Hi Katie-Lou

I was intrigued by this study. However have you really appraised the article for quality and applicability? It seems to me you haven’t. Your comments seem a bit confusing. May I ask do you know how to appraise a clinical trial?

sarahleeandrews - did your group appraise the article? Your only comment seems to be about whether they understood what “Maitland protocol” would be for unilateral neck pain.

I had a bit of a look at the article and to me it has some good points, but a number of limitations. I think the study should be taken seriously.

The article It is a well respected peer-reviewed journal.

Quality:

• Reasonable description of inclusion and exclusion criteria

• Random allocation: Yes = sound method of randomisation

• Concealed allocation: Yes by envelopes

• Baseline comparability: Yes; Well described for age, gender, symptomatic side ( L vs R) Direction of worst movement, Duration of neck pain, Neck Disability Index, Pain intensity at rest on the most painful movement, and cervical ROM. The groups were very much comparable

• Outcome measures chosen have demonstrated reliability and validity with limited data on responsiveness.

• Blinding: the study claims to be a triple blind study which is of course rubbish. It is impossible to blind subjects and as the same therapist carried out both interventions it is impossible to blind the therapists in way or form. However assessors were well blinded which is often the best you can do in a physiotherapy intervention.

• Adequate follow-up: Yes, although the effect was measured immediately (5 minutes) after the session. It would have been interesting to know if there was any longer term benefit – however this was not the aim of the study

• Intention-to-treat analysis: No not stated although flow chart implies people were analysed according to the group they were originally allocated to;

• Between-group comparisons: Yes; However the statistical analysis is limited in that doesn’t indicate the clinical significance of the change in pain level

• Point estimates and variability: Yes. However difficult to interpret, Better if they had give the mean differences and confidence intervals for each outcome measure

• This means it scores about 7/10 on the PEDro scale which is quite a good score. Certainly it indicates the study is worth taking note of.



Applicability

• sarahleeandrews, I think you have to be realistic what in a single RCT can achieve. You have to simplify things in order to standardise the treatment; The question is whether the simplification (in such a complex system of identification and selection of the most appropriate treatment technique as Maitland) is clinically meaningful. The authors decided to tackle unilateral vertebral joint unilateral PAs as a therapeutic technique so this isn’t the whole of Maitland’s array of techniques for a unitlateral problem. For example it excluded the use of rotation mobilisations. I think this is justified because the whole treatment group have to get a similar treatment. They adjusted the grades based on whether predominant impairment was pain or hypomobility – this was in accordance with Maitland’s principles

• Katie-Lou, I agree that the control group treatment was a bit odd. However the therapist couldn’t choose the treatment in anyway as they had to deliver what was randomly selected between the three available techniques. There seemed no sensible rationalisation for this “randomisation” .Some would be allocated a unilateral PA on the same side so some would presumably get a treatment that could have been similar to the active treatment. A more sensible choice might have been to offer a standard sham treatment to all in the control group such as unilateral PAs on the contralateral side or a general massage.

• The outcome measures were appropriate for the clinical problem and for assessing the treatment. Pain and ROM were highly relevant to the technique, particularly as the time frame is immediately after one treatment.

• It was hard to interpret the data analysis (2-way ANOVA). It would have been more useful if the mean difference (along with the 95% Confidence Intervals) between the two groups post treatment was reported for both pain and most painful movement.

• In fact the treated group gained more range of motion than the comparison group. However it was clinically relatively small improvement and as no confidence intervals are given it is hard to say how much we can rely on these results

• The Global Perceived Effect indicates there was no difference in benefit between the groups

• It is a not a large study so this far from “disproves” or proves the efficacy of unilateral PAs as a technique

• It only examined one treatment as far as I could see. And the authors make this clear. So usually one would treat a patient a few times with such a technique or progress the technique/change the technique. So this is a limitation

• It seemed this group was quite chronic (although variable). 800-900 average days since the onset means the patients had these symptoms for years 2-3 years. Given what we know about the value of manual techniques to date such techniques may be more warranted in the acute/subacute presentation. One reason there wasn’t much of a response is that a chronic pain patient may not be expected to respond to such a modality.

In summary

This study was of reasonable quality and does cast some doubt over the immediate effectiveness of the unilateral PA for treating unilateral neck pain. The protocol for the comparison group treatment is a design flaw. The control group didn't get a consistent treatment and I would wonder if if was a true placebo in some cases. It may not have really separated definitively what the two treatment groups got and this could have made the treatment look more useless than it may have been.

The study certainly doesn’t “disprove” Maitland’s approach to unilateral neck pain. One would want to see the study replicated and it may be more useful now to see what a course of treatment produces – as this maybe more clinically meaningful.

And finally I wonder if in such a chronic group of patients starting with a passive modality like mobilisations is really good practice anyway.