0256 infant formula supporting statement for GMP rule

Infant Formula Requirements
OMB Control No. 0910-0256
RIN: 0910-AF27
SUPPORTING STATEMENT, Part A
Terms of Clearance: None.
A. Justification
1. Circumstances Making the Collection of Information Necessary
Statutory requirements for infant formula under the Federal Food, Drug, and Cosmetic
Act (the FD&C Act) are intended to protect the health of infants and include a number of
reporting and recordkeeping requirements. Among other things, section 412 of the FD&C Act
(21 U.S.C. 350a) requires manufacturers of infant formula to establish and adhere to quality
control procedures, notify FDA when a batch of infant formula that has left the manufacturers'
control may be adulterated or misbranded, and keep records of distribution. We have issued
regulations to implement the FD&C Act's requirements for infant formula in 21 CFR parts 106
and 107, which covers the labeling of infant formula under the authority of section 403 of the
FD&C Act (21 U.S.C. 343). Under our regulations, the label of an infant formula must include
nutrient information and directions for use. The purpose of these labeling requirements is to
ensure that consumers have the information they need to prepare and use infant formula
appropriately.
This information collection approval request is for all information requirements imposed by the
regulations listed below:
21 CFR part 106, Infant Formula Requirements Pertaining To Current Good Manufacturing
Practice, Quality Control Procedures, Quality Factors, Records And Reports, And Notifications;
and
21 CFR part 107, Infant Formula.
2. Purpose and Use of the Information Collection
This information is used by the agency and industry to ensure the availability of safe
infant formula through compliance with the requirements associated with infant formula
development, production, and distribution. The information is also used by consumers when
purchasing, storing, and preparing infant formula as well as to confirm that the nutrient
requirements of the FD&C Act have been met.
Description of Respondents: Respondents to this information collection are manufacturers of
infant formula. Respondents are from the private sector (for-profit businesses).

2
3. Use of Improved Information Technology and Burden Reduction
We expect that 100% of the infant formula manufacturers will use electronic means for
the required recordkeeping. For petitions or notifications, we expect that 50% will be
electronically submitted. Because such submissions contain technical material specifically
related to the infant formula, respondents may choose their preferred format for submission as
long as the required information is included.
4. Efforts to Identify Duplication and Use of Similar Information
There is no duplicative information collection as a result of the infant formula
regulations. The data recorded are specific to the individual processing facilities. No other
regulation or information collection duplicates this effort. There are no similar data that can be
used or modified for use.
5. Impact on Small Businesses or Other Small Entities
Using the Small Business Administration’s definition of small business for food
manufacturers of 500 employees or fewer (up to 1,000 employees for some industries), and Dun
& Bradstreet data on firm size, FDA estimates that no small firms will be affected by the
provisions in this information collection. However, FDA does aid small businesses in complying
with its regulatory requirements through the agency’s Regional Small Business Representatives
and through the scientific and administrative staffs within the agency. FDA has provided a
Small Business Guide on the agency’s website at http://www.fda.gov/oc/industry/.
6. Consequences of Collecting the Information Less Frequently
FDA has concluded that the recordkeeping and reporting requirements are necessary to
ensure compliance with the current good manufacturing practice and quality control procedures
(including production aggregate control and distribution), quality factors, audits, and registration
and notification provisions. Records of actions taken due to each requirement are essential for
manufacturers to implement these provisions effectively. We believe these burdens are the
minimum necessary to meet the requirements of the FD&C Act.
7. Special Circumstances Relating to the Guidelines of 5 CFR 1320.5
There are no special circumstances for this collection of information.
8. Comments in Response to the Federal Register Notice and Efforts to Consult Outside the
Agency
8a. Publication in the Federal Register
In accordance with 5 CFR 1320.8(d), FDA provided opportunity for public comment on
the information collection requirements associated with its infant formula regulations in a
proposed rule that published July 9, 1996 (61 FR 36154). Another opportunity for public

3
comment on the information collection requirements came April 28, 2003 (68 FR 22341). Still
another opportunity for public comment was provided February 10, 2014 (79 FR 7934) regarding
the information collection provisions associated with infant formula. Finally, on June 10, 2014
(79 FR 33057), again provided opportunity for public comment regarding its infant formula
regulations, including the information collection requirements. Although some comments were
received, none pertained to the information collection provisions covered by the PRA and are
therefore not discussed in this document.
8b. Outside Consultation
Given the lack of data on current practices in the infant formula industry, we identified
individuals with experience who could inform us on industry’s current good manufacturing
practice and aide in the estimation of the burden of information collection. This information
served to supplement agency expertise regarding burden estimates associated with the regulatory
impact of the infant formula regulations. Specifically, on November 24, 2009, a phone meeting
was conducted between the infant formula GMP rule working group and a representative of
Fujitsu America with extensive experience in infant formula manufacturing. The meeting
consisted of a question and answer discussion on general issues regarding the manufacturing of
infant formula. Subsequent e-mail communications on follow up issues, including questions on
information collection, were made on 04/08/2011, 08/01/2011, and 09/02/2011. Furthermore,
one of CFSAN’s internal experts contacted three former colleagues in the infant formula industry
to anonymously answer a questionnaire regarding practices in infant formula manufacturing.
9. Explanation of Any Payment or Gift to Respondents
FDA does not provide any payment or gift to respondents.
10. Assurance of Confidentiality Provided to Respondents
Information submitted to FDA under the infant formula regulations may contain trade
secret and commercial confidential information. Only information that is releasable under the
agency’s regulations in 21 CFR part 20 would be released to the public. This information is also
safeguarded by Section 301(j) of the FD&C Act and would be protected from disclosure under
the Freedom of Information Act (FOIA) under sections 552(a) and (b) (5 U.S.C. 552(a) and (b)).
11. Justification for Sensitive Questions
This collection of information does not involve questions that are of a personally
sensitive nature.

12. Estimates of Annualized Burden Hours and Costs
12 a. Annualized Hour Burden Estimate
FDA estimates the burden of this collection as follows:

4
Table 1.—Estimated Annual Reporting Burden1
Row
No.

Activity;
21 CFR Section

1

Requirements for
Quality Factors GMS
Exemption; 106.96(c)
Requirements for
Quality Factors —PER
Exemption;
106.96(g)
New Infant Formula
Registration; 106.110
New Infant Formula
Submission; 106.120
TOTAL

2

3
4

1

No. of
Respondents

No. of
Responses per
Respondent

Total
Annual
Responses

Avg. Burden
per Response
(in hours)

Total
Hours

4

9(8.5)

34

20

680

1

34

34

12

408

4

9(8.5)

35

18

4

9 (8.5)

35

.5
(30 min.)
10

350
1,456
(1455.5)

Where necessary, numbers have been rounded to the nearest whole number.

It is estimated that that 34 exemptions will be submitted annually and that each
exemption will take 20 hours to assemble (Silverman Statement 2011). Therefore, 34
exemptions x 20 hours = 680 hours is the total estimated burden for §106.96(c), as presented in
row 1 of Table 1.
FDA estimates that annually the infant formula industry will submit a total of 35 PER
submissions: 34 exemption requests and the results of one PER study (Silverman Statement
2011). For the submission of the PER exemption, it is estimated that infant formula industry will
submit 34 exemptions per year and that each exemption will take supporting staff 12 hours to
prepare (Silverman Statement 2011). Therefore, 34 exemptions x 12 hour per exemption = 408
hours to fulfill the requirements of §106.96(g), as presented in row 2 of Table 1.
FDA estimates that, for each of the four firms in the infant formula industry, one senior
scientist or regulatory affairs professional will need 30 minutes to gather and record the required
information for an infant formula registration made pursuant to §106.110. The annual number of
registrations for a new infant formula and the number of firms that will make such registrations
is not known. However, it is estimated that, annually, the industry could register 35 new infant
formulas (Silverman Statement 2011), or an average of about nine (8.5) registrations per firm.
Therefore, to comply with §106.110, the total annual industry burden is 35 registrations x 30
minutes per registration = 17.5 (18) hours, as presented in row 3 of Table 1.
FDA estimates that, for each of the four firms in the infant formula industry, one senior
scientist or regulatory affairs professional will need ten hours to gather and record information
needed for infant formula submissions made pursuant to §106.120. This estimate includes the
time needed to gather and record the information the manufacturer uses to request an exemption
under §106.91(b)(1)(ii) of the final rule, which states that the manufacturer shall include the
scientific evidence that the manufacturer is relying on to demonstrate that the stability of the new
infant formula will likely not differ from the stability of formula with similar composition,
processing, and packaging for which there are extensive stability data. The annual number of

5
submissions for a new infant formula and the number of firms that will make such submissions is
not known. However, it is estimated that, annually, the industry could make submissions for 35
new infant formulas, or an average of about nine (8.5) submissions per firm (Silverman
Statement 2011). Therefore, to comply with §106.120, the total annual industry burden is 35
submissions x 10 work hours per submission = 350 hours, as presented in row 4 of Table 1.
Thus, the total annual submission burden is 1,455.5 (rounded to 1,456) hours.
Table 2. –Estimated Annual Recordkeeping Burden1
First Year Hourly Burden
Row
No.
1

2

3

4

5

6

7

Activity;
21 CFR Section
Production and In-Process
Control System;
106.6(c)(5) and
106.100(e)(1) and (e)(3)
Controls to Prevent
Adulteration due to
Automatic (mechanical or
electronic) Equipment;
106.35(c) and
106.100(f)(5), 106.94
Requirements for Quality
Factors for Eligible Infant
Formulas—Written Study
Report; 106.96(i)(1)(i) or
109.96(i)(1)(ii),
106.96(i)(4), and
106.100(p)(2)
Requirements for Quality
Factors For Eligible Infant
Formulas—Anthropometric
Data; 106.96(i)(1),
106.96(i)(3), and
106.96(i)(4)
Requirements for Quality
Factors For Eligible Infant
Formulas—Formula Intake;
106.96(i)(1),106.96(i)(3),
and 106.96(i)(4)
Requirements for Quality
Factors For Eligible Infant
Formulas—data plotting;
106.96(i)(1),106.96(i)(3),
and 106.96(i)(4)
Requirements for Quality
Factors For Eligible Infant
Formulas—data
comparison;
106.96(i)(1),106.96(i)(3),
and 106.96(i)(4)

No. of
Recordkeepers
5

No. of Records
per
Recordkeeper
1

Total
Annual
Records
5

Avg. Burden
per
Recordkeeping
40

Total
Hours

5

1

5

6,400

32,000

4

1

4

16

64

672

4

2688

0.5
(30 min.)

1,344

672

4

2688

0.25
(15 min.)

672

672

4

2,688

0.08
(5 min.)

215

672

4

2,688

0.08
(5 min.)

215

200

6
8

Quality Factors—Records;
106.96(i)(1)(iii) and
106.100(p)(2)

46

1

46

20

TOTAL (first year only)

920

35,630
Recurring Hourly Burden

Row
No.

21 CFR Section

9

Controls to prevent
adulteration caused by
facilities – testing for
radiological contaminants2;
106.20(f)(3)
Controls to prevent
adulteration caused by
facilities – recordkeeping
of testing for radiological
contaminants2; 106.20(f)(4)
and 106.100(f)(1)
Controls to prevent
adulteration caused by
facilities – testing for
bacteriological
contaminants
106.20(f)(3)
Controls to prevent
adulteration caused by
facilities – recordkeeping
of testing for
bacteriological
contaminants
106.20(f)(4) and
106.100(f)(1)
Controls to prevent
adulteration by equipment
or utensils; 106.30(d) and
106.100(f)(2)
Controls to prevent
adulteration by equipment
or utensils; 106.30(e)(3)(iii)
and 106.100(f)(3)
Controls to prevent
adulteration by equipment
or utensils; 106.30(f) and
106.100(f)(4)
Controls to prevent
adulteration due to
automatic (mechanical or
electronic) equipment;
106.35(c) and 106.100(f)(5)

10

11

12

13

14

15

16

Number of
Record
keepers
21

Annual
Frequency of
Recordkeeping
1

Total
Records

Hours per
Record

Total
Hours

21

1.5
(90 min.)

32

21

1

21

.08
(5 min.)

2

5

52

260

.08
(5 min.)

21

5

52

260

08
(5 min.)

21

5

52

260

0.21
(13 min.)

55

5

52

260

0.21
(13 min.)

55

5

52

260

0.19
(12 min.)

49

5

1

5

520

2,600

7
17

18

19

20

21

22

23

24

25

26

27
28

29

Controls to prevent
adulteration due to
automatic (mechanical or
electronic) equipment
106.35(c) and 106.100(f)(5)
Controls to prevent
adulteration caused by
ingredients, containers, and
closures; 106.40(g) and
106.100(f)(6)
Controls to prevent
adulteration during
manufacturing; 106.50 and
106.100(e)
Controls to prevent
adulteration from
microorganisms; 106.55(d),
106.100(e)(5)(ii), and
106.100(f)(7)
Controls to prevent
adulteration during
packaging and labeling of
infant formula; 106.60(c)
General quality controltesting
106.91(b)(1), 106.91(b)(2)
and 106.91(b)(3)
General quality control;
106.91(b)(1), 106.91(d),
and 106.100(e)(5)(i)
General quality control;
106.91(b)(2) 106.91(d), and
106.100(e)(5)(i)
General quality control;
106.91(b)(3) 106.91(d), and
106.100(e)(5)(i)
Audit plans and
procedures; ongoing review
and updating of audits;
106.94
Audit plans and procedures
– regular audits; 106.94

5

2

10

640

6,400

5

52

260

0.17
(10 min.)

44

5

52

260

0.2
(14 min.)

60

5

52

260

0.25
(15 min.)

65

1

12

12

.25
(15 min.)

3

4

1

4

2

8

4

52

208

0.15
(9 min.)

31

4

52

208

0.15
(9 min.)

31

4

52

208

0.15
(9 min.)

31

5

1

5

8

40

5

52

260

4

1040

Requirements for quality
factors for infant
formulas—written study
report; 106.96(b),
106.96(d), 106.100(p)(1) ,
106.100(q)(1), and 106.121
Requirements for quality
factors for infant
formulas—anthropometric
data; 106.96(b)(2),
106.96(d), and
106.100(p)(1)

1

1

1

16

16

112

6

672

0.5
(30 min.)

336

8
30

Requirements for quality
factors for infant
formulas—formula intake
106.96(b)(3) and 106.96(d),
and 106.100(p)(1)

112

6

672

0.25
(15 min.)

168

31

Requirements for quality
factors for infant
formulas—data plotting;
106.96(b)(4), 106.96(d),
and 106.100(p)(1)
Requirements for quality
factors for infant
formulas—data
comparison; 106.96(b)(5),
106.96(d), and
106.100(p)(1)
Requirements for quality
factors—per data
collection; 106.96(f)
Requirements for quality
factors—per written report;
106.96(f)
TOTAL Recurring Burden

112

6

672

0.08
(5 min.)

54

112

6

672

0.08
(5 min.)

54

1

1

1

8

8

1

1

1

1

1

32

33

34

TOTAL
1
2

11,225
(11,225.05)
46,855
(46,855.05)

Where necessary, numbers have been rounded to the nearest whole number.
This testing only occurs every four years.

The total one-time estimated burden imposed by this collection of information is 35,630
hours. Thereafter, the total annual estimated burden imposed by this collection of information is
12,680.55 (11,225.05 recordkeeping hours + 1, 455.5) hours. The estimated burden for this final
rule is based on “Evaluation of Recordkeeping Costs for Food Manufacturers,” Eastern Research
Group Task Order No. 5, Contract No. 223-01-2461.
For records pertaining to production and in-process controls, FDA estimates that, at most,
five plants will be required to develop production records to comply with §106.6(c)(5),
§106.100(e)(1), and §106.100(e)(3) (Zink Statement 2010). A team of two senior validation
engineers (or other similarly skilled employees) per plant (2 x 5 plants = 10 workers) will each
need to work 20 hours to provide sufficient initial baseline records and documentation to develop
records pertaining to production and in-process controls in order to comply with §106.6(c)(5)
and §106.100(e)(1) of the final rule, for an industry total of 200 hours (2 workers x 5 plants x 20
hours = 200 hours), as presented in row 1 of Table 2. For the purposes of this table, the number
of recordkeepers is shown as 5 (teams).
For the recordkeeping requirement of §106.35(c), in accordance with §106.100(f)(5),
FDA estimates that a team of ten senior validation engineers (or other similarly skilled
employees) per plant will need to work full time for the 16 weeks (640 work hours per person) to
provide sufficient initial records and documentation to comply with this section. The total
burden for ten senior validation engineers each working 640 hours is 6,400 per plant in the first
year (10 senior validation engineers x 640 hours = 6,400). For five plants, the total one time

9
hourly burden is 5 plants x 6,400 hours = 32,000 hours, as presented in row 2 of Table 2. For the
purposes of this table, the number of recordkeepers is shown as 5 (teams).
It is estimated that the data collection associated with a growth study performed to
comply with §106.96(i)(1) will be assembled into a written study report and that the study report
will be kept as a record in compliance with §106.96(i)(1)(i) or §106.96(i)(1)(ii), §106.96(i)(4),
and §106.100(p)(2). As noted, four growth studies of eligible infant formulas are estimated as a
result of this final rule. Therefore, it is estimated that four growth study reports will be generated
as a result of this rule. It is estimated that one report will require one senior scientist to work 16
hours to compile these data into a comprehensive report. Therefore, four growth study reports x
16 hours = 64 hours for compliance with §106.96(i)(1)(i) or §106.96(i)(1)(ii), as presented in
row 3 of Table 2.
A clinical growth study conducted according to the requirements of §106.96(b)(2) must
include the collection of anthropometric measurements of physical growth and formula intake,
and §106.96(b)(3) requires that the anthropometric measurements be taken six times during the
growth study. It is estimated that in a growth study of 112 infants, two nurses or other health
professionals with similar experience will need 15 minutes each per infant at each of the required
six times to collect and record the required anthropometric measurements. Therefore, 2 nurses x
.25 hours = .5 hour per infant, per visit, and .5 hour x 6 visits = 3 hours per infant. For 112
infants in a study, 3 hours x 112 infants = 336 hours to collect anthropometric information for
one growth study. For four growth studies, this burden is 1,344 hours (336 hours x 4 studies), as
presented in row 4 of Table 2.
In addition, it is estimated that one nurse will need 15 minutes per infant to collect and
record the formula intake information. That is, .25 hour x 6 visits = 1.5 hour per infant, and 1.5
hour per infant x 112 infants = 168 hours to collect information on formula intake for one growth
study. For four growth studies, this burden is 672 hours (168 hours x 4 studies), as presented in
row 5 of Table 2.
Section §106.96(b)(4) requires plotting each infant’s anthropometric measurements on
the CDC-recommended WHO growth charts. This task is estimated to take five minutes per
infant at each study visit. Therefore, six data plots x 112 infants = 672 total data plots, and 672
data plots x .08 hour per comparison = 53.75 total hours. For four growth studies, this burden is
215 hours (53.75 hours x 4 studies), as presented in row 6 of Table 2.
Finally, section §106.96(b)(5) requires that data on formula intake by the test group be
compared to that of the concurrent control group. FDA estimates that one nurse or other health
care professional with similar experience will need five minutes per infant, for each of the six
study visits, to fulfill the requirements of this section. Therefore, six comparisons of data x 112
infants = 672 data comparisons, and 672 data comparisons x .08 hour per comparison = 53.75
total hours. For four growth studies, this burden is 215 hours (53.75 hours x 4 studies), as
presented in row 7 of Table 2.
Final §106.100(p)(2) and §106.100(q)(2) require that, in accordance with §106.96(i)(4), a
manufacturer keep records demonstrating that an eligible infant formula fulfills one or more of

10
the criteria listed in §106.96(i)(1) and one or more of the criteria in §106.96(i)(2). It is estimated
that, for an eligible infant formula for which a growth study is performed, the records required by
§106.100(p)(2) are fulfilled by the growth study data collection and the study report and do not
represent an additional quantifiable hourly burden to these manufacturers (Silverman Statement
2011). In addition, it is estimated that the records required by §106.100(q)(2) are fulfilled by an
infant formula firm by virtue of the current requirement in 21 CFR 106.30(c)(2) to conduct a
PER study, and thus, this requirement does not represent an additional quantifiable hourly
burden (Silverman Statement 2011). For an eligible infant formula for which no growth study is
performed, the recordkeeping burden of §106.100(p)(2) is estimated to be 20 hours per record for
each of 46 estimated formulations due to the need for manufacturers to compile existing data into
a record. Therefore, 20 hours x 46 formulations = 920 hours for this subset of manufacturers to
comply with §106.100(p)(2), as presented in row 8 of Table 2. This 920 hours represents the
total industry burden for compliance with §106.100(p)(2). This burden is estimated also to cover
the requirements of §106.96(i)(1)(iii), which state that an eligible infant formula meets the
quality factor of normal physical growth if the scientific evidence on such infant formula
otherwise demonstrates that such formula supports normal physical growth.
Accordingly, as shown in Table 2, FDA estimates a total first year only hourly burden of
35,630 hours.
It is estimated that the requirement to test at least every four years for radiological
contaminants will represent a new collection of information for 21 infant formula plants (Zink
Statement 2010). In addition, it is estimated that collecting water for all testing in §106.20(f)(3)
takes between one and two hours (Zink Statement 2013). For the purposes of this analysis, it is
conservatively estimated that water collection takes, on average 1.5 hours and that water
collection occurs separately for each type of testing. It is estimated that performing the test
(collecting the information) will take 1.5 hours per test, every four years. Therefore, 1.5 hours
per plant x 21 plants = 31.5 total hours, every four years, as seen in row 9 of Table 2.
Furthermore, §106.20(f)(4) and §106.100(f)(1) require firms to make and retain records of the
frequency and results of water testing. For the 21 plants that are estimated not to currently test
for radiological contaminants, this burden is estimated to be five minutes per record every four
years. Therefore, .08 hour per record x 21 plants = 1.68 hours, every four years for the
maintenance of records of radiological testing, as seen on row 10 of Table 2.
It is estimated that the requirement to test weekly for bacteriological contaminants is a
new burden for five infant formula plants. It is estimated that performing the test (collecting the
information) will take five minutes per test once a week. Annually, this burden is .08 hours x 52
weeks = 4.16 hours per year, per plant, and 4.16 hours per plant x 5 plants = 20.8 total annual
hours, as seen on row 11 of Table 2. Furthermore, for the five plants that are estimated to not
currently test weekly for bacteriological contaminants, this burden is estimated to be five minutes
per record, every week. Therefore, .08 hour per record x 52 weeks =4.16hours, per plant for the
maintenance of records of bacteriological testing. Accordingly, 4.16 hours x 5 plants = 20.8
annual hours, as seen on row 12 of Table 2.
The regulations require that certain instruments be calibrated against a known reference
standard, and that records of these calibration activities be made and retained (§106.30(d) and

11
§100.100(f)(2)) ; these records will be kept at the plant level. FDA estimates that one senior
validation engineer (or other similarly skilled employee) for each of the five (at most) plants will
need to spend about 13 minutes per week to satisfy the ongoing calibration recordkeeping
requirements of §106.30(d) and §100.100(f)(2). Therefore, 5 record keepers x 52 weeks = 260
records; 260 records x .21 hour per record = 55 hours as the total industry annual burden, as
presented in row 13 of Table 2.
The regulations under §106.30(e)(3)(iii) and §106.100(f)(3)) require the making and
retaining records of the temperatures of each cold storage compartment. Based on expert
opinion, FDA estimates that five (at most) plants are not currently adhering to this recordkeeping
provision, and that at each of these five plants, compliance will require one senior validation
engineer (or other similarly skilled employee) about 13 minutes per week. Therefore, 5 record
keepers x 52 weeks = 260 records; 260 records x .21 hours per record = 55 hours as the total
industry annual burden, as presented in row 14 of Table 2.
Sections §106.30(f) and §100.100(f)(4) require the making and retention of records of
ongoing sanitation efforts. Based on expert opinion, FDA estimates that five (at most) plants are
not currently adhering to this recordkeeping provision, and that at each of these five plants,
compliance will require one senior validation engineer (or other similarly skilled employee)
about 12 minutes per week. Therefore, 5 record keepers x 52 weeks = 260 records; 260 records
x .19 hours per record = 49.4 hours as the total industry annual burden, as presented in row 15 of
Table 2.
There will be annual recordkeeping associated with §106.35(c) and §106.100(f)(5). It is
estimated that one senior validation engineer (or other similarly skilled employee) per plant will
need to work ten hours per week (520 work hours per year) to meet the ongoing recordkeeping
requirements of this section. For the estimated five (at most) plants not adhering to the
recordkeeping provisions of §106.35, the total annual burden for this provision is 520 hours per
plant x 5 plants = 2,600 annual hours, as shown in row 16 of Table 2. In addition, an infant
formula manufacturer will need to revalidate its systems when it makes changes to automatic
equipment. FDA estimates that such changes are likely to occur twice a year to any aspect of the
plant’s system, and that on each of the two occasions, a team of four senior validation engineers
(or other similarly skilled employees) per plant will need to work full time for four weeks (4
weeks x 40 hours per week = 160 work hours per person) to provide revalidation of the plant’s
automated systems sufficient to comply with this section. The total annual burden for four senior
validation engineers each working 160 hours twice a year is 1,280 hours ((160 hours x 2
revalidations) x 4 engineers =1280 total work hours), per plant. Therefore, 1280 hours per plant
x 5 plants = 6,400 annual hours, as shown on row 17 of Table 2.
Section 106.40(d) requires written specifications for ingredients, containers, and closures,
and is considered a collection of information. FDA estimates that the infant formula industry
already establishes written specifications for these components. However, the requirements of
§§106.40(g) and 106.100(f)(6) may represent new recordkeeping for five (at most) plants (Zink
Statement 2013). It is not possible to predict how often a specification will not be met or how
often documented reviews of reconditioned ingredients, closures, or containers will occur. FDA
estimates that, on average, one senior validation engineer per plant will work about 10 minutes a

12
week to fulfill the recordkeeping requirements of §106.40(g) and 106.100(f)(6). Therefore, 5
record keepers x 52 weeks = 260 records and 260 records x .17 hour = 45 total annual hours, as
presented in row 18 of Table 2.
Records pertaining to §106.50, the master manufacturing order and any changes to it, will
be kept at the plant level. It is not possible to predict how often changes to the master
manufacturing order will be made or how often deviations from the master manufacturing order
will occur. Based on expert opinion, FDA estimates that each year, 5 (at most) plants will
change a master manufacturing order and that, on average, one senior validation engineer for
each of the 5 (at most) plants will spend about 14 minutes per week on recordkeeping pertaining
to the master manufacturing order, as required by §106.50(a)(1) and §106.100(e). Thus, 5 record
keepers x 52 weeks = 260 records; 260 records x .23 hour = 60 hours as the total annual industry
burden, as presented in row 19 of Table 2.
Sections §106.55(d), §106.100(e)(5)(ii), and §106.100(f)(7)) require infant formula
manufacturers to make and retain records of the testing of infant formula for microorganisms.
Based on expert opinion, the agency estimates that these recordkeeping requirements represent a
new collection of information for, at most, five plants (Zink Statement 2013) and that one senior
validation engineer per plant will spend 15 minutes per week on recordkeeping pertaining to
microbiological testing. Thus, 5 record keepers x 52 weeks = 260 records; 260 records x .25
hour per record = 65 hours as the total annual industry burden, as presented in row 20 of Table 2.
Section §106.60 establishes requirements for the labeling of mixed-lot packages of infant
formula. The agency estimates that §106.60 will require infant formula diverters to label infant
formula packaging (such as packing cases) to facilitate product tracing and to keep specific
records of the distribution of these mixed lot cases. For the purposes of this analysis, it is
estimated that it may take one worker using manual methods 15 minutes, at most, to relabel one
case of infant formula, one time each month (.25 x 12 months = 3 annual hours), to meet the
requirements of §106.60(c)(2), as presented in row 21 of Table 2.
The regulations also establish on-going stability testing requirements (§106.91(b)(1),
§106.91(b)(2), and §106.91(b)(3)). It is estimated that the systems and processes of the formula
industry partially adhere to these provisions in that 80% of infant formula plants (17 of 21 plants)
conduct stability testing as specified in these provisions (Zink Statement 2013). For the 20% of
plants (4 of 21 plants) that do not conduct stability testing as specified in this provision, it is
estimated that these plants do conduct initial stability testing, but may not do so at the intervals
specified in this provision (Zink Statement 2013). For the purposes of this analysis, it is
estimated that the stability testing requirements of §106.91(b) represent a new burden of 2 annual
hours, per plant. Therefore, 2 hours x 4 plants = 8 annual hours to fulfill the testing requirements
of §106.91(b) as shown in row 22 of Table 2.
The requirements of §106.91(d) and §106.100(e)(5) to keep records of tests required
under §106.91(b)(1), §106.91(b)(2), and §106.91(b)(3) represent new information collections
for the four plants that are estimated not to be conducting all of the stability testing specified in
§106.91(b) (Zink Statement 2013). For the purposes of this analysis, FDA estimates that, for the
testing requirements in §106.91(b), one senior validation engineer per plant will spend about

13
nine minutes per week maintaining records to be in compliance with §106.91(d) and
§106.100(e)(5). Thus, 4 record keepers x 52 weeks = 208 records; 260 records x .15 hour per
record = 31.2 hours, per testing requirement, as the annual total industry burden, as presented in
lines 23, 24, and 25 of Table 2.
It is estimated that the ongoing review and updating of audit plans will require a senior
validation engineer eight hours per year, per plant. Therefore, 8 hours x 5 plants = 40 annual
hours to regularly review and update audit plans as shown in row 26 of Table 2.
The regulations do not mandate a frequency of auditing. For the purposes of this
analysis, FDA estimates that a manufacturer will choose to audit once per week. Each weekly
audit is estimated to require a senior validation engineer four hours, or 52 weeks x 4 hours = 208
hours per plant. Therefore, the total annual burden for the estimated five plants not currently
adhering to this provision to update audit plans is 208 hours x 5 plants = 1,040 hours, as shown
in row 27 of Table 2.
FDA estimates that, as a result of the regulations, the industry as a whole will perform
one additional growth study per year (Silverman Statement 2011), in accordance with §106.96.
The regulations require that several pieces of data be collected and maintained for each infant in
the growth study. It is estimated that the data collection associated with the growth study
performed to comply with §106.96(b) will be assembled into a written report and kept as a
record in compliance with §106.96(d) and §106.100(p)(1). Thus, it is estimated that one
additional growth study report will be generated as a result of the regulations, and that this report
will require one senior scientist to work 16 hours to compile the data into a study report.
Therefore, one growth study report x 16 hours = 16 annual hours for compliance with §106.96(d)
and §106.100(p)(1), as presented in row 28 of Table 2.
A study conducted according to the requirements of §106.96(b)(2) must include the
collection of anthropometric measurements of physical growth and information on formula
intake and §106.96(b)(3) requires that the anthropometric measurements be made at six times
during the growth study. It is estimated that in a growth study of 112 infants, two nurses or other
health professionals with similar experience will need 15 minutes per infant at each of the
required six times to collect and record the required anthropometric measurements. Therefore, 2
nurses x .25 hours = .5 hour per infant, per visit, and .5 hour x 6 visits = 3 hours per infant. For
112 infants in a study, 3 hours x 112 infants = 336 hours to collect anthropometric measurement
information, as presented in row 29 of Table 2. In addition, it is estimated that one nurse will
need 15 minutes per infant to collect and record the formula intake information. That is, .25 hour
x 6 visits = 1.5 hour per infant, and 1.5 hour per infant x 112 infants = 168 hours to collect
information on formula intake, as presented in row 30 of Table 2.
Section §106.96(b)(4) requires plotting each infant’s anthropometric measurements on
the CDC-recommended WHO Child Growth Standards. It is estimated that it will take five
minutes per infant to record the anthropometric data on the growth chart at each study visit.
Therefore, 112 infants x 6 data plots = 672 total data plots, and 672 data plots x .08 hour per
comparison = 53.75 total hours, as presented in row 31 of Table 2.

14
Section §106.96(b)(5) requires that data on formula intake by the test group be compared
to the intake of a concurrent control group. FDA estimates that, to fulfill the requirements of this
section, one nurse or other health care professional with similar experience will need five
minutes per infant for each of the six times anthropometric data are collected. Therefore, 6
comparisons of data x 112 infants = 672 data comparisons and 672 data comparisons x .08 hour
per comparison = 53.75 total hours, as presented in row 32 of Table 2.
The requirements of §106.96(f) state that a manufacturer shall meet the quality factor of
sufficient biological quality of the protein by establishing the biological quality of the protein in
the infant formula when fed as the sole source of nutrition using an appropriate modification of
the Protein Efficiency Ratio (PER) rat bioassay. Under §106.96(g)(1), a manufacturer of infant
formula may be exempt from this requirement if the manufacturer requests an exemption and
provides assurances, as required under §106.121, that changes made by the manufacturer to an
existing infant formula are limited to changing the type of packaging. A manufacturer may also
be exempt from this requirement under §106.100(g)(2), if the manufacturer requests an
exemption and provides assurances, as required under §106.121, that demonstrates, to FDA’s
satisfaction, that the change to an existing formula does not affect the bioavailability of the
protein. Finally, a manufacturer of infant formula may be exempt from this requirement under
§106.96(g)(3) if the manufacturer requests an exemption and provides assurances, as required
under §106.121(i), that demonstrate that an alternative method to the PER that is based on sound
scientific principles is available to show that the formula supports the quality factor for the
biological quality of the protein. It is estimated that these requirements represent two
information collections: submission of the PER results or submission of a request for an
exemption when appropriate. FDA estimates that annually the infant formula industry will
submit a total of 35 PER submissions: 34 exemption requests and the results of one PER study
(Silverman Statement 2011).
A PER study conducted according to AOAC Official Method 960.48 will be 28 days in
duration. It is estimated that there will be 10 rats in the control and test groups (20 rats total) and
that food consumption and body weight will be measured at day zero and at 7-day intervals
during the 28-day study period (a total of five records per rat). It is further estimated that
measuring and recording food consumption and body weight will take five minutes per rat.
Therefore, 20 rats x 5 records = 100 records; 100 records x .08 hour minutes per record = 8 hours
to fulfill the requirements of §106.96(f). Furthermore, it is estimated that a report based on the
PER study will be generated and that this study report will take a senior scientist one hour to
generate. Therefore a total of 9 hours will be required to fulfill the requirements for §106.96(f):
8 hours for the PER study and data collection, and 1 hour for the development of a report based
on the PER study, as presented in lines 33 and 34 of Table 2. Therefore, the total recurring
recordkeeping burden is 11,225.05 hours.

15
12b. Annualized Cost Burden Estimate
Table 3.—Annualized Cost Burden Estimate
First Year Only
21 CFR Part

Type of Respondent

Total Burden
Hours

Total
Respondent
Costs
$46.26
$1,480,320
$46.26
$9,252
$54.96
$188,512.80
$1,678,084.80

§106.35(c), §106.100(f)(5)
§106.100(e)(1), (e)(3)
§106.96
Total
Annual Cost
21 CFR Part

Senior Validation Engineer
Senior Validation Engineer
Scientist

§106.20
§106.30(f), §106.100(f)(5)
§106.35(c), §106.100(f)(5)
§106.40(g), §106.100(f)(6)
§106.50, §106.100(e)
§106.55,
§106.100(e)(5)(ii), (f)(7)
§106.60
§106.91(b),§106.91(d),
§106.100(e)
§106.94
§106.96(b),(d),
§1006.100(p)
§106.96(c)
§106.96(f)
§106.96(g)
§106.110
§106.120

Scientist
Senior Validation Engineer
Senior Validation Engineer
Senior Validation Engineer
Senior Validation Engineer
Senior Validation Engineer

74.78
159.4
9,000
45
60
65

$46.26
$46.26
$46.26
$46.26
$46.26
$46.26

Total
Respondent
Costs
$3,459.32
$7,373.84
$416,340
$2,081.70
$2,775.60
$3,006.90

Line Worker
Scientist

3
101.6

$26.24
$46.26

$78.42
$4,700.02

Senior Validation Engineer
Scientist

1,080
627.5

$46.26
$54.96

$49,960.80
$34,487.40

680
9
408
17.5
350

$54.96
$54.96
$27.23
$27.23
$54.96

$37,372.80
$494.64
$11,109.84
$476.53
$19,236

Total Cost

Type of Respondent

Scientist
Scientist
Support Staff
Support Staff
Scientist

32,000
200
3,430

Hourly Wage
Rate

Total Burden
Hours

Hourly Wage
Rate

$592,953.81

16
One-Time Costs
For the recordkeeping requirement of §106.35(c), in accordance with §106.100(f)(5),
FDA estimates that a team of ten senior validation engineers per plant will need to work full
time for the 16 weeks (640 work hours per person) to provide sufficient initial records and
documentation to comply with this section. FDA estimates average compensation for a senior
validation engineer to be about $46.26 per hour, including overhead (Bureau of Labor Statistics
a). The total cost for ten senior validation engineers each working 640 hours at $46.26 per hour
is $301,760 per plant in the first year (10 senior validation engineers x 640 hours = 6,400, and
6,400 work hours x 5 (at most)plants = 32,000 hours, and $46.26/hour x 32,000 hours=
$1,480,320.
It is estimated that a senior engineer will need a total of 200 hours to establish records
that include complete information relating to the production and control of the production
aggregate, (§106.100(e)(1)). A manufacturer is required to maintain records pertaining to
current good manufacturing practice, (§106.100(e)(3)). At a wage rate of $46.26 per hour
(Bureau of Labor Statistics a), this cost is estimated to be $46.26 x 200 = $9,252. It is estimated
that the one-time recordkeeping burden of 106.96(i)(4) is fulfilled by scientists working for a
total of 3,215 hours to gathering data into records. At a wage rate of $54.96 per hour, the total
one-time cost for scientists is 3,215 work hours x $54.96 = $188,512.80 (Bureau of Labor
Statistics b). Accordingly, the total one-time labor cost is $1,678,084.80.
Recurring Costs
It is estimated that scientists will have to work 74.78 total hours to comply with the
testing requirements of §106.20. At a wage rate of $46.26/hour (Bureau of Labor Statistics a),
the labor cost of this information collection is estimated to be $46.26 x 74.78 hours = $3,459.32.
It is estimated that fulfilling the recordkeeping requirements of §106.30(f) will require
senior validation engineers to spend 159.4 hours on these requirements. At a wage rate of
$46.26/hour, the labor cost of this information collection is estimated to be $46.26 x 159.4 =
$7,373,84.
It is estimated that fulfilling the recordkeeping requirements of §106.35 will require
senior validation engineers to work a total of 9,000 hours annually. At a wage of $46.26/per
hour), this cost is estimated to be 9,000 hours x $46.26/hour = $416,340.
It is estimated that fulfilling the recordkeeping requirements of §106.40(g) will require
senior validation engineers to work a total of 46.26 hours annually. At a wage of $46.26/per
hour, this cost is estimated to be 446.26 hours x $46.26/hour = $2,081.70.
It is estimated that fulfilling the recordkeeping requirements of §106.50 will require
senior validation engineers to work a total of 60 hours annually. At a wage of $46.26/per hour,
this cost is estimated to be 60 hours x $46.26/hour = $2,775.60.

17
It is estimated that fulfilling the recordkeeping requirements of §106.55 will require
senior validation engineers to work a total of 65 hours annually. At a wage of $46.26/per hour,
this cost is estimated to be 65 hours x $46.26/hour = $3,006.90.
It is estimated that it may take one worker, at most, 15 minutes to re-label one case of
infant formula, one time each month (.25 x 12 months = 3 annual hours), using manual methods,
to meet the requirements of §106.60(c)(2). At a wage of $26.24 per hour, including overhead
(Bureau of Labor Statistics c), the maximum cost of meeting the requirement of §106.60(c)(2) is
$26.24 x 3 = $78.42 per year.
It is estimated that the testing requirements of §106.91 will require scientists to work at
total of 101.6 hours to fulfill the information collection requirements of this section. At a wage
of $46.26/hour, the total cost is $46.26 x 101.6 = $4,700.02. It is estimated that the ongoing
review and updating of audit plans will require senior validation engineers to spend 1,080 hours
annually at $46.26 per hour to fulfill the requirements of §106.94. Therefore, 1.080 hours x
$46.26/hour = $49,960.80
It is estimated that it will take scientists 627.5 hours annually at $54.96 per hour to fulfill
the requirements of §106.96(b) and (d). Therefore, 627.5 hours x $54.96/hour = $34,487.40.
The agency estimates that scientists will spend 680 hours annually at a wage of $54.96
per hour to comply with the requirements of §106.96(c). Therefore, the cost to comply with
§106.96(c) is 680 x $54.96 = $37,472.80.
For the submission of the PER exemption for non-eligible infant formulas, it is estimated
that industry will submit 34 exemptions per year and that each exemption will take support staff
12 hours (Silverman Statement 2011) at a wage of $27.23 per hour (Bureau of Labor Statistics
c). Therefore, 12 hours x $27.23 wage per hour = $324 per exemption; $324 per exemption x 34
exemptions = $11,109.84 annually to fulfill the requirements of §106.96(g). To comply with
§106.110, the total annual industry burden is 35 registrations x 30 minutes per registration = 17.5
hours, and $27.23 per hour supporting staff wage x 17.5 hours = $476.53 .
It is estimated that, annually, the industry could make submissions for 35 new infant
formulas, or an average of seven submissions per firm (Silverman Statement 2011), pursuant to
§106.120. At a wage of $54.96 per hour, one submission is estimated to cost $549.60, or $54.96
hourly wage x 10 hours of work per submission = $549.60. Therefore, to comply with §106.120,
the total annual industry burden is 35 submissions x $549.60 per submission = $19,236.
The total annual cost related to information collection is $592,953.81
13. Estimates of Other Total Annual Costs to Respondents and/or Recordkeepers/Capital Costs
There are no start-up, operating or maintenance costs associated with this information
collection.

18
14. Annualized Cost to the Federal Government
The total cost to the Federal government is $860,473.90, calculated by adding the total
labor cost of $163,800 for the information collection requirements imposed by the new
regulations to the current cost of $696,673.90.
Specifically, each registration, new infant formula notification, infant formula
submission, quality factor submission, or verification submission will require evaluation by
subject matter experts who organize the evaluation and prepare the final documents. On average,
there will be 2 subject matter experts (SME), a reviewer and a Consumer Safety Officer; the
hours devoted will vary depending on the task. The reviewer would be a GS-14 or 15 level (a
medical officer, a methods expert, etc.), with an average cost of $63/hour. The Consumer Safety
Officer would be a GS-13 or 14 at an average cost of $5/hour. Further, the labor cost to review a
registration submitted in accordance with §106.110 is estimated to be about $85.50 ((.5hour *
$63/reviewer hour) + (1 hour * $54/CSO hour)). For an annual average of 35 registrations, the
total labor cost is 35 registrations x $85.50 = $2,992.50 per year. The labor cost to review a new
infant formula notification submitted in accordance with §106.120 is estimated to be about
$4,680 (40 hours * $63/review hour) + (40 hours * $54/CSO hour)). For an annual average of
35 submissions, the total labor cost is $163,800 per year.
15. Explanation for Program Changes or Adjustments
This information collection request revises the currently approved collection under OMB
Control No. 0910-0256. The new regulatory provisions impose a “first-year” burden of 35,630,
and a subsequent recurring burden of 11,225. While this results in an immediate increase of
48,311 burden hours, ultimately the agency estimates that the recurring burden will result in a
decrease of 18,977 hours. Similarly, the number of annual responses is increased by 16,683.
Upon satisfaction of the “first-year” requirements, however, the agency estimates there will be a
drop in the annual responses to 6,091.
16. Plans for Tabulation and Publication and Project Time Schedule
The agency has no plans for publication of information from this information collection.
17. Reason(s) Display of OMB Expiration Date is Inappropriate
Approval to not display the expiration date of OMB approval is not being sought.
18. Exceptions to Certification for Paperwork Reduction Act Submissions
There are no exceptions to the certification.

19
References:
1. Bureau of Labor Statistics a. Occupational Employment and Wages, 19-1012 Food
Scientists and Technologists. May 2012, Washington, DC; available at
http://www.bls.gov/oes/current/oes191012.htm, accessed April 5, 2013.
2. Bureau of Labor Statistics b. Occupational Employment and Wages, 19--1029 Biological
Scientists, All Others. May 2010, Washington, DC; available at
http://www.bls.gov/oes/current/oes191029.htm, accessed April 5, 2013.
3. Bureau of Labor Statistics c. Occupational Employment and Wages, 19-4011
Agricultural and Food Science Technicians. May 2012, Washington, DC (available at
http://www.bls.gov/oes/current/oes194011.htm, accessed April 5, 2013.
4. Bureau of Labor Statistics d. Occupational Employment and Wages, 43-4199 Information
and Record Clerks, All Other. May 2010, Washington, DC; available at
http://www.bls.gov/oes/current/oes434199.htm, accessed April 5, 2013.
5. Silverman, Benson. Statement of Benson M. Silverman, M.D. “Industry Efforts to Meet
Quality Factor Requirements,” (2011).
6. Zink, Don. Statement of Donald L. Zink, Ph. D.: Infant Formula Manufacturing
Practices 2010.