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...
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 other dietary fats (as was the case for the majority of participants in this trial). Although no significant difference was detected in weight change across groups, those in the coconut oil group gained a mean of 300g in one month, which, if sustained over a longer period, could become problematic. A longer follow up is therefore required before such an intervention can be recommended.
Reference:
1. Ganesan K, Sukalingam K, Xu B. Impact of consumption and cooking manners of vegetable oils on cardiovascular diseases- A critical review. Trends Food Sci Technol [Internet]. 2018 Jan 1 [cited 2018 Apr 15];71:132–54. Available from: https://www.sciencedirect.com/science/article/pii/S0924224417300985#bib89
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.
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...
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 is contentious that the active arms all represent gold standard CR services. Moreover, we argue that in the “current era”, standard care involves many elements of CR, as is demonstrated by the nature of the control arms in the included trials in Table 3. Hence, the review suffers from dilution in the active arms, and contamination in the control arms, rendering conclusions disputable.
On a related note, we would argue that in the current era, no research ethics board would approve a protocol for a trial comparing CR to no CR, given previous reviews demonstrating benefit.[4] Powell et al. call for more such trials to establish efficacy. Hence, ultimately it is not possible to “prove” the effectiveness of CR in developed countries in the current era. We postulate that the benefits of CR in low and middle-income countries would be greater than is observed in Europe and other Western countries, given the high burden of risk factors, which go unrecognized and uncontrolled. Given so few patients access CR in these countries, [13] perhaps it is ethical to run a rigorous test of the effects of CR versus standard care (which likely is not as “enhanced” [14] as it is in high-income countries). There are no trials with mortality or morbidity outcomes in these settings to our knowledge, [15] and hence such trials are encouraged, with an eye to ultimately augmenting the number of patients who access CR in these regions.
Third, and finally, Powell et al. also raise the importance of other key outcomes that were not considered in their review, namely quality of life and cost. We would also argue that additionally the central role of CR in coordination of outpatient, chronic cardiovascular care not be overlooked. It is only through CR programs that all cardiovascular secondary prevention recommendations [16] can be fully implemented, and especially that this can be achieved in a patient-centered way.
References
1. Ghisi GM, Chaves G, Bennett A, et al. The effects of cardiac rehabilitation on mortality and morbidity in women: A meta-analysis attempt. J Cardiovasc Prev Rehab 2018. [in press].
2. Rauch B, Davos CH, Doverty P, et al. The prognostic effect of cardiac rehabilitation in the era of acute revascularisation and statin therapy: A systematic review and meta-analysis of randomized and non-randomized studies – The Cardiac Rehabilitation Outcome Study (CROS). Eur J Prev Cardiol 2016;23:1914-39.
3. Sumner J, Harrison A, Doherty P. The effectiveness of modern cardiac rehabilitation: A systematic review of recent observational studies in non-attenders versus attenders. PLoS One 2017; 12:e0177658.
4. Anderson L, Oldridge N, Thompson DR, et al. Exercise-Based Cardiac Rehabilitation for Coronary Heart Disease Cochrane Systematic Review and Meta-Analysis. J Am Coll Cardiol 2016;67:1–12.
5. van Halewijn G, Deckers J, Tay HY, et al. Lessons from contemporary trials of cardiovascular prevention and rehabilitation: A systematic review and meta-analysis. Int J Cardiol 2017:232:294-303.
6. Santiago de Araújo Pio C, Marzolini S, Pakosh M, et al. Effect of Cardiac Rehabilitation Dose on Mortality and Morbidity: A Systematic Review and Meta-regression Analysis. Mayo Clin Proc 2017;92:1644-59.
7. Price KJ, Gordon BA, Bird SR, et al. A review of guidelines for cardiac rehabilitation exercise programmes: Is there an international consensus? Eur J Prev Cardiol 2016;23:1715-33.
8. Swift DL, Lavie CL, Johannsen NM, et al. Physical activity, cardiorespiratory fitness, and exercise training in primary and secondary coronary prevention. Circ J 2013;77:281-92.
9. Hoffman TC, Glasziou PP, Boutron I, et al. Better reporting of interventions: template for intervention description and replication (TIDieR) checklist and guide. BMJ 2014;348:g1687.
10. Craig P, Dieppe P, Macintyre S, et al. Developing and evaluating complex interventions: the new Medical Research Council guidance. BMJ 2008; 337: a1655.
11. Grace SL, Turk-Adawi KI, Contractor A, et al. Cardiac rehabilitation delivery model for low-resource settings. Heart 2016;102:1449-55.
12. Grace SL, Turk-Adawi KI, Contractor A, et al. Cardiac rehabilitation delivery model for low-resource settings: An international council of cardiovascular prevention and rehabilitation consensus statement. Prog Cardiovasc Dis 2016;59:303-22.
13. Turk-Adawi K, Sarrafzadegan N, Grace SL. Global availability of cardiac rehabilitation. Nature Reviews: Cardiology 2014;11: 586-96.
14. Freedland KE, Mohr DC, Davidson KW, et al. Usual and unusual care: existing practice control groups in control groups in randomized controlled trials of behavioral interventions. Psychosom Med 2011;73:323-35.
15. Turk-Adawi K, Grace SL. Narrative review comparing the benefits of, participation cardiac rehabilitation in high-, middle- and low-income countries. Heart Lung Circ 2015;24:510-20.
16. Piepoli MF, Hoes AW, Agewall S, et al. 2016 European Guidelines on cardiovascular disease prevention in clinical practice: The Sixth Joint Task Force of the European Society of Cardiology and Other Societies on Cardiovascular Disease Prevention in Clinical Practice. Eur Heart J 2016;37:2315-81.
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...
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 over the reduction in hypertensives between 2012 and 2015. But risk factors for hypertension being largely behavioral and life style related how does increased insurance coverage have an impact over reduction of the risk factors and the resulting hypertension as well.
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...
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] also emphasise that Anderson et al[2] were able to demonstrate a profoundly positive effect of CR overall, even if the effect upon total mortality has lessened.
Anderson et al’s Cochrane review[2] and subsequent letter[3] attributes the diminished benefit of CR on mortality to both the ‘changing face of usual care’ and to the recruitment of more mixed populations of cardiac patients to recent studies. Certainly in CR clinical practice there is an ageing demographic presenting with increasingly complex multi-morbidity, and prevalence of heart failure. Whilst a 2014 Cochrane Review[5] clearly demonstrated that CR significantly improves health-related quality of life and reduces the risk of hospital re-admissions in heart failure, there is growing recognition that, with longevity, some patients are willing to trade off improved quality over quantity of life[6]. Although optimal medical therapy and percutaneous intervention importantly add ‘years to life’, CR’s potential to add ‘life to years’ should not be underplayed.
Furthermore, Powell et al[1] cites the 2012 BACPR Standard and Core Components[7]; these were revised and re-published in June 2017[8]. Regardless, both editions highlight the importance of including all core components (a co-ordinated ‘sum’ of activities – not just exercise training) to enable the patient to resume optimal functioning in their community and slow or reverse progression of their disease. Although 16 of the 22 trials reviewed by Powell et al[1] specified inclusion of additional components of a comprehensive CR programme, the article remains focused upon exercise.
Moreover, the authors promote this paper as being contemporary, yet they include papers where recruitment periods are unconfirmed. The CROS study[9] of modern era CR did manage to confirm waiting times in their review, which often required contacting the authors. Powell et al[1] also included stable angina patients, a group that are ‘out of scope’ and not recommended for CR in NICE guidance. The version of ‘CR’ offered includes vague reference to supervised and non-supervised options and inclusion of a pedometer based activity programme. Given the huge variability in intervention type and patient populations, it’s hardly surprising that no mortality benefits were seen.
In the review, the follow-up periods are generally very short – is it reasonable to expect a survival benefit over a matter of weeks? Taylor et al[10] recently demonstrated that long-term participation (>36 months) in supervised CR exercise within an extended, community-based maintenance programme may be associated with a significant stepwise survival benefit, compared to all shorter CR durations, irrespective of patient fitness level at programme entry. Indeed it seems logical that any survival benefit resulting from exercise-based cardiac rehabilitation is likely to be derived from sustained regular patterns of activity. Given that the authors state that several included trials failed to report on exercise ‘dose’, it would have been useful to display other outcomes obtained from these trials, to identify potential ‘under-dosage’ and provide a more comprehensive overall picture of the evidence.
Finally, it must be stated (as a reminder) that despite the conclusions of this review, in our current era where drugs are approved at £20K per quality-adjusted life year, and some established CVD interventions have <50% success rate, CR remains the most cost-effective intervention that we should offer all patients. Cardiac rehabilitation is not just about exercise and mortality!
Aynsley Cowie, Scientific Officer BACPR, Research Officer Association of Chartered Physiotherapists in Cardiac Rehabilitation (ACPICR)
Scott W Murray, President BACPR
Sally Hinton, Executive Director BACPR
Hasnain M Dalal, Council Officer BACPR
Simon J Nichols, Council Officer BACPR
Rod Taylor, Cochrane Rehabilitation Review Coordinator
Patrick Doherty, Director National Audit of Cardiac Rehabilitation (NACR)
Laura Burgess, Chair, BACPR-Exercise Professionals Group; Co-chair-ACPICR
[1] Powell R, McGregor G, Ennis S, et al. Is exercise-based cardiac rehabilitation effective? A systematic review and meta-analysis to re-examine the evidence. BMJ Open 2018; 8(3):e019656
[2] Anderson L, Oldridge N, Thompson DR, et al. Exercise-based cardiac rehabilitation for coronary heart disease: Cochrane systematic review and meta-analysis. J Am Coll Cardiol 2016; 67:1-12
[3] Taylor RS, Anderson N, Oldridge N, et al. The efficacy of exercise-based cardiac rehabilitation. JACC 2017; 69(9): 1204-12
[4]Taylor RS, Sagar VA, Davies EJ, et al. Exercise-based rehabilitation for heart failure. Cochrane Database Syst Rev. 2014 Apr 27;(4):CD003331. doi: 10.1002/14651858.CD003331.pub4.
[5] Lavie CL, Arena R & Franklin BA. Cardiac rehabilitation and healthy life-style interventions: Rectifying program deficiencies to improve patient outcomes. J Am Coll Cardiol 2016; 67:13-15
[6] Lewis EF, Johnson PA, Johnson W, et al. Preferences for quality of life or survival expressed by patients with heart failure. The Journal of Heart and Lung Transplantation 2001; 20(9):1016–24
[7] BACPR, Buckley JP, Furze G, Doherty P, Speck L, Connolly S, et al. BACPR scientific statement: British standards and core components for cardiovascular disease prevention and rehabilitation. Heart 2013; 99(15):1069-71.
[8] BACPR. The BACPR Standards and Core Components for Cardiovascular Disease Prevention and Rehabilitation, 3rd Edition, 2017. London: BACPR.
[9] Rauch B, Davos CH, Doherty P, et al & ‘Cardiac Rehabilitation Section’, European Association of Preventive Cardiology (EAPC). The prognostic effect of cardiac rehabilitation in the era of acute revascularisation and statin therapy: A systematic review and meta-analysis of randomized and non-randomized studies - The Cardiac Rehabilitation Outcome Study (CROS)' European Journal of Preventive Cardiology. DOI: 10.1177/2047487316671181
[10] Taylor C, Tsakirides C,Moxon J, et al. Exercise dose and all-cause mortality within extended cardiac rehabilitation: a cohort study. Open Heart 2017;4:e000623. doi:10.1136/openhrt-2017-000623
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.
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...
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 at least as results of univariate analyses. It looks like two laboratory test parameters with P values < 0.05 have been removed from the model; alanine aminotransferase (ALT) and blood urea nitrogen. Makino et al. reported that liver fibrosis affected the pathogenesis of pseudoaldosteronism via decreased function of multidrug resistance-associated protein 2 (2), and some clinical scores show that ALT is a good marker of liver fibrosis (3,4). Therefore, including these items in the model is both statistically and physiologically reasonable.
Another concern is the handling of continuous variables, baseline serum potassium level and age, in the Cox proportional hazard model.
The authors considered baseline serum potassium level as a binary variable in the analysis, although it was originally a continuous variable, and the threshold value (≧4.1 mEq/L) was established by additional analysis. Conversely, age was considered a continuous variable and the authors did not perform the same additional analysis. The authors could avoid such discrepancy by handling continuous variables in the same manner.
In addition, the range of age in their data is roughly 50-90 years, and the variable has approximately 40 units. It is vital to check the proportionality of age with the Schoenfield residual plot. If outliers exist, especially in the younger age, this variable could have many more units. The Cox proportional hazard model is very prone to being distorted by even a few outliers. Furthermore, it is difficult to compare the hazard ratio of age with that of the other binary variables, which only have one unit (0 or 1), even if there was a significant effect of age.
Two more minor points are the lack of the number at risk in the bottom of Kaplan-Mayer curve and information of observation period in the method section.
Overall, we congratulate the authors on their interesting and clinical-oriented research. The spread of alternative and complementary medicine opens up a new problem, one of safety. Humans now have greater longevity than humans did in the ancient era when many traditional herbal formulas were developed. Healthcare providers should be cautious about side effects of alternative and complementary medicines, especially elderly patients with slower drug metabolism and excretion. More studies should be conducted regarding the pharmacokinetics of compounds made from liquorice, and the clinical risk factors of pseudoaldosteronism in other populations.
References
1. Shimada S, Arai T, Tamaoka A, et al. Liquorice- induced hypokalaemia
in patients treated with Yokukansan preparations: identi cation of the risk factors in a retrospective cohort study. BMJ Open 2017;7:e014218. doi:10.1136/ bmjopen-2016-014218
2. Makino T, Ohtake N, Watanabe A, et al. Down-Regulation of a Hepatic Transporter Multidrug Resistance-Associated Protein 2 Is Involved in Alteration of Pharmacokinetics of Glycyrrhizin and Its Metabolites in a Rat Model of Chronic Liver Injury. Drug Metab Dispos 2008;36:1438-43. doi:10.1124/dmd.108.021089
3. Sterling RK, Lissen E, Clumeck N, Sola R, Correa MC, Montaner J, S Sulkowski M, Torriani FJ, Dieterich DT, Thomas DL, Messinger D, Nelson M, APRICOT Clinical Investigators. Development of a simple noninvasive index to predict significant fibrosis in patients with HIV/HCV coinfection. Hepatology. 2006 Jun; 43(6):1317-25. doi:10.1002/hep.21178
4. Imbert-Bismut F, Ratziu V, Pieroni L, Charlotte F, Benhamou Y, Poynard T. Biochemical markers of liver fibrosis in patients with hepatitis C virus infection: a prospective study. Lancet. 2001;357:1069-1075. doi:10.1016/S0140-6736(00)04258-6
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...
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 exercise on mortality outcomes seems to be in decline. Such a decline is strongly associated with both the general population reductions in premature mortality and improved secondary prevention outcomes from advancing aggressive medical therapies. In the UK in 2013 the British Association of Cardiovascular Prevention and Rehabilitation (BACPR) specifically highlighted this phenomenon in their published Standards and Core Components (4). To negate to comment in any publication about the efficacy of cardiac rehabilitation beyond the single metric of mortality seems contrary to the World Health Organization (WHO)’s goals for CR stated in 1993. The WHO continues to assert that CR aims to help people regain as best as possible their active/productive place in society, and where possible achieve this by supporting people via their own means to achieve this. What this current article could highlight more, even though it was not analysed, that just because CR has had a a more recent declining influence on mortality, it is effective in helping people live better lives whilst managing a major chronic disease. Although more good evidence in this domain is certainly required.
Exercise participation and fitness outcomes
This article comes as no large surprise, as the most recent systematic reviews of the past five years have all demonstrated a reducing influence on mortality. However, this still doesn’t negate the fact that part of this effect could be due to the delivery (not the design) of the trials. The analyses of Powell et al., failed to be critical on probably the most important outcome of exercise-based rehabilitation; changes to fitness and exercise capacity (5). Did the participants in the studies actually “adhere” to the required exercise prescriptions and physical activity guidance, so that such changes in fitness could occur? Although it was good to see Powell et al. had clearly summarised and highlighted the exercise doses reported by the included studies, there was no mention of compliance, which was in fact the biggest criticism of the RAMIT trial (6-8), which raised alarm around the CR community. In 2016 and 2017, Taylor et al., added weight to the argument that maintaining fitness and long-term adherence to frequent physical activity were the strongest influences from exercise on reducing mortality risk(9, 10).
The only true way to determine if participants have complied with the recommendations is to assess whether their aerobic fitness and muscular strength/endurance has actually improved following CR. The lack of reporting on “fitness changes” in the CR literature continues to plague the whole history of outcomes for many trials and audits, as highlighted by Sandercock et al. (8)
An example in-point of accounting for changes in compliance and fitness is the HF-Action multicentre trial on exercise rehabilitation for heart failure(11). As one of the largest single CR trials, they showed no effect on mortality or re-hospitalisaton and other clinical outcomes. Initially there was great surprise, but then a very significant amount of non-compliance to participation was highlighted. In re-analysed data, that accounted for non-compliance, the clinical effects then became quite prominent and revealed a very strong dose-response of MET-hours attained and clinical outcomes(12).
Old evidence for new programmes in developing countries
A main tenet of the ICCPR is to promote and guide the growth of CR in developing countries, especially where cardiology care provision is weak(13). Until such a time when developing countries can deliver modern cardiology services to the same standards of developed nations, is not the “Old CR pill” (3) of mortality prevention still of value? The ICCPR has embarked on a number of initiatives, with the WHO’s ambitions “at heart”, to promote effectiveness in all core components, and hence ensure the benefits of CR (not just for exercise and mortality) are achieved(14-18).
**The ICCPR (www.globalcardiacrehab.com) is an International Associate Member of the World Heart Federation. It is comprised of approximately 35 associations/societies from around the world with an interest in promoting prevention and rehabilitation programmes for people with cardiovascular disease. The comments within this article represent an “overarching” consensus of leaders from ICCPR member associations, and therefore it may not represent the specific (in whole or in part) opinions of any individual member Association/Society
References
1. O'Connor GT, Buring JE, Yusuf S, Goldhaber SZ, Olmstead EM, Paffenbarger RS, Jr., et al. An overview of randomized trials of rehabilitation with exercise after myocardial infarction. Circulation. 1989;80(2):234-44.
2. Powell R, McGregor G, Ennis S, Kimani PK, Underwood M. Is exercise-based cardiac rehabilitation effective? A systematic review and meta-analysis to re-examine the evidence. BMJ open. 2018;8(3):e019656.
3. Oldridge NB, Guyatt GH, Fischer ME, Rimm AA. Cardiac rehabilitation after myocardial infarction. Combined experience of randomized clinical trials. Jama. 1988;260(7):945-50.
4. BACPR, Buckley JP, Furze G, Doherty P, Speck L, Connolly S, et al. BACPR scientific statement: British standards and core components for cardiovascular disease prevention and rehabilitation. Heart. 2013;99(15):1069-71.
5. Myers J, Prakash M, Froelicher V, Do D, Partington S, Atwood JE. Exercise capacity and mortality among men referred for exercise testing. The New England journal of medicine. 2002;346(11):793-801.
6. West RR, Jones DA, Henderson AH. Rehabilitation after myocardial infarction trial (RAMIT). Heart. 2013;99(2):146.
7. Members BEC. RAMIT presents an outdated version of cardiac rehabilitation. Heart. 2012;98(8):672; author reply 3-4.
8. Sandercock GR, Cardoso F, Almodhy M, Pepera G. Cardiorespiratory fitness changes in patients receiving comprehensive outpatient cardiac rehabilitation in the UK: a multicentre study. Heart. 2013;99(11):785-90.
9. Taylor C, Tsakirides C, Moxon J, Moxon JW, Dudfield M, Witte KK, et al. Submaximal fitness and mortality risk reduction in coronary heart disease: a retrospective cohort study of community-based exercise rehabilitation. BMJ open. 2016;6(6):e011125.
10. Taylor C, Tsakirides C, Moxon J, Moxon JW, Dudfield M, Witte K, et al. Exercise dose and all-cause mortality within extended cardiac rehabilitation: a cohort study. Open Heart. 2017;4(2):e000623.
11. O'Connor CM, Whellan DJ, Lee KL, Keteyian SJ, Cooper LS, Ellis SJ, et al. Efficacy and safety of exercise training in patients with chronic heart failure: HF-ACTION randomized controlled trial. Jama. 2009;301(14):1439-50.
12. Keteyian SJ, Leifer ES, Houston-Miller N, Kraus WE, Brawner CA, O'Connor CM, et al. Relation between volume of exercise and clinical outcomes in patients with heart failure. J Am Coll Cardiol. 2012;60(19):1899-905.
13. Grace SL, Warburton DR, Stone JA, Sanderson BK, Oldridge N, Jones J, et al. International Charter on Cardiovascular Prevention and Rehabilitation: a call for action. Journal of cardiopulmonary rehabilitation and prevention. 2013;33(2):128-31.
14. Turk-Adawi K, Sarrafzadegan N, Grace SL. Global availability of cardiac rehabilitation. Nature reviews Cardiology. 2014;11(10):586-96.
15. Babu AS, Lopez-Jimenez F, Thomas RJ, Isaranuwatchai W, Herdy AH, Hoch JS, et al. Advocacy for outpatient cardiac rehabilitation globally. BMC Health Serv Res. 2016;16:471.
16. Grace SL, Turk-Adawi KI, Contractor A, Atrey A, Campbell NR, Derman W, et al. Cardiac Rehabilitation Delivery Model for Low-Resource Settings: An International Council of Cardiovascular Prevention and Rehabilitation Consensus Statement. Progress in cardiovascular diseases. 2016;59(3):303-22.
17. Shanmugasegaram S, Perez-Terzic C, Jiang X, Grace SL. Cardiac rehabilitation services in low- and middle-income countries: a scoping review. The Journal of cardiovascular nursing. 2014;29(5):454-63.
18. Pesah E, Supervia M, Turk-Adawi K, Grace SL. A Review of Cardiac Rehabilitation Delivery Around the World. Progress in cardiovascular diseases. 2017;60(2):267-80.
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.
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.
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...
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 of the difference in trend that relates to the efficacy of hospital treatments in the two countries.”
In Scotland, we have focussed on health and social care integration. The aim of this is to ensure care is delivered to patients in their homes or a homely setting wherever possible, with hospital care reserved for only the most unwell. This would explain the observed higher rate of mortality for hospitals, but not reflect the improved care being delivered early in the patients journey or in the community settings.
The York study has raised important questions about comparative hospital mortality in England and Scotland, but does not allow us to draw strong conclusions about the variation we see. We would be very keen to collaborate to advance our understanding of these trends.
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:
Show More1. 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...
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.
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.
Show MoreFirst, 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...
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 MorePowell 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 MoreThis 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.
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 MoreIn 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
Show MoreIt 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...
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.
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 MorePages