The aim of the study is to determine “the additional effect of... function” for patellofemoral pain syndrom (PFPS). The additional effect of orthotic devices over exercise therapy on pain and function. A systematic literature search was conducted in MEDLINE, CINAHL, EMBASE, Cochrane and PEDro. Randomised controlled trials and controlled clinical trials of patients diagnosed with PFPS evaluating a clinically relevant outcome were included. Treatment had to include exercise therapy combined with orthotics, compared with an identical exercise programme with or without sham orthotics. Data were summarised using a best evidence synthesis. Eight trials fulfilled the inclusion criteria, of which three had a low risk of bias. There is moderate evidence for no additive effectiveness of knee braces to exercise therapy on pain (effect sizes (ES) varied from −0.14 to 0.04) and conflicting evidence on function (ES −0.33). There is moderate evidence for no difference between knee braces and exercise therapy versus placebo knee braces and exercise therapy on pain and function (ES −0.1–0.10). More studies of high methodological quality are needed to draw definitive conclusions.
Statistics from Altmetric.com
Patellofemoral pain syndrome (PFPS) is a common diagnosis in young adults and adolescents. The exact incidence of PFPS in the general population has not been properly evaluated.1 In a military population, the overall risk of PFPS is 3%, with an incidence rate of 22 injuries/1000 person-years.2 In an athlete population, the cumulative incidence risk and rate for the development of new unilateral PFPS is 9.66 per 100 athletes and 1.09 per 1000 athletic exposures, respectively.3 PFPS is defined by a complex of symptoms in which pain around the patella is the most dominant, and the complaint is associated with activities that load the patellofemoral joint, such as squatting, stair ascending and descending, walking, running and jumping.4 Patients with PFPS are often highly limited in physical activities due to pain. It is suggested that the cause of PFPS is a combination of proximal, distal and local factors that influence the movement of the patella within the trochlea of the femur.5
Patients with PFPS are often referred to a physical therapist. There is evidence that some interventions applied by physical therapists have significant beneficial effects on pain and function compared with no treatment.4 6 Treatment strategies that are applied in clinical practice include exercise therapy to increase muscle strength, improve neuromuscular coordination and enhance flexibility. Occasionally, orthotic devices are added to the treatment to relieve pain, and it is suggested that orthotics correct for possible malalignment.
Up to now, no systematic review has been published on the additional value of orthotics on exercise therapy. Therefore, the aim of this study was to determine the effectiveness, in measures of pain and function, of a physiotherapeutical intervention consisting of exercise therapy and orthotic devices or exercise therapy and placebo orthotics compared with exercise therapy only for patients with PFPS.
Orthotic devices in this review include patellar bracing, patellar taping and foot orthotics.
A literature search was conducted in MEDLINE, CINAHL, EMBASE, Cochrane and PEDro by two independent researchers (MvM and NMS). Wherever possible, the Cochrane Library search filter for randomised controlled trials (RCT)and controlled clinical trials (CCT) was applied to make a restriction for the design. If it was impossible to use the filter in the database, the specific limits options were used for RCT and CCT. Studies were collected from 1990 up to January 2010 with a language restriction for English, German and Dutch studies (online appendix 1). Finally, the reference lists of the included articles were searched for more relevant articles.
Criteria for considering studies for this review
Studies were included if they met the following criteria: (1) RCT or CCT. (2) The study population had to include patients diagnosed with PFPS or anterior knee pain. (3) The intervention had to consist of exercise therapy aiming at muscle strengthening and stretching exercises, combined with foot or knee orthotics (including tape, braces and insoles). The control group had to receive an identical exercise programme with or without sham orthotics. (4) The outcome measures assessed were pain and function either reported by the patient using a questionnaire or by the assessor using an objective performance measure of the knee.
Studies were excluded when they did not fulfil the above-mentioned inclusion criteria on design, intervention and outcome.
Two review authors (NMS, MvM) working independently from one another examined all citations (including titles and abstracts) from the electronic search. Full articles were obtained for those citations thought to fulfil the inclusion criteria. A third reviewer (SMAB-Z) was consulted if consensus was not reached.
Assessment of risk of bias in included studies
The risk of bias assessment was conducted using the 12 criteria recommended by the Cochrane Back Review Group and evaluated independently by two researchers (RvL and NMS) (online appendix 2). The 12 questions were answered with ‘yes’, ‘no’ or ‘unclear’. If there was a difference in the scores between the assessors, agreement was reached after discussion. Studies with six or more positive items were considered to have a low risk of bias. This cut-off point is supported by empirical evidence.7
Agreement between the two authors was calculated by Cohen's κ.8 Values of κ between 0.40 and 0.59 have been considered to reflect fair agreement, between 0.60 and 0.74 to reflect good agreement and 0.75 or more to reflect excellent agreement.9
Data extraction was performed independently by two researchers (SMAB-Z and NMS). The following data were extracted: study design, baseline characteristics, duration of the complaints, duration and specific details of the intervention, outcome measures and results. Data were extracted for short (0−12 weeks) and long-term (>12 weeks) outcomes. The change scores over time were extracted from the studies or in the absence calculated for the orthotic (O) and the control (C) groups. Subsequently, the differences in change scores between O and C were extracted or calculated, together with a 95% CI. If the raw data were not presented in the study, data were extracted from the figures. The authors were not contacted for data that were not provided. If studies do not provide enough information to calculate the 95% CI, information about significant differences between the groups is abstracted from the studies. Effect sizes (ES) were calculated by dividing the mean difference between O and C by the pooled SD of the baseline scores, when available.10 Cohen labelled an ES small if d=0.20, medium if d=0.50 and large if d=0.80.11
A qualitative data analysis was applied using a best evidence synthesis, consisting of five levels to assess the power of the results. A modified version recommended by the Cochrane Back Review Group12 was used and outlined below:
▶ Strong evidence—consistent findings among multiple low risk of bias RCT (vonsistency: ≥75% of the trials report the same findings).
▶ Moderate evidence—consistent findings among multiple high risk of bias RCT and/or CCT and/or one low risk of bias RCT.
▶ Limited evidence—one high risk of bias RCT and/or CCT or consistent findings among multiple CCT.
▶ Conflicting evidence—inconsistent findings among multiple RCT and/or CCT.
▶ No evidence from trials—no RCT or CCT could be found.
Information about the number of studies identified from the databases, included and excluded for analysis is shown in figure 1. The search retrieved 269 articles leading to 153 unique articles. Five studies were identified from a hand search of the reference lists. From the 158 publications, 142 studies were excluded based on the title and abstract. Eight studies were excluded based on the full text. The reason for exclusion was a design other than a RCT or CCT, the absence of an intervention consisting of exercise therapy combined with foot or knee orthotics and the absence of outcome measures for pain and function. Finally, seven RCT and one CCT were included and used for analysis.13,–,20 Of the eight included studies, two compared an exercise and orthotics group with both an exercise-only and an exercise and placebo orthotics group. The results of the placebo orthotics group were analysed separately.
Risk of bias in included studies
The interrater reliability of the risk of bias assessment was good (κ 0.73).
The risk of bias of the eight included studies was ‘low’ in three studies13 14 18 and ‘high’ in five studies15,–,17 19 20 (table 1). Six studies used an adequate method of randomisation; in one study the method of randomisation was unclear. In four studies the treatment allocation was concealed. In none of the studies was the care provider, the patient or the outcome assessor blinded to the intervention. In three studies the drop-out rate was described and acceptable, and seven studies analysed the patients in the group to which they were allocated.
Characteristics of included studies
The characteristics of the included studies are described in table 2. The studies included 325 participants, of which 50.3% were women and the mean age of the population was 25.8 years. The duration of complaints varied between the studies from 2 weeks to 15 years. Outcome measures used for pain were: the verbal pain score15 and the visual analogue scale (VAS) for pain at rest or at different activities.21 Questionnaires used to assess function were the Kujala patellofemoral score (KPS),22 the knee function,18 the functional index questionnaire (FIQ)23 and the Western Ontario and McMaster Universities osteoarthritis questionnaire.24 One study also used an additional step test to measure the performance of the knee. Measurements were performed at different time points ranging from 2 weeks to 1 year.
The exercise therapy programmes that were applied are described in table 3. A variety of interventions was used for muscle strengthening and balance or proprioception training. All studies used quadriceps stretching techniques, several used hamstring and iliotibial band, gastrocnemius and hip flexors stretching techniques. In one study the exercise intervention consisted of home exercises only. Four studies appended a home exercise programme to the supervised exercise therapy. One study used a multimodal approach of exercise therapy, patellar mobilisation and patellar taping.14 In addition to the exercise therapy foot or knee orthotics were added to the treatment.
Effectiveness of interventions
Due to the clinical heterogeneity of interventions, outcome measures and time to follow-up, pooling of data from the included studies was impossible.
Knee braces and exercise therapy versus exercise therapy only
One low risk of bias study18 and two high risk of bias studies15,19 described the additional effect of knee braces on exercise therapy at short term (table 2). Two studies did not provide enough information to calculate effect sizes or 95% CI.15 19 Information about significant differences between the groups is abstracted from the studies.
The study of Miller et al19 described three groups, of which two groups used different knee braces. Group 1 used a Palumbo dynamic patellar brace (DynOrthotics, Vienna, Virginia, USA) in addition to exercise therapy. The aim of the dynamic patellar brace is to give an active, medially displacing force on the lateral border of the patella, maintaining constant pressure during flexion, extension and rotation of the knee.25 Group 2o used the Cho-Pat knee strap (Cho-Pat, Hainesport, New Jersey, USA) in addition to exercise therapy. The strap functions dynamically as the knee bends and straightens and improves tracking and assists in spreading pressure uniformly over the surface area.19 For the study of Denton et al15 the Protonics system (Inverse Technology, Lincoln, Nebraska, USA) was used. The Protonics system includes a brace set to resist knee flexion to increase hamstring activity and inhibit the activity of the tensor fasciae latae. The study of Lun et al18 used a Y-shaped patellar brace to help control patellar movement (Special FX knee brace; Generation II Orthotics, Richmond, British Columbia, Canada).
In none of the studies was a significant difference between the knee brace group and the control group found on pain (ES varied from −0.14 to 0.04). On the outcome function, one low risk of bias study18 revealed a significant difference between the knee brace group and the control group (ES −0.33). In contrast to these results, one high risk of bias study15 revealed no significant difference between the knee brace group and the control group on function (ES not available). Therefore, there is moderate evidence that there is no difference in effectiveness between knee braces plus exercise therapy versus exercise therapy only on pain at short term. There is conflicting evidence on the additional effect of knee braces on exercise therapy regarding function. In addition, according to the results of one high risk of bias study,15 there is limited evidence that knee braces have no additional effect on exercise therapy on the performance of the knee.
Knee braces and exercise therapy versus placebo braces and exercise therapy
One low risk of bias study18 used a knee sleeve to measure the additional effect of a placebo knee brace in addition to exercise therapy at short term (table 2). The knee sleeve was constructed with the same material as the patella brace. No hole was made in the sleeve over the patella. No significant difference was found between the knee brace and exercise group and the placebo brace and exercise group on pain and function (ES varied from −0.1 to 0.10). Therefore, there is moderate evidence that there is no difference between knee braces and exercise therapy versus placebo knee braces and exercise therapy on pain and function.
Tape and exercise therapy versus exercise therapy only
One study did not provide enough information to calculate ES or 95% CI.17 Information about significant differences between the groups is abstracted from the study.
Patellar taping was used to pull the patella medially. A significant reduction in pain and improvement in function was found in one high risk of bias study20 at short term after a treatment of exercise therapy and patellar taping compared with exercise therapy alone (ES varied from 1.89 to 2.89). In contrast to these results, one low13 and one high risk of bias study17 found no significant reduction in pain and improvement in function (ES varied from −0.19 to 0) at short term. At long term, one low risk of bias study found no significant difference on pain and function outcomes between the tape and exercise group and the exercise-only group.13
Therefore, there is conflicting evidence on the additional effect of tape on the outcomes pain and function at short term, while there is moderate evidence that there is no difference in effectiveness on pain and function outcomes between exercise and tape versus exercise only at long term.
Tape and exercise therapy versus placebo tape and exercise therapy
One high risk of bias study20 compared an exercise and placebo taping group with an exercise and tape group (table 2). For the placebo taping the tape was placed across the surface of the patella without patella alignment correction. A significant difference between the exercise and tape group compared with the exercise and placebo tape group was found on pain and function at short term (ES varied from 1.0 to 3.0). Therefore, there is limited evidence that taping and exercise therapy improves pain and function significantly better than placebo taping and exercise therapy at short term.
Foot orthotics and exercise therapy versus exercise therapy only
In one low14 and one high risk of bias study16 foot orthotics were applied additionally to an exercise programme (table 2). Collins et al14 used prefabricated orthotics (Vasily International, Broadbeach, Queensland, Australia), which were fitted into the shoes. Comfort was the primary goal of the orthotics, by heat moulding and adding wedge or heel raises. Eng and Pierrynowski16 used soft foot orthotics with medial wedges to position the subtalar joint towards a neutral position (Spenco Sports Medicine Products, Toronto, Ontario, Canada). Collins et al14 found no significant difference between the exercise and foot orthotics group and the exercise-only group on pain at short and long term (ES varied from −0.22 to 0.20). In contrast, Eng and Pierrynowski16 found a significant difference in favour of exercise combined with foot orthotics for pain during running, but no significant difference between the groups was found for pain during walking at short term. Therefore, there is conflicting evidence for the additional effect of foot orthotics on exercise therapy for pain at short term, and there is moderate evidence that there is no significant difference between exercise and foot orthotics versus exercise only on pain at long term. Collins et al14 is the only study available measuring function. There is conflicting evidence within that study on both short and long-term follow-up on the additional effect of foot orthotics over exercise therapy on function.
Knee braces and exercise therapy
According to the results of this review, knee braces or placebo knee braces have no additional effect over exercise therapy on pain and function for patients with PFPS. None of the braces evaluated in this study resulted in a significant improvement when compared with exercise only. These results are supported by a recent systematic review for anterior knee pain and osteoarthritis, which concluded that there was disputable evidence from low-quality studies for patellar bracing benefits.26
The studies included in this review all used different knee braces, ie, the Protonics system,15 the Palumbo dynamic patellar brace and the Cho-Pat knee strap19 and a Y-shaped knee brace.18 Denton et al15 applied Protonics knee braces as an additional intervention superiorly to exercise therapy. A study not included in this review27 examined the effect of the Protonics knee brace compared with a proprioceptive neuromuscular facilitation programme or no treatment. The Protonics system was more effective for the patients with PFPS. However, it is unclear what mechanism is responsible for the effect of the Protonics system. McCrory et al28 concluded that a single application of the Protonics brace did not alter anterior pelvic tilt, hip internal rotation an adduction, or tibial external rotation during a lateral step up and gait. Earl et al29 stated that the Protonics brace may unload the quadriceps and therefore decrease the load of the patellofemoral joint. It appears that the Protonics system decreases pain when compared with proprioceptive neuromuscular facilitation or no treatment, but has no additional effect to exercise therapy on pain and function. Furthermore, the effect of the Protonics system may be attributed to the specific set of exercises performed daily to strengthen the hamstrings that are accompanied by the brace.
Further research in this field should focus on the heterogeneity of exercise and knee brace protocols, to make the studies comparable and to create a body of evidence on the possible additional effect of knee braces on exercise therapy for patients with PFPS.
Tape and exercise therapy
According to the results of this review, there is conflicting evidence for the additional effect of tape superiorly to exercise therapy on pain and function. Furthermore, there is limited evidence that taping and exercise therapy improve pain and function significantly better than placebo taping and exercise therapy in the short term. These results are supported by a review of Overington et al,30 who stated that some studies found pain reduction by the addition of taping, whereas other studies found no additional effect on a general exercise programme by taping. Crossley et al4 concluded in 2001 in their review that RCT had failed to find a beneficial effect of patellar taping in addition to physiotherapy. This review was published before the results of the study of Whittingham et al20 were published. Although the results of two recent reviews26 31 revealed that medially directed tape decreased pain significantly more than no tape, it is still unclear if tape has an additional value in the treatment of PFPS patients. The rationale behind the use of medial-directed tape, first used by McConnell32 is to pull the patella medially to realign the patella within the femoral trochlea. In the literature there is no agreement about the effect of tape on patella position. One study found no significant difference in patellar position in taped and non-taped conditions,33 whereas other studies did find a significant difference in patellar position.34 35 Although patellar taping seems to reduce pain, the mechanism behind the pain reduction is still unclear.
Foot orthotics and exercise therapy
There is conflicting evidence for the additional effect of foot orthotics on pain and function compared with exercise therapy alone. A recent review36 concluded that combining foot orthotics with physiotherapy showed significantly greater improvements than foot orthotics alone. The results of a prospective study37 demonstrate that a disturbance of the normal dynamic foot alignment is a risk factor for the development of PFPS. Another prospective study has identified a pronated foot type (measured as navicular drop) as being a risk factor for the development of PFPS.2 A possible rationale behind the use of foot orthotics is to reduce excessive pronation. In the included studies no diagnostic criteria were formulated for the use of foot orthotics. It is hypothesised that patients who have excessive foot pronation will benefit more from foot orthotics than patients with normal foot posture.36
Collins et al14 found significant differences between the foot orthotics and exercise and the exercise-only groups for function, measured with the FIQ, at the short-term follow-up and on the KPS at long-term follow-up. In contrast to the results that are established in this review, Collins et al14 concluded that there was no difference in effectiveness between foot orthotics and exercise therapy nor was there any benefit of adding foot orthotics to exercise therapy. This contradiction is caused by Collins et al14 using a 99% CI and a significance cut-off value of p≤0.01, whereas we applied a 95% CI and a significance cut-off value of p≤0.05.
This review has some limitations. There is a small number of studies available describing the additional effect of orthotics or placebo orthotics to exercise therapy on the outcomes pain and function. For that reason one CCT19 is included in this review, although this study has a risk of selection bias and confounding.
Furthermore, only three of the eight included studies in this review are classified as ‘low risk of bias’ according to the guidelines of van Tulder et al.7 Because of the lack of low risk of bias studies on this topic it was impossible to draw the conclusion of ‘strong evidence’. Besides, three of the eight studies included did not provide the raw data and therefore the data from figures had to abstracted. This may have caused inaccuracies. Because these three studies did not supply data, it was impossible to calculate the ES.
There are inconsistencies in the pain and function measures used. The VAS is used to evaluate the pain, but there is no agreement in which situation the pain should be evaluated. The most frequently used questionnaires in the included RCT to measure self-reported function are the KPS22 and the modified FIQ. In the literature, the KPS and the VAS for worst pain are described as being the most reliable measures for detecting a treatment effect.21 The modified FIQ also seems to be a valid measure.38
What is already known on this topic
▶ PFPS is a common diagnosis in young adults and adolescents.
▶ Exercise therapy reduces pain and improves function and additional orthotics might be beneficial.
What this study adds
▶ There is moderate evidence for no additive effectiveness of knee braces to exercise therapy on pain and function, and there is also moderate evidence for no difference in effectiveness between knee braces and exercise therapy versus placebo knee braces and exercise therapy.
▶ Conflicting evidence for the additive effectiveness of tape and foot orthotics to exercise therapy on pain and function was found.
Although PFPS results in a variety of limitations in physical activities in young active adults, there are still many uncertainties about the most optimal treatment of the syndrome. To identify subgroups of patients that are most likely to benefit from orthotic devices could be a way forward in clinical research and very useful in clinical practice. For example, knee bracing and taping may be used by patients with malalignment of the patellofemoral joint, and foot orthotics may be applied by patients with an excessive pronation of the foot. However, when subgroups of PFPS patients are analysed, sample sizes have to increase, which affects the feasibility of the studies. To detect subgroups responsive to specific treatment, dedicated trials designed to assess subgroup effects are needed. More low risk of bias studies on this topic are needed, using adequate randomisation and providing enough information to calculate ES and 95% CI. Blinding of the patients and the care providers is difficult when exercise is concerned; however, other potential sources of bias must be considered in future studies. In future research, agreement should be reached about the outcome measures used in order to facilitate comparisons between the studies and pooling of the results. In addition to pain and function outcome measures, a global effect scale rated by the patient should be used to explain patients' perceptions of their improvement following an intervention.21
There is no additional effect of knee braces over exercise therapy regarding pain and function outcomes for patients with PFPS. The evidence for the additional effect of tape and foot orthotics on exercise therapy is conflicting when compared with exercise only. The combination of tape and exercise seems to be preferable when compared with placebo tape and exercise. This conclusion is based on a small number of high risk of bias studies. More studies with high methodological quality are needed to draw definitive conclusions. Future research should address subgroups to reflect the current strategies as they are used in physiotherapeutic practice. As there are uncertainties in the current literature according to the treatment strategies applied to patients with PFPS, practitioners should rely on their clinical reasoning skills (including vigilant follow-up and re-assessment) and patient presentation to arrive at a management plan.
Criteria for a judgement of ‘yes’ for the sources of risk of bias
1. Was the method of randomisation adequate?
A random (unpredictable) assignment sequence. Examples of adequate methods are coin toss (for studies with two groups), rolling a dice (for studies with two or more groups), drawing of balls of different colours, drawing of ballots with the study group labels from a dark bag, computer-generated random sequence, preordered sealed envelopes, sequentially ordered phials, telephone call to a central office and preordered list of treatment assignments.
Examples of inadequate methods are: alternation, birth date, social insurance/security number, date in which they are invited to participate in the study and hospital registration number.
2. Was the treatment allocation concealed?
Assignment generated by an independent person not responsible for determining the eligibility of the patients. This person has no information about the persons included in the trial and has no influence on the assignment sequence or on the decision about eligibility of the patient.
Was knowledge of the allocated interventions adequately prevented during the study?
3. Was the patient blinded to the intervention?
This item should be scored ‘yes’ if the index and control groups are indistinguishable for the patients or if the success of blinding was tested among the patients and it was successful.
4. Was the care provider blinded to the intervention?
This item should be scored ‘yes’ if the index and control groups are indistinguishable for the care providers or if the success of blinding was tested among the care providers and it was successful.
5. Was the outcome assessor blinded to the intervention?
Adequacy of blinding should be assessed for the primary outcomes. This item should be scored ‘yes’ if the success of blinding was tested among the outcome assessors and it was successful or:
for patient-reported outcomes in which the patient is the outcome assessor (eg, pain, disability): the blinding procedure is adequate for outcome assessors if participant blinding is scored ‘yes’.
for outcome criteria assessed during scheduled visit and that supposes a contact between participants and outcome assessors (eg, clinical examination): the blinding procedure is adequate if patients are blinded, and the treatment or adverse effects of the treatment cannot be noticed during clinical examination.
for outcome criteria that do not suppose a contact with participants (eg, radiography, MRI): the blinding procedure is adequate if the treatment or adverse effects of the treatment cannot be noticed when assessing the main outcome.
for outcome criteria that are clinical or therapeutic events that will be determined by the interaction between patients and care providers (eg, co-interventions, hospitalisation length, treatment failure), in which the care provider is the outcome assessor: the blinding procedure is adequate for outcome assessors if item ‘E’ is scored ‘yes’.
for outcome criteria that are assessed from data of the medical forms: the blinding procedure is adequate if the treatment or adverse effects of the treatment cannot be noticed on the extracted data.
Were incomplete outcome data adequately addressed?
6. Was the drop-out rate described and acceptable?
The number of participants who were included in the study but did not complete the observation period or were not included in the analysis must be described and reasons given. If the percentage of withdrawals and drop-outs does not exceed 20% for during follow-up and does not lead to substantial bias a ‘yes’ is scored. (NB these percentages are arbitrary, not supported by literature.)
7. Were all randomly assigned participants analysed in the group to which they were allocated?
All randomly assigned patients are reported/analysed in the group they were allocated to by randomisation for the most important moments of effect measurement (minus missing values) irrespective of non-compliance and co-interventions.
Other sources of potential bias:
8. Were the groups similar at baseline regarding the most important prognostic indicators?
In order to receive a ‘yes’, groups have to be similar at baseline regarding demographic factors, severity of complaints and value of main outcome measure(s).
9. Were co-interventions avoided or similar?
This item should be scored ‘yes’ if there were no co-interventions or they were similar between the index and control groups.
10. Was the compliance acceptable in all groups?
The reviewer determines if the compliance with the interventions is acceptable, based on the reported intensity, duration, number and frequency of sessions for both the index intervention and control intervention(s). For example, physiotherapy treatment is usually administered over several sessions; therefore, it is necessary to assess how many sessions each patient attended. For single-session interventions (for ex: surgery), this item is irrelevant.
11. Was the timing of the outcome assessment similar in all groups?
Timing of outcome assessment should be identical for all intervention groups and for all important outcome assessments.
Note: These instructions are adapted from van Tulder et al,12 and Boutron et al 2005 (CLEAR NPT) and the Cochrane Handbook of Systematic Reviews of Interventions.
Competing interests None.
Provenance and peer review Not commissioned; externally peer reviewed.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.