785 e-Letters

  • Response: Randomised trial of coconut oil, olive oil or butter on blood lipids and other cardiovascular risk factors in healthy men and women

    We would like to congratulate the authors: the pragmatic trial design is novel and interesting, and may have provided a better estimate of the ‘real-life’ effectiveness of the intervention than a conventional RCT. Nevertheless, there are certain aspects of the paper that we wish to highlight. As the authors acknowledge, there are concerns about uncontrolled confounders. However, we would also like to raise the following 3 points of concern to consider:
    1. The study lacked a control group (where no additional lipid was added/replaced in the diet). This is problematic as the article’s conclusion suggests that the changes seen in LDL-C are an effect of extra virgin coconut oil, whereas it might instead be the impact of additional butter (and without a control group we are unable to assess the impact of coconut oil alone).
    2. The study is unclear about whether participants consumed the 50ml of coconut oil or olive oil as cold oils, or whether they cooked them before consumption. Multiple studies have suggested that hot (cooked) vegetable oil has a negative impact on blood lipids and metabolic profiles, increasing LDL-C levels, inflammatory markers and blood pressure (1). We would therefore welcome clarification of this point from the study authors.
    3. Longer follow up may be warranted to determine whether ongoing use of coconut oil would result in significant weight gain. This is of particular concern if coconut oil is added to the diet instead of replacing...

    Show More
  • Response to query regarding use of the Morisky scale

    During the study design period, we planned to use a Chinese version of the Morisky scale to assess medication adherence. This version was previously validated and available for use in China. In 2017, we were informed by Dr.Morisky team that we need to pay to use the scale. After careful consideration, we opted not to use the Morisky scale and declined to pay the fee. These developments occurred during the submission and review process at BMJ Open. As such, the initial version of the manuscript noted the use of the Morisky scale to assess adherence, though this was deleted from the revised and published version. Instead, we will assess adherence using detailed medication information obtained during each follow-up visit.

  • Cardiac Rehabilitation Effectiveness? A response from the Canadian Association of Cardiovascular Prevention and Rehabilitation (CACPR)

    In the January 2018 issue of BMJ Open, Powell et al. published a review which pointedly challenges the effectiveness of cardiac rehabilitation (CR) in reducing mortality. This response is forwarded with: (1) respect for the quality of this study (although we might dispute inclusion of some of the trials),[1] (2) consideration of other recent reviews of CR efficacy in the modern era, [2-5] and (3) the desire to incite a balanced discussion of the merits of CR. CACPR invites readers to take into consideration a few important factors when reading Powell et al.’s article.
    First, Powell et al. identify the importance of CR dose and patient adherence and how it was not considered in their work. Actual dose received may often be too low, such that the impact of CR is reduced. [6] There is wide variation in exercise prescriptions and how they are progressed. [7] Degree of improvement and degree of cardiorespiratory fitness is one of the major mechanisms through which CR can reduce mortality,[8] yet often this is not described in trials. In future trials, these factors need to be better documented and considered statistically. Powell et al. recommend a nice framework (see their reference 63), and the TiDIER framework provides another option.[9]
    Second, CR is a “complex”, [10] multi-component intervention.[5,11,12] As Powell et al. state, many of the included trials were exercise-only, and for those that were comprehensive, there was substantial variation. Therefore, it...

    Show More
  • Comment on: Prevalence and distribution of hypertension and related risk factors in Jilin Province, China 2015: a cross-sectional study

    First I would thank Wu J et al, for publishing their research on prevalence of hypertension & its associated risk factors in your esteemed journal and this work is of high public health priority. There is no mention on the study setting where the blood pressure was actually measured, as the variability can be expected in settings like clinic where there is a chance of white coat hypertension which might not be there if measured at home by the health worker. Though the variability in blood pressure would be taken care by taking average of three readings, maintaining uniformity in study setting would offer high internal validity. In the methodology section, there is no mention on exclusion criteria and if pregnant women were included into the study as they are of different risk profile. Behavioral factors like smoking/alcohol intake were captured through face to face interview; hence social desirability can play a role in the risk factor ascertainment. Another interesting fact about this study was current tobacco smokers were estimated as 37% (in males) vs. 8% (in females) which is different to that of China’s national level estimate of 52% (in males) and 2.7% (in females) based on Global Adult Tobacco Survey (GATS, 2015). Though both GATS and this study adopted same interview technique in assessing smoking behavior, still this region estimates varies with the national level data. In discussion, authors have highlighted increased insurance coverage could have impacted o...

    Show More
  • Response from the British Association for Cardiovascular Prevention and Rehabilitation (BACPR) in collaboration with NACR, the Cochrane Heart Rehabilitation Review Coordination Centre and ACPICR

    Powell et al’s systematic review and meta-analysis acknowledges that previous meta-analyses have included trials undertaken in the 1970s and 80s that may overestimate the contemporary benefit of exercise-based cardiac rehabilitation (CR) in terms of mortality[1]. The authors recognise the impact of the improvement in the acute medical management of patients with coronary artery disease since the turn of the century, which has led to better survival. Whilst we advocate that more reviews should acknowledge the wider clinical context when evaluating interventions, this particularly broad-brushed approach to the overall effectiveness of CR warrants several counter responses.

    Most notably, it is disappointing that the article depicts mortality as the main barometer of the effectiveness of CR, and conveys the message that the contemporary approach to CR generates no effect. In their updated 2016 Cochrane Review and meta-analysis, Anderson et al[2] acknowledge a linear reduction in all-cause mortality effect over time (i.e., with publication date) but importantly stress that promotion of CR should now focus upon reduced hospital admissions and clinically relevant improvements in quality of life, rather than mortality[3]. This view is supported by Lavie et al[4] who reiterate that CR is known to improve cardiorespiratory fitness and quality of life, and reduce cardiovascular disease (CVD) risk factors, providing cost-effective secondary CVD prevention. Lavie et al[4] al...

    Show More
  • Non-disclosure by physicans of payments from the pharmaceutical industry

    This is an interesting piece. Pharmaceutical companies have to publish payments made to individual, named physicians. (These are available at http://www.abpi.org.uk/ethics/ethical-responsibility/disclosure-uk/.) However, recipients can choose to remain unidentifiable (PMCPA Clause 24.9). In the figures for 2016, published in Disclosure UK in June 2017, 35% of health care professionals chose to remain incognito. In the GMC's Good Medical Practice they only recommend that physicians consider the PMCPA code. Perhaps they could be firmer in their recommendation.

  • Re: Liquorice-induced hypokalaemia in patients treated with Yokukansan preparations: identification of the risk factors in a retrospective cohort study

    We read, with great interest, the retrospective cohort study by Shimada et al. on liquorice-induced hypokalaemia in patients treated with Yokukansan (YK) preparations (1). Their findings suggest that monitoring of serum potassium should be performed at least monthly in patients with the following risk factors: YK administration, co-administration of potassium-lowering drugs, hypoalbuminaemia, and administration of a full dose of YK preparation. These recommendations can be easily applied in a daily clinical setting, and might reduce future pseudoaldosteronism cases. However, we would like to highlight two points that we believe should be considered when interpreting the results of the study.

    Our first concern is the uncertainty of the criteria for selecting variables in the Cox proportional hazard model.

    Table 3, “Demographic data of the subjects”, shows the P value of each item.
    The authors included items with P value < 0.05 in the Cox proportional hazard model, namely YK administration, co-administration of potassium-lowering drugs, hypoalbuminaemia, administration of a full dose of YK preparation, and baseline serum potassium. At the same time, sex and age were included in the model shown in table 4 even though they had P values > 0.05.

    As been pointed out by one reviewer, Eiseki Usami, authors should clarify a relationship between table 3 and 4. If authors have analyzed all items in table 3, it should be mentioned and shown in table 4 a...

    Show More
  • Cardiac Rehabilitation Effectiveness? *A commentary from the International Council of Cardiovascular Prevention and Rehabilitation

    In the January 2018 issue of BMJ Open, Powell et al. published a well-performed systematic review/meta-analyses, which pointedly challenges the effectiveness of cardiac rehabilitation (CR) (2). Our commentary to follow is forwarded with respect for the quality of this study, but we challenge how they chose to define the term “effective” CR. We focus on three areas of concern in relation to this article:

    i. the value of mortality as the key metric for judging effectiveness;

    ii. the lack of analysis of exercise participation (compliance/adherence) which impacts on aerobic fitness as probably the single most important exercise component outcome linked with morbidity and mortality, and

    i. the value of CR and lifestyle interventions in developing countries who do not yet have well-structured, aggressive and modern medical approaches to preventing or treating coronary heart disease.

    Mortality Metric
    It would have been more fitting if this article were titled: Does exercise-based cardiac rehabilitation still lead to reduced mortality and (re) hospitalisation? To suggest it is not effective is a bit of an over-generalisation. Over the past four decades CR’s effectiveness has been assessed across numerous outcomes bio-medically, psychosocially and health-economically. Granted, this most recent review has added to an evolving picture that since the first systematic reviews by Oldridge et al. and O’Connor et al in the 1980s (1, 3) the impact of ex...

    Show More
  • BMJ Open response to Prof Marcora

    The editors would like to thank Prof Marcora for his comments.

    We agree that Prof Noakes should have declared his involvement in the original conception and design of this study as a competing interest, and hence we invited Prof Noakes to submit a response on the article to explain his position as one of the three reviewers of the completed study. As stated in our previous response, we do not feel that Professor Noakes' involvement with the protocol has compromised the integrity of the peer review process.

    Professor Noakes has also commented further on his role in the protocol, which you can view at http://bmjopen.bmj.com/content/6/1/e009301.responses#my-contribution-to-....

    As an expert in the field, Professor Noakes has made a valuable contribution as a reviewer on this paper. We do not feel that the value of the review would be diminished had Professor Noakes mentioned his prior involvement with the study protocol.

  • Key differences between England and Scotland Mortality

    We write to highlight some questions which arise from the recently published paper entitled “How do time trends in in-hospital mortality compare? A retrospective study of England and Scotland over 17 years using administrative data” by María José Aragón and Martin Chalkley in BMJ Open 2018;8:e017195.

    Seeking to understand variations in hospital associated mortality is a worthy endeavour and we welcome the authors’ contribution. They are clearly aware of the importance of case mix adjustment when studying comparative mortality, but may not have been able to take account of differences in the way hospital activity data are collected in the two health care systems.

    The recording of comorbidity in English HES data may be more complete than in our SMR01s. The reason for this is the financial incentive to fully cover background risk in England. Whilst Scottish data are getting better, the practical consequence of this for the York study is that they will have underestimated the risk of death for our patients, thereby increasing the relative mortality parameters used to calculate the risk adjusted trends used in the plots shown.

    The paper suggests “If for example alternative settings to which terminally ill patients can be discharged have expanded faster in England than in Scotland, we would observe the kind of differential trend of in-hospital mortality established by our analysis. The second, more worrying possibility is that there remains some element o...

    Show More