Manuscript Writing Dos and Donts

BRUCE FIRST DRAFT

Do’s and don’t of manuscript writing. Manuscript writing is closely tied to trial design, which of course can’t be changed by the time you are writing the manuscript. Here are some pointers.

1. POPULATION STUDIED. MolDx puts high priority on the population studied and if it is a realistic and natural fit to a future clinical population. Depending some on the context (the relevance of age), MolDx will look closely to whether a subpopulation > 65 is adequate, and they may ask questions about that >65 group even if it is a separate, non published analysis.
   1. INDICATION FOR USE. This should be well defined, although it is not so much part of the paper as part of the coverage policy. But the population (and outcomes chosen, if measured) should track to the eventual coveage INDICATION FOR USE.
2. PROSPECTIVE/RETROSPECTIVE STUDiES. There is a long history of retrospective results not necessarily being confirmed in prospective studies, with the result that retrospective (biobank) studies are usually looked at very cautiously and closely. Size of effect helps here (a very large retrospective result increases confidence that it would be durable in a prospective study).
3. REPRODUCIBILITY. THis is one of their highest concerns. Is this a lucky population? Is it a multi center population? Has something similar been done with similar results? Is this paper itself a confirmation (good).
4. ANALYTICAL VALIDITY. While it may involve more the tech assessment spreadsheets they use, MolDx will be very, very interested in details of lab protocols, test-retest accuracy, lifetime of reagents, specimen handling, and other technical aspects. This is usually a level of detail greater than what is publishable and is requested by template spreadsheets that MolDx provides. Often there are several rounds of questions involving work at the lab bench level.
5. COVERAGE DECISIONS. Think of coverage decisions in 3 buckets.
   1. Bucket 1 is standard of care. Probably there is no LCD. And they say, “If you have to ask if it’s standard of care, it isn’t.” For example, testing ER, PR, and Her2 in breast cancer is standard of care.
   2. Bucket 2 is replacing an existing test with the same function in a well accepted pathway. Famously, MolDx accepts MRD tests that are more accuarate or faster than CT/MRI surveillance, in situations where CT/MRI is recommended. Note that not all oncologists and payers agree how much value this is.
   3. Bucket 3 is something “novel” where the clinical use and physician behavior and outcomes (improved outcomes) are not obvious. These require “clinical utility” data, meaning at a minimum the test is used in the process of patient care and decisions.
6. OP EDS ETC. I am not sure how much this is true of MolDx itself, but payers may be influenced by Op Eds accompanying a paper, which help give the “temperature” of the thought leader community.
7. CLEAN WRITING. It goes without saying that the best science writing is both straightforward and easy to read. Sentences not too long, good use of subheadings, logical progression, major issues not skipped over or hidden. (Even if you get past the editor, MolDx may be upset by the claim or omission). Don’t go all fuzzy in the necessary “Limiitations” section. Don’t foret old fashioned skills like opening with good topic sentences for each paragraph.
8. A PATIENT CENTRIC AND PHYSICIAN CENTRIC FOCUS should emerge. Ask yourself, what would a patient or physician want to know (assuming patient could read it.) What would they want to know? This is sort of a crosscheck that may affect and improve a paragraph here and there. This also helps a bit (like good study design) with the concerns about industry-funded clinical science.
9. 
   

REWRITE BY CLAUDE AI

Do's and Don'ts of Manuscript Writing

Manuscript writing is closely tied to trial design, which of course can't be changed by the time you are writing the manuscript. Here are some pointers:

Population Studied

MolDx puts high priority on the population studied and whether it is a realistic and natural fit for a future clinical population. Depending on the context (the relevance of age), MolDx will look closely at whether a subpopulation >65 years old is adequate, and they may ask questions about that >65 group even if it is a separate, unpublished analysis.

Indication for Use

This should be well defined, although it is not so much part of the paper as part of the coverage policy. But the population (and outcomes chosen, if measured) should relate to the eventual coverage INDICATION FOR USE.

Prospective vs. Retrospective Studies

There is a long history of retrospective results not necessarily being confirmed in prospective studies, resulting in retrospective (biobank) studies usually being looked at very cautiously and closely. The size of the effect helps here (a very large retrospective result increases confidence that it would be durable in a prospective study).

Reproducibility

This is one of MolDx's highest concerns. Is this a lucky population? Is it a multi-center population? Has something similar been done with similar results? Is this paper itself a confirmation (good)?

Analytical Validity

While it may involve more of the tech assessment spreadsheets they use, MolDx will be very interested in details of lab protocols, test-retest accuracy, lifetime of reagents, specimen handling, and other technical aspects. This level of detail is usually greater than what is publishable and is requested via template spreadsheets that MolDx provides. Often there are several rounds of questions involving work at the lab bench level.

Coverage Decisions

Think of coverage decisions in 3 buckets:

1. Bucket 1 is standard of care. If you have to ask if it’s standard of care, it isn’t. For example, testing ER, PR, and Her2 in breast cancer is standard of care.
2. Bucket 2 is replacing an existing test with the same function in a well-accepted pathway. MolDx accepts MRD tests that are more accurate or faster than CT/MRI surveillance, in situations where CT/MRI is recommended. Note that not all oncologists and payers agree on the value here.
3. Bucket 3 is something "novel" where the clinical use, physician behavior, and outcomes (improved outcomes) are not obvious. These require “clinical utility” data, meaning at a minimum the test is used in patient care and decisions.

Op Eds, etc.

I'm not sure how much this is true of MolDx itself, but payers may be influenced by Op Eds accompanying a paper, which help indicate the “temperature” of the thought leader community.

Clean Writing

It goes without saying that the best science writing is both straightforward and easy to read - sentences not too long, good use of subheadings, logical progression, major issues not skipped or hidden. Even if you get past the editor, MolDx may be upset by a claim or omission. Don’t go fuzzy in the necessary “Limitations” section. Use old-fashioned skills like opening paragraphs with good topic sentences.

A Patient-Centric and Physician-Centric Focus

A patient or physician focus should emerge. Ask yourself what a patient or physician would want to know (assuming the patient could read it). This sort of cross-check may improve a paragraph here and there. It also helps a bit (like good study design) with concerns about industry-funded clinical science

CHAT GPT RE WRITE

Manuscript Writing Guidelines: Maximizing Impact and Clarity

Introduction to Manuscript Writing and Trial Design Manuscript writing is intrinsically linked to trial design. Although the trial design is set by the time of manuscript writing, understanding its nuances is crucial. Below are key pointers to enhance your manuscript's effectiveness. 1. Population Studied MolDx places significant emphasis on the population studied, evaluating its alignment with potential clinical applications. Age relevance plays a role here, especially in examining the adequacy of subpopulations over 65. Expect MolDx to inquire about this demographic, even in cases where it pertains to a separate, unpublished analysis. 2. Indication for Use Clearly defining the indication for use is essential, though it aligns more with coverage policy than the manuscript itself. Ensure the chosen population and outcomes (if measured) correspond to the indication for use in the coverage policy. 3. Prospective vs. Retrospective Studies Retrospective studies, especially those using biobank data, are scrutinized due to historical inconsistencies with prospective studies. A large effect size in retrospective studies may bolster confidence in their prospective applicability. 4. Reproducibility MolDx highly values reproducibility. Consider these aspects:

• Is the population representative or unique?
• Does the study involve multiple centers?
• Are there similar existing studies with comparable results?
• Does the manuscript confirm previous findings?

5. Analytical Validity MolDx delves deeply into lab protocols, test-retest accuracy, reagent lifespan, specimen handling, and other technical details. These often exceed the scope of typical publications and are addressed through detailed MolDx-provided templates, potentially leading to multiple rounds of lab-based inquiries. 6. Coverage Decisions: Three Tiers Tier 1: Standard of Care - Generally, no Local Coverage Determination (LCD) is involved. A key guideline here is, "If you have to ask if it’s standard of care, it isn’t." Tier 2: Replacement Tests - MolDx acknowledges tests that outperform existing ones in accepted pathways, like MRD tests superseding CT/MRI in certain scenarios. Tier 3: Novel Applications - These require clear demonstration of clinical utility and tangible improvements in patient care and outcomes.

7. Influence of Op-Eds While not directly related to MolDx, payer decisions can be swayed by Op-Eds accompanying a paper, as they reflect the perspectives of thought leaders. 8. Clarity and Cleanliness in Writing Aim for straightforward, readable science writing. Utilize concise sentences, clear subheadings, logical progression, and transparency in addressing major issues. Maintain clarity and precision in the "Limitations" section and ensure each paragraph opens with a strong topic sentence. 9. Patient and Physician Centricity Your manuscript should clearly convey the implications and benefits for patients and physicians. Regularly ask, "What would patients and physicians need to know from this study?" This focus not only enhances readability but also addresses concerns about industry-funded clinical science. By adhering to these guidelines, your manuscript will not only meet MolDx's criteria but also stand out for its clarity, relevance, and impact.