ethnic and socioeconomic disparities in the use of helmets in children involved in bicycle accidents

    Racial/ethnic and socioeconomic disparities in the use of helmets in children involved in bicycle accidents Veronica F. Sullins, Arezou Yaghoubian, Julien Nguyen, Amy H. Kaji, Steven L. Lee PII: DOI: Reference:

S0022-3468(14)00047-5 doi: 10.1016/j.jpedsurg.2014.01.038 YJPSU 56669

To appear in:

Journal of Pediatric Surgery

Received date: Accepted date:

25 January 2014 27 January 2014

Please cite this article as: Sullins Veronica F., Yaghoubian Arezou, Nguyen Julien, Kaji Amy H., Lee Steven L., Racial/ethnic and socioeconomic disparities in the use of helmets in children involved in bicycle accidents, Journal of Pediatric Surgery (2014), doi: 10.1016/j.jpedsurg.2014.01.038

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT

T

Racial/Ethnic and Socioeconomic Disparities in the Use of Helmets in Children Involved in Bicycle Accidents

Department of Surgery, Harbor-UCLA Medical Center, Los Angeles, CA 90509, USA

b

NU

a

SC

RI P

Veronica F Sullinsa, Arezou Yaghoubianb, Julien Nguyenc, Amy H Kajid, Steven L Leea*

Department of Surgery, Division of Plastic Surgery, University of California, Los Angeles, Los

University of California, Los Angeles, Los Angeles, CA 90095, USA

d

ED

c

MA

Angeles, CA 90095, USA

Department of Emergency Medicine, Harbor-UCLA Medical Center, Los Angeles, CA 90509,

CE

*

PT

USA

AC

Corresponding Author:

Steven L. Lee Division of Pediatric Surgery Harbor-UCLA Medical Center 1000 West Carson Street Box 25, Bldg 1-E Torrance, CA 90509, USA Telephone: +1 310 222 2706 Fax: +1 310 782 1562 E-mail: [email protected]

ACCEPTED MANUSCRIPT Abstract Purpose: While bicycle helmet use reduces bicycle-related head injury, few children wear them

RI P

T

regularly. We aimed to describe racial/ethnic and socioeconomic differences in pediatric helmet use in Los Angeles County (LAC) to help target groups for injury prevention programs.

SC

Methods: A retrospective review of all pediatric patients involved in bicycle-related accidents in

NU

LAC between 2006 and 2011 was performed. Our primary analysis examined the association between helmet use and age, gender, insurance status, and race/ethnicity. We also evaluated the

MA

association between helmet use and the need for emergency surgery, mortality, and length of hospital stay (LOH), after adjusting for injury severity score (ISS), age, insurance status, and

ED

race/ethnicity.

PT

Results: Of 1,248 patients, 11.3% wore helmets with decreased use among children 12 years and older, minorities, and those without private insurance. Overall, 5.9% required an emergency

CE

operation, 34.1% returned to their pre-injury capacity, and mortality was 0.7%. On multivariable

AC

analysis, higher ISS increased LOH, the risk for emergency surgery, and mortality. Conclusion: Nearly 90% of children involved in bicycle-related accidents were not wearing helmets. Helmet use was lower among older children, minorities, and those from a low socioeconomic status. Injury prevention programs targeting low-income middle and high schools and minority communities may help increase helmet use in children in LAC.

Keywords: Bicycle Helmet; Bicycle Accidents; Helmet Use; Racial Disparities; Ethnic Disparities

ACCEPTED MANUSCRIPT According to the CDC, an estimated 33 million children ride bicycles nearly 10 billion hours each year. Unfortunately, nearly 200 children die each year from bicycle crashes, and bicycle-

T

related injuries account for over 250,000 emergency department related visits [1]. While helmet

RI P

use is a proven intervention that reduces bicycle-related head injury by 63-88% [2] only 18% of

SC

the population and 15% of children under 15 years of age wear helmets all or most of the time [3]. Although California has had a mandatory bicycle helmet law for those less than 18 years old

NU

since 1994, the state had the second highest number of bicyclist fatalities in 2011 [4]. The problem of ethnic and racial disparities has recently received significant attention in both the

MA

media and the literature; yet there are little data regarding racial/ethnic and socioeconomic differences in the use of helmets among injured patients. Thus, the aim of this study was to

ED

describe racial/ethnic and socioeconomic differences in pediatric helmet use in Los Angeles

PT

County to delineate target groups for developing and disseminating injury control

CE

recommendations.

AC

1. Materials and Methods

This study was approved by the Human Subjects Committee of the Los Angeles Biomedical Research Institute at Harbor-UCLA Medical Center. We evaluated all bicycle-related trauma patients <18 years of age in Los Angeles County between 2006 and 2011. Data on each patient were extracted from the Trauma and Emergency Medicine Information System (TEMIS) in Los Angeles County, a prospectively collected database. Patient data was transferred from the TEMIS registry into an Excel database (Microsoft Excel, Microsoft Corporation, Redmond, WA) and translated into a native SAS format using DBMS/Copy® (Dataflux Corporation, Cary, NC).

ACCEPTED MANUSCRIPT The main outcome measure was helmet use and its association with age, gender, insurance status, and race/ethnicity. Since continuous variables were not normally distributed, medians are

T

reported and the non-parametric Wilcoxon rank sum test was used for comparisons between sub-

RI P

groups. In addition to analyzing age as a continuous variable, it was also stratified as greater than or equal to 12 years. Categorical or nominal variables were compared using the chi square test

SC

and odds ratios with associated 95% confidence intervals (CI) where appropriate. An exploratory

NU

analysis using multivariable logistic and linear regression, was also performed to assess whether helmet use was associated with the need for emergency surgery, mortality, and length of hospital

MA

stay (LOH), after adjusting for injury severity score (ISS), age, insurance status, and race/ethnicity. Due to the small numbers with the outcome of interest and the exploratory nature

ED

of the analysis, no interactions were assessed. Model-fit was assessed using the Hosmer-

PT

Lemeshow fit statistic.

CE

2. Results

AC

After our database search, 1,248 patients were identified for study inclusion. Demographic data are listed in Table 1. The median age of patients was 13 years (IQR 5-15) with 464 (37.2%) Table 1

younger than 12, and 1081 (85.1%) of the study cohort were male. A total of 141 patients (11.3%) wore helmets. In the study group overall, 55.5% were Hispanic, 20.0% Caucasian, 14.6% African American, 3.5% Asian, and 6.5% were identified as other, and 49.1% had private Table 2

insurance. Outcome data are listed in Table 2. The median LOH was 2 days (IQR 1-3) while the median ISS was 5 (IQR 2-10). Overall, 9 (0.7%) children died while 74 (5.9%) required emergency surgery. Six (0.5%) patients suffered permanent disability, while 801 (65.4%) reported temporary disability and 418 (34.1%) returned to their pre-injury capacity.

ACCEPTED MANUSCRIPT After stratifying by age, 464 (37.2%) children were < 12 and 784 (62.8%) children were ≥ 12 years old. Children younger than 12 years old were more likely to wear helmets than those 12

T

years and older (OR 1.47, 95% CI 1.03-2.09, p = 0.03). There were fewer males in the younger

RI P

group (OR 0.63, 95% CI 0.46-0.87, p < 0.005). No differences were identified in insurance status, mortality, need for emergency surgery, and return to pre-injury capacity when comparing

SC

younger to older children.

NU

The univariate analysis examining helmet use versus no helmet use is presented in Table 3.

MA

Among the 9 patients who died, 8 did not wear helmets. All unhelmeted patients who died had a GCS of 3 and the median ISS for this group was 36.5. The one helmeted patient who died had a GCS of 3 and ISS of 26. There were no differences in gender, age, ISS, LOH, mortality, the

ED

Table 3

need for emergency surgery, and return to pre-injury capacity. Table 4 compares differences in

PT

helmet use based on age, race/ethnicity, and insurance status. There was a significant

CE

difference across racial/ethnic groups. Among Caucasians, 35.2% wore helmets, compared to 4.2% among Hispanics, 6.0% among African Americans, 7.0% among Asians, and 12.4% among other races/ethnicities (p < 0.0001). In addition, those with private insurance were more likely to

AC

Table 4

wear helmets (OR 2.19, 95% CI 1.51-3.16, p < 0.0001). On multivariable logistic regression analysis, after controlling for age, race/ethnicity, ISS, and insurance status, helmet use was not associated with need for emergency surgery (OR 1.02, 95% CI 0.47-2.21) or mortality (OR 2.68, 95% CI 0.26-28.17). Multivariable linear regression analysis adjusting for the same previously listed predictors did not demonstrate any additional increase in LOH (0.18 days, 95% CI 0.64-1.0, p = 0.67). Only ISS was associated with need for emergency surgery (OR 1.07, 95% CI 1.05-1.10), mortality (OR 1.24, 95% CI 1.15-1.33), and an additional increase in LOH (0.4 days, 95% CI 0.36-0.44, p < 0.0001). There was an association

ACCEPTED MANUSCRIPT between having private insurance and shorter LOH (0.57 fewer days, 95% CI 0.31-0.78, p = 0.03) but this association did not persist for the remaining outcomes.

RI P

T

3. Discussion

In the US, unintentional injury is the leading cause of death among children and adolescents,

SC

accounting for nearly 50% of deaths among 5- to 14-year olds [5]. Each year, nearly 325,000

NU

children visit the emergency department for bicycle-related injuries and there are over 26,000 associated traumatic brain injuries [6]. In 2011, the National Highway and Traffic Safety

MA

Administration reported over 17,000 bicycle-related injuries and 116 fatalities among those <20 years old [4]. Both the financial and non-monetary cost to society cannot be understated.

ED

In contrast to multiple studies across North America [7–9] there has not been a sustainable

PT

increase in helmet use in Los Angeles County despite California legislation mandating helmet use in those under the age of 18 [10].

CE

While helmet use in children across the United States is estimated to be 48% [11], use among

AC

injured children is much lower [10,12,13]. Perhaps the most striking finding in our study was the extremely low percentage of all injured children wearing helmets in Los Angeles County. Although the cause is likely multi-factorial, it is probably most affected by the large percentage of patients in Los Angeles County belonging to subgroups with lower helmet use, such as minority and low-income patients. Our finding that helmet use in older children was significantly lower is consistent with the findings of many others [11,14,15]. Not surprisingly, this specific age group has the highest rate of unintentional injury [5]. Of the 9 patients who died, 8 were not wearing helmets, but the difference did not reach statistical significance likely because there were very few patients in this subgroup. Although the TEMIS database does not capture the

ACCEPTED MANUSCRIPT specific cause of death, the high ISS and low GCS in unhelmeted patients who died supports

T

traumatic brain injury as a related, but not the only factor.

RI P

The lack of an association between helmet use and mortality has been documented [16], however it is well established that helmet use is associated with lower rates of severe head injury, and

SC

therefore mortality, among helmet users [1,3,4]. It was surprising to find that helmet use was unrelated to mortality in our study, even after multivariable analysis. We attribute this result to

NU

small sample size. In addition, our study population is limited to patients presenting to trauma

MA

centers and does not include those who die prior to reaching the hospital. This also may contribute to the lack of association between helmet use and all other outcomes. As expected for

ED

patients that have more severe injuries, a higher ISS was associated with a greater risk of emergency surgery, mortality, and LOH. Lastly, having private insurance was associated with a

PT

shorter LOH. Although there are many factors contributing to length of hospital stay, this may

CE

reflect the additional inpatient care of uninsured children who cannot be placed into subacute care or rehabilitation facilities. A similar finding of longer hospital stays in orthopedic pediatric

AC

patients with public insurance was reported by Gao and colleagues [17], who theorized that private payers may enhance efficiency of care. Overall, the evidence is mixed regarding insurance status and outcomes, reflecting the complexity of this issue. Racial/ethnic disparities exist in income levels, proportions of people living below the poverty level, education, home ownership, health care access and delivery, and mortality [5,18,19]. Therefore it is not surprising that a recent nationwide survey of helmet use in children found significant differences in age, ethnicity, household income, and education [11]. Specifically, older children, those with Hispanic ethnicity, lower household income, and lower education level were less likely to wear helmets. These discrepancies persist in children treated for bicycle-

ACCEPTED MANUSCRIPT related trauma. We also found a dramatic difference in helmet use across all racial and ethnic backgrounds. While over 35% of Caucasian children wore helmets, less than 10% of Hispanic,

T

black, and Asian children used them. In addition, those with private insurance were twice as

RI P

likely to wear helmets. Our findings are also consistent with Schuster, et al, who demonstrated that among 5th grade children, racial/ethnic disparities exist in many health-related behaviors

SC

including bike helmet use, even after adjusting for sociodemographic characteristics [20]. In fact,

NU

the most significant predictors of disparities in health-related behaviors were socioeconomic status (household income and highest education level) and the location of the child’s school.

MA

Interestingly, a UK study reported that in the absence of legislation, attitudinal variables more strongly influence bike helmet use than socioeconomic factors, although both had significant

ED

effects [15]. It is difficult to ascertain how results of the current study would have changed in the

PT

absence of legislation. Nevertheless, our findings further support the implementation of

schools.

CE

education programs focused in minority and low-income communities and their associated

AC

Injury prevention is a challenging problem to address and reasons for non-use of bicycle helmets are varied. Children in particular are subject to a wide range of influences. Children and adolescents have been found to be more likely to wear helmets when accompanied by a parent or companion wearing a helmet [21-23]. Various strategies including school or community-cased educational programs, free or subsidized helmet distributions, and legislative action and enforcement of active laws have been studied [24-26]. While outcomes have been mixed, one study found that of the non-legislative programs, community-based and free helmet distribution programs were most effective in increasing observed helmet use while school-based programs were effective, but slightly less [27]. Legislative action has been taken in parts of Canada and the

ACCEPTED MANUSCRIPT US. A Toronto-based observational study found that increases in bike helmet use were greatest among middle- and low-income areas after legislation, though over time helmet use returned to

T

pre-legislative levels among the same socioeconomic groups [28,29]. A multi-faceted, long term

RI P

strategy will be necessary to increase bicycle helmet use among children. Our study supports specifically targeting low income and minority communities and middle and high school children

SC

for implementation of bicycle helmet education and injury prevention programs.

NU

Findings of our study may not be applicable to other geographic areas. Nearly 50% of the Los

MA

Angeles County population is Hispanic, compared to 17% nationwide. In addition, in California, a greater percentage of the population lives below the poverty level. Multiple studies have found

ED

that helmet use in the Hispanic population is significantly lower than in other racial/ethnic groups [10,11,20,30]. Our study underscores the importance of targeting these specific

PT

communities and racial/ethnic groups for injury prevention programs. In addition, as a large

CE

metropolitan area with significant diversity, our results actually highlight groups that are not well

AC

represented in other parts of the country. Our study is also potentially limited by spectrum bias, since the study population includes only pediatric patients presenting to trauma centers. Excluded are patients who die on-scene or en route, children seen in emergency departments who do not meet specific trauma criteria, those seen in an outpatient setting, and those who did not seek any medical attention. We elected to focus on the pediatric trauma population because they account for most of the morbidity and health care costs associated with bicycle-related injuries. Another potential limitation of our study was the use of insurance status as a surrogate measure of patients’ socioeconomic status. Although the validity of this approach is not known, currently

ACCEPTED MANUSCRIPT Medicaid covers nearly 50 million low-income individuals [5]. Therefore, it is well accepted in the literature that insurance status serves as a proxy for socioeconomic status [31,32].

RI P

T

Nearly 90% of children treated for bicycle-related injuries in Los Angeles County did not wear helmets. In particular, those 12 years and older, minorities and patients from a lower

SC

socioeconomic class were less likely to wear helmets. These disparities may help target minority and low-income communities as well as middle and high schools for multi-tiered injury

AC

CE

PT

ED

MA

NU

prevention programs.

ACCEPTED MANUSCRIPT

References

Centers for Disease Control and Prevention, National Center for Injury Prevention and

T

1.

RI P

Control. Web-based Injury Statistics Query and Reporting System (WISQARS):

2.

SC

http://www.cdc.gov/injury/wisqars/index.html. Accessed August 20, 2013. Thompson D, Rivara F, Thompson R. Helmets for preventing head and facial injuries in

3.

NU

bicyclists (Review ). The Cochane Database of Systematic Review. 2009;CD001855. US Consumer Products and Safety Commission Bicycle Helmet Safety Institute:

4.

MA

http://www.helmets.org/stats.htm. Accessed August 22, 2013. National Highway Traffic Safety Administration. Traffic Safety Facts, 2011 data -

ED

bicyclists and other cyclists. Washington, DC: US Department of Transportation; 2013.

5.

PT

DOT HS 811 743.

National Center for Health Statistics. Health, United States, 2012: With Special Feature on

Nonfatal traumatic brain injuries related to sports and recreation activities among persons

AC

6.

CE

Emergency Care. Hyattsville, MD. 2013.

aged ≤19 years--United States, 2001-2009. MMWR 2011;60(39):1337–42. 7.

Shafi S, Gilbert J, Loghmanee F, et al. Impact of bicycle helmet safety legislation on children admitted to a regional pediatric trauma center. J Pediatr Surg 1998;33(2):317–21.

8.

Macpherson A, Spinks A. Bicycle helmet legislation for the uptake of helmet use and prevention of head injuries ( Review ). Cochane Database Syst Rev 2008;(3):CD005401.

9.

Karkhaneh M, Rowe BH, Saunders LD, et al. Trends in head injuries associated with mandatory bicycle helmet legislation targeting children and adolescents. Accid Anal Prev 2013;59:206–12.

ACCEPTED MANUSCRIPT 10.

Castle SL, Burke R V, Arbogast H, et al. Bicycle helmet legislation and injury patterns in trauma patients under age 18. J Surg Res 2012;173(2):327–31. Dellinger AM, Kresnow M. Bicycle helmet use among children in the United States: the

T

11.

Thompson DC, Rivara FP, Thompson RS. Effectiveness of bicycle safety helmets in

SC

12.

RI P

effects of legislation, personal and household factors. J Safety Res 2010;41(4):375–80.

preventing head injuries. JAMA 1996;276(24):1968–73.

Persaud N, Coleman E, Zwolakowski D, et al. Nonuse of bicycle helmets and risk of fatal

NU

13.

head injury: a proportional mortality, case–control study. CMAJ 2012;184(17):921–3. Davison CM, Torunian M, Walsh P, et al. Bicycle helmet use and bicycling-related injury

MA

14.

among young Canadians: an equity analysis. Int J Equity Health 2013;12(48):1–8. Lang IA. Demographic, socioeconomic, and attitudinal associations with children’s cycle-

ED

15.

16.

PT

helmet use in the absence of legislation. Inj Prev 2007;13(5):355–8. McDermott FT, Lane J, Brazenore G, et al. The effectiveness of bicyclist helmets:

Gao Y, Johnston RC, Karam M. Pediatric sports-related lower extremity fractures:

AC

17.

CE

a study of 1710 casualties. J Trauma 1993;34(6):834–45.

hospital length of stay and charges: what is the role of the primary payer? Iowa Orthop J 2010;30:115–8. 18.

DeNavas-Walt C, Proctor BD, Smith JC. Income , Poverty , and Health Insurance Coverage in the United States: 2011 Current Population Reports. U.S. Census Bureau, Current Population Reports 2012;Sept:60–243.

19.

Shapiro T, Meschede T, Osoro S. The roots of the widening racial wealth gap: explaining the black-white economic divide. IASP 2013;(February):1-8.

20.

Schuster MA, Elliott MN, Kanouse DE, et al. Racial and ethnic health disparities among

ACCEPTED MANUSCRIPT fifth-graders in three cities. New Engl J Med 2012;367(8):735–45. 21.

Rivara FP, Thompson DC, Thompson RS, et al. The Seattle Children’s Bicycle Helmet

Finnoff JT, Laskowski ER, Altman KL, et al. Barriers to Bicycle Helmet Use. Pediatrics

RI P

22.

T

Campaign : changes in helmet use and head injury admissions. Pediatrics 1994;93:567–9.

23.

SC

2001;108(1):e4–e4.

Khambalia A, MacArthur C, Parkin PC. Peer and adult companion helmet use is

24.

NU

associated with bicycle helmet use by children. Pediatrics 2005;116(4):939–42. Parkin PC, Hu X, Spence LJ, et al. Evaluation of a subsidy program to increase bicycle

25.

MA

helmet use by children of low-income families. Pediatrics 1995;96:283–7. Watts D, O’Shea N, Ile A, et al. Effect of a bicycle safety program and free bicycle helmet

ED

distribution on the use of bicycle helmets by elementary school children. J Emerg Nurs

26.

PT

1997;23(5):417–9.

Gilchrist J, Schieber RA, Leadbetter S, et al. Police enforcement as part of a

Owen R, Kendrick D, Mulvaney C, et al. Non-legislative interventions for the promotion

AC

27.

CE

comprehensive bicycle helmet program. Pediatrics 2000;106(1):6–9.

of cycle helmet wearing by children ( Review ). Cochane Database Syst Rev 2011;(11):CD003985. 28.

Parkin PC, Khambalia A, Kmet L, et al. Influence of socioeconomic status on the effectiveness of bicycle helmet legislation for children: a prospective observational study. Pediatrics 2003;112(3):192–6.

29.

Macpherson AK, Macarthur C, To TM, et al. Economic disparity in bicycle helmet use by children six years after the introduction of legislation. Inj Prev 2006;12(4):231–5.

30.

Allen ML, Elliott MN, Morales LS, et al. Adolescent participation in preventive health

ACCEPTED MANUSCRIPT behaviors, physical activity, and nutrition: differences across immigrant generations for Asians and Latinos compared with Whites. Am J Public Health 2007;97(2):337–43. Ayanian JZ, Kohler BA, Abe T, et al. The relation between health insurance coverage and

T

31.

RI P

clinical outcomes among women with breast cancer. New Engl J Med 1993;329(5):326–

Marcin JP, Schembri MS, He J, et al. A population-based analysis of socioeconomic status

NU

and insurance status and their relationship with pediatric trauma hospitalization and

CE

PT

ED

MA

mortality rates. Am J Public Health. 2003;93(3):461–6.

AC

32.

SC

31.

ACCEPTED MANUSCRIPT Legend:

T

Table 1: Demographics (n = 1248)

RI P

Table 2: Patient outcomes

SC

Table 3: Univariate analysis of outcomes and demographics comparing helmeted and

NU

unhelmeted trauma patients under age 18 treated for bicycle-related injuries in Los Angeles

MA

County.

Table 4: Univariate analysis of helmeted patients treated for bicycle-related injuries comparing

AC

CE

PT

ED

age (stratified), race/ethnicity, and insurance status.

ACCEPTED MANUSCRIPT

Tables:

Male

1061 (85.0)

Age (years)

13 (10-15)

Helmet Use

141 (11.3)

250 (20.0)

Hispanic

692 (55.5)

Black

182 (14.6)

CE

PT

ED

White

43 (6.5)

AC

Asian Other

NU MA

Race/Ethnicity

Private Insurance

81 (6.5) 613 (49.1)

RI P

N (%) or Median (IQR)

SC

Variable

T

Table 1

ACCEPTED MANUSCRIPT

N (%) or Median (IQR)

ISS

5 (2-10)

LOH (days)

2 (1-3)

Mortality

9 (0.7)

Emergency surgery

74 (5.9)

Permanent disability

6 (0.5)

CE

Return to pre-injury capacity

AC

SC NU

MA

ED PT

Temporary disability

RI P

Variable

T

Table 2

801 (65.4) 418 (34.1)

ACCEPTED MANUSCRIPT

Table 3 Helmet (n = 141)

No Helmet (n = OR (95% CI) 1107)

Male

125 (88.7%)

936 (84.6%)

Age (years)

12 (IQR 10-14)

13 (IQR 10-15)

ISS

5 (IQR 3-9)

5 (IQR 2-10)

LOH (days)

2 (IQR1-4)

2 (IQR 1-3)

Mortality

1 (0.71%)

8 (0.72%)

Emergency surgery

8 (5.7%)

66 (6.0%)

Return to preinjury capacity

45 (31.9%)

RI P

SC

1.43 (0.83-2.46) 0.2

NU

MA

ED

373 (33.7%)

PT CE AC

P

T

Variable

0.08 0.11 0.67

0.98 (0.12-7.90) 0.99 0.95 (0.45-2.02) 0.89

0.92 (0.63-1.34) 0.67

ACCEPTED MANUSCRIPT

Table 4 Helmet (n = 141)

OR (95% CI)

< 12 years

64 (45.4%)

1.47 (1.03-2.09)

≥ 12 years

77 (9.8%)

P

T

Variable

SC

RI P

Age

Race/Ethnicity

29 (4.2%)

Black

11 (6.0%)

Asian

3 (7.0%)

Other

10 (12.4%)

NU

Hispanic

9.69 (6.63-14.15)

<0.0001

MA

88 (35.2%)

ED

White

0.03

PT

Insurance Status

93 (15.1%)

None/Public

48 (7.6%)

AC

CE

Private

2.19 (1.51-3.16)

<0.0001