Abstract
Purpose
To identify which preoperative diagnostic findings, individually and in combination, are most strongly associated with a surgeon’s recommendation for carpal tunnel release.
Methods
This retrospective cohort study included 313 wrists from 218 patients evaluated for carpal tunnel syndrome (CTS) by one of three surgeons at a tertiary academic center between 2012 and 2020. Each wrist had a documented surgical recommendation and complete preoperative data, including CTS-6 score, ultrasound median nerve cross-sectional area (CSA), and at least one nerve conduction study parameter. Demographic variables, such as age, sex, and body mass index, were also analyzed. Nerve conduction study parameters included distal motor latency (DML), distal sensory latency (DSL), compound muscle action potential, and sensory nerve action potential (SNAP). Associations with surgical recommendation were examined using generalized estimating equation logistic regression. Models were compared by the area under the receiver operating characteristic curve and evaluated for sensitivity, specificity, positive predictive value, and negative predictive value.
Results
All six diagnostic tests were significantly associated with surgical recommendation on univariable analysis (all P < .05). The multivariable model combining CTS-6, DML, DSL, SNAP, and CSA achieved the highest discrimination. CTS-6, DML, and CSA demonstrated significant independent associations with surgical recommendation in both unadjusted and covariate-adjusted models (all P ≤ .01), whereas DSL, SNAP, and body mass index were not significant. SNAP-based models exhibited high specificity and positive predictive value but were limited by an imbalance in SNAP results between groups.
Conclusions
Surgical recommendation for CTS is influenced by complementary diagnostic information across clinical, physiologic, and anatomic domains. CTS-6, DML, and CSA were most consistently associated with surgical recommendation, highlighting the value of a multimodal diagnostic framework that integrates clinical, electrodiagnostic, and ultrasound findings when assessing surgical candidacy.
Type of study/level of evidence
Diagnostic III.
Keywords: Carpal tunnel syndrome, Diagnostic tests, Nerve conduction studies, Surgical decision-making, Ultrasound imaging
Carpal tunnel syndrome (CTS) is the most common peripheral nerve compression disorder and can result in pain, numbness, and functional deficits that impair daily activities and work participation.1 Although multiple diagnostic tools are available, there is no universally accepted standard for diagnosis. This lack of standardization extends to surgical decision-making, where indications for carpal tunnel release are not strictly defined and often vary among clinicians.2
Nerve conduction studies (NCS), ultrasound measurement of median nerve cross-sectional area (CSA), and clinical scoring systems such as the CTS-6 are frequently used to evaluate suspected CTS.3 Each tool contributes different information: NCS provides physiologic data on nerve conduction, ultrasound visualizes structural enlargement of the median nerve, and CTS-6 integrates symptom and examination findings into a standardized clinical score. However, reported diagnostic performance measures for these tools vary across studies, and interpretation can be influenced by testing thresholds, equipment, and operator technique.4,5 Reflecting this uncertainty, the 2016 American Academy of Orthopaedic Surgeons guideline removed the recommendation to routinely obtain NCS before carpal tunnel release, citing limited evidence to support its necessity.6 National claims data later showed that electrodiagnostic study (EDS) use was already declining before and immediately after the guideline, but subsequently began to rise again in the post-clinical practice guideline (CPG) period.7 This rebound highlights persistent variability in diagnostic testing practices and continued reliance on these tools in clinical decision-making. As a result, understanding how clinical, electrodiagnostic, and ultrasound findings are integrated to inform surgical recommendation has become increasingly important.
Despite the widespread use of these diagnostic modalities, little is known about how individual or combined test results influence a surgeon’s decision to recommend surgery. The existing literature has focused primarily on diagnostic accuracy rather than on the practical question of how test results translate into surgical decision-making.4 Understanding which findings most strongly align with operative recommendation may clarify how surgeons synthesize available data and help standardize evaluation across practices. The purpose of this study was to identify which preoperative diagnostic findings, used individually and in combination, are most strongly associated with a surgeon’s recommendation for carpal tunnel release.
Materials and Methods
Study design and population
This retrospective cohort study included patients evaluated for CTS by one of three surgeons (including author J.F.) at a single academic tertiary care center between 2012 and 2020. Patients were identified from an institutional clinical database comprising evaluations performed by multiple fellowship-trained hand surgeons. Wrists were eligible if CTS was the primary working diagnosis, and the record contained demographic information, a documented surgical recommendation, and complete preoperative testing, including CTS-6 score, ultrasound CSA, and at least one NCS parameter. Distal motor latency (DML), distal sensory latency (DSL), compound muscle action potential (CMAP), and sensory nerve action potential (SNAP) values were included when available. Wrists with prior carpal tunnel release, those evaluated primarily for alternative neuropathies, and those with incomplete diagnostic records were excluded.
All diagnostic test results were dichotomized using established clinical thresholds: CTS-6 score ≥ 12, CSA ≥ 10 mm2, DML ≥ 4.0 ms, DSL ≥ 3.5 ms, SNAP ≤ 6 μV, and CMAP < 10 mV were classified as positive findings. Each wrist was treated as a separate observational unit and generalized estimating equation (GEE) models accounted for within-subject correlation by clustering wrists belonging to the same patient. The primary outcome was whether the evaluating surgeon recommended carpal tunnel release at the clinical visit, as documented in the treatment plan. The documented surgical recommendation used for analysis reflected the surgeon’s final recommendation after all diagnostic information was available at the visit.
This study received approval from the University of Pittsburgh Institutional Review Board (STUDY20070308), with waiver of informed consent and Health Insurance Portability and Accountability Act (HIPAA) authorization due to its retrospective cohort design. The investigation was conducted in accordance with the Strengthening the Reporting of Observational Studies in Epidemiology guidelines.
Statistical analysis
Descriptive statistics were calculated to summarize demographic and diagnostic variables. Wilcoxon rank-sum and Pearson chi-square tests were used to compare continuous and categorical variables, respectively, between surgical recommendation groups. Associations between diagnostic variables and surgical recommendation were evaluated using GEE logistic regression models with a logit link function and an independent working correlation structure, clustering by patient to account for correlation between bilateral wrists. Each of the six diagnostic tests and all possible combinations thereof were analyzed using univariable and multivariable GEE models. Multivariable models were subsequently adjusted for age, sex, ethnicity, and body mass index (BMI) to determine whether these factors influenced the associations between diagnostic findings and surgical recommendation. These covariates were selected a priori based on prior literature and clinical relevance to CTS. Model discrimination was quantified using the area under the receiver operating characteristic curve (AUC). Sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) were calculated using optimal thresholds determined by Youden's J statistic. Odds ratios (ORs) with 95% confidence intervals (CIs) were reported for each variable, with significance level set to α = 0.05. Model performance was compared across unadjusted and covariate-adjusted GEE models, with the model achieving the highest AUC identified as the best-performing model.
Results
A total of 313 wrists from 218 individuals were included in the analysis.
Wrist-level descriptive characteristics
Table 1 summarizes wrist-level demographic and diagnostic characteristics stratified by surgical recommendation status. Body mass index and all six diagnostic tests were significantly different between wrists that were versus were not recommended for surgery (all P < .05). Among wrists not recommended for surgery (n = 98), 97 had negative SNAP findings, resulting in limited variability for SNAP in subsequent models.
Table 1.
Wrist-Level Demographics and Preoperative Diagnostic Tests, Stratified by Surgical Recommendation
| Characteristic | No Surgical Recommendation (N = 98) | Surgical Recommendation (N = 215) | P |
|---|---|---|---|
| Age | 47.90 (14.8) | 50.56 (12.7) | .095 |
| Sex | .975 | ||
| F | 74 (76%) | 162 (75%) | |
| M | 24 (24%) | 53 (25%) | |
| Ethnicity | .472 | ||
| Non-White | 26 (27%) | 49 (23%) | |
| White | 72 (73%) | 166 (77%) | |
| Diabetes | .646 | ||
| N | 81 (83%) | 173 (80%) | |
| Y | 17 (17%) | 42 (20%) | |
| BMI | 29.1 (6.1) | 32.4 (7.5) | <.001 |
| CTS-6 | <.001 | ||
| (−) | 63 (64%) | 58 (27%) | |
| (+) | 35 (36%) | 157 (73%) | |
| Ultrasound Cross-Sectional Area (US CSA) | <.001 | ||
| (−) | 70 (71%) | 52 (24%) | |
| (+) | 28 (29%) | 163 (76%) | |
| DML | <.001 | ||
| (−) | 70 (71%) | 66 (31%) | |
| (+) | 28 (29%) | 149 (69%) | |
| DSL | <.001 | ||
| (−) | 75 (77%) | 80 (37%) | |
| (+) | 23 (23%) | 135 (63%) | |
| SNAP | <.001 | ||
| (−) | 97 (99%) | 184 (86%) | |
| (+) | 1 (1.0%) | 31 (14%) | |
| CMAP | .021 | ||
| (−) | 66 (67%) | 115 (53%) | |
| (+) | 32 (33%) | 100 (47%) |
Bolded values indicate statistical significance (P < .05).
Patient-level descriptive characteristics
Table 2 presents patient-level demographics stratified by surgical recommendation category (none, unilateral, bilateral). Aside from ethnicity and BMI, demographic variables were similar across groups.
Table 2.
Patient-Level Demographics, Stratified by Surgical Recommendation
| Characteristic | No Surgical Recommendation (N = 53) | Unilateral Surgical Recommendation (N = 115) | Bilateral Surgical Recommendation (N = 50) | P |
|---|---|---|---|---|
| Age | 48.8 (16.1) | 51.2 (13.4) | 49.9 (11.8) | .617 |
| Sex | .791 | |||
| F | 38 (72%) | 88 (77%) | 37 (74%) | |
| M | 15 (28%) | 27 (23%) | 13 (26%) | |
| Ethnicity | .023 | |||
| Non-White | 20 (38%) | 21 (18%) | 14 (28%) | |
| White | 33 (62%) | 94 (82%) | 36 (72%) | |
| Diabetes | .841 | |||
| N | 41 (77%) | 91 (79%) | 41 (82%) | |
| Y | 12 (23%) | 24 (21%) | 9 (18%) | |
| BMI | 28.6 (6.0) | 32.3 (8.2) | 32.7 (6.5) | .008 |
Bolded values indicate statistical significance (P < .05).
Univariable associations between demographic factors, diagnostic test results, and surgical recommendation are summarized in Table 3. Positive results for all six diagnostic tests were associated with greater odds of surgical recommendation (all P < .05). Body mass index was also associated with greater odds of surgical recommendation (P < .001), with higher BMI observed among wrists recommended for surgery.
Table 3.
Univariable Associations Between Demographics, Preoperative Diagnostic Tests, and Surgical Recommendation
| Characteristic | OR | 95% CI | P |
|---|---|---|---|
| Preoperative diagnostic tests | |||
| CTS-6 | 4.87 | 2.84, 8.36 | <.001 |
| DML | 5.64 | 3.22, 9.90 | <.001 |
| CMAP | 1.79 | 1.06, 3.02 | .028 |
| DSL | 5.5 | 3.13, 9.69 | <.001 |
| SNAP | 16.3 | 2.22, 120 | .006 |
| CSA | 7.84 | 4.42, 13.9 | <.001 |
| Demographics | |||
| Age | 1.02 | 0.99, 1.04 | .175 |
| Sex (M) | 1.01 | 0.54, 1.90 | .978 |
| Ethnicity (White) | 1.22 | 0.67, 2.23 | .512 |
| Type 2 diabetes mellitus | 1.16 | 0.57, 2.34 | .686 |
| BMI | 1.07 | 1.04, 1.11 | <.001 |
Bolded values indicate statistical significance (P < .05)
Performance of multivariable models combining diagnostic tests is summarized in Table 4. The combination of CTS-6, DML, DSL, SNAP, and CSA demonstrated the highest discrimination (AUC 0.849). The combination of DML, SNAP, and CSA achieved the highest sensitivity (0.926) and NPV (0.778), while SNAP alone yielded the highest specificity (0.990) and PPV (0.969). Because positive SNAP findings were rare among wrists not recommended for surgery, performance estimates for SNAP-dependent models should be interpreted with caution.
Table 4.
Top-Performing Multivariable Diagnostic Test Combinations and Model Performance Metrics
| Combination | AUC | Sensitivity | Specificity | PPV | NPV |
|---|---|---|---|---|---|
| CTS-6, DML, DSL, SNAP, and CSA | 0.849 | 0.833 | 0.735 | 0.873 | 0.667 |
| DML, SNAP, and CSA | 0.814 | 0.926 | 0.571 | 0.826 | 0.778 |
| SNAP alone | 0.567 | 0.144 | 0.990 | 0.969 | 0.345 |
Bold indicates the highest performance metric across diagnostic test combinations.
To determine which variables most strongly contributed to multivariable model performance, ORs and 95% CIs were evaluated for each variable included in the top-performing combinations (Table 5). In the model with the highest AUC, positive CSA, CTS-6, and DML findings were each associated with higher odds of a surgeon recommending carpal tunnel release (CSA and CTS-6 P < .001; DML P = .002), with CSA demonstrating the strongest independent association. DSL and SNAP were not statistically significant in this context. In the model with the highest sensitivity and NPV, positive DML and CSA again were significantly associated with surgical recommendation (both P < .001), whereas SNAP was not significant. Because the model that maximized specificity and PPV consisted solely of SNAP, there were no additional variables to evaluate within that model. These findings identify the diagnostic tests that demonstrated the strongest independent associations within the best-performing multivariable models.
Table 5.
Independent Associations of Diagnostic Tests Within Top-Performing Multivariable Models
| Model Characteristic | OR | 95% CI | P |
|---|---|---|---|
| Max AUC | |||
| CTS-6 | 3.47 | 1.87, 6.44 | <.001 |
| DML | 2.94 | 1.49, 5.81 | .002 |
| DSL | 1.59 | 0.80, 3.16 | .185 |
| SNAP | 6.94 | 0.56, 85.9 | .131 |
| CSA | 4.87 | 2.56, 9.25 | <.001 |
| Max sensitivity and NPV | |||
| DML | 3.93 | 2.13, 7.23 | <.001 |
| SNAP | 8.86 | 0.88, 88.8 | .064 |
| CSA | 6.15 | 3.34, 11.3 | <.001 |
| Max specificity and PPV | |||
| SNAP | 16.3 | 2.22, 120 | .006 |
Bolded values indicate statistical significance (P < .05).
Performance of covariate-adjusted multivariable models is summarized in Table 6. When demographic covariates (age, sex, race, diabetes, and BMI) were added, the only meaningful improvement occurred in the model combining CTS 6, DML, DSL, SNAP, and CSA, where adding BMI increased the AUC from 0.849 to 0.854. Body mass index did not improve sensitivity, specificity, PPV, or NPV, and no other covariate enhanced the performance of any multivariable model. As noted above, SNAP-based model estimates should be interpreted with caution. Overall, while BMI modestly improved discrimination in the best-performing model, adding demographic covariates generally provided no meaningful gains in model performance.
Table 6.
Top-Performing Covariate-Adjusted Multivariable Diagnostic Models and Performance Metrics
| Combination | AUC | Sensitivity | Specificity | PPV | NPV |
|---|---|---|---|---|---|
| CTS-6, DML, DSL, SNAP, and CSA + BMI | 0.854 | 0.758 | 0.857 | 0.921 | 0.618 |
| DML, SNAP, and CSA | 0.814 | 0.926 | 0.571 | 0.826 | 0.778 |
| SNAP | 0.567 | 0.144 | 0.990 | 0.969 | 0.345 |
Bold indicates the highest performance metric across diagnostic test combinations.
To characterize independent associations within the covariate-adjusted model with the highest AUC, ORs and 95% CIs for each diagnostic test are reported in Table 7. In this model (AUC 0.854), positive CSA, CTS-6, and DML results demonstrated significant independent associations with increased odds of surgical recommendation (CSA and CTS-6 P < .001; DML P = .007), with CSA again showing the strongest association. DSL and SNAP were not statistically significant contributors, and BMI showed a borderline association that did not reach statistical significance.
Table 7.
Independent Associations Within the Covariate-Adjusted Model With the Highest AUC
| Model Characteristic | OR | 95% CI | P |
|---|---|---|---|
| Max AUC | |||
| CTS-6 | 3.71 | 1.98, 6.97 | <.001 |
| DML | 2.69 | 1.32, 5.50 | .007 |
| DSL | 1.54 | 0.76, 3.11 | .232 |
| SNAP | 6.96 | 0.57, 85.0 | .129 |
| CSA | 4.52 | 2.37, 8.61 | <.001 |
| BMI | 1.04 | 1.00, 1.09 | .065 |
Bolded values indicate statistical significance (P < .05).
Discussion
Purpose and overview
The goal of this study was to determine which preoperative diagnostic tests, individually and in combination, were most strongly associated with a surgeon’s recommendation for carpal tunnel release. We examined clinical (CTS-6), imaging (CSA), and electrodiagnostic parameters (DML, DSL, CMAP, SNAP) to evaluate how findings from different diagnostic domains influence surgical decision-making. On univariable analysis, positive results across all six diagnostic tests were associated with greater odds of surgical recommendation, and higher BMI demonstrated a similar association, with the surgical cohort averaging BMI values in the obese range compared with overweight values among nonsurgical wrists. These findings provided the basis for subsequent multivariable modeling to evaluate the independent contribution of each measure.
The relative influence of clinical, imaging, and electrodiagnostic findings on surgical decision-making remains poorly defined. Identifying which diagnostic measures align most closely with a surgeon’s recommendation clarifies how diagnostic data are interpreted in practice and may help make evaluation more consistent and efficient. Understanding these relationships also outlines the diagnostic hierarchy surgeons apply when determining surgical candidacy.
Interpretation of multivariable findings
In both the unadjusted and covariate-adjusted multivariable models that produced the highest AUC (“Max-AUC” model), CTS-6, DML, and CSA demonstrated strong independent associations with surgical recommendation, whereas DSL and SNAP did not. Among these variables, CSA showed the strongest independent association, followed by CTS-6 and DML. Although their effect sizes were generally similar, the relative ordering of their contributions remained consistent after covariate adjustment. Adding demographic covariates modestly improved overall model performance (ΔAUC = 0.005), but the same three diagnostic variables continued to drive prediction, suggesting that operative recommendations rely primarily on anatomic, clinical, and physiologic findings, with relatively limited influence from demographic factors.
Role of ultrasound CSA
Ultrasound measurement of the median nerve CSA provides a quick, noninvasive assessment of anatomic nerve enlargement at the carpal tunnel. In our highest-performing models, CSA consistently showed a strong independent association with surgical recommendation. In office use of ultrasound by a skilled provider allows for real time measurement of the median nerve, allowing a provider to make clinical diagnosis of CTS. It also reflects a surgeon’s trust in this tool as a reliable diagnostic modality. Prior ` have shown that CSA is among the most reliable sonographic parameters for identifying and grading median nerve pathology, with larger CSA values linked to more severe clinical presentations of CTS.8,9 At the same time, ultrasound findings are inherently operator- and technique-dependent; image acquisition, probe positioning, and the experience of the musculoskeletal sonographer can influence measurement reliability. In centers with experienced musculoskeletal sonographers, reliability is typically high, but variability across settings may limit generalizability. Ultrasound also enables visualization of anatomic contributors such as transverse carpal ligament thickening, ganglion cysts, or vascular and nerve variants, supporting its role as a valuable adjunct in surgical decision-making.8
Role of CTS-6
The CTS-6 score integrates symptom patterns and physical examination findings into a single clinical construct reflecting overall disease severity.3 CTS-6 was consistently associated with surgical recommendation, underscoring the continued importance of clinical assessment alongside objective testing. Because it incorporates patient-reported symptoms such as nocturnal numbness and examiner findings such as positive provocative tests, CTS-6 captures aspects of functional impairment and symptom burden that electrophysiologic and imaging measures cannot. When interpreted with CSA and NCS parameters, CTS-6 provides essential clinical context that complements anatomic and physiologic data in guiding surgical decision-making.
Role of NCS
NCS quantify median nerve function at the wrist and provide objective physiologic evidence of neuropathy. In our analysis, prolonged DML was consistently associated with surgical recommendation, emphasizing the weight surgeons place on motor conduction delay as a potentially reversible marker of demyelination. Prolonged DML and DSL indicate slowed conduction in intact fibers that may recover following decompression. In contrast, reduced CMAP and SNAP amplitudes reflect axonal injury, which is commonly interpreted as a marker of advanced disease that may prompt surgical consideration, although recovery may be incomplete.10, 11, 12 NCS results primarily measure physiologic dysfunction rather than anatomic compression, and up to 25% of patients with clinically evident CTS demonstrate normal NCS findings.13 Abnormal results may also arise from other neuropathies, such as diabetic neuropathy. Because NCS findings can be influenced by disease chronicity, comorbid neuropathies, and technical factors, they may be most useful when integrated with complementary clinical and imaging assessments.
SNAP interpretation and clinical nuance
Although SNAP was highly specific and demonstrated strong univariable associations with surgical recommendation, these results were driven by the extreme distribution of SNAP findings in the cohort. Among wrists not recommended for surgery, 1 of 98 (1%) was SNAP-positive, while 31 of 215 (14.4%) of surgical wrists were SNAP-positive (<6 μV), a considerable difference between groups. This imbalance limits interpretability but provides insight into how surgeons incorporate sensory testing within a broader diagnostic framework. In this cohort, abnormal SNAPs reinforced the decision to operate but were not required for surgical recommendation.
Clinical implications
These findings show that the core diagnostic modalities for CTS, including clinical assessment, electrodiagnostic testing, and ultrasound, each provide distinct yet complementary information that contributes to surgical decision-making. CTS-6 reflects symptom severity and examination findings, DML quantifies physiologic conduction delay, and CSA identifies structural nerve enlargement. Together, these measures capture the multidimensional process by which surgeons assess disease burden and determine surgical candidacy. The consistent associations of these three parameters across models highlight their collective importance as the foundation of a comprehensive diagnostic approach for CTS.
Limitations
This study has several limitations. Its retrospective design reflects real-world clinical practice but limits the ability to control for all potential confounders or to determine the specific rationale behind each surgeon’s decision for surgical recommendation. Although covariate-adjusted models accounted for demographic variables, other relevant factors, including symptom duration, occupational demands, and prior treatments, were not analyzed. The use of surgical recommendation as the primary outcome also introduces subjectivity, as decisions are influenced by surgeon experience, patient preferences, and institutional norms. Patterns of recommendation may therefore differ across surgeons and practice settings, and the single tertiary academic center design further limits generalizability.
In addition, inclusion required complete preoperative data for CTS-6, ultrasound CSA, and at least one NCS parameter, which likely enriched the cohort for patients with intermediate diagnostic certainty. Individuals with clearly positive or clearly negative clinical presentations may have proceeded directly to surgery or foregone further testing, respectively, limiting generalizability to the broader CTS population. And for situations where all variables are not available (eg, a clinic without ultrasound) our findings may change with regard to which variables are most relied on for surgical decision-making. Furthermore, because this retrospective design captured only the final documented treatment recommendation, it did not allow assessment of how diagnostic results changed a surgeon’s decision. Future prospective studies should evaluate pretest–posttest decision changes to better quantify how diagnostic information influences surgical decision-making.
Beyond this, the imbalance in SNAP findings (particularly in the nonsurgical group) represents an interpretive limitation that may influence the precision of SNAP-related metrics. Finally, while this study identifies diagnostic parameters most associated with surgical recommendation, it does not establish causality or determine which tests should guide surgical decision-making. Prospective, multicenter validation incorporating postoperative outcomes is needed to confirm the generalizability of these findings.
Conflicts of Interest
Given his role as Editor-in-Chief of The Journal of Hand Surgery Global Online, Dr Fowler had no involvement in the peer review of this article and has no access to information regarding its peer review. Full responsibility for the editorial process for this article was delegated to Aviram M. Giladi, MD, MS. Dr Fowler reports consulting fees from Integra LifeSciences, outside the submitted work. No benefits in any form have been received or will be received by the other authors related directly to this article.
References
- 1.Atroshi I., Gummesson C., Johnsson R., Ornstein E. Diagnostic properties of nerve conduction tests in population-based carpal tunnel syndrome. BMC Musculoskelet Disord. 2003;4:9. doi: 10.1186/1471-2474-4-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Sonoo M., Menkes D.L., Bland J.D.P., Burke D. Nerve conduction studies and EMG in carpal tunnel syndrome: do they add value? Clin Neurophysiol Pract. 2018;3:78–88. doi: 10.1016/j.cnp.2018.02.005. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Graham B., Regehr G., Naglie G., Wright J.G. Development and validation of diagnostic criteria for carpal tunnel syndrome. J Hand Surg Am. 2006;31(6):919–924. [PubMed] [Google Scholar]
- 4.Demino C., Fowler J.R. The sensitivity and specificity of nerve conduction studies for diagnosis of carpal tunnel syndrome: a systematic review. Hand (N Y) 2021;16(2):174–178. doi: 10.1177/1558944719855442. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Fowler J.R., Byrne K., Pan T., Goitz R.J. False-positive rates for nerve conduction studies and ultrasound in patients without clinical signs and symptoms of carpal tunnel syndrome. J Hand Surg Am. 2019;44(3):181–185. doi: 10.1016/j.jhsa.2018.11.010. [DOI] [PubMed] [Google Scholar]
- 6.Graham B., Peljovich A.E., Afra R., et al. The American Academy of Orthopaedic Surgeons Evidence-Based Clinical Practice Guideline on: management of carpal tunnel syndrome. J Bone Joint Surg Am. 2016;98(20):1750–1754. doi: 10.2106/JBJS.16.00719. [DOI] [PubMed] [Google Scholar]
- 7.Zhuang T., Shapiro L.M., Schultz E.A., Truong N.M., Harris A.H.S., Kamal R.N. Has the use of electrodiagnostic studies for carpal tunnel syndrome changed after the 2016 American Academy of Orthopaedic Surgeons Clinical Practice Guideline? J Hand Surg Am. 2023;48(1):19–27. doi: 10.1016/j.jhsa.2022.09.019. [DOI] [PubMed] [Google Scholar]
- 8.Ooi C.C., Wong S.K., Tan A.B., et al. Diagnostic criteria of carpal tunnel syndrome using high-resolution ultrasonography: correlation with nerve conduction studies. Skeletal Radiol. 2014;43(10):1387–1394. doi: 10.1007/s00256-014-1929-z. [DOI] [PubMed] [Google Scholar]
- 9.Takata S.C., Kysh L., Mack W.J., Roll S.C. Sonographic reference values of median nerve cross-sectional area: a protocol for a systematic review and meta-analysis. Syst Rev. 2019;8(1):2. doi: 10.1186/s13643-018-0929-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Cartwright M.S., Walker F.O. Neuromuscular ultrasound in common entrapment neuropathies. Muscle Nerve. 2013;48(5):696–704. doi: 10.1002/mus.23900. [DOI] [PubMed] [Google Scholar]
- 11.Johnson E.W. Should immediate surgery be done for carpal tunnel syndrome?—no. Muscle Nerve. 1995;18(6):658–659. doi: 10.1002/mus.880180613. ; discussion 663. [DOI] [PubMed] [Google Scholar]
- 12.Osiak K., Mazurek A., Pękala P., Koziej M., Walocha J.A., Pasternak A. Electrodiagnostic studies in the surgical treatment of carpal tunnel syndrome-a systematic review. J Clin Med. 2021;10(12):2691. doi: 10.3390/jcm10122691. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Caetano M.R. Axonal degeneration in association with carpal tunnel syndrome. Arq Neuropsiquiatr. 2003;61(1):48–50. doi: 10.1590/s0004-282x2003000100008. [DOI] [PubMed] [Google Scholar]