To compare the effectiveness and acceptability of silver ring splints (SRS) and commercial prefabricated thermoplastic splints (PTS) in treating swan neck deformities in patients with rheumatoid arthritis (RA).
To compare the effectiveness and acceptability of silver ring splints (SRS) and commercial prefabricated thermoplastic splints (PTS) in treating swan neck deformities in patients with rheumatoid arthritis (RA).
Consecutive patients with RA and a mobile swan neck deformity were included in a randomized, crossover trial. In 2 different sequences, patients used both splints for 4 weeks, with a washout period of 2 weeks. Afterward, patients used the preferred splint for another 12 weeks. The primary outcome measure was dexterity measured with the Sequential Occupational Dexterity Assessment (SODA). Secondary outcome measures included other measures of hand function, satisfaction with the splints, and splint preference.
Fifty patients were included, and 47 (94%) of those completed the study. Eighteen patients (36%) had 1 swan neck deformity, whereas the other patients had 2 or more. The improvement of the total SODA score with the SRS (11.2; 95% confidence interval [95% CI] 8.1, 14.3) and PTS (10.8; 95% CI 7.5, 14.1) was similar (difference −0.5; 95% CI −2.2, 1.2). In addition, there were no significant differences in change scores regarding the other clinical outcome measures, or satisfaction. Twenty-four patients preferred the SRS, 21 preferred the PTS, and 2 patients chose neither. A comparison in the 12-week followup period yielded similar clinical outcomes, with the exception of a significantly higher score in 3 items of satisfaction in the SRS group.
For patients with RA and a mobile swan neck deformity, SRS and PTS are equally effective and acceptable.
The swan neck deformity is a common finger deformity associated with rheumatoid arthritis (RA). The deformity is characterized by flexion at the metacarpalphalangeal joint, hyperextension of the proximal interphalangeal joint (PIPJ), and flexion of the distal interphalangeal joint (DIPJ) caused by a dominance of the extensor apparatus, which is not counteracted by flexion forces. This can result in a decreased ability to actively flex the PIPJ, leading to impaired dexterity (1–4). With conservative treatment, finger splints aiming to prevent hyperextension and allow flexion of the PIPJ can be employed. Currently, 3 types of finger splints are available in The Netherlands: custom- made silver ring splints (SRS), custom-made thermoplastic splints (CTS), and prefabricated thermoplastic splints (PTS). SRS are made of sterling silver and manufactured according to the patient's ring size. The maximum allowed extension of the PIPJ can be individually adjusted by bending the splint within material limits. SRS cost ∼€CE80 (US $126.40) each. CTS are individually fabricated. Their costs depend on the time needed for fabrication, usually ranging from 21 to 44 minutes (5). PTS are available in kits containing numerous sizes, with minimal time required for individual adjustments. They are made of less material than CTS and their price is ∼€CE15 (US $23.70), exclusive of the therapist time (6).
To date, studies on the effectiveness of finger splints for swan neck deformity are sparse. In an uncontrolled study with 17 patients who had RA and swan neck deformities (7), SRS improved dexterity but did not have an effect on grip strength, self-reported hand function, or hand pain. In a controlled study comparing SRS and CTS in 18 patients with RA (5), both splints were found to be equally effective with respect to digital stability, grip strength, and mobility of the finger joints. However, CTS were considered less comfortable and less attractive. Since PTS appear to be more elegant and comfortable than CTS, they may offer an acceptable, inexpensive alternative to the SRS. Therefore, our study aimed to compare the effectiveness, patient satisfaction, and patient preferences of PTS and SRS in patients with RA and swan neck deformities.
This study used a multicenter, randomized, crossover design. Patients were randomly assigned to the SRS-PTS or PTS-SRS sequence. Randomization was done with a random digit generator with stratification for center and sex. Blocked randomization of 4 patients per block was used to prevent unbalanced distribution. The randomization procedure was completed by an administrative assistant who was not aware of the block size. Patients used each splint for 4 weeks with a washout period of 2 weeks in between. After both treatment periods, patients chose their preferred splint and were followed up for 12 weeks while wearing the preferred splint.
From August 2005 through September 2006, consecutive patients with RA and swan neck deformities were recruited one patient at a time until the target sample size was obtained at the outpatient rheumatology clinics of 3 centers in The Netherlands: Leiden University Medical Center (Leiden), Sint Maartenskliniek (Nijmegen), and Reinier de Graaf Gasthuis (Delft). Patients were eligible for the study if they had been diagnosed with RA according to the 1987 criteria of the American College of Rheumatology (formerly, the American Rheumatism Association) (8), were ≥18 years old, had a mobile swan neck deformity that was manually correctable to ≥45° of PIPJ flexion of an index and/or middle finger, had stable disease activity, had received no corticosteroid injections for the previous 3 months, had no planned surgery for the duration of the study, and had not been treated with swan neck finger splints in the past. Exclusion criteria were the presence of a condition other than RA, or other severe finger deformities that interfered with hand function or the use of finger splints. The study was approved by the hospital medical ethics committees of the 3 participating hospitals, and all participants gave written informed consent.
The study used the SIRIS Swan Neck Splint (Silver Ring Splint Company, Charlottesville, VA) and the Oval-8 Finger Splint (3-Point Products, Stevensville, MD). Both splints were sized according to the manufacturer's recommendations by experienced (>5 years) hand therapists (FJvdG and CK-S), and fitted in slight flexion, which corrected the hyperextension of the PIPJ. All the correctable swan neck deformities of the index or middle finger(s) were splinted. The participants were asked to wear the splints as much as possible, removing them only for cleaning. During the crossover period, participants returned the first splint type worn to ensure the consistent wearing of the second splint type.
Sociodemographic and disease characteristics were collected at baseline (T0). Outcome measurements were gathered at T0, after the first treatment period (T1), after the washout period (T2), after the second treatment period (T3), and after the 12-week followup period (T4). All measurements were collected with patients wearing the splints, except for the PIPJ hyperextension. Adherence and satisfaction were measured at T1, T3, and T4. The patients' preferences were obtained at T3 and at T4 by asking whether they would continue wearing the splints (yes, unsure, no). All clinical measurements were performed by 2 experienced hand therapists (FJvdG and CK-S). To enhance the intrarater reliability, a training session and calibration meeting was scheduled prior to the start of the study and once during the study.
Sociodemographic data included sex, age, living status (alone or with others), and paid employment (yes/no and number of hours per week, if applicable). Disease duration was extracted from the patient records. Disease activity was measured with the Disease Activity Score in 28 joints using 4 variables: the 28-joint counts for swelling and tenderness, the erythrocyte sedimentation rate, and the patient's overall assessment of well-being (9). Functional ability was measured with the Health Assessment Questionnaire (HAQ) (10), which covers 20 activities of daily living in 8 dimensions. The total HAQ score is the average score of the 8 dimensions (where 0 = best possible function and 3 = worst possible function). General health status was measured with a validated version of the Medical Outcomes Study Short Form 36 health survey (11), which includes 8 subscales. These subscales can be converted into 2 summary scales: the physical and mental component summary scales, standardized to a mean ± SD score of 50 ± 10 in the general population. For that purpose, we used the scores from an age- and sex-matched normative sample, which was drawn from a large, random, nationwide sample of adults (n = 1,742) from the general Dutch population (12) and factor score coefficients (13).
The number of swan neck deformities, all deviations from the splinting protocol, and adverse events were recorded. Adherence during the first 2 splinting periods was assessed with a diary, in which the participants recorded the number of hours per day that they used the splint(s). Because the prolonged usage of diaries is usually hampered by low compliance, adherence during the followup period was assessed only once at the end of that period, when patients were asked, “How many days of the week did you wear the splint on average in the past 12 weeks?” and “When you were wearing the splint, for how many hours per day was this in the past 12 weeks?” The results of these questions were multiplied to obtain an average number of splint usage hours per week.
The primary outcome measure was dexterity as measured by the Sequential Occupational Dexterity Assessment (SODA) (14–16). With the SODA, the patient performs 12 standardized tasks and an assessor scores the ability to perform every task (where 4 = able to perform in the requested way, 1 = able to perform in a different way, and 0 = unable to perform) and the level of difficulty in the performance (where 2 = not difficult, 1 = some difficulty, and 0 = very difficult). The SODA score range is 0–108, with a higher number indicating better dexterity. The SODA-pain score is computed by counting the number of painful tasks (range 0–12).
The secondary outcome measures included general hand function as measured by the hand and finger function domain of the Dutch Arthritis Impact Measurement Scales 2 (D-AIMS2) (17) and the Michigan Hand Outcomes Questionnaire (MHQ) (18, 19). The hand and finger function subscale of the D-AIMS2 consists of 5 questions regarding the patient's hand function over the previous week. The scores range from 0 (good hand function) to 10 (bad hand function). The MHQ is a questionnaire covering 6 domains: overall hand function, activities of daily living, pain, work performance, esthetics, and satisfaction with hand function. The total score range (average of all domains) is 0–100, with higher scores indicating better hand function.
Passive PIPJ hyperextension of the splinted fingers was measured unsplinted with a goniometer (20). If more than one finger was splinted, the average of the scores of the involved fingers was used. Cylindrical grip strength and pinch grip strength were determined with a Jamar dynamometer (JA Preston Corporation, Clifton, NJ) (21) and a North Coast hydraulic pinch gauge (Pieksma Medical Supplies, Bussum, The Netherlands), respectively. The patients underwent the tests twice on every side, with the highest score recorded. Scores were obtained for the splinted finger(s) only. In the case of involvement of multiple fingers, average scores were used. After both initial splinting periods, the patient's perceived change in hand function was measured on a 3-point scale (where 1 = worsened, 2 = neither worsened nor improved, and 3 = improved). In the absence of a validated questionnaire to measure the RA patients' satisfaction with hand or finger splints, we used a 13-item questionnaire previously developed by one of the authors (Stern EB: personal communication). Each item was rated on a 5-point Likert scale (where 1 = totally disagree and 5 = totally agree).
The sample size was calculated based on the SODA as the primary outcome measure and using the formula:
with α = 0.05 and 1 minus β = 0.80, then (Zα/2 + Zβ)2 = 7.85. In the absence of data on the clinical relevance of SODA change scores, we used the data from a study (7) executed within a context similar to the present study (22) with respect to the patient population (RA patients with finger deformities), the intervention (finger splints), and the primary outcome measure (dexterity). In that uncontrolled study on custom-made silver splints in 17 patients with RA, a median improvement of 5 points on the SODA after 1 month was observed (7), yielding a value of 5 for the mean difference. From the individual data presented in that study, an SD of the SODA change score (τ) of 8.95 was calculated. With a comparison of 2 groups with equal numbers implying that r = 1, and using the previously mentioned formula, a total of 50 patients per group would be required. To detect a difference of 6 points on the SODA score, in line with the 12-week difference observed by Zijlstra et al (7), 36 patients per group would be needed. As in this study, a crossover design was employed, and the total number of patients required for the study was 50.
For all data, we observed the distribution plots. In case of normally distributed data, we presented mean ± SDs or 95% confidence intervals (95% CIs), and in case of a non-normal distribution, we presented median and interquartile range as the net result of the 75th percentile minus the 25th percentile.
To determine if a period effect was present, the differences in effect between the 2 periods were compared between the 2 sequences (SRS-PTS or PTS-SRS). To determine if a carryover effect was present, the change scores of the 2 splints in the 2 periods were averaged for every patient, and the mean change score was then compared between the 2 splinting sequences (23). These possible effects were tested for all outcome variables by unpaired t-tests. In the absence of statistically significant carryover or period effects, the main analysis would pertain to a comparison of the change scores after wearing them for 4 weeks between the 2 splints by means of the paired t-test. McNemar's test was used to compare numbers of patients reporting the following categories of perceived changes in hand function between the 2 splints: “improved” versus both “neither worsened nor improved” and “worsened.” In addition, a within-splint analysis was performed using paired t-tests to test for significant differences between T0 and followup measurements after 4 weeks. To study the effects after a 12-week followup period, the change scores between T3 and T4 within the groups of patients preferring the SRS and PTS were compared using unpaired t-tests. In addition, the chi-square test was used to assess a trend in differences in willingness to continue using the splints between patients who chose the SRS or the PTS. For that purpose, the outcomes were dichotomized into “yes” versus “unsure and no.” All analyses were computed using the Statistical Package for the Social Sciences, version 14.0 (SPSS, Chicago, IL). The level of significance was set at less than 0.05 for all statistical tests.
A total of 83 patients were invited and screened (Figure 1). Eventually, 50 participants entered the study, and 3 of those patients did not complete the study. Table 1 shows the characteristics of the study sample.
|Age, years||53.8 (21.6)|
|Living with other(s), no.||45|
|Not living with other(s), no.||5|
|Paid employment, no. (%)||13 (26)|
|Hours of paid employment per week (n = 13)||20 (18)|
|Disease duration, years||13.7 (11.5)|
|HAQ (range 0–3)||1.13 (1.1)|
|SF-36 physical component summary (range 0–100)||38.3 (15.2)|
|SF-36 mental component summary (range 0–100)||56.2 (15.9)|
|Number of swan neck deformities, no. (%)†|
In the initial 2 treatment periods, 8 patients did not use one or more of their splints for the full study period. Seven patients, all having more than 1 swan neck deformity, could not use 1 splint during 1 period because it was broken (1 SRS), lost (2 SRS and 3 PTS), or did not fit because of local joint inflammation (1 PTS). In one patient, neither the SRS nor the PTS could be used on one finger because of the development of an interfering nodule, but the patient continued using splints on other fingers. One patient developed skin problems on both splinted fingers (SRS), possibly due to the relatively high hyperextension force. This patient refused further participation after 2 weeks. Two patients reported minor skin problems with the PTS due to perspiration in warm weather, but they could continue to wear the splints throughout the assigned period. In the initial treatment periods, 47 of 49 SRS diaries, and 47 of 48 PTS diaries, were filled in. Adherence rates of the SRS group (mean ± SD 15.3 ± 7.4 hours/week) and the PTS group (mean ± SD 15.4 ± 7.4 hours/week) were similar (difference −0.05; 95% CI −2.1, 1.9).
None of the outcome measures demonstrated a significant period or carryover effect (P > 0.05 for all; data not shown). As shown in Table 2, after 4 weeks of use there were no differences between the change scores of the 2 finger splints for any of the outcome measures. The number of patients considering their hand function “worse,” “worse nor improved,” or “improved” were 5, 18, and 21 for the SRS splint, and 2, 21, and 20 for the PTS splint, respectively (P = 0.42 by McNemar's test).
|SRS (n = 50)||PTS (n = 50)||Difference between splints|
|SODA, range 0–108|
|Baseline, mean ± SD||83.7 ± 18.0||84.3 ± 18.8|
|Change after 4 weeks||11.2 (8.1, 14.3)†||10.8 (7.5, 14.1)†||−0.5 (−2.2, 1.2)|
|SODA-pain, range 0–12|
|Baseline, mean ± SD||1.6 ± 2.6||1.7 ± 2.5|
|Change after 4 weeks||−0.5 (−0.9, −0.1)†||−0.5 (−1.0, −0.1)†||0.4 (−0.1, 1.0)|
|D-AIMS2 hand/finger subscale, range 0–10|
|Baseline, mean ± SD||2.72 ± 2.27||3.07 ± 2.26|
|Change after 4 weeks||−0.14 (−0.55, 0.27)||−0.21 (−0.62, 0.19)||−0.01 (−0.66, 0.65)|
|MHQ total score, range 0–100|
|Baseline, mean ± SD||60.7 ± 15.6||58.7 ± 13.7|
|Change after 4 weeks||−0.1 (−3.0, 2.7)||1.4 (−2.0, 4.8)||−1.3 (−6.4, 3.8)|
|Passive PIPJ hyperextension, degrees|
|Baseline, mean ± SD||27.7 ± 8.2||27.0 ± 9.1|
|Change after 4 weeks||−2.1 (−3.3, −1.0)†||−1.1 (−2.4, 0.1)||−1.1 (−2.6, 0.5)|
|Cylindrical grip strength, kg|
|Baseline, mean ± SD||17.7 ± 9.9||17.5 ± 10.4|
|Change after 4 weeks||−0.13 (−1.10, 0.82)||0.39 (−0.36, 1.15)||−0.53 (−1.67, 0.61)|
|Pinch grip strength, kg|
|Baseline, mean ± SD||2.4 ± 1.7||2.3 ± 1.7|
|Change after 4 weeks||0.03 (−0.20, 0.26)||0.08 (−0.17, 0.32)||−0.02 (−0.24, 0.21)|
|Patient preferences, no. (%)||24 (51)||21 (45)|
Concerning within-group changes with both splints, both SODA scores improved significantly after 4 weeks. Moreover, with the SRS splints, passive PIPJ hyperextension decreased significantly between T0 and 4 weeks. With all other outcomes, a trend toward improvement was seen with both splints; however, none of the changes reached statistical significance. After using both splints for 4 weeks, 24 patients preferred the SRS, 21 the PTS, and 2 chose neither (P = 0.66 by chi-square test).
After choosing a preferred splint, 3 patients, using more than one splint, did not use one because it was found to be uncomfortable (2 SRS) or it was lost (1 PTS). In the followup period, the mean ± SD splinting duration was 11.7 ± 8.1 hours/day with the SRS, and 16.3 ± 6.5 hours/day with the PTS (difference −4.6, 95% CI −9.2, 0.03). Table 3 shows that with 24 patients continuing to wear the SRS and 21 patients the PTS, there are no significant differences between the 2 splints regarding the change scores between 10 and 22 weeks on any of the clinical outcome measures.
|SRS (n = 24)||PTS (n = 21)||Difference between splints|
|SODA, range 0–108||−0.4 (−1.8, 0.9)||−0.2 (−3.4, 3.0)||−0.2 (−3.4, 3.0)|
|SODA-pain, range 0–12||−0.3 (−1.0, 0.5)||0.5 (−0.4, 1.4)||−0.8 (−1.9, 0.4)|
|D-AIMS2 hand/finger subscale, range 0–10||−0.08 (−0.7, 0.6)||0.00 (−0.6, 6.0)||−0.08 (−1.0, 0.8)|
|MHQ total score, range 0–100||0.6 (−3.2, 4.5)||−2.0 (−7.1, 3.1)||2.6 (−3.4, 8.7)|
|Passive hyperextension PIPJ, degrees||−1.0 (−4.6, 2.5)||−2.7 (−5.8, 0.5)||1.6 (−3.1, 6.3)|
|Cylindrical grip strength, kg||0.2 (−1.7, 2.1)||−0.3 (−2.4, 1.9)||0.5 (−2.3, 3.2)|
|Pinch grip strength, kg||0.1 (−0.2, 0.4)||−0.1 (−0.6, 0.4)||0.3 (−0.3, 0.8)|
For the other outcome measures, there were no significant changes during the 12-week followup between and within the splints, indicating that effects obtained after 4 weeks were sustained after 12 weeks. When asked at followup if they would continue wearing the splints after the study, 21 (91%) of the 23 patients using the SRS answered positively, and 2 (9%) answered indifferently or negatively. In the PTS group, 11 patients (61%) answered positively, and 7 patients (39%) answered indifferently or negatively (SRS versus PTS; P = 0.79 by chi-square test).
Table 4 shows the results of the patient satisfaction questionnaire. Overall, after 4 weeks the scores were higher for the SRS than for the PTS, except for item 12, but only the score on item 1 reached statistical significance.
|After 4 weeks||After 12 weeks followup|
|SRS (n = 50)||PTS (n = 50)||Difference mean (95% CI)||SRS (n = 24)||PTS (n = 21)||Difference mean (95% CI)|
|1. Did you like how the splint looked? (1–5)||4.3 ± 0.6||3.7 ± 1.0||0.5 (0.2, 0.8)†||4.5 ± 0.5||3.9 ± 0.6||0.6 (0.2, 0.9)‡|
|2. Did you like how your fingers looked with the splint on? (1–5)||4.0 ± 0.6||3.8 ± 0.8||0.2 (−0.1, 0.5)||4.2 ± 0.6||3.8 ± 0.6||0.4 (0.0, 0.8)‡|
|3. Was the splint easy to put on and take off? (1–5)||4.1 ± 0.8||4.0 ± 0.7||0.1 (−0.2, 0.4)||4.3 ± 0.4||4.0 ± 0.9||0.3 (−0.2, 0.7)|
|4. Was the splint easy to clean? (1–5)||4.4 ± 0.5||4.2 ± 0.6||0.1 (−0.1, 0.3)||4.3 ± 0.4||4.1 ± 0.4||0.2 (−0.1, 0.5)|
|5. Did the splint relieve your finger joint pain? (1–5)||3.2 ± 0.8||3.0 ± 1.0||0.2 (−0.1, 0.5)||3.1 ± 1.0||2.9 ± 0.8||0.1 (−0.5, 0.8)|
|6. Did the splint decrease your finger joint swelling? (1–5)||2.9 ± 0.9||2.8 ± 1.1||0.1 (0.1, 0.4)||3.0 ± 0.9||2.8 ± 0.7||0.2 (−0.3, 0.8)|
|7. Did the splint improve your grasp while you wore it? (1–5)||3.0 ± 0.8||2.9 ± 0.9||0.1 (−0.2, 0.4)||3.6 ± 0.9||3.1 ± 0.9||0.5 (−0.1, 1.1)|
|8. Was your hand stronger with the splint on?||2.8 ± 0.8||2.7 ± 1.0||0.1 (−0.2, 0.4)||3.0 ± 0.9||2.9 ± 0.8||0.2 (−0.4, 0.7)|
|9. Did the splints keep your fingers from hyper extending (bending back) the majority of the time? (1–5)||4.1 ± 0.9||3.9 ± 1.2||0.3 (−0.1, 0.6)||4.6 ± 0.6||4.0 ± 1.0||0.6 (0.1, 1.1)‡|
|10. Did the splint help you manipulate small objects? (1–5)||3.1 ± 0.8||2.9 ± 0.8||0.2 (−0.1, 0.4)||3.3 ± 0.9||3.0 ± 0.7||0.3 (−0.2, 0.8)|
|11. Was it easier to perform daily activities with the splint on? (1–5)||3.0 ± 0.7||2.9 ± 0.8||0.1 (−0.3, 0.3)||3.3 ± 0.8||2.8 ± 0.8||0.5 (0.0, 1.0)|
|12. Did the splint stay on your fingers when reaching into tight areas? (1–5)||2.9 ± 1.1||3.0 ± 1.0||−0.1 (−0.5, 0.2)||3.1 ± 1.1||3.2 ± 0.9||−0.1 (−0.8, 0.6)|
|13. Was the splint comfortable to wear? (1–5)||4.0 ± 1.0||3.3 ± 1.0||0.2 (−0.2, 0.5)||4.0 ± 0.6||3.6 ± 0.8||0.4 (−0.1, 0.8)|
In the followup period, with the patients using the preferred splint, the scores were significantly higher for the SRS than for the PTS, regarding the items 1, 2, and 9.
The current study suggests that SRS and PTS are equally effective and acceptable in patients with RA and mobile swan neck deformities. Both splints improve dexterity (as measure by SODA) and reduce dexterity-related pain (as measured by the SODA-pain score), but only the SRS reduces PIPJ hyperextension. Neither splint significantly improved or interfered with reported hand function, measured grip, or pinch strength. Except for a significantly higher satisfaction score in 3 satisfaction items with the SRS, a followup with the preferred splint yielded equal results with both splints. Almost two-thirds of the patients said they would continue wearing their preferred splints after the study.
With respect to the comparison of the effectiveness of thermoplastic splints with SRS, our results parallel those of the study by Schegget and Knipping (5), where CTS were found to be equally effective as SRS in 18 patients with RA. However, in that study, CTS were found to be far less acceptable than SRS, mainly due to their less attractive appearance and their thickness, making the fingers spread. In contrast, no differences in satisfaction were seen after 4 weeks with the PTS used in the present study, with equal numbers of patients choosing the SRS and the PTS. In the followup period, with the patients wearing the splint of their choice, 2 aesthetics-related items of satisfaction were valued higher in the SRS than in the PTS; however, the absolute difference was small, so its clinical significance is questionable.
The results of our study suggest that the provision of a PTS is worth considering for patients with a mobile swan neck deformity. This is especially the case since PTS are less expensive than SRS, with similar time needed for the hand therapist to measure and adjust the splint to obtain the optimal fit and allowed extension.
Similar to the results of the present study, significant improvements of the SODA dexterity score with SRS were also seen in a previous observational study (7). However, the SODA dexterity score improvements were larger in the current study. Moreover, the SODA-pain score improved, whereas in the study by Zijlstra et al, no improvement was seen. The greater improvement in the present study might be explained by the fact that Zijlstra and colleagues included patients with a longer disease duration (median 21 years) and worse hand function, which is reflected by a lower baseline SODA score (median 71). As suggested by the authors of that study, it could be hypothesized that finger splints for swan neck deformity are more effective in the earlier stages of disease, when correction is relatively easy. To date, there are no studies available with respect to the clinical relevance of the observed changes of the SODA score.
In all other measures of hand function, no improvements were seen either in our study or in the study by Zijlstra et al, except for a significant decrease of passive mobility of the PIPJ in the present study. This latter result is comparable with the outcome of the study by Schegget and Knipping (5), which demonstrated a reduction of passive PIPJ hyperextension of equal magnitude with both SRS and thermoplastic splints after 24 weeks. Since the absolute reduction of passive PIPJ hyperextension was small, its clinical relevance is doubtful.
The absence of effect of finger splints on various outcome measures related to hand function could possibly be explained by the impact of limitations of joints or structures other than the PIPJ and the DIPJ of the index and middle fingers. Although patients with limitations or deformities seriously interfering with hand and finger function were excluded, most patients had other mild limitations such as ulnar deviation, or swan neck or boutonnière deformities of the thumb or fourth and fifth fingers. As these other limitations would probably affect hand function to a similar extent with both splints, they are not likely to have had an impact on the comparisons between the two splints.
Despite the modest effectiveness of both splints, more than 60% of the patients in the current study indicated that they would continue using the splints after the study. This rate is in line with the findings of Zijlstra et al. However, studies with a longer duration of followup assessment would be needed to examine which proportion of the patients would fulfill their intention.
A limitation of the study is its crossover design, where despite statistical testing, carryover or period effects cannot be totally ruled out. Furthermore, when calculating the sample size, we did not take into account a dropout rate. Since 3 patients left the study prematurely, the possibility that this study did not have sufficient power to detect a difference of 5 points or less on the SODA score cannot be totally ruled out. Because it is not possible to blind assessors in this type of research, where measurements are done with splints on, bias toward the effectiveness of finger splints cannot be completely avoided. Nevertheless, having the assessments performed by an independent assessor, rather than 2 of the authors, would probably have been preferable. Another potential source of bias could be the lack of blinding of assessors and patients, due to the nature of the intervention. In addition, we used a satisfaction questionnaire, of which the psychometric properties have not yet been established. Given the fact that specific satisfaction questionnaires for hand and finger splints are scarce, more research into the validation of questionnaires used to evaluate RA patients' satisfaction with orthoses seems justified.
In summary, the results of the present study, and that of the available evidence (5, 7), suggest that wearing SRS or PTS for mobile swan neck deformities is effective in improving dexterity and dexterity-related pain after 4 weeks in patients with RA. The decision about what type of splint to use should therefore not depend on its effectiveness, but merely on patient preferences and costs.
All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved the final version to be submitted for publication. Mr. van der Giesen had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Study conception and design. Van der Giesen, van Lankveld, Stern, le Cessie, Nelissen, Vliet Vlieland.
Acquisition of data. Van der Giesen, Kremers-Selten, Peeters, Nelissen.
Analysis and interpretation of data. Van der Giesen, Stern, le Cessie, Vliet Vlieland.
The authors would like to thank the patients who participated in the study.