Advertisement

Outcome of a Hand Orthosis and Hand Therapy for Carpometacarpal Osteoarthritis in Daily Practice: A Prospective Cohort Study

      Purpose

      Initial treatment for symptomatic carpometacarpal (CMC) osteoarthritis (OA) of the thumb is usually nonsurgical. However, evidence on the effect of a hand orthosis and hand therapy for mid- and long-term results is limited, and it is unknown how many patients undergo additional surgical treatment. Therefore, the aim of this study is to describe the outcome of a hand orthosis and hand therapy for CMC OA in a large cohort study, and to evaluate the conversion rate to surgical treatment.

      Methods

      In this multicenter, prospective cohort study, patients treated with a hand orthosis and hand therapy for primary CMC OA between 2011 and 2014 were included. Pain (visual analog scale) and function (Michigan Hand Questionnaire) were measured at baseline, 6 weeks, 3 months, and at 12 months after the start of treatment. All patients converted to surgery were recorded between 2012 and 2016. Outcome was compared with baseline, and post hoc comparisons were made between patients who were not converted to surgery and patients who were converted to surgery after initially receiving a hand orthosis and hand therapy. Lastly, subgroup analysis was performed based on baseline pain levels.

      Results

      After a mean follow-up of 2.2 ± 0.9 years, 15% of all patients were surgically treated. In the group that was not converted to surgery, pain (visual analog scale) significantly improved from 49 ± 20 at baseline to 36 ± 24 at 12 months. The Michigan Hand Questionnaire score was essentially unchanged from 65 ± 15 at baseline to 69 ± 10 at 12 months. Post hoc testing showed that improvement in pain was only significant between baseline and 6 weeks, and thereafter stabilized until 1 year after the start of treatment. The group that converted to surgery did not show any improvement in pain and function at follow-up.

      Conclusions

      In this cohort of patients with thumb CMC OA who underwent hand therapy including an orthosis, 15% of the patients underwent additional surgical treatment. The patients (85%) who did not undergo surgery improved in pain and function, although only improvements in pain were significant and clinically relevant. Most improvement was seen in the first 6 weeks and stabilized till 1 year after the start of treatment.

      Type of study/level of evidence

      Therapeutic II.

      Key words

      Treatment guidelines for carpometacarpal (CMC) osteoarthritis (OA) of the thumb usually advise a period of nonsurgical treatment before considering surgical treatment for all patients with primary CMC OA.
      • van Uchelen J.
      • Beumer A.
      • Brink S.M.
      • Hoogvliet P.
      • Moojen T.M.
      • Spaans A.J.
      Dutch Guideline for Conservative and Surgical Treatment for Primary Thumb Base Osteoarthritis.
      • Hochberg M.C.
      • Altman R.D.
      • April K.T.
      • et al.
      American College of Rheumatology 2012 recommendations for the use of nonpharmacologic and pharmacologic therapies in osteoarthritis of the hand, hip, and knee.
      National Clinical Guideline Centre (UK)
      Osteoarthritis: care and management in adults.
      • Zhang W.
      • Doherty M.
      • Leeb B.F.
      • 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).
      Nonsurgical treatment for CMC OA can consist of orthosis immobilization, intra-articular steroid injections, hand therapy, or a combination of modalities.
      • Stamm T.A.
      • Machold K.P.
      • Smolen J.S.
      • et al.
      Joint protection and home hand exercises improve hand function in patients with hand osteoarthritis: a randomized controlled trial.
      • Joshi R.
      Intraarticular corticosteroid injection for first carpometacarpal osteoarthritis.
      • Colditz J.C.
      The biomechanics of a thumb carpometacarpal immobilization splint: design and fitting.
      When nonsurgical treatment fails to provide enough pain relief or functional improvement in daily life, a decision may be made to proceed to surgical treatment.
      However, the existing evidence on the effectiveness of nonsurgical treatment is of poor quality, primarily due to small sample sizes, nongeneralizable study samples, or short follow-up time. In addition, most of these studies are limited to only comparing different types of orthoses and not the effect of combination therapy, that is, an orthosis with hand therapy.
      • Egan M.Y.
      • Brousseau L.
      Splinting for osteoarthritis of the carpometacarpal joint: a review of the evidence.
      • Spaans A.J.
      • van Minnen L.P.
      • Kon M.
      • Schuurman A.H.
      • Schreuders A.R.
      • Vermeulen G.M.
      Conservative treatment of thumb base osteoarthritis: a systematic review.
      • Aebischer B.
      • Elsig S.
      • Taeymans J.
      Effectiveness of physical and occupational therapy on pain, function and quality of life in patients with trapeziometacarpal osteoarthritis—a systematic review and meta-analysis.
      For example, a systematic review by Egan and Brousseau
      • Egan M.Y.
      • Brousseau L.
      Splinting for osteoarthritis of the carpometacarpal joint: a review of the evidence.
      showed that hand orthoses may help to relieve pain, but sample sizes of the included studies were very small (N = 10–37) and follow-up times relatively short (1 wk to 6 mo). In addition, hand function was not measured. Another systematic review on comparative studies of hand orthoses or hand therapy of CMC OA
      • Spaans A.J.
      • van Minnen L.P.
      • Kon M.
      • Schuurman A.H.
      • Schreuders A.R.
      • Vermeulen G.M.
      Conservative treatment of thumb base osteoarthritis: a systematic review.
      concluded that a hand orthosis or hand therapy may provide pain reduction, but all studies had a short follow-up time (2 wk to 3 mo) and study samples comprising only older individuals (70–90 y). In addition, none of the studies evaluated outcome after a combination of a hand orthosis and hand therapy.
      • van Uchelen J.
      • Beumer A.
      • Brink S.M.
      • Hoogvliet P.
      • Moojen T.M.
      • Spaans A.J.
      Dutch Guideline for Conservative and Surgical Treatment for Primary Thumb Base Osteoarthritis.
      • Hochberg M.C.
      • Altman R.D.
      • April K.T.
      • et al.
      American College of Rheumatology 2012 recommendations for the use of nonpharmacologic and pharmacologic therapies in osteoarthritis of the hand, hip, and knee.
      • Melvin J.L.
      Therapist's Management of the Thumb Carpometacarpal Joint with Osteoarthritis.
      • van der Giesen F.J.
      • Nelissen R.G.
      • Arendzen J.H.
      • de Jong Z.
      • Wolterbeek R.
      Vliet Vlieland TP. Responsiveness of the Michigan Hand Outcomes Questionnaire-Dutch language version in patients with rheumatoid arthritis.
      A recent meta-analysis of Aebischer et al
      • Aebischer B.
      • Elsig S.
      • Taeymans J.
      Effectiveness of physical and occupational therapy on pain, function and quality of life in patients with trapeziometacarpal osteoarthritis—a systematic review and meta-analysis.
      based on studies on hand orthoses, hand therapy, and nonpharmaceutical treatment for CMC OA concluded that combination therapy is more effective for pain than single interventions. Because of the paucity of available evidence, it is unknown how many patients respond favorably, and how many patients who initially received a hand orthosis and hand therapy are eventually converted to surgical treatment. In addition, the timing of surgical intervention, in relation to receiving a hand orthosis and hand therapy (eg, how long should surgery be delayed if patients do not respond to a hand orthosis and hand therapy), is unknown as well. Therefore, the aim of this study was to describe the long-term outcome of providing a hand orthosis and hand therapy for thumb CMC OA, and to identify when and how many patients need additional surgical treatment.

      Methods

      Study population

      This study was conducted as an observational, prospective cohort study, performed in a private hand surgery clinic (Xpert Clinic, the Netherlands), consisting of 11 locations, with 13 European board-certified hand surgeons delivering care. Hand therapy was given by more than 50 hand therapists at specialized hand therapy clinics, located in or near an Xpert clinic (Handtherapie Nederland, the Netherlands).
      All patients evaluated at the outpatient clinic between January 2011 and November 2014, clinically diagnosed with CMC OA and treated with a hand orthosis and hand therapy, were asked to participate in this study. All patients received an x-ray of their hand to confirm the clinical diagnosis and to grade the severity of CMC OA. However, because the grading of the osteoarthritis was done in a nonsystematic way, we did not further analyze the CMC OA severity based on the x-ray images. The study was approved by the local institutional review board, and all patients signed an informed consent. Exclusion criteria were previous CMC surgery, post-traumatic OA, isolated scaphotrapeziotrapezoid OA on the x-ray, or a history of prior intra-articular corticosteroid injections in the thumb CMC joint. Furthermore, patients with active trigger finger, carpal tunnel syndrome, OA of the interphalangeal joints, or de Quervain tendonitis were excluded when they received simultaneous treatment for these conditions at the start of treatment.

      Intervention

      Treatment was based on the Dutch treatment guideline.
      • van Uchelen J.
      • Beumer A.
      • Brink S.M.
      • Hoogvliet P.
      • Moojen T.M.
      • Spaans A.J.
      Dutch Guideline for Conservative and Surgical Treatment for Primary Thumb Base Osteoarthritis.
      In general, treatment consisted of prescribing a custom-made or prefabricated orthosis (based on the preference of the surgeon, hand therapists, and insurance of the patient) and 2 sessions of hand therapy per week of an average duration of 25 minutes per session. The hand therapists all received the same internal training on how to treat CMC OA with hand therapy. However, this was a pragmatic study in the sense that the hand therapy was not strictly protocolled and controlled, but evaluated, based on clinical practice. Therapy sessions were planned by judgment of the therapist and ability and availability of the patient. In some cases, patients received only a hand orthosis without further treatment, for example because of their insurance or schedules. The treatment was divided into 2 phases: phase I (week 0–6) included instructions to wear the hand orthosis almost 24 hours per day and consisted of hand therapy for optimizing thumb position (training pinch and grasping movements without hyperextension in the metacarpophalangeal thumb joint and without CMC adduction) and using a full thumb range of motion (where the specific coordination of the intrinsic and extrinsic muscles of the thumb is trained); in phase II (week 7–12), the hand orthosis was slowly phased out: the patient was advised to use the hand orthosis only during heavy activities, depending on pain level and the patient’s ability to perform activities with a stable thumb position. The hand therapy during this phase focused on maintaining the pain reduction, introducing the learned stability during daily activities and improving thenar muscle strength. In this phase, fewer hand therapy sessions were scheduled and patients performed more home exercises, up to 4–6 times a day. The number of prescribed home exercises ranged between 3 and 6 exercises per day, with 10–15 repetitions each, depending on the individual patient and the level of pain. After this period of supervised therapy, patients were encouraged to keep doing the exercises, and patients were allowed to use the hand orthosis when necessary. No corticosteroid injections were given for their CMC OA during or after hand therapy, and no anti-inflammatory medication was prescribed by the surgeon.

      Measures

      Baseline demographics of all patients, including duration of complaints, comorbidity, and hand medical history, were collected before the start of treatment. Outcome measures were recorded before the start of the treatment, at 6 weeks, at 3 months, and at 12 months through our web-based outcome registration. All patients had a follow-up appointment with their hand surgeon after approximately 3 months, during which progress was evaluated.

      Pain and function

      Pain was measured with a visual analog scale (VAS) during 2 situations: pain during activities and pain experienced during the last week. To measure patient-rated hand function, the Michigan Hand Questionnaire was used (MHQ, Dutch Language Version; 0 = poorest function, 100 = ideal function).
      • Chung K.C.
      • Pillsbury M.S.
      • Walter M.R.
      • Hayward H.R.
      Reliability and validity testing of the Michigan Hand Outcome Questionnaire.
      • Marks M.
      • Audige L.
      • Herren D.B.
      • Schindele S.
      • Nelissen R.G.H.H.
      • Vliet Vlieland T.P.
      Measurement properties of the German Michigan Hand Outcomes Questionnaire in patients with trapeziometacarpal osteoarthritis.
      • London D.A.
      • Stepan J.G.
      • Calfee R.P.
      Determining the Michigan Hand Outcomes Questionnaire minimal clinically important difference by means of three methods.
      The MHQ is a self-reported questionnaire with 6 domains and 37 items. The Minimal Clinically Important Difference ranges between 9 and 13 points for total MHQ and between 11 and 14 points for the subdomain pain, for nontraumatic hand conditions.
      • Cohen J.
      Statistical Power Analysis for the Behavioral Sciences.

      Surgery

      All patients had a follow-up appointment with their hand surgeon after approximately 3 months; further follow-up was only scheduled when indicated. Surgical intervention was discussed when patients did not respond well to the hand orthosis and hand therapy and had functional impairments and/or residual pain. Together with the surgeon, the decision to operate was made based on the symptoms of the patient. All surgeries performed between January 2012 and February 2016, together with time until surgery, were retrieved from the clinical records, independent of whether patients responded to the questionnaires. These results were separately analyzed and not combined with the results of the questionnaires, which made it possible to report conversion to surgery on all patients eligible for inclusion (Fig. 1).
      Figure thumbnail gr1
      Figure 1Flowchart of the study. CTS, carpal tunnel syndrome; DIP, distal interphalangeal joint; PIP, proximal interphalangeal joint; Quervain, Morbus de Quervain; TVS: tendovaginitis stenosans.

      Statistical analysis

      We performed a sample size calculation to determine the number of patients required to detect a conventional effect size of 0.3 for pain (VAS) after receiving a hand orthosis and hand therapy. The required sample was 90 participants.
      • Little R.J.A.
      • Rubin D.B.
      Statistical Analysis with Missing Data.
      Baseline demographics were available in more than 98% of the patients. Because data were collected during daily clinical practice, we had a substantial proportion of nonresponse during follow-up (Fig. 1). In addition, the data that were missing at 12 months consisted of both patients who did not fill in the questionnaires and patients who had already converted to surgery. Because of this, a thorough nonresponder analysis on the whole group was performed using χ2 statistics or t tests for all variables measured at baseline based on the response at 1 year. No significant differences were found at baseline between patients who filled in the questionnaires at follow-up and patients who did not fill the questionnaires at follow-up. In addition, Little’s MCAR test

      LeeJ H, Huber J Jr. Multiple imputation with large proportions of missing data: how much is too much? United Kingdom Stata Users’ Group Meetings. London: Stata Users Group; 2011:23.

      for all separate outcome variables showed that more than 95% of the outcome variables were missing completely at random. We therefore performed all main analysis with patients who responded at all follow-up measurements (complete case analysis). As a secondary analysis, we performed multiple imputations to compare the outcome for consistency with the complete case analysis. We performed multiple imputation by chained equations by fully conditional specification and used all patients. We imputed 10 times and compared the imputed data with the complete case data using t tests (Appendix A, available on the Journal’s Web site at www.jhandsurg.org). Here again, no significant differences were found between the imputed analysis and the complete case analysis.
      Hereafter, analysis of variance tests with repeated measures were performed to compare baseline and follow-up measurements, combined with Tukey’s post hoc tests, to determine between which follow-up points the significant difference existed. In addition, we compared patients who eventually received surgery with patients who did not convert to surgery using independent samples t tests. Because the decision to operate was made from 3 months onward, patients who eventually received surgery filled in the questionnaires only until 3 months. This allowed us to compare the group that was not operated with the group that converted to surgery up to 3 months without having to impute any data.
      To study the influence of different baseline pain levels on outcome after treatment, we divided patients into 4 subgroups based on baseline pain level (VAS), correcting for regression to the mean, which can occur if a variable is extreme on its first measurement; in that case, it will tend to be closer to the average on its second measurement.
      • Kelly C.
      • Price T.D.
      Correcting for regression to the mean in behavior and ecology.
      Corrections were made based on a test-retest reliability of 0.85 for VAS pain.
      • Visser A.W.
      • Bøyesen P.
      • Haugen I.K.
      • et al.
      Instruments measuring pain, physical function, or patient’s global assessment in hand osteoarthritis: a systematic literature search.
      For all tests, we considered a P value smaller than .05 as statistically significant.

      Results

      Study population

      Between January 2012 and November 2014, 1,033 patients with complaints of CMC OA visited Xpert Clinic, of whom 809 were eligible for inclusion. Of those patients eligible for inclusion, 122 completed all follow-up measurements without undergoing surgery and were used to analyze the primary outcomes: pain and function. In addition, 28 patients who underwent surgery completed all follow-up measurements until 3 months. Figure 1 shows the flowchart and Table 1 shows the baseline characteristics. To study conversion to surgery we included all 809 patients, because the recording of conversion to surgery was independent of the response of the patients to the questionnaires.
      Table 1Baseline Characteristics
      Baseline CharacteristicsTotal (n = 122)
      VariablesNPercentage or Mean ± SD
      Sex
       Female9172
      Treated hand
       Right6350
      Workload
       No work6753
       Light physical work2218
       Moderate physical work3024
       Heavy physical work76
      Dominance
       Left1411
       Right10785
       Both54
      Age (y)60 ± 8
      Duration symptoms (wk)40 ± 72
      SD, standard deviation.
      Patients were followed for a minimum of 1.5 years to verify whether they had undergone surgical intervention. After a mean follow-up of 2.2 ± 0.9 (±standard deviation) years, 124 patients (15%) were surgically treated (Fig. 2). The majority of the surgically treated patients (n = 93; 75%) were operated on within the first year after the start of hand therapy and the median number of days until surgery was 160 (interquartile range, 40–280) days.
      Figure thumbnail gr2
      Figure 2Survival analysis. Chart shows the duration of time until receiving surgery. On the y axis the proportion of patients not operated is shown, and on the x axis the number of days since the start of a hand orthosis and hand therapy. A total of 15.3% converted to surgery with a median number of days until surgery of 5 months.

      Pain

      The patients who did not convert to surgery showed a significant decrease in pain during the last week, VAS from 49 ± 20 (mean ± standard deviation) at baseline to 36 ± 24 at 12 months after the start of treatment (P < .05), and showed a significant decrease in pain during activities (VAS) from 60 ± 21 to 44 ± 27 after 12 months (P < .05) (Fig. 3A). Post hoc tests showed that improvements were only significant between baseline and 6 weeks: 14.5 points improvement (95% confidence interval [CI], 7.2–21.8; P < .05) for pain during the last week and 17.6 points improvement (95% CI, 9.8–25.4; P < .05) for pain during activities. Between 6 weeks and 12 months, no significant change occurred: 1.8 points mean difference (95% CI, −9.1 to 5.5; P = .922) for pain during the last week and 1.7 points mean difference (95% CI, −9.5 to 6.1; P = .945) for pain during activities.
      Figure thumbnail gr3
      Figure 3Outcome in pain (VAS) and function (MHQ). In the group that was not operated, there was a significant improvement in pain between baseline and 12 months. Furthermore, most improvement was seen in the first 6 week. In the group that was eventually operated, there was no significant improvement between baseline and 3 months after receiving a hand orthosis and hand therapy. Error bars indicate standard errors.
      The patients who chose to convert to surgery after 3 months had at baseline (at the start of receiving a hand orthosis and hand therapy) a mean score of pain experienced during the last week of 62 ± 17, and a mean score of pain during activities of 67 ± 25. At follow-up, no significant change was seen between baseline and 3 months in pain experienced during the last week (1.6 points mean difference, 95% CI, −6.1 to 9.2; P = .677) or pain during activities (1.5 points mean difference, 95% CI, −11.0 to 8.0; P = .749).
      When comparing patients whowere converted to surgery with the patients who were not converted, we observed that the converted patients had at baseline 13.0 (95% CI, 5.0–21.0; P < .05) points higher pain experienced during the last week and 6.6 (95% CI, −2.3 to 15.5; P = .143) points higher pain during activities compared with the patients who were not converted. At 3 months the differences increased, with patients who were converted having 22.4 (95% CI, 12.9–32.0; P < .05) points higher pain experienced during the last week and 25.6 (95% CI, 15.9–35.4; P < .05) points higher pain during activities compared with the patients who were not converted.

      Function

      The patients who did not convert to surgery showed a significant change in function (total MHQ score) from 65 ± 15 at baseline to 69 ± 10 after 12 months (P < .05) (Fig. 3B). Post hoc tests showed that for function, improvement was significant between baseline and 6 weeks: 6.2 (95% CI, 1.2–11.2; P < .05) points improvement, but the improvement in function was no longer significant at 1 year after the start of treatment (mean difference ± 3.7 points; 95% CI, −0.95 to 11.2; P = .172) (Fig. 3B).
      The patients who chose to convert to surgery after 3 months had at baseline a mean function score of 58 ± 18. At follow-up, no significant improvement was seen between baseline and 3 months in function (0.2 points mean difference; 95% CI, −4.4 to 4.1; P = .939).
      When comparing patients who were converted with the patients who were not converted, we observed that the converted patients had at baseline 7.0 (95% CI, 0.2–13.7; P = .044) points less function compared with the patients who were not converted. At 3 months the differences increased, with patients who were converted having 15.0 (95% CI, 8.5–21.4; P < .05) points less function compared with the patients who were not converted.

      Subgroup analysis

      When grouping patients who were not converted to surgery based on the severity of baseline pain (Fig. 4), pain only improved significantly in groups where pain at baseline was higher than 50 on average (VAS); the higher the average pain level at baseline, the higher the reduction in pain. In contrast, patients with a baseline level of 25 or lower (VAS) showed a significant increase in pain after a hand orthosis and hand therapy.
      Figure thumbnail gr4
      Figure 4Subgroup analysis for pain during last week (VAS) and pain during activities (VAS) based on baseline pain levels. The figure shows the outcome of treatment on subgroups. Patients with high baseline pain improved in outcome, whereas patients with low baseline pain deteriorated in outcome. Patients with high baseline pain converted relatively more often to surgery compared with patients with low baseline pain. Error bars indicate standard errors.

      Discussion

      In this prospective cohort study using data collected as part of routine clinical care, we found that after a mean follow-up of 2.2 years, 15% of the patients treated with a hand orthosis and hand therapy underwent surgical treatment, after a median duration of 5 months from receiving a hand orthosis and hand therapy. When we divided patients who eventually did or did not convert to surgery, we found that the group that was not converted to surgery showed significant improvement in pain within 1 year after being treated with a hand orthosis and hand therapy for CMC OA. Most of this improvement was gained in the first 6 weeks of treatment, whereafter improvements were maintained. In addition, we saw that both pain and functional outcome were worse in the group that eventually received surgery, both at baseline and at the follow-up measurements. In the group that was converted to surgery, no improvement in pain and function was seen at follow-up measurements. Subgroup analysis, based on baseline pain levels, showed that patients with mean baseline pain levels of 50 or higher had a significant reduction in the amount of pain experienced, whereas patients with mean baseline pain levels of 25 or lower had a significant increase in pain.
      The improvements on a group level in pain and function after a hand orthosis and hand therapy are in line with the limited available evidence. For example, Villafañe et al
      • Villafañe J.H.
      • Cleland J.A.
      • Fernández-de-Las-Peñas C.
      The effectiveness of a manual therapy and exercise protocol in patients with thumb carpometacarpal osteoarthritis: a randomized controlled trial.
      randomized 60 patients with CMC OA to manual therapy or a placebo intervention and found that the manual therapy group had a significant pain reduction after 1 month, whereas the placebo intervention did not reduce pain. Between 1 and 2 months after the start of manual therapy, pain did not change in this study, which is in line with our finding that pain reduced mostly within the first 6 weeks, although manual therapy in their study had a very different treatment protocol compared with our study, including passive nerve mobilization and joint mobilization. Similarly, the small retrospective study of O’Brien and Giveans
      • O’Brien V.H.
      • Giveans M.R.
      Effects of a dynamic stability approach in conservative intervention of the carpometacarpal joint of the thumb: a retrospective study.
      described that, within 90 days, hand therapy significantly reduced pain from 3.3 to 2.7 on a 1–5 Likert scale.
      It should be noted that not all significant improvements in this study were clinically relevant. In our study, we found a clinically relevant improvement of 12 on the MHQ subdomain pain.
      • Cohen J.
      Statistical Power Analysis for the Behavioral Sciences.
      However, the improvement on the total MHQ score of 4 points did not exceed the Minimal Clinically Important Difference of 9–13 points,
      • Cohen J.
      Statistical Power Analysis for the Behavioral Sciences.
      indicating that the improvement in function may not be clinically relevant. Because Frouzakis et al
      • Frouzakis R.
      • Herren D.B.
      • Marks M.
      Evaluation of expectations and expectation fulfillment in patients treated for trapeziometacarpal osteoarthritis.
      found that pain reduction is the primary reason for patients to seek treatment, the clinically relevant pain reduction in this study supports the implementation of a hand orthosis and hand therapy in these patients.
      Although we found that only approximately 15% of our patients received additional surgical treatment, we are not aware of any other studies reporting this outcome after a hand orthosis and hand therapy. Wajon et al
      • Wajon A.
      • Vinycomb T.
      • Carr E.
      • Edmunds I.
      • Ada L.
      Surgery for thumb (trapeziometacarpal joint) osteoarthritis.
      concluded in a Cochrane review that they could not provide any information on the right time to convert to surgical treatment. Berggren et al
      • 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.
      reported that 23 of 33 patients (70%) waiting for operation could be treated successfully with hand therapy within 7 months before surgery, and within 7 years, only 2 more patients underwent additional surgical treatment. However, because patients in this study were already planned for surgery, we cannot compare this rate with our study.
      This study has a number of specific strengths and limitations. An important strength of this study is the large sample size of 122 patients. Another strength is the pragmatic nature of this study, recording how hand therapy is performed in actual clinical practice, outside of the more controlled and potentially less-natural setting of a randomized controlled trial. At the same time, the natural setting is also a limitation of the study because treatment was not completely standardized. Therapists adjusted treatment to the specific condition of the patient, severity of the complaints, time schedule, and type of insurance of the patient. Treatment in the form of purely an orthosis is very different compared with an orthosis and hand therapy. In addition, compliance with the treatment protocol by the participants was not recorded. The natural setting also resulted in the proportion of missing data, another limitation of our study. An important reason for missing data is that patients who had residual pain or functional complaints after being treated with hand therapy and an orthosis received surgical treatment and therefore were “missing” after 12 months. Another possible reason for our missing data is that patients may have gone elsewhere to receive treatment. However, because these patients visited this center seeking treatment and, as a part of protocol, first received hand therapy and orthosis, our experience is that only a very small portion of patients elect to undergo surgery elsewhere when hand therapy leads to insufficient relief of symptoms.
      Another limitation of this study is that it focuses only on the combination of an orthosis and hand therapy, but cannot conclude anything on the outcome of other treatment strategies, such as topical or oral anti-inflammatory medication or intra-articular corticosteroid injections. We did not perform radiological staging, because the Dutch guideline
      • van Uchelen J.
      • Beumer A.
      • Brink S.M.
      • Hoogvliet P.
      • Moojen T.M.
      • Spaans A.J.
      Dutch Guideline for Conservative and Surgical Treatment for Primary Thumb Base Osteoarthritis.
      for the treatment of CMC OA indicates that x-rays can support the diagnosis of CMC OA but that radiological staging according to Eaton and Glickel does not have added value, due to only fair interobserver reliability and only fair correlation with symptoms. In future research, it would be very useful to perform subgroup analyses based on radiological staging. Inherent to the cohort nature of this study is that a control group is lacking. Therefore, this study does not provide information on what the relative effectiveness is compared with, for example, no treatment or direct surgical treatment. Finally, the relatively short follow-up is a major limitation, because decisions regarding surgical treatment in patients with OA usually develop over years, and are influenced by various other factors.
      Our results support clinical guidelines stating that treatment for CMC OA should first be nonsurgical, because, at a group level, outcome significantly improved up to 1 year after treatment and the majority of patients did not undergo additional surgical treatment within the first 2 years. Subgroup analysis indicates that initial nonsurgical treatment with an orthosis and hand therapy is also relevant, particularly for patients with higher baseline pain levels, because this subgroup showed the largest improvement in pain. The implication of our findings for patients with relatively low baseline pain levels is less clear. In this group, pain significantly increased at follow-up. However, they had a relatively low conversion to surgery rate. A possible explanation may be that these patients had only minor impairments before treatment, and became more aware of the pain in their thumb and the impairments in daily living during their treatment, which could contribute to the increased pain at 12 months after receiving an orthosis and hand therapy.
      Although we found that the median duration to surgery was 5 months, this finding was subject to multiple local factors, and therefore, may be less generalizable. For example, as a rule, decision making on additional surgical treatment was scheduled at the outpatient clinic after 3 months, when treatment was completed. In addition, factors such as waiting lists, personal factors, holidays, or financial reasons and insurance policies influenced the timing of surgery. To answer the question about the best timing to convert to surgery, a different study design would be preferred, using more frequent measurements.
      For future research, it would be interesting to study the effect of an orthosis alone versus an orthosis combined with hand therapy versus hand therapy alone. In addition, the effect of patient adherence to therapy on treatment outcome is interesting. Future studies should also focus on identifying prognostic factors to predict which patients will have a good outcome after a hand orthosis and hand therapy and which can benefit more from early surgery. In addition, future research should focus on the optimal timing of this decision. Moreover, other possible predictors that can influence treatment outcome should be evaluated, such as coping mechanisms, catastrophizing, quality of life, emotional, and mental health.

      Acknowledgments

      J.T. and K.R.S. contributed equally to this manuscript.

      Appendix A.

      Appendix AComparison of Outcome of Complete Cases Versus Outcome of Imputed Data
      VariableMean Complete Cases ± SDMean Imputed Data ± SDP Value
      VAS pain during activities baseline57.4 ± 2261.0 ± 22.154
      VAS pain during activities 6 wk45.3 ± 2248.8 ± 23.205
      VAS pain during activities 3 mo44.4 ± 2347.2 ± 23.288
      VAS pain during activities 12 mo42.7 ± 2645.7 ± 22.330
      VAS pain in the last week baseline48.1 ± 1949.6 ± 20.537
      VAS pain in the last week 6 wk35.2 ± 1939.7 ± 21.060
      VAS pain in the last week 3 mo38.3 ± 2240.4 ± 21.396
      VAS pain in the last week 12 mo35.1 ± 2539.8 ± 19.099
      MHQ total baseline64.3 ± 1563.9 ± 14.820
      MHQ total 6 wk70.0 ± 968.7 ± 9.428
      MHQ total 3 mo70.8 ± 1570.3 ± 11.791
      MHQ total 12 mo73.1 ± 1571.9 ± 9.8.502
      MHQ pain baseline57.7 ± 2462.0 ± 26.151
      MHQ pain 6 wk43.8 ± 2049.7 ± 20.011
      MHQ pain 3 mo41.2 ± 2146.3 ± 21.040
      MHQ pain 12 mo39.1 ± 2540.8 ± 20.557
      When applying Bonferroni correction to account for multiple testing, a P value of <.003 was required to find a significant difference. No significant differences were found for all variables.
      MHQ, Michigan Hand Questionnaire; SD, standard deviation; VAS, visual analog scale.

      References

        • van Uchelen J.
        • Beumer A.
        • Brink S.M.
        • Hoogvliet P.
        • Moojen T.M.
        • Spaans A.J.
        Dutch Guideline for Conservative and Surgical Treatment for Primary Thumb Base Osteoarthritis.
        NVPC, NVVH, NOV, VRA, VvBN, Amsterdam, the Netherlands2014
        • Hochberg M.C.
        • Altman R.D.
        • April K.T.
        • 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 (Hoboken). 2012; 64: 465-474
        • National Clinical Guideline Centre (UK)
        Osteoarthritis: care and management in adults.
        NICE Clin Guidel No 177. 2014; (Available at:) (Accessed March 5, 2016)
        • Zhang W.
        • Doherty M.
        • Leeb B.F.
        • 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: 377-388
        • Stamm T.A.
        • Machold K.P.
        • Smolen J.S.
        • et al.
        Joint protection and home hand exercises improve hand function in patients with hand osteoarthritis: a randomized controlled trial.
        Arthritis Rheum. 2002; 47: 44-49
        • Joshi R.
        Intraarticular corticosteroid injection for first carpometacarpal osteoarthritis.
        J Rheumatol. 2005; 32: 1305-1306
        • Colditz J.C.
        The biomechanics of a thumb carpometacarpal immobilization splint: design and fitting.
        J Hand Ther. 2000; 13: 228-235
        • Egan M.Y.
        • Brousseau L.
        Splinting for osteoarthritis of the carpometacarpal joint: a review of the evidence.
        Am J Occup Ther. 2007; 61: 70-78
        • Spaans A.J.
        • van Minnen L.P.
        • Kon M.
        • Schuurman A.H.
        • Schreuders A.R.
        • Vermeulen G.M.
        Conservative treatment of thumb base osteoarthritis: a systematic review.
        J Hand Surg Am. 2015; 40: 16-21.e5
        • Aebischer B.
        • Elsig S.
        • Taeymans J.
        Effectiveness of physical and occupational therapy on pain, function and quality of life in patients with trapeziometacarpal osteoarthritis—a systematic review and meta-analysis.
        Hand Ther. 2016; 21: 5-15
        • Melvin J.L.
        Therapist's Management of the Thumb Carpometacarpal Joint with Osteoarthritis.
        in: Mackin E.J. Callahan A.D. Skirven T.M. Schneider L.H. ALO Rehabilitation of the Hand and Upper Extremity. 5th ed. Mosby, St. Louis, MO2002
        • van der Giesen F.J.
        • Nelissen R.G.
        • Arendzen J.H.
        • de Jong Z.
        • Wolterbeek R.
        Vliet Vlieland TP. Responsiveness of the Michigan Hand Outcomes Questionnaire-Dutch language version in patients with rheumatoid arthritis.
        Arch Phys Med Rehabil. 2008; 89: 1121-1126
        • Chung K.C.
        • Pillsbury M.S.
        • Walter M.R.
        • Hayward H.R.
        Reliability and validity testing of the Michigan Hand Outcome Questionnaire.
        J Hand Surg Am. 1998; 23: 575-587
        • Marks M.
        • Audige L.
        • Herren D.B.
        • Schindele S.
        • Nelissen R.G.H.H.
        • Vliet Vlieland T.P.
        Measurement properties of the German Michigan Hand Outcomes Questionnaire in patients with trapeziometacarpal osteoarthritis.
        Arthritis Care Res (Hoboken). 2014; 66: 245-252
        • London D.A.
        • Stepan J.G.
        • Calfee R.P.
        Determining the Michigan Hand Outcomes Questionnaire minimal clinically important difference by means of three methods.
        Plast Reconstr Surg. 2014; 133: 616-625
        • Cohen J.
        Statistical Power Analysis for the Behavioral Sciences.
        Lawrence Erlbaum Associates, Hillsdale, NJ1988
        • Little R.J.A.
        • Rubin D.B.
        Statistical Analysis with Missing Data.
        2nd ed. Wiley, Hoboken, NJ2002
      1. LeeJ H, Huber J Jr. Multiple imputation with large proportions of missing data: how much is too much? United Kingdom Stata Users’ Group Meetings. London: Stata Users Group; 2011:23.

        • Kelly C.
        • Price T.D.
        Correcting for regression to the mean in behavior and ecology.
        Am Nat. 2005; 166: 700-707
        • Visser A.W.
        • Bøyesen P.
        • Haugen I.K.
        • et al.
        Instruments measuring pain, physical function, or patient’s global assessment in hand osteoarthritis: a systematic literature search.
        J Rheumatol. 2015; 42: 2118-2134
        • Villafañe J.H.
        • Cleland J.A.
        • Fernández-de-Las-Peñas C.
        The effectiveness of a manual therapy and exercise protocol in patients with thumb carpometacarpal osteoarthritis: a randomized controlled trial.
        J Orthop Sports Phys Ther. 2013; 43: 204-213
        • O’Brien V.H.
        • Giveans M.R.
        Effects of a dynamic stability approach in conservative intervention of the carpometacarpal joint of the thumb: a retrospective study.
        J Hand Ther. 2013; 26 (quiz 52): 44-51
        • Frouzakis R.
        • Herren D.B.
        • Marks M.
        Evaluation of expectations and expectation fulfillment in patients treated for trapeziometacarpal osteoarthritis.
        J Hand Surg Am. 2015; 40: 483-490
        • Wajon A.
        • Vinycomb T.
        • Carr E.
        • Edmunds I.
        • Ada L.
        Surgery for thumb (trapeziometacarpal joint) osteoarthritis.
        Cochrane database Syst Rev. 2015; : CD004631
        • 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: 415-417