Medication Non-Adherence in Rheumatology, Oncology and Cardiology: A Review of the Literature of Risk Factors and Potential Interventions

Round 1

Reviewer 1 Report

The authors report a broad narrative review of factors that are associated with adherence to treatment in three areas of therapeutics – rheumatology, oncology and cardiology. They acknowledge that there is a large literature on focussed areas of the three clinical specialties but argue for a broader sweep of the risk factors, including those that are cross-cutting and inform interventions.

In those respects, they have a point. However, their review is not truly systematic and appears to be more influenced by the informal consensus of the expert panel than a rigorous attempt to find all relevant studies and assess their overall strength, including their risk of bias.

My main comments are as follows:

1)      Literature search. The authors don’t say whether had support from an information specialist/ library scientist when developing their list of key words and executing the search. As far as I can tell the principal searches use the key words ‘adherence’ and ‘compliance’. I am unsure if any other key words were tried? Other terms in the literature are ‘persistence’ and ‘concordance’. I accept that most relevant publications will be retrieved with the authors’ terms, but I wouldn’t call the literature search truly comprehensive. I could not find a detailed literature search algorithm.

2)      I appreciate the authors have a depth of experience and knowledge in the relevant clinical specialties but to what extent were they supported by information specialists (see above), clinical epidemiologists with expertise in study designs and appraisal and scientists with content knowledge in adherence?   

3)      To provide some idea of the sheer scale of this literature I ran a brief search in Google Scholar on the terms ‘adherence’ or ‘compliance’ or ‘persistence’ or ‘concordance’ and ‘systematic review’. There were literally hundreds of systematic reviews and even some overviews of systematic reviews. Many included assessments of design and risk of bias in component studies. Many had a narrow focus on a form of treatment or disease area (e.g., diabetes, COPD), but some took a broader review of factors that impair adherence. There were several systematic reviews and meta-analyses of randomised trials of adherence promoting strategies. I appreciate that the authors took a broad cross-cutting approach to their review and added value through their clinical expertise. But I am unclear how their narrative review adds to the existing extensive literature.

4)      The authors use the phrase ‘high degree of evidence’ to refer to the quality of the component studies and systematic reviews that they retrieved. I don’t know what metric they have used to make this judgment. There are many tools to assess risk of bias / study quality for different study designs – randomised trials, cohort and case-control and cross-sectional studies. In addition, AMSTAR is widely used to appraise systematic reviews.  I am unclear if any of these were used, or if the GRADE approach was employed to provide an overall summary of the strength of evidence. Perhaps there is a separate technical document or supplementary file that addresses these issues. If so I apologise but I didn’t see it.

5)      In terms of production, the writing is generally good but there are large sections of unbroken text that reduce the readability of the paper. The figure showing the taxonomy of risk factors for adherence is good (Fig 2) but the tables are unusually formatted with unnecessary centering of text, which reduces legibility.

In conclusion this is a comprehensive review of risk factors for treatment adherence across several major clinical specialty areas. It has been performed by authors with deep clinical expertise, so their interpretation of the literature is valuable as is the cross-cutting approach finding themes that are common across clinical specialties. However, the report makes this sound like a traditional expert narrative review. That is not a deal breaker but the processes do not appear to have been structured and systematic and there is little detail on methods – ranging from the literature review, techniques for assessing strength of evidence and the approach used to achieve expert consensus. While I am sure this report contains useful information I am not able to determine whether it adds usefully to the existing literature.    

Author Response

Response to reviewer #1

The authors would like to thank the receipt of reviewers’ comments, which will certainly improve the quality of the manuscript. Please, find a point-by-point response to the reviewer’s comments below.

Comment 1. Literature search. The authors don’t say whether had support from an information specialist/ library scientist when developing their list of key words and executing the search. As far as I can tell the principal searches use the key words ‘adherence’ and ‘compliance’. I am unsure if any other key words were tried? Other terms in the literature are ‘persistence’ and ‘concordance’. I accept that most relevant publications will be retrieved with the authors’ terms, but I wouldn’t call the literature search truly comprehensive. I could not find a detailed literature search algorithm.

Response: Thank you very much for your appreciation. The terms were chosen based on the authors' own previous experience and a preliminary search for terms. Also note that when you enter search terms, PubMed automatically includes alternate spellings and MeSH terms related to the phrase. In this sense, as in any literature review, some papers may have been excluded by the search methodology itself, although we believe that the manuscripts dealing with the topic under analysis use some of the search words used or that are included in the MeSH terms.

Comment 2. I appreciate the authors have a depth of experience and knowledge in the relevant clinical specialties but to what extent were they supported by information specialists (see above), clinical epidemiologists with expertise in study designs and appraisal and scientists with content knowledge in adherence?  

Response: Thank you for the opportunity to clarify this aspect of the work. The members of the Steering Committee are also experts in adherence issues in their therapeutic areas. Some of them are even part of working groups on this topic in their respective Scientific Societies.

In addition, they also have experience in conducting literature reviews. In this sense, Ascendo Sanidad & Farma, with extensive experience in this type of task, has provided support by organizing the search and offering methodological support in the project.

This information has been added to the text (page 2, lines 95-97; page17, lines 524-525).

Comment 3. To provide some idea of the sheer scale of this literature I ran a brief search in Google Scholar on the terms ‘adherence’ or ‘compliance’ or ‘persistence’ or ‘concordance’ and ‘systematic review’. There were literally hundreds of systematic reviews and even some overviews of systematic reviews. Many included assessments of design and risk of bias in component studies. Many had a narrow focus on a form of treatment or disease area (e.g., diabetes, COPD), but some took a broader review of factors that impair adherence. There were several systematic reviews and meta-analyses of randomised trials of adherence promoting strategies. I appreciate that the authors took a broad cross-cutting approach to their review and added value through their clinical expertise. But I am unclear how their narrative review adds to the existing extensive literature.

Response: We appreciate the reviewer's feedback since it allows us to clarify in further detail how this work differs from other published studies. The purpose of this evaluation is to add the clinical utility perspective of specialists to the previously conducted literature review. It also varies from others published studies in that it explores three domains of medicine in the same work and combines the findings from these distinct branches of medicine together. Therefore, we feel that this review can be beneficial for clinicians in these three therapeutic domains, as it will provide them with a more thorough understanding of the problem of therapeutic adherence as well as an update on what is occurring in their field of management.

Comment 4. The authors use the phrase ‘high degree of evidence’ to refer to the quality of the component studies and systematic reviews that they retrieved. I don’t know what metric they have used to make this judgment. There are many tools to assess risk of bias / study quality for different study designs – randomised trials, cohort and case-control and cross-sectional studies. In addition, AMSTAR is widely used to appraise systematic reviews.  I am unclear if any of these were used, or if the GRADE approach was employed to provide an overall summary of the strength of evidence. Perhaps there is a separate technical document or supplementary file that addresses these issues. If so I apologise but I didn’t see it.

Response: After reading your comment we have realized the error in the use of the term evidence in the context used. For this reason, we have replaced the term evidence with confidence in most of the text, since it is more in line with the methodology used to classify risks and non-adherence interventions.

Comment 5. In terms of production, the writing is generally good but there are large sections of unbroken text that reduce the readability of the paper. The figure showing the taxonomy of risk factors for adherence is good (Fig 2) but the tables are unusually formatted with unnecessary centering of text, which reduces legibility.

Response: We would like to thank the reviewer for pointing out these areas for improvement. We have changed and enhanced the language throughout the entire text, as well as the tables and figure 2, in order to enhance comprehension of the study.

Author Response File: Author Response.pdf

Reviewer 2 Report

Dear authors, thank you for this interesting literature review on medication adherence. This topic is always evolving and a difficult challenge for all. The information your team extracted from the literature is important, but needs to be organized and presented succinctly, only because the target audience are very busy and possibly will not pay attention to the content if not presented succinctly and with the take home (summary) message. If you could provide actionable steps for the clinician and health care managers, that would improve the article immensely. I have attached some line by line recommendations but also embedded in the attached document is the Beers Criteria which exemplifies expert committee lit review and an example of how to effectively present it.  Looking forward to the article should you decide to revise.

Comments for author File: Comments.pdf

Author Response

Response to reviewer #2

The authors would like to thank the receipt of reviewers’ comments, which will certainly improve the quality of the manuscript. Please, find a point-by-point response to the reviewer’s comments below.

Comment 1. Remove “a” in …most clinical scenarios a 50% of patients….

Response: As requested by the reviewer, we have removed “a” in the sentence and also a native English made a review on all the text (page 2, line 53).

Comment 2. Regarding the limitation of the World Health Organization definition of adherence.

Response: We thank the reviewer for the comment. In relation to your comment we have added in the discussion section: “Therapeutic adherence depends on numerous factors, classified by the World Health Organization in five dimensions depending on their origin [1]: socioeconomic, healthcare system, patient, condition, and therapy. It should be noted that WHO classification represents one perspective on adherence but that it may not be all inclusive. However, although this classification may seem limited, it includes a wide variety of risk factors for non-adherence, such as patients’ beliefs and perceptions about treatment or the severity of the disease, or the readability of directions of use.” (page16, lines 442-448)

Comment 3. Rewriting sentence.

Response: We are grateful to the reviewer for bringing this to our attention. We have replaced “In the field of rheumatology, although there are broad reviews [20–22], some issues associated with this medical specialty are still not covered. “ by “Although there are broad reviews on the topic of rheumatological medication adherence [20–22], some issues associated with this medical specialty are still not covered. “ (page2, lines 73-75).

Comment 4. Regarding to the method to conduct the review and derive conclusions.

Response: Due to the question, we believe that the methodology section was not adequately explained; therefore, we have added new wording and phrases to this section (page 2, lines 95-97; page 4 lines 120-126, lines 139-141). Consensus was reached in the various workshops (Figure 1), in which the risks and/or interventions were classified (according to category or confidence level, red, orange or green) and then agreed upon through a voting process of all members of the steering committee.

Comment 5. Regarding the Steering Committee explanation.

Response: Thanks to the reviewer's comment, we have replaced “The project was carried out by a Steering Committee made up of experts on the fields of rheumatology (1 expert), oncology (1 experts), cardiology (1 expert), general medicine (3 experts), hospital pharmacy (1 expert) and community pharmacy (1 expert).” with “The project was carried out by a Steering Committee consisted of 1 rheumatologist, 1 oncologist (1 experts), 1 cardiologist, 3 general practicioners, 1 hospital pharmacist and 1 community pharmacist” (page2, lines 93-95)

Comment 6. Add a discussion section

Response: As suggested by the reviewer, a discussion section was added (page 16, lines 441-490).

Comment 7. Consider reformatting to report what this means for clinicians and how they can use this information versus reporting every detail of the studies reviewed.

Response: Due to this comment a conclusion section with more useful information for clinicians has been rewritten and included (page 17, lines 494-513).

Comment 8. Regarding results section format of figure 2 and tables.

Response: As suggested by the reviewer, Figure 2 and tables have been reformatted.

Comment 9. To my knowledge, every Cochrane review has concluded that an intervention that

improves adherence has not been found to date. As the authors point out in line 481,

the Cochrane reviews have noted that the multifactorial nature of adherence makes it

challenging.

Response: We apologize for not being able to understand your comment, could you please rewrite it? Thank you very much

Comment 10. Regarding rewriting the conclusion based on the data analysis and presented data.

Response: As suggested by the reviewer, a new conclusion has been included, based on the data analysis and presented data (page 17, lines 494-513).

Comment 11. Regarding including in the conclusion the risks and interventions that were the most frequently identified across all categories, and a conclusion statement about the need for future work might fit here.

Response: As suggested by the reviewer, a new conclusion has been included, based on the data analysis and presented data (page 17, lines 494-513).

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

The authors have submitted a minor revision - I think its up to the editorial staff to decide if the resulting manuscript fits with the mission of the Journal. The authors have moved the emphasis a little from an evidence based approach with a systematic review and risk of bias assessment to a narrative summary based on an expert review of the published literature. They now use the word 'confidence' to denote their view of the quality of the component studies. They have improved the presentation tables - they are more legible. 

Author Response

Response to reviewer #1

The authors would like to thank the receipt of reviewers’ comments, which will certainly improve the quality of the manuscript. Please, find a point-by-point response to the reviewer’s comments below.

Comment 1. The authors have submitted a minor revision - I think its up to the editorial staff to decide if the resulting manuscript fits with the mission of the Journal. The authors have moved the emphasis a little from an evidence based approach with a systematic review and risk of bias assessment to a narrative summary based on an expert review of the published literature. They now use the word 'confidence' to denote their view of the quality of the component studies. They have improved the presentation tables - they are more legible.

Response: Thank you very much for your comments. We look forward to the editorial decision.

Thank you very much for taking the time to improve the article.

Author Response File: Author Response.docx

Reviewer 2 Report

The authors wanted help understanding comment 9- the point of comment 9 was to make clear that medication adherence is highly individual.
 
They have addressed this comment in the manuscript around line 549:
 Furthermore, some of the factors identified in the literature (such as patients’ treatment cost related factors) cannot be extrapolated to any patient, disease, or healthcare system, and should therefore be assessed on an individual basis.
 
I recommend another peer reviewer's review of this version before acceptance.  But per my review, I have no further comments or recommendations.  As an editor myself, I believe that two peer reviewers are necessary to make an acceptable manuscript.

Author Response

Response to reviewer #2

The authors would like to thank the receipt of reviewers’ comments, which will certainly improve the quality of the manuscript. Please, find a point-by-point response to the reviewer’s comments below.

Comment 1. The authors wanted help understanding comment 9- the point of comment 9 was to make clear that medication adherence is highly individual. They have addressed this comment in the manuscript around line 549: “Furthermore, some of the factors identified in the literature (such as patients’ treatment cost related factors) cannot be extrapolated to any patient, disease, or healthcare system, and should therefore be assessed on an individual basis.”

Response: Thank you very much for the explanation. We are glad to hear that your comment was answered in the last version.

 

Comment 2. I recommend another peer reviewer's review of this version before acceptance.  But per my review, I have no further comments or recommendations.  As an editor myself, I believe that two peer reviewers are necessary to make an acceptable manuscript.

Response: Thank you for your comments. Another reviewer has been evaluating in parallel the work you recommend.

Thank you very much for the time spent improving this article.

Author Response File: Author Response.docx