Advertisement
Scientific/Clinical Article| Volume 25, ISSUE 1, P89-96, January 2012

Download started.

Ok

A Retrospective Review to Determine the Long-term Efficacy of Orthotic Devices for Trigger Finger

      Abstract

      Purpose: To evaluate the use of orthotic devices (splints) in an attempt to resolve trigger finger. Methods: Data were extracted from 46 charts during a five-year period from January 2005 to December 2010. At ten weeks, patients were seen for follow-up assessment of pain and stage of stenosing tenosynovitis (SST). One-year follow-up was performed to determine if the patients required further surgical intervention or steroid injection. The data were analyzed to determine the efficacy of orthosis intervention. Results: Mean pain score preorthotic is 5.63 and postorthotic is 1.20. Mean SST score preorthotic is 3.93 and postorthotic is 1.21. There was an 87% (40 patients) success rate with the orthotic intervention; 4.3% (two patients) had surgery and 8.5% (four patients) received a steroid injection in the year after orthotic application. Conclusion: This study demonstrated the efficacy of orthoses for the reduction of pain and SST score for patients who have trigger finger.

      Level of Evidence

      3.
      Trigger finger is a common finger condition, thought to be caused by inflammation and subsequent narrowing of the A1 pulley. Primary trigger finger occurs up to six times more frequently in women than men and the onset is usually in the middle fifth to sixth decades of life.
      • Makkouk A.L.
      • Oetgen M.E.
      • Swigart C.R.
      • Dodds S.D.
      Trigger finger: etiology, evaluation, and treatment.
      The lifetime risk of developing trigger finger is between 2% and 3%, but increases to up to 10% in diabetics.
      • Makkouk A.L.
      • Oetgen M.E.
      • Swigart C.R.
      • Dodds S.D.
      Trigger finger: etiology, evaluation, and treatment.
      Trigger finger symptoms can include pain, clicking, catching, and loss of motion of the affected finger. Functional limitations can include limited grip strength and decreased ability to hold narrow handled objects, which can cause “triggering” of the involved digits.
      The goal of orthotic device intervention (splints) is to prevent the friction caused by flexor tendon movement through the affected A1 pulley until the inflammation there resolves.
      • Akhtar S.
      • Bradley M.J.
      • Quinton D.N.
      • et al.
      Management and referral for trigger finger/thumb.
      In completing a literature search regarding orthoses and trigger finger, the following was found: Evans et al.
      • Evans B.
      • Hunter J.
      • Burkhalter W.
      Conservative management of the trigger finger: a new approach.
      demonstrated a 73% success rate using an orthotic device combined with hook and fist exercises. Patel and Bassini
      • Patel M.R.
      • Bassini L.
      Trigger fingers and thumb: when to splint, inject or operate.
      reported a 70% successful outcome with immobilizing the metacarpophalangeal joint (MCP) joint in 10–15° of flexion. They found that orthotic device intervention was successful in 77% of the patients whose symptoms had been present six months or less and in 44% of those patients with symptoms longer than six months. Colbourn et al.
      • Colbourn J.
      • Heath N.
      • Manary S.
      • Pacifico D.
      Effectiveness of splinting for the treatment of trigger finger.
      demonstrated a clinically significant positive relationship between the use of a custom thermoplastic ring orthotic device and the subjective improvement in symptoms and objective outcome measures for an isolated incidence of trigger finger at ten weeks. The purpose of this retrospective study was to evaluate the use of orthoses designed to restrict composite digit flexion in an attempt to resolve isolated trigger finger, trigger thumb, and multiple incidence of trigger finger in patients. The long-term efficacy of orthotic device intervention was also evaluated.

      Methods

      Chart Selection Process

      The eligible participants for the study were identified through a review of the International Classification of Diseases
      The International Classification of Diseases: 9th Revision: Clinical Modification. ICD.9CM.
      diagnosis code 727.03 (Trigger finger, acquired) through the database at a private practice outpatient therapy facility. The participants received services between the dates of January 2005 to December 2010. Seventy-six charts were initially reviewed. Thirty charts of the initial 76 were eliminated because the patients did not meet the inclusion criteria. Inclusion criteria were defined as participants who had trigger finger of one or more digits who received orthosis intervention for trigger finger and whose chart had complete information on the measures of interest at the initial, ten-week, and one-year time points. Exclusion criteria included participants who had previously received a steroid injection in the affected finger. Thirty charts were excluded from this study based on the exclusion criteria (n=30 patients received a steroid injection in the affected finger). All subjects who met the inclusion criteria signed a consent letter to allow the researcher to use their health information for research. Each potential subject was adequately informed of the aims, methods, the anticipated benefits, and potential risks of the study and the discomfort it may entail. Every precaution was taken to protect the privacy of research subjects and the confidentiality of their personal information to minimize the impact of the study on their physical, mental, and social integrity.

      Subjects

      The final review of 46 charts included 17 participants who had trigger finger of more than one digit and 29 participants had isolated trigger finger. Orthosis fabrication and provision was provided at the same facility for all participants by the same certified hand therapist (CHT).

      Data Collected

      Data were collected from the therapy chart. The measurement instruments used in the study included the trigger finger grade using the stages of stenosing tenosynovitis (SST)
      • Patel M.R.
      • Bassini L.
      Trigger fingers and thumb: when to splint, inject or operate.
      (Table 1) and Visual Analog Scale (VAS) pain score. The CHT determined the trigger finger grade using the SST scale. The patient rated their pain using the 10-cm VAS pain scale with 0 indicating no pain and 10 indicating severe pain. The demographic information collected on the patients included digits involved, orthosis provided, age, gender, insurance company, hand dominance, number of subjects employed outside the home, and co-existing conditions (Table 2).
      Table 1Stages of Stenosing Tenosynovitis
      StageSymptoms
      1Normal
      2A painful palpable nodule
      3Triggering
      4The proximal interphalangeal joint (PIP) locks into flexion and is unlocked with active PIP joint extension
      5The PIP joint locks and is unlocked with passive PIP joint extension
      6The PIP joint remains locked in a flexed position
      Table 2Descriptive Characteristics
      CharacteristicsN (Total 46)Percentage
      Mean age68.48
      Gender
       Female2860
       Male1840
      Employed1022
      Affected hand
       Right2860
       Left1840
      Affected dominant hand2860
       Isolated affected finger2963
       Index24
       Long1533
       Ring511
       Small24
       Thumb511
      Combined affected fingers1737
       Index and thumb24
       Index and long24
       Long and ring715
       Long and small12
       Ring and small37
       Ring and thumb12
       Long, ring, and thumb12
      Taking anti-inflammatory medication715
       Co-existing diagnoses2860
       Mallet24
       Diabetes511
       Osteoarthritis1328
       Rheumatoid arthritis37
       Carpal tunnel syndrome511
      N=number.
      At the one-year follow-up phone conversation, the discharged patients were questioned regarding whether or not they had required either surgical intervention or steroid injection for their trigger finger during the past year. They were also asked if they had experienced any adverse side effects from the orthosis.

      Procedure

      All patients received the same treatment provided by the same CHT. During the initial visit, a custom thermoplastic orthosis was fabricated from Polyform® (Roylan Manufacturing, Cedarburg, Wisconsin). If only one digit was involved, the patient received a static finger circumferential orthosis that blocked proximal interphalangeal joint (PIP) motion, but allowed metacarpal phalangeal (MP) and distal interphalangeal joint (DIP) active motion (Figure 1). The decision to use a single digit orthosis was based on the understanding that blocking PIP flexion would restrict flexor tendon movement through the affected A1 pulley, while leaving the palm unrestricted. The thumb orthosis immobilized the interphalangeal (IP) joint but kept the thumb tip as free as possible for improved prehension and tactile discrimination (Figure 2). When individuals had more than one trigger finger, they wore a hand finger orthosis that immobilized the MP joints of the involved digits in 10–15° of flexion that allowed unrestricted IP motion (Figure 3). The orthosis positioned the MCP joint in approximately 10–15° of joint flexion, which optimizes hand function by allowing tip-to-tip prehension, as supported by several authors.
      • Patel M.R.
      • Bassini L.
      Trigger fingers and thumb: when to splint, inject or operate.
      • Colbourn J.
      • Heath N.
      • Manary S.
      • Pacifico D.
      Effectiveness of splinting for the treatment of trigger finger.
      • Ryzewicz M.
      • Wolf J.
      Trigger digits: principles, management, and complications.
      • Lindner-Tons S.
      • Ingell K.
      An alternative splint design for trigger finger.
      Participants were instructed to wear the orthosis day and night for a minimum of six weeks as supported in the literature.
      • Patel M.R.
      • Bassini L.
      Trigger fingers and thumb: when to splint, inject or operate.
      • Colbourn J.
      • Heath N.
      • Manary S.
      • Pacifico D.
      Effectiveness of splinting for the treatment of trigger finger.
      • Ryzewicz M.
      • Wolf J.
      Trigger digits: principles, management, and complications.
      • Lindner-Tons S.
      • Ingell K.
      An alternative splint design for trigger finger.
      Because continuous wearing of the orthosis was recommended, each participant completed a daily record of the number of hours the orthosis was worn, and an indication of whether or not it was a continuous period. The participants were told if they continued to have triggering at six weeks, the orthosis was to be worn ten weeks. The SST score and VAS pain score were obtained during the initial visit. An educational handout on the histopathology of trigger finger was provided to each participant. Participants were also given an exercise sheet with diagrams demonstrating passive IP joint flexion, composite full finger flexion, full finger extension, and active hook exercises.
      • Evans B.
      • Hunter J.
      • Burkhalter W.
      Conservative management of the trigger finger: a new approach.
      • Colbourn J.
      • Heath N.
      • Manary S.
      • Pacifico D.
      Effectiveness of splinting for the treatment of trigger finger.
      • Lindner-Tons S.
      • Ingell K.
      An alternative splint design for trigger finger.
      They were instructed to remove the orthosis to complete these exercises for five repetitions, three times a day. The participants were seen for a final follow-up visit ten weeks (M=10.39 weeks, SD=.99) after their initial visit. On the follow-up visit, the SST score was determined and VAS pain score was obtained.
      One-year postintervention, (M=53.1 weeks, SD=1.74), follow-up phone calls were made. The patient’s response to the question as to whether or not they required either surgical intervention or steroid injection for their trigger finger during the year was recorded. They also indicated whether or not they had any adverse response to the orthotic device.
      The primary outcome comparison for this study was the comparison of pre- and postorthosis SST and VAS pain scores. Statistical evaluation was performed using SPSS 16.0 (SPSS Inc., Chicago, IL, USA). Paired t-tests were used to compare SST and VAS pain scores.

      Data Analysis

      Change scores (the difference between the initial and final assessment) were calculated for all outcome measures and used in the data analysis. Paired t-tests were used to determine significant differences in the change score for each outcome measure. To control for experiment-wise error, all of the tests were evaluated at an alpha level of p<0.05. Effect size (ES) and standardized response mean (SRM) values were determined to calculate the effect of the orthosis intervention clinically. Sample size justification was unable to be performed because a formal hypothesis is required. Because this study was retrospective, a formal hypothesis was not used. Although the sample size could not be justified, the precision of the success rate was estimated with a confidence interval.

      Results

      Mean age of subjects was 68.48 years; 60% of the subjects were female, and 40% were males. Sixty-three trigger fingers were involved. Finger involvement was as follows: index 9.5%, long 41%, ring 27%, small 9.5%, and thumb 13%. Seventeen (37%) of patients had trigger finger of more than one digit (Table 3). Twenty-eight participants wore their orthosis six weeks and 18 wore their orthosis ten weeks in the study (Figure 4). Although the study population was a heterogeneous group, 12 participants (42%) had rheumatoid arthritis with 17 participants (60%) taking nonsteroidal anti-inflammatory drugs.
      Table 3Finger and Hand Involvement
      Subject Trigger Finger InvolvementN (Total 46)Percentage
      Involved hand left1840
      Involved hand right2860
      Involved digit
       Index69.5
       Long2641
       Ring1727
       Small69.5
       Thumb813
      Isolated digit involved2963
      Multiple digits involved1737
      N=number.
      Twenty-four participants (83%), who had isolated occurrence of trigger finger, reported that they wore their orthosis continuously day and night. Twelve participants (71%), who had multiple trigger fingers, reported continuous use of their orthosis day and night.
      The following prepost outcome measures were found to be significant in determining the efficacy of custom thermoplastic orthotic device intervention for trigger finger: SST and VAS pain scale (Table 4). The efficacy of orthotic device intervention was further established through the determination of ES and SRM values for the change scores. Figure 5 presents the pre- and posttreatment frequency distributions for the variable of SST measured on a six-point scale. Initially, the symptoms of all 46 participants were rated as stage 3 (triggering) or higher. The follow-up results indicated that 45 (98%) participants received a grade of stage 1 or 2. The greatest improvement was by four stages. The mean SST score preorthosis was 3.93 and postorthosis was 1.22. The results were analyzed using a pair-wise t-test, which demonstrated a statistical difference for the pre- to postscores for SST (t=31.58; df=45; p<0.001). The results shown in Figure 3 indicate an average improvement of 2.72±0.58. The ES of the SST change score is 0.97 (Cohen’s d 7.84) and the SRM score is 0.96 (Cohen’s d 6.57), which demonstrate a large responsiveness effect using Cohen’s guidelines for interpretation of ES scores.
      Table 4Paired t-Test Examining the Effects of Orthotic Device Intervention on SST and Pain Change Scores from Initial Treatment and Ten Weeks
      Measurement InstrumenttMean Change ScoreStandard DeviationStandard Error MeanLowerUpperdfLevel of SignificanceEffect SizeSRM
      SST score31.582.720.580.092.542.89450.001
      Indicates results were statistically significant.
      0.970.96
      VAS pain score26.654.451.130.174.104.77450.001
      Indicates results were statistically significant.
      0.920.94
      t=t-Test; df=degrees of freedom; SRM=standardized response mean; SST=stages of stenosing tenosynovitis; VAS=Visual Analog Scale.
      Indicates results were statistically significant.
      Figure thumbnail gr5
      Figure 5Visual Analog Scale (VAS) pain rating score’s frequency distribution.
      An evaluation of the VAS pain score was based on pre- and postmeasures of the participants’ rating of hand pain using an 11-point scale ranging from 0=no pain to 10=the most pain the participant has felt. The frequency distributions for the pre- and postratings of pain are presented in Figure 6. Thirty-seven participants (80%) rated their pain as greater than five in the pretreatment rating; no participant rated their pain above four in the posttreatment rating of pain. The statistical comparison of the pre- to post-VAS pain score revealed a significant decrease in pain rating by the participants (t=26.65; df=45; p<0.001). The average improvement in VAS was 4.43±1.13. All 46 participants reported a decrease in the rating of pain. The mean pain score preorthosis was 5.63 and the mean postorthosis pain score was 1.19. The ES of the pain change score is 0.92 (Cohen’s d 4.57) and the SRM score is 0.94 (Cohen’s d 5.59), which demonstrate a large responsiveness effect using Cohen’s guidelines for interpretation of ES scores.
      Figure thumbnail gr6
      Figure 6Stages of stenosing tenosynovitis (SST) rating score’s frequency distribution.
      There was an 87% (40 patients) success rate with orthosis application as determined by the number of participants that required either injection or surgical intervention in the year after orthotic device application (Table 5). This estimate has reasonable precision. The proportion is 85% and the width of the confidence interval is ±10% (that is, 75–95%). Two patients (4.3%) had A1 pulley release surgery and four patients (8.5%) had steroid injections in the year after orthotic device application (Table 6). None of the patients reported any adverse side effects from their orthosis after their discharge from therapy.
      Table 5Results at One-year Follow-up
      Interventions received during the past yearNumberPercentileProportionConfidence Interval of the Difference
      Orthosis sole intervention4087
      Precision of success estimate0.85Lower 0.75Upper 0.95
      Surgery24.3
      Injection48.7
      Table 6Digits Involved Requiring Further Intervention in the Year after Orthosis Intervention
      Digit InvolvementType of Orthosis
      Injection
       LongSingle digit
       LongSingle digit
       LongSingle digit
       Long, ringHand based
      Surgery
       RingSingle digit
       Long, ringHand based

      Discussion

      This study evaluated the long-term efficacy of orthotic device intervention for trigger finger and found that 87% of the participants did not require further intervention in the year after orthosis application. These findings outperform the findings of Rozental et al.
      • Rozental T.D.
      • Zurakowski D.
      • Blazar P.E.
      Trigger finger: prognostic indicators of recurrence following corticosteroid injection.
      that reported 56% of the trigger digits treated with corticosteroid injection had a recurrence of symptoms at a median of 5.6 months after the injection. Huisstede et al.
      • Huisstede B.M.
      • van Middelkoop M.
      • Randsdorp M.S.
      • Glerum S.
      • Koes B.W.
      Effectiveness of interventions of specific complaints of the arm, neck, and/or shoulder: 3 musculoskeletal disorders of the hand. An update.
      also found moderate evidence to support the effectiveness of steroid injections for trigger finger in the short-term (one to four weeks) but not for long-term outcomes. The goal of orthotic device intervention is to prevent the friction caused by flexor tendon movement through the affected A1 pulley until the inflammation there resolves and this may provide a better long-term solution than a steroid injection. This is in agreement with Salim et al.
      • Salim N.
      • Abdullah S.
      • Sapuan J.
      • Haflah N.H.
      Outcome of corticosteroid injection versus physiotherapy in the treatment of mild trigger fingers.
      that concluded that physiotherapy may have a role in the prevention of recurrence of trigger finger.
      Colbourn et al.
      • Colbourn J.
      • Heath N.
      • Manary S.
      • Pacifico D.
      Effectiveness of splinting for the treatment of trigger finger.
      reported that 26 participants (92.9%) felt that their triggering had improved after orthosis intervention. Colbourn et al. did not perform a long-term follow-up to determine if their subjects had a recurrence of symptoms in the year after their orthosis intervention. Ninety-eight percent of the participants in this study achieved a triggering SST score of 1 or 2 after the orthosis intervention, and 87% of the subjects required no further intervention in the year after orthosis intervention. Colbourn et al. studied the isolated incidence of trigger finger and this study included the incidence of trigger thumb, isolated trigger finger, and multiple trigger fingers. The benefits of orthosis application for trigger finger can be generalized to a larger population that includes multiple incidence of trigger finger and trigger thumb based on the results of this study.

      Limitations of the Study

      The use of one protocol and one therapist eliminated bias, but also reduced the ability to generalize the results of this study to a larger population. Small sample size also reduces the ability to generalize the results of this study to a larger population.
      Tung et al.
      • Tung W.
      • Kuo L.
      • Lai K.
      • Jou I.
      • Sun Y.
      • Su F.
      Quantitative evidence of kinematics and functional differences in different graded trigger fingers.
      reported that Quick Disabilities of the Arm, Shoulder, and Hand (DASH) scores demonstrated that there are significant differences in functional performance among individuals with different trigger finger grades. The patients in this study did not use a patient-centered outcome measure to measure change in function. This study provides no evidence that function improves with orthotic device intervention. Unfortunately, there is not a consistent evaluation technique for trigger finger. There are numerous outcome measures and classification systems used in studies that evaluate trigger finger injections and surgical releases. Some studies have used the DASH questionnaire’s
      • Ring D.
      • Lozano-Calderon S.
      • Shin R.
      • Bastian P.
      • Mudgal C.
      • Jupiter J.
      A prospective randomized controlled trial of injection of dexamethosone versus triamcinolone for idiopathic trigger finger.
      various pain scales, the Wolfe scale,
      • Wolfe S.W.
      Tenosynovitis.
      • Lim M.H.
      • Lim K.K.
      • Rasheed M.Z.
      • Narayanan S.
      • Tan A.B.
      Outcome of open trigger digit release.
      Quinnell grade,
      • Ring D.
      • Lozano-Calderon S.
      • Shin R.
      • Bastian P.
      • Mudgal C.
      • Jupiter J.
      A prospective randomized controlled trial of injection of dexamethosone versus triamcinolone for idiopathic trigger finger.
      • Quinnell R.C.
      Conservative management of trigger finger.
      a nominal scale for frequency of triggering,
      • Peters-Veluthamaningal C.
      • Winters J.C.
      • Groenier K.H.
      • Meyboom-de Jong B.
      Corticosteriod injections effective for trigger finger in adults in general practice: a double-blinded randomized placebo controlled trial.
      and subject perceived improvement scales.
      • Ring D.
      • Lozano-Calderon S.
      • Shin R.
      • Bastian P.
      • Mudgal C.
      • Jupiter J.
      A prospective randomized controlled trial of injection of dexamethosone versus triamcinolone for idiopathic trigger finger.
      This fact limits the ability to compare trigger finger study outcomes to one another.
      SST
      • Patel M.R.
      • Bassini L.
      Trigger fingers and thumb: when to splint, inject or operate.
      is a numbered classification system (1–6) used to describe the extent of tendon pathology based on objective physical findings. The data uses an ordinal scale and its validity and reliability has not yet been determined.

      Conclusion

      This study demonstrated a clinically significant positive relationship between the use of a custom thermoplastic orthosis and the subjective reduction of pain and objective SST outcome measure for patients who have isolated trigger finger, multiple trigger fingers, or trigger thumb. It also demonstrated an 87% success rate among patients who required no further intervention in the year after orthosis application. Custom-orthosis treatment is a low-cost, easily applied intervention that may provide patients with long-term relief of trigger finger.

      References

        • Makkouk A.L.
        • Oetgen M.E.
        • Swigart C.R.
        • Dodds S.D.
        Trigger finger: etiology, evaluation, and treatment.
        Curr Rev Musculoskelet Med. 2008; 2: 92-96
        • Akhtar S.
        • Bradley M.J.
        • Quinton D.N.
        • et al.
        Management and referral for trigger finger/thumb.
        BMJ. 2005; 331: 30-33https://doi.org/10.1136/bmj.331.7507.30
        • Evans B.
        • Hunter J.
        • Burkhalter W.
        Conservative management of the trigger finger: a new approach.
        J Hand Ther. 1988; 2: 59-68
        • Patel M.R.
        • Bassini L.
        Trigger fingers and thumb: when to splint, inject or operate.
        J Hand Surg Am. 1992; 17: 110-113
        • Colbourn J.
        • Heath N.
        • Manary S.
        • Pacifico D.
        Effectiveness of splinting for the treatment of trigger finger.
        J Hand Ther. 2008; 21: 336-343
      1. The International Classification of Diseases: 9th Revision: Clinical Modification. ICD.9CM.
        Public health service, Bethesda, MD1980
        • Ryzewicz M.
        • Wolf J.
        Trigger digits: principles, management, and complications.
        J Hand Surg Am. 2006; 31: 135-146
        • Lindner-Tons S.
        • Ingell K.
        An alternative splint design for trigger finger.
        J Hand Ther. 1998; 11: 206-208
        • Rozental T.D.
        • Zurakowski D.
        • Blazar P.E.
        Trigger finger: prognostic indicators of recurrence following corticosteroid injection.
        J Bone Joint Surg Am. 2008; 9: 1665-1672
        • Huisstede B.M.
        • van Middelkoop M.
        • Randsdorp M.S.
        • Glerum S.
        • Koes B.W.
        Effectiveness of interventions of specific complaints of the arm, neck, and/or shoulder: 3 musculoskeletal disorders of the hand. An update.
        Arch Phys Med Rehabil. 2010; 91: 298-314
        • Salim N.
        • Abdullah S.
        • Sapuan J.
        • Haflah N.H.
        Outcome of corticosteroid injection versus physiotherapy in the treatment of mild trigger fingers.
        J Hand Surg Eur. 2011; 8 (doi:10.1177/1753193411415343): 4
        • Tung W.
        • Kuo L.
        • Lai K.
        • Jou I.
        • Sun Y.
        • Su F.
        Quantitative evidence of kinematics and functional differences in different graded trigger fingers.
        Clin Biomech. 2010; 25: 535-540
        • Ring D.
        • Lozano-Calderon S.
        • Shin R.
        • Bastian P.
        • Mudgal C.
        • Jupiter J.
        A prospective randomized controlled trial of injection of dexamethosone versus triamcinolone for idiopathic trigger finger.
        J Hand Surg Am. 2008; 33: 516-522
        • Wolfe S.W.
        Tenosynovitis.
        in: Green D.P. Operative Hand Surgery. 5th ed. Churchill Livingstone, New York2005: 2137-2159
        • Lim M.H.
        • Lim K.K.
        • Rasheed M.Z.
        • Narayanan S.
        • Tan A.B.
        Outcome of open trigger digit release.
        J Hand Surg Eur Vol. 2007; 32: 457-459
        • Quinnell R.C.
        Conservative management of trigger finger.
        Practitioner. 1980; 224: 187-190
        • Peters-Veluthamaningal C.
        • Winters J.C.
        • Groenier K.H.
        • Meyboom-de Jong B.
        Corticosteriod injections effective for trigger finger in adults in general practice: a double-blinded randomized placebo controlled trial.
        Ann Rheum Dis. 2008; 67: 1262-1266