If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Corresponding author: Kevin C. Chung, MD, MS, Comprehensive Hand Center, Michigan Medicine, 2130 Taubman Center, SPC 5340, 1500 E. Medical Center Dr., Ann Arbor, MI 48109-5340.
To evaluate factors that predict the use of electrodiagnostic testing (EDS) for patients undergoing carpal tunnel release (CTR).
Methods
In this cohort study, we analyzed 553 patients who underwent primary CTR from 8 practices between July 1, 2019 and December 1, 2019 by 32 surgeons in the Michigan Collaborative Hand Initiative for Quality in Surgery (M-CHIQS). The M-CHIQS is a collaborative initiative aimed at improving quality in hand surgery. Demographic and clinical characteristics were collected, including the 6-item carpal tunnel symptoms scale (CTS-6) scores and EDS timing. Multilevel logistic regression was used to assess practice and surgeon variation in EDS use related to clinical diagnostic criteria.
Results
Of the 553 patients who underwent CTR during the study period, 461 (83.3%) received preoperative EDS. After controlling for patient clinical and demographic characteristics, CTS-6 scores were not associated with receiving any preoperative EDS (lower probability of CTS: odds ratio [OR], 0.94; 95% confidence interval [95% CI], 0.59–1.51), preconsultation EDS (low probability of CTS: OR, 1.00; 95% CI, 0.73–1.38), or postconsultation EDS (low probability of CTS, OR, 1.10; 95% CI, 0.77–1.60). For use of any EDS, 9.3% of the variation in testing was explained at the practice level and 31.1% of the variation in testing was explained at the surgeon level.
Conclusions
Variation in EDS use is explained primarily at the practice and surgeon levels and is not related to patient clinical criteria. We recommend that providers and practices assess their use of preoperative EDS and limit its use to patients with an unclear clinical CTS diagnosis, as stated in current clinical practice guidelines. Likewise, providers should be encouraged to use the CTS-6 before prescribing EDS.
Clinical relevance
Limiting the use of EDS to patients with an unclear clinical diagnosis of CTS will reduce costs and improve patient care by eliminating the discomfort and time associated with this test.
Dr Graham interviews Dr. Jessica Billig regarding her article "Variation in Use of Electrodiagnostic Testing: Analysis From the Michigan Collaborative Hand Initiative for Quality in Surgery”, the lead article in the March 2021 issue of the Journal of Hand Surgery.
Despite the high prevalence of CTR, there is no universally accepted reference standard for the diagnosis of carpal tunnel syndrome (CTS). Electrodiagnostic testing (EDS) has the lowest specificity (highest false-positive rate) compared with other diagnostic modalities, such as ultrasound and a validated clinical evaluation of carpal tunnel syndrome (CTS-6), and is user-dependent in examination findings, varying based on the ultrasonographer or electrodiagnostician.
In 2009, the American Academy of Orthopaedic Surgeons (AAOS) released their Evidence-Based Clinical Practice Guideline for CTS recommending the use of EDS if the clinical examination is positive and surgical management is being considered.
In 2016, the AAOS released new guidelines stating, “Electrodiagnostic testing may be of most value when the clinical diagnosis is unclear or when atypical features exist.”
A survey of American Society for Surgery of the Hand (ASSH) members revealed that the majority (57%) of respondents are more likely to order EDS because of medical-legal ramifications of the AAOS guidelines, despite feeling that a supporting history and examination, without EDS or cortisone injection, were sufficient to recommend surgery.
it appears that physicians may use it as a common practice rather than on an as-needed basis.
The purpose of this study was to evaluate the use of EDS for patients undergoing CTR using data from a large hand surgery quality collaborative initiative. Our primary outcome was EDS use and timing (ordered pre- or postconsultation with a hand surgeon). We hypothesize that EDS use is not associated with patient clinical characteristics, but is associated with variation at the surgeon and practice levels.
Methods
Data source
The Michigan Collaborative Hand Initiative for Quality in Surgery (M-CHIQS) is a collaborative quality initiative aimed at understanding and implementing measures to improve quality in hand surgery. Similar to other health care collaboratives, the long-term goal of M-CHIQS is to improve the quality of hand surgery at a national level by guiding quality improvement interventions such as changes in clinical practice guidelines and policy. The M-CHIQS consists of 9 practices performing hand surgery within the United States. The M-CHIQS includes both community practices and academic medical centers, in a conscious effort to collect practice patterns and patient data across different settings. The details of the formation of M-CHIQS have been previously reported.
In short, all M-CHIQS sites engage in data collection and feedback, using a universal and common quality-reporting infrastructure, which is housed at the reporting center. Data of interest for M-CHIQS are collected prospectively as part of standard clinical care and extracted from the electronic medical records at each site 6 to 8 weeks after surgery. Study data were collected and managed using the REDCap (Research Electronic Data Capture) tool hosted at the reporting center.
The REDCap is a secure, Web-based software platform that provides an interface for validated data capture, audit trails for tracking data organization, procedures for data integration, and automated export to common statistical software.
Regular data audits were performed to ensure validity and to alert the sites to any missing data, which they were asked to complete. We have no missing data in this study. The institutional review board at every participating site approved the initiative and permitted the collection and use of limited data sets derived from the M-CHIQS database.
Study cohort
We included patients who underwent elective primary CTR at 8 practices (7 academic and 1 private) between July 1, 2019 and December 1, 2019 by 32 hand surgeons. All surgeons within the collaborative completed a hand surgery fellowship and have an active subspecialty certificate in surgery of the hand. Board certification of the 32 participating surgeons included 22 in orthopedic surgery and 10 in plastic surgery. The ninth site did not collect data within our designated study cohort time period and was excluded from the analysis. The sites that collected data for this study included The University of Michigan, Curtis National Hand Center, Emory Healthcare, Indiana University Health, OrthoCarolina, University of Pittsburgh Medical Center, University of Rochester Medical Center, and Wake Forest Baptist Health. All these sites participate in the training of residents and fellows and have substantial research involvement. Patients were included if they were age 18 years or older and underwent open CTR or endoscopic CTR. Patients were excluded if they had undergone a previous CTR of the affected hand or if the CTR was not elective (ie, performed for trauma-related procedures). Trauma-related procedures were entered in our database as history of wrist fracture/trauma along with the date of trauma, and all patients with a history of wrist fracture were excluded. For the final model, we excluded all patients with ulnar neuropathy because patients with ulnar neuropathy will be more likely to have EDS to confirm diagnosis.
Primary outcome
The primary outcome for this study was use of EDS. We further categorized EDS timing in relation to the hand surgeon evaluation: preconsultation and postconsultation. Preconsultation testing occurred before the hand surgeon evaluation, and postconsultation testing occurred between the hand surgeon evaluation and the CTR. Preconsultation testing may have been ordered by a referring provider or by the hand surgeon (we did not collect data regarding who ordered preconsultation testing). Postconsultation testing was ordered by the hand surgeon.
Patient-level variables
Patient-level (considering each of the 553 patients as individual units) variables of interest included age, sex, race, and insurance type. Race was categorized as white non-Hispanic, black non-Hispanic, Asian, Hispanic, and other. For insurance type, we included private/employer-sponsored health insurance, Medicare, Medicaid, workers’ compensation, and other. The CTS-6 scores (Table 1)
were calculated by the authors using symptom criteria entered for each patient for each affected hand and categorized as higher probability (CTS-6 score ≥ 12, 80% probability) or lower probability (CTS-6 score < 12). We also collected data regarding the presence of comorbidities associated with CTS including diabetes, hypothyroidism, and rheumatoid arthritis.
To determine variation at the practice level (considering each of the 8 sites as individual units) and surgeon level (considering each of the 32 surgeons as individual units), we calculated the intraclass correlation coefficient and the median odds ratio (OR).
A brief conceptual tutorial of multilevel analysis in social epidemiology: using measures of clustering in multilevel logistic regression to investigate contextual phenomena.
We conducted descriptive analyses of patient demographic and clinical characteristics stratified by practice. We then calculated unadjusted associations using chi-square test and Fisher exact test for categorical variables and Student 2-tailed t test for continuous variables.
We used 3 different models to examine the association of patient-level characteristics and variation at the practice and surgeon levels with EDS use. Given the clustering of patients within practices and surgeons, a multilevel logistic regression model was used with random intercepts at the practice and surgeon levels. We expect that surgeons treat their patients similarly, thus patients treated by the same surgeon would be more related to each other than patients treated by other surgeons. We also would expect that surgeons within the same group would be more likely to have similar practice patterns than surgeons from other groups. Therefore, to account for this within-surgeon and within-practice similarity, we placed random intercepts at the practice and surgeon levels. We used fixed slopes in the model because we did not find any heterogeneity within how each covariate affected the surgeon or practice level. To get unbiased estimates using multilevel modeling, it is recommended that the sample contain at least 50 observations.
Covariates included age as a quadratic, sex, race, insurance type, CTS-6 score, and presence of diabetes. Postestimation margins were used to obtain the probability of EDS use. A significance level of P less than .05 was used for all analyses.
Results
A total of 655 patients underwent CTR during the study period. Within this cohort, 102 patients had concomitant ulnar neuropathy and were excluded, resulting in a final cohort of 553 patients. Participating surgeons within each practice ranged from 1 to 11 surgeons. The distribution of patients within each practice is found in Table 2.
Of the 553 patients who underwent CTR, 461 (83.3%) had preoperative EDS. Of these patients, 298 (64.6%) received preconsultation EDS, 147 (31.9%) received postconsultation EDS, and 16 (3.5%) received repeat testing, both preconsultation and postconsultation EDS. When comparing patients with preoperative EDS with those without preoperative EDS, we found no significant differences in age, sex, race, insurance type, presence of diabetes, hypothyroidism, or rheumatoid arthritis (Table 3). The overall distribution of CTS-6 scores in this cohort is shown in Figure 1. The mean CTS-6 score of patients undergoing preoperative EDS was 13.1 (SD, 5.2) compared with 13.4 (SD, 6.2) in patients who did not undergo preoperative EDS (P = .53). Figure 2 describes the distribution of CTS-6 scores stratified by EDS timing: preconsultation and postconsultation.
Table 3Patient Characteristics Stratified by Receipt of EDS (n = 553)
Chi-square test used for categorical variables and Student 2-tailed t test used for continuous variables. An exception was made for rheumatoid arthritis in which Fisher exact test was used because of the small numbers.
Mean age (SD)
57.8 (14.1)
57.0 (15.8)
58.0 (13.7)
.56
Sex
.09
Male
185 (33.5)
32 (34.8)
153 (33.2)
Female
368 (66.6)
60 (65.2)
308 (66.8)
Race
.77
White non-Hispanic
407 (73.6)
71 (77.2)
336 (72.9)
Black non-Hispanic
101 (18.3)
14 (15.2)
87 (18.9)
Asian
7 (1.3)
2 (2.2)
5 (1.1)
Hispanic
12 (2.2)
2 (2.2)
10 (2.2)
American Indian/Pacific Islander
4 (0.7)
1 (1.1)
3 (0.7)
Other
22 (4.0)
2 (2.2)
20 (4.3)
Insurance type
.70
Private/employer Sponsored
257 (46.5)
47 (51.1)
210 (45.6)
Medicare
168 (30.4)
29 (31.5)
139 (30.2)
Medicaid
83 (15.0)
11 (12.0)
72 (15.6)
Workers’ compensation
27 (4.9)
3 (3.3)
24 (5.2)
Self-pay
7 (1.3)
0 (0)
7 (1.5)
Other
2 (0.4)
0 (0)
2 (0.4)
Unknown
9 (1.6)
2 (2.2)
7 (1.5)
Mean CTS-6 score (SD)
13.2 (5.4)
13.4 (6.2)
13.1 (5.2)
.53
Diabetes
118 (21.3)
15 (16.3)
103 (22.3)
.20
Hypothyroid
71 (12.8)
13 (14.1)
58 (12.6)
.69
Rheumatoid arthritis
28 (5.1)
6 (6.5)
22 (4.8)
.44
∗ Patients with concomitant ulnar neuropathy were removed.
† Chi-square test used for categorical variables and Student 2-tailed t test used for continuous variables. An exception was made for rheumatoid arthritis in which Fisher exact test was used because of the small numbers.
Figure 3 illustrates the frequency of EDS use at the practice level, stratified by timing. Substantial variation in the use of EDS was seen at the practice level, ranging from 66.7% to 98.3% of patients in a specific practice receiving any EDS. More specifically, the timing of EDS use was substantially different among the practices. For preconsultation EDS stratified by practice, the range of patients receiving testing was 28.1% to 71.9%.
Figure 3Substantial variation in the use of any EDS is seen at the practice level. In addition, the timing (preconsultation versus postconsultation) of EDS was substantially different among the practices. The site numbering corresponds to the same sites in Table 2.
After controlling for patient, surgeon, and practice characteristics, women were more likely to receive preconsultation EDS (OR, 1.50; 95% confidence interval [95% CI], 1.10–2.05) (Table 4). However, no other patient characteristics were significantly associated with receiving preconsultation EDS. Patients with low probability of CTS by CTS-6 scores were not associated with receiving preconsultation EDS (OR, 1.00; 95% CI, 0.73–1.38). For preconsultation EDS, 7.5% of the variation was explained at the practice level with a median OR of 1.65. In addition, 9.4% of the variation was explained at the surgeon level (median OR, 1.28). For postconsultation EDS, women had decreased odds of receiving testing (OR, 0.60; 95% CI, 0.42–0.87). No other patient characteristic was significantly associated with receiving postconsultation EDS. Patients with a lower probability of CTS, as determined by CTS-6 scores, were not associated with receiving postconsultation EDS (OR, 1.10; 95% CI, 0.77–1.60) (Table 4). Approximately 3.7% of the variation in receiving postconsultation EDS was at the practice level (median OR, 1.45), and 20.8% of the variation was explained at the surgeon level (median OR, 2.24).
Table 4Multilevel Logistic Regression for Predictors for Use of EDS
Covariates in the model included age as a quadratic, race, sex, insurance, CTS-6 score, and diabetes with intercepts at the practice and surgeon levels.
Covariate
Preconsultation EDS
Postconsultation EDS
Any EDS
OR (95% CI)
P Value
OR (95% CI)
P Value
OR (95% CI)
P Value
Sex
Male
1 (Reference)
-
1 (Reference)
-
1 (Reference)
-
Female
1.50 (1.10–2.05)
<.05
0.60 (0.42–0.87)
<.05
1.09 (0.70–1.71)
.70
Diabetes
No
1 (Reference)
-
1 (Reference)
-
1 (Reference)
-
Yes
1.31 (0.90–1.91)
.15
0.93 (0.62–1.42)
.76
1.72 (0.95–3.13)
.07
CTS-6 score
Lower probability
1.00 (0.73–1.38)
>.999
1.10 (0.77–1.60)
.59
0.94 (0.59–1.51)
.82
Higher probability
1 (Reference)
-
1 (Reference)
-
1 (Reference)
-
Practice level
ICC
7.5
-
3.7
9.3
-
Median OR
1.65
-
1.45
1.89
-
Surgeon level
ICC
9.4
-
20.8
-
31.1
-
Median OR
1.28
-
2.24
-
2.64
-
ICC, intraclass correlation coefficient.
∗ Covariates in the model included age as a quadratic, race, sex, insurance, CTS-6 score, and diabetes with intercepts at the practice and surgeon levels.
Lastly, using multivariable modeling for receiving any EDS, lower probability CTS-6 scores were not associated with EDS use (OR, 0.94; 95% CI, 0.82–1.51) (Table 4). At the practice level, 9.3% of the variation was explained (median OR, 1.89). The variation at the surgeon level was 31.1% (median OR, 2.64). The predicted probability of receiving any EDS over the range of CTS-6 scores stratified by sex is illustrated in Figure 4. Overall, CTS-6 scores did not significantly contribute to the predicted probability of EDS use, and there were no significant differences between males and females.
Figure 4CTS-6 scores did not significantly contribute to the predicted probability of receiving EDS, and there were no significant differences between males and females. Predicted probability was obtained by postestimation margins command.
In this national quality improvement study, we found that the majority of patients receive EDS prior to undergoing CTR. However, patient clinical criteria, including CTS-6 scores, were not associated with receiving EDS for the diagnosis of CTS. Moreover, the variation in EDS use was explained primarily at the surgeon level. These findings underscore the multifactorial influences in the use of EDS for CTS diagnosis and highlight that physicians are not adhering to current clinical practice guidelines that indicate EDS is most useful for patients with an unclear clinical diagnosis of CTS.
Our findings that the majority of patients received preoperative EDS are consistent with previous research. Approximately 90% of ASSH members routinely use EDS.
evaluated whether EDS changed the ultimate treatment plan for CTS and found that only 19% of patients had a shift in recommendations from surgical management to nonsurgical treatment because of EDS results. However, debate continues as to whether all patients with a possible diagnosis of CTS should undergo EDS. In another ASSH survey with clinical scenarios, 47% of members responded that EDS was not necessary for patients with supporting clinical history and physical examination and symptomatic relief with corticosteroid injections.
Despite the lack of consensus for EDS use, some insurers, specifically for workers’ compensation, may require EDS prior to CTR. This is understandable given the high cost of CTS cases including days away from work compared with other musculoskeletal disorders.
The reasons why physicians continue to routinely order EDS are unknown, but it has been speculated that it has become “entrenched in the diagnostic algorithm” and also may serve as a financial incentive for many providers.
However, the lack of consensus regarding any universal reference standard for diagnosis of CTS limits the ability to compare diagnostic tests. In fact, the AAOS acknowledges that a reference standard is the “most important research goal in this area.”
found that the addition of EDS to history and physical examination did not change the probability of a CTS diagnosis, arguing that EDS should be used in situations in which the clinical diagnosis of CTS is uncertain. Multiple other studies have shown that the sensitivity and specificity of EDS is substantially lower than clinical diagnosis, questioning the application of EDS in all patients with CTS.
Comparison of ultrasound and electrodiagnostic testing for diagnosis of carpal tunnel syndrome: study using a validated clinical tool as the reference standard.
More specifically, in patients with CTS-6 scores greater than or equal to 12, where the probability of CTS is over 80%, there may not be a clinical need to obtain additional testing. With a pretest probability of over 80% from clinical examination, the addition of confirmatory diagnostic tests including EDS provides little guidance in treatment. Even when a diagnosis is not in doubt, some providers may use EDS for prognostication when counseling patients about symptom resolution and recovery after CTR, for a comparison in the event that symptoms continue after CTR, owing to insurance requirements, or for counseling patients who are interested in nonsurgical options. However, research has shown that there is little evidence to support that a positive EDS or EDS severity are predictors of symptom resolution after CTR.
In this study, we found that clinical criteria including CTS-6 scores were not associated with receiving EDS. Thus, it appears that providers are not using pretest probabilities to determine whether EDS will adjust the ultimate treatment, but rather, they may routinely order this test regardless of patient symptoms and physical examination findings. The wide variation (28%–72%) of patients who received preconsultation EDS in this study seems to support the notion that certain providers order EDS as common practice. Alternatively, providers may assess clinical symptoms but not calculate a CTS-6 score or follow the 80% guideline. The reasons behind EDS use must be further explored to fully understand why providers are ordering these tests in the majority of patients. As discussed previously, possible reasons include insurance requirements, a financial incentive for providers, an unclear diagnosis, to improve provider efficiency, or owing to a common practice at the surgeon or practice level. Moreover, the variation in EDS use was primarily at the surgeon level, revealing that the choice for EDS may be provider-driven. However, a large proportion of variance remains unexplained by factors that are not at the practice or surgeon level. Our findings highlight the importance of providers assessing their use of EDS, specifically limiting additional diagnostic tests in instances in which the pretest probability of CTS using CTS-6 is greater than 80% and EDS is not required by the insurance carrier. Moreover, all providers who diagnose or treat CTS, including nonsurgeons, should be encouraged to use the CTS-6 and not EDS as a screening tool prior to consultation with a hand surgeon.
Controversy exists regarding the optimal timing of EDS: before or after hand surgical consultation. However, there continues to be overuse in diagnostic testing, specifically prior to consultation. In a study by Hartzell et al,
approximately 74% of patients presenting for hand surgical evaluation underwent preconsultation testing, in which 70% of those tests were considered unnecessary. In our study, of the patients who received preoperative EDS, 64.6% received testing prior to hand surgery consultation and 3.5% underwent repeat testing. The purpose of repeat testing in this study is unknown, but its elimination appears to be an easy way to reduce wasteful spending. The reasons behind receiving preconsultation EDS are multifactorial including hand surgical practice requirements, primary care testing and referral patterns, or patient preference. In a study by Colombo and Shah,
approximately 67% of patients referred to hand surgeons by general practitioners had received EDS prior to hand surgery consultation, obtained at the discretion of the general practitioners. For surgeons, preconsultation EDS may improve clinical efficiency by expediting diagnosis during initial hand surgical evaluation. Sears et al
conducted a secret shopper study of hand surgeons in the state of Michigan and found that 57% required diagnostic testing prior to consultation. However, preconsultation EDS is associated with delays in time to hand surgeon evaluation and ultimate treatment.
The efficiency that may be afforded to the surgeon with preconsultation EDS is associated with potentially unnecessary inefficiencies to patients, including pain and losses of time and cost. Therefore, understanding when EDS is beneficial, specifically preconsultation EDS, is imperative to improving the quality of hand surgical care.
Our study has some limitations. First, this study is subject to selection bias. The M-CHIQS includes surgeons who voluntarily decided to join a quality collaborative, which may affect specific practice patterns. Moreover, the results may not be generalizable to all hand surgeons and all practices. The M-CHIQS includes surgeons from 7 academic centers and 1 large private practice, all participating in the education of trainees. These surgeons may have different practice patterns that are not generalizable to other hand surgeons. Last, our cohort was limited to surgical patients. Thus, it is possible that EDS was used as a diagnostic tool for patients who were not surgical candidates. However, our results show that the majority of EDS were obtained before consultation with the hand surgeon, meaning that the test may have been used as a screening tool rather than for clinical confirmation of CTS diagnosis. In addition, we cannot determine the reasons behind EDS use and whether they are patient-, practice-, or surgeon-driven. Although CTS-6 scores were calculated by the authors using symptoms provided for each patient, it is unknown whether participating surgeons calculated the CTS-6 score to determine diagnosis. Thus, we cannot determine whether EDS was ordered to confirm diagnosis. We did not collect information on referral patterns and cannot determine which provider prescribed preconsultation EDS. However, we found substantial variation at the practice and surgeon levels with no association with patient clinical criteria, indicating that EDS use is more likely to be practice- and provider-related. The data presented here are from a cohort study and not subject to randomization. Therefore, there may be unmeasured confounders that contribute to the total variation in EDS use. However, the data from M-CHIQS contain rich clinical data from electronic medical records to obtain patient-level characteristics, thus minimizing confounding. In addition, we did not have missing data in this study, eliminating any biases caused by missingness.
Despite these limitations, this study highlights the variation in EDS use for the diagnosis of CTS. Patient clinical criteria were not predictive of EDS use, and there was substantial variation in EDS use explained at the practice and surgeon levels. Collectively, these findings suggest that practices and providers need to assess whether their EDS use follows clinical practice guidelines in order to minimize overuse and improve the quality of hand surgical care.
Acknowledgments
Research reported in this publication was supported by funds from the American Foundation for Surgery of the Hand. K.C.C. receives research funding from the National Institutes of Health, book royalties from Wolters Kluwer and Elsevier, and funding from Axogen. The funding organizations had no role in the design and conduct of the study, including collection, management, analysis, and interpretation of the data. The content is solely the responsibility of the authors and does not necessarily represent the official views of the U.S. government or Veterans Administration.
The M-CHIQS Collaborators are Joshua Adkinson, MD; John Fowler, MD; R. Glenn Gaston, MD; Aviram Giladi, MD, MS; Michael Gottschalk, MD; Warren Hammert, DDS, MD; Ryan Katz, MD; Zhongyu John Li, MD, PhD; Marco Rizzo, MD; and Eric Wagner MD, MS.
We want to thank the M-CHIQS study coordinators Natalie Baxter (University of Michigan, Ann Arbor, MI); Christian Cuevas (Indiana University Health, Indianapolis, IN); Karen Wasil and Brooke Marshall (University of Pittsburgh, Pittsburgh, PA); Danielle Drossart and Carter Gunn (OrthoCarolina Hand Center, Charlotte, NC); Ike Fleming, Imran Yousaf, Kezia Alexander, Pragna Shetty, Lumanti Manandhar, Flossine Brown, Alessandra Butanis (Curtis National Hand Center, Baltimore, MD); Alexander Dawes (Emory Orthopaedics, Sports & Spine, Atlanta, GA); Raechel Argento (University of Rochester, Rochester, NY); Rachel Bordelon (Wake Forest Baptist Health, Winston-Salem, NC); and Taylor Trentadue (Mayo Clinic, Rochester, MN).
References
Jain N.B.
Higgins L.D.
Losina E.
Collins J.
Blazar P.E.
Katz J.N.
Epidemiology of musculoskeletal upper extremity ambulatory surgery in the United States.
A brief conceptual tutorial of multilevel analysis in social epidemiology: using measures of clustering in multilevel logistic regression to investigate contextual phenomena.
Comparison of ultrasound and electrodiagnostic testing for diagnosis of carpal tunnel syndrome: study using a validated clinical tool as the reference standard.
30711-5&id=gr1.jpg){kind=link}
30711-5&id=gr2.jpg){kind=link}
30711-5&id=gr3.jpg){kind=link}
30711-5&id=gr4.jpg){kind=link}