Effects and Working Mechanisms of a Multilevel Implementation Program for Applying Shared Decision-Making while Discussing Systemic Treatment in Breast Cancer

Round 1

Reviewer 1 Report

General comment: This manuscript applied an implementation program to establish shared decision making in medical consultations with breast cancer patients. Both topics (SDM and implementation) are important and thus the paper is innovative and interesting. However, the presentation of the methods, the results and the discussion is hard to follow given the high amount of outcomes and the lack of a clear structure according to the research questions. Therefore, my main concerns are

(i) to better structure each section according to well-defined research questions and (ii) to streamline the outcomes/ the research questions. Abstract: - Generally, the abstract is well written - The outcomes “redistribution of tasks” and “discussing feedback from consultations” in the results are not mentioned in the methods before. Accordingly, the conclusions are hard to follow Introduction: - It is stated that “Chemotherapy or hormone therapy improves survival, but this benefit only applies to a small proportion of patients” – please give a citation - It is stated that SDM “is especially challenging for clinicians to achieve in decisions about systemic therapy with (older) breast cancer patients” – please report some reasons for this - In the second last paragraph, some previous implementation studies are mentioned: are these from the same working group, i.e., is the current manuscript building upon a previous project? Please clarify - The authors cite many previous studies, but there should be more information why your study is needed: What is still unknown? What are methodological drawbacks of previous studies/approaches that limit their interpretation? Please add within the manuscript - the second study aim is to dectect “whether applying more SDM influences the consultation duration, the number of options discussed, and the type of decisions made” – however, this first requires that SDM is adopted at all by the implementation program, right? So I think this may be rather a side question of the paper; the two main questions in may opinion are to investigate whether SDM can be adopted and which components contributed to the adoption Study population and selection - Please clearly indicate in- and exclusion criteria for the sites and the study population Ethical approval - Except for the last sentence, the content should be shifted to other sections, e.g., to 2.1. and 2.3 - Ist the study registered? If not, this should be discussed as a limitation Implementation program - “(5) Collaborative meetings: two collab-141 orative team meetings were organized (two fewer than in the previous trajectory).” Does that mean that meetings were reduced? Please clarify and justify this adaptation - “Patients were involved during all phases of the program, i.e., as part of the research 145 team, and in training and providing feedback on consultations.” This is an interesting issue – please clearly indicated at which stage they participated in which manner Data collection - Please better describe the logbook and which type of information is entered here - Please clarify which outcomes are gathered from the audiotapes. Please also clarify how you extracted the information from this qualitative data. - Please clarify (also in the study design section) whether the same patients and/or cliniciains are assessed before and after the implementation? Sample size calculation - “Therefore, a mean score of 39 before the intervention was used” – used as what? Please clarify - For which statistic do you need the 120 recordings? Why was this statistic chosen for sample size calculation and not one of the others? Do you need 120 at each time point? Is it important to have a equal number at each time point? - Are the required data points also sufficient for the other statistics? If not, this needs to be discussed in the limitations section - the order of the infos provided for the sample size calculation seems a bit unlogical – please re-structure 2.7 Data analysis - this section does not present information on data analysis and the term of the heading can hardly be differentiated from 2.8 (statistical analysis) - therefore, I suggest to shift this information to the data collection section and to remove the whole 2.7 section 2.8 statistical analysis - “After merging variables with ex-214 actly equal scores for each of the 5 teams, it concerned the following variables:” I did not get this sentence – please re-formulate - generally, the statistical analysis section is hard to follow: I strongly suggest to re-structure the whole section according to the research questions and to clearly present which analyses are applied for which research question and how this is done Results in general - Generally, the results section is hard to follow: the information should be clearly structured according to the research questions.

- Given the many subheadings and the vague presentation of the statistical analysis above, it is not possible to get the main message of the results - Correlate variables: is this important for one of the main research questions or just a sensitivity analysis? If the latter is true, it may shifted in the appendix

- Also, please present the number of included cases in any table to check the robustness of each statistic - I also suggest to re-consider the outcomes of the study: are all of these outcomes necessary? My recommendation is to streamline the outcomes and/or re-define the research questions to enhance readibility and to enable a better manuscript presentation Results, hospital teams - 5 out of 7 hospitals participated, but another two hospitals outside Amsterdam also agreed to participate? Please clarify - Ll. 235-242: this reads as if you have applied statistical tests or other comparisons– please simply state the distribution before and after the implementation without a direct comparison - I did not get why you report the info between lines 244 and 249 in this section – please clarify or remove/shift Results, patients - The info on recruitment needs to be shifted to the data collection section - I do not understand why options discussed are reported here; I would have expected this information below, e.g. under 3.2 Discussion: - The first pargaraph is too detailed – it should only briefly summarize two or three main findings of the study - By now, the discussion is hard to follow – again, please re-structure according to the research questions and clearly differentiate between (i) integration of findings into previous research and implications drawn from the current data; please also add a paragraph on strenghts of the study - Third paragraph: where was the superiority of the OPTION-5 instrument investigated in the current article? Please clarify; otherwise, the paragraph should be removed

- “A limitation of the regression analysis was that variables with exactly equal scores 382 for each of the 5 teams, had to be merged in the model.” I did not get this sentence; please clarify, also in the statistical analysis section - Limitations: why is it problematic that pre-implementation scores were relatively high? Please clarify - Limitations: the lack of a registration is of concern: please elaborate on this issue and how you addressed it

 

Author Response

Dear editor and reviewers,

Please find enclosed the revision of our manuscript "Effects and working mechanisms of a multilevel implementation program for applying shared decision-making while discussing systemic treatment in breast cancer". We appreciate very much that we have the opportunity to revise and submit our manuscript. We would like to thank you for the time and effort that you invested into commenting on our work. Your comments were relevant to us and helpful in improving the paper.

We are grateful that you acknowledge that the question how we can help clinicians to better adopt Shared Decision Making (SDM) is important and that our paper is generally well-written. In the table below, we respond to the comments and explain how we addressed them in our manuscript. Changes in the revised text are underlined.

Yours sincerely, on behalf of the authors,

Haske van Veenendaal

 

Comments reviewer

Revisions

1.     Abstract: - Generally, the abstract is well written - The outcomes “redistribution of tasks” and “discussing feedback from consultations” in the results are not mentioned in the methods before. Accordingly, the conclusions are hard to follow 2.     Introduction: It is stated that “Chemotherapy or hormone therapy improves survival, but this benefit only applies to a small proportion of patients” – please give a citation 3.     It is stated that SDM “is especially challenging for clinicians to achieve in decisions about systemic therapy with (older) breast cancer patients” – please report some reasons for this 4.     In the second last paragraph, some previous implementation studies are mentioned: are these from the same working group, i.e., is the current manuscript building upon a previous project? Please clarify - The authors cite many previous studies, but there should be more information why your study is needed: What is still unknown? What are methodological drawbacks of previous studies/approaches that limit their interpretation? 5.     Please add within the manuscript - the 2nd study aim is to detect “whether applying more SDM influences the consultation duration, the number of options discussed, and the type of decisions made” – however, this 1st requires that SDM is adopted at all by the program, right? So this may be a side question of the paper; the two main questions in may opinion are to investigate whether SDM can be adopted and which components contributed to the adoption 6.     Study population and selection - Please clearly indicate in- and exclusion criteria for the sites and the study population 7.     Ethical approval - Except for the last sentence, the content should be shifted to other sections, e.g., to 2.1. and 2.3 - Ist the study registered? If not, this should be discussed as a limitation Implementation program - “ 8.     Collaborative meetings: two collaborative team meetings were organized (two fewer than in the previous trajectory).” Does that mean that meetings were reduced? Please clarify this adaptation. 9.     “Patients were involved during all phases of the program, i.e., as part of the research team, and in training and providing feedback on consultations.” This is an interesting issue – please clearly indicated at which stage they participated in which manner 10.  Data collection - Please better describe the logbook and which type of information is entered here - Please clarify which outcomes are gathered from the audiotapes. Please also clarify how you extracted the information from this qualitative data. - Please clarify (also in the study design section) whether the same patients and/or clinicians are assessed before and after the implementation? 11.  Sample size calculation - “Therefore, a mean score of 39 before the intervention was used” – used as what? Please clarify - For which statistic do you need the 120 recordings? Why was this statistic chosen for sample size calculation and not one of the others? Do you need 120 at each time point? Is it important to have a equal number at each time point? - Are the required data points also sufficient for the other statistics? If not, this needs to be discussed in the limitations section - the order of the infos provided for the sample size calculation seems a bit unlogical 12.  please re-structure 2.7 Data analysis - this section does not present information on data analysis and the term of the heading can hardly be differentiated from 2.8 (statistical analysis) - therefore, I suggest to shift this information to the data collection section and to remove the whole 2.7 section 13.  2.8 statistical analysis - “After merging variables with exactly equal scores for each of the 5 teams, it concerned the following variables:” I did not get this sentence – please re-formulate - generally, the statistical analysis section is hard to follow: I strongly suggest to re-structure the whole section according to the research questions and to clearly present which analyses are applied for which research question and how this is done 14.  Results in general - Generally, the results section is hard to follow: the information should be clearly structured according to the research questions. Given the many subheadings and the vague presentation of the statistical analysis above, it is not possible to get the main message of the paragraph 15.  Correlate variables: is this important for one of the main research questions or just a sensitivity analysis? If the latter is true, it may shifted in the appendix 16.  Also, please present the number of included cases in any table to check the robustness of each statistic - I also suggest to re-consider the outcomes of the study: are all of these outcomes necessary? My recommendation is to streamline the outcomes and/or re-define the research questions to enhance readability and to enable a better manuscript presentation 17.  Results, hospital teams - 5 out of 7 hospitals participated, but another two hospitals outside Amsterdam also agreed to participate? Please clarify 18.  Ll. 235-242: this reads as if you have applied statistical tests or other comparisons– please simply state the distribution before and after the implementation without a direct comparison 19.  I did not get why you report the info between lines 244 and 249 in this section – please clarify or remove/shift Results, patients 20.  The info on recruitment needs to be shifted to the data collection section. 21.  I do not understand why options discussed are reported here; I would have expected this information below, e.g. under 3.2 22.  Discussion: - The first paragraph is too detailed – it should only briefly summarize two or three main findings of the study - By now, the discussion is hard to follow – again, please re-structure according to the research questions and clearly differentiate between (i) integration of findings into previous research and implications drawn from the current data; please also add a paragraph on strenghts of the study - Third paragraph: where was the superiority of the OPTION-5 instrument investigated in the current article? Please clarify; otherwise, the paragraph should be removed 23.  “A limitation of the regression analysis was that variables with exactly equal scores for each of the 5 teams, had to be merged in the model.” I did not get this sentence; please clarify, also in the statistical analysis section 24.  Limitations: why is it problematic that pre-implementation scores were relatively high? Please clarify 25.  Limitations: the lack of a registration is of concern: please elaborate on this issue and how you addressed it

1.  The “redistribution of tasks” and “discussing feedback from consultations” are program components introduced under the ‘Results’ in the abstract. We added: ‘regression analysis showed that’ for better clarification’. 2. We added:  Effects of chemotherapy and hormonal therapy for early breast cancer on recurrence and 15-year survival: an overview of the randomised trials Lancet, 365 (2005), pp. 1687-1717. 3. This has been added. 4. We added: ‘In a previous study on which this research builds,..’ and better clarified the need for this study. 5. We made the adjustments as suggested. 6. We made the adjustments as suggested. 7.  We made the adjustments as suggested. 8.  We made the adjustments as suggested. 9. We specified the patient involvement. 10.  We have incorporated all suggestions, with some of the information from the removed paragraph 2.7 (see comment 12). 11. This is necessary to determine the adequate inclusion numbers in a pre/post measurement (power calculation for a paired t-test). We clarified this in this section and under limitations. 12. This paragraph is integrated in paragraph 2.5 See comment 10. 13.  We modified our sentence and adapted the structure to the order in the research questions: 1. SDM Adoption (OPTION-5), 2. Patient Perceptions (SDMQ-9), 3 Working Mechanisms (Logbook/Regression) and 4 Key Consultation Characteristics (from audio recordings). 14. Also the result section is restructured to the order in the research questions. 15.  The correlated variables describe the main variables that influence the increase in the primary outcome measure. This shows, among other things, that participation in the implementation has by far the greatest effect. We believe this is important information for the reader and prefer to include it in the main text. 16. The numbers have been added to the table. 17. This has been clarified. 18.  We have adjusted this as suggested. 19.  We report this to show whether the group of participants - which should preferably be unchanged between the pre- and post-test - differs between before and after the implementation. We clarified the presentation of this data and shortened this passage. 20. This information has been shifted. 21. This information has been shifted. 22. The discussion is restructured in line with the research questions and strengths were added and the part about the OPTION-5 has been removed. 23. This has been clarified in both sections. 24. This has been clarified. 25. Since we did not find the passage 'lack of registration', we assumed it was the passage 'lack of an obligation'? We added to address this.

Reviewer 2 Report

The manuscript describes an investigation of an adapted multilevel shared decision making (SDM) implementation program in breast cancer care. This is an important topic and will be of interest to many readers, including those working in other therapeutic areas. The experimental design is single-arm non-randomised pre vs post test, which is known to be prone to bias. Nevertheless, these results provide sufficiently robust information to design a randomised trial.

There are a few points that need to be addressed:

Introduction: The final paragraph describes well the aims of the study, but does not entirely align with the final sentence of the Abstract or with the Conclusions. For example, the aim "helps clinicians to adopt SDM" (assessed through the primary outcome measure the OPTION-5 instrument) were clinically relevant (>10 points) and significantly greater after implementation.

Data collection: The instrument is not a Likert scale, and is probably better described as an ordinal numeric rating scale.

Table 2: It would be helpful to indicate the direction of the differences in the final column; for example, the change in OPTION-5 scores was +20.4, whereas the change is SDM-Q-9 scores was -2.2

Discussion: The authors might consider adding some discussion on the use of their SDM approach during remote consultations, as these are frequently done in this post-pandemic era.

Author Response

Dear editor and reviewers,

Please find enclosed the revision of our manuscript "Effects and working mechanisms of a multilevel implementation program for applying shared decision-making while discussing systemic treatment in breast cancer". We appreciate very much that we have the opportunity to revise and submit our manuscript. We would like to thank you for the time and effort that you invested into commenting on our work. Your comments were relevant to us and helpful in improving the paper.

We are grateful that you acknowledge that the question how we can help clinicians to better adopt Shared Decision Making (SDM) is important and that our paper is generally well-written. In the table below, we respond to the comments and explain how we addressed them in our manuscript. Changes in the revised text are underlined.

Yours sincerely, on behalf of the authors,

Haske van Veenendaal

Comment Reviewer	Revision
Introduction: The final paragraph describes well the aims of the study, but does not entirely align with the final sentence of the Abstract or with the Conclusions. For example, the aim "helps clinicians to adopt SDM" (assessed through the primary outcome measure the OPTION-5 instrument) were clinically relevant (>10 points) and significantly greater after implementation.	We added a sentence in the abstract and we changed the sentence in our conclusion, mentioning the successful adoption and clinical relevance.
Data collection: The instrument is not a Likert scale, and is probably better described as an ordinal numeric rating scale.	We reread the original article (ref. 39) and the article describing the translation into Dutch of the instrument (ref. 40). In the latter article, the authors refer to the scales of the SDMQ-9 as Likert scales. However, to avoid any confusion, we have deleted 'Likert'
Table 2: It would be helpful to indicate the direction of the differences in the final column; for example, the change in OPTION-5 scores was +20.4, whereas the change is SDM-Q-9 scores was -2.2	We changed this in table2.
Discussion: The authors might consider adding some discussion on the use of their SDM approach during remote consultations, as these are frequently done in this post-pandemic era.	We added this to the discussion.

Reviewer 3 Report

Article show important decision making for patients with breast cancer. 

1.The research treats an important topic about  multiple shared decision in breast cancer patients. 2.The topic is very original and relevant for every day  clinical practice considddering that the participation of the patient in  therapeutical decision may improve patient compliance and oncologic outcome. 
3. They particularised this new tumor decission in patients with breast cancer and they follow important clinical and  medical resource  features like time to patients consultation, no of consultation, reaching a decision in breast cancer patients. This may increase  the accesibility of the patients to the doctor and minimaising waiting time for the patient.  
4. Correct statistical methods were used 
5. The conclusions  are consistent with the evidence and arguments presented and they answear to the main question rise by the study. 
6. The references are appropriate and  includes reccent articles from 2021-2022
7. There are no figures in the in the manuscript and the table are complete and relevant for the research.

Author Response

Dear editor and reviewers,

Please find enclosed the revision of our manuscript "Effects and working mechanisms of a multilevel implementation program for applying shared decision-making while discussing systemic treatment in breast cancer". We appreciate very much that we have the opportunity to revise and submit our manuscript. We would like to thank you for the time and effort that you invested into commenting on our work. Your comments were relevant to us and helpful in improving the paper.

We are grateful that you acknowledge that the question how we can help clinicians to better adopt Shared Decision Making (SDM) is important and that our paper is generally well-written. In the table below, we respond to the comments and explain how we addressed them in our manuscript. Changes in the revised text are underlined.

Yours sincerely, on behalf of the authors,

Haske van Veenendaal

 

Comments Reviewer #2

Revision

1.     The research treats an important topic about  multiple shared decision in breast cancer patients. 2.The topic is very original and relevant for every day  clinical practice considering that the participation of the patient in  therapeutical decision may improve patient compliance and oncologic outcome. 2.     They particularized this new tumor decision in patients with breast cancer and they follow important clinical and  medical resource  features like time to patients consultation, no of consultation, reaching a decision in breast cancer patients. This may increase  the accesibility of the patients to the doctor and minimalizing waiting time for the patient.  3.     Correct statistical methods were used 4.     The conclusions  are consistent with the evidence and arguments presented and they answer to the main question rise by the study. 5.     The references are appropriate and  includes recent articles from 2021-2022 6.     There are no figures in the in the manuscript and the table are complete and relevant for the research.

Thank you very much for the careful review of the manuscript. As we've interpreted the comments, no adjustments are needed.