ARTICLE IN PRESS Seminars in Arthritis and Rheumatism 000 (2018) 1 13 Contents lists available at ScienceDirect Seminars in Arthritis and Rheumatism journal homepage: www.elsevier.com/locate/semarthrit Orthosis for rhizarthrosis: A systematic review and meta-analysis Sandra Mara Meireles, Anamaria Jones, Jamil Natour* ~o Paulo, Escola Paulista de Medicina, Sa ~o Paulo, SP, Brazil Rheumatology Division, Universidade Federal de Sa TAGEDPA R T I C L E I N F O Keywords: Hand Splint Orthosis Osteoarthritis Thumb Rehabilitation TAGEDPA B S T R A C T Objective: investigating the effectiveness of orthosis for rhizarthrosis by means of a systematic review and meta-analysis. Methods: A systematic review was carried out using eight electronic databases. The randomized controlled trials included were those presenting subjects using orthosis for rhizarthrosis compared with individuals without orthosis or other rehabilitation interventions, as well as studies that compared different types of orthosis. The systematic review was performed according to the Cochrane methodology. The statistical software Review Manager 5.3 was employed to analyze the data. Results: Fourteen studies were included in the review and three of them participated in the meta-analysis. The orthosis group had a reduction in pain in the long term as compared to the control group with a statistically significant difference, a medium effect size, and low-quality evidence [Effect size -0.52, Confidence Interval 95% -0.94 to -0.11, p = 0.01), I2 = 50%]. The orthosis group presented improvements regarding function in the long term as compared to control group, with a statistically significant difference, a medium effect size, and moderate quality of evidence [Effect size -0.44, Confidence Interval 95% -0.72 to -0.15, p = 0.002), I2 = 0%]. Conclusion: the orthosis for rhizarthrosis presents low-quality evidence for reducing pain in the long term and moderate evidence for an increase in function in the long term. Since imprecision and inconsistency of the data were aspects which influenced the quality of the evidence, future studies with larger samples and standardized data are needed. © 2018 Elsevier Inc. All rights reserved. Introduction Rhizarthrosis or thumb base osteoarthritis (OA) refers to OA that affects the trapeziometacarpal (TMC) and scaphotrapezial joints. Typically, patients have pain in the base of the thumb, loss of thenar muscles strength, and deformity of the thumb with subluxation of the first carpometacarpal joint which leads to an inability to abduct the thumb from the second finger. With the progression of the disease, rhizarthrosis’ patients present loss of grip and pinch strength which leads to difficulties in hand function and a reduction in quality of life [1 5]. Orthoses for rhizarthrosis are employed to reduce pain and inflammation as well as to increase joint stability [6 9]. The traditional concept of orthosis for rhizarthrosis is based on immobilizing the joint at the base of the thumb in palmar abduction to a neutral position that enables all ligaments and muscles to be in a resting position and at maximum joint spacing [10 13]. Spaans et al. [14] recently reported that studies on orthosis for rhizarthrosis show marked heterogeneity in the types of orthoses used, durations of treatments and types of study designs. The authors ~o Paulo Research Foundation (FAPESP) [grant Funding: This work was supported by Sa number 2013/11720 4]. * Corresponding author. E-mail address: jnatour@unifesp.br (J. Natour). https://doi.org/10.1016/j.semarthrit.2018.07.013 0049-0172/© 2018 Elsevier Inc. All rights reserved. indicated that an orthosis may reduce pain in patients with rhizarthrosis; however, it does not improve function, dexterity and strength. Moreover, in 2015, Bertozzi et al. [15] published a systematic review and meta-analysis that included three randomized controlled studies. The authors concluded that there was moderate evidence regarding improvements in the function of patients with rhizarthrosis who used the orthosis in the long term. Few studies of good methodological quality on orthoses in thumb base OA are available in the literature, and different authors recommend the need for additional studies [14 24]. There is no current systematic review in the literature that solely assesses the effects of orthoses on rhizarthrosis. This study aimed to conduct a systematic review and a metaanalysis of randomized controlled studies on efficacy of orthosis in rhizarthrosis. Methods The systematic review was performed according to the Cochrane methodology [25]. Inclusion criteria for studies The inclusion criteria comprised randomized controlled clinical trials that involved subjects with rhizarthrosis classified using clinical ARTICLE IN PRESS 2 S.M. Meireles et al. / Seminars in Arthritis and Rheumatism 00 (2018) 1 13 and radiological criteria [26], in which orthosis was used for rhizarthrosis compared with individuals without orthosis or other rehabilitation interventions, as well as studies that compared different types of orthosis. Outcomes Primary: pain in the thumb base assessed with the Visual Analog Scale (VAS) and hand function assessed with the Australian/Canadian hand osteoarthritis index (AUSCAN), the Michigan Hand Outcomes Questionnaire (MHQ), the Disabilities of the Arm, Shoulder and Hand (DASH), the Cochin Hand Functional Scale (CHFS) and the Canadian Occupational Performance Measure (COPM). Secondary: range of movement of the thumb measured with a goniometer, palmar grip strength measured with a Jamar hydraulic dynamometer, pinch strength measured with a pinch gauge, hand dexterity assessed with the O'Connor test, quality of life assessed with the Short Form 36 Health Survey Questionnaire, inflammation, stiffness, and patient satisfaction assessed with a Likert scale, use of analgesics and adverse effects. Therapy Systematic Evaluation of Evidence (OTseeker) and Physiotherapy Evidence Database (PEDro) (Appendix 1). Research was also conducted with the databases ClinicalTrials.gov and World Health Organization International Clinical Trials Registry Platform (WHO ICTRP) to identify ongoing randomized controlled studies. Moreover, searches were performed with respect to the references of the studies electronically identified and non-indexed journals in the databases described, in addition to searches into annals of scientific events in the area. EndNote online (Clarivate Analytics, Philadelphia, PA) was used to manage the references selected. Study selection A list was created that contained the titles and abstracts of the potentially relevant studies identified by the literature search. Two independent reviewers (SMM and AJ) applied the pre-determined inclusion criteria to the full texts. Conflicts could be solved by a third researcher (JN); however, there were no conflicts at this stage. Data extraction and management Search method for study identification A search strategy was developed in the following databases with no restriction on the date of publication or language until December 2017: Cochrane Central Register of Controlled Trials (CENTRAL), Medical Literature Analysis and Retrieval System Online (MEDLINE), PubMed, Excerpta Medica Database (EMBase), Literatura Latino^ncias da Sau de (LILACS), Indice BibAmericana e do Caribe em Cie fico Espanhol de Cie ^ncias da Sau de (IBECS), Occupational liogra The same two reviewers, in an independent manner, used a standardized form to extract data, which included: methods, characteristics of participants, intervention, outcomes and results. Assessment of the risk of bias in included studies The two independent reviewers also assessed the risk of bias in the studies using the Cochrane Collaboration's Risk of Bias tool [25]. Fig. 1. Flow chart of the screening and inclusion of studies. Table 1 Qualitative analysis and PEDro score of included studies Outcome measures and follow up Reported results** PEDro score Adams et al. [29] RCT 4 weeks Baseline and 4 weeks: 1 AUSCAN pain, function, and stiffness subscales 2 MHQ ADL and Aesthetics subscales 3 Global rating of change scale No information about drop-outs or intention-to-treat analysis The splint group showed significantly less improvement than the group under OT care. Pain - AUSCAN pain 0 20 (higher scores - worse pain): The scores increased 3.66 points in the splint group The scores increased 0.89 points in the OT group The scores decreased 3 points in the placebo group (p = 0.011 among groups) Function - MHQ ADL 0 100 (higher scores - better function): The scores decreased 14.28 points in the splint group The scores increased 9.52 points in the OT group The scores increased 21.43 points in the placebo group (p < 0.001 among groups) Splinting had a significant effect on pain and function in patients with thumb base OA. Pain - VAS 0 10: The scores decreased 1 point in the splint group The scores decreased 0.11 points in the control group (p = 0.001 between groups) Function - MHQ total 0 100 (higher scores - better function): The scores increased 0.85 points in the splint group The scores decreased 4.11 points in the control group (p = 0.048 between groups) Both splints reduced pain and increased function, and pinch strength when compared to the control group. The custom made thumb splint showed superior results in pain reduction. Pain - VAS 0 10: The scores decreased 4.5 points in the custom made splint group The scores decreased 3.8 points in the prefabricated splint group (p = 0.024 between groups) Function DASH 0 100 (higher scores - worse function): The scores increased 11.9 points in the custom-made splint group The scores increased 17.2 points in the prefabricated splint group (p = 0.136 between groups) There was no difference between the prefabricated neoprene splint group and the customized thermoplastic splint group. The prefabricated neoprene splint was less expensive, and patients found it more comfortable. Pain - VAS 0 10: The scores decreased 0.7 points in the neoprene group The scores decreased 0.8 points in the thermoplastic group (p = 0.78 between groups) Function DASH 0 100 (higher scores - worse function): The scores decreased 2.5 points in the neoprene group The scores decreased 3.8 points in the thermoplastic group (p = 0.71between groups) There was no difference among the groups. After 7 months of treatment, 10 of the 33 participants chose to be operated. During the following seven years, 2 more patients opted for surgery. 4/10 OT care 18 subjects OT care along with a biomechanically active splint OT care along with a biomechanically inactive (placebo) splint Arazpour et al. [30] RCT 25 subjects 4 weeks Intervention group: Custom made thermoplastic thumb splint Baseline and 4 weeks: 1 Diameter and cross-sectional area of the thenar muscles 2 MHQ total 3 Pain VAS No drop-outs Control group: no intervention Bani et al. [31] RCT crossover 10 weeks Custom made thumb splint 35 subjects Prefabricated thumb splint Baseline, 4, 6, and 10 weeks: 1 Pain VAS 2 DASH 3 Pinch gauge 4 Grip strength dynamometer No drop-outs Control group: No intervention Becker et al. [32] RCT 119 subjects 15 weeks Prefabricated neoprene thumb TMC restriction splint Customized thermoplastic hand-based thumb spica splint Berggrent al. [33] RCT 7 years Technical accessories 33 subjects Baseline, 5 and 15 weeks: 1 Pain VAS 2 DASH 3 Pinch gauge 4 Grip strength dynamometer 5 11-points ordinal satisfaction scale Drop-outs: 57 Intention-to-treat analysis Baseline, 7 months, and 7 years: 1 Number of patients who needed CMC1 joint replacement Drop-outs: 14 No intention-to-treat analysis 6/10 7/10 6/10 5/10 Technical accessories plus semistable textile splint Technical accessories plus nonstabilizing leather splint ARTICLE IN PRESS Duration of study and intervention S.M. Meireles et al. / Seminars in Arthritis and Rheumatism 00 (2018) 1 13 Type of study and sample size* 3 4 Table 1 (continued) Type of study and sample size* Duration of study and intervention Outcome measures and follow up Reported results** PEDro score Buurke et al. [34] RCT crossover 12 weeks Baseline, 4, 8, and 12 weeks: 1.<?show -aptara_TEMP_aptara-?>Pain VAS 2.<?show -aptara_TEMP_aptara-?>Function, comfort and cosmesis VAS 3.<?show -aptara_TEMP_aptara-?>Pinch Gauge 4.<?show -aptara_TEMP_aptara-?>Green Test 5.<?show -aptara_TEMP_aptara-?>Utility research Six participants preferred the supple elastic splint and 2 chose the semi-rigid. Eight patients preferred the permanent use of the splint. Values of changes for the outcomes in each group not provided. The supple elastic material thumb splint scores were significantly better than the other two splints for the factors: Comfort (p = 0.05) Function (p = 0.09) The thermoplastic polyethylene semi-rigid thumb splint scores were significantly better than the other two splints for the aspect: Cosmesis (p = 0.01) The splint group presented significant reduction in pain when compared to the control group. No changes were found in other outcomes. Pain - VAS 0 10: The scores decreased 2.6 points in the splint group The scores decreased 0.7 points in the control group (p = 0.009 between groups) Function - DASH 0 100 (higher scores - worse function): The scores decreased 17.1 points in the splint group The scores decreased 9.8 points in the control group (p = 0.225 between groups) A soft splint had an immediate pain-relieving effect when worn. No differences between groups were found for other outcomes including pain after 8 weeks. Pain - NRS 0 10: The scores decreased 0.08 points in the splint group The scores increased 0.4 points in the control group (p = 0.90 between groups) Pinch strength - in Newtons: The scores decreased 0.7 points in the splint group The scores decreased 0.9 points in the control group (p = p = 0.70 between groups) Values of changes for function in each group not provided. 4/10 10 subjects Thermoplastic polyethylene semi-rigid thumb splint Firm elastic material thumb splint Supple elastic material thumb splint No drop-outs Carreira et al. [35] RCT Intervention group: Custom made Thermoplastic thumb splint No drop-outs Hermann et al. [36] RCT 59 subjects 8 weeks Intervention group: Soft, elastic fabrifoam covering the wrist, the CMC joint, and the MCP joint of the thumb, reinforced with extra material on the volar side of the thumb to give extra support to the TMC joint Control group: No splint Kjeken et al. [37] RCT 70 subjects Both groups received instructions on hand exercises 12 weeks Intervention group: Assistive devices and/or splints Control group no intervention Both groups received information about hand osteoarthritis Rannou et al. [38] RCT 112 subjects 48 weeks Intervention group: Custom made neoprene splint Control group: Usual care Baseline and 8 weeks: 1.<?show -aptara_TEMP_aptara-?>Conventional radiography 2.<?show -aptara_TEMP_aptara-?>Pain NRS during measure of grip and pinch strength 3.<?show -aptara_TEMP_aptara-?>Grippit electronic instrument for Grip and pinch strength 4. AUSCAN Drop-outs: 4 8/10 8/10 No intention-to-treat analysis Baseline and 12 weeks: 1. COPM 2. Pain VAS 3. Fatigue and disease activity VAS 4. AUSCAN 5. mHAQ Drop-outs: 4 Intention-to-treat analysis Baseline, 1, 6, and 12 months: 1. Pain VAS 2. CHFS 3. Patient-perceived disability VAS 4. Pain level during pinch strength VAS 5. Kallman score 6. Pinch electronic dynamometer 7. Kapandji index thumb opposition 8. Kapandji score for index thumb counter-opposition 9. Closure of the first web Drop-outs: 14 Intention-to-treat analysis Activity performance and satisfaction with performance had better results in the intervention group. No changes were found in other outcomes. Activity performance - COPM 0 10 (higher scores - better function): The scores increased 1.4 in the intervention group The scores decreased 1 point in the control group (p < 0.001 between groups) Satisfaction with performance - COPM 0 10: The scores increased 2.5 in the intervention group The scores increased 0.6 points in the control group (p = 0.001 between groups) After one year, the splint group showed a significant difference when compared to the control group in reducing pain and improving function. Pain - VAS 0 100: The scores decreased 22.2 points in the splint group The scores decreased 7.9 points in the control group (p = 0.002 between groups) Function - CHFS 0 90 (higher scores - worse function): The scores decreased 1.9 points in the splint group The scores increased 4.3 points in the control group (p = 0.008 between groups) 8/10 8/10 ARTICLE IN PRESS Control group: No intervention Baseline, 45, 90, and 180 days: 1.<?show -aptara_TEMP_aptara-?>Pain VAS 2.<?show -aptara_TEMP_aptara-?>DASH 3.<?show -aptara_TEMP_aptara-?>Pinch gauge 4.<?show -aptara_TEMP_aptara-?>Grip strength dynamometer 5.<?show -aptara_TEMP_aptara-?>O'Connor Test S.M. Meireles et al. / Seminars in Arthritis and Rheumatism 00 (2018) 1 13 40 subjects 25 weeks Sillem et al. [39] RCT crossover 20 weeks Prefabricated neoprene splint 56 subjects Custom made neoprene splint Wajon and Ada [40] RTC 40 subjects 6 weeks Custom made thermoplastic thumb strap splint Prefabricated short opponens splint 26 subjects Prefabricated long opponens splint Weiss et al. [42] RCT crossover 25 subjects 2 weeks Custom made thermoplastic short opponens splint Prefabricated neoprene short splint Baseline, 1, and 2 weeks: 1. Pain VAS 2. Satisfaction VAS 3. Self-rating scale of 22 ADL items 4. Tip-pinch strength 5. Pain VAS during tip-pinch strength No information about drop-outs or intention-to-treat analysis Baseline, 1, and 2 weeks: 1. Pain VAS 2. Satisfaction VAS 3. Tip-pinch strength 4. Pain VAS during tip-pinch strength 5. Self-rating scale of 22 ADL items No information about drop-outs or intention-to-treat analysis Both splints reduced pain; however, the results were better in the prefabricated neoprene splint group. Pain - VAS 0 10: The scores decreased 1.83 points in the thermoplastic splint group The scores decreased 3.13 points in the neoprene splint group (p = 0.019 between groups) Although the authors did not provide the exact values of changes for function in each group, they reported that more than the double of the patients reported that activities were harder with the thermoplastic splint than with the neoprene splint. 8/10 7/10 5/10 5/10 RCT: Randomized controlled trial, OT: Occupational therapy, AUSCAN: Australian/Canadian Hand Osteoarthritis Index, MHQ: Michigan Hand Outcomes Questionnaire, ADL: Activities daily living, VAS: Visual analogic scale, DASH: Disabilities of the Arm, Shoulder and Hand, TMC: Trapeziometacarpal, OA: Osteoarthritis, CMC1: First Carpometacarpal, NRS: Numerical rating scale, COPM = Canadian Occupational Performance Measure, HAQ = Health Assessment Questionnaire, CHFS: Cochin Hand Functional Scale. * All the patients were diagnosed with thumb base osteoarthritis ** All the authors considered the p value 0.05 statistically significant except Buurke et al. (1999) who considered p value 0.10 as significant. ARTICLE IN PRESS Weiss et al. [41] RCT crossover Both groups received instructions on hand exercises 2 weeks Baseline, 2, and 6 weeks: 1. Pain VAS 2. Pinch gauge 3. Sollerman Test of Hand Function Drop-outs: 6 No intention-to-treat analysis The best result was observed after 4 weeks of treatment. The custom-made splint showed significant reduction in pain when compared to the prefabricated splint. Pain: AUSCAN pain 0 50 (higher scores - worse pain): The scores decreased 5.69 points in the custom-made splint group The scores decreased 2.05 points in the prefabricated splint group (p < 0.02 between groups) Function: AUSCAN function 0 90 (higher scores - worse function): The scores decreased 5.54 points in the custom-made splint group The scores decreased 2.69 points in the prefabricated splint group (p < 0.15 between groups) There were no differences between the thumb strap and the short opponens splints regarding pain, pinch strength, and function. Pain - VAS 0 10: The scores decreased 1.8 points in the thumb strap splint group The scores decreased 2.3 points in the short opponens splint group (p = 0.5 between groups) Function - Sollerman Test 0 80 (higher scores - better function): The scores increased 6.8 points in the thumb strap splint group The scores increased 6.1 points in the short opponens splint group (p = 0.7 between groups) Both splints reduced thumb base pain with no difference between groups. Patients preferred to use the prefabricated short opponens splint. Values of changes for the outcomes in each group not provided. For pain, there was a statistically significant difference between pre and post treatment in both groups with p = 0.001. Regarding function, the authors did not provide the p-value between pre and post treatment. S.M. Meireles et al. / Seminars in Arthritis and Rheumatism 00 (2018) 1 13 Custom made thermoplastic short opponens splint Baseline, 4, 5, 9, and 12 weeks: 1. AUSCAN pain and function subscale 2. Pinch gauge 3. Grip strength dynamometer 4. Comfort, appearance, convenience and durability 5-poins Likert scale Drop-outs: 2 No intention-to-treat analysis 5 ARTICLE IN PRESS 6 S.M. Meireles et al. / Seminars in Arthritis and Rheumatism 00 (2018) 1 13 This tool is based on seven domains as follows: random sequence generation, allocation hiding, blinding of participants and professionals, blinding outcome assessors, incomplete outcomes, report of selective outcomes and other sources of biases. The judgement of the risk of bias for each domain analyzed may be classified into three categories: low-risk, high-risk or uncertain risk of bias. Disagreements between the reviewers were solved by the third person in charge. The methodological quality of the studies included was also analyzed using the PEDro scale [27]. All studies included in the systematic review were qualitatively described. Incomplete data The authors were contacted in cases of incomplete data for analysis. The complete data were considered if at least 85% of the participants were included in the final analysis [25]. Assessment of heterogeneity To verify the heterogeneity, the statistical test I2 was employed. The data were gathered using a fixed-effect model for cases of low heterogeneity (I2 < 50%) and a random effect model for high heterogeneity among the studies (I2 > 50%) [25]. Statistical analysis The statistical analysis of the data was performed using Review Manager (RevMan) [Software] version 5.3. Measure of the effect size The outcomes of the studies included in the meta-analysis were continuous. For each outcome, we calculated the effect size with 95% Confidence Interval (CI) by means of the software RevMan 5.3 using Hedges's g, which is defined as the difference between two means divided by their pooled standard deviation. According to Cohen [28], it is agreed that the values of the effect size (g) are considered small if g < 0.50, medium if 0.50 g < 0.80 and large if g 0.80. Unit of analysis We considered the individual patient to be the unit of analysis. Data synthesis The outcomes of the studies were combined in a meta-analysis when comparable in relation to the population, intervention, assessment and study design. We also applied the tool Grading of Recommendations Assessment, Development and evaluation (GRADE) to grade the quality of the evidence and the strength of the treatment recommendations [25]. Analysis of subgroups We analyzed the following outcomes: pain, function and pinch strength in the short and long term for the studies included in the meta-analysis. Fig. 2. Judgement of the risk of bias made by the reviewers for each item of the cochrane tool for all the studies included in the systematic review. ARTICLE IN PRESS S.M. Meireles et al. / Seminars in Arthritis and Rheumatism 00 (2018) 1 13 Results Six hundred five articles were initially identified. After excluding duplicates and applying the inclusion and exclusion criteria, 14 randomized controlled studies were included in the systematic review [29 42]. Three studies were considered homogeneous and were included in the meta-analysis [35,37,38]. There was no conflict between the reviewers regarding this issue. The flowchart in Fig. 1 shows the selection of articles based on the search strategies. In the data extraction, we determined that all studies included were randomized controlled, with five studies comprising the crossover type [31,34,39,41,42]. The selected articles included 668 patients, and the sample size varied between 10 and 119 patients. The duration of the patient follow-up varied between two weeks and seven years (Table 1). Four studies randomized the participants into three groups [29,31,33,34] whereas the other ten studies randomized the participants into two groups [30,32,35 42]. Of these studies, four studies compared groups with and without orthosis [30,35,37,38], one study analyzed orthosis added to exercises and exercises only [36], three studies compared prefabricated and custom-made orthoses [32,39,42], another study verified long and short orthosis [41] and the last study analyzed two types of custom-made orthoses [40]. The four studies that randomized patients into three groups compared occupational therapy, orthosis and placebo orthosis [29]; prefabricated orthosis, custommade and without orthosis [31]; technological devices, textile and leather orthoses [33]; and thermoplastic, firm and flexible elastic orthoses [34]. 7 The orthoses were manufactured with different types of materials: thermoplastic, neoprene, leather, elastic, and fabric. With the exception of Weiss et al. [41], who compared the effectiveness of long and short orthosis, all the authors only studied short orthoses. Regarding the outcomes, substantial heterogeneity was identified among the studies. Most studies included pain, function, and muscle strength in their evaluations; the Visual Analogical Scale was the most widely used tool to measure pain. The Australian/Canadian hand osteoarthritis index, the Michigan Hand Outcomes Questionnaire, the Disabilities of the Arm, Shoulder and Hand, the Cochin Hand Functional Scale, and the Canadian Occupational Performance Measure were used to gauge function. In most cases, for pinch and grip strength assessment, the authors used the pinch gauge and the Jamar hydraulic dynamometer. Following the information extraction, a table was created with all descriptive data for the selected studies (Table 1). Fig. 2 shows the judgement of the risk of bias made by the reviewers. There was a single conflict between the reviewers; one reviewer concluded the assessor was blinded and there was no blinding of the assessor in the study by Bani et al. [31]. This issue was resolved with the opinion of the third reviewer, JN, who deemed there was a high risk of bias regarding this parameter. Table 1 also presents the study classification according to the criteria of the PEDro scale. Seven studies had a low risk of bias for most items in the Cochrane tool, and the same studies had a general score greater than or equal to 7 in the PEDro scale [31,35 40]. All studies presented in this review used orthosis to treat rhizarthrosis; however, they were very heterogeneous regarding the type of Fig. 3. Meta-analysis of the variable pain in the short and long terms. ARTICLE IN PRESS 8 S.M. Meireles et al. / Seminars in Arthritis and Rheumatism 00 (2018) 1 13 orthosis, study design, follow-up time and assessments. We selected three studies that were similar in these parameters and presented a low risk of bias to perform the meta-analysis: Carreira et al. [35] registered in clinicaltrials.gov (ID: NCT00612248), Kjeken et al. [37] registered in the ISRCTN register (ISRCTN40357804), and Rannou et al. [38] registered in ClinicalTrials.gov (ID: NCT00350896). We used the intention to treat analysis provided in the results of the three studies for performing the meta-analysis. The meta-analysis for the pain variable in the short term (treatment that lasted up to 45 days), with 141 patients, indicated that there was no statistically significant difference between the groups and a small effect size [Effect size ¡0.29, CI 95% ¡1.00 to 0.42, p = 0.42), I2 = 73%]. The meta-analysis for the pain variable in the long term (treatment that lasted more than three months), with 203 patients, indicated that the group using orthosis had a reduction in pain compared with the control group, with a statistically significant difference and a medium effect size [Effect size ¡0.52, CI 95% ¡0.94 to ¡0.11, p = 0.01), I2 = 50%] (Fig. 3). The meta-analysis for the function variable in the short term, with 141 patients, indicated that there was no statistically significant difference between the groups and a small effect size [Effect size 0.11, CI 95% ¡0.22 to 0.44, p = 0.53), I2 = 0%]. The meta-analysis for the function variable in the long term, with 201 patients, indicated that the group that used orthosis had an improvement in function compared with the control group, with a statistically significant difference and a medium effect size [Effect size ¡0.44, CI 95% ¡0.72 to ¡0.15, p = 0.002), I2 = 0%] (Fig. 4). The meta-analysis for the pinch strength variable in the short term, with 142 patients, indicated there was no statistically significant difference between the groups and a very small effect size [Effect size ¡0.02, CI 95% ¡0.35 to 0.31 p = 0.91), I2 = 47%]. The metaanalysis for the pinch variable strength in the long term, with 136 patients, indicated there was no statistically significant difference between the group that used orthosis and the control group with a small effect size [Effect size ¡0.18, CI 95% ¡0.52 to 0.16, p = 0.30), I2 = 0%] (Fig. 5). Table 2 presents a summary of the findings and the GRADE results. Discussion Summary of main results This review analyzed 14 studies on orthosis for rhizarthrosis, which were qualitatively analyzed [29 42] and included a metaanalysis of three studies [35,37,38]. These were taken into consideration when it was possible to compile the results of similar studies. According to the numerical analyses, this study presents a lowquality evidence suggesting that the orthosis group decreased pain in the long term in patients with rhizarthrosis compared with a control group. Furthermore, it presents low-quality evidence that the orthosis and control groups behave in the same way, with no differences regarding pain in the short term. It also presents moderate evidence that function increases in the long term; moreover, there is no difference between the orthosis and control groups concerning the function outcomes in the short term and pinch strength in the short and long term. Besides improving function in the long term, the orthosis is a low cost and low risk intervention which should be considered in the treatment of rhizarthrosis. The Fig. 4. Meta-analysis of the variable function in short and long terms. ARTICLE IN PRESS S.M. Meireles et al. / Seminars in Arthritis and Rheumatism 00 (2018) 1 13 reliability in the effect estimations varies from very low to medium; therefore, future studies may modify this perspective and impact the results. The heterogeneity among the studies and the low number of samples available for the analyses account for these findings. Quality of the evidence The three studies included in the meta-analysis presented a low risk of bias assessed by the Cochrane collaboration tool and the PEDro scale [35,37,38]. The only items of these instruments considered to have a high risk of bias for the three studies were the blinding of the patients and those applying the intervention. In the case of the randomized controlled studies with orthosis as the intervention, it is impossible to blind patients and appliers because they are compulsorily aware of the intervention application. We considered this to be an item not to be taken into account in the assessment of these types of studies. Therefore, we consider that the three studies that were included in the metaanalysis presented a low risk of bias [35,37,38]. For the other studies included in the qualitative description of this review, we identified a similar scenario. However, some tool items were considered uncertain because of the lack of description of the parameters identified in the articles. When we applied the GRADE, we determined that the quality of the evidence varies from low to moderate for the outcomes quantitatively analyzed. The low number of the sample (imprecision) and the heterogeneity (inconsistency) of the data limited the strength of the conclusion of our review. 9 Agreements and disagreements with other reviews In 2007, Egan and Brousseau [18] published a systematic review with the same aim as the current study, i.e., to assess the effectivity of orthosis in OA of the TCM joint. This is the only systematic review identified that exclusively analyzes this intervention. Although Egan and Brousseau [18] conducted a review with the same purpose as ours, their review is a decade older. In recent years, the volume of studies with better methodological quality has increased, and in our study, we included more complete, recent and better designed articles than the previous authors. They included seven studies in their review; however, only four randomized controlled studies were included [33,34,41,42], which were also included in our research. The other studies were not controlled, which characterizes the inaccuracy of their results. In contrast to our study, this review does not present a meta-analysis, which is likely a result of the lack of similarity among the articles included. In contrast to these authors, our research did not have a restriction regarding languages, and although we only identified studies in English, this is an important inclusion criterion which should be taken into account. In the study by Egan and Brousseau [18], a single author verified the risk of bias in the works included and used a tool rarely cited in the literature. Our assessment, however, had the risk of bias assessed by two authors (SMM and AJ), and when there was a conflict of opinion, a third author (JN) was consulted. Our study used the Cochrane collaboration tool to analyze the risk of bias, which is currently considered the most complete form for this type of analysis [25]. Furthermore, our research presents an assessment of the methodological quality using the PEDro scale, which is the most widely used tool in rehabilitation studies. Corroborating our findings, despite the difference in conduction and the lack Fig. 5. Meta-analysis of the variable pinch strength in the short and long terms. ARTICLE IN PRESS 10 S.M. Meireles et al. / Seminars in Arthritis and Rheumatism 00 (2018) 1 13 of quantitative data, the researchers also concluded there is reasonable evidence for using orthosis to alleviate rhizarthrosis pain. However, they classified the studies included as low-quality concerning the methodological characteristics, thus suggesting, as in our research, that additional studies should be conducted regarding the subject. In 2009, Moe et al. [19] published a study on the overview type, compiling systematic reviews on non-pharmacological interventions for OA of the hands, including three reviews that contained orthosis for rhizarthrosis [6,16,18]. They concluded there were no high quality studies on the theme for any assertion, and considering the impact this condition causes, there was an urgent need to conduct novel studies. Their analysis is important to evidence the relevance of our Table 2 Summary of findings findings, as only seven years ago, this issue had been rarely explored in detail. In 2011, Kjeken et al. [22] conducted a systematic review of exercises and orthoses in OA of the hands and included seven studies that assessed the effectivity of orthosis for rhizarthrosis [33 35,38,41,42,43]. Of these studies, only McKee and Eason-Klatt (2006) [43] was not considered in our review because it was not a randomized controlled study. As in our study, these authors state that research studies on orthosis and rhizarthrosis are very different, which hinders the analysis of a set. This was the first review that included a meta-analysis with the studies by Carreira et al. [35] and Rannou et al. [38]. Similar to our group, the authors believe that these two studies are the best published, with a low risk of bias, thus GRADE Exercise compared to control for osteoarthritis of the hand OutcomeFollow up N of participants (studies) Assumed riskwithout exercises The mean hand pain was Short-term pain 47.7 pointsa Follow up: mean 37,5 days Number of participants: 141 (2 RCTs) Corresponding riskwith exercises Anticipated absolute effects (95% CI) Difference Mean hand pain in intervention groups was 5.7 points lower (8.3 points higher to 19.8 points lower) Effect size 0.29 lower (1.00 lower to 0.42 higher) The mean hand pain was Long-term pain* 47.7 pointsa Follow up: mean 6 months Number of participants: 203 (3 RCTs) Mean hand pain in intervention groups was 10.3 points lower (18.6 to 2.2 points lower) The mean hand pain was Short-term function 17.7 pointsd Follow up: mean 37,7 days Number of participants: 141 (2 RCTs) Mean hand function in intervention groups was 1.4 points higher (2.8 points lower to 5.6 points higher) Quality LOW Comments b Effect size 0.52 lower (0.94 lower to 0.11 lower) LOWc Effect size 0.11 higher (0.22 lower to 0.44 higher) MODERATEe Absolute reduction in hand pain 5.7% (8.3% increase to 19.8% reduction) on a 0 100 mm scale Relative change 12% (17% increase to 41% reduction). Absolute reduction in hand pain 10.2% (2.1 to 18.6%) on a 0 100 mm scale Relative change 22% (5 to 39%) Absolute worsening in hand function 1.5% (3.1% improvement to 6.3% worsening) on a 0 90 scale Relative change 8% (16% improvement to s32% worsening) The mean hand pain was Long-term function* 17.7 pointsd Follow up: mean 6 months Number of participants: 201 (3 RCTs) Effect size 0.44 lower (0.72 Mean hand function in lower to 0.15 lower) intervention groups was 5.6 points lower (9.2 to 1.9 points lower) MODERATEe Absolute improvement in hand function 6% (2 to 10%) on a 0 90 scale Relative change 32% (10 to 52%) Effect size 0.02 points lower Short-term pinch strength (0.35 lower to 0.31 higher) Follow up: mean 37,5 days MODERATEe Number of participants: 142 (2 RCTs) Effect size 0.18 points lower Long-term pinch strength (0.52 lower to 0.16 higher) Follow up: mean 7,5 MODERATEe months Number of participants: 136 (2 RCTs) GRADE: Grading of Recommendations Assessment, Development, and Evaluation RCT: Randomized controlled trial GRADE Working Group grades of evidence High quality: We are very confident that the true effect lies close to that of the estimate of the effect Moderate quality: We are moderately confident in the effect estimate: The true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different Low quality: Our confidence in the effect estimate is limited: The true effect may be substantially different from the estimate of the effect Very low quality: We have very little confidence in the effect estimate: The true effect is likely to be substantially different from the estimate of effect * a b c d e P < 0.05 Control group baseline pain mean (SD) 47.7 (19.8) from Rannou et al. (2009). Downgraded for inconsistency, and imprecision. Downgraded for inconsistency, and imprecision. Control group baseline pain mean(SD) 17.7 (12.9) from Rannou et al. (2009). Downgraded for imprecision. ARTICLE IN PRESS S.M. Meireles et al. / Seminars in Arthritis and Rheumatism 00 (2018) 1 13 presenting similarities in the design, assessment and follow-up time. Kjeken et al. [22] performed a similar meta-analysis to those presented in the current paper and obtained similar results, with the exception of the function in the long term, which did not present evidence of improvement. This difference occurred because we had one additional article for this analysis. In 2015, a systematic review of conservative treatment for the base of the thumb in patients with OA was published by Spaans et al. [14] These authors employed a search strategy different from the current strategy. They searched only in the MEDLINE and EMbase databases, whereas we also included the CENTRAL, PubMed, LILACS, OTseeker and PEDro bases. This is a recommendation of the Cochrane collaboration, and we sought to conduct a more comprehensive search. In contrast to our study, the research by Spaans et al. [14] did not employ the guidelines given by the Cochrane collaboration. These authors described the diversity in the study design and types of orthosis identified, which suggests that orthosis presents evidence of decreasing pain in patients with rhizarthrosis, with no impact on function, strength and dexterity. We agree with the finding that there is evidence of a decrease in pain after the use of orthosis. Nevertheless, with respect to function, we obtained different results. This discrepancy may be explained by the fact that we performed a metaanalysis of the data, whereas Spaans et al. [14] solely conducted a qualitative analysis of the studies. The systematic review by Spaans et al. [14] is the most recent and complete in the literature to date. It presents inclusion criteria similar to our research and similar results; however, it does not present a quantitative analysis of the data or assessments of the risk of bias in the studies. In 2015, Bertozzi et al. [15] published a systematic review and meta-analysis regarding conservative interventions for OA of the thumb carpometacarpal joint. These authors included three studies on orthosis, including Kjeken et al. [37], Carreira et al. [35] and Rannou et al. [38], and considered, as in the present review, that they presented a low risk of bias, as the most adequate in the literature, in assessing the effectivity of the orthosis in rhizarthrosis. They verified DATABASE LILACS and IBECS 18 ? records PubMed/ MEDLINE ? 375 ? records 11 that orthosis improved the function of patients in the long term and considered this moderate evidence. As in the present study, they used GRADE to assess the quality of the evidence. The conclusions of Bertozzi et al. [15] are somewhat different from our conclusions. The results also show that the patients improved function in the long term with a moderate quality of evidence. However, we verified that pain decreased in the long term with the use of orthosis. For the meta-analysis, we used the same works as Bertozzi et al. [15]; however, we believe that the assessment tools used in the quantitative analysis by these authors differed from our tools. All published reviews suggest that additional randomized controlled studies should be conducted, an opinion we agree with given that the quality of the evidence is related, among other variables, to the number of subjects who participate in the studies. Conclusion Fourteen studies on orthosis and rhizarthrosis were included in this review. The identified studies were heterogeneous regarding the study design, type of orthosis, follow-up time and assessments. Based on a numerical analysis, we concluded that orthosis for rhizarthrosis presents low-quality evidence for pain reduction in the long term and moderate evidence for an increase in function in the long term. No side-effect was described regarding the use of orthosis for rhizarthrosis. We therefore consider it a safe intervention. Declarations of interest None. Appendix 1. Search strategies in databases and results #1 Mh:Osteoartrite OR Osteoartritis OR Osteoarthritis OR (Artrite Degenerativa) OR Osteoartrose OR (Osteoartrose Deformante) OR Artrose OR MH:C05.550.114.606$ OR MH:C05.799.613$ OR rhizarthros$ OR rizartrose ~o" OR (Articulaciones de la Mano) OR (Hand Joints) OR MH:A02.835.583.405$ OR MH:"Articulaç o ~ es Carpometacarpais" #2 Mh:"Articulaç ~ao da Ma ~o carpo-metacarpica) OR OR (Articulaciones Carpometacarpianas) OR (Carpometacarpal Joints) OR MH:A02.835.583.405.200$ OR (articulaç a (Carpometacarpal Joints) OR (Articulation trapezo-metacarpienne) OR (Trapeziometacarpal joint) cnicas)) OR (Te cnicas #3 Mh:"Modalidades de Fisioterapia" OR (Modalidade$ de Fisioterapia) OR (Physical Therapy Modalities) OR (Fisioterapia (Te picas) OR MH:E02.779$ OR MH:Reabilitaç a ~o OR Rehabilitacio n OR Rehabilitation OR Habilitaç ~ao OR MH:E02.831$ OR MH: Fisiotera H02.403.680.600$ OR MH:N02.421.784$ OR MH:SP4.046.442.633.869.155$ OR MH:SP8.946.117.208$ OR MH:VS4.002.001.002.003$ OR MH: ^utica OR TERAPEUTICA$ OR Exercício OR EXERCICIO$ OR Ejercicio OR Exercise OR MH:G11.427.590.530.698.277$ OR MH:I03.350$ OR MH:Terape utica OR Therapeutics OR Terapia$ OR Mh:E02$ OR Mh:VS3.003.001.006$ OR splint$ OR electrotherapy OR massage$ OR shortwave OR Terape tese$ or orthosi$ or Brace$ or "Orthotic Devices" or (Ortope dic$ and (Aparato$ or aparelh$)) or Ortese$ rehabilitation or orto #4 ((Pt:"randomized controlled trial" OR Pt:"controlled clinical trial" OR Mh:"randomized controlled trials as topic" OR Mh:"random allocation" OR Mh:"double-blind method" OR Mh:"single-blind method") AND NOT (Ct:animals AND NOT (Ct:humans AND Ct:animals))) OR ((Pt:"clinical trial" OR Mh:E05.318.760.535$ OR (clin$ AND (trial$ OR ensa$ OR estud$ OR experim$ OR investiga$)) OR ((singl$ OR simple$ OR doubl$ OR doble$ OR duplo$ OR trebl$ OR trip$) AND (blind$ OR cego$ OR ciego$ OR mask$ OR mascar$)) OR Mh:placebos OR placebo$ OR (random$ OR randon$ OR casual$ OR acaso$ OR azar OR aleator$) OR Mh:"research design") AND NOT (Ct:animals AND NOT (Ct:humans and Ct:animals))) OR ((Pt:"comparative study" OR Mh:E05.337$ OR Mh:"follow-up studies" OR Mh:"prospective studies" OR control$ OR prospectiv$ OR volunt$ OR volunteer$) AND NOT (Ct:animals AND NOT (Ct:humans and Ct:animals))) #1 AND #2 AND #3 AND #4 #1 "Osteoarthritis"[Mesh] OR Osteoarthritides OR Osteoarthros* OR (Arthritis, Degenerative) OR (Arthritides, Degenerative) OR (Degenerative Arthriti*) OR (Osteoarthrosis Deformans) OR rhizarthros* #2 "Hand Joints"[Mesh] OR "Carpometacarpal Joints"[Mesh] OR (Carpometacarpal Joint*) OR (Joint, Carpometacarpal) OR (Joints, Carpometacarpal) OR (Joints, Hand) OR (Intermetacarpal Joint*) OR (Joint*, Intermetacarpal) OR (trapeziometacarpal joint*) #3 "Physical Therapy Modalities"[Mesh] OR "Rehabilitation"[Mesh] OR Rehabilitation OR "Therapeutics"[Mesh] OR Therapeutic* OR Physical Therapy Modalit* OR (Physiotherap* (Techniques)) OR Physical Therapy Technique* OR splint* OR orthotic devices OR orthosis OR orthoses OR devices, orthotic OR device, orthotic OR Brace* #4 (randomized controlled trial [pt] OR controlled clinical trial [pt] OR randomized controlled trials [mh] OR random allocation [mh] OR doubleblind method [mh] OR single-blind method [mh] OR clinical trial [pt] OR clinical trials [mh] OR ("clinical trial" [tw]) OR ((singl* [tw] OR doubl* [tw] OR trebl* [tw] OR tripl* [tw]) AND (mask* [tw] OR blind* [tw])) OR ( placebos [mh] OR placebo* [tw] OR random* [tw] OR research design [mh: noexp] OR comparative study [mh] OR evaluation studies [mh] OR follow-up studies [mh] OR prospective studies [mh] OR control* [tw] OR prospectiv* [tw] OR volunteer* [tw]) NOT (animals [mh] NOT humans [mh]) #1 AND #2 AND #3 AND #4 (continued) ARTICLE IN PRESS 12 S.M. Meireles et al. / Seminars in Arthritis and Rheumatism 00 (2018) 1 13 EMBase 100 ? records #1 'osteoarthritis'/exp OR (arthritis, degenerative) OR (arthritis, noninflammatory) OR arthrosis OR (degenerative arthritis) OR (degenerative joint disease) OR (osteo-arthritis) OR (osteo-arthrosis) OR osteoarthrosis OR (primary osteoarthritis) OR (rheumatoid arthrosis) OR rhizarthros* #2 'hand joint'/exp OR (hand joint*) OR (joint, hand) OR 'carpometacarpal joint'/exp OR (carpo metacarpal joint) OR (carpometacarpal joint*) OR (joint, carpometacarpal) OR (metacarpocarpal joint i) OR (trapeziometacarpal joint*) #3 'physiotherapy'/exp OR 'rehabilitation'/exp OR 'therapy'/exp OR Physical Therapy Technique* OR 'orthosis'/exp OR 'orthosis' OR 'splint'/exp OR 'splint' #4 (randomized controlled trial) OR (controlled clinical trial) OR (randomized) OR (placebo) OR (drug therapy) OR (randomly) OR (trial) OR random* OR factorial* OR crossover* OR cross AND over* OR placebo* OR 'placebo'/exp OR doubl* AND blind* OR singl* AND blind* OR assign* AND allocate* OR volunteer* OR 'volunteer'/exp OR crossover AND procedure OR double AND 'blind'/exp AND procedure OR randomized AND controlled AND trial OR single AND 'blind'/exp AND procedure OR (controlled AND clinical AND trial) OR (clinical AND trial) #1 AND #2 AND #3 AND #4 CENTRAL #1 "Osteoarthritis" OR Osteoarthritides OR Osteoarthros* OR (Arthritis, Degenerative) OR (Arthritides, Degenerative) OR (Degenerative Arthriti*) 77 ? records OR (Osteoarthrosis Deformans) OR rhizarthros* #2 "Hand Joints" OR "Carpometacarpal Joints" OR (Carpometacarpal Joint*) OR (Joint, Carpometacarpal) OR (Joints, Carpometacarpal) OR (Joints, Hand) OR (Intermetacarpal Joint*) OR (Joint*, Intermetacarpal) OR (trapeziometacarpal joint*) #3 "Physical Therapy Modalities" OR "Rehabilitation" OR Rehabilitation OR "Therapeutics" OR Therapeutic* OR Physical Therapy Modalit* OR (Physiotherap* (Techniques)) OR Physical Therapy Technique* OR splint* OR orthotic devices OR orthosis OR orthoses OR devices, orthotic OR device, orthotic OR Brace* #1 AND #2 AND #3 FILTER: trials PEDro Abstract & Title: osteoarthritis 18 ? records Therapy: orthosis, taping, splinting Body Part: hand or wrist Method: clinical trial OTseeker Abstract & Title: osteoarthritis 17 ? records Intervention: physical modalities/orthotics/splinting Method: Randomized controlled trial ^ncias da Sau de; IBECS: Indice Bibliografico Espanhol de Cie ^ncias da Sau de; MEDLINE: Medical Literature Analysis and LILACS: Literatura Latino-Americana e do Caribe em Cie Retrieval System Online; EMBase: Excerpta Medica Database; CENTRAL: Cochrane Central Register of Controlled Trials; PEDro: Physiotherapy Evidence Database; Otseeker: Occupational Therapy Systematic Evaluation of Evidence. References [1] Dieppe P. Osteoarthritis: a clinical features. In: Klippel JH, ed. Primer on the rheumatic diseases, thirteenth ed, New York (NY): Springer; 2008. [2] Nelson AE, Jordan JM. Osteoarthritis: epidemiology and classification. In: Hochberg MC, ed. Rheumatology, sixth ed, Philadelphia (PA): Mosby; 2015. [3] Siviero P, Zambon S, Limongi F, et al. How hand osteoarthritis, comorbidity and pain interact to determine functional limitation in older people: observations from the EPOSA study. Arthritis Rheumatol 2016;68(11):2662–70. [4] Flores RH, Hochberg MC. Definition and classification of osteoarthritis. In: Brandt KD, Doherty M, Lohmander LS, eds. Osteoarthritis, second ed., New York (NY): Oxford; 2003. [5] Hochberg MC, Altman RD, April KT, et al. American College of Rheumatology 2012 recommendations for the use of nonpharmacologic and pharmacologic therapies in osteoarthritis of the hand, hip, and knee. Arthritis Care Res 2012;64(4):465–74. [6] Zhang W, Doherty M, Leeb BF, et al. EULAR evidence based recommendations for the management of hand osteoarthritis: report of a task force of the EULAR Standing Committee for International Clinical Studies Including Therapeutics (ESCISIT). Ann Rheum Dis 2007;66(3):377–88. [7] Rillo O, Riera H, Acosta C, et al. PANLAR Consensus recommendations for the management in osteoarthritis of hand, hip, and knee. J Clin Rheumatol 2016;22 (7):345–54. [8] Kloppenburg M. Hand osteoarthritis-nonpharmacological and pharmacological treatments. Nat Rev Rheumatol 2014;10(4):242–51. [9] Colditz JC. The Biomechanics of a thumb carpometacarpal immobilization splint: desing and fitting. J Hand Ther 2000;13(3):228–35. [10] Melvin JL. Rheumatic disease in the adult and child: occupational therapy and rehabilitation. third ed. Philadelphia (PA): F.A. Davis; 1989. [11] Fess EE, Genttle K, Philips C, Janson R. Hand and upper extremity splinting: principles and methods. third ed. St Louis (MO): Mosby; 2005. € ld U, Lundgren Nilsson A. Effects of a hand-joint protec[12] Boustedt C, Nordenskio tion programme with an addition of splinting and exercise: one year follow-up. Clin Rheumatol 2009;28(7):793–9. [13] Jacobs MS. Splint classification. In: Jacobs MS, Austin N, eds. Splinting the hand and upper extremity: principles and process, Philadelphia (PA): Lippincott Williams Wilkins; 2003. [14] Spaans AJ, van Minnen LP, Kon M, Schuurman AH, Schreuders AR, Vermeulen GM. Conservative treatment of thumb base osteoarthritis: a systematic review. J Hand Surg Am 2015;40(1):16–21. ~ e JH. Investigation of [15] Bertozzi L, Valdes K, Vanti C, Negrini S, Pillastrini P, Villafan the effect of conservative interventions in thumb carpometacarpal osteoarthritis: systematic review and meta-analysis. Disabil Rehabil 2015;37(22):2025–43. [16] Towheed TE. Systematic review of therapies for osteoarthritis of the hand. Osteoarthrit Cartil 2005;13(6):455–62. [17] Mahendira D, Towheed TE. Systematic review of non-surgical therapies for osteoarthritis of the hand: an update. Osteoarthrit Cartil 2009;17(10):1263–8. [18] Egan MY, Brousseau L. Splinting for osteoarthritis of the carpometacarpal joint: a review of the evidence. Am J Occup Ther 2007;61(1):70–8. [19] Moe RH, Kjeken I, Uhlig T, Hagen KB. There is inadequate evidence to determine the effectiveness of nonpharmacological and nonsurgical interventions for hand osteoarthritis: an overview of high-quality systematic reviews. Phys Ther 2009;89(12):1363–70. [20] Valdes K, Marik T. a systematic review of conservative interventions for osteoarthritis of the hand. J Hand Ther 2010;23(4):334–51. [21] Ye L, Kalichman L, Spittle A, Dobson F, Bennell K. Effects of rehabilitative interventions on pain, function, and physical impairments in people with hand osteoarthritis: a systematic review. Arthritis Res Ther 2011;13(1):R28. [22] Kjeken I, Smedslund G, Moe R, et al. Systematic review of desing and effects of splints and exercise programs in hand osteoarthritis. Arthritis Care Res 2011;63 (6):834–48. [23] Bennell KL, Hall M, Hinman RS. Osteoarthritis year in review 2015: rehabilitation and outcomes. Osteoarthrit Cartil 2016;24(1):58–70. [24] Mejjad O, Maheu E. Therapeutic trials in hand osteoarthritis: a critical review. Osteoarthrit Cartil 2000;8(Suppl A):S57–63. [25] Higgins J.P.T., Green S. Cochrane handbook for systematic reviews of interventions. Version 5.1.0 [updated March 2011]. The Cochrane Collaboration; 2011. Disponível em: http://handbook.cochrane.org/. Data de acesso: 11 de novembro de 2016. [26] Kennedy CD, Manske MC, Huang JI. Classifications in brief: the eaton-littler classification of thumb carpometacarpal joint arthrosis. Clin Orthop Relat Res 2016;474(12):2729–33. [27] Verhagen AP, de Vet HC, de Bie RA, Kessels AG, Boers M, Bouter LM, Knipschild PG. The Delphi list: a criteria list for quality assessment of randomized clinical trials for conducting systematic reviews developed by Delphi consensus. J Clin Epidemiol 1998;51:1235–41. [28] Cohen J. Statistical power for the behavioral sciences. second ed Hillsdale, NJ: Lawrence Erlbaum Associates; 1988. [29] Adams J, Bouc JB, Hislop K, Dziedzic K, et al. The effectiveness and efficacy of splints for thumb base osteoarthritis: a pilot randomized controlled trial. Rheumatology 2014;53(suppl 1):i41–2. [30] Arazpour M, Soflaei M, Ahmadi Bani M, Madani SP, Sattari M, Biglarian A, Mosallanezhad Z. The effect of thumb splinting on thenar muscles atrophy, pain, and function in subjects with thumb carpometacarpal joint osteoarthritis. Prosthet Orthot Int 2017;41(4):379–86. [31] Bani MA, Arazpour M, Kashani RV, Mousavi ME, Hutchins SW. Comparison of custom-made and prefabricated neoprene splinting in patients with the first carpometacarpal joint osteoarthritis. Disabil Rehabil Assist Technol 2013;8 (3):232–7. [32] Becker SJ, Bot AG, Curley SE, Jupiter JB, Ring D. A prospective randomized comparison of neoprene vs thermoplast hand-based thumb spica splinting for trapeziometacarpal arthrosis. Osteoarthrit Cartil 2013;21(5):668–75. [33] Berggren M, Joost-Davidsson A, Lindstrand J, Nylander G, Povlsen B. Reduction in the need for operation after conservative treatment of osteoarthritis of the first carpometacarpal joint: a seven year prospective study. Scand J Plast Reconstr Surg Hand Surg 2001;35(4):415–7. [34] Buurke JH, Grady JH, de Vries J, Baten CT. Usability of thenar eminence orthoses: report of a comparative study. Clin Rehabil 1999;13(4):288–94. [35] Carreira ACG, Jones A, Natour J. Assessment of the effectiveness of a functional splint for osteoarthritis of the trapeziometacarpal joint of the dominant hand: a randomized controlled study. J Rehabil Med 2010;42(5):469–74. [36] Hermann M, Nilsen T, Eriksen CS, Slatkowsky-Christensen B, Haugen IK, Kjeken I. Effects of a soft prefabricated thumb orthosis in carpometacarpal osteoarthritis. Scand J Occup Ther 2014;21(1):31–9. [37] Kjeken I, Darre S, Smedslund G, Hagen KB, Nossum R. Effect of assistive technology in hand osteoarthritis: a randomised controlled trial. Ann Rheum Dis 2011;70(8):1447–52. ARTICLE IN PRESS S.M. Meireles et al. / Seminars in Arthritis and Rheumatism 00 (2018) 1 13 [38] Rannou F, Dimet J, Boutron I, et al. Splint for base-of-thumb osteoarthritis. a randomized trial. Ann Internal Med 2009;150(10):661–9. [39] Sillem H, Backman CL, Miller WC, Li LC. Comparison of two carpometacarpal stabilizing splints for individuals with thumb osteoarthritis. J Hand Ther 2011;24 (3):216–25. [40] Wajon A, Ada L. No difference between two splint and exercise regimens for people with osteoarthritis of the thumb: a randomised controlled trial. Aust J Physiother 2005;51(4):245–9. 13 [41] Weiss S, Lastayo P, Mills A, Bramlet D. Prospective analysis of splinting the first carpometacarpal joint: an objective, subjective, and radiographic assessment. J Hand Ther 2000;13(3):218–26. [42] Weiss S, Lastayo P, Mills A, Bramlet D. Splinting the degenerative basal joint: custom-made or prefabricated neoprene? J Hand Ther 2004;17(4):401–6. [43] McKee P, Eason-Klatt M. A multi-center study comparing two styles of orthoses for individuals with thumb carpometacarpal osteoarthritis. J Hand Ther 2006;19 (4):446–7.
