{EditHealth.ai} templates are Microsoft Word documents, created one at a time per project, that incorporate overlays for the unique regulatory requirements of each:

• population
• modality
• mechanism
• form of administration
• phase or life cycle
• regulatory authority
• setting
• disease

They differ from a “starting point” template because {EditHealth.ai} templates are not reusable foundations, as they change with document type, study design, and life cycle (updates within a life cycle are reusable). The reusable base is not a document, it is a database called the Evidence Envelope.

Similar to “starting point” templates, they begin with a structured foundation designed to ensure consistency, compliance, and efficiency in document creation or workflow setup.

We issue an interrelated template set using a system of hundreds of bases and overlays. The bases and overlays are pre-reviewed by experts in the functional area. At the time your project starts up, we saturate the document with external information that fits your program information.

An advantage of the {EditHealth.ai} approach is that your templates interrelate closely and arrive with the latest external guidance and precedents integrated. You also pay by the project, not for a system or a seat fee.

A Microsoft Word sidebar is a panel that opens beside your document. Regulatory documents are written, reviewed, and tracked in Word, so an assistant in the sidebar works where the document already lives, with no copying text between windows. The sidebar also allows you to accelerate work the way you know, using tracked changes, comments, and superpowered grammar and style check.

Your specified {EditHealth.ai} template is for use with Word sidebars such as Claude for Word and Copilot for Word. Asthra AI is coming soon. Your {EditHealth.ai} template holds the settled information and the questions for your experts, and the sidebar helps you with the rest as you write.

A deterministic process gives the same output every time it receives the same input. A probabilistic process, such as text generation by a large language model, chooses among likely next words, so the same prompt can produce different wording from one run to the next.

Your one of a kind {EditHealth.ai} template is the deterministic option. Rules built into your {EditHealth.ai} template handle the settled parts, so information is collected and described the same way every time, in the structure the guidance requires.

Settled information for your program is maintained in one place, the Evidence Envelope, which accompanies your bespoke template. The Evidence Envelope comes prefilled with external information that matches your program.

As you complete each document, you settle more information, and the program-related updates compile in the Envelope. The Evidence Envelope becomes your single source of truth until your primary sources of truth mature, such as your prescribing information.

Settled information is information whose meaning no longer needs a decision. Settled information comes from definitions and approved facts across every function:

• Safety: what counts as an unexpected adverse reaction
• Terminology: standard terms from controlled vocabularies such as MedDRA^® and CDISC^®
• Company decisions: approved facts such as the drug’s dosage form
• Toxicology: the no-adverse-effect level and exposure from each final study report
• Clinical pharmacology: half-life and exposure values from final pharmacokinetic reports
• Manufacturing: strength, storage conditions, and specifications
• Clinical: the study population, dose levels, and schedule in the approved protocol
• Statistics: the analysis populations and endpoints in the approved statistical analysis plan
• Regulatory: the development international birth date and the reporting period it sets

Every item above is recorded with its source and its approval when settled.

Judgement is information whose meaning still needs a decision.

• Toxicologist: whether an animal finding matters for people and whether it reverses
• Clinical pharmacologist: the starting dose, the exposure margin behind it, and the risk of drug interactions
• Medical lead: which risks investigators should watch for and which participants should not take part
• Safety physician: whether a new signal changes the balance of benefit and risk
• Statistician: which analyses support a conclusion and how missing data are handled
• Manufacturing lead: what specifications and stability results are needed to support the clinical supply
• Regulatory lead: what to ask the FDA and what the program commits to
• Medical writer: which messages, benefits, and risks to emphasize, and how to present them

Only a human can provide a statement of judgement. The template provides context but no suggested wording. Each judgement is made by the named expert who owns it, with the evidence alongside.

Curating is the process of setting your own style and presentation.

Style and presentation are set by default to representative industry standards, but can be modified. Your organization will know its preferred style (for example, “e.g.” or “eg”), headings (for example, initial capitalization or sentence capitalization), or arrangement.

No statement is fixed. You can rearrange and even lightly edit deterministic statements that are connected to primary sources.

Your team authors “judgement” statements on their own, with context and light LLM editing. Summaries tied to those judgements do not require tracking. Your judgements are summarized in the same words unless you override and choose different wording.

You can highlight a section and switch among text, bulleted, diagram, table, or cross reference to an appendix. You arrange the order of presentation to highlight, for example, key secondary endpoints. You are not limited to presets.

With less copy and paste, medical writers put their effort into their expertise... curation.

Your precision {EditHealth.ai} template is built to specification, with settled information prefilled. It leaves room for curation and for the judgements only your team can make.

Each specification comes with an analysis of regulatory precedents, recent studies, and key literature for similar compound classes, treated populations, and more.

Precedent supports your judgements as you select.

The guidelines that apply to your document are analyzed with each specification, so you can hone in on your requirements.

Information remains flagged, and the document cannot be finalized, until a single responsible party decides that unsettled information is settled and that the document checklist of predetermined requirements is complete.

Your functional reviewers are called to action, with requests integrated into their messaging of choice.

A master database in your regulatory system of record, the Evidence Envelope, holds settled information.

The Evidence Envelope distributes the single source of truth across your documents as “snapshots”.

Your one of a kind {EditHealth.ai} template is built to match your named workflow conventions, but every {EditHealth.ai} template has nine stages in common:

• configure
• background and rationale
• collection of resources
• methods
• results
• evaluation
• summary
• conclusions
• format

These common names allow you to match your report preparation workflow with workflows from Veeva Vault, OpenText Documentum, Microsoft SharePoint, and project management tools.

Your bespoke {EditHealth.ai} template provides significant resources, but not everyone has a statistician or toxicologist on hand. The {EditHealth.ai} template will suggest a resource to review the “Judged information” so that you are confident with your submission.

Specialist SMEs were consulted for the base {EditHealth.ai} templates and they are available for additional, focused review of your work.

Integrated human SMEs, also called insight resources, are individuals who are prepared to review sections of your report on demand. Example sections of your {EditHealth.ai} template where they are available:

• Statistics: the protocol’s statistical considerations, including the estimand and how missing data are handled; the clinical study report’s statistical methods
• Toxicology: the investigator’s brochure’s nonclinical toxicology summary; the protocol’s nonclinical findings behind the dose rationale
• Pharmacovigilance: the DSUR’s overall safety assessment and summary of important risks; the investigator’s brochure’s reference safety information
• Clinical research: the protocol’s eligibility criteria and schedule of assessments; the clinical study report’s description of study participants