Thearticle being reviewed is titled, “Comparison of effects of eccentric training,eccentric-concentric training, and eccentric-concentric training combined withisometric contraction in the treatment of lateral elbow tendinopathy” byDimitrios Stasinopoulos PhD and Ioannis Stasinopoulos MD, taken from theJournal of Hand Therapy. This article looks into the comparative trainingmethods available for treatment post-diagnosis. There are three forms oftreatment explored; an eccentric, eccentric-concentric, andeccentric-concentric training method with an isometric contraction.
Inthe abstract for the article, Stasinopoulos and Stasinopoulos clearly statethat all three approaches to rehabilitation have been shown to have value, but nowwant to produce a trial to compare their effectiveness. The abstract outlines conciselythe main points of the method, results and a brief conclusion. Bisset et al(2011) indicate the validity of the study, with around 3% of the populationsuffering from lateral elbow pain, with at least 20% of these chronically. AnRCT (randomised control trial) was used to explore this research, with 34patients participating, split into 3 groups; A, B and C. Bulpitt (2013) statesthe RCT study design is “necessary to prove the effectiveness of new healthcarestrategies or treatment and to prevent the introduction of new but uselesstreatments”.
Bulpitt (2013) also suggests that the RCT has been presentthroughout the history of medicine in differing forms, and has throughout thistime shown its necessity. The determination of results, through using the VAS(visual analogue scale) for pain and function, in addition to using a pain-freegrip strength test are also clearly indicated in the abstract. Regarding theVAS scale, Van den Broeck and Brestoff (2013) suggest “The VAS scale is veryeasy to apply, is usually considered cognitively easy to respond to, andprovides results that can be interpreted straightforwardly”. However, the sameeditors also reason that, “…VAS scales are considered to be overly simplisticby researchers”, and “…VAS scales tend to result in disease states beingweighed as more severe than with other instruments”. This may cast doubt on thisform of assessment, with the VAS scale potentially being overly subjective. Theresults of the trial are stated plainly, with the eccentric-concentric pairedwith isometric contraction treatment yielding the best results. A potentialoversight in this section could be the inclusion of a hypothesis and nullhypothesis, a very important factor in qualitative data and the production of agood study result (Gad, 2009).
Inthe introduction, there are supported references relating to the causes oflateral elbow tendinopathy (LET), both anatomically and physiologically. Theresearchers also give a good insight into clinical presentations and potential triggersof LET, for example, work or sport-related repetitive strain movements. Theintroduction shows further necessity of the trial, with the researchersstating, “such a variety of treatment options suggests that the optimaltreatment strategy is not known, and more research is needed”. Stasinopoulosand Stasinopoulos also indicate the preferred treatment of the LET conditioncurrently lies with an eccentric training programme, which is echoed in conditionssimilar to LET such as patellar and Achilles tendinopathies. Hans Hoppeler(2014) agrees, suggesting that “eccentric exercise is recognised to be theconservative treatment of choice for tendinopathies”. Stasinopoulos andStasinopoulos also summarise with references, two alternative treatments and previousstudies surrounding them, showing their relevance to the study. They alsooutlined this physiological movement, “forceful grip activities requiringisometric contraction” which would validate the grip test as a meaningful testfor outcomes.
Overall, the introduction states the relevance of this trial,presenting well-informed scientific knowledge, supporting references andclinical relevance; early indicators that the method may also be wellconsidered and precise. Stasinopoulosand Stasinopoulos explain in the method, their reasoning for using theparallel-group design for their RCT over the crossover design. With thecrossover design, there is the possibility of all treatment types having aneffect on participants, potentially confusing results. Therefore, there isjustification that each participant randomly receives one of three treatmentssimultaneously. This will give a more controlled set of results, and reducesthe possibility of attrition due to being ‘cured’ before completion. There weretwo external professionals assigned for the trial; a doctor who was responsiblefor diagnosis, baseline and follow-up assessments, as well as collecting informedconsent, and a therapist carrying out all of the selected treatments for eachgroup. This minimalizes external factors affecting results. For example, the therapistwill have a practised routine of treatment and will have a very low percentageof variance in technique.
However, under scrutiny, the therapist may have aparticular bias toward a form of treatment and may be more practised in thisarea, giving a better treatment. A proposed alternative could be to use threespecialist physiotherapists for the three differing treatments. Stasinopoulosand Stasinopoulos took many precautions through testing and eliminatingalternate pathologies from the trial, which could affect treatment outcomes. Acondition such as LET can be very common and not always related to articular surfaces(Koot et al, 2016). Due to the population pool of the study being simply overthe age of 18, with no maximum age, influences such as sensitivity of growthplates during puberty, and increases in onset of poorer health conditions inlater life are minimalised due to the researcher’s rigorous testing.Additionally, the researchers show good, referenced knowledge of pathologiesthat could have an effect on the study, which aided to filter them from thetrial.
Dawe and Poulter (2011) state that elbow pain, although very common, maywell reveal itself as a “multisystem disorder”, meaning it can be verydifficult to diagnose a definitive pathology without extensive testing, whichis evidently achieved in the study. Onceall diagnoses were confirmed and informed consent was gained to make the studylegitimate, the control group of 43 was then randomly split into 3 separate groupsso no selection bias was shown (Kennes et al, 2011). All individuals were giveninstructions to continue using the affected arm, with the following actions tobe avoided; grasping, lifting, knitting, handwriting, driving a car and using ascrewdriver. All heavily activate the extensor carpi radialis which willprovoke their symptoms and affect the study. However, there were no absolutechecks undertaken to ensure compliance with this. Forthe avoidance of potential bias, the participants weren’t given any prognosisregarding each of the treatments before entering into the study, andcommunication between therapist and patient was minimal.
Hróbjartsson et al(2011) said of response bias, “Patients may desire to please the researcher, orjust give a “correct” or expected answer that fits with the experimentalsituation”. The doctor responsible for initial testing and diagnosis was alsokept blind in relation to which patient received which treatment type, meaningthere was also no professional bias in relation to assessment when determiningresults. All three forms of treatment began with the sameinstructed starting position, as well as the same amount of sets, reps andfrequency per week, showing continuity within this section. However, it couldbe argued that the technique of the individual treatments will differ inworkload, most notably from the purely eccentric movement to theeccentric-concentric with the addition of isometric being the largest shift ineffort load, including a potential progressive inclusion of hand-held weights.This has the potential to affect results/attrition, as the higher the workload,the more aggravation and fatigue to the injured site, leading to a higherinjury risk factor (McCall, 2016). Drew, Cook and Finch (2016), however,suggest by their Tertiary Prevention concept, that although training loads andinjury incidence are linked, a phased return programme is recommended forfuture injury prevention rather than ruling out all but the lightest exercisevariations. Gatheringthe results of the trial, as mentioned in the abstract, was taken in the formof patient experienced function and pain over the course of 24 hours previousto the final assessment, measured on a VAS scale (0 – 10).
Kersten et al (2012)published an article arguing the point regarding the VAS scale that “VAS changescores may seriously over or underestimate changes resulting fromrehabilitation”, as well as “a given change in one patient may be of differentmagnitude than the same apparent change in another”. This may lead to theresults process, and subsequent findings not given the credit intended. Thepain-free grip strength test in the results can be determined as more credible,with a physical piece of equipment giving a reading. However, this is onlygiven after the patient experiences their subjective pain level. It can bedetermined that there is a high factor of subjectivity in regard to therecorded results, only from the patients, not the assessor.
A more dependablefeature of the pain-free grip strength test was that the patients weren’tallowed to see their results, so as to prevent a potentially forced result. The results were produced by comparing the initialtests carried out before treatments began, to those taken once a course oftreatments had been completed (4 weeks) and again at 8 weeks. Due to the samescales being used in both instances, this a logical and consistent way ofproducing results. The mean average was calculated from each group’s givenscores, giving each group one mean total for each test. Stasinopoulos andStasinopoulos indicate well the relevant differing statistics of the group,from the average age of the participants (43), their approximate timescale ofthe condition upon treatment (6 months) and the percentage of participantssuffering with the tendinopathy in their dominant arm (85%). The most relevantstatement being that between the three groups, the mean average age andtimescale of condition was very similar. These factors aid in the results beingmore dependable. Attrition was monitored by use of a flow chart,with the patients that dropped out doing so before the treatment phase began.
This means that attrition bias is not applicable here (Hróbjartsson et al, 2011). Stasinopoulosand Stasinopoulos indicate their results clearly in a table showing adefinitive decrease in pain on the VAS scale, a definitive increase in function(VAS) and pain-free grip strength (lbs) for all treatment types. There are afew clear points highlighted by the table; the eccentric-concentric alongsideisometric contraction treatment has yielded the largest threshold ofimprovement, and that the largest improvement was recorded after week 4, with avery slight improvement made in comparison when results were taken at week 8.There was less than 0.
5 units of statistical significance in results betweeneccentric and eccentric-concentric treatment methods, with Stasinopoulos andStasinopoulos basing their p-level at 5%. Generally, the lower the p-value, thestronger the evidence (du Prel et al, 2009). However, due to the apparent lackof a hypothesis, the p-value cannot be based on an appropriate source. Duringthe discussion, the researchers openly admit that “future well-designed studiesare needed to confirm the results of the present trial”.
This would suggestthat, although there is some well-thought evidence shown in this trial,Stasinopoulos and Stasinopoulos may not be truly confident that their resultsbe concrete findings until further trials concur with their result. This couldbe a causality of their decision to only have a target group of 34 people, assmaller trial results cannot be truly reliable until further testing of largergroups can confirm the same conclusion (Hackshaw, 2009). Another issue with thetarget group could be that they were all amateur athletes taking part, and itis suggested that when treating athletes, “all aspects of the kinetic chain” betaken into consideration (Ellenbecker et al, 2010), which may be a furtherprogression of this particular study. Alongside this, although it can be up forethical debate, it is recommended that a placebo/sham component be introducedto test the legitimacy of the trial and its results (Brim and Miller 2013). Inconclusion, although the trial itself has been well thought out in its validity,it has been now suggested there are many components lacking before this trialcan be fully accepted for the results it has produced.