group showed greater improvement (the mean BTCQ improved with 55.8%)    Padua et al also used the BCTQ index but showed the values in scales: the Symptom Severity Scale (SSS) and Functional Status Scale (FSS) of the BCTQ (►Fig.6). 8In 2 months, the SSS improved by 73.4% in the CI group and by 53.6% in the control group, compared with the preoperative scores.The FSS improved by 68.1% in the CI group and by 42.8% in the control group.The changes from baseline to 15 and 60 days showed a significant difference in the SSS in both groups (p ¼ 0.0035 and 0.005, respectively) compared with the preoperative scores.The postoperative FSS differences were not significant between the study groups.

Infection
Two studies mentioned if any infection occurred.Hanssen et al reported two (12.5%)infections in the preoperative CI group and two (4.5%) in the control group (►Table 5). 14There was no statistically significant correlation between preoperative CI and infection in this study.There was a difference between incidence of infections in male (0.87%) and female (0.25%) patients, which was significant (p < 0.02).In total, 19 patients received prophylactic antibiotics, which did not change the statistical analysis; the difference was still not significant.
Kirby et al analyzed patients with postoperative infection retrospectively and compared these to control patients (►Table 5). 15Patients in the infection group had a significantly shorter period between CI and surgery (77 AE 52 days) than in the control group (133 AE 89 days) (p ¼ 0.05).Thirty-six patients (92%) had a superficial postoperative infection, and three patients (8%) had a deep postoperative infection (►Table 6).
Only one study reported on BCTQ scores in patients receiving preoperative CIs.Bland and Ashworth used the BCTQ to measure the symptom severity and functional status in 942 patients: 278 patients underwent a preoperative CI and 664 patients did not (►Fig.7). 12The postoperative SSS and FSS scores showed more improvement in the control group when compared with the preoperative corticosteroid   group but the improvements were not significant (SSS p ¼ 0.37 and FSS p ¼ 0.40).

Patient-Reported Outcomes
Two studies used patient-reported outcomes of 1,170 patients to evaluate the success of preoperative CI. 11,12 A total of 182 patients (39.7%) receiving a preoperative CI were free of symptoms postoperatively, compared with 268 patients (37.7%) in the control group.Also, 277 patients (60.3%) receiving a preoperative CI still experienced symptoms such as pain and paraesthesia postoperatively, compared with 443 patients (62.3%) in the control group.Bland and Ashworth found no significant difference in patientreported outcomes after CTR in the injection group compared with the control group (p ¼ 0.82). 12ahi et al found that patients who received a preoperative CI had a higher occurrence of complaints (pain [p ¼ 0.04], paraesthesia [p ¼ 0.007], and nocturnal awakenings [p ¼ 0.003]) when compared with the control group. 11Within the injection group, there was an association between patients receiving more injections and the reporting of the specific complaints.However, this difference was not significant (p ¼ 0.09).Other studies did not mention if there were any pain complaints postoperatively.
In Edgell et al, patients were considered cured if they did not have any complaints after 6 months postoperatively. 13In total, 13 patients (87%) with some relief after CI underwent surgery, while 21 patients (54%) without relief after CI underwent surgery.
In Vahi et al, the patients were also asked to fill in the Numeric Rating Scale (NRS) from 0 to 100: 100 points meaning total regression of symptoms and 0 meaning no effect. 11In total, 120 patients had regression of symptoms, meaning they scored 90 or more points in the NRS.In the control group, 33 patients (70%) had a score of 90 or more points, while in the steroid group 86 patients (68%) scored 90 points or more.The pain was still present in 14 patients (30%) in the control group and in 41 patients (32%) in the steroid group.The difference in NRS scores between patients receiving preoperative injections and the control group was not significant.

Nerve Conduction Studies
Edgell et al measured median sensory values of 31 patients with preoperative CI, in which 24 (77%) patients had diminished values, 16 (67%) patients were symptom-free after surgery, and 7 (23%) patients had a normal median sensory value; 3 (43%) of these patients showed complete relieve by CTR. 13 The number of patients who were symptom-free after CTR did not differ significantly between the group with retarded and normal median sensory values preoperatively (p ¼ 0.255).The median motor measurement was performed on 37 patients: 24 (65%) patients had a diminished value, of whom 15 (63%) recovered after CTR, and 13 (35%) patients had a normal value, of whom 6 (46%) recovered after CTR (p ¼ 0.338).The number of patients who were symptom-free after CTR was not significantly different between the groups with retarded or normal median motor measurements preoperatively (►Table 7).

Discussion
][9][10][11][12][13][14][15] A meta-analysis could not be performed on any of the outcome measures due to the lack of data and heterogeneity of the data in the included studies.
Administering intraoperative CI might be beneficial for reduced postoperative pain after CTR and lead to earlier recovery of the hand function. 9,10However, no benefit is seen in using intraoperative CI for the final outcomes in terms of symptoms, function, and electrophysiology. 7,8,10anssen et al found that an intraoperative CI increases the risk of having a postoperative infection compared with the control group. 14This study, however, associates the infection rates with extended operative techniques, such as flexor tenosynovectomy and the insertion of a surgical drain, and not with the steroid usage.9][10] Table 6 Infections after preoperative corticosteroid use in Hanssen et al 14  In Mottaghi et al, all patients received postoperative prophylactic cefazolin 500 every 6 hours for 3 days, which can cause unnecessary antibiotic resistance due to the fact that intraoperative CI does not significantly induce infections. 7irby et al showed that the usage of preoperative CI increases the risk of having a postoperative infection significantly. 15However, the number of infections seems to be underrated, due to limited reporting or diagnosis of postoperative wound infections in different clinics.They recommend informing patients about the risk of infection after CTR when CIs are given in the prior 2 to 3 months. 15][13] There is no evidence for improved recovery after CTR when using preoperative CIs.Specifically, Bland and Ashworth found no difference in postoperative complaints after CTR between patients receiving a preoperative CI and the control group. 12A possible advantage of preoperative CI could be a holdup for a CTR.Disadvantages of this study were the small sample size, a big loss in follow-up, and heterogeneity between the study groups due to the direct offer to undergo CTR to patients with serious complaints.Vahi et al found that preoperative CIs lead to more postoperative complaints, such as pain, paresthesia, and numbness. 112][13] In conclusion, there is no evidence for improved recovery after CTR when using preoperative CIs.Additionally, the adverse effects of CIs such as skin depigmentation and subcutaneous atrophy should not be forgotten. 16 limitation of this systematic review is the heterogeneity of the included studies, especially in the preoperative CI studies.The time between injection and surgery are different in all the studies, such as the number of CI per Furthermore, the quality of four of the included studies addressing preoperative CIs is poor.A prospective randomized trial with large cohort of patients and a longer follow-up in the future could increase our knowledge of CI usage in patients after CTR.

Fig. 2
Fig. 2 Postoperative recovery of symptoms in patients (%) after carpal tunnel release in Stepić et al.

Fig. 4
Fig. 4 Motor and sensory latency values (m/s) in Mottaghi et al.

Table 1
Characteristics of included intraoperative corticosteroid injection studies and risk-of-bias assessment Archives of Plastic Surgery Vol.50 No.4/2023 © 2023.The Author(s).

Table 2
Characteristics of included preoperative corticosteroid injection studies and quality assessment Abbreviation: Nm, not mentioned.

Table 3
Summary of data of studies included Archives of Plastic Surgery Vol.50 No.4/2023 © 2023.The Author(s).Corticosteroid Injections and Carpal Tunnel Release Kumaş et al. 401

Table 5
Overview of results of intraoperative corticosteroid injections Abbreviations: CI, corticosteroid injection; Nm, not mentioned; NRS, Numeric Rating Scale.

Table 7
and Kirby et al 15 Corticosteroid Injections and Carpal Tunnel Release Kumaş et al. 405 Overview of results of preoperative corticosteroid injections Fig. 7 BCTQ SSS and FSS scores preoperatively and postoperatively in Bland and Ashworth.BCTQ, Boston Carpal Tunnel Questionnaire; SSS, Symptom Severity Scale.Archives of Plastic Surgery Vol.50 No.4/2023 © 2023.The Author(s).Abbreviations: CI, corticosteroid injection; Nm, not mentioned.