Medical Records and Your Health Profile

This is a long page. Use the table of contents to jump around, or skim the section headers. Every question has a full answer directly below it.

Important Notes Before You Connect

• New medical record connections currently need to be started from desktop web. You cannot initiate a new EHR connection from mobile web or the iOS app today.
• After records are connected and processed, your Health Profile can still help you across the app. The desktop limitation is about starting the connection.
• Connecting medical records is a Premium feature. The Health Profile itself is still available on the free plan through conversation, manual entry, uploads, and photos.
• Not every provider or lab company is supported as a direct connection. LabCorp is one known example. Many members can still see LabCorp results if the test was ordered by a clinician whose clinic or health system shares results through a supported portal, such as Epic or another traditional health system portal.

Table of Contents

1. Important Notes Before You Connect
2. The Basics
3. About Fasten Health
4. Privacy, Security, and HIPAA
5. The Setup Wizard
6. Connecting Your Records
7. Processing Time and the Two Stages
8. Your Health Profile at a Glance
9. The Medications Section
10. The Conditions Section
11. Diagnostic Certainty
12. Data Provenance and Source Tracking
13. Other Profile Sections
14. The Rebuild from Chats Button
15. Refreshing Your Records
16. Connecting Multiple Providers
17. Reviewing and Correcting Your Profile
18. Disconnecting a Provider
19. Troubleshooting
20. Availability, Plans, and Support

The Basics

What is the Medical Records feature?

My Doctor Friend can connect directly to your hospital or clinic's patient portal and pull in your complete medical record. Instead of typing every condition, medication, allergy, and lab result by hand, you sign in once to your patient portal, and your records flow into your Health Profile automatically. From that moment on, the AI has context about you when you chat: your actual diagnoses, your real medication list, your lab trends, your visit notes. It can answer questions about your own history, not just general medical knowledge.

What's the difference between "records" and the "Health Profile"?

Think of records as the raw file cabinet: everything your clinicians wrote down, every lab report, every visit summary. The Health Profile is the organized, editable summary built from those records plus anything you've told us. Records are what your hospital has on you. The Health Profile is your personal, structured view, one that you can correct.

Why bother connecting records instead of just telling the AI what I have?

Two reasons. First, your actual records are often more complete than what you can remember: old lab values, visit notes from years ago, imaging reports, immunization dates, specialist impressions. Second, imported records carry weight: when the AI sees a condition confirmed by your EHR, it treats it as established rather than a possibility. That matters for the quality of guidance you get back.

Is connecting records required?

No. You can use My Doctor Friend perfectly well without connecting any records. You can tell the AI your history in conversation, upload photos of lab reports or prescriptions, and type information directly into your Health Profile. Connecting records is a shortcut and an accuracy boost, not a requirement.

Is this feature free?

Connecting medical records is a Premium feature. If you are on the free plan, you can still build a Health Profile by chatting, by entering information manually, and by uploading documents and photos.

What kinds of questions can I ask after my records are connected?

A few examples:

• "What did my doctor say at my last cardiology visit?"
• "How has my A1C trended over the past two years?"
• "Have I ever had a positive strep test?"
• "What does this new result mean given my history?"
• "Am I due for any vaccines or screenings?"
• "Which of my current medications could be causing this side effect?"

The AI can reference your actual notes, labs, and imaging reports when it answers. When it does, it tells you where the information came from.

About Fasten Health

What is Fasten Health?

Fasten Health is the company we partner with to securely connect your patient portal to My Doctor Friend. They specialize in one thing: being a secure bridge between consumer health apps and thousands of hospitals, clinics, and health systems across the United States. Fasten speaks the technical language of EHRs (Epic, Cerner, Athena, Allscripts, and many more), handles the authentication flow, and delivers your records in a standardized format.

Practically, when you start a connection from desktop web, a small widget from Fasten opens inside My Doctor Friend. You search for your health system, log in with your patient portal credentials, and authorize the connection. Fasten handles the secure handshake with your hospital; we handle everything else.

Why Fasten? Why not connect directly?

There are thousands of patient portals in the U.S., each with its own login flow, its own authorization rules, and its own quirks. Building direct connections to every one of them would take years, and maintaining those connections as hospitals change systems would be a full-time engineering effort. Fasten has already done that work. Partnering with them means we can support most U.S. patient portals today rather than a handful a year from now.

Fasten is also a specialized security vendor. They are HIPAA-compliant, they hold the relevant legal agreements with health systems, and their entire business is built around safely handling this exact problem. We trust them with the connection so we can focus on what comes next: turning records into a useful Health Profile.

Does Fasten see my health data?

Fasten is the pipe that carries your records from your hospital to us. Because of that, your data passes through their systems in transit. They are HIPAA-compliant and bound by the same medical privacy rules as your hospital and as us. They do not sell your data, share it with advertisers, or use it for anything other than delivering it to My Doctor Friend on your instructions.

Does Fasten see my patient portal password?

Your portal password travels from you, through Fasten's secure connection, directly to your health system. Fasten does not store it, and neither do we. The authorization that results from the login is what gets saved, and that authorization can be revoked at any time.

Which hospitals and clinics does Fasten support?

Fasten supports thousands of U.S. health systems, including all the major networks (Kaiser, Sutter, Mayo Clinic, Cleveland Clinic, Mass General Brigham, NYU Langone, UCSF, Providence, Intermountain, Geisinger, and many more), along with regional hospitals, clinic networks, and smaller independent practices. If your hospital has a patient portal (MyChart, Cerner HealthELife, FollowMyHealth, Athenahealth, or a custom branded portal), there is a good chance it is supported.

What if my provider isn't listed?

If your health system does not show up in the Fasten search, there are a few things you can do:

• Search for the parent health system instead of the specific clinic. A local clinic is often owned by a larger network (for example, your doctor's office might be part of Sutter Health).
• Try the generic name. Some systems are listed under branded names ("MyChart") rather than hospital names.
• Know that some providers and lab companies are not supported as direct connections today. LabCorp is one known example.
• If your lab work was ordered by a doctor or clinic that uses a supported portal, the result may still come through that health system even if the lab company itself is not available as a direct connection.
• Skip the connection for now and enter your information manually. You can always come back to this.
• Reach out to us. We can flag the gap with Fasten for future support, and in some cases we can help you work around it.

What file format do the records come in?

The records are delivered in a standardized healthcare format called FHIR (Fast Healthcare Interoperability Resources). You do not need to know what that means as a user. It is how health systems share structured data with apps like ours. FHIR covers conditions, medications, allergies, labs, imaging reports, immunizations, clinical notes, and visit summaries.

Privacy, Security, and HIPAA

Is My Doctor Friend HIPAA compliant?

Yes. The Medical Records feature is built on HIPAA-compliant infrastructure. Our partner Fasten Health is HIPAA-compliant. Your records are encrypted in transit and at rest. Access is controlled so that only you can see your own data. We maintain the technical and administrative safeguards required for handling Protected Health Information (PHI).

Where does my patient portal password go?

Straight to your hospital or clinic, through Fasten's secure connection. It does not touch our servers. We never see it, store it, or log it. You could search our systems high and low and you would not find it.

How is my data encrypted?

In transit: every connection between your phone, our servers, Fasten's servers, and your health system uses HTTPS with modern TLS. That is the same encryption used by online banking.

At rest: your data lives in encrypted databases and in encrypted cloud storage. If someone physically took the hard drives our data sits on, they would not be able to read anything.

Who can see my health records?

You. That is the short answer. Your records are scoped to your account, and the only way to access them is through your login. Our team does not browse member records for any reason. If we ever need to debug a technical issue on an account, we follow an internal process that minimizes exposure and requires your permission or a strong, documented operational reason.

Are my records ever sold or shared?

No, not to insurance companies, not to pharmaceutical companies, not to advertisers, not to anyone. We do not run an ad-funded business. Your records are used to build your Health Profile, to power your conversations with My Doctor Friend, and for nothing else.

What happens to my records if I disconnect a provider?

Disconnecting immediately revokes the connection so that no further data flows from that provider. The records that were already imported remain inside your Health Profile, where you can still edit or delete them. If you want everything gone, you can also ask to have the raw imported data purged, and we will delete the underlying files and processed records on our side.

What happens to my data if I delete my account?

When you delete your account, we delete your personal health data. This includes your Health Profile, your raw EHR imports, your uploaded documents, your conversations, and your indexed search data. Some minimal records (such as billing history) may be retained for accounting and legal purposes, but nothing that identifies your health condition.

Do you have a Business Associate Agreement with Fasten?

Yes. Any vendor that handles PHI on our behalf has the appropriate legal agreements in place. Fasten's entire business is structured around being a HIPAA-compliant Business Associate to consumer health apps.

Do you keep audit logs?

Yes. Every step of your record import, every refresh, every connection change, and every access is logged internally. That lets us investigate issues, verify that processing completed correctly, and satisfy the audit-trail requirements of HIPAA.

Can I export my data?

You can view your Health Profile in the app, and your records are always accessible to you. If you need a full data export for portability reasons, reach out to support and we will help you get it.

Why can I trust this?

A fair question. Our answer is not "take our word for it," it is a combination of things: our choice of vendor (a specialist with the right agreements), our technical architecture (encryption end to end, isolated per-user data), our policies (no ads, no data sales), and the fact that we are a small team that handles your data with the seriousness it deserves. If something feels off, tell us, and we will answer you honestly.

The Setup Wizard

What is the setup wizard?

In this FAQ, "setup wizard" means the short first-time onboarding flow that appears when a new profile is created. Its job is to gather enough medical context to make My Doctor Friend useful right away: basic demographics, current health conditions, and what you want help with.

This is separate from the Health Profile Setup Wizard on the Health Profile page. That later wizard is a profile-editing flow for filling in or reviewing more Health Profile sections, such as medications, allergies, family history, and lifestyle.

Is the wizard mandatory?

No. Each step has a Skip option. You can skip the whole wizard if you want to jump right into a conversation, and you can always come back and fill in the details later from your Health Profile page.

What are the steps?

The first-time profile setup wizard has three short steps, plus one optional Apple Health step on iPhone for primary profiles:

1. Apple Health, iOS only and optional. On iPhone, primary profiles may be asked whether My Doctor Friend can read from Apple Health. If you agree, it can pre-fill basics such as date of birth, biological sex, height, and weight when available. You can decline and skip this step.
2. Demographics. Date of birth or birth year, biological sex assigned at birth, gender identity if you want to share it, height, and weight.
3. Conditions. Any current health conditions you want My Doctor Friend to know about. You can select from grouped suggestions or add your own.
4. Goals. What brings you to My Doctor Friend, such as understanding symptoms, managing conditions, preparing for doctor visits, getting another perspective, or caregiving support.

Each step has a progress indicator so you can see where you are. There is a Back button if you want to change something, and a Skip All button if you want to finish immediately.

Does the wizard connect my medical records?

No. The first-time setup wizard does not connect EHR records and does not open Fasten. It is for information you enter or approve yourself.

You may see separate prompts to connect records if you are eligible for the Medical Records feature, but that is a different flow. Today, starting a new EHR connection requires desktop web.

Can I re-enter the wizard later?

You do not need to re-enter the exact first-time onboarding flow. From the Health Profile page, you can edit individual sections directly or use the Health Profile Setup Wizard to fill in more detail. That later wizard is re-entrant and covers a broader set of Health Profile sections.

Why does the wizard ask for date of birth, sex assigned at birth, height, and weight?

Those basics affect many ordinary health conversations: age-based screenings, medication context, growth or weight changes, and whether a lab value or symptom pattern means something different for you. You can skip what you do not want to share.

I'm setting this up for a parent, not for myself. Does the wizard change?

Yes. When you create a caregiving profile, for example for a parent or child, the wizard is framed from your perspective. The profile belongs to the person you are helping, and you can switch between profiles freely. Apple Health pre-fill is for primary profiles, because Apple Health data belongs to the phone owner.

Connecting Your Records

Where can I connect records from?

Today, new EHR connections need to be started from desktop web. On desktop, there are several places that can start or point you to the same Fasten connection:

• From the Health Profile page. Look for a Connect Records button near the top. If you already have a connection, the button is replaced by a strip of connected providers with their logos and sync dates.
• From Settings. Under Profile and EHR Management, each profile has an EHR status badge. Use this area to check, refresh, or manage existing connections. If you need to start a new connection, use desktop web.
• During profile creation. Eligible Premium members may see a separate records prompt around profile creation. This is not the same thing as the first-time setup wizard.

What does the connection flow look like?

Once you start the connection from desktop web, here is what happens, step by step:

1. A short informational screen appears explaining what is about to happen and who Fasten is.
2. You tap Connect, and the Fasten widget opens inside the desktop web app.
3. You search for your health system. Type the name (for example, "Kaiser Permanente") and pick the match.
4. You log in with your patient portal credentials. These are the same username and password you use for MyChart or your hospital's portal.
5. You authorize My Doctor Friend to receive your records.
6. The widget confirms the connection and closes.

From your perspective, that is the whole thing. It usually takes less than a minute. Processing the records themselves takes longer, see the next section.

Can I start the connection from iPhone, iPad, or mobile web?

Not today. You cannot initiate a new EHR connection from the iOS app, iPad, or mobile web right now. Start the connection from desktop web. After your records are connected and processed, you can use the resulting Health Profile context across the app.

How do I know it worked?

You will see a "Records connected!" confirmation screen right after the widget closes. That card shows two options: Start chatting (the AI has enough to help with general questions about your history within minutes) and I'll do this later (dismisses the card). Processing continues in the background either way.

Processing Time and the Two Stages

Why does processing take so long?

This is the question we get most often, so the answer here is detailed. The short version: after you connect, your records do not instantly become a Health Profile. We pull them, we read them carefully, and we build a structured summary that the AI can reason about. That second part is the time-consuming one, because we do it thoroughly rather than quickly.

The long version is that your records go through two stages.

Stage 1: Indexing and Ingestion

As soon as you finish connecting, we request your records from Fasten. Fasten in turn pulls them from your health system. This handoff takes a few minutes on average, sometimes longer for large health systems that respond slowly. We then download the raw records, store them securely, and begin to index them.

Indexing means we make every part of your record searchable. Your clinical notes, your encounter summaries, your imaging reports, and your visit discharge documents become individually retrievable. This is what lets the AI later say "According to your cardiology note from March 15, 2025..." rather than guessing from a summary. Indexing runs in parallel with Stage 2 and finishes in a few minutes to half an hour depending on volume.

During Stage 1 you can begin chatting: the AI has enough information from the indexed records to answer many questions, even before the structured profile is fully built.

Stage 2: Building the Structured Health Profile

This is the longer stage, and the reason the total wait can reach an hour or more. Here we read every part of your record carefully and assemble your Health Profile. The profile is organized into sections (conditions, medications, allergies, immunizations, family history, labs, imaging, clinical notes, encounters) and each section is built by a specialized AI routine that reads your records and extracts, normalizes, and deduplicates the relevant information.

You will see progress tick by in a banner at the top of your Health Profile page as each section is processed:

Each step is doing real work: reading source documents, cross-referencing codes, resolving ambiguity (for example, the same condition written two ways in two notes), deciding whether something is still active or has resolved, and writing clean entries.

How long does the full pipeline take?

Most profiles are ready in 60 to 90 minutes from the moment you connect. Profiles with a lot of history (many decades of records, many specialists, complex conditions) can take up to two hours. Profiles with a few encounters and basic labs can finish in under half an hour.

Why does it need to be that careful?

Because your Health Profile becomes the foundation for every conversation afterwards. If the profile says you are allergic to penicillin when you are not (or worse, misses that you are), the AI might make bad recommendations. If it says a medication is active when you stopped taking it years ago, the guidance you get will be wrong. We would rather take an hour and get it right than deliver something fast and sloppy.

The same logic applies to conditions: we want to distinguish between "this is a confirmed diagnosis in your chart" and "this was a transient note from a triage form years ago." That distinction requires careful reading.

Can I use the app while records are processing?

Yes. You can leave the Health Profile page, close the app, and come back later. Processing continues in the background on our servers, not on your phone. You can also chat with the AI during Stage 1, and the AI will tell you when your profile is still building.

How will I know when records are ready?

When processing completes, you will see a notification: "Your Records Are Ready!" From there, you can tap View Health Profile to see everything in one place, or keep chatting.

What if processing seems stuck?

Processing is supposed to progress through the steps. If you see the banner stuck on the same step for more than 20 minutes, something has probably gone wrong (a temporary service issue, an unusually large document, or an LLM timeout). In that case, we automatically mark the job as failed and offer you a Retry button. You can retry without re-connecting the provider.

What if processing fails?

If the system cannot complete processing even after retries, it will surface a clear error in the app. You can retry manually, or you can reach out to support and we will investigate on our side. In most cases, a failed run is recoverable by retry. If not, we may need to pull fresh data from your provider.

Is my data lost if processing fails?

No. The raw records are stored as soon as they arrive from Fasten. A failure only means the Health Profile did not finish assembling. The records are still safe, and a retry picks up where the previous attempt left off.

Why do I sometimes see a progress bar that pauses, then restarts?

Processing is cloud-based, and our systems sometimes coordinate by handing a job off from one worker to another. When that happens, the progress indicator can look like it reset briefly. It is still the same job, still making progress. If it truly stalls for 20+ minutes, that is when to retry.

Your Health Profile at a Glance

What is the Health Profile?

The Health Profile is the personalized, structured view of your health that My Doctor Friend maintains for you. It includes everything the AI knows about you: your demographics, your conditions, your medications, your allergies, your family history, your lab results, your visit summaries, your health priorities, and more. It is organized into sections so you can quickly see any part of your health at a glance.

How is the Health Profile different from my medical record?

Your medical record is what your hospital or clinic keeps. It is their record of your interactions with them, organized for billing, clinical decisions, and legal purposes. Your Health Profile is your own: organized for understanding, editable by you, and brought together across every provider you connect plus every conversation you have with the AI.

Can I edit the Health Profile?

Yes, completely. Every item in the profile is editable by you. You can correct, add, delete, mark as resolved, update dosages, adjust dates, and so on. Your edits are respected: when future imports happen, they do not overwrite the changes you made by hand.

How do I open it?

Tap the profile icon or navigate to Health Profile in the main navigation. On web the URL is /health-profile.

How do I know where each piece of data came from?

Every item in your Health Profile has a source badge. At a glance you can see whether something was imported from an EHR, entered by you, extracted from a conversation with the AI, or drawn from multiple sources. The section on Data Provenance goes into the details.

Is everything editable?

Yes. Even information imported from your EHR can be corrected, deleted, or overridden. Your Health Profile is yours.

The Medications Section

What appears in the Medications section?

Every medication we know about: current prescriptions, past prescriptions, and in some cases over-the-counter medications and supplements you have mentioned. Each entry contains the medication name and, where available, the dose, the frequency, the purpose, who prescribed it, the start date, and a status badge.

What are the medication statuses?

There are four statuses:

• Active. You are currently taking this medication. This is the default status for any prescription that has not been marked otherwise.
• Discontinued. You have stopped taking this medication. The entry is kept in the profile as historical context rather than a current prescription.
• On Hold. The medication is paused, not permanently stopped. You or your clinician intend to resume it. Examples: a medication paused before a procedure, or one you stopped temporarily to see how you feel without it.
• Completed. The medication was prescribed for a defined course (for example, a 10-day antibiotic or a 5-day steroid taper), and the course is finished. It is no longer needed because it was never meant to be ongoing.

The difference between Discontinued and Completed is a subtle one but useful: Completed means "the job is done," while Discontinued means "we stopped this one."

What is the difference between Discontinued and Completed?

Completed applies to medications with a planned endpoint, like a week of antibiotics. Discontinued applies to medications where continuation was possible but was stopped (side effects, lack of effect, a switch to a different drug, or a lifestyle change).

This distinction matters when the AI reasons about your history. If the AI sees a Completed amoxicillin course, it understands that to be a normal conclusion. If it sees a Discontinued lisinopril, it might ask why: side effects, switch to another ACE inhibitor, resolution of hypertension, or something else.

Can I add a medication manually?

Yes. Every section has an "Add" button. You can enter a medication you started outside of your connected providers (for example, a supplement you began on your own, or a prescription from a clinic that is not connected to your profile). Manually added medications carry the "You" source badge so you can see later that you entered them yourself.

What about over-the-counter medications and supplements?

The system supports them, and the AI will ask about them during conversations that are relevant (for example, any mention of pain, sleep, or supplements in general). They are not always imported from EHRs because they are not always recorded there. If you want the AI to factor in a daily vitamin or pain reliever, add it manually or mention it in conversation.

If I stop taking a medication, does the status update on its own?

Sometimes. If you mention to the AI that you have stopped something, it may propose an update to the Medications section in your "Suggested Changes" review. You then accept or reject the proposal, so nothing is changed behind your back. If you want to update the status yourself, you can do it directly by tapping the medication and changing the status dropdown.

Can I see who prescribed a medication?

If the data is present in your EHR, yes. Prescriber name and specialty often come through in medication records. If they are missing, it is because the underlying EHR record did not include them (some systems omit this information to simplify their data feeds).

Can I see why a medication was prescribed?

Sometimes, yes. If the purpose is noted in your record, or if the AI can reliably link the medication to an active condition (for example, metformin to type 2 diabetes), the purpose field is populated. If not, the field stays empty rather than guessing wrong.

What about dosing changes over time?

If your dose has changed (for example, atorvastatin increased from 20mg to 40mg), the current dose reflects the most recent prescription. Prior doses are reflected in the source notes and visit summaries that the AI can search when it needs to answer a question like "has my dose changed recently?"

How do dates work for medications?

• Start date is when the prescription first appeared in your records (or when you added it yourself).
• End date is when you (or your clinician) stopped it, if it was stopped.
• Duration applies to planned courses (for example, "14 days" for an antibiotic).

Dates are best-effort: if your EHR does not record a start date, we cannot invent one. In that case the field is left blank.

What happens when I delete a medication?

Deleting moves the entry to a dismissed list. The AI will not bring it back on its own, even if it appears in future EHR imports. If you deleted it by mistake, you can restore it.

The Conditions Section

This is the most detailed section of the Health Profile, and the one most worth understanding carefully. Conditions drive how the AI reasons about risk, tests, and recommendations, so the system tracks conditions along several independent axes: clinical status, diagnostic certainty, chronicity, body system, and provenance.

What is a condition?

Any diagnosis, health problem, or tracked concern that is part of your medical picture. Examples: hypertension, type 2 diabetes, hypothyroidism, migraine, GERD, depression, chronic kidney disease stage 3, osteoarthritis of the right knee. Each condition in your profile is an entry with its own metadata.

What information is tracked for each condition?

At minimum, the name. In practice, most entries also have:

• Details (a short description of the condition in your context)
• Severity (mild, moderate, or severe, when meaningful)
• Clinical status (Active, Managed, Monitoring, or Resolved)
• Diagnostic certainty (Verified, Documented, Self-Reported, or Unconfirmed)
• Chronicity (acute, chronic, episodic, or unknown)
• Body system (for grouping, e.g., Cardiovascular, Endocrine)
• Diagnosed date and resolved date (where known)
• Source badges (EHR, You, AI, Mixed)
• Supporting evidence icons (labs, imaging, prescription, specialist, and so on)

What do the four clinical status values mean?

The Clinical Status axis tells you where the condition stands today in terms of what it needs from you or your care team. The four values are intentional and distinct. Here is how to read each one.

Active

"Active" means the condition is currently affecting you in a way that requires attention: you are symptomatic, you are being worked up for it, or it is not yet under control. It is the default state for a newly identified condition that has not been clearly stabilized or resolved.

Examples:

• Newly diagnosed hypertension, before medication has brought it into range.
• Acute low back pain currently limiting your activity.
• A kidney stone being actively passed.
• Depression being worked up and not yet stably treated.

The AI treats Active conditions with priority when it thinks about your care. They tend to dominate current recommendations.

Managed

"Managed" means the condition is real and present, but it is stable and under control through ongoing treatment or lifestyle. It is not gone, but it is not currently causing problems because you are doing the things that keep it quiet.

Examples:

• Type 2 diabetes with A1C in target range on metformin.
• Hypertension with blood pressure in range on lisinopril.
• Asthma with no current symptoms on a controller inhaler.
• Hypothyroidism well-replaced on levothyroxine with normal TSH.

Managed conditions still need the AI's awareness (they affect lab interpretation, drug interaction checks, and long-term risk), but they do not dominate current conversations unless something changes.

Monitoring

"Monitoring" means the condition is being watched without active treatment. It is not clearly active, but it is not gone either, and someone (you, your clinician, or both) is tracking it over time to see what happens.

Examples:

• A small pulmonary nodule found incidentally on a chest CT, being followed with periodic imaging.
• Prediabetes (A1C 5.8 to 6.4) without medication, being rechecked periodically.
• Elevated liver enzymes of unclear cause, being tracked.
• Slightly enlarged thyroid noted on exam, being watched for change.

Monitoring is the state in between "confirmed and treated" and "ruled out." The AI factors it in when you bring up related concerns (for example, a new cough would reasonably prompt a question about that lung nodule).

Resolved

"Resolved" means the condition is no longer present. It was cured, it passed on its own, or it has been stable and inactive long enough that it is no longer clinically relevant. You can optionally add resolution notes explaining how it ended.

Examples:

• Strep throat from a year ago, treated and done.
• A kidney stone that passed.
• A skin cancer fully excised with clean margins.
• Postpartum depression that resolved after a defined period.

Resolved conditions stay in the profile because they still give the AI historical context (someone who had one kidney stone is at higher risk for another, someone who had melanoma needs surveillance), but they are not treated as active concerns.

How is the clinical status decided?

For imported conditions, we read the clinical status from your EHR if it is recorded, and we infer it from the surrounding context if it is not. For AI-extracted conditions (things picked up from conversation), we start with Active as the default and adjust based on subsequent evidence (for example, if you mention you stopped taking a medication because the condition resolved, the AI will propose moving it to Resolved).

For manually added conditions, you choose the status when you add it.

Can I change the status myself?

Yes, at any time. Tap the condition, open the detail view, and change the status. Your change takes effect immediately. When future imports happen, the system respects your choice and does not overwrite it.

What is severity?

Severity is a separate axis from status. It describes how intense the condition is when active:

• Mild means it is present but tolerable and not significantly impairing.
• Moderate means it has a meaningful impact on your life.
• Severe means it is substantially limiting, dangerous, or both.

Severity is useful for acute conditions (mild vs severe migraine, mild vs severe asthma attack) but less meaningful for some chronic ones (there is no "mild" hypothyroidism in a clinically useful sense).

What is chronicity?

Chronicity tells you how the condition behaves over time:

• Acute means sudden onset, short duration. A sprain. A cold. An ear infection.
• Chronic means long-standing and ongoing. Hypertension. Rheumatoid arthritis. Chronic kidney disease.
• Episodic means it comes and goes in episodes. Migraine. Gout. IBS flares.
• Unknown means we do not have enough information to classify it yet.

Chronicity and status are related but not the same: a chronic condition can be active (poorly controlled) or managed (well controlled), and an episodic condition can be in a quiet phase (no current symptoms, but the underlying tendency remains).

What is a body system?

Each condition is mapped to a body system so the profile can be organized, viewed by system, and summarized. The body systems are the standard clinical groupings: Cardiovascular, Endocrine, Musculoskeletal, Neurological, Respiratory, Gastrointestinal, Renal and Urinary, Dermatological, Psychiatric, Blood and Immune, Oncological, Ophthalmological, Reproductive, and so on.

You can view conditions grouped by body system, which is especially useful when you want to get a sense of what is going on in a particular area.

Can I see the source evidence for a condition?

Yes. Expanding a condition shows you the evidence behind it: which FHIR resources in your EHR support it, which clinical notes mention it, what medications you are on for it, what labs support it, and whether a specialist has been involved. This is the "why is this condition in my profile" view, and it is the best answer to the question "how confident should I be about this entry?"

Can I add a condition manually?

Yes. Tap Add in the Conditions section and fill in the fields. When you add a condition manually, it gets the "You" source badge and is treated as Self-Reported in terms of certainty (see the next section for what that means).

Can I delete a condition?

Yes. Deleted conditions are moved to a dismissed list, which prevents the AI from re-extracting the same condition from your conversation history and adding it back automatically. If you delete by mistake, you can restore.

For conditions imported from an EHR, we recommend editing the status (to Resolved, for example) rather than deleting. That preserves the historical record.

Diagnostic Certainty

Diagnostic Certainty is a separate axis from Clinical Status. Status tells you where a condition is now; Certainty tells you how confident we (or your records) are that the diagnosis is correct in the first place.

Why do you track certainty separately?

Because there is a real difference between "my records document coronary artery disease with angiography" and "I told the AI I think I might have sleep apnea." Both can appear in your Health Profile, but they deserve very different weight when the AI reasons about them. Tracking certainty explicitly lets the AI do that, and lets you see at a glance which entries are solid and which are still being evaluated.

What are the certainty badges?

There are four, in descending order of confidence:

• Verified. Objectively supported. The diagnosis is backed by the kind of evidence that settles the question: a definitive lab result, a positive biopsy, a confirmatory imaging study, or a formal diagnosis in your problem list backed by matching treatment and specialist care. You can trust this entry.
• Documented. Strong clinical picture, likely correct. The diagnosis is present in your records and the surrounding context is consistent (matching medication, matching specialist care, supportive labs), but there is no single smoking-gun piece of evidence that makes it airtight. Most common chronic diagnoses live here.
• Self-Reported. You told us. Something you mentioned in conversation or added to your profile manually, without corresponding documentation from an EHR. Not lower in truth value necessarily, just lower in the external verification sense. You might know your diagnosis perfectly well, but our records do not independently confirm it.
• Unconfirmed. Not yet established. Something the AI has noted as a possibility based on symptoms or context, but which is not strongly supported. The AI treats these as hypotheses worth investigating rather than established facts.

How is the certainty level decided?

For each condition, the system looks at the evidence available:

• Is the condition on your EHR problem list, with an ICD code?
• Are you on a medication that specifically treats it?
• Are there relevant labs, imaging, or procedure records that support it?
• Have you seen a specialist who focuses on this area of medicine?
• Did you manually confirm it?

The more of these are present, the higher the certainty. A condition that is on your problem list, has matching medication, and is followed by a relevant specialist lands at Verified. A condition you mentioned once in conversation, without any other signal, lands at Unconfirmed.

What is "Confirmed by you"?

If you manually confirm a condition (by editing it and selecting the Confirmed option), the entry is locked at its current certainty level. The AI will not downgrade a condition that you have explicitly confirmed. This is useful when you have a diagnosis that is well-established in your mind (and maybe in records outside of what we have imported) and you want the AI to treat it as settled without arguing about evidence.

"Confirmed by you" does not upgrade anything to Verified on its own: if the underlying evidence is thin, the certainty does not magically become strong. But it prevents the AI from second-guessing what you have said.

The AI changed the certainty on one of my conditions. What do I do?

If you disagree, change it back. Your edit locks the value from further automatic updates. If you want to be extra sure, use Confirmed by you.

If you are unsure whether the AI was right to downgrade (for example, it moved something from Documented to Self-Reported because it noticed the condition is not on your EHR problem list), consider whether adding documentation would help. You can upload a photo of a letter or report from your clinician, and the AI will factor it in.

Is "Unconfirmed" a bad thing?

Not at all. Unconfirmed is the honest label for something the AI is considering. Medicine often works in possibilities. The point of showing you an Unconfirmed condition is to let you decide: is this a possibility worth investigating, or is it a miss that should be dismissed?

What happens if I dismiss an Unconfirmed condition?

It is removed from your profile and added to the dismissal list. The AI will not re-introduce the same condition on its own based on future conversations, unless new and different evidence comes up.

Can certainty change over time?

Yes. As new records are imported, as you have more conversations, as you upload documents, the system may propose certainty changes. Those changes appear in the Suggested Changes review if they are substantive, or happen automatically in the case of clear upgrades (a condition that was Self-Reported becomes Verified once we see it in your imported problem list with treatment).

Downgrades never happen to conditions you have manually confirmed or edited.

Data Provenance and Source Tracking

What is data provenance?

Provenance means "where did this piece of information come from?" Every item in your Health Profile (every condition, every medication, every allergy, every lab, every note) has a provenance record. That record tells you, at a glance and in detail, what source contributed the item, and often exactly which document or conversation.

What are the main source badges?

There are four:

• EHR. Imported from your connected medical records. This is the highest-trust source for clinical data because it reflects what your care team actually wrote down.
• You. Entered, confirmed, or edited by you manually. Trust here is whatever you bring to it: you know yourself.
• AI. Extracted by the AI from a conversation you had with My Doctor Friend. Useful for capturing things you told the AI in passing (for example, "I get heartburn when I eat spicy food"), but always marked as AI-derived so you can tell it apart from verified clinical data.
• Mixed. More than one source contributed. For example, an allergy that was originally imported from your EHR but that you then corrected the reaction description on. Or a condition that was AI-extracted first and then corroborated by an EHR import.

Why does this matter?

Because the same fact can be true or false depending on where it came from. "Penicillin allergy" from an EHR entry written by a clinician who verified it with you is different from "penicillin allergy" inferred from a casual mention in one conversation. Provenance lets you see the difference and decide how much to trust each entry.

What about deeper source details?

For EHR-sourced items, you can often see:

• The specific FHIR resource ID in your EHR (useful for advanced users who want to trace an entry back to a specific record).
• The date the information was first documented in your EHR.
• Which EHR connection it came from (important when you have connected multiple providers).
• The clinical notes that mention the item, so the AI (and you) can retrieve the exact passage.

For AI-extracted items, you can see:

• Which conversation the information came from.
• The surrounding context that led the AI to extract it.
• The date and time it was first extracted.

For manually entered items, you see:

• When you added it.
• When you last edited it.

What is "Evidence" on a condition?

Evidence is a more specialized form of provenance specific to conditions. Alongside the source badge, conditions carry a row of evidence icons that answer "what makes us believe this diagnosis is real?" The icons include:

• Medical Records (EHR problem list): the condition appears as a coded entry in your EHR problem list.
• Labs: there are laboratory values consistent with the diagnosis.
• Imaging: imaging or pathology confirms it.
• Specialist: you have been under the care of a specialist who focuses on this condition.
• Medication: you are on a medication that is specifically used to treat it.
• Procedure: a relevant procedure has been performed (for example, a cardiac catheterization for coronary artery disease).
• Confirmed by you: you have manually confirmed the diagnosis.

More icons mean stronger evidence. An entry with five or six filled icons is close to rock-solid. An entry with zero filled icons is a note on a hypothesis.

Which source wins when there is a conflict?

In general: your manual edits win, then your EHR, then AI extraction. Here is what that means in practice:

• If your EHR says you take 10mg lisinopril and you have edited it to 20mg, the edit stands. Future imports will not overwrite your 20mg.
• If your EHR says a condition is Active and you have marked it Resolved, your status stands.
• If the AI extracts something from a conversation that contradicts something in your EHR, the EHR entry is kept and the conversation entry is flagged for your review rather than automatically applied.

You are always the final authority over your own profile.

Can I see the original source document?

For EHR-sourced items, yes. Tapping into an item's details often reveals the underlying note, report, or record excerpt. For items derived from a clinical note, you can see the relevant passage that led to the extraction.

For AI-extracted items, you can see the conversation that produced the extraction.

For items you entered yourself, there is no external document: you are the source.

What does the "indexed" indicator mean?

After an EHR import, your clinical notes and documents are made searchable by the AI. The indexed indicator on a record means that the AI can retrieve it on demand during a conversation ("what did my cardiologist say at my last visit?"). Indexing happens alongside the structured profile build and is usually complete within minutes of the raw records arriving.

Other Profile Sections

Allergies

Your drug, food, and environmental allergies, each with:

• The allergen (e.g., penicillin, peanuts, latex).
• Severity (mild, moderate, severe).
• The reaction you experience (e.g., hives, difficulty breathing).
• Who diagnosed or confirmed it.
• The date it was diagnosed, if known.

Allergies are safety-critical. The AI checks your allergy list before making any medication-related suggestion. A missing penicillin allergy could lead to dangerous guidance; a wrongly listed allergy could lead you to avoid a medication you actually tolerate. Review this section carefully after any import.

Immunizations

Your vaccination history, with the vaccine name, date administered, dose number (for multi-dose series), and where it was administered. The AI uses this to flag due and overdue vaccines based on standard schedules for your age and risk factors.

Family History

Hereditary conditions in your first- and second-degree relatives, with the relationship (mother, father, sibling, grandparent, etc.), the condition, and the age of onset if known. This drives risk-based reasoning: a family history of early-onset heart disease, for example, changes how the AI weighs your own cardiovascular symptoms.

Active Issues

Current health concerns that are being tracked but have not risen to the level of a diagnosed condition. For example, "intermittent shoulder pain for the past three weeks" or "new headaches starting last month." These are usually extracted from conversations and exist to track evolving situations before they become full entries.

Resolved Issues

Past acute issues that have resolved. A sinus infection last winter. A sprained ankle that healed. Keeping these preserves context without cluttering your active issues view.

Test Results (Labs and Imaging)

Lab values, imaging results, pathology reports, and vital signs from your records or from uploads. Each result has:

• The test name (e.g., Hemoglobin A1c, chest X-ray).
• The type (lab, imaging, pathology, procedure, vital).
• The date performed.
• The value and units.
• The normal range for comparison.
• An interpretation (normal, borderline, high, low, abnormal).
• Clinical significance, when relevant.

The Labs sub-page provides interactive charts so you can see how values like your A1c, LDL, or TSH have trended over time. Results can be grouped by body system.

Encounters and Visits

A structured summary of visits: when you went, where, who you saw, what was discussed, and what was decided. These are used to give the AI conversational memory about your care: "at my last cardiology appointment we decided to add a statin" becomes a known fact rather than something you have to repeat.

Clinical Notes

The underlying notes your clinicians wrote at visits. These are indexed so the AI can retrieve specific passages when you ask about a specific visit. You can read the notes themselves if you want to see what your clinicians actually wrote.

Health Priorities and Goals

What matters most to you: staying active, being there for grandkids, managing a chronic condition well, avoiding a specific complication, preparing for a specific goal (a hike, a surgery, a pregnancy). The AI frames recommendations in terms of these priorities when it can.

Engagement Patterns and Tone

Preferences the AI picks up from your conversation style: whether you like lots of detail or a quick summary, whether you prefer autonomy or guidance, whether you enjoy reading studies or just want the takeaway. The AI applies these silently, so conversations feel tailored to you without feeling surveilled.

Uploaded Images and Documents

Every photo and PDF you share with the AI (lab screenshots, prescription photos, MRI reports, symptom photos, medication labels) is stored permanently and accessible in two places: inline in the conversation where you shared it, and in the Uploaded Images area of your profile (organized by category). These files are also indexed, so the AI can answer questions about them later without asking you to re-upload.

Wearables (iOS only)

If you are on iPhone and have granted Apple Health access, the Wearables page shows heart rate, sleep, and activity data synced from HealthKit. The AI can use these data points to ground conversation: if you mention fatigue, it can look at your recent sleep patterns.

The Rebuild from Chats Button

What does Rebuild from Chats do?

There is a button on your Health Profile page labeled Rebuild from Chats. When you tap it (and confirm), My Doctor Friend re-analyzes your entire conversation history and updates the AI-generated portions of your Health Profile. That means it re-reads everything you have ever told the AI and freshens up any extractions that depend on that.

When should I use it?

You do not need to use it routinely. It is there for two main situations:

1. You have had a lot of conversations recently that changed your situation (new symptoms, changes in treatment, lifestyle changes, a new diagnosis you told the AI about), and you want the profile to catch up to the current picture in one pass.
2. Something looks off in the AI-derived parts of your profile, and you want the system to reconsider given the full history. A Rebuild sometimes resolves stale or inconsistent entries.

Most people go months without pressing it. That is fine.

Does it overwrite my EHR data?

No. EHR-imported data is preserved exactly as it is. Rebuild from Chats only touches the AI-derived portions of the profile. Anything that came from your connected records stays put.

Does it overwrite my manual edits?

No. Your manual edits are locked and preserved. If you have corrected a medication dose or changed a condition's status, that change survives the Rebuild. The same goes for dismissals: things you have explicitly removed stay removed.

What does it actually change, then?

The AI-derived entries: conditions the AI inferred from conversation, active issues it has been tracking, tone and engagement patterns, and the narrative summaries it maintains. Think of it as asking the AI to re-read your whole journal with you and update its understanding.

How long does it take?

Usually 30 to 60 seconds. In rare cases (very long conversation histories), up to a few minutes. You will see a spinner while it runs, and a confirmation when it is done.

Will it re-read all my old conversations?

Yes, it is comprehensive. It does not skip older conversations. This is deliberate: older context matters for things like "when did this symptom first appear" or "has this been an issue for years."

Is there a confirmation before it runs?

Yes. When you tap the button, a modal appears asking you to confirm. The modal explains what will happen and reminds you that manual edits and EHR data are preserved. You have to tap OK to proceed, and Cancel to back out.

Can I use the app while it runs?

It only takes half a minute or so, and the page is in a loading state during that time. We recommend just waiting for it to finish. If you navigate away, it will continue in the background and your profile will reflect the new state the next time you load it.

Does Rebuild from Chats replace Refresh Data?

No. They do different things:

• Refresh Data pulls new information from your EHR provider.
• Rebuild from Chats re-synthesizes the AI-derived parts of your profile from your conversation history.

You can use either independently.

Refreshing Your Records

What is Refresh Data?

A button next to each connected provider that tells My Doctor Friend to pull the latest records from that provider. Your records do not sync automatically. You control when to pull updates.

Where do I find it?

Two places:

• Health Profile page. Tap the provider logo in the connected providers strip to expand options, and tap Refresh Data.
• Settings. Under Profile and EHR Management, each connected provider row has a Refresh button.

When should I refresh?

Whenever you expect there is new information to pull. Good moments:

• A day or two after a doctor's visit (to capture the visit note, new orders, and any new diagnoses).
• After getting lab results back.
• After a hospital stay or procedure.
• When you know something was updated in your patient portal.

You do not need to refresh on a schedule. Only when something new has happened.

Is there a cooldown?

Yes. Once you refresh a provider, you need to wait 24 hours before refreshing the same provider again. This is because patient portals often take time to update their own records (a visit note may not appear in MyChart until a day later), so back-to-back refreshes typically return the same data and waste everyone's time and resources.

Is there a monthly limit?

Yes. You get 10 refreshes per month across all your profiles and providers. Your remaining refreshes show up next to the refresh button, for example "7/10 refreshes left." If you hit the cap, the button is grayed out with a tooltip explaining why.

Why 10 per month?

Each refresh costs us real money (Fasten is a paid service, and every pull uses their quota). We set a cap that is generous enough for normal use (a new visit, new labs, a specialist consult, a procedure, and then some) while keeping the service sustainable. If you find yourself hitting the cap regularly, reach out. Power users may have different needs and we can discuss options.

Does refreshing pull everything again, or just new records?

Only what is new. Refresh is incremental: we pull the delta since the last sync. That makes refreshes much faster than the initial import.

Does refreshing re-trigger the full profile build?

Not from scratch. Stage 2 (building the profile) runs again only for the sections that have new or changed records. Most of your profile remains stable, and we only update what has changed. This is why a refresh typically processes much faster than the initial import.

Does the app refresh automatically in the background?

No. Refresh is manual by design. Automated refreshing would pull data on schedules that do not match what users actually want and would quickly burn through the monthly cap without clear benefit. You decide when to pull.

What happens if my patient portal password changes?

Fasten's connection uses an authorization token that persists until you revoke it or your health system invalidates it. If your password changes and your health system revokes your token, the next refresh will fail with an authorization error. In that case, disconnect and reconnect the provider to re-authorize with your new password.

What happens if my health system changes systems (for example, migrates from Cerner to Epic)?

You may need to reconnect. A system migration usually invalidates existing authorizations. When that happens, your existing data is preserved in your profile, but you will need to do the connect flow again to resume syncing.

Why is the Refresh button grayed out sometimes?

Three possible reasons:

1. You are still within the 24-hour cooldown window. The button shows something like "Available in ~18h."
2. You have used all 10 monthly refreshes. The cap resets at the start of the next month.
3. The provider is currently processing a sync. Only one refresh can run at a time per provider.

Connecting Multiple Providers

Can I connect more than one provider?

Yes. Every profile supports multiple connected providers. If you see doctors at different health systems (primary care at one network, specialists at another, labs at a third), you can connect each of them.

How do I add another provider?

From desktop web, use + Add Another Provider in the connected providers strip on the Health Profile page, or the relevant provider management area in Settings. The connection flow is the same as the first one.

Do data from multiple providers get combined?

Yes. All imported data merges into your unified Health Profile. Conditions from your primary care system and conditions from your cardiologist's system appear together in the Conditions section. Medications from both sources appear together in the Medications section. And so on.

How are duplicates handled?

When two providers report the same condition or medication, the system tries to recognize that and consolidate them into a single entry with multi-source provenance (the Mixed badge). This is not always perfect: sometimes the same diagnosis is coded slightly differently by different EHRs, or the same medication appears under a brand name in one feed and a generic name in another. The system errs on the side of keeping both rather than silently discarding, and then proposes consolidations for your review.

What is the Suggested Changes view?

When a new import (initial or refresh) produces updates that the AI thinks you should review (not automatically apply), those show up as a Suggested Changes indicator on your Health Profile page. Tapping it opens a review modal where each suggested change is listed with:

• What kind of change is proposed (remove, update, mark resolved, etc.).
• A confidence level (high, medium, low).
• The reason the change is proposed.
• An evidence snippet from your conversation or records.

For each suggestion, you have two buttons: Accept (applies the change) and Dismiss (rejects it). You can go through them one at a time.

Are multi-EHR merges reviewed here too?

When a second connected provider introduces items that conflict with or duplicate existing items, the AI will typically surface the significant ones in Suggested Changes for your review, rather than silently auto-merging. Minor or clearly non-conflicting items (for example, a new lab result from the second provider that has no counterpart elsewhere) are merged into the profile automatically.

Should I connect multiple providers if I only see one regularly?

Not necessarily. The main reason to connect multiple providers is to capture care that lives in different systems (specialists at one hospital network, primary care at another, labs at an independent lab). If all your care is in one system and that system shares labs and specialist notes, one connection is enough.

What if two providers disagree about something?

It happens. Let's say your cardiologist's system lists "paroxysmal atrial fibrillation" and your primary care's system lists "history of atrial fibrillation," and the distinction matters (current vs. past). The system will typically keep both and let you reconcile in review. You can edit the entry to reflect the correct current state.

Does the 10-refresh monthly limit apply across all providers?

Yes. The cap is shared. If you refresh provider A three times and provider B seven times, you have used all 10.

Reviewing and Correcting Your Profile

Why should I review after an import?

Because medical records are often out of date. A resolved condition may still be listed as active. A medication you stopped two years ago may still appear on your med list. A lab value might have been transcribed incorrectly. Reviewing is the step that turns an imported record into an accurate, current profile.

It is the same reason a doctor's office asks you to confirm your medication list at every visit: records drift.

What should I check first?

We recommend this priority order:

1. Allergies. Safety-critical. Make sure everything you are allergic to is listed, and nothing is listed that you are not. The AI will use this before every medication-related suggestion.
2. Active medications. Is the list correct, are the doses current, and are any medications you no longer take still marked Active?
3. Active conditions. Are any of these resolved or managed? Is anything current missing?
4. Everything else. Family history, immunizations, labs, and so on.

How do I mark a condition as resolved?

Tap the condition, open the detail view, change the Clinical Status to Resolved. You can optionally add a resolution date and notes (for example, "resolved after gallbladder removal in 2022").

How do I mark a medication as discontinued?

Tap the medication, change Status to Discontinued. You can add an end date and a reason (side effects, lack of effect, switched to another medication).

How do I correct a wrong entry?

Tap the entry, edit the fields, save. Your change is saved immediately, and future imports will not overwrite it.

Does updating my records in My Doctor Friend change my records at my health system?

No. Corrections you make in My Doctor Friend do not write back to your hospital, clinic, lab company, patient portal, or EHR. The corrected version will live in your My Doctor Friend Health Profile, and My Doctor Friend will use that corrected version when helping you.

If the original record at your health system is wrong, the long-term fix is to ask your clinician, clinic, or medical records department to correct it at the source. We wish we could send corrections back automatically, but health systems generally do not allow consumer apps to update their official records this way.

How do I delete an entry entirely?

Tap the entry, tap Delete. The entry is removed and added to the dismissal list, so the AI will not re-extract it automatically. If you change your mind, deleted items can be restored.

What is the Section Review feature?

Some sections (especially the ones populated after an EHR import) have a Review button at the top. Tapping it walks you through the items in that section one at a time, letting you confirm, edit, or delete each one. When you complete a section review, the profile tracks that you have done so, and the review badge disappears.

The purpose is to encourage a thorough first-pass review after an import, so the AI is working from a profile you have actually verified.

Will my edits survive future imports?

Yes. The system keeps a record of every manual edit you make, and future imports from the EHR will not overwrite those edits. If your EHR later updates the same field and you want to see the update, you can tap Restore AI (or Restore EHR) on the item to revert to the imported value.

I deleted something by accident. Can I get it back?

Yes. Deleted items live in a dismissal list for a period of time, and you can restore from there. If you need help finding it, contact support.

What about items the AI has "inferred"?

AI-inferred conditions (and other items) are marked as Unconfirmed by default and are highlighted for your review. If you confirm, they move up in certainty. If you dismiss, they are removed. The AI does not reinforce an inferred condition into a confirmed diagnosis on its own.

I keep seeing the same suggestion every time I chat. Why?

Usually because something in your conversation is strongly suggesting that fact to the AI, and you have not yet acted on the suggestion. Either confirm it (if it is correct) or dismiss it (if it is not). Dismissing prevents re-suggestion of the same item.

Disconnecting a Provider

How do I disconnect?

From Settings, find the provider you want to disconnect, tap Disconnect, and confirm. The connection is revoked immediately. No further data flows from that provider.

What happens to the data I already imported?

By default, your imported data stays in your Health Profile. Disconnecting stops future syncs; it does not erase the past. You can still edit, keep, or remove items from your profile as you normally would.

Can I remove the imported data as well?

Yes. When you disconnect, you can choose to also purge the underlying imported data. This deletes:

• The raw EHR files we stored from this provider.
• The processed records we created from them.
• The indexed versions of your clinical notes from this provider.
• The per-connection provenance records.

Your Health Profile entries remain, but their backing evidence from this provider is gone. If other providers also contributed to the same entry, their evidence is unaffected.

What if I change my mind and want to reconnect?

You can reconnect at any time. The flow is the same as the first connection: open Fasten, search for your provider, log in, authorize. If you had previously purged data, you will need to re-pull everything.

Will disconnecting affect my other connected providers?

No. Each connection is independent. Disconnecting one does not affect the others.

Does disconnecting cancel my subscription?

No. These are separate. Disconnecting just stops the data flow from one provider. Your subscription continues as normal.

Troubleshooting

I can't find my health system in the Fasten search

Try searching for the parent network (for example, "Sutter" instead of "Sutter West Valley"). Try the portal brand (for example, "MyChart" or "FollowMyHealth"). If it is still not there, your provider may not be supported yet. Some lab companies, including LabCorp, are not supported as standalone direct connections today. If the lab was ordered by a connected clinic or health system, the result may still appear through that health system's portal. Reach out to us and we will check whether we can do anything on our side.

The Fasten widget closed before I finished

Sometimes the widget times out, or a redirect fails, or the login flow gets interrupted. The fix is usually to try again: tap Connect and restart the flow. Your progress is not saved, so you will go through the login again, but it is quick.

I got an "already connected" error

This means the provider is already linked to your profile. Check your Settings page. The existing connection is the one to use. If something looks wrong with it (stale data, an error), try a Refresh or, if that fails, disconnect and reconnect.

I got a "connection exists for another account" error

The credentials you are using are linked to a different My Doctor Friend account. This most often happens when a household shares credentials and one person already set up the connection. Use a different account, or reach out to us to help sort it out.

It says "records connected," but I don't see anything in my profile

Processing takes time. Immediately after connecting, the raw records are being downloaded and parsed. The Health Profile entries start appearing as each processing step completes. Give it time: 60 to 90 minutes on average, up to two hours for very large records.

If two hours pass and the profile is still empty, something has gone wrong. Look for an error banner, check the processing status, and retry or contact support.

Processing is stuck at Step X

If the banner has not advanced for more than 20 minutes, the job has probably stalled. We auto-detect this and will typically mark the job failed with a Retry button. Tap Retry. If retries keep failing, contact support.

I see wrong information in my profile

Edit it. Your edits are preserved. If the wrong information keeps reappearing, it is coming from somewhere specific (an EHR entry that is actually incorrect in your record, for example), and editing locks your correction in place.

For EHR-sourced items that are genuinely wrong in your underlying medical record, the long-term fix is to correct the record at its source (ask your clinician to update it). In the meantime, your edit in My Doctor Friend gives you the correct view.

I see duplicates

Common when multiple providers are connected and both report the same condition or medication. The system tries to deduplicate, but does not always succeed. You can delete the duplicate manually; the remaining entry will carry the merged context.

If duplicates are appearing between conversation-extracted and EHR-imported entries, this is a signal that the AI has not recognized that the two are the same. You can delete one.

My lab results aren't showing

Lab and immunization imports are being actively developed and may not be 100 percent complete for every EHR. What is present is what your EHR surfaces; what is missing is typically either not yet supported for your specific provider or not recorded in the EHR format we consume.

LabCorp is not currently supported as a standalone direct connection. Many LabCorp results can still arrive through a health system portal if the test was ordered by a doctor or clinic that uses a supported portal.

In the meantime, the underlying lab documents that do import are indexed and the AI can reference them in conversations. If you ask "what was my last A1c?", the AI can look it up even if the Labs section does not yet show it in a structured way.

The Refresh button is grayed out

One of three reasons:

1. You are within the 24-hour cooldown on that provider.
2. You have used all 10 refreshes this month.
3. A sync is currently in progress.

The tooltip next to the button explains which one.

I got a "timed out" error during connection

Sometimes your health system is slow to respond to the authorization, especially during peak hours. Try again later from desktop web. If it keeps failing, contact support.

I'm getting a bunch of duplicate conditions

This often means two sources (for example, one EHR and your conversations) are reporting the same thing and the system has not yet recognized them as the same. You can manually delete duplicates. Future AI behavior will improve in recognizing these patterns over time.

The AI seems to have forgotten something I told it

Two things to check: first, the Dismissed list. If you dismissed an item earlier, the AI will not re-extract it. Second, make sure the item is in your profile. Telling the AI something does not always result in a profile entry, especially if the mention was very casual or brief. You can add it manually.

I'm a caregiver and I'm seeing the wrong person's data

Make sure you are on the right profile. The profile switcher at the top of the app lets you toggle between profiles. If you are on "Mom" and you see your own data, something is misrouted. Log out and back in, or contact support.

Something else is wrong

Reach out to us. We are a small team and we read every support message. Tell us what happened, what you expected, and (if possible) any error messages you saw. We will investigate and get back to you.

Availability, Plans, and Support

Who can use the Medical Records feature?

Premium members can use the Medical Records feature. New EHR connections currently need to be started from desktop web.

Is it on the free plan?

Connecting medical records is a Premium feature. The Health Profile itself, including the structured view, AI extraction from conversations, manual entries, uploads, and photos, is available on the free plan. You can use My Doctor Friend well without ever connecting records. Connecting is a shortcut and an accuracy boost for people who have records to connect.

Does it work on Android?

You can use My Doctor Friend on Android through the web app, but you cannot start a new EHR connection from mobile web today. Start the connection from desktop web.

Does it work on iPhone?

You can use My Doctor Friend in the native iOS app, and Apple Health integration is iOS-only. You cannot start a new EHR connection from the iOS app today. Start the connection from desktop web.

Are caregiving profiles supported?

Yes. You can create a profile for a family member you care for, such as a parent, child, or spouse, and connect their records under that profile if you are authorized to do so. Their data stays on their profile and is not mixed with yours. The AI addresses you as the caregiver when you are viewing their profile.

How many family members can I manage?

Multiple profiles are supported. If you are on Premium, you can connect EHRs for multiple profiles you manage. If you run into a profile limit or have a larger caregiving situation, contact support.

Is there a separate family plan for medical records?

No. There is no separate family plan for this feature today. The key distinction is free plan versus Premium. Premium unlocks medical record connections, including for multiple profiles you manage.

How do I get help?

• Reply to any email we have sent you, and it will reach our support team.
• Email us directly (address in your account).
• Use the in-app contact option if available.

We read every message. If something is confusing, broken, or just frustrating, tell us.

Can someone walk me through the setup?

Yes. If you would like a guided setup (for example, you want help finding your health system in Fasten, or you want us to walk through your first profile review with you), reach out and we will schedule a call. This is free. Connecting health records is genuinely tricky for a lot of people, and we would rather help than leave anyone stuck.

What's new? How do I find out about updates to this feature?

We post updates to our changelog and send periodic emails to active users. Major improvements to the Health Profile, Medical Records, or any of the processing pipelines are announced explicitly so you know when behavior changes.

I have feedback on this FAQ

Tell us. If something is unclear, missing, or just wrong, we want to know. This page exists to help; if it is not helping, we will fix it.