# Safety Standard for Clothing Storage Units

> Briefs, arguments, decisions, and more.

URL: https://www.frixlaw.com/law-library/documents/fr%3A2022-01689

## Record

- **Collection:** Federal Register
- **Document type:** Proposed Rule
- **Published:** February 3, 2022
- **Citation:** 87 FR 6246

## Text

CONSUMER PRODUCT SAFETY COMMISSION
16 CFR Parts 1112 and 1261
[Docket No. CPSC-2017-0044]
Safety Standard for Clothing Storage Units

AGENCY:

Consumer Product Safety Commission.

ACTION:

Notice of proposed rulemaking.

SUMMARY:

The U.S. Consumer Product Safety Commission (Commission or CPSC) has determined preliminarily that there is an unreasonable risk of injury and death, particularly to children, associated with clothing storage units (CSUs) tipping over. To address this risk, the Commission proposes a rule addressing the stability of CSUs. Specifically, the proposed rule would require CSUs to be tested for stability, exceed minimum stability requirements, be marked and labeled with safety information, and bear a hang tag providing performance and technical data about the stability of the CSU. The Commission issues this proposed rule under the authority of the Consumer Product Safety Act (CPSA). The Commission requests comments about all aspects of this notice, including the risk of injury, the proposed requirements, alternatives to the proposed rule, and the economic impacts of the proposed rule and alternatives.

DATES:

Submit comments by April 19, 2022.

ADDRESSES:

Direct comments related to the Paperwork Reduction Act aspects of the proposed rule to the Office of Information and Regulatory Affairs, the Office of Management and Budget, Attn: CPSC Desk Officer, fax to: 202-395-6974, or email
oira_submission@omb.eop.gov.
Submit other comments, identified by Docket No. CPSC-2017-0044, by any of the following methods:

Electronic Submissions:
Submit electronic comments to the Federal eRulemaking Portal at:
https://www.regulations.gov.
Follow the instructions for submitting comments. CPSC does not accept comments submitted by electronic mail (email), except through
https://www.regulations.gov,
and as described below. CPSC encourages you to submit electronic comments by using the Federal eRulemaking Portal, as described above.

Mail/Hand Delivery/Courier Written Submissions:
Submit comments by mail/hand delivery/courier to: Division of the Secretariat, Consumer Product Safety Commission 4330 East-West Highway, Bethesda, MD 20814; telephone: (301) 504-7479. Alternatively, as a temporary option during the COVID-19 pandemic, you can email such submissions to:
cpsc-os@cpsc.gov.

Instructions:
All submissions must include the agency name and docket number for this notice. CPSC may post all comments without change, including any personal identifiers, contact information, or other personal information provided, to:
https://www.regulations.gov.
Do not submit electronically: Confidential business information, trade secret information, or other sensitive or protected information that you do not want to be available to the public. If you wish to submit such information, please submit it according to the instructions for mail/hand delivery/courier written submissions.

Docket:
To read background documents or comments regarding this proposed rulemaking, go to:
https://www.regulations.gov,
insert docket number CPSC-2017-0044 in the “Search” box, and follow the prompts.

FOR FURTHER INFORMATION CONTACT:

Kristen Talcott, Project Manager, U.S. Consumer Product Safety Commission, 5 Research Place, Rockville, MD 20852; telephone (301) 987-2311; email:
KTalcott@cpsc.gov.

SUPPLEMENTARY INFORMATION:

I. Background

CSUs are freestanding furniture items, typically used for storing clothes. Examples of CSUs include chests, bureaus, dressers, chests of drawers, drawer chests, door chests, chifforobes, armoires, and wardrobes. CPSC is aware of numerous deaths and injuries resulting from CSUs tipping over, particularly onto children. CPSC identified 226 fatalities associated with CSUs tipping over that were reported to have occurred between January 1, 2000 and December 31, 2020.
1

Of these, 193 (85 percent) involved children (
i.e.,
under 18 years old), 11 (5 percent) involved adults (
i.e.,
18 to 64 years old), and 22 (10 percent) involved seniors (
i.e.,
65 years and older). In addition, there were an estimated 78,200 nonfatal CSU tip-over injuries that were treated in U.S. hospital emergency departments (EDs) between January 1, 2006 and December 31, 2019. Of these, an estimated 56,400 (72 percent) involved children, and the remaining estimated 21,800 (28 percent) involved adults and seniors.

1
Reporting is considered incomplete for the years 2018-2020 because reporting is ongoing.

To address the hazard associated with CSU tip overs, the Commission has taken several steps. In June 2015, the Commission launched the Anchor It! campaign. This educational campaign includes print and broadcast public service announcements; information distribution at targeted venues, such as childcare centers; social media; blog posts; videos; and an informational website (
www.AnchorIt.gov
). The campaign explains the nature of the risk, provides safety tips for avoiding furniture and television tip overs, and promotes the use of tip restraints to anchor furniture and televisions.

In addition, CPSC's Office of Compliance and Field Operations has investigated and recalled CSUs. Between January 1, 2000 and March 31, 2021, 40 consumer-level recalls occurred to address CSU tip-over hazards. The recalled products were responsible for 328 tip-over incidents, including reports of 149 injuries and 12 fatalities.
2

These recalls involved 34 firms and affected approximately 21,500,000 CSUs.

2
For the remaining incidents, either no injury resulted from the incident, or the report did not indicate whether an injury occurred.

In 2016, CPSC staff prepared a briefing package on furniture tip overs, looking at then-current levels of compliance with the voluntary standards, and the adequacy of the voluntary standards.
3

3
Massale, J., Staff Briefing Package on Furniture Tipover, U.S. Consumer Product Safety Commission (2016), available at:
https://www.cpsc.gov/s3fs-public/Staff%20Briefing%20Package%20on%20Furniture%20Tipover%20-%20September%2030%202016.pdf.

In 2017, the Commission issued an advance notice of proposed rulemaking (ANPR), discussing the possibility of developing a rule to address the risk of injury and death associated with CSU tip overs. 82 FR 56752 (Nov. 30, 2017).
4

The ANPR began a rulemaking proceeding under the CPSA (15 U.S.C. 2051-2089). CPSC received 18 comments during the comment period, as well as five additional correspondences after the comment period, which staff also considered.

4
The briefing package supporting the ANPR is available at:
https://www.cpsc.gov/s3fs-public/ANPR%20-%20Clothing%20Storage%20Unit%20Tip%20Overs%20-%20November%2015%202017.pdf?5IsEEdW_Cb3ULO3TUGJiHEl875Adhvsg. After issuing the ANPR, the Commission extended the comment period on the ANPR. 82 FR 2382 (Jan. 17, 2018).

The Commission is now issuing a notice of proposed rulemaking (NPR), proposing to establish requirements for CSU stability.
5

The information discussed in this preamble is derived

from CPSC staff's briefing package for the NPR, which is available on CPSC's website at:
https://www.cpsc.gov/s3fs-public/Proposed%20Rule-%20Safety%20Standard%20for%20Clothing%20Storage%20Units.pdf.
This preamble provides key information to explain and support the rule; however, for a more comprehensive and detailed discussion, see the NPR briefing package.

5
The Commission voted 4-0 to approve this notice.

II. Statutory Authority

CSUs are “consumer products” that the Commission can regulate under the authority of the CPSA.
See
15 U.S.C. 2052(a)(5). Section 7 of the CPSA authorizes the Commission to issue a mandatory consumer product safety standard that consists of performance requirements or requirements that the product be marked with, or accompanied by, warnings or instructions.
Id.
2056(a). Any requirement in the standard must be “reasonably necessary to prevent or reduce an unreasonable risk of injury” associated with the product.
Id.
Section 7 requires the Commission to issue such a standard in accordance with section 9 of the CPSA.
Id.

Section 9 of the CPSA specifies the procedure the Commission must follow to issue a consumer product safety standard under section 7.
Id.
2058. Under section 9, the Commission may initiate rulemaking by issuing an ANPR or NPR.
Id.
2058(a). As noted above, the Commission issued an ANPR on CSU tip overs in November 2017. 82 FR 56752 (Nov. 30, 2017). When issuing an NPR, the Commission must comply with section 553 of the Administrative Procedure Act (5 U.S.C. 553), which requires the Commission to provide notice of a rule and the opportunity to submit written comments on it. 15 U.S.C. 2058(d)(2). In addition, the Commission must provide interested parties with an opportunity to make oral presentations of data, views, or arguments.
Id.

Under section 9 of the CPSA, an NPR must include the text of the proposed rule, any alternatives the Commission proposes, and a preliminary regulatory analysis. 15 U.S.C. 2058(c). The preliminary regulatory analysis must include:

• A preliminary description of the potential costs and benefits of the rule, including costs and benefits that cannot be quantified, and the analysis must identify who is likely to receive the benefits and bear the costs;

• a discussion of the reasons any standard or portion of a standard submitted to the Commission in response to the ANPR was not published by the Commission as the proposed rule or part of the proposed rule;

• a discussion of the reasons for the Commission's preliminary determination that efforts submitted to the Commission in response to the ANPR to develop or modify a voluntary standard would not be likely, within a reasonable period of time, to result in a voluntary standard that would eliminate or adequately reduce the risk of injury addressed by the proposed rule; and

• a description of alternatives to the proposed rule that the Commission considered and a brief explanation of the reason the alternatives were not chosen.

Id.

In addition, to issue a final rule, the Commission must make certain findings and include them in the rule.
Id.
2058(f)(1), (f)(3). Under section 9(f)(1) of the CPSA, before promulgating a consumer product safety rule, the Commission must consider, and make appropriate findings to be included in the rule, concerning the following issues:

• The degree and nature of the risk of injury the rule is designed to eliminate or reduce;

• the approximate number of consumer products subject to the rule;

• the need of the public for the products subject to the rule and the probable effect the rule will have on the cost, availability, and utility of such products; and

• the means to achieve the objective of the rule while minimizing adverse effects on competition, manufacturing, and commercial practices.

Id.
2058(f)(1). Under section 9(f)(3) of the CPSA, the Commission may not issue a consumer product safety rule unless it finds (and includes in the rule):

• The rule, including the effective date, is reasonably necessary to eliminate or reduce an unreasonable risk of injury associated with the product;

• that issuing the rule is in the public interest;

• if a voluntary standard addressing the risk of injury has been adopted and implemented, that either compliance with the voluntary standard is not likely to result in the elimination or adequate reduction of the risk or injury, or there is unlikely to be substantial compliance with the voluntary standard;

• that the benefits expected from the rule bear a reasonable relationship to its costs; and

• that the rule imposes the least burdensome requirement that prevents or adequately reduces the risk of injury.

Id.
2058(f)(3). At the NPR stage, the Commission is making these findings on a preliminary basis to allow the public to comment on the findings.

Section 9(g)(2) of the CPSA allows the Commission to prohibit manufacturers of a consumer product from stockpiling products subject to a consumer product safety rule to prevent manufacturers from circumventing the purpose of the rule. 15 U.S.C. 2058(g)(2). The statute defines “stockpiling” as manufacturing or importing a product between the date a rule is promulgated and its effective date at a rate that is significantly greater than the rate at which the product was produced or imported during a base period ending before the date the rule was promulgated.
Id.
The Commission is to define what constitutes a “significantly greater” rate and the base period in the rule addressing stockpiling.
Id.

Section 27(e) of the CPSA authorizes the Commission to issue a rule to require manufacturers of consumer products to provide “such performance and technical data related to performance and safety as may be required to carry out the purposes of [the CPSA].” 15 U.S.C. 2076(e). The Commission may require manufacturers to provide this information to the Commission or, at the time of original purchase, to prospective purchasers and the first purchaser for purposes other than resale, as necessary to carry out the purposes of the CPSA.
Id.
Section 2(b) of the CPSA states the purposes of the CPSA, including:

• Protecting the public from unreasonable risks of injury associated with consumer products; and

• assisting consumers in evaluating the comparative safety of consumer products.

Id.
2051(b)(1), (b)(2).

III. The Product and Market

A. Description of the Product

The proposed rule defines a “CSU” as a freestanding furniture item, with drawer(s) and/or door(s), that may be reasonably expected to be used for storing clothing, that is greater than or equal to 27 inches in height, and that has a total functional volume of the closed storage greater than 1.3 cubic feet and greater than the sum of the total functional volume of the open storage and the total volume of the open space. Common names for CSUs include, but are not limited to: Chests, bureaus, dressers, armoires, wardrobes, chests of drawers, drawer chests, chifforobes, and

door chests. CSUs are available in a variety of designs (
e.g.,
vertical or horizontal dressers), sizes (
e.g.,
weights and heights), dimensions, and materials (
e.g.,
wood, plastic, leather, manufactured wood or fiber board). Consumers may purchase CSUs that have been assembled by the manufacturer, or they may purchase CSUs as ready-to-assemble furniture.

The proposed definition includes several criteria to help distinguish CSUs from other furniture. As freestanding furniture items, CSUs remain upright without requiring attachment to a wall, when fully assembled and empty, with all extension elements closed. As such, built-in units or units intended to be permanently attached to a building structure (other than by tip restraints) are not considered freestanding. In addition, CSUs are typically intended and used for storing clothing and, therefore, they are commonly used in bedrooms. However, consumers may also use CSUs in rooms other than bedrooms and to store items other than clothing in them. For this reason, whether a product is a CSU depends on whether it meets the criteria in the proposed definition, rather than what the name of the product is or what is the marketed use for the product. The criteria in the proposed definition regarding height and closed storage volume (
i.e.,
storage space inside a drawer or behind an opaque door) aim to address the utility of a unit for holding multiple clothing items. Some examples of furniture items that, depending on their design, may not meet the criteria in the proposed definition and, therefore, may not be considered CSUs are: Shelving units, office furniture, dining room furniture, laundry hampers, built-in closets, and single-compartment closed rigid boxes (storage chests).

CSUs may be marketed, packaged, or displayed as intended for children 12 years old and younger. Examples of such products include CSUs with pictures or designs on them that would appeal to children; CSU designs that would be useful for children; or CSUs that are part of a matching set with a crib, or similar infant product. However, CSUs are more commonly general-use products that are not specifically intended for children 12 years old and younger. The proposed rule applies to both children's products and non-children's products.

B. The Market

CPSC staff estimated the annual revenues and shipments of CSUs, using estimates of manufacturer and importer revenue, and estimated sales, by using data on retail sales. The shipment value of chests of drawers and dressers combined for an estimated $5.15 billion in 2018, and combined shipments of dressers and chests totaled 43.6 million units. Average manufacturer shipment value was $118 per unit in 2018 (about $104 for chests of drawers and $144 for dressers).

Retail prices of CSUs vary substantially. The least expensive units retail for less than $100, while more expensive units may retail for several thousand dollars. The estimated retail value of U.S. bedroom furniture sales in 2019 totaled $60.3 billion, of which $20.8 billion was sales of closets (which likely includes wardrobes and armoires), nightstands (some of which may be considered CSUs), and dressers (which likely includes chests of drawers).

According to data from the U.S. Census Bureau, in 2017, there were a total of 3,404 firms classified in the North American Industrial Classification System (NAICS) as non-upholstered wood household furniture manufacturing, upholstered household furniture manufacturing, metal household furniture manufacturing, or household furniture (except wood and metal) manufacturing. Of these firms, 2,024 were primarily categorized in the non-upholstered wood furniture category. However, these categories are broad and include manufacturers of furniture other than CSUs, such as tables, chairs, bed frames, and sofas. As such, it is likely that not all of the firms in these categories manufacture CSUs. Production methods and efficiencies vary among manufacturers; some use mass production techniques, and others manufacture their products one at a time or on a custom-order basis.

The number of U.S. firms that are primarily classified as manufacturers of non-upholstered wood household furniture has declined over the last few decades, as retailers have turned to international sources of CSUs and other wood furniture. Additionally, some firms that formerly produced all of their CSUs domestically have shifted production to foreign plants. More than half (64 percent) of the value of apparent consumption of non-upholstered wood furniture (net imports plus domestic production for the U.S. market) in 2019 was comprised of imported furniture, which may be true for CSUs as well. In addition to manufacturers, according to the Census Bureau data, in 2017, there were 5,117 firms involved in household furniture importation and distribution. According to the Census Bureau, there were 13,826 furniture retailers in 2017. Wholesalers and retailers may obtain their products from domestic sources or import them from foreign manufacturers.

IV. Risk of Injury

A. Incident Data
6

6
For more details about incident data, see Tab A of the NPR briefing package.

CPSC staff analyzed reported fatalities, reported nonfatal incidents and injuries, and calculated national estimates of injuries treated in EDs that were associated with CSU instability or tip overs. Each year, CPSC issues an annual report on furniture instability and tip overs.
7

The information provided for this rulemaking is drawn from a subset of data from those annual reports, as well as from the National Electronic Injury Surveillance System
8

(NEISS), which includes reports of injuries treated in U.S. EDs, and the Consumer Product Safety Risk Management System
9

(CPSRMS). For this rulemaking, staff focused on incidents that involved products that would be considered CSUs.
10

Staff considered incidents that involved the CSU tipping over, as well as incidents of CSU instability with indications of impending tip over. Tip-over incidents are a subset of product instability incidents, and involve CSUs actually falling over. Product instability incidents are a broader category that includes tip-over incidents, but may also include incidents where CSUs did not fully tip over. Staff considered instability incidents relevant because product instability can lead to a tip over, and the same factors, such as product design, can contribute to instability and tip overs.
11

7
These annual reports are available at:
https://www.cpsc.gov/Research--Statistics/Furniture-and-Decor-1.

8
Data from NEISS is based on a nationally representative probability sample of about 100 hospitals in the United States and its territories. NEISS data can be accessed from the CPSC website under the “Access NEISS” link at:
https://www.cpsc.gov/Research--Statistics/NEISS-Injury-Data.

9
CPSRMS is the epidemiological database that houses all anecdotal reports of incidents received by CPSC, “external cause”-based death certificates purchased by CPSC, all in-depth investigations of these anecdotal reports, as well as investigations of select NEISS injuries. Examples of documents in CPSRMS include: Hotline reports, internet reports, news reports, medical examiner's reports, death certificates, retailer/manufacturer reports, and documents sent by state/local authorities, among others.

10
Staff considered incidents that involved chests, bureaus, dressers, armoires, wardrobes, portable clothes lockers, and portable closets.

11
This section refers to tip-over incidents and instability incidents collectively as tip-over incidents.

The data presented here represent the minimum number of incidents or

fatalities during the time frames described. Data collection is ongoing for CPSRMS, and is considered incomplete for 2018 and after, so CPSC may receive additional reports for those years in the future.
12

12
Among other things, CPSRMS houses all in-depth investigation reports, as well as the follow-up investigations of select NEISS injuries. As such, it is possible for a NEISS injury case to be included in the national injury estimate, while its investigation report is counted among the anecdotal nonfatal incidents, or for a NEISS injury case to appear on both the NEISS injury estimate and fatalities, if the incident resulted in death while receiving treatment.

1. Fatal Incidents

Based on NEISS and CPSRMS, CPSC staff identified 193 reported CSU tip-over fatalities to children (
i.e.,
under 18 years old),
13

11 reported fatalities to adults (
i.e.,
ages 18 through 64 years), and 22 reported fatalities to seniors (
i.e.,
ages 65 years and older) that were reported to have occurred between January 1, 2000 and December 31, 2020.
14

Of the 193 reported CSU tip-over child fatalities, 89 (46 percent) involved only a CSU tipping over, whereas, 104 (54 percent) involved a CSU and a television tipping over. Of the child fatalities, 190 (98 percent) involved a chest, bureau, or dresser, 2 involved a wardrobe, and 1 involved an armoire. Of the 33 reported adult and senior fatalities, 32 (97 percent) involved only a CSU tipping over, whereas, 1 (9 percent) involved both a CSU and a television tipping over. Of the adult and senior fatalities, 29 involved a chest, bureau, or dresser, 2 involved a wardrobe, 1 involved an armoire, and 1 involved a portable storage closet.

13
Of the 193 reported fatalities, there was one tip-over incident that resulted in two deaths, making the number of fatal incidents 192.

14
Different time frames are presented for NEISS, CPSRMS, fatal, and nonfatal data because of the timeframes in which staff collected, received, retrieved, and analyzed the data. One example of the reason for varied timeframes is that staff drew data from previous annual reports and other data-collection reports (which used varied start dates), and then updated the data set to include more recent data. Another example is that CPSRMS data are available on an ongoing basis, whereas NEISS data are not available until several months after the end of the previous calendar year.

For the years for which reporting is considered complete—2000 through 2017—there have been from 3 to 21 child fatalities each year from CSU tip overs, and from 0 to 5 fatalities each year to adults and seniors.

Of the 193 reported child fatalities from tip overs, 166 involved children 3 years old or younger; 12 involved 4-year-olds; 7 involved 5-year-olds; 4 involved 6-year-olds; 1 involved a 7-year-old; and 3 involved 8-year-olds. Of the 89 reported child fatalities from tip overs involving only CSUs (
i.e.,
no televisions), 84 involved children 3 years old or younger; 2 involved 4-year-olds; 1 involved a 5-year-old; 1 involved a 6-year-old; and 1 involved a 7-year-old. Thus, 94 percent of these fatalities were children 3 years old and younger; 97 percent were 4 years old and younger; 98 percent were 5 years old and younger; and 99 percent were 6 years old and younger. Therefore, regardless of television involvement, the most reported CSU tip-over fatalities happened to children 3 years old or younger. Among children 4 years and older, a television was more frequently involved than not involved.

CSU tip-over fatalities to children were most commonly caused by torso injuries when only a CSU was involved, and were more commonly caused by head injuries when both a CSU and television tipped over. For the 89 child fatalities not involving a television, 58 resulted from torso injuries (chest compression); 13 resulted from head/torso injuries; 12 resulted from head injuries; 4 involved unknown injuries; and 2 involved a child's head, torso, and limbs pinned under the CSU. For the 104 child fatalities that involved both a CSU and television tipping over, 91 resulted from head injuries (blunt head trauma); 6 resulted from torso injuries (chest compression resulting from the child being pinned under the CSU); 2 resulted from head/torso injuries; 4 involved unknown injuries; and 1 involved head/torso/limbs.

2. Reported Nonfatal Incidents

CPSC staff identified 1,002 reported nonfatal CSU tip-over incidents for all ages that were reported to have occurred between January 1, 2005 and December 31, 2020.
15

CPSRMS reports are considered anecdotal because, unlike NEISS data, they cannot be used to identify statistical estimates or year-to-year trend analysis, and because they include reports of incidents in which no injury resulted. Although these anecdotal data do not provide for statistical analyses, they provide detailed information to identify hazard patterns, and provide a minimum count of injuries and deaths.

15
Nonfatal incident reports submitted to CPSC come from reports entered into CPSC's CPSRMS database no later than December 31, 2020, and includes completed NEISS investigations. All of the investigation reports based on NEISS injuries that occurred from 2006 through 2020 appear in the reported nonfatal incidents.

Of the 1,002 reported incidents, 64 percent (639 incidents) involved only a CSU, and 36 percent (363 incidents) involved both a CSU and television tipping over. Of the 1,002 incidents, 99.5 percent (997 incidents) involved a chest, bureau, or dresser; less than 1 percent (4 incidents) involved an armoire; and less than 1 percent (1 incident) involved a wardrobe.

For the years for which reporting is considered complete—2005 through 2017—there were from 6 to 256 reported nonfatal CSU tip-over incidents each year, with 2016 (256 incidents) and 2017 (101 incidents) reporting the highest number of incidents. Each year, there were from 5 to 232 reported nonfatal incidents involving only a CSU, with the highest number (232 incidents) occurring in 2016.

Of the 1,002 nonfatal CSU tip-over incidents reported, 362 did not mention any specific injuries; 628 reported one injury; and 12 reported two injuries, resulting in a total of 652 injuries reported among all of the reported nonfatal incidents. Of these 652 reported injuries, 64 (10 percent) resulted in hospital admission; 296 (45 percent) were treated in EDs; 28 (4 percent) were seen by medical professionals; and the level of care is unknown
16

for the remaining 264 (40 percent). Of 293 reports of nonfatal CSU tip-over injuries where only a CSU was involved; 7 resulted in hospital admission (of which 6 were children
17

); 23 were treated in the ED (of which 22 were children); 27 were seen by a medical professional (of which 19 were children); and the level of care is unknown for the remaining 236.

16
These reports include bruising, bumps on the head, cuts, lacerations, scratches, application of first-aid, or other indications of at least a minor injury that occurred, without any mention of aid rendered by a medical professional. There were three NEISS cases in which the victim went to the ED, but then left without being seen.

17
Incidents involving children include those in which the age of the victim was reported as well as those in which the age was not reported, but the report included indications that the victim was a child (
e.g.,
a sibling of a small child, or referred to as a “child,” “daughter,” or “son”). For the remaining incidents, the victim was either an adult, or the age was unknown.

Of the victims whose ages were known, there were more injuries suffered by children 3 years old and younger, than to older victims; and the injuries suffered by these young children tended to be more severe, compared to older children and adults/seniors. The severity of injury ranged from cuts and bumps to concussions and skull fractures. Of the 7 victims admitted to the hospital, 5 were 3 years old or younger; 1 was a child of unknown age; and 1 was an adult. Of the 23 victims treated in the ED, 8 were 3 years old or younger; 4 were 4 to 5 years old; 4 were 6 to 17 years old; and 6 were children of unknown age.

3. National Estimates of ED-Treated Injuries
18

18
Estimates are rounded to the nearest hundred and may not sum to total, due to rounding. NEISS estimates are reportable, provided the sample count is greater than 20, the national estimate is 1,200 or greater, and the coefficient of variation (CV) is less than 0.33.

According to NEISS, there were an estimated 78,200 injuries,
19

an annual average of 5,600 estimated injuries, related to CSU tip overs for all ages that were treated in U.S. hospital EDs from January 1, 2006 to December 31, 2019. Of the estimated 78,200 injuries, 56,400 (72 percent) were to children, which is an annual average of 4,000 estimated injuries to children over the 14-year period. For the remaining estimated 21,800 injuries to adults and seniors, about 3,200 (15 percent) were to seniors (
i.e.,
65 years and older).

19
Sample size = 2,629, coefficient of variation = .0667.

An estimated 61,700 (79 percent) of ED-treated injuries involved only a CSU tipping over, whereas, an estimated 16,500 (21 percent) involved both a CSU and television tipping over. This ratio was similar for injuries to children, with an estimated 40,700 (72 percent) of child incidents involving only a CSU, and an estimated 15,700 (28 percent) involving both a CSU and a television. In contrast, nearly all (an estimated 21,000 or 96 percent) of the estimated injuries to adults and seniors involved only a CSU. For each year from 2006 through 2019, there have been more estimated ED-treated injuries to children involving only a CSU tipping over, compared to incidents involving a CSU and a television tipping over.

For all ages, an estimated 77,000 (98 percent) of the ED-treated injuries involved a chest, bureau, or dresser. Similarly, for child injuries, an estimated 55,800 (99 percent) involved a chest, bureau, or dresser.
20

Of the ED-treated injuries to all ages, 93 percent were treated and released, and 4 percent were hospitalized. Among children, 93 percent were treated and released, and 3 percent were hospitalized.

20
Data on armoires, wardrobes, portable closets, and clothes lockers were insufficient to support reliable statistical estimates.

For each year from 2006 through 2019, there were an estimated 2,500 to 5,900 ED-treated injuries to children from CSU tip overs. The estimated annual number of ED-treated injuries to adults and seniors from CSU tip overs is fairly consistent over most of the 14-year period, with an overall yearly average of 1,600 estimated injuries, although data were insufficient to support reliable statistical estimates for adults and seniors for 2014, 2015, and 2019.

CPSC focused on ED-treated injuries involving children because these make up the majority of ED-treated CSU tip-over injuries. For 2010 through 2019, there is a statistically significant linear decline in child injuries involving CSU tip overs (both with and without televisions);
21

however, there is no linear trend detected in injuries to children involving only CSUs tipping over. This indicates that the statistically significant decrease in all CSU tip overs involving children is driven by the decline in tip overs involving televisions, while the rate of ED-treated incidents involving CSUs without televisions has remained stable.

21
There were not enough CSU ED-treated incidents to children involving both a CSU and a television to make reliable estimates for the most recent 5 years, 2015 through 2019.

Of the estimated ED-treated injuries to children, most involved 2- and 3-year-olds, followed by 1- and 4-year-olds. An estimated 7,900 ED-treated injuries involved 1-year-olds;
22

an estimated 15,000 involved 2-year-olds;
23

an estimated 13,000 involved 3-year-olds;
24

and an estimated 7,500 involved 4-year-olds.
25

There were an estimated 2,300 injuries to 5-year-olds that involved only a CSU, and an estimated 1,800 injuries to 6-year-olds that involved only a CSU, but data were insufficient to support reliable statistical estimates for incidents involving CSUs and televisions for these ages. For children 7 to 17 years old,
26

there were an estimated 4,700 ED-treated injuries involving only a CSU, and an estimated 1,600 involving a CSU and a television.

22
An estimated 6,300 involved only a CSU and the remaining 1,600 involved a CSU and television.

23
An estimated 10,600 involved only a CSU, and the remaining 4,400 involved a CSU and television.

24
An estimated 9,200 involved only a CSU, and the remaining 3,800 involved a CSU and television.

25
An estimated 5,100 involved only a CSU, and the remaining 2,400 involved a CSU and television.

26
These ages are grouped together because data were insufficient to generate estimates for any single age within that range.

Of the estimated 56,400 ED-treated CSU tip-over injuries to children, an estimated 20,800 (37 percent) resulted in contusions/abrasions;
27

an estimated 14,900 (26 percent) resulted in internal organ injury (including closed head injuries);
28

an estimated 7,600 (13 percent) resulted in lacerations;
29

an estimated 5,200 (9 percent) resulted in fractures;
30

and the remaining estimated 7,800 (14 percent) resulted in other diagnoses.

27
Seventy-six percent of these involved only a CSU, and the remainder involved a CSU and television tipping over.

28
Sixty-one percent of these involved only a CSU, and the remainder involved a CSU and television tipping over.

29
Eighty-two percent of these involved only a CSU, and the remainder involved a CSU and television tipping over.

30
Sixty-nine percent of these involved only a CSU, and the remainder involved a CSU and television tipping over.

Overall, an estimated 33,700 (60 percent) of ED-treated tip-over injuries to children were to the head, neck, or face; and an estimated 10,300 (18 percent) were to the leg, foot, or toe. The injuries to children were more likely to be head injuries when a television was involved than when no television was involved. Of the estimated number of ED-treated injuries to children involving a CSU and a television, 73 percent were head injuries, compared to 55 percent of injuries involving only a CSU. In addition, of the estimated injuries to children involving only a CSU, 20 percent were leg, foot, or toe injuries, and 14 percent were trunk or torso injuries. Data were insufficient to generate estimates of trunk/torso or arm/hand/finger injuries when both a CSU and television tipped over.

B. Details Concerning Injuries
31

31
For more details about injuries, see Tab B of the NPR briefing package.

To assess the types of injuries that result from CSU tip overs, CPSC staff focused on incidents involving children, because the vast majority of CSU tip overs involve children. The types of injuries resulting from furniture tipping over onto children include soft tissue injuries, such as cuts and bruises (usually a sign of internal bleeding); skeletal injuries and bone fractures to arms, legs, and ribs; and potentially fatal injuries resulting from skull fractures, closed-head injuries, compressional and mechanical asphyxia, and internal organ crushing leading to hemorrhage. These types of injuries can result from tip overs involving CSUs alone, or CSUs with televisions.

As explained above, head injuries and torso injuries are common in CSU tip overs involving children. The severity of injuries depends on a variety of factors, but primary determinants include the force generated at the point of impact, the entrapment time, and the body part impacted. The head, neck, and chest are the most vulnerable. The severity of injury can also depend on the orientation of the child's body or body part when it is hit or trapped by the CSU. Sustained application of a force that affects breathing can lead to compressional asphyxia and death. In most CSU tip-over cases, serious injuries and death are a result of blunt force trauma to the head and intense pressure on the chest causing

respiratory and circulatory system impairment.

Head injuries are produced by high-impact forces applied over a small area and can have serious clinical consequences, such as concussions and facial nerve damage. Such injuries are often fatal, even in cases where the child is immediately rescued and there is rapid intervention. An incident involving blunt head trauma can result in immediate death or loss of consciousness. Autopsies from CSU tip-over fatalities to children reported crushing injuries to the skull and regions of the eye and nose. Brain swelling, deep scalp hemorrhaging, traumatic intracranial bleeding, and subdural hematomas were often reported. These types of injuries are typical of crush injuries caused by blunt head trauma and often have a fatal outcome. Children who survive such injuries may suffer neurological deficits, require neurosurgical interventions, and can face lifelong disabilities.

Compressional and mechanical asphyxia is another potential cause of injury and death in CSU tip-over incidents. Asphyxia can be fatal within minutes. In multiple CSU tip-over incidents, there was physical evidence of chest compression visible as linear marks or abrasions across the chest and neck, consistent with the position of the CSU. Compressional and mechanical asphyxia can result from mechanical forces generated by the sheer mass of an unyielding object, such as furniture, acting on the thoracic and abdominal area of the body, which prevents thorax expansion and physically interferes with the coordinated diaphragm and chest muscle movement that normally occurs during breathing. Torso injuries, which include compressional and mechanical asphyxia, are the most common form of injury for non-television CSU fatalities. External pressure on the chest that compromises the ability to breathe by restricting respiratory movement or on the neck can cause oxygen deprivation (hypoxia). Oxygen deprivation to the brain can cause unconsciousness in less than three minutes and may result in permanent brain damage or death when pressure is applied directly on the neck by the CSU or a component of the CSU (such as the edge of a drawer). The prognosis for a hypoxic victim depends on the degree of oxygen deprivation, the duration of unconsciousness, and the speed at which cardiovascular resuscitation attempts are initiated relative to the timing of cardiopulmonary arrest. Rapid reversal of the hypoxic state is essential to prevent or limit the development of pulmonary and cerebral edema that can lead to death or other serious consequences. The sooner the CSU (compression force) is removed and resuscitation initiated, the greater the likelihood that the patient will regain consciousness and recover from injuries.

In addition to chest compression, pressure on the neck by a component of the CSU can also result in rapid strangulation due to pressure on the blood vessels in the neck. The blood vessels that take blood to and from the brain are relatively unprotected in the soft tissues of the neck and are vulnerable to external forces. Sustained compression of either the jugular veins or the carotid arteries can lead to death. Petechial hemorrhages of the head, neck, chest, and the periorbital area were reported in autopsy reports of CSU tip-over incidents.

Pediatric thoracic trauma has unique features that differ from adult thoracic trauma, because of differences in size, structure, posture, and muscle tone. While the elasticity of a child's chest wall reduces the likelihood of rib fracture, it also provides less protection from external forces. Impact to the thorax of an infant or small child can produce significant chest wall deflection and transfer large kinetic energy forces to vital thoracic organs such as the lungs and heart, which can cause organ deflection and distention and lead to traumatic asphyxia, or respiratory and circulatory system impairment or failure. In addition, a relatively small blood volume loss in a child, due to internal organ injuries and bleeding, can lead to decreased blood circulation and shock.

The severity of the injury or likelihood of death can be reduced if a child is quickly rescued. However, children's ability to self-rescue is limited because of their limited cognitive awareness of hazards, limited skills to react quickly, and limited strength to remove the fallen CSU. Moreover, many injuries can result in immediate death or loss of consciousness, making self-rescue impossible.

C. Hazard Characteristics
32

32
For additional information about hazard patterns, see Tab C of the NPR briefing package.

To identify hazard patterns associated with CSU tip overs, CPSC focused on incidents involving children and CSUs without televisions because the majority of fatal and nonfatal incidents involve children and, in recent years, there has been a statistically significant decrease in the overall number of ED-treated CSU tip-over incidents that appears to be driven by a decline in incidents involving CSUs with televisions, while the rate of ED-treated incidents involving CSUs without televisions has remained stable. Staff used NEISS and CPSRMS reports to identify hazard patterns, including In-Depth-Investigation (IDI) reports, and also considered child development and capabilities, as well as online videos of real-life child interactions with CSUs and similar furniture items (including videos of tip-over incidents).

1. Filled Drawers

Of the 89 fatal CPSRMS incidents involving children and only CSUs, 53 (59 percent) provided information about whether the CSU drawers contained items at the time of the tip over. Of those 53 incidents, 51 (96 percent) involved partially filled or full drawers. Of the 263 nonfatal CPSRMS tip overs involving children and only CSUs, drawer fill level was reported for 67 incidents (25 percent). Of these 67 incidents, 60 (90 percent) involved partially filled or full drawers.
33

CPSRMS incidents show that most items in the drawers were clothing, although a few mentioned other items along with clothing (
e.g.,
diaper bag, toys, papers).

33
Nonfatal NEISS incident reports did not contain information on drawer fill level or contents.

2. Interactions

Of the 89 fatal CPSRMS tip overs involving children and only a CSU, 47 reported the type of interaction the child had with the CSU at the time of the incident. Of these 47 incidents, 35 (74 percent) involved a child climbing on the CSU; 8 (17 percent) involved a child sitting, laying, or standing in a drawer; and 4 (9 percent) involved a child opening drawers. Climbing was the most common reported interaction for children 3 years old and younger.

Of the 263 nonfatal CPSRMS tip-over incidents involving children and only CSUs, the type of interaction was reported in 160 incidents. Of these, 101 (63 percent) involved opening drawers; 32 (20 percent) involved climbing on the CSU; 10 (6 percent) involved putting items in/taking them out of a drawer; 9 (6 percent) involved pulling on the CSU; 5 (3 percent) involved leaning or pushing down on an open drawer; 2 (1 percent) involved another interaction; and 1 (less than 1 percent) involved a child in the drawer. Opening drawers was the most common reported interaction for children 6 years old and younger, and was particularly common for 2- and 3-year-olds.

Of the 1,463 nonfatal NEISS incidents involving children and only CSUs, the type of interaction was reported in 559 incidents. Of these, the child was injured because of another person's interaction with the CSU in 22 incidents; the remaining 537 incidents involved the child interacting with the CSU. Of these 537 incidents, 412 (77 percent) involved climbing on the CSU; 42 (8 percent) involved opening drawers; and the remaining 83 incidents (15 percent) involved a child in the drawer, pulling on the CSU, putting items in or taking items out of a drawer, reaching, hitting, jumping, a child on top of the CSU, playing in a drawer, pulling up, swinging, or other interaction. For children 3 years old or younger, climbing constituted almost 80 percent of reported interactions. Overall, 81 percent (438 of 537) of the reported interactions in the nonfatal NEISS tip-over incidents involving children and only CSUs are those in which the child's weight was supported by the CSU (
e.g.,
climbing, in drawer, jump, on top, swinging), and 12 percent (64 of 537) were interactions in which the child's strength determines the force (
e.g.,
hit, opening drawers, pulled on, pulled up).

Thus, in fatal incidents, a child climbing on the CSU was, by far, the most common reported interaction; and in nonfatal incidents, opening drawers and climbing were the most common reported interactions. These interactions are examined further, below.

To learn more about children's interactions with CSUs during tip-over incidents, CPSC staff also reviewed videos, available from news sources, articles, and online, that involved children interacting with CSUs and similar products, and CSU tip overs. Videos of children climbing on CSUs and similar items show a variety of climbing techniques, including stepping on the top of the drawer face, stepping on drawer knobs, using the area between drawers as a foothold, gripping the top of an upper drawer with their hands, pushing up using the top of a drawer, and using items to help climb. Videos of children in drawers of CSUs and other similar products include children leaning forward and backward out of a drawer; sitting, lying, and standing in a drawer; and bouncing in a drawer. Some videos also show multiple children climbing a CSU or in a drawer simultaneously.

a. Climbing

As discussed above, climbing on the CSU was one of the primary interactions involved in CSU tip overs involving children and only a CSU. It was the most common reported interaction (74 percent) in fatal CPSRMS incidents; it was the most common reported interaction (77 percent) in nonfatal NEISS incidents; and it was the second most common reported interaction (20 percent) in nonfatal CPSRMS incidents.

Children as young as 9 months, and up to 13 years old were involved in climbing incidents. Fatal climbing incidents most often involved 1-, 2-, and 3-year-old children, and nonfatal climbing incidents most often involved 2- and 3-year-old children. Of climbing incidents with a reported age, the children were 3 years old or younger in 94 percent (33 of 35) of the fatal CPSRMS incidents; 73 percent (301 of 412) of the nonfatal NEISS incidents; and 63 percent (17 of 27) of the nonfatal CPSRMS incidents.

The prevalence of children climbing during CSU tip overs is consistent with the expected motor development of children. Between approximately 1 and 2 years old, children can climb on and off of furniture without assistance, use climbers, and begin to use playground apparatuses independently; and 2-year-olds commonly climb. The University of Michigan Transportation Research Institute (UMTRI) focus groups on child climbing (the UMTRI study is described in section VII.B. Forces and Moments During Child Interactions with CSUs of this preamble) demonstrated these abilities, with child participants showing interest in climbing CSUs and other furniture.

b. Opening Drawers

As discussed above, opening the drawers of a CSU was a common interaction in CSU tip overs involving children and only a CSU. It was the most common reported interaction (63 percent) in nonfatal CPSRMS incidents; it was the second most common reported interaction (8 percent) in nonfatal NEISS incidents; and it was the third most common reported interaction (9 percent) in fatal CPSRMS incidents.

Children as young as 11 months, and up to 14 years old were involved in incidents where the child was opening one or more drawers of the CSU. In nonfatal CPSRMS incidents, opening drawer incidents most commonly involved 2-year-olds; in nonfatal NEISS incidents, opening drawer incidents most commonly involved 3-year-olds, followed by 2-year-olds, followed by 4-year-olds, followed by children under 2 years old; and in nonfatal CPSRMS incidents, opening drawer incidents most commonly involved 3-year-olds, followed by 2-year-olds. Children of all ages were able to open at least one drawer.

Looking at both fatal and nonfatal CPSRMS tip overs involving children and only CSUs, where the interaction involved opening drawers, overall, about 53 percent involved children opening one drawer, 10 percent involved opening two drawers, and almost 17 percent involved opening “multiple” drawers. In several incidents (23 CPSRMS incidents), children opened “all” of the drawers; it is possible that additional incidents, mentioning a specific number of open drawers (between 2 and 8), also involved all the drawers being opened. In incidents where all of the drawers were open, the CSUs ranged from 2-drawer to 8-drawer units. The youngest child reported to have opened all drawers was 13 months old.

Consistent with these incident data, the UMTRI child climbing study found that caregivers commonly reported that their children opened and closed drawers when interacting with furniture.

It is possible for CSUs to tip over from the forces generated by open drawers and their contents, alone, without additional interaction forces. However, pulling on a drawer to open it can apply increased force that contributes to instability. Once a drawer is fully opened, any additional pulling is on the CSU as a whole. The pull force, and the height of the drawer pull location, relative to the floor, are relevant considerations. To examine this factor, staff assessed 15 child incidents in which the height of the force application could be calculated based on descriptions of the incidents. Force application heights ranged from less than one foot to almost four feet (46.5 inches), and children pulled on the lowest, highest, and drawers in between.

c. Opening Drawers and Climbing Simultaneously

CPSC staff also examined incidents in which both climbing and open drawers occurred simultaneously. Of the 35 fatal CPSRMS climbing incidents, 13 reported the number of drawers open; in all of these incidents, the reported number of drawers open was one, although, based on further analysis, the number of open drawers could be as high as 8 in one incident.
34

Of the 32 nonfatal CPSRMS climbing incidents, 15 gave some indication of the number of open drawers. Of these, 7 reported that one drawer was open, 2 reported

that half or less of the drawers were open, 4 reported that multiple drawers were open, and 2 reported that all the drawers were open. In the 2 cases where all drawers were open, the children were 3 and 4 years old. Of the 412 climbing incidents in the nonfatal NEISS data, 28 gave some indication of the number of open drawers. Of these, 11 reported that one drawer was open, 12 reported that multiple drawers were open, 1 reported that two drawers were open, and 2 reported that all drawers were open. These data are consistent with the videos staff reviewed, which show a range of drawer positions when children climbed on units, including all drawers closed, one drawer open, multiple drawers open, and all drawers fully open.

34
CPSC staff analysis suggests that 7 or more drawers of an 8-drawer unit were open and the child was in a drawer leaning out over the edge in a fatal incident. This analysis is described in Tab M of the NPR briefing package, as Model E.

There is limited information in the incident data about children's interaction with doors on CSUs, as opposed to interactions with drawers. Staff found two fatal CPSRMS and four nonfatal CPSRMS tip-over incidents involving wardrobes and armoires, which include doors. In one of the fatal incidents, the victim was found inside a wardrobe that had two doors and one drawer, suggesting that the child opened the doors of the wardrobe. In the other fatal incident, the victim was found under a two-door wardrobe. In most of the nonfatal incidents involving wardrobes or armoires, children were reportedly interacting with items inside the unit, which would require them to open the doors. The ages of the children in these incidents ranged from 3 to 11 years, although opening doors is easily within the physical and cognitive abilities of younger children.

These incidents indicate that children can and do open CSU doors. There is no direct evidence in the incident data that, once CSU doors are open, children put their body weight on the open doors (
i.e.,
open and climbing). However, this is a plausible interaction based on child capabilities, provided that the child has a sufficient hand hold.

d. Differences in Interactions by Age

Based on the incident data, children 3 years old and younger climb, open drawers without climbing, get items in and out of drawers, lean on open drawers, push down on open drawers, sit or lie in bottom drawers, or stand on open bottom drawers. Among fatal CPSRMS tip-over incidents involving children and only CSUs, climbing was the most common interaction for children 3 years old and younger; this drops off sharply for 4-year-olds. Starting at 4 years old, children do not appear to sit or lie in bottom drawers of a CSU. Among nonfatal CPSRMS tip-over incidents involving children and only CSUs, opening drawers was, by far, the most common interaction for children 7 years old and younger; and climbing was also common among 3-year-olds and, to a lesser extent, among 2- and 4-year-olds. Among nonfatal NEISS tip overs involving children and only CSUs, climbing was common for 2- and 3-year-olds, slightly less common for 4-year-olds and children under 2 years, and dropped off further for children 5 years and older.

3. Flooring

Of the 89 fatal CPSRMS tip overs involving children and only CSUs, the type of flooring under the CSU was reported for 55 incidents. Of these, 45 (82 percent) involved carpeting, which includes rugs; 8 (15 percent) involved wood, hardwood, or laminate wood flooring; and 2 (4 percent) involved tile or linoleum flooring. The reports for 30 of the fatal CPSRMS tip-over incidents involving carpet included photos with visible carpet. All carpet in these pictures appeared to be typical wall-to-wall carpeting. Four appeared to be a looped pile carpet, and 26 appeared to be cut pile. Staff also identified two incidents with reported “shag” carpeting, including one fatal incident. Staff found one report mentioning a rug, although the thickness of the rug is unknown.

Of the 263 nonfatal CPSRMS tip overs involving children and only CSUs, the type of flooring under the CSU was reported for 60 incidents. Of these, 48 (80 percent) involved carpeting, which includes rugs; 10 (17 percent) involved wood, hardwood, or laminate wood flooring; 1 (2 percent) involved tile or linoleum flooring; and 1 (2 percent) indicated that the front legs of the CSU were on carpet while the back legs were on wood flooring.
35

35
Flooring type was not reported in nonfatal NEISS incident reports.

Thus, for incidents where flooring type was reported, carpet was, by far, the most prevalent flooring type.

4. Characteristics of Children in Tip-Over Incidents

a. Age of Children

Children in fatal CPSRMS tip-over incidents involving only CSUs were 11 months through 7 years old. A total of 33 fatal incidents involved children under 2 years old; 30 involved 2-year-old children; 21 involved 3-year-olds; 2 involved 4-year-olds; and 1 incident each involved 5-, 6-, and 7-year-old children. Among the nonfatal CPSRMS tip-over incidents involving children and only CSUs where age was reported, 3-year-olds were involved in the highest number of incidents (59 incidents), followed by 2-year-olds (47 incidents).

Nonfatal NEISS tip-over incidents involving children and only CSUs follow a similar distribution, with the highest number of reported incidents involving 2-year-olds, followed by 3-year-olds, and children less than 2 years. Further details regarding the age of children involved in CSU tip overs is available in the discussion of incident data, above.

b. Weight of Children

Among the 89 fatal CPSRMS tip-over incidents involving children and CSUs without televisions, the child's weight was reported in 49 incidents and ranged from 18 pounds to 45 pounds. Where weight was not reported, staff used the most recent Centers for Disease Control and Prevention (CDC) Anthropometric Reference to estimate the weight of the children.
36

Staff used the 50th percentile values of weight that correspond to the victims' ages to estimate the weight range of the children. For the remaining 40 fatal CPSRMS incidents without a reported weight, the estimated weight range was 19.6 pounds to 45.1 pounds.

36
Fryar, C.D., Carroll, M.D., Gu, Q., Afful, J., Ogden, C.L. (2021). Anthropometric reference data for children and adults: United States, 2015-2018. National Center for Health Statistics. Vital Health Stat 3(46). The CDC Anthropometric Reference is based on a nationally representative sample of the U.S. population, and the 2021 version is based on data collected from 2015 through 2018. CPSC staff uses the CDC Anthropometric Reference, rather than the CDC Growth Chart, because it is more recently collected data and because the data are aggregated by year of age, allowing for estimates by year. CDC growth charts are available at:
https://www.cdc.gov/growthcharts/clinical_charts.htm.

Among the 263 nonfatal CPSRMS incidents involving children and only CSUs, the weights of 47 children were reported, ranging from 26 pounds to 80 pounds. Where it was not reported, staff again estimated the weight of the children using the 50th percentile values of weight that correspond to the victims' ages from the most recent CDC Anthropometric Reference. The estimated child weights for the 164 nonfatal CPSRMS incidents without a reported child weight, but with a reported age (which included a 17-year-old), ranged from 19.6 pounds to 158.9 pounds.

Although nonfatal NEISS incident data did not include the children's weights, staff again estimated the children's weights by age, determining that for tip overs involving only CSUs, the estimated weights of the children ranged from 15.8 pounds to 158.9 pounds (this covered children from 3

months to 17 years old). The weighted average of children's estimated weight in nonfatal NEISS incidents was 40.26 pounds.
37

37
Weighted average is equal to the sum of the product of the number of reported incidents for that age times the estimated weight for that age divided by the total number of reported incidents.

Overall, the weighted average of children's reported weight for CPSRMS incidents is 34.23 pounds; whereas, the weighted average of children's estimated weight was 38.8 pounds.

The weight of a child is particularly relevant for climbing incidents because weight is a factor in determining the force a child generates when climbing. For this reason, CPSC staff looked at the weights of children involved in climbing incidents, specifically. Of the 35 fatal CPSRMS child climbing incidents, the weight of the child was reported for 23 incidents, and ranged from 21.5 to 45 pounds. For the remaining 12 climbing incidents in which the child's weight was not reported, CPSC staff estimated their weights, based on age, and the weights ranged from 23.8 to 39 pounds. Of the 32 nonfatal CPSRMS child climbing incidents, the weight of the child was reported in 8 incidents, and ranged from 26 to 80 pounds. For the remaining 24 incidents, staff estimated the weights based on age, and the weights ranged from 25.2 to 45.1 pounds. Weight was not reported in the nonfatal NEISS data, however, using the ages of the children in the 412 nonfatal NEISS child climbing incidents (9 months to 13 years old), staff estimates that their weights ranged from 19.6 to 122 pounds, and the weighted average was 34.2 pounds.

5. Televisions

Of the 104 child fatalities involving a CSU and television tipping over, 85 (90 percent) involved a box or cathode ray tube (CRT) television, 2 involved a flat-panel television, and 16 did not provide information about the television. Of the incidents that provided information about television size, the most common television size was 27 inches. The approximate weight range of the CRT televisions, when provided, was between 70 pounds and 150 pounds.

Although televisions are involved in CSU tip overs, and the Commission raised the possibility of addressing televisions in the ANPR, the proposed rule does not focus on television involvement. This is primarily because, in recent years, there has been a decline in the overall number of CSU tip-over incidents that appears to be driven by a decrease in tip overs involving televisions, while the rate of ED-treated incidents involving CSUs without televisions has remained stable.

V. Relevant Existing Standards
38

38
For additional information about relevant existing standards, see Tab C, Tab D, Tab F, and Tab N of the NPR briefing package.

In the United States, the primary voluntary standard that addresses CSU stability is ASTM F2057-19,
Standard Consumer Safety Specification for Clothing Storage Units
. In addition, CPSC staff identified three international consumer safety standards and one domestic standard that are relevant to CSUs:

• AS/NZS 4935: 2009, the Australian/New Zealand Standard for
Domestic furniture—Freestanding chests of drawers, wardrobes and bookshelves/bookcases—determination of stability
;

• ISO 7171 (2019), the International Organization for Standardization
International Standard for Furniture—Storage Units—Determination of stability
;

• EN14749 (2016), the European Standard,
European Standard for Domestic and kitchen storage units and worktops—Safety requirements and test methods
; and

• ANSI/SOHO S6.5-2008 (R2013),
Small Office/Home Office Furniture—Tests American National Standard for Office Furnishings
.

This section describes these standards and provides CPSC staff's assessment of their adequacy to address CSU tip-over injuries and deaths.

A. ASTM F2057-19

ASTM first approved and published ASTM F2057 in 2000, and has since revised the standard seven times. The current version, ASTM F2057-19, was approved on August 1, 2019, and published in August 2019. ASTM Subcommittee F15.42, Furniture Safety, is responsible for this standard. Since the first publication of ASTM F2057, CPSC staff has participated in the F15.42 subcommittee and task group meetings and worked with ASTM to improve the standards; however, ASTM has not addressed several issues CPSC has identified.

1. Scope

ASTM F2057-19 is intended to reduce child injuries and deaths from hazards associated with CSUs tipping over and aims “to cover children up to and including age five.” The standard covers CSUs that are 27 inches or more in height, freestanding, and defines CSUs as: “furniture item[s] with drawers and/or hinged doors intended for the storage of clothing typical with bedroom furniture.” Examples of CSUs provided in the standard include: Chests, chests of drawers, drawer chests, armoires, chifforobes, bureaus, door chests, and dressers. The standard does not cover “shelving units, such as bookcases or entertainment furniture, office furniture, dining room furniture, underbed drawer storage units, occasional/accent furniture not intended for bedroom use, laundry storage/sorting units, nightstands, or built-in units intended to be permanently attached to the building, nor does it cover `Clothing Storage Chests' as defined in Consumer Safety Specification F2598.”

2. Stability Requirements

ASTM F2057-19 includes two performance requirements for stability. The first is in section 7.1 of the standard,
Stability of Unloaded Unit
. This test consists of placing an empty CSU on a hard, level, flat surface, opening all doors (if any) to 90 degrees, and extending all drawers and pull-out shelves to the outstop (which is a feature that limits outward motion of drawers or pull-out shelves). In the absence of an outstop, all drawers and pull-out shelves are opened to two-thirds of the operational sliding length (which is the length from the inside face of the drawer back to the inside face of the drawer). All flaps and drop fronts are opened to their horizontal position or as near to horizontal as possible. If the CSU tips over in this configuration, or is supported by any component that was not specifically designed for that purpose, it does not meet the requirement.

The second stability requirement is in section 7.2 of the standard,
Stability with Load
. This test consists of placing an empty CSU on a hard, level, flat surface, and gradually applying a 50±2-pound test weight. The 50-pound test weight is intended to represent the weight of a 5-year-old child. For units with drawers, the test requires opening one drawer to the outstop, or in the absence of an outstop, to two-thirds of its operational sliding length, and gradually applying the test weight to the front face of the drawer. For units with doors, the test requires opening one door to 90 degrees and gradually applying the test weight. All other drawers and doors remain closed, unless they must be opened to access other components behind them (
e.g.,
a drawer behind a door). Each drawer and door is tested individually. If the CSU tips over in this configuration, or is supported by any component that was not specifically designed for that

purpose, it does not meet this requirement.

3. Tip Restraint Requirements

ASTM F2057-19 requires CSUs to include a tip restraint that complies with ASTM F3096-14,
Standard Performance Specification for Tipover Restraint(s) Used with Clothing Storage Unit(s)
.
39

ASTM F2057-19 and F3096-14 define a tip restraint as a “supplemental device that aids in the prevention of tip over.” ASTM F3096-14 provides a test protocol to assess the strength of tip restraints, but does not evaluate the attachment to the wall or CSU. The test method specifies that the tester attach the tip restraint to a fixed structure and apply a 50-pound static load.

39
Approved October 1, 2014 and published October 2014.

4. Labeling Requirements

ASTM F2057-19 requires CSUs to be permanently marked in a conspicuous location with warnings that meet specified content and formatting. The warning statements address the risk of children dying from furniture tip overs; not allowing children to stand, climb, or hang on CSUs; not opening more than one drawer at a time; placing the heaviest items in the bottom drawer; and installing tip restraints. For CSUs that are not intended to hold a television, this is also addressed in the warning. Additionally, units with interlock systems must include a warning not to defeat or remove the interlock system. An interlock system is a device that prevents simultaneous opening of more drawers than intended by the manufacturer (like is common on file cabinets). The standard requires that labels be formatted in accordance with ANSI Z535.4,
American National Standard for Product Safety Signs and Labels
.

The standard also includes a performance requirement and test method for label permanence, which are consistent with requirements in other ASTM juvenile furniture product standards. The warning must be “in a conspicuous location when in use” and the back of the unit is not considered conspicuous; the standard does not define “conspicuous location when in use.”

5. Assessment of Adequacy

CPSC does not consider the stability requirements in ASTM F2057-19 adequate to address the CSU tip-over hazard because they do not account for multiple open and filled drawers, carpeted flooring, and dynamic forces generated by children's interactions with the CSU, such as climbing or pulling on the top drawer. As discussed earlier in this preamble, these factors are commonly involved in CSU tip-over incidents; and, as discussed later in this preamble, testing indicates that these factors decrease the stability of CSUs.

Although ASTM F2057-19 includes a test with all drawers/doors open, the unit is empty and no additional force is applied during this test. Consumers are likely to fill drawers with clothing, since that is the intended purpose of the product, and a CSU with filled drawers is likely to be less stable than an empty unit when more than half of the drawers are open. In addition, although ASTM F2057-19 includes a static weight applied to the top of one open drawer or door (intended to represent a 5-year-old child), this 50-pound weight does not include the additional moment
40

due to the center of gravity of a child climbing, dynamic forces, and horizontal forces when a child climbs, even when only considering the forces generated by very young children. As the UMTRI study described in this preamble found, the forces children can exert while climbing a CSU exceed their static weights. Finally, the testing does not account for the effect of carpeting, which is common flooring in homes (particularly in bedrooms), is commonly present in tip-over incidents, and decreases CSU stability. Thus, by testing CSUs with open drawers empty, a 50-pound static weight, and on a hard, level, flat surface, ASTM F2057-19 does not reflect real-world use conditions that decrease the stability of CSUs.

40
Moment, or torque, is an engineering term to describe rotational force acting about a pivot point, or fulcrum.

Staff also looked at whether CSUs involved in tip-over incidents complied with ASTM F2057-19 because it would give an indication of whether F2057 is effective at preventing tip overs and, by extension, whether it is adequate. Of the 89 fatal CPSRMS tip-over incidents involving children and only CSUs, CPSC staff determined that 1 of the CSUs complied with the ASTM F2057-19 stability requirements, 1 CSU met the stability requirements when a test weight at the lower permissible weight range was used, and 11 units did not meet the stability requirements. For the remaining 76 units, staff was unable to determine whether they met the ASTM F2057-19 stability requirements, although staff did determine that an exemplar of one of these CSUs complied with the requirements. Of 263 nonfatal CPSRMS incidents involving children and CSUs without televisions for which staff assessed the compliance of the CSU, staff determined that 20 met the ASTM F2057-19 stability requirements, and 95 did not. For the remaining 148 units, staff was unable to determine whether the units met the ASTM F2057-19 stability requirements.
41

41
Staff did not assess whether NEISS incidents involved ASTM-compliant CSUs because the reports do not contain specific information about the products.

Based on a limited review of the tip restraint requirements in ASTM F2057-19 and ASTM F3096-14, CPSC is concerned that these requirements may not be adequate either. ASTM F3096-14 does not address the whole tip-restraint system, which includes the connection to the CSU and the connection to the wall. The standard assumes an ideal connection to both the furniture and the wall, but incidents suggest that both of these are potential points of failure. In addition, ASTM F3096-14 uses a 50-pound static force. Based on the UMTRI study, this force may not represent the force on a tip restraint from child interactions, especially for interactions that can generate large amounts of force, including from older children. For example, the UMTRI study found that when a child bounced, leaned, or yanked on a CSU, the forces generated were equivalent to 2.7, 2.7, and 3.9 times the child's body weight, respectively, at a distance of 1 foot from the fulcrum. However, staff did not evaluate the tip restraint requirements in ASTM F2057-19 and ASTM F3096-14 because, as discussed in this preamble, several research studies show that a large number of consumers do not anchor furniture, including CSUs, and there are several barriers to the use of tip restraints. As such, even if tip restraint requirements were effective, CSUs should be inherently stable to account for the lack of consumer use of tip restraints and additional barriers to proper installation and use of tip restraints.

CPSC also has some concerns with the effectiveness of the content in the warning labels required in ASTM F2057-19. For example, the meaning of “tipover restraint” may not be clear to consumers, and directing consumers not to open more than one drawer at a time is not consistent with consumer use. In addition, focus group testing discussed in this preamble indicated that consumers had trouble understanding the child climbing symbol required by the standard. CPSC staff also believes that greater clarity about the required placement of the label would make the warning more effective.

6. Compliance With ASTM F2057

CPSC staff assessed compliance with the stability requirements in ASTM F2057-19. In 2016,
42

staff tested 61 CSU samples and found that 50 percent (31 of 61) did not comply with the stability requirements in ASTM F2057.
43

In 2018, CPSC staff assessed a total of 188 CSUs, including 167 CSUs selected from among the best sellers from major retailers, using a random number generator; 4 CSU models that were involved in incidents;
44

and 17 units assessed as part of previous test data provided to CPSC.
45

Of the 188 CSUs, 171 (91 percent) complied with the stability requirements in ASTM F2057. One CSU (0.5 percent) did not comply with the Stability of Unloaded Unit test, and 17 (9 percent) did not meet the Stability with Load test. The unit that did not meet the requirements of the Stability of Unloaded Unit test also did not meet the requirements of the Stability with Load test.

42
Although this testing involved ASTM F2057-14, the stability requirements were the same as in ASTM F2057-19. The test results are available at:
https://www.cpsc.gov/s3fs-public/2016-Tipover-Briefing-Package-Test-Results-Update-August-16-2017.pdf?yMCHvzY_YtOZmBAAj0GJih1lXE7vvu9K
.

43
This testing also found that 91 percent of CSUs (56 of 61) did not comply with the labeling requirements in ASTM F2057-14, and 43 percent (26 of 61) did not comply with the tip restraint requirements.

44
Staff tested exemplar units, meaning the model of CSU involved in the incident, but not the actual unit involved in the incident.

45
The CSUs were identified from the Consumer Reports study “Furniture Tip-Overs: A Hidden Hazard in Your Home” (Mar. 22, 2018), available at:
https://www.consumerreports.org/furniture/furniture-tip-overs-hidden-hazard-in-your-home/
.

In addition, as part of staff's incident recreation and modeling (discussed in section VII.D. Incident Recreation and Modeling of this preamble), staff determined that two of the seven tested CSU models that had been involved in tip-over incidents complied with the stability requirements in ASTM F2057, and one additional CSU was borderline on whether it complied with the standard. This suggests that the stability requirements in ASTM F2057-19 do not adequately reduce the risk of tip overs.

B. AS/NZS 4935: 2009

AS/NZS 4935 is a voluntary standard prepared by Standards Australia's and Standards New Zealand's Joint Technical Committee CS-088/CS-091, Commercial/Domestic Furniture. There is only one version of the standard, the current version AS/NZA 4935:2009, which was approved on behalf of the Council of Standards Australia on August 28, 2009, and on behalf of the Council of Standards New Zealand on October 23, 2009. It was published on November 17, 2009.

1. Scope

AS/NZS 4935 aims to address furniture tip-over hazards to children. It describes test methods for determining the stability of domestic freestanding chests of drawers over 500 mm (19.7 inch) high, freestanding wardrobes over 500 mm high (19.7 inch), and freestanding bookshelves/bookcases over 600 mm (23.6 inch) high. It defines “chest of drawers” as containing one or more drawers or other extendible elements and intended for the storage of clothing, and may have one or more doors or shelves. It defines “wardrobe” as a furniture item primarily intended for hanging clothing that may also have one or more drawers, doors or other extendible elements, or fixed shelves. It defines bookshelves and bookcases as sets of shelves primarily intended for storing books, and may contain doors, drawers or other extendible elements.

2. Stability Requirements

Similar to ASTM F2057-19, AS/NZS 4935 includes two stability requirements. The first requires the unit, when empty, to not tip over when a 29-kilogram (64-pound) test weight is applied to a single open drawer. The 64-pound test weight is based on the 95th percentile body mass of a 5-year-and-11-month-old child (which is 27 kilograms or 59.5 pounds), adjusted to reflect trends of increasing body mass. The test weight is applied to the top face of a drawer, with the drawer opened to two-thirds of its full extension length. The second test requires the unit not tip over when all of the extension elements are open and the unit is empty. Each drawer or extendible element is open to two-thirds of its extension length, and doors are open perpendicular to the furniture. Units do not pass the stability requirements if they cannot support the test weight, if they tip over, or if they are only prevented from tipping by an extendible element.

3. Tip Restraint Requirements

The standard does not require, but recommends, that tip restraints be included with units, along with attachment instructions.

4. Labeling Requirements

The standard requires a warning label, and provides example text that addresses the tip-over hazard. The standard also requires a warning tag with specific text and formatting. The label and tag include statements informing consumers about the hazard, warning of tip overs and resulting injuries, and indicating how to avoid the hazard. These requirements do not address the use of televisions. The standard includes label permanency requirements and mandates that the warning label be placed “inside of a top drawer within clear view when the drawer is empty and partially opened, or on the inside face of a drawer” for chests of drawers and wardrobes.

5. Assessment of Adequacy

CPSC does not consider the stability requirements in AS/NZS 4935 adequate to address the CSU tip-over hazard because they do not account for multiple open and filled drawers, carpeted flooring, and dynamic forces generated by children's interactions with the CSU, such as climbing or pulling on the top drawer. As discussed in this preamble, these factors are commonly involved in CSU tip-over incidents and testing indicates that they decrease the stability of CSUs.

AS/NZS 4935 requires drawer extension to only two-thirds of extension length for both stability tests. This partial extension does not represent real-world use because children are able to open drawers fully, incidents involve fully open drawers, and opening a drawer further decreases the stability of a CSU. In addition, it does not account for filled drawers, which are expected during real-world use, are common in tip-over incidents, and contribute to instability when multiple drawers are open. It also does not account for carpeted floors, which are common in incidents and contribute to instability. Although AS/NZS 4935 uses a heavier test weight than ASTM F2057-19, it is inadequate because neither stability test accounts for the moments children can exert on CSUs during interactions, such as climbing. Considering additional moments, the 64 pounds of weight on the drawer face is equivalent to a 40-pound child climbing the extended drawer. A 40-pound weight corresponds to a 75th percentile 3-year-old child, 50th percentile 4-year-old child, and 25th percentile 5-year-old child.
46

46
Fryar, C.D., Carroll, M.D., Gu, Q., Afful, J., Ogden, C.L. (2021). Anthropometric reference data for children and adults: United States, 2015-2018. National Center for Health Statistics. Vital Health Stat 3(46).

C. ISO 7171 (2019)

The International Organization for Standardization (ISO) developed the voluntary standard ISO 7171 through the Technical Committee ISO/TC 136,
Furniture
and published the first version in May 1988. The current 2019

version was published in February 2019.

1. Scope

ISO 7171 (2019) describes methods for determining the stability of freestanding storage furniture, including bookcases, wardrobes, and cabinets, but the standard does not define these terms.

2. Stability Requirements

ISO 7171 (2019) includes three stability tests, all of which occur on a level test surface. The first uses a weight/load on an open drawer. The second involves all drawers being filled and a load/weight placed on a single open drawer. In the loaded test, one drawer is opened to the outstop, and if no outstops exist, the drawer is opened to two-thirds of its full extension length. The test weight is applied to the top face of the opened drawer, and varies depending on the height of the unit (either 200 N (44 pounds) or 250 N (55 pounds)). The fill weight is also variable, depending on the clearance height and volume of the drawer (fill density ranges from 6.25 lb/ft
3
to 12.5 lb/ft
3
). The third test is an unloaded test with all drawers open. For this test, drawers and extendible elements are open to the outstop and doors are open 90 degrees. If there are no outstops, then the extension elements are open to two-thirds of their extension length. Existing interlock systems are not bypassed for this test.

ISO 7171 (2019) does not include criteria for determining whether a unit passed or failed the loaded stability test. However, it includes a table of “suggested” forces, depending on the height of the unit.

An additional unfilled, closed drawer test is required for units greater than 1000 mm in height, where a vertical force of 350 N (77 pounds) along with a simultaneous 50 N (11 pounds) outward horizontal force is applied to the top surface of the unit.

3. Tip Restraint Requirements

ISO 7171 (2019) does not require tip restraints to be provided with units, but does specify a test method for them. The tip restraints are installed in both the wall and unit during the test and a 300 N (67.4 lbf) horizontal force is applied in the direction most likely to overturn the unit. The force is maintained between 10 and 15 seconds.

4. Labeling Requirements

The standard does not have any requirements or test methods related to warning labels.

5. Assessment of Adequacy

CPSC does not consider the stability requirements in ISO 7171 (2019) adequate to address the CSU tip-over hazard because they do not account for carpeted flooring, or dynamic and horizontal forces generated by children's interactions with the CSU, such as climbing or pulling on the top drawer. In addition, although ISO 7171 (2019) includes a stability test with filled drawers, the multiple open drawer test does not include filled drawers, and the simultaneous conditions of multiple open and filled drawers during a child interaction are not tested. As discussed in this preamble, these factors are commonly involved in CSU tip-over incidents and testing indicates that they decrease the stability of CSUs. Finally, test weights are provided only as recommendations and there are no criteria for determining whether a unit passes.

D. EN 14749: 2016

EN 14749: 2016 is a European Standard that was prepared by Technical Committee CEN/TC 207 “Furniture.” This standard was approved by the European Committee for Standardization (CEN) on November 21, 2015, and supersedes EN 14749:2005, which was approved on July 8, 2005, as the original version. EN 14749:2016 is a mandatory standard and applies to all CEN members.

1. Scope

EN 14749: 2016 describes methods for determining the stability of domestic and non-domestic furniture with a height ≥ 600 mm (23.6 in) and a potential energy, based on mass and height, exceeding 60 N-m (44.25 ft-lbs). Kitchen worktops and television furniture are the only furniture types defined. The test methods in this standard are taken from EN 16122: 2012,
Domestic and non-domestic storage furniture-test methods for the determination of strength, durability and stability,
which covers “all types of domestic and non-domestic storage furniture including domestic kitchen furniture.”

2. Stability Requirements

EN 14749: 2016 includes three stability tests, which are conducted with the units freestanding. In the first loaded test, a 75 N (16.9 lbf) test weight is applied to the top of the drawer face, when pulled to the outstop. However, if no outstops exist, the extension element is open to two-thirds of its full extension length. In the second test, all drawers and extendible elements are open to the outstop and doors are open 90 degrees. If no outstops are present, then the extension elements are open to two-thirds of their extension lengths. Existing interlock systems are not bypassed for this test. The third test involves filled drawers and a load; all storage areas are filled with weight and the loaded test procedure (above) is carried out but with a test weight that is 20 percent of the mass of the unit, including the drawer fill, not exceeding 300 N (67.4 pounds). Similar to ISO 7171, an additional unfilled, closed drawer test is required for units greater than 1000 mm in height, where a vertical force of 350 N (77 pounds) along with a simultaneous 50 N (11 pounds) outward horizontal force are applied to the top surface of the unit.

Relevant to the portions of stability testing that involve opening drawers, the standard also accounts for interlock systems, requiring one extension element to be open to its outstop, or in the absence of an outstop, two-thirds of its operational sliding length, and a 100 N (22 lbf) horizontal force to be applied to the face of all other extension elements. This is repeated 10 times on each extension element and all combinations of extension elements are tested.

3. Tip Restraint Requirements

EN 14749: 2016 does not include any requirements regarding tip restraints.

4. Labeling Requirements

EN 14749: 2016 does not include any requirements regarding warning labels.

5. Assessment of Adequacy

CPSC does not consider the stability requirements in EN 14749: 2016 adequate to address the CSU tip-over hazard because they do not account for carpeted flooring, or dynamic and horizontal forces generated by children's interactions with the CSU, such as climbing or pulling on the top drawer. In addition, although the standard includes a stability test with filled drawers, the multiple open drawer test does not include filled drawers, and the simultaneous conditions of multiple open and filled drawers during a child interaction are not tested. Moreover, the fill weight ranges from 6.25 lb/ft
3
to 12.5 lb/ft
3
, which includes fill weights lower than staff identified for drawers filled with clothing (discussed in section VII.A. Multiple Open and Filled Drawers of this preamble). As discussed in this preamble, these factors are commonly involved in CSU tip-over incidents and testing indicates that they effect the stability of CSUs.

E. ANSI/BIFMA SOHO S6.5-2008 (R2013)

ANSI/SOHO S6.5 does not address CSUs, but rather, applies to office furniture, such as file cabinets. However, CPSC considered this standard because it addresses interlock systems, which some CSUs include and are relevant to stability testing. This standard was completed by BIFMA Engineering Committee and its subcommittee on Small Office/Home Office Products in 2000. The first version was approved by ANSI on August 4, 2008. The current version of the standard was approved on September 17, 2013.

This standard specifies tests for “evaluating the safety, durability, and structural adequacy of storage and desk-type furniture intended for use in the small office and/or home office.” ANSI/BIFMA SOHO S6.5 includes testing to evaluate interlock systems. The test procedure calls for one extendable element to be fully extended while a 30 lbf horizontal pull force is applied to all other fully closed extendable elements. Every combination of open/closed extendable elements
47

must be tested. The interlock system must be fully functional at the completion of this test and no extendable element may bypass the interlock system.

47
Excluding doors, writing shelves, equipment surfaces, and keyboard surfaces.

As discussed in section VIII.B.2.a.ii Interlocks of this preamble, child strength studies show that children between 2 and 5 years old can achieve a mean pull force of 17.2 pounds. Therefore, CPSC considers a 30-pound horizontal pull force adequate to evaluate the strength of an interlock system. However, because ANSI/SOHO S6.5 does not include stability tests or requirements reflecting the real-world factors involved in CSU tip overs, the standard would not adequately address the CSU tip-over hazard.

VI. Technical Background

This preamble and the NPR briefing package include technical discussions of engineering concepts, such as center of gravity (also referred to as center of mass), moments, and fulcrums. Tab D of the NPR briefing package provides detailed background information on each of these terms, including how staff applies them to CSU tip-over analysis. This section provides a brief overview of that information; for further information, see Tab D of the NPR briefing package.

A. Center of Gravity and Center of Mass

Center of Gravity (CG) or Center of Mass (CM)
48

is a single point in an object, about which its weight (or mass) is completely balanced. In terms of freestanding CSU stability, if the CSU's CG is located behind the front foot, the CSU is stable and will not tip over on its own. Alternatively, if the CSU's CG is in front of the front foot, the CSU is unstable and will tip over. The CG (and CM) of an object is dependent on its geometry and materials. For example, CSU drawers typically have a front that is thicker and larger than the back, which causes the drawer's CG to be closer to the front. The CSU's CG is defined by the position and weight of the CSU cabinet (without drawers), combined with the position and weight of each drawer. A CSU's CG is equal to the sum of the products of the position and the weight of each component, divided by the total weight.

48
For CSU-sized objects, CG and CM are effectively the same. Therefore, CG and CM are used interchangeably in this preamble.

The CG of a CSU will change as a result of the position of the drawers, doors, and pull-out shelves (open or closed). Opening extendable elements, such as drawers, shifts the CG towards the front of the CSU. The closer the CG is to the front leg, the easier it is to tip forward if a force is applied to the drawer. Therefore, CSUs will tip more easily as more drawers are opened. The CG of a CSU will also change depending on the position and amount of clothing in each drawer. Closed drawers filled with clothing tend to stabilize a CSU, but as each filled drawer is pulled out, the CSU's CG will shift further towards the front.

B. Moment and Fulcrum

Moment, or torque, is an engineering term to describe rotational force acting about a pivot point, or fulcrum. The moment is created by a force or forces acting at a distance, or moment arm, away from a fulcrum. One simple example is the moment or torque created by a wrench turning a nut. The moment or torque about the nut is due to the perpendicular force on the end of the wrench applied at a distance (moment arm) from the fulcrum (nut). Likewise, a downward force on an open CSU drawer creates a moment about the fulcrum (front leg) of the CSU. A CSU will tip over about the fulcrum due to a force (
e.g.,
weight of a child positioned over the front of a drawer) and the moment arm (
e.g.,
extended drawer).

Downward force or weight applied to the drawer tends to tip the CSU forward around the fulcrum at the base of the unit, while the weight of the CSU opposes this rotation. The CSU's weight can be modeled as concentrated at a single point: The CSU's CG. The CSU's stability moment is created by its weight, multiplied by the horizontal distance of its CG from the fulcrum. A child can produce a moment opposing the weight of the CSU, by pushing down or sitting in an open drawer. This moment is created by the vertical force of the child, multiplied by the horizontal distance to the fulcrum. The CSU becomes unbalanced and tips over when the moments applied at the front of the CSU exceed the CSU's stability moment.

Horizontal forces applied to pull on a drawer also tend to tip the CSU forward around the front leg (pivot point or fulcrum) at the base of the unit, while the weight of the CSU opposes this rotation. In this case, the moment produced by the child is the horizontal pull force transmitted to the CSU (for example, through a drawer stop), multiplied by the vertical distance to the fulcrum. The CSU becomes unbalanced and tips over when the moments applied at the front of the CSU exceed the CSU's stability moment.

When a child climbs a CSU, both horizontal forces and vertical forces acting at the hands and feet contribute to CSU tip over. Figure 1 shows a typical combination of forces acting on a CSU while a child is climbing, and it describes how those forces contribute to a tip-over moment. Note that when the horizontal force at the hands and feet are approximately equal, which will occur when the child's CM is balanced in front of the drawers, the height of the bottom drawer becomes irrelevant when determining the tip-over moment. In this case, only the height of the hands above the feet matters. As Figure 1 shows, a child climbing on drawers opened distance A1 from the fulcrum, with feet at height B1 from the ground and hands at height B2 above the feet, will act on the CSU with horizontal forces F
H
and vertical forces F
V
. The CSU's weight at a distance A2 from the CSU's front edge touching the ground creates a stabilizing moment. The CSU will tip if Moment 1 is greater than Moment 2.

EP03FE22.000

VII. Technical Analysis Supporting the Proposed Rule

In addition to reviewing incident data, CPSC staff conducted testing and analysis, analyzed tip-over incidents, and commissioned several contractor studies to further examine factors relevant to CSU tip overs. This section describes that testing and analysis.

A. Multiple Open and Filled Drawers
49

49
Further details about the effect of open and filled drawers on CSU stability is available in Tab D, Tab L, and Tab O of the NPR briefing package.

Staff's technical analysis, as confirmed by testing, indicates that multiple open drawers decrease the stability of a CSU, and filled drawers further decrease stability when more than half of the drawers by volume are open, but increase stability when more than half of the drawers by volume are closed. Thus, while multiple open drawers, alone, can make a unit less stable, whether the drawers are full when open is also a relevant consideration. When filled drawers are closed, the clothing weight contributes to the stability of the CSU, because the clothing weight is behind the front legs (fulcrum). However, open drawers contribute to the CSU being less stable, because the clothing weight is shifted forward in front of the front legs (fulcrum).

To assess the effect of open drawers and filled drawers on CSU stability, CPSC staff conducted testing to evaluate the effect of various combinations of open/closed and filled/empty drawers using a convenience sample of CSUs.
50

Staff conducted two phases of testing (Phase I and Phase II). The purpose of the testing was to assess the weight at which a CSU became unstable and tipped over with various configurations of drawers open/closed and filled/empty.

50
Because of the limited number of units tested, this study provides useful information, but the results are limited to the tested units.

The primary variable of interest in the Phase I study was the influence of multiple open/closed drawers. The 11 CSUs tested in Phase I were primarily units with a single column of drawers. The Phase II study examined the influence of multiple open/closed drawers and filled/empty drawers. The 15 CSUs tested in Phase II included more complex units with multiple columns of drawers. Staff used the stability test methods in ASTM F2057-19, with some alterations, to collect information about variables that ASTM F2057-19 does not address (
i.e.,
the effect of open/closed drawers, filled/empty drawers, and tip weight). Filled drawers contained weight bags to simulate a drawer filled with clothing, based on the interior volume of the drawer and 8.5 pounds per cubic foot (the explanation for this fill volume is provided below). In addition to various configurations of open/closed and filled/empty drawers, staff also varied the drawer on which the tip weight mechanism was applied, referred to as the “tip weight application location.”

The primary goal of the Phase I study was to gain insight into the influence of multiple open or closed drawers on CSU stability as a function of tip weight. Additionally, this study was designed to test and ideally confirm that identical drawer open/closed patterns (
e.g.,
two open drawers) yielded nearly identical tip weights, particularly when drawers were identical in size, regardless of the specific configuration (drawers open/closed and tip weight application location). The Phase I study confirmed that comparable tip weights existed for similar open/closed drawer configurations in the tested CSUs when considering a simple single column of drawers that are identically sized.

The primary goal of the Phase II study was to examine additional complexities with respect to real-world scenarios of CSUs. This included more complex CSUs and combinations of filled and/or empty drawers (including partially filed configurations, in which some drawers were filled and some were empty) within the same CSU, in addition to open/closed drawers. Staff also modified the test method to decrease

test-to-test variability, for example, by adding cross hatches on the drawer and the weight bag to ensure weight bags were centered within drawers.

Based on this testing, lighter and shorter units appear to be less stable, although a taller and heavier unit was also unstable; and similar units passed and failed ASTM's stability requirements. This suggests that specific heights or weights of a CSU do not correlate with stability or instability. Similarly, the footprint ratio (depth-to-width ratio) of the CSU, alone, did not appear to affect tip weight.

From the 26 CSUs tested, CPSC staff analyzed 1,777 data points for a variety of combinations (filled/empty drawers, open/closed drawers, and tip weight application location),
51

and supplemented this data with results from other CSU testing CPSC staff had performed. The results of this testing indicated that individual CSUs vary in stability, depending on the configuration of open/closed drawers, and filled/empty drawers, and that different CSU drawer structures (
e.g.,
number of columns, relative drawer sizes) have an influence on tip weight. In general, the results indicated that CSUs were less stable as more drawers were opened, and that filled drawers have a variable effect on stability. A filled closed drawer contributes to stability, while a filled open drawer decreases stability. Depending on the percent of drawers that are open and filled, having multiple drawers open decreased the stability of the CSU.

51
Staff excluded some data points for reasons explained in Tab O of the NPR briefing package.

To determine the appropriate method for simulating CSU drawers that are partially filled or fully filled, staff considered previous analyses, and conducted additional testing. Although ASTM F2057-19 does not include filled drawers as part of its stability testing, the ASTM F15.42 subcommittee has considered a “loaded” (filled) drawer requirement and test method. The ASTM task group used an assumed clothing weight of 8.5 pounds per cubic foot in testing and other discussions of filled drawers. Kids in Danger and Shane's Foundation found a similar density (average of 8.9 pounds per cubic foot) when they filled CSU drawers with boys' t-shirts in a 2016 study on furniture stability.
52

52
Kids in Danger and Shane's Foundation (2016). Dresser Testing Protocol and Data. Data set provided to CPSC staff by Kids in Danger, January 29, 2021.

To assess whether 8.5 pounds per cubic foot reasonably represents the weight of clothing in a drawer, CPSC staff conducted testing. As part of this assessment, staff looked at four drawer fill conditions. Staff considered folded and unfolded clothing with a total weight equal to 8.5 pounds per cubic foot of functional drawer volume in the drawer; and the maximum amount of folded and unfolded clothing that could be put into a drawer that would still allow the drawer to open and close. For these tests, staff used an assortment of boys' clothing in sizes 4, 5, and 6. Staff used a CSU with a range of drawer sizes to assess small, medium, and large drawers; the functional drawer volume of these 3 drawer sizes was 0.76 cubic feet, 1.71 cubic feet, and 2.39 cubic feet, respectively. Staff determined the calculated clothing weight for the 8.5 pounds per cubic foot drawer fill conditions by multiplying 8.5 by the drawer's functional volume, defined as:
53

53
“Clearance height” is the height from the interior bottom surface of the drawer to the closest vertical obstruction in the CSU frame. “Functional height” is clearance height minus
1/8
inch.

EP03FE22.001

For all three drawer sizes, staff was able to fit 8.5 pounds per cubic foot of folded and unfolded clothing in the drawers. When the clothing was unfolded, the clothing fully filled the drawers, but still allowed the drawer to close. Because the unfolded clothing was stuffed into the drawer fairly tightly, it was not easy to see and access clothing below the top layer. When the clothing was folded, the clothing also fully filled the drawers and still allowed the drawer to close. The folded clothing was tightly packed, but allowed for additional space when compressed. The maximum unfolded clothing fill weight was 6.52, 14.64, and 21.20 pounds for the three drawer sizes, respectively; and the maximum folded clothing fill weight was 7.72, 16.08, and 22.88 pounds for the three drawer sizes, respectively.

Staff also compared the calculated clothing weight (
i.e.,
using 8.5 pounds per cubic foot), maximum unfolded drawer fill weight, and maximum folded drawer fill weight for each drawer. The maximum unfolded clothing fill weight was slightly higher than the calculated clothing fill weight for all tested drawers. The difference between the maximum unfolded clothing fill weight and the calculated clothing weight ranged from 0.08 pounds to 0.87 pounds. The maximum unfolded clothing fill weight was 101 to 104 percent of the calculated clothing weight, depending on the drawer. The maximum folded clothing fill weight was higher than both the maximum unfolded clothing fill weight and the calculated clothing fill weight for all tested drawers; however, the differences were relatively small. The difference between the maximum folded clothing fill weight and the calculated clothing weight ranged from 1.28 to 2.55 pounds. The maximum unfolded clothing fill weight was 111 to 120 percent of the calculated clothing weight, depending on the drawer. The maximum unfolded clothing fill density was slightly higher than 8.5 pounds per cubic foot for all tested drawers; and the maximum unfolded clothing fill density ranged from 8.56 to 8.87 pounds per cubic foot, depending on the drawer. The maximum folded clothing fill density was higher than both the maximum unfolded clothing fill density and 8.5 pounds per cubic foot for all tested drawers. The maximum folded clothing fill density ranged from 9.40 to 10.16 pounds per cubic foot, depending on the drawer. Thus, there does not appear to be a large difference in clothing fill density based on drawer size.

Based on this testing, staff found that 8.5 pounds per cubic foot of clothing will fill a drawer; however, this amount of clothing is less than the absolute maximum amount of clothing that can be put into a drawer, especially if the clothing is folded. The maximum amount of unfolded clothing that could be put into the tested drawers was only slightly higher than 8.5 pounds per cubic foot. Although staff achieved a clothing density as high as 10.16 pounds per cubic foot with folded clothing, consumers may be unlikely to fill a drawer to this level because it requires careful folding, and it is difficult to remove and replace individual pieces of clothing. On balance, staff concluded that 8.5 pounds per cubic foot of functional drawer volume is a reasonable approximation of

the weight of clothing in a fully filled drawer.

B. Forces and Moments During Child Interactions With CSUs
54

54
Further information about the study described in this section, and forces and moments generated by children's interactions with CSUs, is available in Tab C, Tab D, and Tab R of the NPR briefing package.

As indicated above, some of the common themes that staff identified in CSU tip-over incident data involve children interacting with CSUs, including climbing on them and opening drawers. To determine the forces and other relevant factors that exist during these expected interactions between children and CSUs, CPSC contracted with UMTRI to conduct research. The researchers at UMTRI, in collaboration with CPSC staff, designed a study to collect information about children's measurements and proportions, interest in climbing and climbing behaviors, and the forces and moments children can generate during various interactions with a CSU. Forty children, age 20 months to 65 months old, participated in the study. For additional details about the study, see UMTRI's full report in Tab R of the NPR briefing package.

1. Overview of Interaction Portion of UMTRI Study

The interaction portion of the study included children interacting with a CSU test apparatus with instrumented handles and a simulated drawer and tabletop (to simulate the top of a CSU or other tabletop or furniture unit). Researchers measured the forces of the children acting on the test apparatus and calculated moments generated by the children based on the location of the CSU's front leg tip point (fulcrum). The researchers based the fulcrum's location on a dataset of CSU drawer extensions and heights provided by CPSC staff.
55

55
CPSC staff provided UMTRI researchers with a dataset of drawer extensions and drawer heights from the ground from a sample of approximately 180 CSUs. The researchers selected the 90th percentile drawer extension (12 inches) and drawer height (16 inches) as the basis for placing the moment fulcrum in most of their analysis.

The interaction portion of the study looked at forces associated with several climbing-related interactions of interest, which staff and researchers selected based on CSU tip-over incidents, videos of children interacting with CSUs and similar furniture items, and plausible interactions based on children's developmental abilities. Staff focused on the ascent/climbing
56

interaction for this rulemaking because climbing incidents were the most common interaction among fatal CPSRMS incidents and nonfatal NEISS incidents, where the interaction was reported, and they were the second most common interaction in nonfatal CPSRMS incidents, where the interaction was reported; and because climbing begins with ascent, which is a child's initial step to climb up on to the CSU, and therefore, is considered an integral part of all climbing interactions.

56
Ascending is a subcategory of climbing, and is described as a child's initial step to climb up on to a CSU. Therefore, ascending is an integral part of climbing. The UMTRI study provided information about forces children generate during ascent, because that testing measured forces children generate during an initial step onto the CSU test fixture. Those forces can be used to model children climbing because ascent is the first and integral step to climbing, but not all climbing interactions can be modeled with ascent, as forces associated with some other behaviors can exceed those for ascent. The term “climbing” is often used in this preamble and the NPR briefing package because that is the general behavior described in many incidents. Both climbing and ascending are used to refer to the force children generate on a CSU, for purposes of the proposed rule.

2. Test Apparatus and Data Acquisition

UMTRI researchers created the test apparatus shown in Figure 2, which used a padded force plate to measure interactions with the floor and included a column to which the various instrumented test fixtures were attached. Tests were conducted with a pair of handlebars (simulating drawer handles or fronts), a simulated drawer, and a simulated tabletop. In preparation for the study, CPSC staff worked with UMTRI researchers to develop a test fixture that modeled the climbing surfaces of a CSU. CPSC staff provided information to UMTRI researchers on drawer extension and heights from the sample of dressers used in CPSC staff's evaluation (Tab N of the NPR briefing package). Researchers selected and constructed a parallel bar test fixture, representing a lower foothold and an upper handhold. These bars represent a best-case CSU climbing surface, similar to the top of a drawer.

UMTRI researchers configured the test fixtures based on each child's anthropometric measurements. Researchers set the upper bar to three different heights relative to the padded floor surface: Low (50 percent of the child's upward grip reach), mid (75 percent of the child's upward grip reach), and high (100 percent of the child's upward grip reach); researchers set the lower bar to two different heights: Low (4.7 inches from the padded floor surface) and high (the child's maximum step height above the padded floor). The heights for the bars were within plausible heights for CSU drawers. Researchers set the horizontal position of the upper bar to two different positions: “aligned” with the lower bar, or “offset” from the lower bar, at a distance equal to 20 percent of the child's upward grip height. Tabs C and R of the NPR briefing package contain more information about the test fixture configurations. The bars, drawer, and tabletop, as well as the floor in front of the test fixture, had force measurement instrumentation that recorded forces over time in the horizontal (fore-aft, x) and vertical (z) directions.

BILLING CODE 6355-01-P

EP03FE22.002

3. Target Behaviors of Children Interacting With a CSU

CPSC staff worked with UMTRI researchers to develop a set of scripted interactions. Staff focused on realistic interactions in which the child's position and/or dynamic interactions were the most likely to cause a CSU to tip over. The interactions were based on incident data and online videos of children interacting with CSUs and other furniture items. The interactions UMTRI researchers evaluated included:

•
Ascend:
Climb up onto the test fixture;

•
Bounce:
Bounce vigorously without leaving the bar;

•
Lean back:
Lean back as far as possible while keeping both hands and feet on the bars;

•
Yank:
From the lean back position, pull on the bar as hard as possible;

•
1 hand & 1 foot:
Take one hand and foot (from the same side of the body) off the bars and then lean as far away from the bars as possible;

•
Hop up:
Hold the upper bar and try to jump from the floor to a position where the arms are straight and the hips are in front of the upper bar, an action similar to hoisting oneself out of a swimming pool;

•
Hang:
Hold onto the upper bar, lift feet off the floor by bending knees, hang still for a few seconds, and then straighten legs to return to the floor; and

•
Descend:
Climb down from the test fixture.

As described above, the ascend interaction best models the climbing behavior commonly seen in incidents, and is analogous to a child's initial step to climb up on to the CSU, which is an integral climbing interaction. The other, more extreme interactions, such as bounce, lean, and yank, were identified as plausible interactions, based on child behavior; but these interactions were not directly observed in the incident data.

After the children performed the interaction, the researchers reviewed video from each trial to isolate and characterize interactions of interest. Interactions of interest for the handle trials were categorized as: Ascent, Bounce, Lean (lean back), Yank, and One Hand (see Figure 3). Researchers analyzed forces from each extracted behavior to identify peak forces and moments.

EP03FE22.003

4. Image-Based Posture Analysis

Participant postures have strong effects on the horizontal forces exerted by the child and the subsequent calculated moments, due to the location of the child's CM during each behavior. Thus, the CM of the child is important when evaluating the stability or tip-over propensity of the child/CSU-combined system. UMTRI researchers used the images of the subjects to estimate the location of the child's CM. The UMTRI researchers extracted video frames at time points of interest (typically when the child produced the maximum moment during the interaction) and manually digitized the series of landmarks on the image of the child, as shown in Figure 4. The location of the CM was estimated, based on anthropometric information on children,
57

as 33 percent of the distance from the buttock landmark to the top-of-head landmark.

57
Snyder, R.G., Schneider, L.W., Owings, C.L., Reynolds, H.M., Golomb, D.H., Schork, M.A., Anthropometry of Infants, Children and Youths to Age 18 for Product Safety Design (Report No. UM-HSRI-77-17), prepared for the U.S. Consumer Product Safety Commission (1977).

EP03FE22.004

The UMTRI researchers estimated the location of the child's CM by examining the side-view images from the times of maximum moment, as shown in Figure 5. Table 1 shows the average estimated CM location for each behavior.
58

The children in the study extended their CM an average of about 6 inches from the handle/foothold while ascending.

58
Graphs are available in Tab R of the NPR briefing package (page 59, Figure 54).

EP03FE22.005

Table 1—Estimated CM Horizontal Offset From the Handles for Aligned Trials
[Inches]

Behavior
N subjects
N trials
Mean
SD

10th
percentile

50th
percentile

90th
percentile

Ascent
36
109
6.1
2.0
4.3
6.1
8.6

Bounce
32
80
6.0
2.5
4.0
5.8
9.1

Lean Back
30
81
11.3
3.4
8.5
11.6
15.9

Yank
25
53
10.9
3.4
7.3
11.5
15.9

5. Handle Trial Force Results

Figure 6 shows side-view images of examples of children interacting with the handle fixture. The frames were taken at the time of peak tip-over moment. Forces exerted by the child at the hands and feet are illustrated using scaled vectors (longer lines indicate greater force magnitude; arrow direction indicates force direction). Digitized landmarks and estimated CM locations are shown. The images demonstrate that forces at both the hands and feet often have substantial horizontal components, and usually, but not always, the foot forces are larger than the hand forces. The horizontal components at the hands and feet are also in opposite directions: The horizontal foot forces are forward (toward the test fixture), while the hand forces are rearward (toward the child).

EP03FE22.006

Figure 17 in Tab D of the NPR briefing package shows an exemplar time-history plot of the horizontal and vertical forces for the Ascent behavior of the depicted child. As that figure illustrates, the child's body weight transitions from the force plate to the bars, with the lower bar bearing nearly all of the weight. The horizontal forces on the upper and lower bars are approximately equal in magnitude and opposite in direction, consistent with the posture being approximately static toward the end of the test, where the child completed the ascend maneuver. Under these conditions, the behavior is no longer dynamic, and the vertical forces sum to body weight.

UMTRI researchers modeled a child interacting with a CSU with opened drawers, by measuring forces at instrumented bars representing a drawer front or handle. Figure 7 is the free-body diagram of the child climbing the CSU. The horizontal and vertical forces at the hands and feet correspond to the positive direction of the measured forces. The CSU drawers were modeled using the top handle and bottom handle height, and the drawer extension was modeled from 0 inches to 12 inches.
59

The UMTRI researchers calculated the moment about the CSU's front foot or fulcrum, using the measured forces, vertical location of the top and bottom handles, and the defined drawer extension length (Fulcrum X).

59
Here, 0 inches corresponds with a closed drawer when the fulcrum lines up with the drawers. Additionally, 12 inches represents the 90th percentile drawer extension length in a dataset of approximately 180 CSUs.

EP03FE22.007

Figure 7 shows that the child's body weight will generally be distributed between the two bars, but that the child's CM location will also typically be outboard of the bars (farther from the fulcrum than the bars). The quasi-static

climbing moment is approximately equal to the location of the child's CM (the horizontal distance of the CM to the fulcrum), multiplied by the child's weight. In reality, the moment created by dynamic forces generated by the child during the activities in the UMTRI study, such as during ascend, exceed the moment created by body weight alone as a result of the greater magnitude horizontal and vertical forces.

6. Moment About the Fulcrum

UMTRI researchers analyzed the force data as generating a moment around a tip-over fulcrum. The UMTRI researchers calculated the maximum moment about a virtual fulcrum, based on the measured force data for each test and the location of the force. Figure 8 shows the test setup and the forces measured. Note that the test setup mimics a CSU with the drawers closed an

[Text truncated at 120,000 characters. The full text is on the page linked above.]

---

Source: Frix Law Library, https://www.frixlaw.com/law-library/documents/fr%3A2022-01689. Public record. Not legal advice.