Indian standards come first.
+ The questionHow does CDSCO's framework handle biological safety, as of November 2026?
A path from class to license
India's Medical Devices Rules send a maker first to Indian standards, and to international ones where none exists. The Ministry of Health and Family Welfare notified the Rules on 31 January 2017, and they have been in force since 1 January 2018. They set a path in four steps. First the device is classed and the standards it must conform to are chosen. Its tests are then run in laboratories that hold the right recognitions, and its maker or importer applies for a license or, for some devices, a registration.
The class decides which authority licenses the device. Rule 4 classes devices other than in vitro diagnostics by Part I of the Rules' First Schedule, from class A, the lowest risk, to class D, the highest. A State Licensing Authority licenses the manufacture of class A and B devices, on an application in Form MD-3 for a license in Form MD-5. The Central Licensing Authority licenses the manufacture of class C and D devices, in Forms MD-7 and MD-9, and the import of every class, in Forms MD-14 and MD-15. An amendment of 14 October 2022 exempts class A devices that are neither sterile nor measuring from the license, and they are registered instead.
India's essential principles ask of a device's materials what Chapters 4 and 14 measure. Materials are to be chosen with regard to "toxicity, biodegradability" and to their compatibility with the tissues and cells they meet, and "substances that may leach or leak from the device" are to be addressed. "Special attention shall be given to substances which are carcinogenic, mutagenic or toxic to reproduction." Unlike the EU rule of Chapter 27, the principles set no concentration threshold and cite no ISO 10993 part, and they leave the choice of standard to Rule 7. They are a Ministry guidance of July 2017, and the file that carries them is labeled a draft enclosure. The Rules' Fourth Schedule sets out the application's device master file.
Rule 7 orders the standards in three tiers
Rule 7 decides which standards a device must conform to, in three tiers. The first tier holds the standards of the Bureau of Indian Standards, BIS, or those the Ministry notifies "from time to time". Where no relevant Indian standard exists, the device conforms to a standard of ISO or the IEC, or to a pharmacopoeia standard. Where neither exists, "the device shall conform to the validated manufacturer's standards". The order matters wherever an Indian standard differs from the international edition a maker tested to, because the Indian standard comes first.
For biological evaluation, India's first tier is mostly ISO 10993 adopted unchanged. Most of its parts are published as identical Indian standards, among them part 5 on cytotoxicity in its 2009 edition, part 17 in its 2002 edition and part 18 in its 2020 edition, as the list of BIS's committee MHD 19 stood on 13 November 2026. The evaluation itself and several tests sit in a separate series, IS 17932, which BIS lists as modified or technically equivalent rather than identical. Its Part 1:2023 is a modified adoption of ISO 10993-1:2018, and its Parts 6 and 7 modify the 2021 editions of ISO 10993-10, on sensitization, and ISO 10993-23, on irritation. The list marks these parts as modified and does not say what the modifications change.
India's first tier therefore rests on older editions than some that this book used. No Indian adoption of ISO 10993-1:2025 is listed as published, so the Indian standard that comes first for the evaluation is the 2018 edition as India modified it. Chapters 26 and 27 set out how FDA and the EU treat the 2025 edition. India's part 17 is the 2002 edition, while Chapter 14 built its tolerable intake on the 2023 edition. Part 7 of IS 17932 rests on ISO 10993-23 as published in 2021, before its 2025 amendment added further reconstructed human epidermis models (Chapter 18). ISO 18562, the standard for the gas pathway, is not on the committee's list, which holds only MHD 19's own standards.
A laboratory in India holds three recognitions
A test laboratory in India can hold three separate recognitions, and each decides a different thing for a file. Registration with CDSCO decides whether the laboratory may test devices on a maker's behalf under the Rules. The Rules define a medical devices testing laboratory as one "registered under sub-rule (3) of rule 83", and CDSCO registers such laboratories on Form MD-40. As of 13 November 2026, the latest list of registered laboratories that CDSCO's page links is dated 15 September 2023. Government laboratories are notified separately, and a notification of 1 June 2018 named the Central Medical Device Testing Laboratories.
Accreditation by NABL decides whether the laboratory is competent for each method on its scope. Chapter 25 gives that question to ISO/IEC 17025, and NABL accredits Indian laboratories to that standard. At least one provision of the Rules ties testing to the accreditation, and a notification of 16 October 2019 exempted State and Central Government testing laboratories from NABL accreditation for two years. Registration on Form MD-40 and NABL accreditation are separate recognitions, so a laboratory's accreditation for a biological test is checked on its own NABL scope, method by method.
GLP certification decides whether a facility's safety studies carry GLP status, and with it whether their data are accepted beyond India. The OECD's Mutual Acceptance of Data in GLP is a scheme whose purpose is that GLP data from a certified facility are accepted in the OECD's member countries, and India became a full adherent to it on 3 March 2011. Certification in India comes from the National GLP Compliance Monitoring Authority, NGCMA, which works under the Department of Science and Technology. NGCMA certifies against the OECD Principles of GLP, which its application form cites as documents 1 to 22. Each certificate is scoped by area of expertise, category of test item and type of test system, such as animals or cell lines. A GLP certificate is not a NABL accreditation, since each answers its own question (Chapter 25).
The running device's Indian file
This book's Indian file for the running device is an illustration, not any maker's file, and it keeps the evaluation plan of Chapter 22. CDSCO's device-wise classification list places the CPAP and BPAP nasal and oral masks and the heated respiratory humidifier in class B. The same list places a "Respiratory CPAP Device" in class B too: a portable, powered device for continuous positive airway pressure that treats mainly adults with obstructive sleep apnea, "intended for use in homecare or health care settings". The running device's machine is therefore a class B device, and a State Licensing Authority licenses its manufacture. What changes from the US file of Chapter 26 is the standard each document cites.
- The documents. The device's evaluation; the cushion's chemistry, toxicological risk assessment and three tests; and the foam's gas-pathway evaluation: 7 documents.
- Tier 1, the edition this book used or no edition named. Chemistry, IS/ISO 10993-18:2020; cytotoxicity, IS/ISO 10993-5:2009. That makes 2.
- Tier 1, older or modified. The evaluation, IS 17932 (Part 1):2023; the risk assessment, part 17 of 2002; irritation, IS 17932 (Part 7):2024; sensitization, IS 17932 (Part 6):2023. That makes 4.
- Tier 2. The foam, ISO 18562, assuming no Indian standard exists, since the committee's list holds none. That makes 1.
- Tier 3. None.
- The result. 6 of the 7 documents cite an Indian standard first, and 4 of those 6 need a statement of how work done to another edition meets the Indian one. That statement is this book's practice, since Rule 7 asks for conformity.
For the laboratories, what changes is what the file checks. For each laboratory it records its MD-40 registration, its NABL scope for each method and its NGCMA certificate, whose test systems should include cell lines for cytotoxicity and animals for the local lymph node assay. Each recognition is recorded with the date of the list it was checked against. Recording all three is this book's choice, since each recognition decides a different thing for the file.
Status as of November 2026
India's documents each carry their own date, and the dates lie years apart. As of 13 November 2026, CDSCO's rules page lists a notification of 2 June 2023, on notifying State testing laboratories, as the latest amendment to the Rules. BIS's committee list of 13 November 2026 shows no Indian adoption of ISO 10993-1:2025, of ISO 10993-17:2023 or of the 2026 editions of parts 3, 6 and 7 as published. NGCMA's list of GLP-certified test facilities states that it was last updated on 12 May 2026. The certificates on that list run about three years by their dates, and some end before November 2026, so the list does not show whether those certificates were renewed.
Part IX has set one device and one evaluation plan against three jurisdictions, which differ in the edition a file cites, the recognitions its laboratories hold and the authority that reviews it. The closing pages return to the record with which this book opened.
The Form 483 dates each kind of evidence
This book opened at Murrysville, on the Form 483 that FDA's investigator issued on 9 November 2021, a list of observations and not a final determination. The form lists the firm's own reports and assessments, dated from April 2016 to January 2021, and that list holds each kind of evidence that this book describes. The earliest examine the foam and what heat and humidity did to it (Chapter 5), and the emissions reports of January 2020 measure what the machine gave off into the breathed gas (Chapter 19). A risk assessment weighs compounds of concern for patients of 30 kg and 70 kg (Chapter 15), and the laboratory reports of January 2021 test the foam or its extracts on cells and for mutations (Chapters 11 and 18). The reports name the field's own standards: a biological risk assessment of December 2020 follows ISO 10993-1:2018, and a cytotoxicity report of January 2021 judges its result against ISO 10993-5.
The same list dates the firm's findings and its decisions. A report of August 2016 found that polyester urethanes "show bad resistance against high humidity in combination with high temperature", the attack that Chapter 5 traces to water cutting the ester link. In 2018 a field correction went out, and FDA observed that it was neither verified for its effectiveness nor reported in writing to FDA. An email of the same year recorded the decision to keep the polyester foam in the Trilogy platform after testing confirmed its breakdown in high heat and high humidity. A health hazard evaluation of that year assumed a humidifier, a bacterial filter and a healthy lung, while the form quotes the ventilators' intended use as the care of "individuals who require mechanical ventilation" (Chapters 19 and 20). The recall followed in June 2021, and the form gives its cause as "polyester polyurethane foam degradation and volatile organic compound emission".
The silicone foam that replaced the polyester has its own record. An A Series CPAP device containing silicone foam failed testing for VOCs in a test report of 24 August 2021, which the form records. FDA's foam-testing page tells of a "similar device marketed outside the U.S., which failed one safety test for the release of VOCs", and states that "devices authorized for marketing in the U.S. had demonstrated acceptable results". FDA requested further testing of the silicone foam by an independent laboratory, and as of 13 November 2026 no result of the independent testing was posted on that page (Chapter 20).
A plastic that is safe in a water bottle has been judged for one use, what it releases into water that a person then drinks (Chapter 1). In a breathing circuit the same plastic meets warm, humid air every night for years, and what leaves it reaches the lungs directly. In a catheter it meets blood, which draws out fat-loving additives and coats the surface with proteins (Chapters 4 and 8). A change of use therefore changes how much leaves the material, by which route it reaches the body and at what dose, and the dose is what the body answers.
To see harm from a device's materials before patients meet it, a laboratory tests the finished device as it will be used, aged and cleaned where use changes it. The laboratory names and counts what comes out, down to a reporting line set from a daily threshold of toxicological concern. The evaluator turns each amount into a dose for the lightest patient over the whole contact, and tests in cells or animals cover the effects that the thresholds do not, such as sensitization (Chapters 12 to 18).
The running device's foam shows what a reader can now do with a material choice. This book chose a polyether polyurethane for it, to keep ester links out of warm, humid air (Chapter 5). The polyether resists hydrolysis but is weak to oxidants, ozone from a cleaner among them, so the foam is still evaluated as aged, cleaned and used (Chapters 5 and 7). The foam sits in the gas pathway, so its evaluation runs on the delivered gas under ISO 18562 (Chapter 19). A later change of foam would reopen that evaluation, as the silicone foam that replaced the recalled polyester foam reopened the gas pathway's questions (Chapter 23).
Any other material choice goes through the same questions, in the chain's order: the contact (Chapter 2), the recipe and what processing and use make of it (Chapters 3, 6 and 7), and then the release, route, dose and tests of Part V. The answers enter the design file as inputs. A change at a supplier or a maker reopens them (Chapters 22 and 23), and each jurisdiction asks for them against the edition of each standard that it accepts (Chapters 26 to 28).
No material is inert at the scale the body works on, so a material's name, its grade or its certificate settles none of these questions (Chapter 3). An evaluation can still end with its deciding row open, as the running device's illustrative report did on its unknown peak (Chapter 24). Naming that row, with the analysis or the test that would close it, is what carries a result into a decision about the device's risk, the step with which this book's opening ended.
For a file in India, take the device's class from the First Schedule and CDSCO's device-wise classification list before naming the licensing authority, and note that non-sterile, non-measuring class A devices are registered rather than licensed. Set each standard the plan names beside the Indian standard that Rule 7 puts first, with its edition and whether BIS lists it as identical or modified. Check each laboratory's three recognitions on the day of the study.
- India's Medical Devices Rules class devices from A to D, and the class decides which authority licenses their manufacture, while non-sterile, non-measuring class A devices are registered instead.
- Rule 7 puts Indian standards first, then ISO, IEC or pharmacopoeia standards, then the manufacturer's validated standards, and India's standard for the evaluation is ISO 10993-1:2018 as modified in 2023.
- A test laboratory in India can hold CDSCO registration on Form MD-40, NABL accreditation to ISO/IEC 17025 and NGCMA certification for GLP, three recognitions that each decide something different for a file.
- As of November 2026, the running device's illustrative Indian file holds 7 documents, and 4 of them rest on work done to a different edition from the Indian standard that Rule 7 puts first. Each of those 4 needs a statement of how that work meets the Indian standard. The Form 483 with which this book opened lists the firm's own reports and assessments, and they hold each kind of evidence the chapters describe, from the earliest report on the foam that the form lists, of April 2016, to the laboratory reports of January 2021.
Let's Talk