Infant Walker–Related Injuries in the United States

The National Electronic Injury Surveillance System (NEISS) is managed by the CPSC and collects data on the injuries of patients who are treated in US EDs. The NEISS obtains data from ∼100 hospitals, representing a stratified probability sample of >5300 hospitals with a 24-hour ED and at least 6 beds in the United States and its territories.^22,23 Data are weighted to provide national injury estimates.

In this study, we retrospectively examined infant walker–related injuries among children <15 months old who were treated in US EDs from January 1, 1990, to December 31, 2014. Injuries that involved infant walkers were identified by using the NEISS product code 1508 for “baby walkers and jumpers.” Injuries that were not directly related to the infant walker, such as “fell off couch and struck walker,” were excluded. Injuries involving non–infant walker products, such as “baby bouncer,” “baby jumper,” “stroller,” or “infant swing,” were excluded.

NEISS case narratives were used to identify the mechanism of injury and, when applicable, the number of steps that the child fell down. The mechanism of injury was classified as (1) fell down the stairs (including fell from any elevated height, such as porches or decks), (2) fell out of infant walker (including tipped over, flipped over, and climbed out), (3) proximity mechanism (including pulled on object and direct thermal contact), and (4) other (including struck surface, struck by object or person, entanglement, product malfunction, removed from infant walker by person, and other specified mechanisms). If multiple injury mechanisms were referenced in the narrative, the first mechanism in the sequence of events was coded as the mechanism of injury. For example, if a child fell out of an infant walker and struck the edge of a table, the injury mechanism was coded as fell out of infant walker rather than struck object or person. The number of steps that the child fell down was grouped into 2 categories: ≤10 steps or >10 steps. For narratives specifying a range for the number of steps (eg, 4–10 steps), the median number of steps was used.

Body region that was injured was categorized as (1) head or neck (including head, neck, face, eyeball, mouth, and ear), (2) upper extremity (including shoulder, upper arm, elbow, lower arm, wrist, hand, and finger), (3) lower extremity (including upper leg, knee, lower leg, ankle, foot, and toe), and (4) other (including all remaining NEISS body region categories). Diagnoses were grouped as (1) concussion or closed-head injury (CHI) (including internal organ injury to the head), (2) laceration (including laceration, amputation, puncture, and nondental avulsion), (3) burn (all types), (4) soft-tissue injury (including contusion, abrasion, crush injury, hematoma, and strain or sprain), (5) fracture (including skull fracture), and (6) other (including all remaining NEISS diagnoses). If a patient sustained multiple injuries to multiple body regions, only the most severe injury and the most seriously injured body region were included in the data set. Disposition from the ED was classified as (1) treated and released (including examined and released without treatment), (2) admitted (including treated and transferred to another hospital, treated and transferred for hospitalization, treated and admitted for hospitalization within the same facility, and held for observation), and (3) left against medical advice.

Injury location was categorized as (1) home (including house, apartment or condo, farm or ranch, and mobile or manufactured home) and (2) other (including street or highway, industrial place, school or day care, sports or recreational place, and other public property). Other variables included patient age and sex.

Data were analyzed by using IBM SPSS Statistics for Windows version 21.0 (IBM Corporation, Armonk, NY) and SAS Enterprise Guide version 7.15 (SAS Institute Inc, Cary, NC). To account for the NEISS sampling design, complex survey procedures were used to calculate national estimates, and the Taylor series linearization method was used to calculate the variance of estimates. All estimates reported in this study are stable national estimates unless stated otherwise. An estimate is potentially unstable if the sample size is <20 cases, the estimate is <1200 cases, or the coefficient of variation is >33.0%. Preliminary analyses of scatter plots indicated that the number of infant walker–related injuries peaked in 1991 and then decreased through 2014, with the rate of the decrease changing in 2002 for the number of concussions or CHIs and in 2003 for infant walker–related injuries overall. Therefore, trend analyses were performed by using piecewise linear regression with breakpoints at 2002 for concussions or CHIs and at 2003 for infant walker–related injuries overall. The estimated slope from the regression model was reported along with the associated P value. To assess the effectiveness of the 2010 federal safety standard, a t test was used to compare the average annual number of injuries during a 4-year period before (2006–2009) and after (2011–2014) the standard went into effect. Other statistical analyses included the Rao–Scott χ² test for association and calculation of relative risks (RRs) with 95% confidence intervals (CIs). Statistical significance was determined at α = .05. This study was judged exempt by the institutional review board of the authors’ institution.

From 1990 to 2014, an estimated 230 676 (95% CI: 189 808–271 544) children <15 months old were treated in US EDs for infant walker–related injuries, averaging 9227 (95% CI: 7592–10 862) children annually. The median age of children who were injured was 8.2 months (interquartile range: 6.9–9.7), with children 7 to 10 months old accounting for 68.3% of infant walker–related injuries (Fig 1). Most (59.1%) of the reported injuries were among boys, and 98.8% of the injuries occurred at home (Table 1).

The majority (90.6%) of children sustained an injury to the head or neck, and 4.5% of children sustained an injury to the upper extremities (Table 1). Most of the injuries were a soft-tissue injury (50.5%), concussion or CHI (26.5%), or laceration (10.1%). Fractures accounted for 5.1% of the injuries, and of these, 55.4% were skull fractures. Most (95.4%) of the children who were injured were treated and released, and 4.5% were admitted. Among those admitted, 37.8% had a skull fracture, and 29.2% had a concussion or CHI.

Almost three-fourths (74.1%) of infant walker–related injuries resulted from a fall down the stairs (Table 1). The percentage of injuries associated with falling down the stairs decreased from 80.4% during 1990–1996 to 66.4% during 1997–2009 and to 40.6% during 2010–2014. Most (95.9%) of the children who fell down the stairs sustained injuries to the head or neck and were diagnosed with a soft-tissue injury (54.0%) or concussion or CHI (29.8%). The number of steps that the child fell down was available for 78 270 children who fell down the stairs, and among these, 66.3% of children fell down ≤10 steps, and the median number of steps was 5.7 (interquartile range: 2.9–11.8). Among the 6539 children who were diagnosed with a skull fracture, 91.4% fell down the stairs. Patients who fell down >10 steps were more likely to be diagnosed with a fracture (RR: 2.51; 95% CI: 1.69–3.72), especially a skull fracture (RR: 5.04^*; 95% CI: 3.06–8.29), compared with those who fell down ≤10 steps.

A fall out of an infant walker was the second most common mechanism of injury, accounting for 14.7% of injuries. Similar to falling down the stairs, most of these injuries occurred to the head or neck (95.1%) and were diagnosed as a soft-tissue injury (48.1%), concussion or CHI (21.4%), or laceration (21.2%). Although proximity mechanisms accounted for only 2.8% of infant walker–related injuries, 74.0% of these were burn injuries and occurred mostly to the upper extremities (42.8%) or head or neck (39.8%). Among children who were injured via proximity mechanisms, 14.5%^* were admitted, and all were admitted for burns.

The annual number of infant walker–related injuries decreased significantly by 84.5% (slope: −1766.5; P < .001), from 20 650 injuries in 1990 to 3201 injuries in 2003 followed by a nonsignificant 37.5% decline from 2003 to 2014 (slope: −66.0; P = .556) (Fig 2). The overall reduction in injuries was primarily a result of a decline in falls down the stairs. The number of injuries associated with this mechanism decreased significantly by 91.0% (slope: −1500.8; P < .001) from 1990 (16 386 injuries) to 2003 (1473 injuries) and continued to decline by 59.7% from 2003 to 2014 (593^* injuries), although this decrease was not statistically significant (slope: −47.4; P = .628). The number of injuries associated with other mechanisms also decreased significantly by 58.9% (slope: −265.2; P < .001) from 4200 injuries in 1990 to 1728 injuries in 2003 and then decreased nonsignificantly (slope: −20.0; P = .413) by 18.9% to 1402 injuries in 2014. The number of infant walker–related concussions or CHIs decreased by 82.4% (slope: −457.6; P < .001) from 4698 in 1990 to 828^* in 2002 and remained constant (slope: −5.3; P = .881) from 2002 to 2014 (858^* injuries; Fig 3).

There was a 22.7% (P = .019) decrease in the average annual number of infant walker–related injuries during the 4-year period (2165 injuries annually for 2011–2014) after the federal mandatory safety standard went into effect in 2010 compared with the 4-year period before the standard (2801 injuries annually for 2006–2009; Fig 4). Although the average annual number of injuries associated with falls down the stairs decreased by 40.7% (1479 annual injuries in 2006–2009 compared with 878^* annual injuries in 2011–2014; P = .020), the average annual number of injuries associated with other mechanisms increased slightly by 2.3% (1225 annual injuries in 2006–2009 compared with 1253^* annual injuries in 2011–2014; P = .857) after the implementation of the mandatory standard.

From 1990 to 2014, an estimated 230 676 children <15 months old were treated in US EDs for infant walker–related injuries. The characteristics of these injuries are in agreement with those found in studies from other countries.^{8,14,24,–28} Similar to what was reported in previous studies, the majority of infant walker–related injuries resulted from falls down the stairs, were sustained to the head or neck, and were diagnosed as a soft-tissue injury or concussion or CHI.^{2,10,11,24,28} 

There was an 84.5% decline in infant walker–related injuries from 1990 to 2003, which was driven by a 91.0% decrease in falls down the stairs during this period. This reduction is most likely attributable to (1) a decrease in infant walker use after an increase in public awareness of infant walker–related injury risks when advocacy groups petitioned the CPSC in 1992 to ban infant walker sales in the United States, (2) the increasing use of stationary activity centers as an alternative to infant walkers, and (3) improvements in the voluntary infant walker safety standard.⁷ Introduced in the United States in 1994, stationary activity centers resemble infant walkers but are without wheels and allow a child to rock, bounce, and swivel without the mobility to fall down the stairs or approach other hazards.^1,2,20 As use of stationary activity centers increased, infant walker use and injuries began to decrease (Fig 2).²⁰ 

The voluntary safety standard ASTM F977 for infant walkers was first established in 1986.²⁹ It initially addressed tip overs and structural failures, but in 1996, it was revised to address falls down the stairs. Effective in 1997, ASTM F977-96 required that infant walkers be wider than a standard doorway or brake to a halt if ≥1 wheels drop over the edge of a step.^1,2,4,9,20 There were subsequent minor revisions to the voluntary standard until June 2010, when the CPSC issued a mandatory standard (16 CFR 1216) for infant walkers under section 104(b) of the CPSIA, which modified ASTM F977-07 and included more stringent requirements.^7,21 It standardized the evaluation method for falls down the stairs by creating specifications for testing equipment and a new formula for calculating launch distance that incorporated the weight of the infant walker. The mandatory rule also added a parking brake test for infant walkers that had a parking brake, which was adopted from European standards.²¹ These new requirements in the mandatory standard would be expected to primarily affect the likelihood of falls down the stairs. In addition, a mandatory standard makes it easier for the CPSC to stop noncompliant infant walkers at entry points to the United States before they enter the marketplace. This is important because all of the 10 infant walker recalls from 2001 to 2010 were imported products.⁷ 

The average annual number of infant walker–related injuries decreased by 22.7% during the 4-year period after the implementation of the federal mandatory safety standard compared with the previous 4-year period, whereas the average annual number of injuries associated with falls down the stairs decreased by 40.7% between these 2 periods. Although the federal standard may have contributed to this decline, its impact is uncertain because of a number of potentially confounding factors and the lack of a comparison group. The frequency of injuries had been decreasing before its implementation. Other factors, such as changes in infant walker sales and the number in use, the life span of old infant walkers in homes, and infant walker recalls, may have influenced the observed decline in injuries.

Although there was a decrease in the percentage of infant walker–related injuries associated with falling down the stairs from 80% during the beginning of the study period to 41% during the end of the period, falls down the stairs continue to be the leading injury mechanism associated with these products. Friction strips attached to the bottom of infant walkers are the most commonly employed braking feature and are designed to stop falls by gripping the floor if ≥1 wheels go over the edge of the top step. However, they may become ineffective over time because of overuse and accumulated debris, and strong children may be able to bypass the friction strips by lifting up on the walker.^7,17 In addition, the type of floor surface and the child’s weight can influence the effectiveness of braking systems.⁶ Although federal regulations also require infant walkers to meet forward and backward tip-resistance tests to ensure proper stability, 14.7% of infant walker–related injuries in the current study were caused by children falling out of the walker, including from tip overs.¹⁷ 

From 2001 to 2010, there were 10 infant walker recalls because of failure to meet the stair-fall prevention requirement, and given that only an estimated 10% of recalled children’s products are properly corrected, replaced, or returned by consumers, many noncompliant infant walkers remain in homes today.^7,30 In addition, although infant walkers that were purchased before 2010 will eventually be discarded, the mandatory standards will not be able to eliminate the dangers associated with these older products that may still remain in use.

Parental motivation to use infant walkers stems from the perceptions that they offer entertainment, promote walking, and provide an infant with an activity while parents are engaged in other activities.^1,5,20,31 Contrary to some parents’ belief that infant walkers accelerate walking skills or motor development, studies reveal that their use can transiently delay mental and motor development.^{1,2,12,28,31,–34} Infant walkers can give parents a false sense of security. Because children in infant walkers can travel at speeds up to 1 m/second, adults have little time to react to dangerous situations.^1,2 The increased mobility that infant walkers provide to young children increases their risk of proximity injuries from access to items such as heaters, hot beverages, and poisons.⁹ 

Warning labels and educational campaigns have not been shown to be effective strategies for reducing infant walker–related injuries.^1,2 The authors of 1 study found that 59% of families whose child was treated in an ED for an infant walker–related injury reported knowing of the potential dangers associated with the product before the incident, and 32% indicated that they used the device again after the injury.¹⁰ Because infant walkers provide no substantial benefit to children and pose a significant injury risk, Canada banned the manufacture, sale, and importation of infant walkers in 2004, and the AAP recommends that the United States also adopt such a policy.^1,2,7,21,35 Combining a ban with a public education campaign and offering incentives for the return of infant walkers could improve the success of such an initiative.^19,25 

This study has several limitations. In this study, we underestimate the true number of infant walker–related injuries because the NEISS only identifies injuries of patients who are treated in EDs and does not include those who are treated in urgent care centers and other health care settings or those for whom medical care is not sought. The NEISS also does not capture fatalities that are not transported to the ED or that occur after inpatient admission. The level of detail that is contained in NEISS narratives is based on information documented in ED medical records and extracted by NEISS coders and therefore can be inconsistent. Information was not available regarding the safety features that were incorporated into the individual infant walkers in this study. In addition, some estimates in this study are potentially unstable because of small numbers of injuries. We were also unable to quantify the effects of potential confounding factors, such as infant walker sales and use over time. Despite the limitations of this study, we provide a comprehensive epidemiologic update on infant walker–related injuries over a 25-year period using a large, nationally-representative sample with an investigation of the effect of the mandatory federal infant walker safety standard on these injuries.

Infant walker–related injuries decreased after the implementation of the federal mandatory safety standard in 2010. This decrease may, in part, be attributable to the standard as well as other factors, such as decreased infant walker use and fewer older infant walkers in homes. Despite the decline in injuries, infant walkers remain an important and preventable source of injury among young children, which supports the position of the AAP to call for a ban on their manufacture and sale in the United States.

AAP

American Academy of Pediatrics

CHI

closed-head injury

CI

confidence interval

CPSC

Consumer Product Safety Commission

CPSIA

Consumer Product Safety Improvement Act

ED

emergency department

NEISS

National Electronic Injury Surveillance System

RR

relative risk