ankle sprains and in patients with unctional instability.This systematic review was perormed to collect evidenceto update the Clinical Practice Guideline Ankle Injury o theRoyal Dutch Society or Physical Therapy (KNGF).
Two reviewers (TL and PvdW) searchedcomputerised databases independently. The ollowingdatabases were searched: MEDLINE (1966 to March2005), EMBASE (1988 to March 2005), CINAHL (1982to March 2005). In addition the Cochrane Central Registero Controlled Trials (2005, Issue 1), the PEDro database (toMarch 2005) and the DocOnline database (to March 2005)o the Dutch Institute o Allied Health Proessions weresearched. Subject-specic search was based on combinationso ‘ankle’, ‘sprains’, ‘injuries’, ‘prevention’, ‘ligamentous’,‘lateral’, ‘unctional instability’, ‘rehabilitation’,‘physiotherapy’, ‘physical therapy’. Finally, reerencesrom retrieved articles were screened.
Types of studies
This review includes randomised controlledtrials. Full text articles until March 2005, published inEnglish, German or Dutch, were considered or this study. Todetermine whether a study should be included, the abstractso all identied articles were assessed by both reviewers. I there was any doubt, the ull text article was retrieved andread independently by both reviewers. Disagreement wasresolved by consensus.
Types of participants
Trials that include patients with acuteankle sprain, or with unctional instability, or trials with arecognisable subgroup with a history o ankle injury wereconsidered. In acute ankle sprain symptoms may be pain,swelling and unctional disability. Functional instabilityincludes residual problems ater acute injury such asrecurrent sprains, a eeling o giving way or a eeling o apprehension.
Types of interventions
At least one o the interventionsin the trial had to be an intervention aimed at exercisetherapy (including proprioceptive training, co-ordinationtraining, strength training or unctional exercises), or atmanual mobilisation o the ankle joint. The interventionhad to be compared with placebo, no treatment or otherinterventions.
At least one o the ollowing outcomemeasures had to be used or inclusion in this study: recurrentsprains, unctional disability, gait pattern, subjectiveinstability, postural control, ankle joint range o motion,pain.
Two reviewers (PvdW and TL) assessedthe methodological quality independently. A slightlymodied version o the Amsterdam-Maastricht consensuslist (Smidt et al 2002) was used to assess the quality o the internal validity o the studies. The list o criteria orassessment o the methodological quality is shown in Table1. The reviewers scored each item with a ‘Yes’ (sucientinormation is available and bias is considered to beunlikely), ‘No’ (bias was considered to be likely), or ‘Don’tknow’ (insucient inormation is given, the criterion israted as inconclusive). Positive scores (Yes) were added.A study was considered high quality when it had a minimumscore o 4. Disagreement was ollowed by discussion,ollowed i necessary by scrutiny rom another reviewer(EH).
Two reviewers (PvdW and TL) extractedthe data independently rom the studies using a standarddata-extraction orm and cross checked or accuracy.Disagreement was resolved by a consensus procedure,ollowed i necessary by scrutiny rom a third reviewer(EH). When appropriate and possible, additional data wereobtained rom authors o the studies.
Criteria or assessment o methodological qualityo studies. Adequate randomisation: adequate procedure orgeneration o a random number sequence.2 Baseline similarity: treatment and control group arecomparable at entry. Co-intervention: standardised or avoided. Adherence: > 70% in intervention and control group(s),5 Blinding o the therapist.6 Blinding o the patient.7 Withdrawals and drop outs: < 20% or short termollow up; and < 0% or intermediate and long termollow-up.8 Identical timing o outcome measures or all groups.9 Intention to treat analysis.0 Blinding o the outcome assessors.Based on the Amsterdam-Maastricht consensus list (Smidtet al 2002).
Hierarchy o quality o individual studies andstrength o evidence.Hierarchy o evidenceA Systematic reviews, which include trials at quality levelA2, and have consistent resultsA2 Randomised controlled trials o good quality andsucient power and consistencyB Randomised controlled trials o moderate quality orinsucient power, or other non-randomised controlledstudiesC Non-controlled studiesD Expert opinion, such as working group members
Strength of evidence
Level systematic review (A) or 2 studies at level A2Level 2 2 studies o level BLevel study o level A2 or B or CLevel Expert opinion, such as working group members
Formulation of recommendations
Level It has been shown that …Level 2 It is likely that …Level There are indications that …Level The opinion o the working group is …Source: CBO 2005
Australian Journal of Physiotherapy 2006 Vol. 52 – © Australian Physiotherapy Association 200628
van der Wees et al: Systematic review of treatments for ankle sprain