Editor/Reviewer comment
Consider placing it within the early phases of the 2008 MRC Framework for the
design and evaluation of complex interventions
Start any discussion with a warning about the limitations of interpreting such
data. I would suggest reducing the interpretations of the detailed results and
instead focusing on how the findings might inform the future development and
evaluation of the intervention.
Shift your focus to the qualitative data, though I think you should downgrade
rather than completely lose the quantitative results
Avoid abbreviations
Statement about the ethics approval should appear at the beginning of the
methods
Description of the participants on page 7 should be in the results section
Include the aims in the first column in Tab 1 ?
Recruitment of advisors in phase 1
Recruitment of smokers in phase 2
More detail on questionnaire which assessed belief
Response rates need to be included at the beginning of the phase 2 results · The
reviewers, and I agree, seem to be having problems fully understanding the
results. I think it might be clearer if you tabulated the numerical results,
indicating the sample size, and just including the ‘headline findings’ in the text.
Please also provide confidence intervals where appropriate
What did they stop doing to make space for this? Or did it just extend the length
of the programme
Details on accelerometer
More discussion of the findings in relation to other published data
Response
We have added text in the Introduction which references the MRC Guidelines (MRC, 2008;
Campbell et al, 2000; Anderson, 2008).
We have done this.
We have done this by adding text in the Results Section.
Done
Done
Done
We have revised the aims in line with suggestions for refocusing the manuscript. We appreciate the
suggestion and now feel the paper draws out some of the key aspects of our work.
Done
Done
Perhaps we should append it?
We have included response rates in the text and added Tables 5 and 6 with the required data.
The length of clinic did not change but it would appear (from our conversations with advisors) that
the group discussions, which form a considerable part of the clinics, focused on physical activity
more often and for longer durations than previously thereby reducing discussion about other aspects
of the client experiences. Our observations in Phase 1 clearly supported that as we note on P. 9, but
we did not directly observe clinics in Phase 2.
We did not use an accelerometer but pedometers were used as motivational devices, with goal
setting and self-monitoring. We have added details about the pedometer (i.e., Yamax, Digi-Walker).
We have added reference to other studies through the Discussion
Amanda’s comments
add “UK” to end of the sentence in background section of abstract
In the results section of the abstract it should state the point at which there was a significant
advancement in clients’ stage of readiness – as it stands it is rather vague because we are not
told in the methods when assessments took place
Page 3 paragraph 3: Methodological issues are not the only reason why studies have shown
mixed results – it is possible that there is no relationship between exercise and smoking
cessation and we have to be prepared to accept this might be the case? There seems to be the
assumption that a relationship does exist, its just that studies have been poor. But an
alternative explanation might be that exercise is not effective. At the moment we do not
know which is true and a balanced view should be presented.
Page 4 paragraph 1: Change “fairly” to “relatively”
Page 4 paragraph 3: Who thought it would be unpopular – participants/clients?
Page 7. Confusion about the number of advisors
Page 8 paragraph 1: Did participants given informed consent or written informed consent?
Page 9 paragraph 1: In the sentence “for clients who did not complete a follow-up survey we
imputed their follow score” change to “baseline score”
Page 10 paragraphs 2 and 3: no data on advisors’ 7-day physical activity recall reported,
only given 7-day physical activity results for clients. Did the advisors become more active?
Also, did you measure 7 day physical activity behaviour AND time spent promoting
physical activity by advisors?
Change the paragraphs on page 10 so that phase 2 data is presented in a consistent order for
both advisors and clients. For example, start both paragraphs with self-efficacy data and so
on…
Page 11 Paragraph 2: Are the programmes in Birmingham and Plymouth typical of other
programmes in the UK? What reassurances do we have that your results are in anyway
generalisable? This is key if subsequent decisions about the viability of future research and
ultimately services, are to be made on the basis of the results presented in this paper?
Page 11 paragraph 3: Clarify meaning of “The smoking cessation advisors were enthusiastic
about implementing the refined version of the intervention”. Also, on what basis are the you
making the comment “which quickly became almost effortless to deliver”? This line seems
rather subjective but it is a critical statement when talking about implementing services
because it could be somewhat misleading without evidence to support it
Page 12 paragraph 3: The intervention did not significantly change clients self-reported
physical activity levels? Therefore, does this study add to the growing body of mixed
It is already there
Added on p.7.
To address this point we include several new references that support the contention that
exercise can help to facilitate smoking cessation as follows:
Everson-Hock, E.S., Taylor, A.H., Ussher, M., & Faulkner, G. (in press). A qualitative
perspective on multiple health behaviour change: views of smoking cessation
advisors who promote physical activity. Journal of Smoking Cessation
Taylor, A.H., Ussher, M., & Faulkner, G. (2007). The acute effects of exercise on
cigarette cravings, withdrawal symptoms, affect and smoking behaviour: A
systematic review. Addiction, 102, 534-543.
There appears to be little doubt that for some people, becoming more physically active is
an effective aid for smoking cessation. The challenge is to design interventions that
facilitate more quitters to increase physical activity, with the likelihood of a greater
impact on smoking cessation rates.
Done
Advisors – we have clarified this.
We now make reference to just the 7 advisors involved in data collection and training in
Phase 2, and that the S. Birmingham NHS Stop Smoking Service was chosen through
GLOBALink – a Smoking Cessation practitioner research network.
Written informed consent – added at the start of the Methods section.
Done
All this data now presented in Table 5 & 6, and in the text.
Now redundant since we tabulated all the findings then only discussed headline findings
in the text, as the editor suggested.
Technically they should be as they both follow the standard 6/7-week smoking cessation
clinic format (with advisors following gold standard Maudsley training guidance) and
both include diverse geographical areas with differing SES profiles. For example, in
Plymouth clients attending the clinics from post codes in which smoking prevalence
ranged from 20-48% of the population. We address the issue of generalisability in the
Discussion.
We now write: ‘The smoking cessation advisors were enthusiastic about implementing
the refined version of the intervention, quickly increased their confidence to promote
physical activity (based on survey responses and field notes)…’
As another reviewer suggested, this was not a primary outcome and indeed the study was
probably underpowered to detect a difference. We note in the Results the number of
literature that you referred to in the introduction, which has failed to provide favourable
outcomes between exercise and smoking cessation.
Page 14 paragraph 14: Given the data, the conclusions about promoting physical activity in
smoking clinics being “certainly possible” and “could be undertaken in a wide scale ”are far
too strong.
Page 15 paragraph 1: The grounds on which you claim an RCT is justified are currently
weak and unclear. What type of intervention would you use in your RCT then given there
was no change in clients’ physical activity behaviour, self-efficacy, outcome expectancy
beliefs regarding physical activity as a cessation aid and no change in advisors’ stage of
readiness for promoting physical activity as a smoking cessation aid, outcome efficacy or
pro and con beliefs? You did not find a change in any of the key outcomes so it is hard to see
how an RCT is now warranted. Should the next step be to test out different types of exercise
interventions in this population/setting, as you have suggested earlier in the discussion,
rather than a RCT to test effectiveness?
clients who decreased, increased and stayed the same in terms of physical activity.
We have attenuated such strong assertions, but note that some clients were responsive.
The next logical step seems to be to apply the intervention that was developed through
the process of collaboration action research to a sample large enough and with enough
power to detect an effect in the variables of interest. The quantitative findings are
preliminary and a change in some variables is encouraging. We feel that the qualitative
data support further pilot work with the intervention, with more rigorous assessment of
quantitative client measures, prior to a larger randomised trial. Structured exercise
interventions have been adequately assessed and we feel the counselling intervention we
have developed should be rigorously evaluated.
James’s comments
I am particularly concerned with the use of quantitative statistics to evaluate the
‘effectiveness’ of the self-help book in increasing advisers promotion of physical activity in
stop smoking clinics and increasing self-efficacy in relation to PA and quitting in the
smokers themselves. With only 7 advisers the analysis on page 10 seems fairly arbitrary and
at worse potentially misleading. May be more appropriate to analyze the qualitative
comments from the advisers regarding PA use and its promotion in stop smoking services
comments rather than performing t-tests on 7 people on self-report measures.
The statistics performed on data from the smokers themselves is better however the only
significant findings are related to willingness to use PA as a cessation aid, and self-efficacy
for smoking cessation (I am unsure how this is related to PA?). I assume these are based on
Transtheoretical model constructs. I point out to the authors there is a plethora of research
that seriously questions the value of ‘stages’ of change and how little intention is related to
actual behavioural change.
The authors make too much of relatively weak findings (in terms of whether there
intervention increased PA etc) and mention little about the limitations of the study (or the
analysis). This needs to be much more explicit.
I recommend the write up be changed, putting the emphasis onto the qualitative elements of
the study, i.e. developing PA promotion tools for smoking cessation advisers and also
qualitative analysis of the 7 stop smoking advisers comments.
The authors need to be clear that the analysis of the smokers themselves is severely limited
i.e. power, no control group, theoretical issues with measuring intention to change and
limitations of this.
Add on page 3 at least a few sentences on the work (lab based etc) that has found significant
effects of PA on smoking withdrawal.
As the editor suggests, we have downplayed, rather than eliminated, the quantitative
findings, relative to the qualitative findings and developmental process.
Constructs that have been linked to the Transtheoretical Model (TTM) have been drawn
from other theoretical frameworks such as Social Cognitive Theory (eg, self-efficacy and
outcome expectancy). Our published cross-sectional research suggests that these
constructs are stage dependent as the TTM predicts. Stage is also associated with selfreported physical activity behaviour and promotion of physical activity.
We now emphasise that the quantitative findings are preliminary and there may not be
sufficient power to detect an effect. Much more emphasis is placed on the qualitative
data and field notes.
See above.
See above
Added
Wendy’s comments
Abstract – It would be beneficial to state upfront that this article does not explore the impact
of PA promotion within the SSS on engagement with services or smoking outcomes rather it
explores the feasibility of developing and embedding physical activity promotion in SSS.
Conclusion (abstract and main article) currently states that PA was included in the SSS with
‘no apparent distraction’ from the clients attempt to quit. I do not think this statement can be
completely evidenced by the data presented in the paper. The authors should be more
cautious in making this assertion as the number of clients who completed a questionnaire at
T2 highlighted substantial drop-out and there was no exploration of impact on cessation
attempts.
The introduction would benefit from some re-structuring from the paragraph beginning ‘The
challenge then is to develop…’ For example, an additional heading indicating that the paper
is moving from literature review to ‘the purpose of the reported work’ would add clarity.
The next paragraph (5th in the introduction) begins ‘in the survey’ – this ‘familiarity’ is
repeated later in the paragraph. The reader requires further explanation of the context of this
work!
Present actual p values not p<0.05
Results Phase 2 Paragraph 1 – confusing findings in relation to item score for perceived cons
of promoting PA (significant) and later – cons for promoting PA (not significant)? The
difference between the two measures could be clarified.
The substantial drop-put rate at T2 requires further acknowledgment and discussion of how
this may have impacted on findings. (72 clients at T1 to 23 at T2). If possible comparable
data on drop-out rates for this service prior to SSS or other comparable services without the
introduction of PA as a cessation aid should be presented and discussed. Without this
information presented it could in fact be the PA intervention had a part to play in low
retention rates?
Related to point 3 above: The authors need to be explicit about what was meant by the
‘intent to treat’ analysis – a term coined in clinical trials- it is often misused in the literature
The representiveness of the SSS, the client sample and the relatively small number of
advisors needs further acknowledgment and consideration in terms of drawing conclusions.
For example from an inequalities angle – what kind of areas did these clients come from?
What was there socio-economic status etc. More contextual information is required on where
the SSS were based. Reference to South Birmingham clinics much later in the paper with no
geographical info upfront.
More could be made of the qualitative data gathered to back up process findings re
feasibility. At present comments are presented in a table for illustrative purposes but there is
limited thematic analysis and presentation of quotes to directly support process findings.
Discussion page 13: The authors state: ‘Where a structured exercise programme may
require…cognitive overload for quitters.’ The results do not actually indicate this/provide
evidence for this statement- if this pertains to evidence found elsewhere in the literature –
reference needs to be made.
Done
We have removed reference to ‘no apparent distraction’
We have added a new heading: ‘How should physical activity best be promoted?’
We have changed the wording to: ‘In a national survey…’ Both references are now
published.
Done, in the Tables.
There should be no confusion, now that the data is presented in the Tables.
We now consider this in more detail in the Discussion. We estimate that much of the
dropout was due to smokers not returning for a final assessment since they had relapsed.
We provide some data on typical 4 week quit rates in South Birmingham.
We have clarified this.
We have added to the Discussion a comparison between the present sample and data
reported in our national survey.
We have extensively revised the Results in line with this suggestion.
We feel that the additional qualitative data and reference to a recent qualitative paper
now provides greater support for this statement.