BMJ - Decision on Manuscript ID BMJ

Transcription

1 BMJ - Decision on Manuscript ID BMJ

2 Body: 28-Nov-2017 Dear Ms. Exarchakou Manuscript ID BMJ entitled "Population-based cancer survival trends and socioeconomic inequalities in England, : do national cancer policy initiatives have an impact?" Thank you for sending us your paper. We sent it for external peer review and discussed it at our manuscript committee meeting. We recognise its potential importance and relevance to general medical readers, but I am afraid that we have not yet been able to reach a final decision on it because several important aspects of the work still need clarifying. We hope very much that you will be willing and able to revise your paper as explained below in the report from the manuscript meeting, so that we will be in a better position to understand your study and decide whether the BMJ is the right journal for it. We are looking forward to reading the revised version and, we hope, reaching a decision. Kristina Fišter kfister@bmj.com *** PLEASE NOTE: This is a two-step process. After clicking on the link, you will be directed to a webpage to confirm. *** 62ef4a **Report from The BMJ s manuscript committee meeting** These comments are an attempt to summarise the discussions at the manuscript meeting. They are not an exact transcript. Members of the committee were: Wim Weber (chair), Julie Morris (statistician), Georg Roeggla, John Fletcher, Tiago Villanueva, Sophie Cook, Rubin Minhas, Daoxin Yin, Kristina Fišter. Decision: Put points Detailed comments from the meeting: First, please revise your paper to respond to all of the comments by the reviewers. Their reports are available at the end of this letter, below. Please also respond to these additional comments by the committee: * Justification should be provided for the use of one-year net survival rates as the main outcome measure. Is this appropriate? Why are 5-year survival rates not used? We thought a longer term survival figure would be more indicative of treatment success as well as being of greater patient interest. Is it possible to have 5-year survival rates in the paper?

3 * More details on how the net survival rates were calculated would be useful. * The deprivation gap (differences between the first and fifth quintile) is used to measure socioeconomic inequality. But it would be of interest to see how the absolute deprivation levels vary over the time period. That is, how is the gap representing deprivation changes over time? * The presentation of the figures could be improved. At present Figure 2 is difficult to read. * Would you please try and improve the presentation of the Methods section, with a general reader in mind. For example, what is the source of cancer registrations and how complete are these? Where did the information come from that was used to define deprivation and how was this mapped to post codes? Is "net survival" a widely understood term and is the description of adjustment for competing causes of death adequate? * For a non-uk reader who doesn t know enough about the UK cancer initiatives, it isn t clear what happened in these initiatives. * The research question needs a clearer focus. What exactly did you investigate? * You seem to claim your methodology is accepted by several bodies - are there counter opinions? * There was some criticism of the assumptions made: 1. Changes take places immiediately or soon after the set dates 2. Like other analyses of CVD Plans, it is not a safe assumption that trends in cancer survival' will just carry if there are no efforts to continue them. Using a ruler and measuring the trend is a bit simplistic. * There are many data, but it is hard to get a good feel for the actual differences. Can you provide some sort of European comparison figure, to illustrate the English gap? * The title should not contain a question. It should state the research question and study design. * It would be useful if you could rewrite the paper for an international audience and explain why this update is important. * Can you do more to put this in an international context? Can you describe some of the problems or successes with this system and contrast it with others? In your response please provide, point by point, your replies to the comments made by the reviewers and the editors, explaining how you have dealt with them in the paper. Comments from Reviewers Reviewer: 1 Recommendation: Comments:

4 This paper reveals important trends in cancer survival in the U.K. and how those trends have and have not changed over the last two decades--especially for those who are more socio-economically deprived. Its value lies in the feedback it can provide to government agencies who have been mandated to improve survival in general and among deprived populations in particular. From a patient perspective, the paper is important because it highlights systemic rather than individual factors as likely causes for the lack of significant progress in the worthwhile goals of these government programs, putting the burden of responsibility for disappointing outcomes somewhere other than on individual cancer patients' backs (e.g., accusing them of non-compliance with treatments, failure to follow up on suspicious symptoms in a timely fashion, etc.). In that sense, it serves as a mandate for government health planners to go back to the drawing board, review the implementation to date of their program goals, and come up with more effective methods of achieving them--or to dismantle the existing programs and replace them with something that takes a different tack in addressing the disparities in cancer survival in the U.K. Additional Questions: Please enter your name: Brenda Denzler Job Title: Retired Institution: Retired Reimbursement for attending a symposium?: No A fee for speaking?: No A fee for organising education?: No Funds for research?: No Funds for a member of staff?: No Fees for consulting?: No Have you in the past five years been employed by an organisation that may in any way gain or lose financially from the publication of this paper?: No Do you hold any stocks or shares in an organisation that may in any way gain or lose financially from the publication of this paper?: No If you have any competing interests <A HREF=' lists/declaration-competing-interests'target='_new'> (please see BMJ policy) </a>please declare them here: Reviewer: 2 Recommendation: Comments: Population Based Cancer Survival Trends paper Between 2008 and years - I cared for an elderly family member with uterine cancer (adenocarcinoma of the endometrium) throughout their illness

5 trajectory and their death. Their cancer was the only cause of death. (page 3, line 46.) They were treated according to the Two Week Wait guidelines (page 16, line 48) having presented immediately at their GP with symptoms of post-menopausal bleeding. Following biopsy and full radial hysterectomy (surgery carried out within 6 weeks) metastatic Ca ( I use the abbreviation throughout) was found in lymph nodes at grade/stage 3,2. A year post-diagnosis they were alive. Using the online IMD (2010,IMD-by-geo.opendatacommunities.org)their postcode has an index value of 10. They were treated at a specialist tertiary centre for Cancer Care throughout. In reading this paper I think I behaved predictably for a reader who has had close contact with a Ca patient who has died of the disease and has been screened for the same. I started by searching for whether the Ca was included in one of the most primary, common, malignant neoplasms (Page 1, line 26) and was delighted to see that it was. I then searched for whether there was a positive changed in Ca 1-year survival rates and at that point struggled with the way the information was presented in this paper. I read the paper 4 times making line by line notes throughout as a means of understanding the information presented and picking through the detail. In the interests of transparency I have a post-graduate research qualification which includes statistical analysis and come from a professional background in writing/education, I m also a potential patient due to genetic risk. The paper demonstratively represents a vast amount of very detailed and expert work and is impressive and its altruism and intention to inform the public is clear. However as a lay-person I would have found it challenging to comprehend without a background in statistics and a working familiarity with reading research papers. My feedback represents where I found those challenges and suggests some solutions. 1) Define terms: there is no definition of the terms poor or intermediate diagnosis (page 5, 39-40) or mid prognosis (page16, line 27). These variables have to be defined. I have no idea what those variable meant to the research team. 2) Figures 1, 2 and 3 are too small to read either on screen or when printed out. They need to be larger or re-worked to make the data represented there possible to read (I have no eyesight difficulties.) 3) Presenting information: Is it possible to code each of the cancers and present this at the start of the paper. E.g. Lymphoma: Ca 1, Testicular Ca 2, Uterine Ca 3. Using this coding insert a table at the beginning indicating which cancers were included and what their coding was. This is a communication / reading tool for the reader. After that point if they could indicate which coding they are referring to in the long hand text it would help distinguish between the mass of diagnosies that are referred to. It is very difficult to keep flicking between the long hand discussion of a series of Cancers, or truncated series of Cancers, back to the beginning into the tables and on to the end. This will also help distinguish between the different results associated with the different Cancers in table 3. 4) There are some positives, however the paper concludes that little evidence was found of a direct impact..no evidence for a reduction in socio-economic inequalities in cancer survival (page 1, lines ) In the case of uterine cancer, table 3 indicates a difference of between If the data is averaged, yes, it does represent disappointing overall results but there is a glimmer of light when looking at the detail for individual Ca s which for Ca patients is really important to hear. It also helps the lay person when reading what could be seen to be contradictory information about Ca survival rates at 5 years post-diagnosis from national and international charities, for instance. 5) Using the 2004 IMD socio-economic tool is interesting (page 3, 31) as there is a more up to date tool available. It does not explain why an older version was selected and that affects reliability. In discussion (page 15,6) that is not covered. 6) Could there be a table which summarises the outcome measures within the various Cancer Plans referred to in the paper or what they said and when they were published. A lay person reading won t automatically know what this was / is and why selecting the 1 year data point as opposed to the more standard 5 year data point

6 was necessary, as one for instance. They won t also do the additional work necessary to do the background reading to inform themselves. 7) I m not sure that the discussion is balanced or that some conclusions can be drawn. It focuses on socio-economic issues but the paper only evaluates, it doesn t drill down into the data in that way. Correlation is not causation. There is an obvious point and that is that for some Cancers, survival will be affected by the fact that Cancer still kills - there is no cure. Until a cure is discovered unfortunately people will still die of it despite all the best treatment options at the NHS s disposal. Some will die faster than others because that s what some Cancers do, some of those will die within 1 year post-diagnosis. 8) Some Ca s are then subdivided within the overarching diagnostic term and those subdivisions can explain greater mortality risk. E.g. there are 2 types of Uterine Ca. Uterine Ca 1 is less of a killer. Uterine Ca 2 is typically far more invasive far more quickly and kills quickly too. This issue is not discussed and that also affect prognosis, outcome measures and validity. Although this paper may be published in the BMJ, results and conclusions will be read and communicated far wider. Patients and patient s families and friends will read that information via the media and used by healthcare and charities to educate and advise. This paper is so dense with complex and interrelated information there is a risk that unless aspects of it are made clearer and plain English is used, the complexities will get lost in favour of reading the summary conclusion on page 1 which is, in fact, quite a depressing picture. Not everyone will be able to have a presentation given by the research group in order to resolve those complexities ( page 2, 40.) I do trust the data is correct and the work that has been carried out is accurate but I also falter. I wonder if I ve understood those interrelationships properly despite reading it 4 times. A good paper should be clear, accessible and understood by all ( whether they ve got a background in research statistics or not.) If those conclusions are discussed in this paper and I ve missed them somewhere in the text or tables, it s an indication of how it needs to be reworked to make the key points clear. I also though about the following: what factors affected my family members death? They were at the top of the socio-economic scale, the Two Week Plan happened, they had treatment quickly and at an expert tertiary cancer centre. Why did they still die? Why didn t they beat a cancer that was diagnosed so quickly? Elements of the Cancer Plan, in their case worked, it really did. Their care was excellent throughout. What it tells me is that despite all positive moves to make a change, some cancers will always be more deadly than others. Perhaps that s not the point the paper wishes to make, but if not, it needs to be more explicit. Thank you. Additional Questions: Please enter your name: Ms C A Henderson Job Title: NA Institution: NA Reimbursement for attending a symposium?: No

7 A fee for speaking?: No A fee for organising education?: No Funds for research?: No Funds for a member of staff?: No Fees for consulting?: No Have you in the past five years been employed by an organisation that may in any way gain or lose financially from the publication of this paper?: No Do you hold any stocks or shares in an organisation that may in any way gain or lose financially from the publication of this paper?: No If you have any competing interests <A HREF=' lists/declaration-competing-interests'target='_new'> (please see BMJ policy) </a>please declare them here: NA Reviewer: 3 Recommendation: Comments: Important topic, and of interest to a general audience. It s a well presented overview of valuable data on progress in cancer survival and the persistent socioeconomic inequality of this issue (in England in this case), and as such it definitely deserves public and professional attention. The scope of data and statistics used are impressing but nevertheless this is an observational study of an already well-known problem. I have some troubles in identifying the precise hypothesis / intervention (Any particular aspect of the NHS Cancer Plan? Were all plans succesfully implemented, everywhere?). The identified time periods differ from a previous publication (2010) and the notion of "acceleration of positive survival trends" is not well described or defined in the Methods. The analysis is based on a comparison with historical controls and I can t see how this design could be used to assess the effectiveness of a national cancer plan concerning socioeconomic differences in cancer survival; too high a risk of bias. The assumption of assessing a direct impact (causality) of national cancer "strategies" (documents) on one-year survival of all forms of cancer seems very uncertain and just presenting survival and discussing the Cancer Plan would be better (in a scientific journal) as I see it. Also, the deprivation gap was presented as the difference between the most affluent and most deprived quintiles. There is no information on what happened to those in between. It s difficult to pinpoint what is the problem - the poorest lagging behind or the most affluent doing much better than all others? Some improvement for large groups of patients might be hidden in here and the reader is left uninformed on this important aspect. Louise Olsson

8 Additional Questions: Please enter your name: Louise Olsson Job Title: Consultant, head of unit Institution: Dep of Molecular Medicine and Surgery, KI Reimbursement for attending a symposium?: No A fee for speaking?: No A fee for organising education?: No Funds for research?: No Funds for a member of staff?: No Fees for consulting?: No Have you in the past five years been employed by an organisation that may in any way gain or lose financially from the publication of this paper?: No Do you hold any stocks or shares in an organisation that may in any way gain or lose financially from the publication of this paper?: No If you have any competing interests <A HREF=' lists/declaration-competing-interests'target='_new'> (please see BMJ policy) </a>please declare them here: Nothing Reviewer: 4 Recommendation: Comments: An interesting examination of changes in UK cancer survival. However, there are some issues that needs to be improved. Mainly in writing. Major: Deprivation gap is a primary covariate but not well described. I cannot understand what the LSOA is. It must be possible to describe this better. If nothing else, you can produce a table describing average income, occupation (blue white collar.), sex, etc etc for the 5 deprivation groups? Sex? In the results section you list separately by sex and have separate tables, but it s not an objective. If it is not an objective I would downplay sex. Just put all in one table and write male or female where needed, e.g. female breast cancer. Alternatively, (makes more sense to me) make sex a pronounced category and give some more weight on this. You may want to include a recent study from Sweden as a reference, Radkiewicz C, Johansson ALV, Dickman PW, Lambe M, Edgren G (2017). Sex differences in cancer risk and survival: A Swedish cohort study. European Journal of Cancer 84, Figure 2 could be removed: (a) Fonts to small so can only read when increased to 200% (b) To me it looks like this has been presented already in table 1 (c) What do the red bars represent? Where are the confidence intervals? In the text the x-axis is described as survival in late 1990s but the figure says 1996!

9 Writing. Too much hand waving a) e.g. fairly good, nearly all, majority of cancers, most cancer etc etc. Please be specific b) Page 8, line 47,it continues, lower among more deprived, lower than what?? more deprived, than what?? Please specify c) Next sentence: situation was similar, which situation?? similar??? which definition are you using for similar? d) Next sentence: lower survival was more common, lower survival than what? More common than what? (lower survival should be shorter survival?) e) Next sentence: Narrowing for cancers with survival in 1996 higher than 65%. I don t think this is correct, e.g. table 2 say something different for at least larynx, bladder and testis (perhaps others as well, I did not scrutinize all data). Or I m I reading the table wrong? f) Page 17, conclusion, the acceleration in cancer survival. Acceleration can be in positive or negative direction. You mean increase? Page 4, line 6. I don t understand why there are 41 combinations, 18 combinations etc. Please describe. For sex it should be 2x24 if it s the 24 most common as you write earlier, perhaps excluding a few sex specific, e.g. prostate cancer and female breast cancer? Other comments: Aims: Page 2, line 5, less or more deprived? Are all deprived? This comes again in methods section on page 4. Perhaps correct? Sounds strange to me though. Page 3 Net survival. I m not sure I would say net-survival is the only method, but it is **a** method, which do not require knowledge about death causes. You could add a sentence or two describing this better since it would help a reader to better understand the choice of method, and convincing them you have used the appropriate method. You should also briefly describe the method implemented in the stns procedure. What is meant with consistent in this context? Spell out statistically significant when this is this is the case (and not just significant which can be something else). Page 2, line 26 improving one-year survival : Was the NHS cancer plan aiming for improvement only in one-year survival? You probably mean survival? What is the purpose with the method section on page 2 before Data? Remove or motivate what you are trying to say and why it s important Page 4, confidence interval should be (I assume) 95% or perhaps 90%? Page 4, line 27, why linear regression splines? Later you check this by fitting cubic spline. Why not cubic spline only? Page 4, line 48, were milestones. They were important? Table 1, column 2, says (header) "Survival (%)". How should I interpret the information in that column, e.g (28.14 to 30.7)? It's not clear (and you may want to use the same number of decimals throughout?) Additional Questions: Please enter your name: Sven Sandin Job Title: Assistant Professor, Statistician Institution: Dpt of Psychiatry, Icahn School of Medicine at Mount Sinai, New York Reimbursement for attending a symposium?: No A fee for speaking?: No A fee for organising education?: No

10 Funds for research?: No Funds for a member of staff?: No Fees for consulting?: No Have you in the past five years been employed by an organisation that may in any way gain or lose financially from the publication of this paper?: No Do you hold any stocks or shares in an organisation that may in any way gain or lose financially from the publication of this paper?: No If you have any competing interests <A HREF=' lists/declaration-competing-interests'target='_new'> (please see BMJ policy) </a>please declare them here: