Table 1

Methodological characteristics, inclusion and exclusion criteria, and follow-up completion rates of the included studies

AuthorStudy typeRandomisation methodBlinding methodAllocation concealmentStatistical power calculationBaseline comparisonInclusion criteriaExclusion criteriaFollow-up completion (%)
Gatz et al37RCT.Numbered envelopes.None.Sealed envelopes.Yes; sample size adequate for 82% power.No difference.≥18 years, insertional or midportion Achilles tendinopathy diagnosed with history and examination (provoked pain by palpation), physical ability to perform exercises.Pregnancy, very high (>35) or very low (<17) BMI, previous rupture or operation in area of symptoms, injections in last 6 months.71
Clifford et al17Pilot RCT.Random selection of sealed, opaque envelopes from box.None.Sealed, opaque envelopes.No.No difference.≥18 years, lateral hip pain >3 months, pain over greater trochanter and one of the following: (1) single leg stand 30 s, (2) FADER, (3) FADER and resisted IR, (4) FABER, and (5) resisted hip abduction at end-range adduction.Physiotherapy in previous 6 months, corticosteroid injection in previous 3 months, hip or lumbar spine surgery in previous 12 months, moderate to severe hip osteoarthritis.77
Holden et al22RCT single-blind (cross-over).Computer-generated allocation sequence.Researcher blinded to sequence allocation.
Participants blinded to study hypothesis.
Sealed, opaque envelopes, sequence randomised by independent researcher.Yes; sample size adequate for 90% power.No difference.18–40 years of age, patellar tendinopathy clinically and radiologically.Concurrent knee pathologies, previous knee surgery, corticosteroid injection in last 6 months.95
Vuvan et al34RCT.Computer-generated.Not stated.Sealed, opaque envelopes.Yes; sample size adequate for 80% power.No difference.18–70 years, symptoms for >6 weeks, NPRS ≥2/10, at least two of the following: (1) gripping, (2) palpation of lateral epicondyle, (3) stretching of forearm muscles, (4) resisted wrist, second or third finger extension, and (5) reduced pain-free grip strength.Other sources of elbow pain, major neurological, inflammatory or systemic conditions, treatment by a healthcare practitioner within the previous 3 months, injections within the preceding 6 months, or major trauma, fracture or surgery in the last year.98
Dupuis et al31RCT single-blind.Random number generator and block design.Blinded outcome assessment (participants asked not to discuss their treatment with assessor).Sealed, opaque envelopes.Yes; sample size adequate for 80% power.Cryotherapy group older.16–65 years, symptoms <6 weeks, painful arc of movement, positive Neer’s or Kennedy-Hawkins and pain on resisted isometric external rotation, abduction, or positive Jobe’s all present.Upper limb fracture, previous neck or shoulder surgery, cervical spine involvement, frozen shoulder, sign of full cuff tear, rheumatological, inflammatory or neurological disease.77
Parle et al32RCT single-blind pilot.Random number generator.Chief investigator and sonographer blinded to groups.Sealed, opaque envelopes.No.Not done.Unilateral shoulder pain <12 weeks, symptoms aggravated with active or resisted movement, unaccustomed increase in shoulder activity preceding symptoms, evidence of bursitis or tendinosis on ultrasound.Dominant biceps pain, frozen shoulder, full thickness or large partial thickness tears, and traumatic onset of pain.100
Rio et al35RCT single-blind.Random number generator function of Microsoft Excel.Not stated.Unmarked, opaque envelopes.No.No difference.Elite and subelite volleyball and basketball athletes over 16 years of age.Existence of other knee pathology, previous patellar tendon rupture, previous patellar tendon surgery, inflammatory disorders, metabolic bone diseases and type II diabetes, use of fluoroquinolones or corticosteroids in the last 12 months, known familial hypercholesterolaemia and fibromyalgia.62
Diagnosis made clinically and radiologically.
Stasinopoulos and Stasinopoulos33RCT single-blind.‘By drawing lots’.Outcome assessor blinded to groups.Not stated.No.No difference.Amateur tennis athletes.Dysfunction in the shoulder, neck (radiculopathy) and/or thoracic region, local or generalised arthritis, neurological deficit, radial nerve entrapment, limitations in arm functions, the affected elbow had been operated on, had received any conservative treatment in the 4 weeks before entering the study.100
Clinical diagnosis of lateral elbow tendinopathy for at least 4 weeks.
van Ark et al36RCT single-blind.Random number generator function of Microsoft Excel.Not stated.Unmarked, opaque envelopes.No.No difference.Elite and subelite volleyball and basketball athletes over 16 years of age.Existence of other knee pathology, previous patellar tendon rupture, previous patellar tendon surgery, inflammatory disorders, metabolic bone diseases and type II diabetes, use of fluoroquinolones or corticosteroids in the last 12 months, known familial hypercholesterolaemia and fibromyalgia.62
Diagnosis made clinically and radiologically.
Rio et al20RCT single-blind (cross- over).Participants chose an envelope for order of intervention.Not stated.Unmarked, opaque envelopes.No.No difference.Volleyball players who were taking no medications.
Diagnosis made clinically and radiologically.
100
  • BMI, body mass index; FABER, flexion abduction external rotation; FADER, flexion adduction external rotation; IR, internal rotation; NPRS, Numerical Pain Rating Scale; RCT, randomised controlled trial.