Import your data
Coming from another app, a spreadsheet, or a paper diary? Bring your history into TrackDX in a few minutes.
Easiest: do it in the app
In TrackDX, go to Me → Import from another tracker. It gives you the template and the AI prompt, then checks the converted file and shows you exactly what will import before anything is saved. Requires TrackDX 2.0 or later.
Prefer to use a computer?
Converting is often easier on a laptop. You'll use an AI assistant such as Claude, ChatGPT or Gemini to reformat your old records into TrackDX's layout. It works for conditions that come and go in episodes, like migraine, and for ones you rate every day, like sciatica.
- Download the TrackDX Import Assistant Get the workbook below. Attaching it to the chat is strongly recommended — it shows the AI the exact layout TrackDX reads.
- Start a new AI chat Attach the workbook and your old data, then paste in the prompt below.
- Import the file the AI sends back Save it to your phone, then in TrackDX go to Me → Import from another tracker → Import the converted file.
TrackDX Import Assistant
Templates, format guide, and Claude prompt, all in one workbook — covering episodic and daily conditions, procedures, and medication history.
Download the workbookI'm moving my personal health-tracking data into an app called TrackDX. I've attached two files: 1. My existing data — my history from another app, a CSV export, or notes I've typed up describing my symptoms, triggers, and medications. 2. "TrackDX Import Assistant.xlsx" — a workbook with a "Format Guide" tab (explains every column), worked examples, blank templates, and templates for the extra sheets. FIRST, decide how each of my conditions is tracked. TrackDX supports two kinds, and they use DIFFERENT column layouts: - EPISODIC — the condition comes and goes in discrete attacks. Migraine, seizures, cluster headaches, gout flares. My data will show individual events with a start time, or a "did it happen today?" flag, and long stretches with nothing. - DAILY — the condition is present most days and I rate how bad it is. Sciatica, arthritis, neuropathy, fibromyalgia, chronic back pain. My data will show a severity number on most days, often broken down by body location. Rule of thumb: if my data has a pain or severity score on most days, it's DAILY. If it records separate attacks with gaps between them, it's EPISODIC. Tell me which you chose for each condition and why. If I have both kinds, put each condition on its own sheet in the correct layout — one workbook can hold a mix. Then convert my data. For BOTH kinds: - Create ONE sheet per condition, named with the condition's name (max 15 characters — that's Excel's tab-name rule TrackDX follows). - ONE ROW PER CALENDAR DAY, from the earliest date in my data through today, INCLUDING days with nothing logged. The empty days are real information to TrackDX's statistics, so don't skip them. - PUT THE COLUMN HEADERS IN THE VERY FIRST ROW of each sheet. No title, caption or blank row above them — TrackDX reads row 1 as the header and stops, so anything above it makes the whole sheet import as empty. - Date format must be YYYY-MM-DD. Times must be HH:MM in 24-hour format. - Don't add columns of your own, and don't keep unmapped fields "just in case" — anything outside the layouts below is ignored on import. - Delete the template's example rows before handing the file back. - LIMITS: keep every condition, trigger and medication name to 30 characters or fewer, and use at most 10 different triggers per condition. If my data has more, keep the 10 I logged most often and tell me which you dropped — TrackDX keeps only the 10 most frequent anyway. - "Drinks" is the NUMBER of alcoholic drinks that day. "Drink Type" is the main drink, and it must be exactly one of these four words: Red wine, White wine, Beer, Spirits. Map whatever my data says onto one of them — vodka, whisky, gin, tequila, rum, a margarita or any other cocktail are all "Spirits"; lager, ale, IPA, stout and cider are all "Beer"; rosé, prosecco and champagne are "White wine"; name a grape and use its colour (merlot → Red wine, chardonnay → White wine). If a drink genuinely doesn't fit — sake, mead, hard seltzer, port, sherry — leave Drink Type BLANK and keep the count. Never invent a fifth value: TrackDX ignores anything outside the four, so a row saying "Margarita" imports with the count but no type and silently drops out of the drink-type breakdown. Tell me how you mapped my drinks. === EPISODIC LAYOUT === Columns, in this order: Date, Episode Occurred, Episode Count, Triggers Logged, Trigger Start Time, Trigger End Time, Drinks, Drink Type, then a repeating group of 8 columns per episode (Episode 1 Start Time, Episode 1 End Time, Episode 1 Duration (h), Episode 1 Intensity (1-10), Episode 1 Relief Medication, Episode 1 Relief Helped, Episode 1 Suspected Cause, Episode 1 Notes — then Episode 2 …, Episode 3 …, if more than one episode ever happens on the same day). - "Episode Occurred" is 1 if any episode happened that day, else 0. "Episode Count" is how many. - "Episode X Relief Medication" can list SEVERAL medications separated by "; " (e.g. "Excedrin; Maxalt"). Only medications taken to relieve that episode — preventatives go on the Med Periods sheet. - "Episode X Relief Helped" is 0–10 for how much it helped: 0 = not at all, 10 = completely resolved. One score per medication in the SAME ORDER, separated by "; " (e.g. "4; 9"). Convert whatever my data uses: a percentage divides by 10 (80% → 8); "Yes"/"Somewhat"/"No" become 10/8/0; a 1–5 scale multiplies by 2. Tell me which conversion you used. - "Episode X Intensity (1-10)" is 1–10; convert other scales proportionally and say how. - "Episode X Notes" is optional free text about that specific episode. Keep it short; don't invent notes. - A row with intensity 0 and no episode is NOT an episode. Many trackers record 0 or blank on symptom-free days — leave the episode columns blank and set "Episode Occurred" to 0. Such a day can still be a trigger day, so keep its Triggers Logged and Drinks values. === DAILY LAYOUT === Columns, in this order: Date, Triggers Logged, Trigger Start Time, Trigger End Time, Drinks, Drink Type, then one "Severity: <location>" column per body location I track, then Notes. - The "Severity: " columns are what tell TrackDX this is a daily condition, so the spelling matters exactly — "Severity: Lower back", "Severity: Left calf". There is no "Episode Occurred" or "Episode …" column on a daily sheet at all. - MAXIMUM 6 locations. If my data has more, pick the 6 most frequently recorded and tell me which you dropped. - If my data records only one overall score with no body locations, use a single column named "Severity: Overall". - Severity values are 0 to 10, in steps of 0.5. Convert other scales proportionally and tell me how. - CRITICAL — blank and 0 mean different things and must not be swapped. A BLANK cell means "not recorded that day". A 0 means "I recorded it, and there was nothing there". Never fill a missing day with 0, and never blank out a genuine recorded 0. If my source can't distinguish them, leave it blank and tell me. - "Notes" is optional free text for that day. - PHYSICAL THERAPY: if a condition is physical therapy for one body part, name it "PT - " followed by the condition, with L or R for the side — e.g. "PT - R Rotator Cuff", "PT - Low Back Pain" (a hyphen, not a colon). Always use the DAILY layout for it. Record the days I had a therapy session as a trigger named exactly "PT session" in Triggers Logged. The sheet tab may be cut to 15 characters, so always put the full name on the "Conditions" sheet. === EXTRA SHEETS (add these when the data supports them) === "Conditions" — one row per condition. Columns: Condition, Tracking Mode, Sites. - "Tracking Mode" is the word "Daily" or "Episodes". - "Sites" is the full list of body locations for a daily condition, separated by "; " — in the order I'd want them shown. Include locations even if they were rarely recorded, so none get lost. Leave blank for episodic conditions. - Include this sheet whenever any condition is daily. "Procedures" — surgeries, injections, blocks, ablations. Columns: Condition, Procedure, Date, Site, Notes, Recalled Overall, Recalled By Site. - One row per procedure. "Site" is the body location it targeted, if any. - "Recalled Overall" and "Recalled By Site" are ONLY for a procedure that happened BEFORE my logging starts, where I've told you my remembered severity beforehand. "Recalled By Site" format is "Location=Value; Location=Value" (e.g. "Lower back=7; Left calf=5"). - Do NOT estimate or invent these. Leave them blank unless I explicitly gave you remembered numbers — TrackDX labels anything derived from them as recalled rather than recorded, and a made-up baseline would silently distort how well a procedure appears to have worked. "Med Periods" — which preventative medications I was on, and when. Columns: Condition, Medication, Start, End (blank End = still taking it). - Preventative medication history goes here, never in the daily columns. - IF A DOSE CHANGED, SPLIT IT INTO SEPARATE MEDICATIONS — "Depakote 500mg" and "Depakote 1000mg" — each with its own start and end, rather than one continuous period. TrackDX compares every preventative in a single model, so separate entries let it report each dose on its own; kept as one medication, a dose that worked gets averaged with one that didn't and both look mediocre. Only split when a dose lasted a month or more; ignore brief changes that were reversed. Tell me which doses you split. When you're done, give me the finished file as a downloadable .xlsx, and summarize the tracking mode you chose per condition, any scale conversions, and any judgment calls — so I can review before importing.