# FixFirst — Full Reference > FixFirst is a free AI-powered blood test analyzer. Upload any blood test PDF or > image and get your top 3 health priorities in plain English — ranked by clinical > urgency and personalised by age, sex, and lifestyle. No account required, no data > stored. FixFirst is an educational tool, not a medical device. It does not diagnose > conditions or replace clinical consultation. This file is the fuller companion to /llms.txt. It consolidates the explanatory content already published across the site (methodology, clinical approach, and common questions) into one reference an AI assistant can cite directly. Every statement here is drawn from existing on-site copy; nothing new is asserted. ## What FixFirst is FixFirst reads blood test reports — PDFs or images from any lab (LabCorp, Quest Diagnostics, NHS, private clinics) — and returns the three findings that matter most, in order, with a plain-English explanation and specific next steps. It is built for the person who has a "normal" or "mostly normal" report but still wants to know what, if anything, to act on first. It is free, requires no account, and does not store uploaded reports or results. Reports are processed in memory and discarded after analysis. ## How the analysis works: four stages FixFirst is a hybrid of AI extraction and a deterministic rules engine. The AI reads the document; deterministic, medically reviewed rules make every clinical decision. 1. Extraction (AI). A large language model (Claude, by Anthropic) reads the uploaded PDF or image and identifies every blood marker: name, value, unit, lab reference range, and any H/L flags printed by the lab. The AI handles the wide variation in lab report formats. It makes no clinical judgement at this stage — it produces structured data only. 2. Threshold correction (deterministic). Extracted values pass through a rules engine that applies evidence-based threshold overrides. This handles two problems: (a) lab reference ranges vary between labs and can be wider than current clinical guidelines recommend; (b) some markers require sex- or age-adjusted interpretation the raw lab range does not reflect. For example, a TSH of 3.8 mIU/L may sit within a lab's printed range but warrant borderline classification under ATA guidelines. All threshold decisions are hardcoded, traceable to a specific guideline, and medically reviewed. 3. Ranking (deterministic). Each non-optimal marker is scored across six dimensions: severity of deviation, clinical impact, actionability with lifestyle/supplement interventions, expected response time, questionnaire context (sex, age, medications, diet), and lab-flag weighting. The three highest composite scores become the Top 3 priorities. Markers ranking 4th or lower are shown in a secondary list with graded status labels but without full protocol recommendations — the intent is to direct attention, not overwhelm. Questionnaire context is a priority multiplier only — it never changes a guideline threshold, only how urgently a result on the wrong side of that threshold is surfaced. Each factor (activity level, diet, age/sex, supplements, family history) is explained with its guideline evidence in the Context & personalization guides at https://fixfirst.health/guides 4. Recommendations (deterministic). Protocol items — diet, lifestyle, supplements, and expected response timelines — are drawn from a static database keyed to each marker's identity, severity tier, and direction (high or low). No AI generates these recommendations. Every item was written with reference to published clinical guidelines and reviewed by Dr. Prahlad Rai Gupta, MD. When two top-3 markers need conflicting dietary advice, a cross-marker reconciliation step flags the conflict explicitly. ## What is AI vs deterministic AI does: reading marker names from report text; identifying values, units, and ranges across variable formats; recognising H/L/H*/L* flags; handling OCR variation in scanned documents. Deterministic rules do: all status classifications (optimal / borderline / abnormal / critical); all threshold values and sex/age adjustments; all ranking scores and priority decisions; all protocol items (diet, lifestyle, supplements). ## Status levels FixFirst classifies each marker into four levels: Critical, Abnormal, Borderline, and Optimal. The Top 3 are the highest-priority non-optimal findings. ## Clinical approach - Reference ranges are sourced from clinical guidelines and adjusted to current evidence rather than relying solely on a lab's printed range. - Guidelines anchored to include ADA, ATA, ACC/AHA, NICE CKS, NIH ODS, and WHO. - Thresholds are sex-adjusted for hemoglobin, ferritin, HDL cholesterol, ALT, and uric acid. - Priority ranking considers severity, direction (high vs low), co-occurring markers, and user lifestyle context. - Content is reviewed by a qualified medical advisor (Dr. Prahlad Rai Gupta, MBBS, MD Pulmonary Medicine). ## Biomarkers covered Hemoglobin, ferritin, vitamin D (25-OH), vitamin B12, TSH, HbA1c, LDL cholesterol, HDL cholesterol, triglycerides, ALT/AST, creatinine/eGFR, CRP/hsCRP, uric acid, fasting insulin, magnesium, and the full CBC panel (WBC, neutrophils, lymphocytes, MCV, MCH, MCHC, RDW, platelets), among others. 138 markers in total. Some marker pairs share a single root cause rather than being independent findings — for example, an underactive thyroid both impairs iron absorption (raising TSH while lowering ferritin) and slows LDL clearance (raising TSH alongside LDL). FixFirst's combination guides cover these mechanisms and which marker to address first; see /llms.txt for the current list. ## Common questions What should I fix first in my blood test? FixFirst ranks every out-of-range finding by clinical urgency and actionability and returns the three that matter most, so you start with the highest-impact item rather than guessing. Why do I feel unwell when my results are "normal"? Lab reference ranges are wide and vary between labs. A value inside the printed range can still be suboptimal — for example, ferritin or vitamin D at the low end, or TSH in the upper-normal zone. FixFirst applies evidence-based thresholds to flag these. Is FixFirst free? Yes. There is no account and no payment required. Does FixFirst store my data? No. Reports are processed in memory and discarded after analysis. No account or personal data is required or retained. Is this medical advice? No. FixFirst is an educational tool, not a medical device. It does not diagnose conditions or replace consultation with a clinician. ## Trust & authorship - Founder: Ankit Agarwal. - Medical reviewer: Dr. Prahlad Rai Gupta (MBBS, MD Pulmonary Medicine). - Founder and reviewer credentials are detailed at /about. Methodology, medical review policy, and editorial policy are published at /methodology, /medical-review-policy, and /editorial-policy. ## Additional content - Free calculators (/tools): Browser-based eGFR (2021 CKD-EPI race-free equation) and LDL cholesterol (Friedewald and Sampson/NIH equations) calculators. Values never leave the device. - Research commentary (/blog): Dated commentary on new blood-test and longevity research, explained the week it breaks, with links to the studies and labs it affects: - [Why Your Watch's Blood Pressure Never Matches Your Cuff](https://fixfirst.health/blog/cuffless-blood-pressure-wearable-accuracy) (Sep 2026): Cuffless wearable blood pressure estimates a pulse waveform, not a real reading; the FDA is still drafting a formal accuracy standard for it - [You Lost Weight and Your Testosterone Dropped Anyway](https://fixfirst.health/blog/weight-loss-testosterone-drop-energy-deficit) (Sep 2026): Weight loss usually raises testosterone; when it falls with LH, that pairing points at energy deficit suppressing the pituitary, not the testes - [A New Pill Mimics One Exercise Hormone, Not Exercise](https://fixfirst.health/blog/exercise-pill-env-308-lac-phe) (Sep 2026): Enveda's ENV-308 passed a Phase 1 trial mimicking Lac-Phe, a hormone released during sprints, and lowered leptin — without reproducing what exercise does - [David Sinclair's Self-Testing Protocol, and the Pushback](https://fixfirst.health/blog/david-sinclair-blood-test-protocol) (Sep 2026): Sinclair tracks homocysteine and B12 on an off-label metformin stack; a peer-reviewed critic disputes the sirtuin science behind it - [Three Minutes of Sprints Changed 721 Blood Proteins](https://fixfirst.health/blog/sprint-exercise-blood-protein-study) (Aug 2026): A Cell Reports Medicine study found sprint intervals shifted 721 of 2,884 blood proteins immediately, versus 7 after 90 minutes of moderate cycling - [Biological Age Clocks: What Moves Them](https://fixfirst.health/blog/epigenetic-aging-clocks-longevity-study) (Aug 2026): A 51-study Yale synthesis found prescription drugs like metformin and semaglutide moved epigenetic aging clocks; supplements mostly did not - [FDA Clears an Alzheimer's Blood Test for Age 40+](https://fixfirst.health/blog/precivityad2-alzheimers-blood-test-age-40) (Aug 2026): PrecivityAD2 is the third FDA-cleared Alzheimer's amyloid blood test, and the first cleared for adults under 55 - [A Lung Cancer Blood Test Hit an FDA Milestone](https://fixfirst.health/blog/lung-cancer-blood-test-fda-designation) (Aug 2026): Freenome's SimpleScreen Lung got a Breakthrough Device Designation, not clearance — and 90.7% sensitivity came at 50% specificity - [Cialis as a Longevity Drug: What the Data Says](https://fixfirst.health/blog/tadalafil-longevity-claims) (Aug 2026): Online clinics market tadalafil for heart and brain benefit on one observational study of men treated for ED, with no randomised trial behind it - [The FTC Sued Hims & Hers Over Your Health Data](https://fixfirst.health/blog/health-data-privacy-hims-hers) (Aug 2026): FTC alleges Hims & Hers shared health data with Meta and Snap despite promising privacy — what to check before uploading your own - [A New Study Ties Coffee to Testosterone and Insulin](https://fixfirst.health/blog/coffee-testosterone-insulin-study) (Aug 2026): A 2,264-person Finnish study found coffee drinkers had better glucose, insulin, and testosterone numbers at the same BMI - [Your Watch Guesses Insulin Resistance. Should You Trust It?](https://fixfirst.health/blog/pixel-watch-insulin-resistance-blood-test) (Aug 2026): Pixel Watch 5 estimates insulin resistance with no blood draw — what that number can and cannot tell you - [She Left OpenAI to Fix What "Normal" Blood Work Misses](https://fixfirst.health/blog/chroniclebio-ai-blood-biomarkers) (Aug 2026): An ex-OpenAI exec used AI on 3,500 blood vials to sort POTS patients into biomarker groups standard panels missed - [A Cholesterol Treatment Also Clears PFAS Forever Chemicals](https://fixfirst.health/blog/apheresis-pfas-forever-chemicals) (Aug 2026): A small German study found apheresis cut PFAS forever chemicals in blood by up to 25% as a side effect - [The New Colon Cancer Blood Test Misses Most Polyps](https://fixfirst.health/blog/colon-cancer-blood-test-2026) (Aug 2026): FDA-approved SimpleScreen CRC finds 81% of colorectal cancers but only 13.7% of advanced precancerous lesions - [Bryan Johnson Says He Took Longevity Too Far](https://fixfirst.health/blog/bryan-johnson-autoimmune-gastritis) (Aug 2026): An autoimmune gastritis diagnosis — a condition that hides behind ordinary bloodwork for years — prompted him to question his own protocol - [Exercise Erases Half of Muscle Aging, Study Finds](https://fixfirst.health/blog/exercise-reverses-muscle-aging) (Jul 2026): A Nature Aging study found trained older adults missing over half the molecular signature of muscle aging - [Blood Tests Can Finally Diagnose Endometriosis](https://fixfirst.health/blog/endometriosis-blood-tests-2026) (Jul 2026): Two non-FDA-cleared blood tests reached US clinics in 2026, cutting a diagnosis that used to require surgery and years - [Diet Soda and Faster Memory Decline: What the Study Found](https://fixfirst.health/blog/artificial-sweeteners-cognitive-decline) (Jul 2026): A study of 12,772 adults linked artificial sweeteners to 62% faster memory decline under age 60 - [A Fourth Alzheimer's Test Simplifies the Other Three](https://fixfirst.health/blog/ptau217-alzheimers-blood-test-single-cutoff) (Aug 2026): Roche and Eli Lilly's pTau217 is the first Alzheimer's blood test cleared for both rule-in and rule-out with one cutoff, in primary or specialty care - [Why Your GP Won't Explain Your Blood Test Results](https://fixfirst.health/blog/nhs-blood-test-results-gp-gap) (Sep 2026): A UK audit found only 26.6% of GP-ordered blood panels come back fully normal, yet the NHS App just flags "abnormal, contact your doctor" with no explanation - [Your GP Can Now Order an Alzheimer's Blood Test](https://fixfirst.health/blog/alzheimers-blood-test-diagnostics) (Jul 2026, updated Aug 2026): Four FDA-cleared amyloid blood tests, one orderable in primary care nationwide since Feb 2026, one cleared for adults as young as 40, one a single-biomarker test cleared for both rule-in and rule-out; symptoms are the gate, not a specialist referral - Longevity biomarker targets (/longevity): The blood markers named longevity researchers and public figures track in their own protocols, attributed to whoever cited each target — not a FixFirst clinical recommendation. - Peptides & hormones (/peptides-and-hormones): What to test before starting a peptide or hormone therapy and when to retest, for TRT, GLP-1, and sermorelin. Guideline-backed monitoring panels, not dosing advice. ## About - URL: https://fixfirst.health - Full page index: https://fixfirst.health/llms.txt - Last updated: 2026-09-12