Background and rationale {6a}

The carpal tunnel is an anatomical space located at the base of the palm of the hand, delimited by a rigid floor and walls, the carpal bones, and a roof with a certain capacity for distension, the transverse carpal ligament or flexor retinaculum. The entrance of the tunnel is located between the scaphoid and pyramidal bones, and its exit between the trapezium and hamate bones. A number of anatomical structures pass through this relatively narrow space as it transitions from the forearm to the hand: the tendons of the superficial and deep flexor muscles of the fingers, the tendon of the long flexor muscle of the first finger, and the median nerve [1]. Thus, an increase in volume within this space, however small, can lead to compression of the nerve [2].

The dimensions of the carpal tunnel are 20 mm at the level of the hook of the hamate, narrow compared to the proximal (24 mm) or distal (25 mm) dimensions [3]. In healthy individuals the intratunnel pressure ranges between 3–5 mmHg in neutral position [4]; vascular compromise of the nerve begins when the pressure exceeds 20–30 mmHg [5], which could be favored by certain postures or gestures as common as using a computer mouse [6]. Adjacent musculature may also play a role in the etiology; in a cadaver study, the muscle belly of the flexor digitorum superficialis was observed to access the proximal tunnel during wrist extension, similar to the lumbrical muscles observed in contact with the distal tunnel during metacarpophalangeal flexion [7].

The median nerve is responsible for the sensory innervation of the first three fingers and the radial border of the fourth, which is why patients who suffer compression of the median nerve inside the tunnel report pain and hypoesthesia in these fingers. The palm of the hand, however, usually remains unaltered, as it is innervated by a sensory cutaneous branch that emerges 6 cm proximal to the entrance of the nerve into the tunnel [5].

Carpal Tunnel Syndrome (CTS) is the most common compressive neuropathy, accounting for 90% of all neuropathies [1,8]. It is estimated that 1 in 5 patients with symptoms of pain, paresthesia and numbness of the hand will be diagnosed with CTS based on examination and an electrodiagnostic test [8].

Its prevalence ranges from 3.8% to 7.8% of the general population, being more prevalent in females than in males (10% vs. 5.8%), with an incidence of 9.3 new cases per 100 persons per year [9,10]. It leads to increased absenteeism, morbidity and, depending on the country, a cost of up to 2 billion dollars per year [10].

It is characterized by a combination of motor and sensory symptomatology leading to functional impairment, manifested by weakness of the intrinsic hand musculature with decreased grip strength, pain and paresthesia [4]. The etiology of this entity remains uncertain, and most cases are labelled idiopathic; recently, magnetic resonance imaging, biomechanical and histological studies have proposed a close relationship between neuronal vascular dysfunction, synovial tissue, flexor tendons and their sheaths, with the onset of the syndrome [11].

Diagnosis has traditionally been established by anamnesis and physical examination; electrodiagnostic testing is frequently used as a method of confirmation and assessment of severity. Although these studies offer 85% sensitivity and 95% specificity [12], they have the disadvantage of being invasive and unpleasant for patients. Recent advances in ultrasonography techniques have considerably improved the quality of ultrasound imaging of the nerve while providing more compact and economical equipment that makes it more accessible to the clinician, making ultrasonography a potential tool for the evaluation of entrapment neuropathies.

The high prevalence of CTS in the general population causes a high demand for neurophysiological studies, which to date is considered the gold standard in its diagnosis; this leads to long waiting times for the test to be carried out as well as an increase in direct and indirect costs derived from the disease. The availability of an additional technique that allows CTS to be diagnosed quickly, non-invasively and inexpensively would represent a clear advance in healthcare and an improvement in the efficiency of the healthcare system. In this sense, the present study aims to investigate the usefulness of ultrasound in the diagnosis of CTS in comparison with the gold standard, the specific ultrasound parameters that predict the existence of the syndrome in comparison with neurophysiological studies, and thus try to standardize a study protocol and reference values that determine the presence or absence of the syndrome.

Explanation for the choice of comparators {6b}

In our study, we included patients with CTS (confirmed by neurophysiological testing) and patients considered "healthy" (with negative neurophysiological testing), in order to analyze the ultrasound differences between the two groups.

Study setting {9}

We compared the NPT based on the recommendations of the American Association of Electrodiagnostic Medicine (AAEM) protocol with the ultrasound study protocol described in Chen et al. [10] for assessing its effectiveness. The study protocol included different variables: cross-sectional area at the entrance and exit of the tunnel, range of nerve thinning, wrist-forearm index, flexor retinaculum bulging, power Doppler uptake and the existence of adjacent wrists or masses.

Patients will come consecutively from the Neurophysiology Department of the Virgen Macarena Hospital in Seville (Spain), with clinical suspicion of CTS and fulfilling the inclusion/exclusion criteria. Electroneurography or ultrasound diagnosis techniques will be performed to compare their effectiveness. This study has been approved by the Andalusian Biomedical Research Ethics Committee (Comité de Ética de la Investigación Biomédica de Andalucía) on August 3^rd, 2020 (a462739eb0fd339704749b3269feed7855a3aecd).

Eligibility criteria {10}

We establish the following inclusion criteria in order to be able to take part in the study:

1. Presenting clinical symptoms suggestive of CTS, at least one of the following:
   • Neuropathic profile pain in wrist-hand at the level of the median nerve territory (wrist, first-second-third, and radial half of the fourth finger by palmar and/or volar side), possibility of irradiation of the pain towards the forearm and/or arm.
   • Nocturnal symptoms.
   • Hypotrophy or atrophy of the thenar eminence.
   • Self-described sensory alteration (sensation of cramping, numbness, tingling, hypoalgesia or analgesia).
2. Age: 18 to 75 years.
3. Consent to participate in the study.

Similarly, the exclusion criteria established are as follows:

1. Previous surgery on the wrist
2. Acquired or hereditary pathology causing peripheral neuropathies
3. Polyneuropathy of any etiology
4. Rheumatoid arthritis
5. Diabetes Mellitus

Intervention description {11a}

We work side by side with the neurophysiology service of the Virgen Macarena hospital (Seville, Spain); they will apply the neurophysiological evaluation protocol between 1 and 3 months previous to the echography study.

• Neurophysiological:

The neurophysiological evaluation protocol is based on the recommendations of the AAEM [12], listed in Table 1.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Recommendations of the American Association of Electrodiagnostic Medicine (AAEM).

https://doi.org/10.1371/journal.pone.0281221.t001

• Ultrasound:

In our study we followed the ultrasound protocol described by Chen et al. [10] (Table 2), discarding dynamic parameters and adding others based on the most recent literature.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. Ultrasound protocol.

https://doi.org/10.1371/journal.pone.0281221.t002

Criteria for discontinuing or modifying allocated interventions {11b}

Not applicable since the intervention comprises a one-time diagnostic procedure.

Strategies to improve adherence to interventions {11c}

Not applicable since the intervention comprises diagnostic procedure.

Relevant concomitant care permitted or prohibited during the trial {11d}

No concomitant care was prohibited during the study.

Outcomes measures {12}

Primary outcome measure

Secondary outcome measure

1. Demographic variables: sex, age, time of evolution of the clinic, manual dominance, hand affected, body mass index, employment status, type of work (pink, white or blue collar), time worked, triggers of symptoms, what symptoms are relieved with, as well as toxic habits, whether they are smokers (to be answered affirmative or negative) or alcohol consumers (to be answered affirmative or negative).
2. Electrodiagnostic variables: motor conduction velocity of the median nerve (calculating distal motor latency and wrist-elbow motor conduction velocity), motor conduction velocity of the ulnar nerve (calculating distal motor latency, lower and upper wrist-elbow motor conduction velocity), orthodromic sensory conduction of the median, ulnar and radial nerve, as well as antidromic sensory conduction of the median, ulnar and radial nerve.
3. Ultrasound variables:
   • The cross-sectional area (CSA), to be performed with linear tracing, measured over the inner border of the epineurium of the median nerve at the level of the entrance and exit of the carpal tunnel. We take the average of 3 consecutive measurements.
   • The range of median nerve thinning understood as the ratio of the transverse and anteroposterior diameters of the nerve in the proximal-medial intracarpal region, at the level of the pisiform.
   • The wrist-forearm index, understood as the ratio of the CSA at the tunnel entrance to the CSA at 12 cm from the distal wrist crease.
   • The flexor retinaculum bulge, which translates an increase in intracarpal pressure. We measure this by tracing a horizontal from the superior pole of the scaphoid to the superior pole of the pisiform, and a vertical from the superior end of the flexor retinaculum to the horizontal.
   • Power Doppler uptake, which determines the hypervascularization of the nerve, especially in the periphery, at the expense of low-flow intraneural vessels. The uptake is at mid-tunnel level, between the pisiform and hamate nerves.
   • The existence of notches in the nerve in the longitudinal ultrasound view, including the entrance, interior and exit of the tunnel.
4. Boston Carpal Tunnel Questionnaire (BCTQ), to assess both the symptoms and the subjective functional impact of patients.
5. Cost evaluation: To carry out a cost-effectiveness analysis, we will evaluate the costs of the consumables established by the Andalusian Health Service used in both tests and the average time for each of the tests, in order to calculate the cost of the professional when performing each of them.

Participant timeline {13}

As shown in Fig 1, we have established a first period for the study, from July 2020 to December 2020, where the initial literature search is carried out, the study protocol is drawn up and submitted to the ethics committee for approval. During this time, a training period is also planned for the researcher in the performance of the ultrasound protocol, a study carried out in the Physical Medicine and Rehabilitation Services of the Virgen Macarena, Virgen del Rocio and Valme University Hospitals, where ultrasound studies are performed on patients with CTS.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. Schedule of enrolment, interventions, and assessments.

https://doi.org/10.1371/journal.pone.0281221.g001

During the second period of the study, from January 2021 to December 2021, we planned to have performed ultrasound studies on patients. However, due to the Covid-19 pandemic situation, we were forced to prolong this period until September 2022.

We propose a third and final period from September 2022 to May 2023 for data collection, statistical analysis and writing of the study, as well as the relevant corrections.

Sample size {14}

To calculate the sample size (EPIDAT program) we proposed:

• Expected sensibility: 78%
• Expected specificity: 87%
• Level of confidence: 95%

In order to evaluate ultrasound as a diagnostic test versus neurophysiological study in patients with a high clinical suspicion of suffering from CTS, expected values of sensitivity and specificity of the ultrasound test of 78% and 87% respectively, a maximum precision of 5% and a confidence level of 95% are considered, resulting in the need to evaluate a minimum of 264 patients with a positive result in the nerve conduction study as well as a minimum of 174 patients with a negative result, which means a total sample of 438 patients (Table 3). The number of positive and negative tests were known because the results were analysed by a secondary investigator. Therefore, the secondary investigator would indicate whether it was necessary to continue appointing patients for ultrasound scans until the total number of patients in both groups were obtained.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 3. Sample size.

https://doi.org/10.1371/journal.pone.0281221.t003

The calculation was performed using the EPIDAT software, considering the parameters that underlie the calculation of the required sample size for conducting a diagnostic test evaluation study. The parameters requested by the software are based on sensitivity, specificity, and precision, which were obtained from previous studies.

Recruitment {15}

The patients included in the study will be obtained consecutively through the consultations of the Neurophysiology Department of the Virgen Macarena Hospital (Seville, Spain), provided that they have clinical suspicion of CTS.

Sequence generation {16a}

Not applicable, since there are no intervention groups. All patients we recruited from the neurophysiology department of the Vigen Macarena Hospital (Seville, Spain), who underwent an ultrasonography following a specific protocol.

Concealment mechanism {16b}

Not applicable, since there are no intervention groups.

Implementation {16c}

Not applicable, since there are no intervention groups.

Who will be blinded {17a}

Not applicable. This protocol is for an evaluation study of diagnostic tests comparing electroneurography and ultrasound. Nevertheless, the result of the neurophysiological study will be blinded so the principal researcher does not know the expected outcome of the ultrasound (results of the neurophysiological testing).

Procedure for unblinding if needed {17b}

Not applicable. No unblinded procedure will be carried out due to the diagnostic nature of the study but the results of neurophysiological testing are analysed by a secondary researcher who determines when the sample is fully collected.

Data collection methods

Data management {19}

The data will be entered by one of the two principal investigators, in batches of 50 patients, performing a subsequent review of the data to ensure that there are no errors when entering them.

We will not use the patient’s personal data but the personal health code (NUSHA).

The storage will be done in Excel format on a computer and on an external storage system (USB).

Statistical methods for primary and secondary outcomes {20a}

After an initial exploration of the data, quantitative variables are expressed as means and standard deviations or medians and quartiles in case of asymmetric distributions, and qualitative variables as percentages.

The calculation of the measures of test accuracy of the ultrasound test as a diagnostic test for CTS: sensitivity, specificity, predictive values and likelihood ratios, are carried out considering the neurophysiological study as the reference test (gold standard).

Likewise, the reliability of the ultrasound is analyzed by calculating the degree of agreement between the diagnoses of the two tests using Cohen’s Kappa coefficient. The concordance of both tests with the established clinical severity is also studied.

The analysis of the quantitative variables did not present normality (Gaussian assumptions), therefore, for statistical inference, the Mann Whitney test (violations of Gaussianity) was used for independent quantitative variables. Shortly afterwards, a multivariate logistic regression model will be created to try to determine the best subset of them that predict CTS [20].

Finally, ROC curves will be performed to evaluate cut-off points for certain quantitative ultrasound parameters that discriminate well against CTS. In the ROC curves, the optimal cut-off point for greater sensitivity and specificity was determined by using the Youden index. In those ultrasound parameters where the ROC curve did not allow adequate discrimination, the median value of the data distribution was used.

In summary, the univariate analysis aims to identify which ultrasound parameters are related to the result of the CTS, thus providing elements of analysis to determine the cut-off points to be evaluated by means of ROC or medians, as well as the main variables to be considered in the multivariate model.

Data analysis is performed with SPSS 25.0 statistical software.

Methods for additional analyses (e.g. subgroup analyses) {20b}

Not applicable, no additional analysis will be performed.

Methods in analysis to handle protocol non-adherence and any statistical methods to handle missing data {20c}

The main analysis will be the concordance between the results of electroneurography and ultrasound in the diagnosis of CTS. Different parameters will be considered and in case of missing values these will be imputed by multiple imputation, selecting some of the methods that best suits the data set, the maximum interaction method, monotonic method, linear regression, predictive mean matching (PMC) among others.

Data monitoring

Harms {22}

Although the diagnostic tests evaluated in this study are harmless for humans, we have to consider the possibility or adverse effects, and how we will collect them when applying the test throughout the study.

On the one hand, at the end of the ultrasound, we will collect the appearance of any adverse effect by answering a Yes / No question, and if so, a brief description of it. If the adverse effect come up hours after the test was carried out, all patients have, in their informed consent copy, an email address of the principal investigator that they can attend.

Regarding the possible side effects derived from the neurophysiological test, being a different department from Rehabilitation, Regarding the possible side effects derived from the neurophysiological test, being a different department from Rehabilitation, it is they who manage the appearance of these adverse effects as part of their daily clinical activity.

Auditing {23}

Of the two main researchers, one of them will carry out the ultrasounds and data collection, while the other one will carry out the quality control of the study, a well-done data collection, the proper functioning of the ultrasound system, the appointment of patients, etc.

Research ethics approval {24}

This study has been approved by the Andalusian Biomedical Research Ethics Committee (Comité de Ética de la Investigación Biomédica de Andalucía). Written informed consent will be obtained from all participants before entering in the study.

Protocol amendments {25}

We will communicate it by email with the relevant parts, as the secondary investigators or the statistician.

Consent or assent {26a}

The patients included in the study will be obtained consecutively through the consultations of the Neurophysiology Department of the Virgen Macarena Hospital (Seville), provided that they have clinical suspicion of CTS. These patients will be invited to participate in the study and will be asked to sign the informed consent form, which reflects the documentary commitment to confidentiality and safeguards in the treatment of their data by the researchers, as well as information on what their participation in the study will consist of.

Additional consent provisions for collection and use of participant data and biological specimens {26b}

Not applicable, as these will not be collected.

Confidentiality {27}

In order to protect participants personal information, we did not use their names, but their personal health code called “NUSHA” (e.g.: AN0123456789).

Declaration of interests {28}

The authors declare that they have no conflict of interest

Access to data {29}

Only the two main investigators will have access to the final trial dataset as well as the statistician who will handles the database.

Ancillary and post-trial care {30}

Being the echography an innocuous test, we do not stand any provision for post-trial test. Not applicable.

Dissemination plans {31a}

The trial results will be communicated to the scientific community through journal as well as communications at scientific meetings or congresses. As for the patients participating in the trial, again within the informed consent they will have an email address to contact with in case they want to know the results; this way we do not bother those participants who do not want to know the results.

Authors’ contributions {31b}

MPMC is the principal investigator, having developed the idea and protocol with the help of the research team. She will be responsible for collecting the sample. MRPD led the study proposal and developed the protocol, being the main methodologist. ASJS participated in the development of the protocol until the final version was reached. MAL facilitated the collection of the sample by performing the neurophysiological tests in the first stage, and by bringing the patients to the principal investigator. In turn, she provided the neurophysiological protocol used to be included in the manuscript. GJJ facilitated the collection of the sample by performing the neurophysiological tests at an early stage, getting the patients to the principal investigator.

Plans to give access to the full protocol, participant level-data and statistical code {31c}

The full study protocol can be accessed by contacting the trial team at (e-mail). Anonymized participants level data and statistical codes used for the analysis are available on request and subject to data sharing agreements. Data sharing may only be possible after relevant embedded trials have been published.