Essential Drug Delivery

Outputs for Devices Intended to

Deliver Drugs and Biological
Products
Guidance for Industry
DRAFT GUIDANCE
This guidance document is being distributed for comment purposes only.

Comments and suggestions regarding this draft document should be submitted within 90 days of
publication in the Federal Register of the notice announcing the availability of the draft
guidance. Submit electronic comments to https://www.regulations.gov. Submit written
comments to the Dockets Management Staff (HFA-305), Food and Drug Administration, 5630
Fishers Lane, Rm. 1061, Rockville, MD 20852. All comments should be identified with the
docket number listed in the notice of availability that publishes in the Federal Register.

Additional copies are available from:

Office of Combination Products, Food and Drug Administration
WO32, Hub/Mail Room #5129 10903 New Hampshire Avenue Silver Spring, MD 20993
(Tel) 301-796-8930, (Fax) 301-847-8619; https://www.fda.gov/combination-products

For questions regarding this draft document, contact the Office of Combination Products,
Patricia Love at [email protected] or [email protected].

U.S. Department of Health and Human Services

Food and Drug Administration
Office of Combination Products
Center for Biologics Evaluation and Research (CBER)
Center for Devices and Radiological Health (CDRH)
Center for Drug Evaluation and Research (CDER)

June 2024
Combination Products
 Contains Nonbinding Recommendations
Draft — Not for Implementation
TABLE OF CONTENTS

I. INTRODUCTION............................................................................................................. 1
II. SCOPE AND DEFINITION ............................................................................................ 2
III. BACKGROUND ............................................................................................................... 3
IV. OVERVIEW OF ESSENTIAL DRUG DELIVERY OUTPUT PROCESSES ........... 5
V. IDENTIFYING ESSENTIAL DRUG DELIVERY OUTPUTS ................................... 5
VI. VERIFYING AND VALIDATING ESSENTIAL DRUG DELIVERY OUTPUTS ... 8
A. Design Verification for EDDOs .................................................................................................... 8
1. Preconditioning ............................................................................................................................... 8
2. Design Verification Testing ........................................................................................................... 10

B. Design Validation for EDDOs ..................................................................................................... 11

1. Additional Validation Considerations for ANDA Submissions and BLAs Submitted under Section
351(k) of the PHS Act for Combination Products .......................................................................... 12
2. Additional Validation Considerations for Premarket Notifications [510(k)] ................................ 13

VII. CONTROL STRATEGIES FOR ESSENTIAL DRUG DELIVERY OUTPUTS .... 13

VIII. INFORMATION TO PROVIDE IN APPLICATIONS .............................................. 14
A. IND and IDE Applications .......................................................................................................... 15
B. Marketing Applications ............................................................................................................... 17
C. Submissions for Post-Market Changes That May Impact Essential Drug Delivery Outputs
…………………………………………………………………………………………………….20
IX. INTERACTION WITH FDA ........................................................................................ 21
APPENDIX A – ESSENTIAL DRUG DELIVERY OUTPUT IDENTIFICATION
EXAMPLE – PREFILLED SYRINGE .................................................................................... 23
APPENDIX B – ESSENTIAL DRUG DELIVERY OUTPUT IDENTIFICATION
EXAMPLE – AUTOINJECTOR .............................................................................................. 25
APPENDIX C – EXAMPLES OF POTENTIAL ESSENTIAL DRUG DELIVERY
OUTPUTS BASED ON PRODUCT TYPE .............................................................................. 27
APPENDIX D – EXAMPLE OF A CONTROL STRATEGY FOR THE ESSENTIAL
DRUG DELIVERY OUTPUT OF NEEDLE EXTENSION LENGTH FOR AN
AUTOINJECTOR ...................................................................................................................... 34
 Contains Nonbinding Recommendations
Draft — Not for Implementation

1 Essential Drug Delivery Outputs for Devices Intended to Deliver

2 Drugs and Biological Products
3 Guidance for Industry
4
5
6 This draft guidance, when finalized, will represent the current thinking of the Food and Drug
7 Administration (FDA or Agency) on this topic. It does not establish any rights for any person and is not
8 binding on FDA or the public. You can use an alternative approach if it satisfies the requirements of the
9 applicable statutes and regulations. To discuss an alternative approach, contact the FDA staff responsible
10 for this guidance as listed on the title page.
11
12
13
14
15 I. INTRODUCTION
16
17 This guidance addresses key aspects of drug delivery performance information for devices, 1 and
18 combination products 2, 3 that include device constituent parts, 4 intended for delivery of a human
19 drug, including a biological product 5 (herein referred to as drug delivery devices). 6 The
20 guidance describes FDA’s recommendations related to the device design outputs that are
21 essential for establishing and assessing drug delivery performance. The guidance includes
22 recommendations for the information and data to submit in investigational, marketing, and post-
23 market change applications. 7 Generally, as discussed further in this guidance, essential drug
24 delivery output (EDDO) refers to the device drug-delivery design outputs necessary to ensure the
25 drug delivery function. 8 This guidance recommends an approach to identifying EDDOs,
26 provides examples of EDDOs for specific types of devices, and describes the information and
27 data related to EDDOs provided in an application.

1
The term device refers to a device as defined in section 201(h)(1) of the Federal Food, Drug, and Cosmetic
(FD&C) Act (21 U.S.C. 321(h)(1)).
2
See 21 CFR 3.2(e) for the definition of combination product.
3
For the purpose of this guidance, the term product is used to refer to both stand-alone devices (i.e., not part of a
combination product) and combination products.
4
See 21 CFR 4.2 for the definition of constituent part. For the purpose of this guidance, the terms device and device
constituent part are used interchangeably.
5
The term drug refers to a drug as defined in section 201(g)(1) of the FD&C Act (21 U.S.C. 321(g)(1)), and
includes biological products as defined in section 351(i) of the Public Health Service (PHS) Act (42 U.S.C. 262(i)).
6
For the purpose of this guidance, the terms drug and drug constituent part are used interchangeably.
7
For the purpose of this guidance, unless otherwise stated, the terms applications and submissions are used
interchangeably and include, as applicable, initial, supplements to, and amendments to: investigational new drug
applications (IND), new drug applications (NDA), abbreviated new drug applications (ANDA), investigational
device exemption (IDE) applications, premarket approval applications (PMA), De Novo requests submitted under
section 513(f)(2) of the FD&C Act, premarket notifications (510(k) submission), and biologics license applications
(BLA), including BLAs submitted under section 351(k) of the PHS Act.
8
Prior to this guidance, the term essential performance requirements (EPR) was generally used in communications
between FDA and applicants for the EDDOs described herein. FDA is now using the term EDDO as we believe it is
more descriptive.

1
 Contains Nonbinding Recommendations
Draft — Not for Implementation
28 Device drug-delivery performance information is intended to demonstrate that the device drug-
29 delivery function consistently performs as intended. FDA is providing recommendations for
30 development and organization of this information and these data to improve the consistency of
31 this information in applications. Ultimately, these recommendations are intended to facilitate
32 and streamline development of drug delivery devices.
33
34 This guidance does not address all of the data and information to be submitted in support of drug
35 delivery devices (e.g., it does not address drug-device compatibility, biocompatibility, sterility,
36 human factors, electrical safety, electromagnetic compatibility, radio frequency wireless
37 technology, or cybersecurity). Applicants should refer to applicable regulations and guidance for
38 more information on what is required or recommended to be submitted to FDA in applications
39 for such products. 9
40
41 In general, FDA’s guidance documents do not establish legally enforceable responsibilities.
42 Instead, guidances describe the Agency’s current thinking on a topic and should be viewed only
43 as recommendations, unless specific regulatory or statutory requirements are cited. The use of
44 the word should in Agency guidances means that something is suggested or recommended, but
45 not required.
46
47
48 II. SCOPE AND DEFINITION
49
50 Scope: The focus of this guidance is the information and data developed and submitted to FDA
51 regarding EDDOs for devices and device constituent parts of CDER-led and CBER-led
52 combination products 10 intended for delivery of a human drug, including a biological product. 11
53 Examples of products that are within the scope of this guidance include syringes, injectors (e.g.,
54 autoinjector, on body injector), infusion products (e.g., infusion pumps), nasal sprays, inhalers,
55 nebulizers, and vaginal systems.
56
57 Definition: EDDOs are the design outputs necessary to ensure delivery of the intended drug
58 dose to the intended delivery site. 12 Drug delivery includes successful product preparation 13 and

9
For additional guidance related to product development, see the FDA Guidance webpage to search for an FDA
guidance document or to browse FDA guidance documents by topic. Information on these considerations may be
found in applicable guidances. For example, for more information regarding injector delivery devices, see the
guidance for industry and FDA staff Technical Considerations for Pen, Jet, and Related Injectors Intended for Use
with Drugs and Biological Products (June 2013). We update guidances periodically. For the most recent version of
a guidance, check the FDA guidance web page at https://www.fda.gov/regulatory-information/search-fda-guidance-
documents.
10
CDER-led and CBER-led combination products refer to combination products for which CDER or CBER has
primary jurisdiction (i.e., is the lead Center). For information regarding Center assignment, see the guidance for
industry and FDA staff Principles of Premarket Pathways for Combination Products (January 2022).
11
If an applicant has a question about whether the concepts in the guidance apply to a specific product, contact the
applicable review division in the lead Center.
12
See 21 CFR 820.30(d), which states that “[e]ach manufacturer shall establish and maintain procedures for
defining and documenting design output in terms that allow an adequate evaluation of conformance to design input
requirements. Design output procedures shall contain or make reference to acceptance criteria and shall ensure that
those design outputs that are essential for the proper functioning of the device are identified.”
13
See Appendices A and C for examples related to product preparation.

2
 Contains Nonbinding Recommendations
Draft — Not for Implementation
59 the initiation, progression, and completion of dose delivery. 14 EDDOs are system level outputs
60 for which device drug-delivery function is dependent on the device design (see section V for
61 more information).
62
63
64 III. BACKGROUND
65
66 Design control requirements include establishing and maintaining procedures relating to design
67 inputs 15 and design outputs 16 (21 CFR 820.30(c) and (d)). 17, 18 The design must be verified and
68 validated throughout the lifecycle of the product as needed (see 21 CFR 820.30(f) and (g)). 19
69 The design outputs are driven by the design inputs that address the intended use of the device (21
70 CFR 820.30(c)). EDDOs are a subset of design outputs. EDDOs are part of the information that
71 is “essential for the proper functioning of the device” to deliver the drug (i.e., the intended use of
72 the drug delivery device) (21 CFR 820.30(d) (emphasis added)). In accordance with this
73 provision, manufacturers shall ensure that EDDOs are identified and approved before release (21
74 CFR 820.30(d)). In addition, to ensure the quality of the drug delivery function throughout the
75 product lifecycle, control and maintenance of the EDDOs are also necessary. 20
76

14
Drug and drug constituent part attributes may impact appropriate drug delivery.
15
Design input means the physical and performance requirements of a device that are used as a basis for device
design (21 CFR 820.3(f)).
16
Design output means the results of a design effort at each design phase and at the end of the total design effort.
The finished design output is the basis for the device master record. The total finished design output consists of the
device, its packaging and labeling, and the device master record (21 CFR 820.3(g)).
17
21 CFR 820.30(c) states that “[e]ach manufacturer shall establish and maintain procedures to ensure that the
design requirements relating to a device are appropriate and address the intended use of the device, including the
needs of the user and patient.” 21 CFR 820.30(d) states that “[e]ach manufacturer shall establish and maintain
procedures for defining and documenting design output in terms that allow an adequate evaluation of conformance
to design input requirements.”
18
On February 2, 2024, FDA issued a final rule amending the device quality system (QS) regulation, 21 CFR part
820, to align more closely with international consensus standards for devices and making conforming amendments
to 21 CFR part 4 (89 FR 7496, available at https://www.federalregister.gov/d/2024-01709). This final rule will take
effect on February 2, 2026. Once in effect, this rule will withdraw the majority of the current requirements in part
820 and incorporate by reference the 2016 edition of the International Organization for Standardization (ISO)
13485, Medical devices – Quality management systems – Requirements for regulatory purposes, in part 820. As
stated in the final rule, the requirements in ISO 13485 are, when taken in totality, substantially similar to the
requirements of the current part 820, providing a similar level of assurance in a firm’s quality management system
and ability to consistently manufacture devices that are safe and effective and otherwise in compliance with the
FD&C Act. When the final rule takes effect, FDA will also update the references to provisions in 21 CFR part 820
in this guidance to be consistent with that rule.
19
21 CFR 820.30(f) states that “[e]ach manufacturer shall establish and maintain procedures for verifying the device
design. Design verification shall confirm that the design output meets the design input requirements. The results of
the design verification, including identification of the design, method(s), the date, and the individual(s) performing
the verification, shall be documented in the DHF.” 21 CFR 820.30(g) states that “[e]ach manufacturer shall
establish and maintain procedures for validating the device design. Design validation shall be performed under
defined operating conditions on initial production units, lots, or batches, or their equivalents. Design validation shall
ensure that devices conform to defined user needs and intended uses and shall include testing of production units
under actual or simulated use conditions.”
20
See 21 CFR parts 820 and 4.

3
 Contains Nonbinding Recommendations
Draft — Not for Implementation
77 EDDO-related information, including verification and validation data, is provided in
78 investigational and marketing applications for drug delivery devices and combination products
79 with drug delivery devices to demonstrate that the drug delivery device appropriately delivers the
80 intended drug dose to the intended delivery site. 21, 22 In addition to being part of design control
81 activities, the EDDO processes discussed in this guidance can also be used for defining a control
82 strategy.
83
84 Drug-device combination products may be more complex than their individual constituent parts
85 because, in addition to the individual constituent parts, the interactions of the constituent parts
86 also need to be assessed, characterized, and controlled during product design, development, and
87 production. The final rule Current Good Manufacturing Practice Requirements for Combination
88 Products, codified at 21 CFR part 4 (CGMP requirements), 23 provided clarity regarding the
89 applicability of CGMP requirements to combination products, along with a streamlined
90 regulatory framework for demonstrating compliance with applicable requirements. Core
91 requirements in the CGMP regulations provide for systems that assure proper design,
92 monitoring, and control of manufacturing processes and facilities. The guidance for industry and
93 FDA staff: Current Good Manufacturing Practice Requirements for Combination Products (part
94 4 CGMP guidance) 24 further describes and explains the manufacturing requirements and
95 regulatory framework for combination products. It includes a detailed discussion of design
96 controls, including design inputs and design outputs for a combination product (section IV.A.2 of
97 the part 4 CGMP guidance). As described in the CGMP requirements and part 4 CGMP
98 guidance, the design control requirements at 21 CFR 820.30 apply as appropriate to combination
99 products with device constituent parts (see 21 CFR 4.4(b)(1)(ii) and (2)).
100
101 For combination products, FDA acknowledges that the terms design output (see 21 CFR
102 820.3(g) and 820.30(d)) and EDDO could be interpreted as analogous to the ICH guidance for
103 industry Q8(R2) Pharmaceutical Development (November 2009) drug product terminology for a
104 critical quality attribute (CQA), “[a] physical, chemical, biological, or microbiological property
105 or characteristic that should be within an appropriate limit, range, or distribution to ensure the
106 desired product quality.” 25 Similar to the CQA concept, as noted above, EDDOs are essential for
107 the appropriate functioning of the device, and in some instances, an applicant could expand
108 CQAs to include device drug-delivery function features for a combination product. Likewise, a
109 quality target product profile (QTPP), 26 which is similar to design inputs (see 21 CFR 820.3(f)
110 and 820.30(c)), may assist an applicant in identifying CQAs, including those for drug delivery.
111 As appropriate, studies conducted to verify that the CQAs are met may address EDDO

21
These data are typically distinct from the manufacturing information documented and available for review as part
of an establishment assessment or inspection to establish compliance with CGMP and QS requirements, as
applicable.
22
The data needed to support approval/clearance of a particular product depends on the application type.
23
78 FR 4307 (January 22, 2013).
24
January 2017.
25
See ICH Q8(R2) annex glossary at p. 18.
26
A QTPP is “[a] prospective summary of the quality characteristics of a drug product that ideally will be achieved
to ensure the desired quality, taking into account safety and efficacy of the drug product.” ICH Q8(R2) annex
glossary at p. 18.

4
 Contains Nonbinding Recommendations
Draft — Not for Implementation
112 verification and validation (see 21 CFR 820.30(f) and (g)). 27 Applicants may be able to leverage
113 CQA information to support EDDO identification, control, and maintenance processes. When
114 the EDDO is amenable to verification and validation through analytical methods (see section
115 VI), the chemistry, manufacturing, and controls (CMC) information may address these design
116 control requirements.
117
118 EDDO identification, control, and maintenance processes also may facilitate development by:
119
120 • Informing the determination of which data to submit in an investigational or marketing
121 application to demonstrate the drug delivery performance;
122
123 • Ensuring the appropriate device design attributes and manufacturing process steps are
124 evaluated during lifecycle changes; and
125
126 • Providing a basis for comparing the drug delivery performance and facilitating
127 assessment of EDDOs for bridging or leveraging data across products. 28
128
129
130 IV. OVERVIEW OF ESSENTIAL DRUG DELIVERY OUTPUT PROCESSES
131
132 In developing EDDOs, there are three primary processes: identification, control, and
133 maintenance.
134
135 • Identification of the EDDO defines the device drug-delivery function of the product and
136 focuses design and development efforts to ensure appropriate drug delivery.
137
138 • Control of the EDDO ensures the product meets the device drug-delivery function
139 quality standards. See section VII for information on control strategy.
140
141 • Maintenance of the EDDO ensures that any changes to the product made during clinical
142 development or post-market that could adversely impact the EDDO are evaluated to help
143 preserve the quality of the drug-delivery function.
144
145 The following sections provide information on identifying EDDOs (section V), EDDO
146 verification and validation (section VI), and EDDO control strategies (section VII).
147
148
149 V. IDENTIFYING ESSENTIAL DRUG DELIVERY OUTPUTS
150
151 EDDOs can be identified from existing design controls by using a filtering process illustrated in
152 Figure 1 to identify specific design outputs.

27
For additional discussion of the relationship of drug development and device development terms related to design
control, see the part 4 CGMP guidance at section IV.A.2.
28
For additional information on bridging data across combination products, see the draft guidance for industry
Bridging for Drug-Device and Biologic-Device Combination Products (December 2019). When final, this guidance
will represent the FDA’s current thinking on this topic.

5
 Contains Nonbinding Recommendations
Draft — Not for Implementation
153 The EDDO identification process begins with all design outputs and then uses filtering steps to
154 eliminate outputs that do not meet the EDDO definition, which results in remaining outputs that
155 are the EDDOs. As noted in Section III, in some instances, the CQAs of a combination product
156 may include EDDOs, and thus, the below EDDO identification process may be addressed by the
157 determination of those CQAs.
158
159 Figure 1 – Illustration of the EDDO Identification Process
160
161 The figure illustrates, at a high level, the process for identification of EDDOs.
162

Design Outputs

Drug Delivery
Design Outputs

System Level Design

Outputs

Device
Dependent
Design Outputs

EDDOs

163
164
165 The following steps are useful in identifying EDDOs:
166
167 (1) Design Outputs – Begin by defining the proposed intended use, consider, e.g., the
168 indications for use, population, and condition and frequency of use, and design inputs
169 (e.g., user requirements, design specifications, route of administration, drug
170 characteristics, dosage form, and delivery volume). This information should be used to
171 identify the design outputs.
172
173 (2) Drug Delivery Design Outputs – Identify those design outputs related to the delivery of
174 the drug (e.g., related to the intended dose; delivery to target site; method of delivery;
175 product preparation; and the initiation, progression, and completion of dose delivery).
176
177 (3) System Level Design Outputs – Identify the drug delivery design outputs that are
178 system level design outputs (i.e., design outputs that are the functions necessary for the
179 performance of the final finished product). For more information, see the discussion
180 below following step 4 and in Figure 2.
181

6
 Contains Nonbinding Recommendations
Draft — Not for Implementation
182 (4) Device Dependent Design Outputs – Identify the system level drug delivery design
183 outputs that are independent of the user and dependent on the device design. This step is
184 to assure that the design and manufacture of the product are adequately controlled. (This
185 step is not intended to address usability because drug delivery performance that depends
186 on the user is not an EDDO).
187
188 As described in step 3, an EDDO is a system level design output. We note that there are other
189 design outputs known as component level outputs that are different from system level outputs.
190 Component level outputs work together to achieve a system level output and are not EDDOs.
191 Component level outputs support, but are subordinate to, system level outputs (see Figure 2).
192
193 Figure 2 – Example of System Level and Component Level Outputs (Step 3)
194
195 To help in identifying the EDDOs, Figure 2 illustrates the relationship between system level and
196 component level outputs for a prefilled syringe (PFS) that includes a needle, silicone oil as a
197 lubricant, and other components.
198
199 System Level Output Component Level Outputs
200
201

silicone quantity
silicone oil
silicone
distribution

needle length
Glide force needle
needle inner
diameter

component 1
etc
component 2
202
203
204 See Appendix A for a narrative illustration of the process concepts for identifying EDDOs for a
205 PFS. Appendix B illustrates the distinction between EDDOs and other design outputs for an
206 autoinjector. In addition, this document provides examples of design outputs for common
207 combination products with drug delivery devices that are likely to be considered EDDOs (see
208 Appendix C).
209
210

7
 Contains Nonbinding Recommendations
Draft — Not for Implementation
211 VI. VERIFYING AND VALIDATING ESSENTIAL DRUG DELIVERY OUTPUTS
212
213 While design verification 29, 30 and validation 31 activities are intended to address all design inputs
214 and outputs, the following recommendations are specific to EDDOs. Appropriate verification
215 and validation activities for EDDOs depend on the conditions (e.g., environmental conditions) to
216 which the product will be exposed during production, shipping, storage, and preparation and the
217 conditions associated with use. Examples of conditions that may impact performance include,
218 but are not limited to, temperature, pressure, humidity, vibration and shock, and physical
219 orientation. Also, during storage or shipping, a product may be exposed to more than one
220 variation of a condition or sequences of conditions (e.g., ground by truck to air, ground by truck
221 to boat, ground by rail to air to ground by truck). In addition to the storage and shipping
222 stressors, there are stressors associated with the use environments (e.g., health care facility,
223 school, home, first response environment). Verification and validation test reports provided in a
224 submission should provide information on how the tests conducted, including the conditions and
225 methods selected, are adequate to verify and validate the EDDOs.
226
227 Further, although the test methods used to verify and validate an EDDO are beyond the scope of
228 this guidance, the test methods (e.g., mechanical or analytical) may differ for active (e.g.,
229 autoinjector, metered dose inhaler) and passive (e.g., some implants, vaginal systems) drug
230 delivery devices. 32 See section VIII for details on verification and validation information
231 included in applications.
232
233 A. Design Verification for EDDOs
234
235 It is important that prior to initiation of any clinical studies (or any in vivo bioequivalence
236 studies, as applicable) or commercial distribution, applicants verify the performance of the
237 product. How applicants conduct design verification testing is dependent on device design,
238 intended use, and applicable regulations, standards, and guidances. Design verification generally
239 includes preconditioning and subsequent verification testing to confirm that the device drug-
240 delivery function is maintained in accordance with the instructions for use. The following
241 sections provide considerations when developing a design verification approach. If an applicant
242 intends to use an alternative design verification approach, we recommend providing an
243 explanation of and rationale for the alternative approach and requesting Agency feedback in a
244 formal meeting or communication with FDA (see section IX).
245
246 1. Preconditioning
247
248 During the product lifecycle, the product is exposed to multiple stressors that may influence
249 performance of the device drug-delivery function. Preconditioning is a method to simulate
250 exposing the product to stressors to which the product will likely be exposed during shipping,
29
Verification means “confirmation by examination and provision of objective evidence that specified requirements
have been fulfilled” (21 CFR 820.3(aa)).
30
Design verification shall confirm that the design output meets the design input requirements (21 CFR 820.30(f)).
31
Design validation means “establishing by objective evidence that device specifications conform with user needs
and intended use(s)” (21 CFR 820.3(z)(2)).
32
Different analytical methods, including assays, and other mechanical methods can be used or leveraged, as
applicable, for the specific product design.

8
 Contains Nonbinding Recommendations
Draft — Not for Implementation
251 storage, and use (e.g., cleaning, reprocessing, storage (see section VI.A.1.a), or repeat use, as
252 applicable). Applicants should identify preconditions applicable to the specific product, and
253 verification testing should assess the ability of the product to withstand those stressors.
254 Applicants should identify preconditions based on the risk of the product design, research and
255 development characterization testing, intended use, how the product will move from the finished
256 product manufacturer to the end user, and/or the conditions associated with use (see section
257 VI.A.1.c). Because of the risk to the patient should the device fail, sequential preconditioning is
258 generally expected for emergency-use injectors, 33 and applicants should identify the sequence in
259 which the preconditions should be applied. In developing the preconditioning methods,
260 applicants may leverage test methods, acceptance criteria, and statistical analysis techniques
261 from recognized standards.
262
263 a. Storage
264
265 Preconditions to simulate storage may include, for example, specific temperatures, temperature
266 fluctuations, pressure changes, and humidity. A product may be subjected to different storage
267 conditions throughout the product lifecycle (e.g., conditions at the manufacturing facility,
268 warehouse, health care facility, pharmacy, or home setting), and these should all be considered
269 before EDDO design verification.
270
271 b. Shipping
272
273 Preconditions to simulate shipping may include, for example, vibration, shock, and pressure
274 changes, that may impact device drug-delivery performance. Some EDDOs (e.g., physical
275 dimensions) would not be impacted by shipping and would not warrant preconditioning for
276 shipping conditions. For the EDDOs that may be influenced by shipping conditions, shipping
277 preconditioning representative of likely exposures (e.g., during shipping via air, ground by truck,
278 ground by rail, boat) should occur before EDDO design verification.
279
280 c. Other Conditions
281
282 In addition to the preconditions associated with storage and shipping, preconditioning may be
283 warranted to assess stressors related to conditions associated with use. For drug delivery devices
284 that do not have a recognized standard that includes preconditions or for which there may be
285 preconditions in addition to those in a recognized standard, preconditioning should be conducted
286 in the sequence of steps as specified in the instructions for use in the proposed labeling (e.g.,
287 storage and warm up time).
288
289 Additionally, depending on the device design and instructions for use, more than one
290 preconditioning method may be needed to account for potential failure modes. For example, for
291 infusion pumps, verifying the EDDO of flow rate accuracy should account for any before-use
292 instructions to precondition the pump and all aspects of the infusion pump system, including

33
For additional information regarding preconditioning, see section V.4.a of the draft guidance for industry and
FDA staff Technical Considerations for Demonstrating Reliability of Emergency-Use Injectors Submitted under a
BLA, NDA, or ANDA (April 2020). When final, this guidance will represent the FDA’s current thinking on this
topic.

9
 Contains Nonbinding Recommendations
Draft — Not for Implementation
293 accessories (e.g., infusion sets). Preconditions for the infusion pump and infusion pump
294 accessories may be different. Therefore, the various sequential preconditioning methods should
295 be applied as appropriate for the components of the infusion pump system and in accordance
296 with the proposed labeling.
297
298 For reusable devices, the preconditioning methods should simulate the worst-case number of
299 repeat use and reprocessing cycles. For example, preconditioning of reusable drug delivery
300 devices after reprocessing should include cleaning and sterilization or disinfection methods
301 identified in the proposed labeling. 34 As a second example, the labeling for some metered dose
302 inhalers calls for periodic cleaning of the actuator to prevent orifice blockage. Additionally,
303 labeling for metered dose inhaler and reusable nebulizer components often calls for periodic
304 cleaning to prevent contamination and/or changes in electrostatic properties to minimize capture
305 of small particles/droplets and any change in respirable drug.
306
307 2. Design Verification Testing
308
309 Overall, the design verification assessment of EDDOs should occur after appropriate
310 preconditioning. Depending on the product, the EDDO, and the specific types of preconditions,
311 there may be different types of testing to verify that an EDDO is maintained throughout the
312 range of conditions and use environments described in proposed labeling. This testing may
313 include functional testing after preconditioning or it may be part of a design verification shelf-
314 life testing program (see section VI.A.2.b). For example, an applicant may follow a protocol for
315 assessing the impact of shipping preconditions on a specific EDDO. When used to evaluate a
316 product that has been fully preconditioned, the protocol should enable assessment of the impact
317 of actual preconditions associated with use (e.g., repeat use following the instructions for use
318 including any reprocessing steps) and verification that the EDDO is maintained following
319 preconditioning.
320
321 Regardless of the type of testing method, EDDO evaluation during design verification should
322 address the applicable preconditions as described in section VI.A.1. If preconditioning is
323 omitted, the associated protocol should provide a rationale for the omission.
324
325 a. Sampling considerations
326
327 Sampling plans for design verification testing for EDDOs should be risk-based, taking into
328 consideration the indication for use, patient population, drug being delivered, context of use, and
329 complexity of design and manufacturing. For example, a product with a higher risk profile
330 would warrant a more robust sampling plan than a product with a lower risk profile. Sampling
331 recommendations in recognized standards may be used in developing sampling plans, as
332 appropriate, based on product-specific risk considerations. A design verification testing protocol
333 should include a statistical sampling plan with the number of lots to be tested and acceptance
334 criteria. The tested lots should be manufactured using principles that are representative of the
335 commercial process (e.g., materials and methods of manufacture).

34
Additional information regarding the reprocessing of medical devices can be found in the guidance for industry
and FDA staff Reprocessing Medical Devices in Health Care Settings: Validation Methods and Labeling (March
2015).

10
 Contains Nonbinding Recommendations
Draft — Not for Implementation
336 b. Shelf-life and stability testing considerations
337
338 Design verification testing should include an evaluation of EDDOs that may change over time or
339 have age-related failure modes. Data provided to support a proposed expiration date should
340 demonstrate that EDDO performance is maintained. For a combination product, such data can
341 be derived from design verification shelf-life testing, 35 stability testing, 36 or both. For a
342 combination product, the final determination of the expiration date is informed by the
343 maintenance of EDDO performance and drug stability testing. EDDOs that would not change
344 over time (e.g., physical dimensions such as needle length) would not warrant evaluation.
345
346 The testing to support a proposed expiration date should include consideration of the number of
347 repeat uses of the product to deliver the drug and the potential impact of any associated cleaning
348 and reprocessing cycles and interim storage between uses. This testing should consider and
349 address the applicable preconditions as described in section VI.A.1 and include a justification for
350 any applicable preconditions omitted during the shelf-life or stability testing. This justification
351 may include other testing information and an explanation as to how such testing information
352 addresses or supports the omission of any identified precondition during shelf-life or stability
353 testing.
354
355 For combination products, to support the proposed expiration date, verifying the performance of
356 certain EDDOs over time may be accomplished by relying on or leveraging drug stability testing
357 results (e.g., for EDDOs subject to chemical degradation such as EDDOs for a vaginal system). 37
358 For EDDOs subject to physical or mechanical degradation, additional data to address design
359 verification shelf-life testing of the device drug-delivery function may be appropriate to support
360 expiration dating. Such design verification shelf-life testing should be conducted using the final
361 finished product under real-time aging conditions. As appropriate, accelerated aging data may
362 be used to establish the shelf life. When used, accelerated aging data should be confirmed by
363 real-time aging data. Applicants who are considering this approach should discuss their
364 proposals with the Agency (see section IX for more information).
365
366 B. Design Validation for EDDOs
367
368 To ensure appropriate validation, the applicant must ensure that devices conform to defined user
369 needs and intended uses and must include testing of production units under actual or simulated
370 use conditions. 38
371
372 The most appropriate method may depend on the application type, stage of development, and
373 EDDO. For these studies, it is important that the protocol be designed with endpoints that have

35
Shelf life is “the term or period during which a commodity remains suitable for the intended use” (FDA guidance
Shelf Life of Medical Devices (April 1991)).
36
Stability testing refers to a requirement for drugs and combination products addressed in 21 CFR 211.166 and 21
CFR 4.4, respectively.
37
For EDDOs that are evaluated as part of stability testing, information and data as described in ICH Q1A(R2)
Stability Testing of New Drug Substances and Products (November 2003) and related FDA stability guidances are
appropriate to support the proposed expiration date.
38
Design validation shall be performed under defined operating conditions on initial production units, lots, or
batches, or their equivalents (21 CFR 820.30(g) and 4.4(a)).

11
 Contains Nonbinding Recommendations
Draft — Not for Implementation
374 the capability of validating device performance. For certain application types, examples of
375 methods available to validate the EDDO specifications may include the studies identified below.
376
377 • Clinical studies (and/or reliance on FDA’s finding of safety/effectiveness for a reference
378 listed drug (RLD) or reference product)
379
380 • Pharmacokinetic/pharmacodynamic (PK/PD) or bioequivalence/bioavailability studies
381
382 As an example, when developing an infusion pump that is submitted in an NDA, the EDDO of
383 flow rate would be evaluated through effectiveness data, PK/PD data, and safety data (e.g.,
384 adverse events, occlusion rates, infusion site reactions) in appropriately designed clinical studies.
385
386 As appropriate, for certain applications, some EDDOs may be validated using alternative
387 methods, such as:
388
389 • Literature: e.g., injection site and patient population information to support the proposed
390 injection depth specification
391
392 • Simulated bench testing: 39 studies designed to evaluate whether users are capable of
393 using the prototype devices (e.g., exerting forces, hearing sounds) over the range of the
394 EDDO specification 40
395
396 • Anthropometric data: e.g., simulated strength testing of specific patient populations and
397 postures, capability of specific populations’ ability to hear specific tones
398
399 For example, for an NDA for a PFS, the EDDO validation testing might include clinical studies
400 appropriately designed to validate deliverable volume and injection to the target site, analysis of
401 anthropometric data 41 for cap removal force, and simulated bench testing using prototype PFS
402 devices for breakloose and glide forces.
403
404 1. Additional Validation Considerations for ANDA Submissions and BLAs Submitted
405 under Section 351(k) of the PHS Act for Combination Products
406
407 For an ANDA submission or a BLA submitted under section 351(k) of the PHS Act for a
408 combination product, it may be possible to rely on an RLD or reference product, respectively, by
409 providing a side-by-side comparison of EDDOs or EDDO performance 42 for the combination
410 product against that for the RLD or reference product.
411

39
Applicants would have to manufacture devices that function at the limits of the specification to effectively
validate the EDDO.
40
These simulated bench tests are different than human factors user validation studies.
41
ANSI/AAMI HE 75:2009 (R2018) Human factors engineering – Design of medical devices, addresses how to
perform this analysis.
42
For example, if the specifications of the RLD or reference product are unknown to the applicant, then the
application can use a side-by-side comparison of performance testing to verify that the ANDA or BLA submitted
under 351(k) device performs similarly to the RLD or reference product, respectively.

12
 Contains Nonbinding Recommendations
Draft — Not for Implementation
412 If the comparison shows there are differences from the EDDO specifications in the RLD or
413 reference product, then applicants may provide additional data to support that the specification is
414 appropriate for an ANDA or BLA submitted under section 351(k) of the PHS Act, as applicable.
415 For example, for the EDDO of activation force, if the specification or performance was different
416 from the RLD or reference product, the applicant may provide data demonstrating the activation
417 force is not too high such that the proposed combination product would not be therapeutically
418 equivalent to, or biosimilar to or interchangeable with, as applicable, the RLD or reference
419 product. 43
420
421 FDA recognizes that in some instances applicants may not have certain information for
422 comparison (e.g., if the RLD or reference product is discontinued or if the proposed combination
423 product has different design features). In such instances, the applicant should contact the
424 Agency through a controlled correspondence or pre-ANDA meeting request, or following
425 procedures described in the draft guidance for industry Formal Meetings Between the FDA and
426 Sponsors or Applicants of BsUFA Products, 44 as appropriate, (see section IX) with a proposal for
427 how to validate the EDDO.
428
429 2. Additional Validation Considerations for Premarket Notifications [510(k)]
430
431 For a 510(k), submitters compare the intended use (including indications for use) and the
432 technological characteristics of their device, including EDDOs, to the predicate device. When
433 there are differences in technological characteristics, which do not raise different questions of
434 safety and effectiveness from the predicate device, FDA may request additional performance
435 data, including clinical data as necessary, to assess whether the device is as safe and effective as
436 the predicate device. 45
437
438
439 VII. CONTROL STRATEGIES FOR ESSENTIAL DRUG DELIVERY OUTPUTS
440
441 After completion of the design verification and validation processes described in section VI, a
442 control strategy is used to ensure that each lot of the final finished product is manufactured to
443 conform to the design outputs. For a given EDDO, an appropriate control strategy may consist
444 of one or more types of control steps at different stages of the manufacturing process. Some
445 control steps are performed earlier in the manufacturing process (e.g., upstream controls such as
446 in-process controls, control of process parameters, control of incoming materials, and purchasing
447 controls). Other control steps are performed at the end of the manufacturing process (e.g., a
448 downstream control such as lot release testing).
449

43
Information on differences in the user interface between an ANDA and its RLD is beyond the scope of this
guidance. For additional information on identifying and analyzing differences between a proposed generic
combination product and the RLD, see the draft guidance for industry Comparative Analyses and Related
Comparative Use Human Factors Studies for a Drug-Device Combination Product Submitted in an ANDA (January
2017). When final, this guidance will represent FDA’s current thinking on this topic.
44
August 2023. When final, this guidance will represent the FDA’s current thinking on this topic.
45
Applicants submitting 510(k)s for their devices should address the submission recommendations included in the
guidance for industry and FDA staff The 510(k) Program: Evaluating Substantial Equivalence in Premarket
Notifications [510(k)] (July 2014).

13
 Contains Nonbinding Recommendations
Draft — Not for Implementation
450 The control strategy developed should be risk-based. Therefore, the number and types of
451 controls implemented, and the amount of information regarding the control strategy to include in
452 an application should correspond to the product risks. For a lower risk product with less
453 complex manufacturing processes, certain EDDOs may be adequately controlled with
454 downstream controls. A possible example is release testing of glide force and breakloose force
455 on a PFS with a non-emergency use drug for administration by a health care provider. In
456 contrast, for a higher risk product, a combination of upstream and downstream controls may be
457 needed to ensure consistent EDDO performance.
458
459 When developing an upstream control strategy for an EDDO, it is important to consider the
460 attributes and manufacturing process steps that can influence the EDDO, and describe why these
461 attributes or process steps are the only ones that influence the EDDO. Appropriate controls
462 should be identified for each attribute or process step, and the documentation used to verify these
463 controls should be identified. For certain attributes, purchasing controls or an incoming
464 test/examination, in combination with design assessment verification testing, may be sufficient
465 (e.g., for the length of the syringe and needle that cannot change over the expiration dating
466 period). For other attributes or process steps, a combination of controls may be appropriate. The
467 description of the control strategy should include an assessment of how the effectiveness of an
468 upstream control is impacted by the downstream manufacturing steps.
469
470 Information to support that in the context of the manufacturing process, the control strategy is
471 adequate to ensure consistent EDDO performance is submitted as part of the marketing
472 application, submitted upon request, or available during facility inspection, depending on
473 application type (see section VIII.B). The supporting information may consist of batch analysis
474 (i.e., testing the final finished product against the specification) or EDDO evaluation conducted
475 as part of stability testing (see section VI.A.2.B). For more information, see regulations and
476 guidance on specific application type submission requirements for manufacturing information.
477 For products subject to premarket approval (e.g., under an ANDA, BLA, NDA, PMA), control
478 strategy information is included in the application. 46 In contrast, 510(k) submissions generally
479 do not include control strategy information as part of the 510(k) itself; however, information
480 demonstrating control of the 510(k) device’s manufacturing process must be available for FDA
481 review during inspections. 47
482
483 See Appendix D for an illustration of a control strategy for the final assembly of a syringe based
484 autoinjector.
485
486
487 VIII. INFORMATION TO PROVIDE IN APPLICATIONS
488
489 This section describes the EDDO information and performance data included in investigational,

46
Information on manufacturing controls must be included in such applications, see 21 CFR 314.50(d)(1)(ii),
314.94, 601.2(a), and 814.20(b)(4).
47
See section 704(a) of the FD&C Act. Such records must be made available for FDA review during an inspection
conducted under section 704(a)(1) of the FD&C Act or when requested by FDA in advance of or in lieu of an
inspection as described in section 704(a)(4) of the FD&C Act.

14
 Contains Nonbinding Recommendations
Draft — Not for Implementation
490 marketing, and post-market change applications to demonstrate the device drug-delivery
491 function.
492
493 A. IND and IDE Applications
494
495 The data provided in IND and IDE applications for drug delivery devices should reflect the
496 development stage of the product. Bio-INDs under 21 CFR 320.31 for ANDAs are out of scope
497 of this guidance. 48 It is understood that for combination products, the drug delivery device may
498 be introduced after the investigational drug development begins. We understand that the device
499 design may evolve based on study results. The device drug-delivery data submitted in an IND or
500 IDE application 49 should reflect the development process, and applicants should consider the
501 device design, the drug being delivered, the patient populations, study design, and risks to the
502 study participants when determining the data to provide.
503
504 To support the first use of the device or device constituent part in a clincial study, it is important
505 that applicants provide information demonstrating that the safety and performance of the device
506 is adequate for the proposed investigation. Information from EDDO verification testing is
507 provided to demonstrate that the device performs and the drug is delivered (dose and delivery
508 site) as intended.
509
510 As development proceeds and the product design evolves, additional information may be needed
511 to support the use of the final finished product in pivotal 50 clinical studies. As applicable, the
512 submission should provide more comprehensive information that builds upon earlier safety
513 information and study results, including information that applies to the device drug-delivery
514 function. As product development continues, for any planned changes to the EDDOs of the
515 device for commercialization, it may be possible to use the EDDO performance data from the
516 clinical study design iteration to help bridge between prototypes.
517
518 For any required IDE and IND applications for drug delivery devices, requirements for
519 submission content relevant to EDDOs include, but are not limited to the following, respectively:
520
521 • A complete report of prior investigations of the device and a summary of certain sections
522 of the investigational plan, or, in lieu of such a summary, the complete investigational
523 plan; 51 a description of the methods and controls used for the manufacture of the device;
524 risk analysis; and description of the device. 52

48
A sponsor of a bio-IND under 21 CFR 320.31 with a drug delivery device may submit a controlled
correspondence (see section IX) with any questions regarding the applicability of the information described in this
section. For additional information on bio-INDs, see Manual of Policies and Procedures 5210.5, Review of
Investigational New Drug Applications (Bio-INDs) by the Office of Generic Drugs (Rev. 3, April 14, 2022).
49
For combination products, see footnote 10 for information on application types.
50
For the purpose of this guidance, a pivotal study is a definitive study in which data are evaluated to establish the
safety and effectiveness of the product; this study is submitted in a marketing application to support its intended use.
Other terms for such studies include key, critical, and major studies.
51
We note that the sponsor shall submit to FDA the complete investigational plan and a complete report of prior
investigations of the device if no IRB has reviewed them, if FDA has found an IRB’s review inadequate, or if FDA
requests them. 21 CFR 812.20(b)(2).
52
See 21 CFR 812.20(b)(2) and (3), 812.25(c) and (d), and 21 CFR 812.27(a).

15
 Contains Nonbinding Recommendations
Draft — Not for Implementation

525 • A brief summary of previous human experience with the product, an investigational plan;
526 and information describing the composition, manufacture, and control of the product. 53
527
528 As described above, depending on the application type, certain information on EDDOs is a
529 required part of the application, while other information is recommended for submission. This
530 section outlines considerations for this information that the applicant or submitter, depending on
531 the context, is either required or recommended to “provide,” “identify,” “describe,” or
532 “compare,” for example:
533
534 (1) Device description documentation –
535
536 (a) Provide a description of the device design, including any novel features and
537 functionalities, including engineering drawings or diagrams of the device with all
538 dimensions labeled, descriptions of the individual device components, or any other
539 available information to explain the device design.
540
541 (b) Describe the principles of operation of the device and how it functions throughout
542 use.
543
544 (c) Describe any accessories or other devices labeled for use with the device.
545
546 (2) Device safety – Identify EDDOs that are necessary for patient safety during the study.
547 For example, a device may cause harm if the dose accuracy performance is not adequate
548 (e.g., by delivering a larger dose than intended). For safety-related EDDOs, provide
549 verification and validation data prior to the start of a clinical study. See Performance
550 data for data recommendations.
551
552 In the overall device risk analysis section, include EDDO-related risks.
553
554 (3) Performance data 54 – Include summary test results using established test methods for the
555 device (e.g., recognized standards, test methods discussed in FDA guidance) or complete
556 verification test reports for unique or unrecognized test methods. Recommendations
557 regarding the summaries and/or documentation can be found in the guidance for industry
558 and FDA staff Recommended Content and Format of Non-Clinical Bench Performance
559 Testing Information in Premarket Submissions (December 2019).
560
561 The following considerations apply when the clinical study results are part of the EDDO
562 validation:
563
564 • If the clinical study is intended to obtain data to validate one or more EDDOs, it is
565 appropriate for the clinical study protocol to include endpoints relevant to the
566 performance of the device (e.g., infusion rate, dose range, injection time). Where

53
See 21 CFR 312.23(a)(3)(ii), 312.23(a)(3)(iv), and 312.23(7)(i).
54
As discussed in the part 4 CGMP guidance, investigational combination products that include device constituent
parts may be subject to design controls under 21 CFR 820.30, which includes design verification of an
investigational product.

16
 Contains Nonbinding Recommendations
Draft — Not for Implementation
567 possible, to support provision of evaluable EDDO data, applicants should submit such
568 protocols for Agency feedback on the EDDO validation endpoints in a formal meeting or
569 communication with FDA (see section IX). Also, such clinical studies should be
570 conducted with the final finished drug delivery device.
571
572 • It is important that the clinical study protocol include safety endpoints that capture drug
573 delivery device failures, malfunctions, and adverse events. 55
574
575 B. Marketing Applications
576
577 For marketing applications for devices and combination products, requirements for submission
578 content relevant to EDDOs include, but are not limited to the following:
579
580 • A PMA must include a complete description of the device, each of the functional
581 components or ingredients (as described in FDA regulations), and the methods used in,
582 and the facilities and controls used for, the manufacture of the device; and nonclinical
583 laboratory studies. 56
584
585 • A 510(k) submission must include a description of the device, a statement indicating the
586 device is similar to and/or different from other comparable products including data and
587 information to support that statement, and as applicable, appropriate supporting data to
588 show that the manufacturer has considered what consequences and effects any changes
589 or modifications or new uses might have on the safety and effectiveness of the device. 57
590
591 • A De Novo request must include a complete description of the device and nonclinical
592 laboratory studies. 58
593
594 • NDA, ANDA, and BLA applications must include a description of the product, data to
595 show that the final finished product meets specifications, a description of the
596 manufacturing processes and controls, and stability data. 59
597
598 Other data and information relevant to EDDOs, such as the establishment of design inputs and
599 outputs and completion of design verification and validation testing, are needed to comply with
600 design control requirements. 60
601
602 For products subject to premarket approval (e.g., under an ANDA, BLA, NDA, or PMA), as
603 described above, manufacturing control strategy information is required to be included in the
604 application. In contrast, 510(k) submissions generally do not include control strategy
605 information as part of the 510(k) itself; however, information demonstrating control of the

55
Note that human factors considerations are outside the scope of this guidance; see section I.
56
See 21 CFR 814.20(b)(4)(i), (ii), and (v); 21 CFR 814.20(b)(6)(i).
57
See 21 CFR 807.92(a)(4); 21 CFR 807.87(f) and (g).
58
See 21 CFR 860.220(a)(6) and 21 CFR 860.220(a)(15)(i).
59
See 21 CFR 314.50(d)(1), 314.94(a)(9), and 601.2(a).
60
See 21 CFR 820.30 for devices and 21 CFR 4.4 for combination products.

17
 Contains Nonbinding Recommendations
Draft — Not for Implementation
606 510(k) device’s manufacturing process as part of complying with applicable GMP
607 requirements 61 must be available for FDA review during inspection. 62
608
609 As described above, depending on the application type, certain information on EDDOs is a
610 required part of the application, while in other instances that same information is recommended
611 for submission. This section outlines considerations for this information that the applicant or
612 submitter, depending on the context, is either required or recommended to “provide,” “include,”
613 “identify,” “describe,” or “compare,” for example. 63
614
615 (1) Device description documentation –
616
617 (a) Information in section VIII.A, item #1
618
619 (b) When applicable, include a side-by-side comparison of the to-be-marketed device
620 with any earlier versions utilized in the provided verification and validation testing,
621 identifying any design and manufacturing changes.
622
623 (2) Performance data 64, 65 – Include acceptance criteria and performance data verifying and
624 validating the final finished product. Applicants should use recognized standards and
625 FDA guidance to inform design and testing, as applicable. 66 Provide the following data:
626
627 (a) Design input requirements (i.e., the physical and performance requirements of a
628 device that are used as a basis for device design)
629
630 (b) Design output specifications (e.g., device description, drawings, specifications,
631 materials)
632
633 (c) Design verification plan/summary report, supporting data, and traceability
634
635 (d) Design validation plan/summary report, supporting data, and traceability
636

61
See 21 CFR 820.70(a) and 820.180.
62
See section 704(a) of the FD&C Act (21 U.S.C. 374(a)). Also see footnote 47.
63
Master files may be useful tools to help preserve the trade secrets of a third party that are not known to the
applicant. For more information on biologics, device, and drug master files, see CBER’s Master Files for CBER-
Regulated Products web page (available at https://www.fda.gov/vaccines-blood-biologics/development-approval-
process-cber/master-files-cber-regulated-products), CDRH’s Master Files web page (available at
https://www.fda.gov/medical-devices/premarket-approval-pma/master-files), and CDER’s Drug Master Files (DMF)
web page (available at https://www.fda.gov/drugs/forms-submission-requirements/drug-master-files-dmfs),
respectively.
64
When submitting this information in an NDA, ANDA, or BLA, see section 5 in the FDA eCTD Technical
Conformance Guide: Technical Specifications Document: Guidance for Industry Providing Regulatory Submissions
in Electronic Format - Certain Human Pharmaceutical Product Applications and Related Submissions Using the
eCTD Specifications (July 2020).
65
For additional information regarding information included in NDAs, ANDAs, and BLAs, see the guidance for
industry Container Closure Systems for Packaging Human Drugs and Biologics (July 1999).
66
For questions about design control documentation, see the guidance for industry Design Control Guidance for
Medical Device Manufacturers (March 1997).

18
 Contains Nonbinding Recommendations
Draft — Not for Implementation
637 (i) If changes are made after design validation, include a discussion, along with
638 supportive data, of either why the validation data is still applicable or how the
639 applicant validated the changes
640
641 (e) Risk analyses to evaluate the adequacy of the design verification and design
642 validation plan
643
644 (3) EDDO – Identify EDDO(s) for the drug delivery device
645
646 To facilitate submission and review of the EDDO information, provide a summary of the
647 EDDO information included in the application in the device description documentation,
648 referencing the performance data section and supportive information. For NDAs, BLAs,
649 and ANDAs, if the EDDO performance relies on or leverages CQA and related
650 information in the CMC sections, the summary should reference the relevant sections. 67
651
652 When verifying and validating the EDDO(s), include the EDDO(s) in the following
653 evaluations, when applicable as discussed in sections VI and VII:
654
655 (a) Preconditioning – Provide documentation that demonstrates that the device EDDOs
656 are met after preconditioning testing. See section VI.A.1 for additional
657 considerations.
658
659 (b) Stability/Shelf-life testing – Include endpoints in the final finished product stability
660 and shelf-life testing program to verify that EDDOs that could change with aging are
661 maintained at expiry. See section VI.A.2.b for additional considerations.
662
663 (c) Control strategy – Provide a control strategy based on the product risk profile that
664 ensures that the final finished product maintains its EDDO(s). 68 Include a summary
665 of the controls implemented (upstream and/or at release), including a justification
666 describing how the controls are sufficient to assure the quality of the EDDO is
667 achieved. In the control strategy description, include supporting evidence such as
668 engineering drawings, tolerance stack-up analysis, and manufacturing flow diagrams.
669 FDA may request additional specific documentation referenced in the control strategy
670 during the review. 69 Applicants can consult with the appropriate product office for
671 questions regarding control documentation to include in a submission.
672
673 See section VII for additional information on developing a control strategy.

67
If the application includes proposed established conditions (ECs) for the device constituent part, and the EDDOs
constitute a subset of the proposed ECs, we recommend referencing the EC-related sections of the application for
the EDDO information. See the ICH guidance for industry Q12 Technical and Regulatory Considerations for
Pharmaceutical Product Lifecycle Management (May 2021) and the draft FDA guidance for industry ICH Q12:
Implementation Considerations for FDA-Regulated Products (May 2021) (when final, this guidance will represent
FDA’s current thinking on this topic).
68
See section VII for additional information.
69
See Appendix D for an example of a control strategy for an EDDO where documentation to verify the control was
either included (i.e., documentation from the applicant) or referenced (i.e., documentation from sources other than
the applicant) in the submission.

19
 Contains Nonbinding Recommendations
Draft — Not for Implementation
674 C. Submissions for Post-Market Changes That May Impact Essential Drug
675 Delivery Outputs
676
677 When modifying the product design or manufacturing process of an approved or cleared product,
678 applicants should evaluate whether there are any new EDDOs and verify and validate the new
679 EDDOs, as appropriate. Applicants should also perform an analysis of the impact of the change
680 on the verification and validation of the previously identified EDDOs.
681
682 For post-market change submissions for devices and combination products, requirements for
683 submission content relevant to EDDOs include, but are not limited to the following:
684
685 • PMA, for changes requiring a submission/notice: 70 a supplement/notice must include
686 identification of each change and the reason for each change (PMA supplements); must
687 provide a full explanation of the basis for the changes (special PMA supplements); or
688 must describe in detail the changes and summarize the data or information supporting the
689 change (30-day notices). 71
690
691 • 510(k) submission, for changes requiring a new 510(k): 72 a 510(k) must include a
692 summary of how the technological characteristics of the device compare to a predicate
693 device and what consequences and effects the difference(s) might have on the safety and
694 effectiveness of the device. 73
695
696 • NDA/BLA/ANDA, for changes requiring approval prior to distribution: 74 a supplement
697 must include a detailed description of the proposed change; a description of the methods
698 used and studies performed to assess the effects of the change, and the data derived from
699 such studies. 75
700
701 As described above, depending on the application type, certain information on EDDOs is a
702 required part of the postapproval submission, while in other instances that same information is
703 recommended for submission. This section outlines considerations for this information that the
704 applicant or submitter, depending on the context, is either required or recommended to
705 “provide,” “include,” “identify,” or “describe,” for example:
706
707 (1) Description of changes in comparison to the approved or cleared product
708
709 (2) Potential impact of the change to EDDOs
710
711 (3) Verification, validation, or both of potentially impacted EDDOs
712

70
See 21 CFR 814.39(a), (c), or (f).
71
See 21 CFR 814.39(c)(1), 814.39(d)(1)(ii), 814.39(f).
72
Changes to a previously cleared device may require the submission of a new 510(k). See the guidance for
industry and FDA staff Deciding When to Submit a 510(k) for a Change to an Existing Device (October 2017).
73
See, e.g., 21 CFR 807.92(a)(6), 21 CFR 807.87(f) and (g).
74
See 21 CFR 314.70(b), 314.97(a), and 601.12(b).
75
See 21 CFR 314.70(b)(3), 314.97(a), and 601.12(b)(3).

20
 Contains Nonbinding Recommendations
Draft — Not for Implementation
713 If the specifications are unchanged, then provide verification data, including stability and
714 preconditioning (e.g., storage, shipping, conditions associated with use) information
715 when applicable. If an applicant does not believe that the change would be impacted by
716 preconditioning, then provide a scientific justification and/or evidence that the
717 performance is not expected to shift outside the validated specifications over time or after
718 preconditioning.
719
720 To support the change, provide a side-by-side comparison between the new design or
721 new manufacturing process/site EDDO and the original device design or manufacturing
722 process/site EDDO to assess for potential shifts in performance that may lead to potential
723 out-of-specification results. For an ANDA, in the case of shifts in performance, a
724 comparison to the RLD may be appropriate.
725
726 If the specifications are changing, provide new EDDO validation data or a rationale for
727 why the validation from the original application can be leveraged (e.g., tightening a
728 specification). 76
729
730 (4) Updated control strategy
731
732 To help ensure EDDO maintenance throughout the product lifecycle, include a re-
733 evaluation of the control strategy to determine whether it is still applicable given the
734 design changes, manufacturing changes, or both, including information to demonstrate
735 that the control strategy is still effective.
736
737 Alternatively, if an applicant is making changes that necessitate a change to the control
738 strategy, provide documentation for the changes made to the EDDO control strategy and
739 information to demonstrate that the new strategy is as effective as the original control
740 strategy.
741
742
743 IX. INTERACTION WITH FDA
744
745 During product development we strongly encourage applicants to begin developing and
746 characterizing the EDDOs that will need to be controlled in the final finished product and to
747 begin to develop a device control strategy for the product the applicant plans to market. We
748 recommend applicants submit the proposed EDDOs and control strategy for Agency feedback.
749 The EDDOs and control strategy could be discussed at multiple types of formal meetings or in
750 other communications with FDA, depending on application type. These meetings should be used
751 consistently with their intended purpose. Also, as appropriate for the type of submission and the
752

76
See the draft guidance for industry Bridging for Drug-Device and Biologic-Device Combination Products, which
includes information on how to determine the type of bridging data that may be appropriate. When final, this
guidance will represent the FDA’s current thinking on this topic.

21
 Contains Nonbinding Recommendations
Draft — Not for Implementation
753 combination product lead center, the meeting request information may vary. 77
754
755 When seeking feedback on proposed EDDOs, the meeting background package should provide a
756 clear description of the product and its drug delivery function. This may include figures,
757 instructions for use, description of the intended user population, and context of use (e.g.,
758 immediate or emergency use versus chronic therapy, home versus clinical setting). The meeting
759 background package should provide specific EDDO questions and may request the participation
760 of an FDA office or staff with the appropriate expertise. If there are questions regarding the
761 control of EDDOs, the control strategy proposal should be provided. These meetings would
762 include discussion of the proposed control strategy and the proposed approach and timing of
763 EDDO validation (e.g., type of studies and completion before beginning the pivotal clinical
764 studies).

77
For information on requesting FDA feedback, see the guidance for industry and FDA staff Requesting FDA
Feedback on Combination Products (December 2020). For additional information on the process to request a
meeting under user fee programs, see the draft guidance for industry Formal Meetings Between the FDA and
Sponsors or Applicants of BsUFA Products (August 2023) (when final, this guidance will represent the FDA’s
current thinking on this topic) and the draft guidance for industry Formal Meetings Between the FDA and Sponsors
or Applicants of PDUFA Products (September 2023) (when final, this guidance will represent the FDA’s current
thinking on this topic). ANDA applicants may submit a pre-ANDA meeting request or a controlled correspondence
to obtain Agency feedback. See the guidances for industry Formal Meetings Between FDA and ANDA Applicants of
Complex Products Under GDUFA (October 2022) and Controlled Correspondence Related to Generic Drug
Development (March 2024). See also the guidance for industry and FDA staff Requests for Feedback and Meetings
for Medical Device Submissions: The Q-Submission Program (June 2023).

22
 Contains Nonbinding Recommendations
Draft — Not for Implementation
765 APPENDIX A – ESSENTIAL DRUG DELIVERY OUTPUT IDENTIFICATION
766 EXAMPLE – PREFILLED SYRINGE
767
768 In this hypothetical example, the applicant is developing a drug in a prefilled syringe (PFS)
769 intended for subcutaneous injection that will be delivered by a health care provider (HCP) to a
770 patient in a health care setting. The PFS is a glass barrel containing a 12mm prestaked needle.
771 The labeling directs the HCP to administer drug using the PFS at a 30-45 degree angle while
772 pinching the skin. The PFS is also manufactured with a needle safety device to mitigate the risk
773 of accidental needle sticks. The device design requires the needle safety feature to be activated
774 before completion of injection. During development, the applicant considers the design inputs in
775 identifying the design outputs and identifies which design outputs are essential drug delivery
776 outputs (EDDOs).
777
778 Step 1 – Identify Design Outputs
779 The applicant first identifies the device design outputs as part of design control activities.
780 Design output requirements include all attributes of the device necessary to meet intended user
781 needs and include, for example, specifications for deliverable volume, clarity of the barrel,
782 biocompatibility, sterility, color of plunger rod, markings on the barrel, material performance
783 characteristics, needle safety activation force, and needle guard characteristics.
784
785 Step 2 – Identify Drug Delivery Design Outputs
786 As design outputs are being developed, the applicant analyzes the tasks needed to deliver the
787 intended drug dose with the PFS to the intended delivery site, including the successful product
788 preparation and the initiation, progression, and completion of dose delivery, and identifies design
789 outputs related to these tasks. These are the drug delivery design outputs.
790
791 Other design outputs unrelated to delivering the intended drug dose (e.g., biocompatibility and
792 sterility) are further filtered to eliminate outputs that are not EDDOs (see Figure 1).
793
794 Step 3 – Identify System Level Design Outputs
795 The applicant analyzes the drug delivery outputs to identify system level drug delivery outputs.
796 For example, the applicant determines that glide force is a system level drug delivery output
797 because it is necessary for the progression and completion of the dose. (See Figure 2 for an
798 illustrative example of the relationship between system level and component level design
799 outputs.)
800
801 After assessing the device, the applicant identifies the following system level drug delivery
802 outputs: cap removal force, deliverable volume, injection depth, and injection forces (breakloose
803 force, glide force, needle safety activation force).
804
805 Step 4 – Identify Device Dependent Design Outputs
806 The applicant further analyzes the system level drug delivery outputs to determine which are
807 dependent on the device design and which, if any, are dependent on the user. The applicant
808 determines that the following system level drug delivery outputs are device dependent:
809

23
 Contains Nonbinding Recommendations
Draft — Not for Implementation

810 • Cap removal force – The force to remove the cap is dependent on the design and
811 manufacture of the cap and its interface to the nozzle PFS barrel tip
812
813 • Deliverable volume – The volume of drug extracted during injection is dependent on the
814 device design and fill volume (independent of the user)
815
816 • Injection forces – The injection forces, which include the force to initiate the injection
817 (i.e., breakloose force), the force to sustain the injection (i.e., glide force), and needle
818 safety activation force 78 are dependent on the device design (independent of the user)
819
820 The applicant determines that while the target injection site is subcutaneous tissue and the PFS is
821 prestaked with a 12mm needle, the user controls the injection depth through the injection
822 technique. The injection depth is dependent on the user and independent of the device; therefore,
823 it is not an EDDO.
824
825 Based on these assessments, the applicant determines that the following are EDDOs for the PFS
826 product. For this illustrative example, the EDDOs are categorized in the table by the different
827 aspects of drug delivery (top row) to which they are related.
828
Delivery of Delivery to Product Initiation of Dose delivery Dose delivery
intended the target preparation dose delivery progression completion
dose site
Deliverable N/A Cap removal Breakloose Glide force • Glide force
volume force force • Needle safety
activation force
829

78
If a device design does not require the user to overcome the needle safety feature to be activated before
completion of injection, needle safety activation force would not be an EDDO.

24
 Contains Nonbinding Recommendations
Draft — Not for Implementation
830 APPENDIX B – ESSENTIAL DRUG DELIVERY OUTPUT IDENTIFICATION
831 EXAMPLE – AUTOINJECTOR
832
833 As an illustrative example of the distinction between essential drug delivery outputs (EDDOs)
834 and other design outputs, the table below lists design outputs that were identified following
835 consideration of hypothetical design inputs. The resulting hypothetical product is an autoinjector
836 with a prefilled syringe (PFS) subassembly device constituent part that is also the primary
837 container closure for the drug constituent part. The table shows the outcome of applying the
838 filtering steps to each design output. The design outputs that meet the criteria of each of the
839 filtering steps are EDDOs. The design outputs with gray shading in one or more columns do not
840 meet the criteria of a filtering step(s) and are not EDDOs (i.e., gray shading in the System level
841 column means the design output does not meet the system level criteria and is therefore, not an
842 EDDO). As an example, dose accuracy meets each of the criteria for an EDDO and therefore is
843 an EDDO. In contrast, the design output of PFS-fill volume/container content, meets the criteria
844 of a drug delivery design output but does not meet the criteria of system level or device
845 dependent and therefore is not an EDDO. 79
846
Design Drug System Device EDDO (yes or no) and rationale
outputs delivery level dependent
design
outputs
Yes. The automated dose delivery is
Dose accuracy necessary to ensure appropriate drug
delivery.
PFS-fill No. It is a component level output
volume/container because the fill volume influences but is
content subordinate to dose accuracy.
Biocompat- No. It is not an output intended for drug
ibility delivery.
Container No. It is not an output intended for drug
closure delivery.
integrity
Drop No. It is not an output intended for drug
testing/ delivery.
Shipping
Packaging No. It is not an output intended for drug
integrity delivery.
Extended needle length Yes. It ensures the needle is at the right
depth and is dependent on the device.
No. It is a component level output
PFS-needle length
because it influences but is subordinate
to the extended needle length function.

79
It should be noted that the EDDO identification is distinct from CQA identification; however, some design
outputs in the table may also be considered critical from a drug quality and manufacturing perspective. Refer to
section III for further explanation regarding the relationship between EDDOs and CQAs.

25
 Contains Nonbinding Recommendations
Draft — Not for Implementation
Design Drug System Device EDDO (yes or no) and rationale
outputs delivery level dependent
design
outputs
No. It is a component level output
Needle inner diameter
because it influences but is subordinate
to the injection time function.
Activation force Yes. It initiates drug delivery and is
dependent on the device not the user.
No. It is a component level output
Breakloose force
because it influences but is subordinate
to the injection time function.
No. It is a component level output
Glide force
because it influences but is subordinate
to the injection time function.
Yes. It ensures the drug is delivered to
Injection time the intended space within the appropriate
time and is dependent on the device not
the user.
No. It is not an output intended for drug
Override delivery because it is the force to
force overcome the needle safety mechanism,
which is activated post drug delivery.
No. It is not an output intended for drug
Drug
delivery because the drug appearance
visibility
assessment is a user action.
No. It is not an output intended for drug
delivery because it is the depth that the
Stopper stopper should be placed at in the syringe
height for compatibility with the autoinjector
plunger rod, filling process, and to ensure
the stopper does not come out.
No. It is not an output intended for drug
Needle
delivery because this attribute is specific
pullout
to the device durability ensuring the
force
needle does not detach.
Audible feedback/clicks Yes. It signals that the injection is
complete and is dependent on the device.
Visual feedback Yes. It signals that the injection is
complete and is dependent on the device.
Yes. Cap removal needs to be completed
Cap removal force
before the injection can be administered
and it is dependent on the device.
Rigid needle shield No. It is a component level output
removal force because it influences but is subordinate
to the cap removal force function.
847

26
 Contains Nonbinding Recommendations
Draft — Not for Implementation
848 APPENDIX C – EXAMPLES OF POTENTIAL ESSENTIAL DRUG DELIVERY
849 OUTPUTS BASED ON PRODUCT TYPE
850
851 The following tables provide examples of potential essential drug delivery outputs (EDDOs) for
852 different product types. Note that the EDDOs for specific products may vary due to product
853 specific differences (e.g., indication, design inputs, product design and technology).
854
855 Table 1: Prefilled Syringes
856
Delivery of Delivery to Product Initiation Dose Dose delivery
intended dose the target preparation of dose delivery completion
site delivery progression
Deliverable Needle • Cap removal Breakloose Glide force • Glide force
volume length 80 force force • Needle safety
• Withdrawal activation force 82
force 81
857
858

80
If a device is designed such that needle length is validated to reach a specific delivery space or tissue independent
of user technique (e.g., pinching, skin test bleb method), needle length is an EDDO.
81
For some co-packaged combination products, an empty syringe or a diluent prefilled syringe is used to withdraw
drug directly from a vial, or to reconstitute or mix a drug. The withdrawal force (a type of injection force) of this
empty or prefilled syringe is an EDDO because it is necessary for the preparation of the drug for injection and the
drug cannot be delivered to the patient unless preparation is successfully completed. If the empty or prefilled
syringe used in preparation is also used to deliver the drug, breakloose force and glide force are EDDOs.
82
If a device design requires the needle safety feature to be activated to complete drug delivery, needle safety
activation force is an EDDO. Conversely, if a device design does not require the needle safety feature to be
activated to complete drug delivery, then needle safety activation force is not an EDDO.

27
 Contains Nonbinding Recommendations
Draft — Not for Implementation
859 Table 2: Injectors
860
Device Delivery of Delivery Product preparation Initiation Dose Dose
platform intended to the of dose delivery delivery
dose target delivery progression completion
site
Auto- Dose Extended • Cap removal force • Activation N/A • Injection
injector accuracy needle • Activation force force time
length (e.g., to start (shield) • Audible/
reconstitution • Activation visual/
process) 83 force tactile
• Audible/visual/ (button) feedback
tactile feedback of
successful drug
preparation/reconstit
ution 84
On body • Dose Cannula • Cap removal force Activation Adhesion • Injection
injector accuracy/ length • Activation force force force time
dose (e.g., to start • Audible/
efficiency reconstitution visual/
• Injection process) 87 tactile
time 85 • Audible/visual/ feedback
• Interval of tactile feedback of • Adhesion
delivery 86 successful drug force
preparation/reconstit
ution 88
861
862
863
864
865

83
Depending on the design of the automated mixing/reconstitution feature, there may be alternate or additional
EDDOs.
84
If the injector design automates reconstitution of the drug prior to injection, features that ensure and/or indicate
successful mixing of the drug are EDDOs because if the drug is not mixed then the efficacy is diminished. If the
device does not allow for drug delivery when the drug is not well mixed, this may be considered adequate feedback
depending on the drug/indication to address the EDDO of audible/visual/tactile feedback.
85
On-body injectors are often intended to deliver a large volume (e.g., 1-3mL) of drug. Therefore, injection time is
an EDDO for delivery of intended dose because if the speed is too high then there is a higher risk of tolerability
issues impacting the delivered dose. In addition, injection time is an EDDO for dose delivery completion similar to
an autoinjector.
86
On-body injectors may be used to replace two separate autoinjectors and may be designed to deliver injections at
intervals. Therefore, that interval of delivery would be an EDDO.
87
See footnote 83.
88
See footnote 84.

28
 Contains Nonbinding Recommendations
Draft — Not for Implementation
866 Continued – Table 2: Injectors
867
Device Delivery Delivery Product preparation Initiation of Dose Dose
platform of to the dose delivery delivery
intended target delivery progress completion
dose site -ion
Pen Dose N/A • Cap removal force Injection N/A • Injection
91
injector accuracy • Activation force (e.g., to force time 92
start reconstitution • Audible/
process) 89 visual/
• Audible/visual/tactile tactile
feedback of successful feedback
drug
preparation/reconstitution
90

Jet Dose Penetra- • Cap removal force • Activation N/A • Injection

injector accuracy tion • Activation force (e.g., to force time
depth start reconstitution (shield) • Audible/
93
process) • Activation visual/
• Audible/visual/tactile force tactile
feedback of successful (button) feedback
drug
preparation/reconstitution
94

868
869
870
871

89
See footnote 83.
90
See footnote 84.
91
If the pen injector is automated (i.e., does not manually deliver the dose), the EDDO may be activation force
rather than injection force.
92
If the pen injector is automated, then injection time is an EDDO.
93
See footnote 83.
94
See footnote 84.

29
 Contains Nonbinding Recommendations
Draft — Not for Implementation
872 Table 3: Nasal Sprays 95
873
Delivery of intended dose Delivery Product Initiation Dose Dose
to the preparation of dose delivery delivery
target site delivery progression completion
• Pump delivery (spray N/A N/A Activation N/A N/A
weight) force
• Spray content uniformity
(SCU)
• Spray pattern
• Droplet size distribution
• Particle size distribution
(suspensions)
874
875
876 Table 4: Inhalation Products
877
Device Delivery of Delivery to Product Initiation Dose Dose
Platform intended the target site preparation of dose delivery delivery
dose delivery progression completion
Metered dose Emitted Aerodynamic Priming or Device N/A • Counter
inhaler drug particle size repriming 97 actuation accuracy 98
(MDI) - non- distribution force • Counter
breath (APSD) 96 actuation
actuated force
878
879
880

95
For additional information on nasal spray products, see the guidance for industry Nasal Spray and Inhalation
Solution, Suspension, and Spray Drug Products – Chemistry, Manufacturing, and Controls Documentation (July
2002).
96
Aerodynamic particle size distribution (APSD) characterizes the particle/droplet size distribution of delivered
aerosolized drug. APSD is a critical parameter, and its control is crucial for maintaining the quality of inhalation
drugs. This parameter is dependent on the formulation, container closure system, and device delivery mechanism.
The optimum APSD for most oral inhalation products has generally been recognized in the range of 1 to 5 um. For
more guidance conducting these tests, please refer to USP <601> for aerosols, sprays, and powders and USP
<1601> for nebulizer products. For guidance on information to provide to FDA regarding test setups and data
presentation in submissions, please refer to the draft guidance for industry, Metered Dose Inhaler (MDI) and Dry
Powder Inhaler (DPI) Drug Products—Quality Considerations (April 2018). When final, this guidance will
represent the FDA’s current thinking on this topic.
97
This attribute is applicable to most MDIs where a priming and repriming function is present.
98
If the counter is not accurately measuring the doses dispensed, the patient may dispense the drug and not receive a
dose because the device is empty (under counting). For additional information to avoid undercounting, see design
recommendations included in section III.B of the guidance for industry Integration of Dose-Counting Mechanisms
into MDI Drug Products (March 2003).

30
 Contains Nonbinding Recommendations
Draft — Not for Implementation
881 Continued – Table 4: Inhalation Products
882
Device Delivery Delivery to Product Initiation Dose Dose delivery
platform of the target preparation of dose delivery completion
intended site delivery progression
dose
Dry powder • Metered APSD 99 Dose Trigger Trigger Counter
inhaler (DPI) dose loading 100 mechanism mechanism accuracy 102
101
- both pre- • Emitted
metered and drug
device
metered
Nebulizers 103 • Total • Patient N/A Breath Breath Audible/visual
dose Interface synchroniza synchronizat /tactile
• Rate of 104 tion 106 ion 107 feedback
delivery • APSD 105
883
884
885
886
887

99
See footnote 96.
100
Single-dose DPIs are designed to require loading of doses and multi-dose DPIs are designed with internal
mechanisms to cycle to new dose blisters or load drug from an internal reservoir into a dosing chamber.
101
Additionally, some device features dictate the appropriate airflow for activation (e.g., triggering flow rate) and
may be considered an EDDO.
102
See footnote 98.
103
Applicable to inhalation solutions, inhalation suspensions, solutions for inhalation, and drugs for inhalation
solutions dosage forms, as opposed to inhalation powder and aerosol.
104
The amount of drug delivered to the lung will depend on the route of inhalation (e.g., full-face mask or
mouthpiece) as well as seal and resistance to flow. Additionally, different materials which may have different
electrostatic properties which can attract and capture smaller droplets, add sources of leak and dead space, and may
decrease respirable dose.
105
See footnote 96. For nebulizers, APSD is dependent on the patient interface, in addition to the formulation,
container closure system, and device delivery mechanism. See also footnote 95.
106
Some nebulizers are designed to attempt to nebulize only during patient inhalation. Design implementation of
synchronization includes both simple one-way valve (selective flow during inhalation or exhalation) and more
complex sensor-driven electromechanical actuation systems.
107
See footnote 106.

31
 Contains Nonbinding Recommendations
Draft — Not for Implementation
888 Table 5: Vaginal Systems (VS)
889
Delivery of Delivery to the target Product Initiation Dose delivery Dose
intended site preparation of dose progression delivery
dose delivery completion
Drug • VS physico-mechanical N/A N/A • VS physico- N/A
release rate properties, such as: 109 mechanical
and - Dimensions properties, such
consistency - Durometer/ as: 110
108
hardness - Dimensions
- Tensile strength - Durometer/
- % elongation/force hardness
at break - Tensile strength
- Compression - % elongation/
strength (including force at break
fatigue resistance) - Compression
- Twisting during strength
compression (including
- Seal integrity (if fatigue
applicable) resistance)
- Twisting during
• Applicator/inserter compression
physico-mechanical - Seal integrity (if
properties, such as: applicable)
- Dimensions
- Force needed for
delivery
- Assessment of
correct deployment
890
891
892

108
This includes the initial burst effect, if applicable. Release rate is a reflection of dose accuracy over time.
109
For vaginal systems, the VS physico-mechanical properties ensure that the vaginal system can be inserted,
remains in place for the duration of the use period, and can be removed. If these fail after the vaginal system is
inserted, the vaginal system may fall out, leading to an incomplete dose, or may cause physical harm to the user.
110
See footnote 109.

32
 Contains Nonbinding Recommendations
Draft — Not for Implementation
893 Table 6: Infusion Products
894
Device Delivery of Delivery to the Product Initiation Dose Dose
platform intended target site preparation of dose delivery delivery
dose delivery progression completion
Infusion • Flow rate • Needle/ Drug cartridge • Button • Adhesion Dose status
pumps 111, accuracy/ cannula depth insertion force activation force (if
112
consisten- • Connection (if applicable) force applicable)
113
cy stability to IV (mechani- • Dose status
• Dose or to separate cal)
accuracy administration • Dose
114
(if set for SQ, status 115
applicable) etc.
• Needle/
cannula
insertion force
Subdermal Dose Implantation Implant Applicator N/A Dose status
implants accuracy depth indicators compatibility ejection (if
(and or with force applicable)
applicators Drug applicator
if release rate (e.g.,
applicable) and dimensional
116
consisten- compatibility)
cy
895

111
When bolus feature is present the EDDOs for on-body injectors also apply.
112
The form factor of infusion pumps can vary significantly. However, this table is meant to apply to a wide array
of form factors such as body-worn, body-attached, and conventional infusion pumps. Additional EDDOs may exist
depending on the functionality and intended use of the infusion pump.
113
Accuracy and/or consistency depending on the drug characteristics, indication, and/or safety profile.
114
For pumps that include a bolus dose function, in addition to continuous infusions.
115
Dose status is applicable throughout the initiation, progress, and completion of the infusion; however, whether
dose status communication is achieved through audio, visual, tactile or other means is dependent on the pump
design. Furthermore, this guidance does not address the causes that might result in interruption of a dose such as
occlusion, motor failure, etc., since each pump design is different. Therefore, depending on the design and
technology of the pump, outputs other than dose status may be EDDOs for dose initiation, progression, and
completion.
116
The examples of potential EDDOs identified for subdermal implants and applicators are based on the two most
common forms: passive delivery moderated by polymer degradation and active delivery using an osmotic engine.

33
 Contains Nonbinding Recommendations
Draft — Not for Implementation
896 APPENDIX D – EXAMPLE OF A CONTROL STRATEGY FOR THE ESSENTIAL
897 DRUG DELIVERY OUTPUT OF NEEDLE EXTENSION LENGTH FOR AN
898 AUTOINJECTOR
899
900 The following is an example control strategy for the needle extension length for an autoinjector
901 assembled with the prefilled syringe (PFS) at the final finished combination product 117
902 manufacturing facility. The applicant is ultimately responsible for the control of the essential
903 drug delivery output (EDDO). This example shows how the applicant can leverage information
904 regarding the design and manufacturing from suppliers and purchasing agreements (e.g.,
905 certificate of analysis), which are part of the purchasing controls (21 CFR 820.50), to adequately
906 control the EDDO upstream.
907
908 In this example, supplier A is the syringe and needle manufacturer, supplier B is the autoinjector
909 subassembly manufacturer, and the applicant is the manufacturer of the final finished
910 combination product. The applicant performs the following manufacturing steps, which include:
911
912 (1) Manufacturing the primary container (PFS) – filling and stoppering the syringes from
913 supplier A with drug,
914
915 (2) Manufacturing the autoinjector – Inserting and securing the PFSs into autoinjector
916 subassemblies from supplier B, and
917
918 (3) Final packaging and labeling of autoinjector.
919
920 As shown in the table below, the length of the syringe and needle are not subject to change after
921 filling or assembly into the autoinjector, or over the autoinjector shelf life. Therefore, control of
922 the length of the syringe and needle is effective at supplier A (the syringe and needle
923 manufacturer). The needle cover retraction distance is also a dimensional function that is not
924 impacted by aging and subsequent assembly steps at the final manufacturing stage. Therefore,
925 control of the needle cover retraction distance is effective at supplier B.
926
927 The applicant provides an engineering analysis demonstrating that the needle cover retraction
928 distance specification is compatible with the syringe/needle and would yield an adequate needle
929 extension length. Therefore, this control is effective at the component level at supplier B.
930 Validation of the specification is through the engineering analysis performed by the applicant.
931
932 The applicant determines that the final needle extension length for the autoinjector is dependent
933 on the process steps to assemble the syringe into the subassemblies. Therefore, the applicant
934 provides a process risk assessment to support which steps impact this EDDO, the applicant’s in-
935 process work instructions relating to this process step, and validation data to show this process
936 step is controlled and yields autoinjectors with adequate needle extension lengths.
937
938 See sections VIII.B and VIII.C for control strategy information to include in applications.

117
For purposes of this guidance, the final finished combination product is the product intended for marketing and
submitted in the marketing application.

34
 Contains Nonbinding Recommendations
Draft — Not for Implementation
939 Table 1 – Example of a control strategy for needle extension length, an EDDO for an
940 autoinjector
941
Design Control at Control at Control at final Documentation used
attribute or supplier A 118 supplier manufacturer to verify the control
manufacturing B 119 at final
process step manufacturer/applic
ant and submitted
Length of 100% N/A Incoming Engineering drawing
syringe and dimensional acceptance per (#XXXX)
needle check of COA
overall
syringe/needle
In-process None
control (IPC)
check of
syringe
dimensions
Certificate of COA
analysis
(COA) for
each syringe
batch released
Needle cover N/A IPC Tolerance analysis • Engineering report
retraction dimensional ensuring needle explaining tolerance
distance check for extension is analysis and
autoinjector compatible with justification of CpK
components design to ‘X’ CpK values
/subassem- level • Engineering drawing
blies (Doc# XXXX) of
component A
• Engineering drawing
(Doc# XXXX) of
component B

118
In a marketing application, a proposed control strategy does not need to include the referenced documentation
(e.g., from sources other than the applicant) to verify the control. However, the Agency may request the referenced
documentation. Additionally, the Agency may request protocols, test methods, or work instructions for controls at
suppliers during review of the application depending on the criticality of the EDDO and the method of control
upstream (e.g., if the applicant is relying on the supplier’s control information). When requested, the supplier can
provide this information directly to the applicant or through reference to a master file with an appropriate Letter of
Authorization.
119
Ibid.

35
 Contains Nonbinding Recommendations
Draft — Not for Implementation
Design Control at Control at Control at final Documentation used
attribute or supplier A 118 supplier manufacturer to verify the control
manufacturing B 119 at final
process step manufacturer/applic
ant and submitted
Subassembly N/A 100% check Adequate change Purchasing control
assembly step to ensure control and agreement between
#x key purchasing control final manufacturer and
components agreements are in supplier B
were place to ensure
assembled notification of any
correctly changes to
and interact components
to assure
needle
extension
Step #x - N/A N/A Process risk Process risk
assembling assessment to assessment
subassemblies determine
and syringe assembly steps
impacting needle
extension
Push/snap force of • Work instruction
fixture • Validation of snap
force
Fixture height • Tolerance analysis of
fixture height and
impact on overall
needle extension
length
• Work instruction for
fixture set up and
calibration
Performed process Process validation
validation and report (includes
verified that device autoinjector extended
manufacturing needle length testing)
from PQ
engineering runs 120
met needle
extension
specifications
942
943

120
An engineering run is representative of the intended manufacturing process for the to-be-marketed product.

36