Gosens 2011.

Study characteristics
Methods	Design: multi‐centre (2 sites) parallel‐group 2‐arm double‐blind randomised controlled trial Setting: St. Elisabeth Hospital, Tilburg, The Netherlands, and Haga Hospital, The Hague, The Netherlands Timing: May 2006 to January 2008. Interventions: platelet‐rich plasma injection vs glucocorticoid injection Sample size: 42 participants required per group to detect a 28% difference between groups in the proportion with treatment success (25% reduction in VAS pain score), based on 2‐sided type 1 (alpha) error rate of 0.05 and power of 90%. To account for loss to follow‐up, a minimum of 50 participants were randomised per group Analysis: intention‐to‐treat
Participants	Number of participants Number of participants screened for eligibility: 106 Number excluded: 6 (did not meet inclusion criteria) Number randomised: 100 (51 to platelet‐rich plasma group, 49 to glucocorticoid group) Number included in analysis: 100 (51 in platelet‐rich plasma group, 49 in glucocorticoid group). There were 6 re‐interventions in the PRP group: 3 had an operation, 3 had re‐injection with glucocorticoids. There were 14 re‐interventions in the glucocorticoid group: 6 had an operation, 1 had re‐injection with glucocorticoids, and 7 crossed over to the PRP group Inclusion criteria Minimum 18 years of age Pain over lateral epicondyle on direct palpation and pain in that area during resisted wrist extension Minimum symptom duration of 6 months Minimum pain score of 50 on a VAS scale (0 to 100) Exclusion criteria Age < 18 years Pregnancy Glucocorticoid injection or surgical treatment for lateral epicondylitis in the past 6 months History of carpal tunnel syndrome History of cervical radiculopathy Systemic disorder such as diabetes mellitus, rheumatoid arthritis, or hepatitis Baseline characteristics Platelet‐rich plasma group Mean (SD) age (years): 46.8 (8.5) No. male:female: 23:28 Mean (SD) DASH (0 to 100; 0 is no disability): 56.3 (17.7) Mean (SD) VAS pain score (0 to 100; 0 is no pain): 70.2 (15.2) No. (%) dominant arm affected: 38 (74.5) Glucocorticoid group Mean (SD) age (years): 47.3 (7.8) No. male:female: 23:26 Mean (SD) DASH (0 to 100; 0 is no disability): 44.1 (16.2) Mean (SD) VAS pain score (0 to 100; 0 is no pain): 67.1 (13.5) No. (%) dominant arm affected: 37 (75.5) Pretreatment group differences: DASH scores were higher in the PRP group than in the glucocorticoid group
Interventions	Platelet‐rich plasma injection Whole blood (27 mL) was drawn from the contralateral arm into a 30‐mL syringe containing 3 mL of sodium citrate, and the platelet‐rich fraction was isolated in the Biomet Recover system; 3 mL of PRP was obtained and buffered to physiologic pH; 8.4% sodium bicarbonate and bupivacaine hydrochloride 0.5% with adrenaline (1:200,000) was added. No activating agent was used. Approximately 1 mL of PRP was injected directly into the area of maximum tenderness (from a covered syringe to maintain blinding), and the remaining 2 mL was injected with a 22‐gauge needle and a peppering technique into the common extensor tendon via a single skin portal and 5 penetrations of the tendon. A supervised orthopaedic resident or a consultant performed all injections in both groups Glucocorticoid injection Whole blood (27 mL) was drawn from the contralateral arm and discarded; 1 mL of glucocorticoid (kenacort 40 mg/mL; triamcinolone acetonide) with bupivacaine hydrochloride 0.5% with adrenaline (1:200,000) was injected directly and 2 mL peppered, via the same injection procedure as in the PRP treatment group Post intervention Participants were kept supine without moving the arm for 15 minutes immediately after the injection, then were instructed to rest the arm for 24 hours. Paracetamol was permitted for pain relief, but non‐steroidal anti‐inflammatory medications were prohibited. A standardised stretching protocol was followed for 2 weeks under physiotherapy supervision, followed by eccentric muscle and tendon strengthening. Participants were permitted to return to normal sporting activities after 4 weeks
Outcomes	Outcomes were reported at baseline and at 4 weeks, 8 weeks, 12 weeks, 26 weeks, 52 weeks, and 104 weeks Outcomes Mean pain measured on a visual analogue scale (scale from 0 (no pain) to 100 (maximum pain)) Mean disability measured on Disabilities of the Arm, Shoulder and Hand questionnaire (scale from 0 (no disability) to 100 (most severe disability)) Treatment success as defined by 25% reduction in VAS or DASH score without a re‐intervention after 1 years and after 2 years Adverse events Outcomes included in this review Mean pain on a visual analogue scale Mean disability measured on Disabilities of the Arm, Shoulder and Hand questionnaire Treatment success as defined by 25% reduction in VAS score without a re‐intervention after 2 years Adverse events Time points included in this review 4 weeks, 12 weeks, 26 weeks, 1 year, and 2 years
Notes	Funding: Biomet sponsored the study by supplying the Recover system at a discounted rate; study authors declared that Biomet did not have any influence on collection and analysis of data for this study Trial registration: http://www.clinicaltrials.gov trial identifier 2007‐004947‐31 Withdrawal: 3/49 in the glucocorticoid group and 3/51 in the PRP group were lost to follow‐up Adverse events PRP group Total adverse events No. (%) of events: not reported Nature of the event: "no complications were seen concerning the use of PRP, except for the initial worsening of pain because of the activation of the inflammation cycle, which usually lasted for 1 to 2 weeks" Serious adverse events None Glucocorticoid group Total adverse events: none reported Serious adverse events: none reported
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	A computer using block randomisation of 10 participants was used to create a randomisation schedule
Allocation concealment (selection bias)	Low risk	Allocations were placed in sequentially numbered opaque envelopes and were assigned by trial managers
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Blood was taken from both groups to blind participants. Tubes were masked with opaque tape
Blinding of outcome assessment (detection bias) for self reported subjective outcomes (pain, function, treatment success, quality of life)	Low risk	Participants were blinded to treatment assignment; hence there is low risk of bias in the measurement of pain and function
Blinding of outcome assessment (detection bias) objective outcomes	Low risk	No objective outcomes were measured in this study
Incomplete outcome data (attrition bias) All outcomes	Low risk	3/49 in the glucocorticoid group and 3/51 in the PRP group were lost to follow‐up, and 'last observation carried forward' was planned for any missing data
Selective reporting (reporting bias)	Low risk	Trial was registered and all outcomes planned in the protocol were reported
Other bias	Unclear risk	There is risk of bias due to contamination of results caused by re‐interventions (6 re‐interventions in the PRP group ‐ 3 underwent surgery and 3 had re‐injection of glucocorticoids; 5 out of these 6 re‐interventions were performed in the first year of follow‐up; 14 re‐interventions in the glucocorticoid group ‐ 6 underwent surgery and 1 required re‐injection with glucocorticoids every 3 months and declined surgery, 7 crossed over to the PRP group; 13 out of these 14 re‐interventions were performed in the first year of follow‐up)