Before we start.
The foam in the Murrysville files was put into breathing machines for a mechanical reason, to reduce sound and vibration. The firm's own assessments came to judge the degraded foam not biocompatible. That record raises a question about every plastic, rubber, foam and coating that a device places against the body or in the path of what enters it, and the whole of this book is built around that question:
Why can a plastic that is safe in a water bottle harm a patient through a breathing circuit or a catheter, and what must a laboratory measure to see that harm before patients meet it?
A plastic judged safe for a water bottle has been judged for one use: what it releases into water that a person then drinks. Made into a breathing circuit or a catheter, the same plastic would meet warm air or blood, for hours or for days, and whatever it released would reach the lungs or the circulation directly. Chapter 1 therefore defines biocompatibility as a judgment about a device in a particular use, not a property of its material.
The answer starts with what the body meets, and the body never meets a material as a whole. It meets what leaves the material, molecules and particles, and it meets the material's surface. Proteins from blood or tissue fluid coat that surface as soon as the two touch, so cells meet a layer of proteins rather than the bare material (Chapter 8). A molecule that leaves the material does harm only at a dose.
Prediction therefore rests on three kinds of knowledge. Chemistry finds what can come out of the material, exposure estimates how much of it reaches the patient over the contact, and toxicology sets the dose that is tolerable. Biological tests then answer what that arithmetic cannot settle: how cells respond to a surface, whether a substance can cause allergy, how a mixture acts and what particles do. The Murrysville files hold each kind of evidence: emissions testing of a machine's output, a risk assessment that weighed compounds of concern for patients of 30 kg and 70 kg, and tests on cells and for mutations.
Prediction fails when the material in the patient's device is no longer the material that was evaluated. Processing, sterilization, cleaning, aging and a change at a supplier can each alter what a material releases, and so can use itself: the firm's reports describe a foam whose chains were cut in heat and humidity. A supplier can also deliver the wrong material. The same Form 483 records that an incorrect polyester polyurethane foam, not the one specified, came from the firm's raw foam supplier. Chapter 23 takes up how supplier control is meant to catch such a material.
Every chapter rests on one fact, the invariant of this book: no material is inert at the scale the body works on. In everyday speech an inert material is one that does nothing, and no material meets that standard. When a supplier or a paper calls a material inert, this book reads the word as a claim to be checked: that what leaves the material is too little to matter for the use in question. Molecules leave every material, proteins arrive on its surface wherever blood or tissue fluid touches it, and both happen at a rate. The rate, added up over the time of contact, sets how much leaves the material, and the route decides how much of that reaches the patient. That amount, set against the size of the patient, is the dose (Chapter 14), and the dose is what the body answers.
The rate itself is not fixed. When diffusion in a plastic limits release, the layer under its surface empties, so release slows and the amount released grows only with the square root of time (Chapter 4). A material that breaks down starts to release substances that were not in it before (Chapter 5), and a test of the material as made cannot find a substance that forms only in use. The firm's assessment of December 2020 rested in part on the toxicological significance of such substances: the "foam degradants", the products of the foam's breakdown.
Words with a narrower sense
Several words in this field look ordinary and carry a narrower sense, and a reader who takes one of them in its everyday sense misreads the argument that uses it. Medical grade is the clearest case. On a resin it sounds like a status that a regulator grants, but Chapter 3 shows that it records choices made by the resin's supplier. Each word in the table is explained in the text where it first matters, and the glossary at the end gives every term as this book uses it.
| Word | Its sense in this guide | Explained in |
|---|---|---|
| biocompatible | a judgment about a device in a particular use, not a property of its material | Chapter 1 |
| inert | in a supplier's or a paper's use, a claim that a material releases too little to matter; no material is inert in the everyday sense | above |
| contact | the kind of tissue a part touches and for how long, with repeated use added up | Chapter 2 |
| endpoint | one biological effect that the evaluation must address | Chapter 2 |
| medical grade | a supplier's description of a resin, not a status a regulator grants | Chapter 3 |
| extractable, leachable | what can come out under laboratory conditions harsher than use; what does come out in use | Chapter 4 |
| residue | whatever a process leaves in or on a part | Chapters 3 and 6 |
| residual | what is left of the sterilant in a device after sterilization with ethylene oxide | Chapter 7 |
| sensitization, irritation | immune memory that lets a later small dose cause an allergy; local damage that depends on the dose each time | Chapter 11 |
| systemic | acting away from the contact site, after absorption | Chapter 11 |
| worst case, simulated use, exaggerated extraction | test conditions, each chosen for its own purpose | Chapter 12 |
| dose, exposure | the amount that reaches a patient, set against the patient's size and the time; how much reaches the patient over the contact | Chapters 14 and 15 |
| threshold | a dose below which harm is judged very unlikely | Chapters 14 and 16 |
| cytotoxicity | harm to cells grown in a dish, from a material's extract | Chapter 18 |
| VOC, particulate | volatile organic compounds and particles that a device adds to the air a patient breathes | Chapter 19 |
How to read this guide
The nine parts follow one chain, from a material to what leaves it or what its surface does, then to the route into the body, the dose and the body's response. Part I, The question, explains why the same material can be safe in one use and harmful in another, and sorts what each part of a device touches and for how long. Part II, Materials, explains what a medical plastic contains besides its polymer, why molecules leave it, and how water, heat, oxygen and ozone break it down. Part III, Resin to device, follows the material through molding, bonding, cleaning, sterilization, packaging and years of use, each of which can change what it releases.
Part IV, The body, describes the body's side: the proteins that settle on a surface before any cell arrives, and the response of tissue to an implant or a particle. It then turns to why contact with blood needs its own questions, and to the allergy, irritation and harm at a distance that a small dose can cause. Part IV ends with Chapter 11, which completes the first half of the answer: why a material can harm a patient. Part V, Chemistry first, follows the method in use today, from what comes out of a material to the dose that reaches the patient and the dose that is tolerable. Chapters 17 and 18 complete the second half: how low a laboratory must look, and what only a living system can show.
Part VI, Breathing gas, applies the chain to the air that a breathing machine delivers, and Chapter 20 tells the foam recall from the record, as the case in which prediction failed. Part VII, The system, places the evaluation in the design file: how the method grew from a checklist of tests into a risk estimate, why a change at a supplier or a maker reopens it, and how to read one. Part VIII, The field in 2026, describes who makes the materials, who tests them and who writes the evaluations. Part IX, Regulation, sets out what the United States, the European Union and India require, as of November 2026.
Three kinds of reader come to the subject most often, and each has a shorter route through the book. Every route begins in Part I, because the chain it sets out runs through every later part.
| Reader | Needs most | Parts | Shortest path |
|---|---|---|---|
| An electronics or mechanical engineer new to materials and biology | why a material releases things and how the body answers; how to choose and process materials | I to V, then VI | Chapters 1, 3 to 5, 8, 12 to 17, 19 |
| A founder or product lead choosing materials and a test laboratory | what drives cost and delay; what a supplier's certificate does not cover; when a change reopens the evaluation | I, II, III, VI, VII, VIII | Chapters 1, 3, 7, 20, 22, 23, 24, 25 |
| A regulatory affairs or quality professional writing a biological evaluation plan | the method of the 2025 edition of ISO 10993-1; the chemistry-first arithmetic; reading reports; the three jurisdictions | I, V, VII, IX | Chapters 2, 17, 21, 22, 24, 26 to 28 |
Expect a change of material type, formulation, processing or supplier to be analyzed even when the material stays within its specification. Such a change reopens the evaluation when it can alter what reaches the patient, and it can require a new premarket notification to FDA (Chapter 23). Budget for samples of the finished device in its worst case, tested new and again after aging and cleaning as the instructions set (Chapters 12 and 19). Count the animal tests: for the breathing machine's mask cushion, a checklist reading gives at least five tests, two of them in animals, and the risk estimate gives three, one in animals. The risk estimate adds the genotoxicity battery's two tests only if the chemistry leaves a compound above its threshold (Chapters 18 and 21). Check a laboratory's report for three of the deficiencies FDA staff list as most common in extraction: unclear extraction conditions, manipulated extracts and unclear calculations (Chapter 24). Write the file to the 2025 edition of ISO 10993-1 unless it reaches FDA by 1 July 2029, the last date for declarations to the 2018 edition (Chapter 26).
Four kinds of box recur, and none carries the explanation on its own: a reader who skips every box still gets the whole argument from the prose.
A blue edge marks the structural point of a section, or a worked calculation. Every chapter has a solved example, with real values from its sources, each step shown and its assumptions stated.
An orange edge marks a common misreading, a trap, or the point at which a claim stops holding.
A green edge maps the idea onto design, the bench, verification or a regulatory file.
- Four sentences close each chapter with what it established, and none says more than the chapter does.
Most of the arithmetic stays within the ceiling that this series of guides keeps: doubling, powers of ten, percentages, ratios and squares, with logarithms only as tenfold steps. Dose arithmetic needs more, so this book raises the ceiling for it to amounts in micrograms a day and in milligrams for each kilogram of body weight, margins of safety as ratios, and uncertainty factors multiplied together. Each worked number appears once, in a box, with real values and stated assumptions, and the prose uses no algebra.
The chemistry is school chemistry plus polymers as chains, additives held between the chains, and the bonds that water, heat, ozone or radiation break, named as ester, urethane, ether and siloxane bonds. Structures appear only as simplified drawings, and no reaction mechanism is drawn. The biology starts from cells, proteins and the immune system at school level, and Part IV builds it up.
Inside the plates, color is a legend and never decoration, and five classes of thing keep one color each throughout the book. Violet marks the device's own material: its polymer and its parts. Gold marks what can leave the material, an additive, a molecule or a particle, and red marks the body's side: its tissue, cells and proteins. Teal marks the route a substance travels, breathed gas, blood or saline, and magenta marks a limit or threshold, such as a tolerable dose or concentration. Blue marks every result read off a plate, a curve or a value, and one orange mark in each plate points at the detail that matters most. A key strip under every plate lists only the colors that plate uses.
This book is a reference for understanding, not a test protocol, a regulatory opinion or clinical advice. Regulatory status, standards and the companies of Part VIII are stated as of November 2026 and will date. A figure, test result or statement that comes from a manufacturer is labeled as the manufacturer's where the text uses it. The record of the Philips Respironics case is given as it stood in November 2026. A Form 483 lists observations and is not a final FDA determination. FDA notes that the cause of an event reported to it cannot typically be determined from those reports alone.
Let's Talk