Diagnostic criteria and outcome measures in randomized clinical trials on carpal tunnel syndrome: a systematic review

ABSTRACT BACKGROUND: The diagnostic criteria for carpal tunnel syndrome (CTS) lack uniformity. Moreover, because CTS is a syndrome, there is no consensus as to which signs, symptoms, clinical and complementary tests are more reproducible and accurate for use in clinical research. This heterogeneity is reflected in clinical practice. Thus, establishing effective and comparable care protocols is difficult. OBJECTIVE: To identify the diagnostic criteria and outcome measures used in randomized clinical trials (RCTs) on CTS. DESING AND SETTING: Systematic review of randomized clinical trials carried out at the Federal University of São Paulo, São Paulo, Brazil. METHODS: We searched the Cochrane Library, PubMed, and Embase databases for RCTs with surgical intervention for CTS published between 2006 and 2019. Two investigators independently extracted relevant data on diagnosis and outcomes used in these studies. RESULTS: We identified 582 studies and 35 were systematically reviewed. The symptoms, paresthesia in the median nerve territory, nocturnal paresthesia, and special tests were the most widely used clinical diagnostic criteria. The most frequently assessed outcomes were symptoms of paresthesia in the median nerve territory and nocturnal paresthesia. CONCLUSION: The diagnostic criteria and outcome measures used in RCTs about CTS are heterogeneous, rendering comparison of studies difficult. Most studies use unstructured clinical criteria associated with ENMG for diagnosis. The Boston Questionnaire is the most frequently used main instrument to measure outcomes. REGISTRATION: PROSPERO (CRD42020150965- https://www.crd.york.ac.uk/prospero/display_record.php?RecordID=150965).


INTRODUCTION
Carpal tunnel syndrome (CTS) is the most prevalent peripheral neuropathy in the world. 1,2 Although some patients are treated conservatively, most require surgical treatment, which generates spending more than US$ 2 billion/year. 2 The socioeconomic impact of the disease drove numerous randomized clinical trials (RCTs) to determine the best treatment for CTS. To identify effective interventions, accurate and relevant outcomes for the patient are needed. [3][4][5][6] There is extensive literature about objective outcomes (e.g., loss of strength) and variables derived from nerve conduction studies. However, how these outcomes translate into tangible benefits to the patients remains unclear. [3][4][5][6] In CTS, the lack of uniform criteria poses a challenge in diagnosis. Thus, Graham proposed well-defined criteria, based on a robust methodology. 7 Moreover, because CTS is a syndrome, experts do not agree on which signs, symptoms, clinical and complementary tests are more reproducible and accurate in clinical research. 8 This heterogeneity is reflected in clinical practice and has led to difficulty in establishing effective and comparable care protocols. 9,10 Thus, studies must use precise diagnostic methods and clarify the main post-surgical outcomes to be evaluated in patients with CTS.
Systematic reviews promote synthesis, provide a comprehensive view, and recommend the best available evidence on a topic. Diagnostic and rational management criteria are of interest. 8,10 Importantly, evaluation outcomes should reflect the impact of treatments on body structure and function, including activity limitations and participation restrictions, through a broad evaluation model. [11][12][13] The Classification of Functioning and Disability and Health (ICF), approved in 2001 by the World Health Organization, proposed a comprehensive assessment from both individual and social perspectives. 14,15 The model aimed to recognize the abnormalities in the body structure, identify the consequences on function, and describe the repercussions and adaptations to such changes in the individual's social dynamics. 14 A previous systematic review on the subject was published in 2006. 8 Given the importance of the topic, we sought to give an update.

OBJECTIVE
The objective of this systematic review is to compare the diagnostic criteria and outcome measures based on ICF used in CTS over the past 15 years.

Criteria for selection of studies and procedures
After the initial screening based on the title and abstract, the full-text articles were independently reviewed by RLS and AFZ. These were included if they met the eligibility criteria enumerated below. Disagreements were resolved by a third author, VYM.
The inclusion criteria were: 1. Type of study: randomized clinical trials with follow-up longer than three months; 2. Patients: adults (>18 years) with initial diagnosis of CTS.

Exclusion criteria
1. Studies not published in English.

Methodological quality assessment
We use the Cochrane Collaboration Risk-of-Bias tool, 51 which evaluates: 1. random sequence generation (selection bias); 2.
Other sources of bias (other biases).

Assessment of outcomes based on the International Classification of Functioning, Disability and Health (ICF)
This classification facilitates understanding of health determinants and health-related effects through a standardized and comprehensive terminology. 15 Correlating the pathophysi-

Data analysis
The data collected were presented in tables. Each study was labelled according to its author. The data was managed in Excel 2020 (Microsoft Corporation, Redmond, Washington, United States).

RESULTS
From the 582 studies screened, 35 were included in the systematic review (Figure 1).   Table 1 provides a meta-summary of the characteristics of the studies included.  The outcome measures reported in the RCTs were classified according to the domains of the ICF: A) Body functions and structures ( Table 2); 16-50 B) Activity limitations ( Table 3)

Characteristics of studies and evaluated outcomes
We analyzed studies that evaluated the effectiveness of different The studies analyzed the following clinical diagnostic criteria for CTS: paresthesia in the territory of the median nerve, night paresthesia, and Phalen's and Tinel's tests (part of the six criteria described by Graham) (18 studies; 51%), the Katz diagram (3; 9%) and all the Graham criteria -CTS-6 (2; 6%). Other studies (12; 34%) did not specify the diagnostic method used ( Table 1). Studies that used only part of the six criteria described by Graham 7 were classified as paresthesia and special tests. Electroneuromyography (ENMG) was a complementary examination in 31 studies (89%) and ultrasonography in only one (3%). The studies that used ENMG were then classified based on the use of the Padua criteria, 52 used by 22 (71%). Three other studies (9%) did not use any type of complementary examination ( Table 1).     16 Figure 2 presents the risk of study bias.  Because surgical intervention was involved, blinding was difficult; most were classified as having uncertain or high risk of bias.

Categorization of the outcomes analyzed based on the International Classification of Functionality, Disability and Health (ICF)
The outcomes reported in the ECR were categorized according to the three domains of the ICF. Only one study used more than one questionnaire.   Motor functions, included in 16 studies (46%), were operationally defined as manual grasping force (14; 88%), tweezers (10;
Sensory function was evaluated in 11 studies (31%). The most studied variable was two-point discrimination (10; 91%), followed by the monofilament test (8; 73%) and vibration (1; 10%). Finally, the body structures were analyzed in 18 studies (51%), through sensory and motor nerve conduction (12; 67%), local sensitivity disorders (9; 50%), wound complications (3; 17%) and causalgia (1; 6%). (Table 3) 17 x Not applied x Zyluk 18 Applied Forward 19 Not applied Atroshi 20 x Applied x Huemer 21 Not applied Pomerance 22 Not applied x Atroshi 23 Not applied x Gordon 24 x Applied Faraj 25 Not applied x Nabhan 26 x Not applied x x x Eriji 27 x Not applied Uçar 28 Applied Larsen 29 Not applied Tarallo 30 Applied x x Ullah 31 Not applied Andreu 32 x Not applied Vanni 33 Applied x Peñas 34 Not applied SadatsuneI 35 Not applied Rojo-Manaute 36 Not applied x Acar 37 Not applied Gumustas 38 x Applied Cho 39 Applied x Herold 40 x x Not applied Peñas 41 x Applied x Logli 42 Not applied Peñas 43 Applied Peñas 44 Applied Boriani 45 Applied Kleermaeker 46 Applied Kanchanathepsak 47 Applied Oh 48 Applied Rimdeika 49 x Not applied Zhang 50 Applied  53 Graham's criteria are widely recommended. 7 However, systematic reviews challenge the use of two-point discrimination (one of Graham's criteria), due to its low diagnostic sensitivity for CTS. 54 Our results suggest the same, because most of the studies do not use two-point discrimination.

B) Limitations of activity
ENMG is widely used as a complementary quantitative method and is considered an important tool for analyzing and monitoring CTS intensity. 52,54,55 Few studies utilized ENMG to predict outcomes for CTS. 52,55,56 The ENMG Padua criteria (Electroneuromyography classification for stratification of median nerve involvement in CTS), is one of the most widely used tools. 52 However, although ENMG was a predominant complementary examination in the included studies, most did not use the quantitative criteria of Padua fully.
In addition to effective diagnostic methods, the correct definition of primary and secondary outcomes in RCTs allows the generation of responses to the hypotheses previously defined in these studies. 8,10 The focus of the included studies was the outcomes of body function and structure, with less attention to activity limitation and participation restriction. BQ was the most widely used, being an important outcome measure for assessing symptoms (body function and structure) and functional capacity (activity limitations).
BQ has good psychometric properties in patients with CTS. [57][58][59] Thus, its use should replace other non-standardized methods. 59 Similarly, previous systematic reviews were less focused on outcomes related to activity limitations and participation restriction. 8  This is the first systematic review aimed at identifying the diagnostic criteria and the outcome measures used in ECR on CTS.
The protocol was previously published in the PROSPERO database, restricting biases in methodology and enhancing credibility. 6 In addition, in order to improve the quality of the report of this systematic review, the PRISMA statement was used.
Our review has several limitations. We only looked for studies Contrary to the literature, most studies do not prioritize patientreported outcomes as relevant or primary outcomes. A task force is needed to standardize CTS research.