Stroke continues to affect many people across
the world. Estimates put the number of survivors at around 50 million (van de
port et al., 2012). Those who survive a stroke often find that they have to
depend on others for activities that they previously carried out by themselves.
For instance, it is estimated that only
20% of stroke survivors retain their ability to walk. This is a great cause of
concern given the important role that walking plays in the daily life of any
normal human being. Those recovering from stroke need to undergo rehabilitation
in addition to the usual medical rehabilitation. The usual physiotherapy for
survivors has been individualized task specific training. Recent studies,
however, indicate that these training can be more effective when done in
groups. The new approach is called circuit as the training takes place in
different workstations. What follows is a critical appraisal of a randomized a
study by van de port et al on the effects of circuit training as an alternative
to usual physiotherapy.
Beginning with some disturbing statistics on
the prevalence of stroke across the world, the authors proceed on a brief
review of the literature concerning rehabilitation for survivors (van de port
et al., 2012). Literature supporting the suitability of task specific rehabilitation
is cited to build this point. It is from this review of literature that the
authors justifying their own study. They note that previous studies on this
issue have been on people with chronic stroke. It was, therefore, reasonable
for the article under review to focus on another area instead of duplicating
previous work. They did this by concentrating on the effectiveness of circuit
training on people within the first six months of suffering a stroke. It also
comes out from the hypothesis that the authors were keen to understand relative
effectiveness between circuit training and normal physiotherapy. In order to
achieve these goals, only patients with stroke were taken as participants of
the study.
Randomized controlled trials (RCT) remains
one of the strongest research designs for measuring the effectiveness of health
interventions. Such a design is advantageous in various ways. For one,
randomization adds validity to statistical tests. The differences observed
between the treatment group and the control group in a randomized trial can be
taken to be fairly represented what would ordinarily be observed in the general
population. It is perhaps for this and many other reasons that the authors in
the present study chose RCT. All patients participating in the trial were
recruited following criteria that only identified them as suffering a stroke
(van de port et al., 2012). Once eligible, allocation to treatment was random.
The study authors have also been successful
in accounting for all those patients who entered the study (van de port et al.,
2012). Following a predesigned eligibility criteria, a total of 250 patients
signed up for the study of which 126 underwent circuit training while 124
underwent usual physiotherapy. Accounting for study participants is an
important way of ensuring against making erroneous conclusions from the study.
The paper under review seems to have surpassed the usual standards of
accountability. For instance, the study accounts even for those who remained in
the study but nevertheless failed to complete certain aspects. A case in point
is the report of a patient who missed the 12 week assessment visit.
It sometimes happen that randomization fails
to eliminate all the potential biases in a study. Those responsible for
designing a study can strengthen their design through blinding. A single blind
is where those receiving interventions do not know that they are doing so. On
the other hand, a double blind assumes that another additional group is also
blind to the intervention being given. That additional group may range from
those carrying out the assessment to the investigators themselves. More
important, however, is the fact that any study design should do more than
merely indicate that it is either a single or double blind. The ambiguity
inherent in any of those phrases necessitates a detailed explanation from a
design as to what exactly is meant. The study simply indicates that it was
single blind without explaining which particular group was blinded (van de port
et al., 2012). This is a serious weakness in a study design given that the
phrase often contains ambiguities. The
only information that comes out clearly is that the research assistants
responsible for assessing outcomes were blinded to treatment allocation.
Even more important for a valid study is the
requirement that both the intervention group and the control group be similar. Both
randomization and blinding are just some of the ways in which study designs try
to eliminate the possibility that these two groups are dissimilar at the start
of a study. Similarity ensures that no group has features that the other group
does not possess. For instance, it could be the case that a particular
intervention is only effective when administered to people of a particular
gender. The study under analysis makes no reference to any differences among
the study participants (van de port et al., 2012). To the extent that the study
makes not mention of the demographic features of participants, then this could
be a challenge to its validity. For instance, those at relatively higher social
classes may achieve better outcomes even when receiving usual physiotherapy
than some of their counterparts in the treatment group.
Another related issue to the study validity
relates to equality of treatment. It implies that, besides the relevant
intervention measures, no group in a randomized controlled trial should receive
an advantage in treatment that the other groups do not. It is simply a
requirement that the conditions under which the study is carried be similar. Thus,
the validity a study investigating the effect of noise on students would be
questionable if one group under investigation is placed in an environment where
there are other confounding factors. There seems to be no evidence of any
differential treatment of the two groups in the present study (van de port et
al., 2012). The authors used the same approach to measuring final outcomes of
the different treatments. For instance, the measurement of the primary outcome
remained the mobility domain of stroke. Similarly, secondary outcome measures
were the other domains of the stroke impact scale for both groups.
In planning any study on intervention
effects, it is also crucial for the study design to consider the number of
participants. This number would determine whether there can be a reliable
answer to the study questions. Including an excessively high number of
participants leads to a waste of scarce resources. It may also be harmful to
those participants in some risky studies. This is fortunately not the case in
the present study as the interventions given are both beneficial to the
patients. In the same vein, a sample
size that is not large enough is incapable of produce precise results. The challenge, however, is that there is no
clear rule for determining an appropriate sample size (van de port et al.,
2012). The study under review had a sample size of 250 divided into 126 and 124
for the intervention and control groups respectively. This is no doubt a
relatively small sample considering the implication of such a study. The
authors have, nevertheless, catered for this shortcoming by introducing a power
calculation.
The authors presented the results of the
study as standard deviations (SD) of baseline characteristics. There were no
significant differences between the two groups on the stroke mobility impact scale
(−0.05 (SE 0.68), P=0.94).There were, however, those areas in which the circuit
group showed significant effects (van de port et al., 2012). These were in the
five meter working speed test, the six minute walk test and the modified stairs
test. All these effects were during the intervention. On the other hand, the
usual care group registered significant effects in areas such as memory and
thinking domain of the stroke impact scale as well as the leisure domain of the
Nottingham extended activities of daily living scale. The other secondary
outcomes did not register any significant differences between the groups. It is
evident from these results that no important outcome was left out.
Considering that the authors were able to
tell when the treatment effect in one group was significant relative to the
other, this is an indication that the estimates were fairly precise as to
enable one to make a decision (van de port et al., 2012). This was particularly so when the usual care
group registered significant effects in memory and thinking domain of the
Nottingham extended activities of daily life. Besides, there are several instances where
confidence intervals are reported. This is in addition to instances where P values intervals are also reported.
For example, there is P>0.001 when reporting the significant effects found
in circuit training for five meter walking speed test.
In addition, the results of any study would
not mean much if they can not be put to practical use. This would entail conducting a cost benefit analysis
(van de port et al., 2012). One
difference between circuit training and normal physiotherapy is that the former
relies on fewer therapists. This is itself a strong case for adopting circuit
training in those areas where it produces significant effects. Besides,
patients get to do their trainings while at home, thereby, freeing the limited
rehabilitation centers for other important uses. The only objection would seem
to be the harms incurred but even here there are no significant differences
between the two methods. It follows from these that the benefits are worth the
costs and harms.
Although full of important insights, I would
be very reluctant to apply the outcome of this study to my patients (van de
port et al., 2012). For one, the study has not mentioned the demographic
features of its patients. There is a possibility that those participations are fundamentally
different from my patients. The authors themselves admit of this problem in
their limitations when they caution against the generalisability of the study
findings. The caution stems from their admission that most of the participants
were patients suffering from mild stroke.
With healthcare costs continuing to soar,
efforts must be made to find cheaper alternatives to health interventions.
Circuit training offers a promise to lowering the cost of caring for stroke
survivors. Research in this area is, however, still in the infant stage. The
paper under review was just another attempt to increase knowledge in that area.
More research will have to be done to make this promise a reality for many
stroke survivors the world over.
Reference
Van de port, I.G.L., Wevers, L.E.G., Lindeman,
E.,& Kwakkel, G. (2012).Effects of circuit training
as alternative to usual physiotherapy after stroke: randomised controlled trial.BMJ,1-10.
0 comments:
Post a Comment