Abstract
Falls from buildings, including houses, are an important cause of childhood injury in the United States; however, no study has previously examined the impact of this problem in Hawai‘i. The objective of this study is to categorize the demographics and injury circumstances of pediatric falls from buildings in Hawai‘i and compare to other US cities. Patients age 10 and under who were injured in nonfatal accidental falls from buildings in Hawai‘i between 2005 and 2011 were identified retrospectively from a statewide repository of hospital billing data. The Hawai‘i death certificate database was searched separately for deaths in children age 10 and under due to falls from buildings, with data available from 1991 through 2011. Data was reviewed for demographics, circumstances surrounding the injury, and level of hospital treatment. During the 7-year period for nonfatal injuries, 416 fall-related injuries were identified in children age 10 and younger. Of these, 86 required hospitalization. The rate of nonfatal injury in Hawai‘i County was twice that of Honolulu and Maui Counties, and three times that of Kaua‘i County. There were 9 fatal falls over a 21-year period. The population based incidence for nonfatal injuries was three-fold higher than that reported in the city of Dallas. The rate of hospitalizations following building falls was more than twice as high as the national average, and that of New York City, but similar to that of California. Strategies for education and environmental modification are reviewed, which may be helpful in reducing the incidence of pediatric falls from buildings in Hawai‘i.
Introduction
Falls are the leading cause of nonfatal unintentional injury in children.1 Each year in the United States, falls from the windows of buildings account for approximately 8 deaths and 3,300 injuries in children age 5 and under.2 Falls from buildings can result in serious consequences for children, including fractures, lacerations of the liver/spleen, and permanent neurological impairment.3,4 In several US cities, educational initiatives and legislation pertaining to the installation of window guards have been very successful in significantly decreasing, and in some cases nearly eliminating, the number of falls from windows involving children.5 This study sought to define the risk of similar fall related injuries to children in Hawai‘i, in order to determine if such preventive strategies are worthy of consideration locally.
Methods
Fatal and nonfatal injuries were analyzed separately. For data regarding fatal falls, the Hawai‘i Department of Health Death Certificate Database was searched for the underlying cause of death of unintentional falls from 1991 to 2011. Deaths due to falls from buildings in children age 10 and under were included, and information on whether the fall was from a window or balcony was identified.
For data regarding nonfatal falls, the Hawai‘i Health Information Corporation (HHIC) database and the Hawai‘i Trauma Registry (HTR) were searched for records of patients 10 years of age or younger whose reason for visit was identified as “injury-related” (ICD-9 800-995.85), with the E-code E882 (“Fall from or out of building or other structure) from 2005–2011.
The HHIC database receives information from all but one of the hospital-based emergency departments (ED) in the state and all of the admitting hospitals. Patient county of residence was also recorded in this database. Records of patients who died in the hospital or were transferred to another hospital or hospice facility at discharge were excluded to prevent double-counting. The HHIC data did not provide detailed information regarding if the fall was from a window or a balcony.
The HTR was also searched for cases, using the 882 E-code and a subsequent search of a narrative text field that described the cause of injury. The HTR includes data from a smaller subset of patients: those who met criteria for a Trauma Activation — the assembly of a rapid response team consisting of ED physicians, trauma surgeons, and ancillary staff prepared to care for trauma patients - upon arrival to the ED. From 2009 to 2011, the HTR was expanded to include data from eight facilities throughout Hawai‘i; prior to 2009, only data from the major trauma center in the state, Queen's Medical Center, was included. The HTR provides more detailed information on injury circumstances among patients who were admitted to the hospital, and for most patients, information surrounding the circumstances of the fall (from the window, lanai, etc).
Statistical Analyses
Population denominator data for annual injury rate calculations were obtained from the US Census.6 In order to compare the incidence rates in this study with published rates, age-specific incidence rates reported in the literature were converted to weighted averages using the Census 2000 US population data for Dallas County, the State of California, and the US population respectively. Poisson regression analysis was used to compare the incidence rate of pediatric falls among counties and over time in the state of Hawai‘i, and to compare rates between Dallas County, Texas, and Honolulu County, Hawai‘i. All analyses were conducted using SAS software version 9.2 (SAS Institute, Inc., Cary, North Carolina), with statistical significance defined as P < .05.
Results
Nonfatal Injuries
During the 7-year investigation period, there were 416 total nonfatal injuries attributable to falls from buildings or other structures in the state of Hawai‘i among children aged 10 years or younger. Fourteen of these patients were non-residents of Hawai‘i, and were excluded from injury rate calculations. There were a total of 264 boys and 152 girls, with a nearly 2:1 male to female ratio. Over half of the patients (217, or 52%) were 1–3 years of age, with the number of patients decreasing with increasing age. Most (328, or 79%) of the injuries were treated in the ED setting (Figure 1). There was no statistically significant trend in falls from one year to the next (P > .05 by Poisson regression). From 2005 through 2011, the average annual injury rate was 39 per 100,000 children age 10 or younger statewide (Figure 1), and 34 per 100,000 children in Honolulu County alone (Table 1), with no significant trend from one year to the next. While there was no statistically significant difference in the average annual injury rate between the counties of Honolulu, Kaua‘i, and Maui, the rate of injury in Hawai‘i County was twice that of Honolulu (70 vs 34 per 100,000, P < .001) and Maui (70 vs 30 per 100,000, P<.001), and three fold greater than the injury rate in Kaua‘i (70 vs 23 per 100,000, P < .001) (Table 1). In each year from 2005 to 2011, with the exception of 2009, the injury rate in Hawai‘i County was higher than that of the other counties (Figure 2).
Figure 1.
Nonfatal injuries from unintentional falls in the entire state of Hawaii among children ≤ 10 years of age, 2005–2011. Incidence is reported per 100,000 children 10 years of age or younger. Incidence data was taken from the Hawai‘i Health Information Corporation (HHIC).
Table 1.
Incidence rates of injuries due to unintentional falls from buildings among children ≤ 10 years of age, by county, 2005–2011. Incidence data was taken from the HHIC.
| County | Population* | Total Injuries† | Annual Average Injuries‡ | Rate per 100,000§ |
| Honolulu | 106,874 | 252 | 36 | 34 |
| Hawai‘i | 20,614 | 101 | 14 | 70¶ |
| Kaua‘i | 7,605 | 12 | 2 | 23 |
| Maui | 17,753 | 37 | 5 | 30 |
Total number of injuries over the 7-year study period.
Average annual number of injuries.
Annual rate per 100,000 children in the studied age group.
Hawai‘i County had a significantly higher annual injury rate compared to Honolulu, Kaua‘i, and Maui counties (P < .001 by Poisson regression).
Figure 2.
Incidence of nonfatal injuries due to unintentional falls from buildings among children ≤ 10 years of age, by county, 2005–2011. There were no falls in Kauai‘ in 2009. Incidence data was taken from the HHIC.
A total of 77 relevant hospitalizations were identified separately from the HTR. More than half (58%, or 45) of the injuries resulted from falls out of windows, 16% (12) were from balconies, 8% (6) from roofs, and specific information was not provided for the remaining 14 falls (18%) (Table 2). About half (40/77, or 52%) of the patients were under age 3 (data not shown). Based on the HHIC data, the average annual incidence rate for children age 10 and under requiring admission to the hospital following a fall from a building in the state of Hawai‘i was 8.6 per 100,000 children. For children under age 3, the incidence rate was 10.8 per 100,000 children.
Table 2.
Nonfatal falls from buildings by site among inpatients age 10 and younger, 2005–2011. Incidence data was taken from the Trauma Registry (TR).
| Site | No. (%) |
| Total falls | 77 (100%) |
| Falls from windows | 45 (58%) |
| Falls from non-window sites | 18 (24%) |
| Balcony | 12 (16%) |
| Roof | 6 (8%) |
| Not documented | 14 (18%) |
While the demographics and injury circumstances of Hawai‘i pediatric falls from buildings are similar to that in other US cities, the annual injury rate was significantly higher than that for similarly-aged children in at least one other urban area. Istre, et al, reported in 2003 that the annual injury rate due to falls from buildings in Dallas County, TX, was 15.4 per 100,000 for children aged 0–4, and 1.5 per 100,000 for children aged 5–99 (weighted average = 8.7 per 100,000) (Table 3). In contrast, the average annual injury rate in Honolulu County for the 0–10 year olds was 34, over a three-fold difference (P < .001).
Table 3.
Calculated averages of published rates, combining two or more age groups.
| Author, Year | Age | Published Rate (per 100,000) | Population* | Location | Weighted Average | Rate in Hawai‘i State |
| Injury Rate from Building Falls | ||||||
| Istre, et al9, 2003 | 0–4 | 15.6 | 181,951 | Dallas County, TX | 8.7† | 30.7† |
| 5–9 | 1.5 | 175,763 | ||||
| Hospitalization Rate from Building Falls | ||||||
| Pressley, et al7, 2005 | 0–4 | 4.6 | 19,176,154 | United States | 3.2† | 6.9† |
| 5–9 | 1.98 | 20,549,855 | ||||
| Agran, et al12, 2003 | 0 | 1 | 483,140 | California | 10.7‡ | 10.8‡ |
| 1 | 14 | 486,582 | ||||
| 2 | 17 | 489,335 | ||||
Population of each age group obtained from 2000 US Census.
Calculated average annual rate for children ages 0–9.
Calculated average annual hospitalization rate for children ages 0–2.
The annual rate of hospital admissions following a fall from a building in Hawai‘i was over two times higher than the national average for the same age-group as reported by Pressley, et al, in 2005 (8.6 per 100,000 vs 3.2 per 100,000)7 (Table 3), and by extension, higher than the incidence in New York City which was reported by the authors to be half of the national rate among children 0–17.7 Among children under 3 years of age hospitalized after falling from a building, Hawai‘i rates were comparable to those reported in California in 2003 (10.8 per 100,000 vs 10.7 per 100,000)12 (Table 3).
From 1991 to 2011, there were nine pediatric fatalities resulting from falls from buildings in Hawai‘i. All victims were under 5 years of age. The annual number of fatalities varied from none to four, with only one death occurring after 2006.
Discussion
The demographics of pediatric fall victims in Hawai‘i are similar to those previously reported in the literature. In concordance with published data, the male-to-female ratio approximated 2:1.3,7,8 Ninety one percent of patients admitted as a result of falling from a window were younger than 5 years of age; the percentages reported in other geographic areas in the literature vary from 60 to 90%.1,4,9–11 Also, the number of patients treated for fall-related injuries decreased with age, similar to previous reports.1,3,7 The majority of falls from buildings occurred from windows (58%), similar to prior studies reporting a range of 36–62%.3,8,9
Environmental factors which are unique to Hawai‘i may contribute to the higher rate of unintentional falls from windows and balconies. Several reports from other major cities have documented a higher incidence of injuries among children during the warm summer months, likely reflecting increased falls during seasons when the windows are left open for temperature regulation.3,4,7,8,11 In cities with less temperate climates, residents may be more likely to leave windows closed, instead using central heating or air conditioning for temperature control. In contrast, the warm weather year round in Hawai‘i favors leaving windows open for tradewinds and energy conservation. The relative abundance of high-rise residential structures with lanais/balconies in Hawai‘i may also be a contributing factor. The comparable incidence of building falls among children under 3 years of age in Hawai‘i and California may reflect similarly moderate climates and a tendency to use natural ventilation instead of air conditioning for climate control.
In other US cities, legislation mandating the use of window safety devices has significantly reduced the morbidity and mortality of window falls among children. In 1976, the New York City Board of Health passed a law requiring owners of multiple dwellings to provide window guards in apartments housing children 10 years and younger.5 This resulted in an average decrease in window fall incidents by 50%, and a 35% reduction in deaths attributable to falls from windows.13 A Boston voluntary ordinance which encouraged landlords to install window guards resulted in an 83% decrease in window falls in the two years following its initiation.5 In 2009, the Minnesota Senate passed a window fall prevention law known as Laela's Law which requires the State Building Code to be updated to mandate fall-prevention devices on windows that are used in residences.
The 2006 International Building Code (IBC), which has been adopted by the counties of Honolulu, Hawai‘i, Maui, and Kaua‘i, requires window safety devices on floors more than 5 feet above the ground with windows less than 42” from the floor, but also states that guards are not required in windows with installed “insect” screens.14 Historically, in the majority of window fall incidents nationwide (55–82.8%),3,4,9,11 a window screen is present prior to the fall, but splits or falls as a result of the child's weight. Although the IBC specifies that rail spacing cannot be greater than 4” (which should protect most children over age 1) 1,15 with the overall railing height at least 42”, an amendment to the code exempts older buildings — some with guardrails spaced 6” apart or greater.16 It has been shown that most children under the age of 6 years can fit through a 6-inch opening, and few over 1 year can fit through a 4-inch opening.1,15,16 Without detailed knowledge of these codes, exemptions, and the age of their building, caregivers of children living in older buildings may be unaware of the risk of fall-related injuries in their particular dwelling. Additionally, current building codes do not regulate window height, window depth or placement directly over concrete.
The finding of significant county-level differences in average annual injury rates within the state suggests that the problem is not limited to high-rise tenements or urban areas. The incidence in Hawai‘i County, a predominantly rural community with few high-rise residences, was twice that of Honolulu and Maui, and three times that of Kaua‘i. However, county differences were based almost entirely on ED visits, and the number of patients who required hospitalization were too few to calculate county-specific differences. Additionally, due to socioeconomic differences between counties, the higher incidence of ED visits in Hawai‘i County may reflect a tendency for patients in this rural area to access the ED for more routine care. As noted earlier, all four Hawai‘i counties in this study have adopted similar building code legislation based on the 2006 IBC.
Given the morbidity and mortality associated with these injuries, and a history of very successful prevention strategies in other communities, experts have advocated for public awareness, caregiver education and environmental modifications as a method of reducing the incidence of fall injuries in several publications.1,4,5,9,11 The American Academy of Pediatrics made several recommendations in 2001 regarding pediatric falls from heights, to include constant adult supervision of children, discouraging children from playing near windows/roofs/balconies, installation of window safety devices to prevent window openings greater than 4 inches, placement of furniture away from windows and balconies, and planting grass/shrubbery at the base of windows to soften falls (Table 4).5 Window safety devices fall into two categories: stops and guards. Window stops prevent sliding windows from opening wider than 4 inches. Window guards consist of removable bars with 4″ spacing which fit over the window and allow the window to be fully open or closed. In some communities, fire protection professionals have expressed concern that fixed window bars could prevent egress in the event of a fire. However, a follow-up study performed in New York City 20 years after window guard legislation showed that rates of fire-related deaths did not increase over this time.17
Table 4.
Key educational points for parents and caregivers regarding fall prevention. Recommendations by the American Academy of Pediatrics, 2001.5
| KEY EDUCATIONAL POINTS |
|
Study Limitations
This study is a retrospective analysis of three prospectively collected databases, and as such, the data was not recorded specifically to fulfill the objectives outlined in this paper. Therefore, the study's data are subject to inconsistencies and deficiencies in recording which are difficult to quantify. The practice and accuracy of E-coding may also vary across facilities and may therefore have influenced the county rate comparisons. Patients who were not seriously injured may have either not presented for care at all or presented to an outpatient clinic instead of an emergency department. Additionally, Hawai‘i has a relatively high annual rate of tourism compared to other urban areas, which affects the accuracy of reported rates in two ways. First, the number of patients who were injured and did not seek treatment may be higher than in other areas, due to parents waiting until getting home to present the child for care. Also, the large transient population at any given time is not reflected in the numbers reported by the US Census Bureau, which only captures the number of residents in each area; therefore, the actual population may be much larger than reported.
Since the HTR only reported data from one facility prior to 2009, statistics for patients admitted between 2005 and 2008 is most likely an underestimation of the actual number of injuries. Additionally, bias may have been introduced if the eight trauma facilities included since 2009 were not representative of the entire Hawai‘i State population. In this analysis of injury circumstances (ie, percentage of patients who fell from a window versus a lanai or rooftop), we used a subset of data from the HTR, examining only records of patients who required admission. However, in this database, information on whether a patient was discharged from the ED after a Trauma Activation was only available from 2008 forward. Therefore, numbers from 2005 to 2007 obtained from the Trauma Registry may include patients who were seen in the ED via a Trauma Activation and then discharged.
A limitation in comparing injury rates in Hawai‘i to those of reported elsewhere is incidence rates in the literature are only available for 0–4 and 5–9 year age groups, but Hawai‘i's injury data also included children 10 years of age. The US Census provides a Hawai‘i State population estimates for each year of age, so in order to obtain the most accurate denominator for rate calculations, we extrapolated the population of 0–10 year old children in each county based on the percentage of 10 year olds in the entire state. Bias may have been introduced if the percentage of 10-year-old children varied significantly among counties.
Due to variations in reporting strategies and target demographics, the authors were limited in their ability to make incidence comparisons with injury rates of other major areas reported in the literature aside from Dallas, New York City, and California.4,7,12 Studies that included children older than 10 without analyzing age groups individually (eg, 0–4, 5–9, etc) were could not be directly compared to the data in this study, as the inclusion of older children lowers the average rate,7,11 precluding a valid comparison. Also, a restriction which prevented comparison of injury rates in other major cities was the unavailability of population data at the city/town level for the specific age groups examined in this study. Another limitation is the small sample size in the Hawai‘i data; hospitalization rates are based on very small numbers, particularly when stratified by age (eg, under 3 years).
Finally, the study assumed that all patients given an E code of “fall from a building or structure” sustained a non-intentional injury. It is certainly possible the some of the falls were due to violent crimes. It is also possible that some of the falls identified in the HHIC data were from structures other than buildings, eg, walls, bridges, etc. However, a review of the open text field describing the fall in the HTR suggests this is a relatively rare occurrence (approximately 5%), at least at the hospital admission level.
Conclusions
Pediatric falls from buildings are preventable occurrences which may result in serious injury and, occasionally, death. This study reveals a burden of injury from this mechanism which is significantly higher for the children of Hawai‘i than for the city of Dallas, and among patients hospitalized after building falls, higher than the published national average and for New York City. Within the state of Hawai‘i, the rate of nonfatal injury in Hawai‘i County was twice that of Honolulu and Maui Counties, and three times that of Kaua‘i County. Preventive measures should be considered, including increasing public awareness, caregiver education on window safety, and use of appropriate window safety devices and balcony guardrails. Consideration should also be given to legislative strategies that have shown success in other US cities and states.
Footnotes
This manuscript addresses a public health issue. The views expressed in this manuscript are those of the authors and do not reflect the official policy or position of the Department of the Army, Department of Defense, or the US Government.
Conflict of Interest
The authors have no disclosures nor conflicts of interest to report.
References
- 1.Judy K. Unintentional injuries in pediatrics. Pediatr Rev. 2011;32(10):431–438. doi: 10.1542/pir.32-10-431. [DOI] [PubMed] [Google Scholar]
- 2.United States Consumer Product Safety Commission, author. CPSC: Parents, caregivers should consider safety before opening windows. Washington, DC: Office of Information and Public Affairs; 2011. Apr 8, [Google Scholar]
- 3.Lehman D, Schonfeld N. Falls from heights: a problem not just in the northeast. Pediatrics. 1993;92(1):121–124. [PubMed] [Google Scholar]
- 4.Vish NL, Powell EC, Wiltsek D, Sheehan KM. Pediatric window falls: not just a problem for children in high rises. Inj Prev. 2005;11(5):300–303. doi: 10.1136/ip.2005.008664. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Committee on Injury and Poison Prevention, author. American Academy of Pediatrics: Falls from heights: windows, roofs, and balconies. Pediatrics. 2001;107(5):1188–1191. doi: 10.1542/peds.107.5.1188. [DOI] [PubMed] [Google Scholar]
- 6.U.S. Census Bureau, author. Intercensal Estimates of the Resident Population by Five-Year Age Groups, Sex, Race, and Hispanic Origin for the States and United States: April 1, 2000 to July 1, 2010. < http://www.census.gov/popest/data/intercensal/state/files/ST-EST00INT-ALLDATA.csv> and Intercensal Estimates of the Resident Population by Single Year of Age, Sex, Race, and Hispanic Origin for the United States: April 1, 2000 to July 1, 2010; < http://www.census.gov/popest/data/intercensal/national/files/US-EST00INT-ALLDATA.csv> (3 June 2013)
- 7.Pressley JC, Barlow B. Child and adolescent injury as a result of falls from buildings and structures. Inj Prev. 2005;11(5):267–273. doi: 10.1136/ip.2004.007724. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Musemeche CA, Barthel M, Cosentino C, Reynolds M. Pediatric falls from heights. J Trauma. 1991;31(10):1347–1349. doi: 10.1097/00005373-199110000-00004. [DOI] [PubMed] [Google Scholar]
- 9.Istre GR, McCoy MA, Stowe M, Davies K, Zane D, Anderson RJ, Wiebe R. Childhood injuries due to falls from apartment balconies and windows. Inj Prev. 2003;9(4):349–352. doi: 10.1136/ip.9.4.349. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Barlow B, Niemirska M, Gandhi RP, Leblanc W. Ten years of experience with falls from a height in children. J Pediatr Surg. 1983;18(4):509–511. doi: 10.1016/s0022-3468(83)80210-3. [DOI] [PubMed] [Google Scholar]
- 11.Harris VA, Rochette LM, Smith GA. Pediatric injuries attributable to falls from windows in the United States in 1990–2008. Pediatrics. 2011;128(3):455–462. doi: 10.1542/peds.2010-2687. [DOI] [PubMed] [Google Scholar]
- 12.Agran PF, Anderson C, Winn D, Trent R, Walton-Haynes L, Thayer S. Rates of pediatric injuries by 3-month intervals for children 0 to 3 years of age. Pediatrics. 2003;111:e683. doi: 10.1542/peds.111.6.e683. [DOI] [PubMed] [Google Scholar]
- 13.Spiegel CN, Lindaman FC. Children can't fly: a program to prevent childhood morbidity and mortality from window falls. Am J Public Health. 1977;67(12):1143–1147. doi: 10.2105/ajph.67.12.1143. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.City and County of Honolulu, author. Revised Ordinances of Honolulu 1990: Adoption of the International Building Code. Chapter 16, Section 1.1, Amendment 144. [1 June 2013]. Available at http://www1.honolulu.gov/council/ocs/roh/rohchapter16a1vb.pdf. Updated 20 June 2007.
- 15.Stephenson EO. The silent and inviting trap. Building Off Code Administrator. 1988:28–33. [Google Scholar]
- 16.City and County of Honolulu, author. Revised Ordinances of Honolulu 1990: Adoption of the International Building Code. Chapter 16, Section 1.1, Amendment 5. [1 June 2013]. Available at http://www1.honolulu.gov/council/ocs/roh/rohchapter16a1vb.pdf. Updated 20 June 2007.
- 17.Bijur PE, Spiegel C. Window fall prevention and fire safety: 20 years of experience in New York City. Pediatric Research. 1996;39:102. [Google Scholar]