Hospitalizations at Nonfederal Facilities for Lower Respiratory Tract Infection in American Indian and Alaska Native Children Younger than 5 Years of Age, 1997-2012

Hospitalizations at Nonfederal Facilities for Lower Respiratory Tract Infection in American Indian and Alaska Native Children Younger than 5 Years of Age, 1997-2012

Hospitalizations at Nonfederal Facilities for Lower Respiratory Tract Infection in American Indian and Alaska Native Children Younger than 5 Years of ...

479KB Sizes 0 Downloads 9 Views

Hospitalizations at Nonfederal Facilities for Lower Respiratory Tract Infection in American Indian and Alaska Native Children Younger than 5 Years of Age, 1997-2012 Bethany A. Weinert, MD, MPH, and M. Bruce Edmonson, MD, MPH Objectives To evaluate hospitalizations at nonfederal facilities for lower respiratory tract infection (LRTI) in American Indian/Alaska Native (AI/AN) children and to compare associated rates and risk factors in AI/AN children and white children. Study design We used Kids’ Inpatient Database samples from 1997-2012 to identify discharges in non-Hispanic AI/AN and white children ages <5 years with a principal or secondary diagnosis code indicating LRTI. To address systematic underreporting and misclassification of race in administrative databases, population rates were estimated by deriving race- and year-specific denominators from hospital births. Results During the study period, LRTI-associated discharge rates (per 1000) declined for white children (from 14.8 to 10.9; P < .001 for trend). For AI/AN children, rates varied widely by census region and were highest in the West, where they ranged from 38.6 in 1997 to 26.7 in 2012 (P = .35 for trend). Discharges in AI/AN children were associated with low household income, Medicaid insurance, and rural residence. In a case-cohort analysis of infants hospitalized with LRTI in 2012, discharge rates were higher for AI/AN infants than for white infants only in the West (72.8 vs 22.2; aOR, 2.5; 95% CI, 1.8-3.4). Conclusions Among young children who use nonfederal hospitals, LRTI-associated hospitalizations occur at substantially higher rates for AI/AN children than for white children. These hospitalizations occur at rates that are particularly high for AI/AN infants in the West, where rates are comparable with those reported for Indian Health Service enrollees. (J Pediatr 2016;-:---). See editorial, p 

A

merican Indian/Alaska Native (AI/AN) children enrolled in the Indian Health Service (IHS) suffer disproportionately higher morbidity and mortality from lower respiratory tract infections (LRTIs) than other children in the US.1-9 As recently as 2008, LRTI-associated hospitalizations occurred in AI/AN children enrolled in the IHS at rates comparable with those seen in developing countries.3,5 The rate for infants enrolled in the IHS was >3 times higher than in industrialized countries.3,5 Most information about LRTI-associated hospitalizations in AI/AN children is based on IHS data; there is little information available about rates in the broader population of AI/AN children in the US. Of the 5.2 million people (1.7% of the US population) who self-identified as AI/AN either alone or in combination with $1 other races in the 2010 US Census,10 only about 2 million were enrolled in the IHS.11 There are only 50 IHS and tribe-operated hospitals nationally, and most of these are located in western states12 where IHS enrollment is higher and LRTI-associated hospitalization rates in enrolled children are believed to be particularly high.5-7,11 As a result, a sizeable proportion of AI/AN children has not been represented in previous studies of racial disparities in LRTI-associated hospitalization rates. To address this knowledge gap, we used a large, national discharge database to describe LRTI-associated hospitalizations of AI/AN children at nonfederal (and non-IHS) hospitals. We sought to determine whether the rates of LRTI-associated hospitalization previously reported for children enrolled in the IHS are similarly elevated for AI/AN children who receive care outside the IHS. Our first objective was to examine national trends in associated hospitalization rates by race and region over the period 1997-2012. Our second objective was to compare information about LRTIassociated hospitalizations in AI/AN children and white children. Our third objective was to investigate relationships between socioeconomic factors and From the Department of Pediatrics, University of racial disparities on the risk of LRTI-associated hospitalization in children. Wisconsin School of Medicine and Public Health, Madison, WI

AI/AN HCUP IHS KID LRTI

American Indian/Alaska Native Healthcare Cost and Utilization Project Indian Health Service Kids’ Inpatient Database Lower respiratory tract infection

B.A.W. was supported by a National Research Service Award from the Health Resources and Services Administration (T32HP10010) to the University of Wisconsin Department of Family Medicine and Community Health. The authors declare no conflicts of interest. 0022-3476/$ - see front matter. Copyright ª 2016 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.jpeds.2016.03.019

1

THE JOURNAL OF PEDIATRICS



www.jpeds.com

Methods This retrospective study analyzed hospital discharge information from 1997, 2000, 2003, 2006, 2009, and 2012 in the Kids’ Inpatient Database (KID), a public use administrative dataset maintained by the Healthcare Cost and Utilization Project (HCUP) under the sponsorship of the Agency for Healthcare Quality and Research.13 For each data year, a KID sample was constructed from: (1) a 10% random sample of discharge records for uncomplicated births; and (2) an 80% random sample of discharge records for all other pediatric inpatient stays in each nonfederal, short-term general and specialty hospital in each participating state. From 1997 to 2012, the number of participating states increased from 22 states (accounting for 53.2% of the AI/AN population based on 2000 US Census data10) to 44 states (96.8%, based on 2010 census data10).13 Additional states with substantial AI/AN populations that were added to KID samples during the study period included Texas (2000), Oklahoma (2006), New Mexico (2009), and Alaska (2012).10,13 For 2012, the KID contained information on almost 3.2 million (unweighted) discharges of persons <21 years of age from 4179 US hospitals. The KID includes discharge-level information on patient demographics, admission source, payers, hospital-assigned diagnostic and procedural codes, length of stay, total hospital charges, and discharge disposition.13 Information on hospital location in the 2012 KID was limited to US Census region and could not be aligned with IHS-defined administrative regions because the sampling design precludes analysis at the state level.14 The states included in each US Census region are shown in the Figure. By accounting for its complex sampling design and incorporating sampling weights, each KID sample can be used to generate national estimates for

Volume stays at short-term nonfederal hospitals by children and adolescents in the US.15 IHS and tribally operated hospitals were not included in KID because of their federal status. These hospitals are not the only type of facility available to IHS enrollees, however. A proportion of hospital care for IHS enrollees is provided as contracted care at nonfederal hospitals when direct IHS services are unavailable.11,12 A contract health hospital is a nonfederal hospital that can accept IHS payment for services; at these hospitals, the IHS is considered “payer of last resort” and is accepted only if a form of health insurance (eg, Medicaid) is unavailable as the primary payer.16 In 2013, 24.8% of all hospital discharges funded by the IHS were associated with stays at nonfederal contract health hospitals.11 The KID includes hospitalizations that represent IHS contracted health care. As a result, there is some overlap between discharges captured in KID and in IHS databases.4,5,7 The International Classification of Diseases, 9th edition, Clinical Modification17 discharge diagnosis codes were used to define LRTI, respiratory failure, and asthma. According to previously published methods,4,5,7 we used a principal or secondary diagnosis of LRTI to define a LRTI-associated discharge (any one of 15 diagnoses for 1997-2006 and 25 diagnoses for 2009 and 2012). LRTI was defined by codes for pneumonia (480.x—486.x), influenza with pneumonia (487.0), bronchiolitis (466.0, 466.1), empyema (510.x), pulmonary mycobacteria (031.0), pulmonary tuberculosis (011.x), pulmonary anthrax (022.1), whooping cough (033.x), pulmonary syphilis (095.1), pleurisy with effusion (511.1), pulmonary abscess (513.x), and rheumatic pneumonia (517.1).4,5 Respiratory failure, a measure of LRTI illness severity, was defined by diagnosis codes 518.81, 518.82, 518.84, or 799.1.18 Asthma, a potential comorbid

Figure. aORs for LRTI-associated hospitalization in AI/AN vs white infants, by US Census Region, 2012. ORs were adjusted for sex, household income, rural-urban residence, and primary expected payer.

2

Weinert and Edmonson

- 2016 condition, was defined by code 493.x.17 Neither asthma nor respiratory failure were included in the definition of LRTI. For sensitivity analyses, we used previously published coding algorithms to separately define pneumonia and pneumococcal pneumonia.19,20 Data Analyses All analyses were conducted at the discharge level and were restricted to discharge records of children 0-4 years of age. We confined our analysis to records of children for whom information on race (as defined by HCUP) was nonmissing and indicated either non-Hispanic white or non-Hispanic AI/AN. Missing information on race in administrative databases is particularly important for small populations such as AI/AN children; we did not impute information about race because it is likely not missing at random and because trends in race classification are changing. From 1997 to 2012, the proportion of discharge records in KID for which race information was missing decreased from 19.1%21 to 8.2% and the proportion indicating AI/AN race increased from 0.4% to 1.4%. We also excluded records for which: (1) patient age was missing (<0.5% in the KID 2012 sample); or (2) discharge diagnosis and procedure codes indicated both a hospital birth and a LRTI during the same stay. To estimate LRTI-associated hospitalization rates by race, we used KID weighted samples from 1997, 2000, 2003, 2006, 2009, and 2012 to estimate numerators and hospital births to derive population denominators. Hospital births, rather than US census population estimates, were used for consistency of race classification, under the assumption that classification of race is similar for LRTI-associated discharges and birth discharges in data that hospitals and states report to HCUP for any given sample year. We used same-year hospital births as rate denominators because race classification and state participation changed over time. It was also assumed that hospital births were an appropriate basis for estimating the size of the underlying population of AI/AN children likely to be admitted to nonfederal hospitals. Trends over the study period were evaluated by generating rate estimates and proportions (and their SEs) from each KID sample and then evaluating linear trends with a 2-sided Cochran-Armitage test.22,23 To compare information about LRTI-associated hospitalizations across racial groups, we used weighted estimates from the 2012 KID sample. These stays were categorized by age, region, income quartile, rural-urban residence, certain diagnosis codes (pneumonia, bronchiolitis, asthma), and measures of illness severity (length of stay, respiratory failure, and hospital deaths). To examine the relative contributions of region and patient demographics to racial disparities on the risk of LRTIassociated hospitalization, we carried out an analysis that was restricted to infants. We used a case-cohort design24,25 in which cases were defined as LRTI-associated discharges and the cohort was represented by hospital birth discharges in the same year. Variables included in multivariate logistic regression models included: race, sex, household income, rural-urban residence, and primary expected payer. We car-

ORIGINAL ARTICLES ried out regression analysis on complete case data, after excluding discharges for which information was missing at a low (<0.5%) frequency (sex, hospital region, and ruralurban residence). Information was missing at higher rates for primary payer status (1.2%) and at particularly high rates for household income (10.5%) for AI/AN children, and we treated missing information for these variables categorically. Information on household income was missing, in part, because of HCUP confidentiality protections; household income is estimated by zip code, and related information is suppressed when the population within a specific zip code (eg, an AI/AN reservation) drops below a minimum threshold.26 All routine statistical analyses were conducted in SAS version 9.4 (SAS Institute, Cary, North Carolina) using procedures that account for complex survey design and incorporate post-stratification weights into national estimates.15 Specifically, race-specific hospitalization rates were estimated using PROC SURVEYMEANS (with Ratio and Domain statements) to account for variance associated with both numerators (LRTI-associated discharges) and denominators (hospital births) in subpopulations.27 Weighted estimates were suppressed when the point estimate was #10 discharges or the relative SE was >0.3, in accordance with HCUP guidelines. The University of Wisconsin-Madison Health Sciences Institutional Review Board determined that formal review of the study protocol was not required.

Results Analysis of national trends in rates of LRTI-associated discharges from nonfederal hospitals during 1997-2012 showed steady declines in rates for white children ages <5 years, but rates for AI/AN children fluctuated widely by year and US census region. National rates (per 1000) declined from 14.8 in 1997 to 10.9 in 2012 (P < .001 for linear trend) for white children and ranged from 19.7 to 22.5 (P = .45) for AI/AN children. Table I shows that national rates declined for both white infants (P < .001) and white children ages 1-4 years (P < .001) but not for AI/AN infants (P = .68) or AI/AN children 1-4 years of age (P = .014 for an increase in rates). In the West census region, where rates for AI/AN children generally were highest throughout the study period, we found a similar decline in rates for white children (P = .008) and rates that ranged from 38.6 in 1997 to 26.7 in 2012 for AI/AN children (P = .35; Table II). There was a noticeable decline in LRTIassociated hospitalization rates for both racial groups in 2012, yet racial disparities persisted. Hospitalization rates were particularly high in AI/AN infants living in the West, where rates in 2012 were more than 3 times higher than in white infants (Table II). When trend analysis was restricted to discharges coded specifically for pneumonia, there also was a statistically significant decline in rates (per 1000) for white children (from

Hospitalizations at Nonfederal Facilities for Lower Respiratory Tract Infection in American Indian and Alaska Native Children Younger than 5 Years of Age, 1997-2012

3

THE JOURNAL OF PEDIATRICS



Volume -

www.jpeds.com

Table I. LRTI-associated hospital discharge rates in children, by race and year, 1997-2012 AI/AN Year

Age (y)

2012

<1 1-4 <1 1-4 <1 1-4 <1 1-4 <1 1-4 <1 1-4

2009 2006 2003 2000 1997

White

Weighted frequency (SE)

Population estimate* (SE)

Rate per 1000 (95% CI)

Weighted frequency (SE)

Population estimate* (SE)

Rate per 1000 (95% CI)

1694 (245) 1747 (232) 1789 (250) 2136 (273) 1069 (158) 1165 (167) 1083 (198) 903 (140) 848 (144) 663 (107) 845 (199) 507 (101)

30 527 (3119) 122 108 (12 476) 29 857 (2796) 119 428 (11 184) 17 760 (1577) 71 040 (6308) 13 178 (1396) 52 712 (5584) 14 805 (2017) 59 220 (8068) 13 698 (3297) 54 792 (13 188)

55.5 (41.9-69.0) 14.3 (11.1-17.5) 59.9 (47.7-72.2) 17.9 (14.3-21.5) 60.2 (46.0-74.4) 16.4 (12.2-20.6) 82.2 (57.7-106.7) 17.1 (13.2-21.1) 57.3 (37.9-76.7) 11.2 (7.7-14.7) 61.7 (22.6-100.8) 9.2 (4.0-14.5)

49 781 (1942) 48 473 (1763) 51 771 (1908) 55 264 (1920) 51 451 (1954) 47 540 (1689) 55 336 (2041) 48 607 (1697) 69 496 (2629) 53 358 (1886) 68 132 (2892) 54 876 (2214)

1 790 879 (37 305) 7 163 516 (149 220) 1 732 854 (40 400) 6 931 416 (161 600) 1 551 313 (41 361) 6 205 252 (165 444) 1 486 381 (41 854) 5 945 524 (167 416) 1 847 434 (66 830) 7 389 736 (267 320) 1 662 590 (54 406) 6 650 360 (217 624)

27.8 (25.6-30.0) 6.8 (6.3-7.3) 29.9 (27.7-32.0) 8.0 (7.4-8.5) 33.2 (30.8-35.5) 7.7 (7.2-8.2) 37.2 (34.7-39.8) 8.2 (7.6-8.7) 37.6 (35.2-40.1) 7.2 (6.7-7.7) 41.0 (37.9-44.1) 8.2 (7.6-8.9)

*Population estimates derived from hospital births (see Methods).

between AI/AN and white children. Hospital deaths were rare for both racial groups. The point estimate for hospital mortality was suppressed for AI/AN children owing to small numbers, but the 95% CI was 0.00%-0.11%. Mortality for white children was also low (0.19%; 95% CI, 0.16%-0.23%). Pneumonia and bronchiolitis were the most common LRTI diagnoses. The proportion of LRTI-associated discharges coded for bronchiolitis was similar across racial groups. However, the proportion of discharges in AI/AN children coded for pneumonia was slightly greater than that in white children (51.6% vs 47.5%; P < .001). Many LRTI-associated discharges also included a diagnostic code for asthma as a secondary diagnosis; however, the proportion of discharges was similar across racial groups. To investigate whether the observed racial differences in LRTI-associated hospitalization rates were attributable to confounding by socioeconomic status, we used multivariate regression models and restricted our analysis to infants hospitalized in 2012. In preliminary models, there was a significant interaction (P < .01) between race and hospital census region on the risk of LRTI-associated hospitalization; therefore, final models were stratified by region. In these models, AI/AN race was a significant predictor for LRTI-associated

7.9 to 5.2; P < .001) but not for AI/AN children (12.2 to 11.6). Although pneumonia hospitalization rates declined for both racial groups in 2012, racial disparities persisted, especially in the West where rates were >4 times greater for AI/AN infants than for white infants in that region (Table III; available at www.jpeds.com). Similar race-specific rates were observed for discharges coded for pneumococcal pneumonia (Table IV; available at www.jpeds.com). Crude incidence rate ratios were elevated for pneumonia and for pneumococcal pneumonia hospitalization in AI/AN children living in the West, and these were also elevated in the South (Tables V and VI; available at www.jpeds.com). Based on descriptive information about LRTI-associated discharges at nonfederal hospitals in 2012, there were differences between white and AI/AN patients in socioeconomic status, as indicated by primary payer, mean household income, and rural-urban residence (Table VII). AI/AN patients were more likely to have Medicaid insurance, to be in the lowest income quartile, and to live in a more rural county. We failed to find any racial disparities in measures of illness severity. Median length of stay was 2 days (IQR, 1-3) for both AI/AN and white patients. The proportion of discharges associated with respiratory failure was similar

Table II. LRTI-associated hospital discharge rates per 1000 children*, by race, US census region, and year, 1997-2012 AI/AN rate (95% CI) Year Age (y) 2012 <1 1-4 2009 <1 1-4 2006 <1 1-4 2003 <1 1-4 2000 <1 1-4 1997 <1 1-4

Northeast

Midwest

20.7 (10.3-31.1) 55.3 (25.3-85.3) 8.4 (5.2-11.5) 10.3 (4.8-15.8) 27.4 (17.1-37.7) 56.0 (36.0-75.9) 7.4 (4.7-10.2) 11.8 (7.9-15.6) 31.3 (21.1-41.6) 92.1 (55.0-129.2) 6.7 (4.2-9.3) 19.0 (11.2-26.8) 38.1 (21.3-54.8) 147.3 (82.7-211.9) 10.2 (7.0-13.4) 29.3 (18.5-40.1) 35.4 (24.9-45.8) 76.0 (33.2-118.8) 8.3 (5.8-10.9) 13.9 (7.2-20.5) 20.6 (0.3-40.8) 68.3 (37.4-99.2) 3.7 (0.1-7.4) 9.0 (3.4-14.6)

South

White rate (95% CI) West

37.6 (23.7-51.6) 72.8 (45.8-99.8) 16.2 (10.2-22.3) 15.2 (10.0-20.4) 34.2 (20.9-47.4) 90.1 (63.7-116.4) 20.8 (8.2-10.4) 21.5 (15.2-27.8) 37.6 (24.2-51.0) 87.6 (49.3-125.9) 21.1 (11.5-30.8) 16.0 (8.9-23.1) 52.5 (35.3-69.7) 108.2 (38.6-177.8) 11.5 (7.9-15.1) 21.2 (10.3-32.0) 28.8 (15.8-41.8) 106.8 (46.6-167.1) 6.6 (4.4-8.7) 19.0 (7.6-30.4) 69.8 (0-141.3) 119.5 (31.8-207.2) 7.2 (2.0-12.4) 18.4 (8.3-28.5)

Northeast

Midwest

South

West

25.2 (20.1-30.4) 5.9 (4.8-7.1) 26.3 (21.7-30.9) 6.6 (5.5-7.8) 28.2 (24.0-32.4) 6.2 (5.3-7.1) 29.1 (24.5-33.7) 6.0 (5.1-7.0) 32.4 (26.9-37.9) 5.5 (4.6-6.4) 36.1 (30.9-41.2) 6.3 (5.4-7.2)

31.8 (25.7-38.0) 7.2 (6.0-8.4) 30.1 (25.8-34.4) 8.2 (7.2-9.3) 40.6 (33.9-47.3) 9.3 (8.0-10.7) 42.6 (36.4-48.8) 10.0 (8.7-11.3) 38.0 (33.0-43.1) 7.0 (6.0-8.0) 47.9 (40.8-55.0) 9.8 (8.3-11.3)

29.6 (25.8-33.4) 7.4 (6.5-8.3) 34.3 (30.0-38.7) 9.3 (8.2-10.4) 34.9 (30.6-39.2) 8.6 (7.7-9.6) 44.0 (38.6-49.3) 9.7 (8.6-10.8) 43.3 (38.2-48.4) 9.1 (8.0-10.1) 43.3 (36.8-49.8) 9.3 (8.0-10.6)

22.2 (18.1-26.4) 5.8 (4.7-7.0) 25.1 (20.5-29.7) 6.6 (5.3-7.9) 28.6 (22.8-34.4) 5.8 (4.6-7.0) 28.7 (23.7-33.6) 6.2 (5.0-7.3) 33.0 (27.2-38.8) 6.4 (5.3-7.5) 35.9 (28.6-43.1) 7.2 (5.9-8.5)

*Population denominators estimated from race-specific hospital birth discharges (see Methods).

4

Weinert and Edmonson

- 2016

ORIGINAL ARTICLES

Table VII. Selected characteristics of LRTI-associated hospital discharges in children, by race, 2012 AI/AN (n = 3441) discharges Characteristics Age group (y)† <1 1-4 Sex, male† Census Region† Northeast Midwest South West Rural-urban residence† <50 000 county population 50 000-999 999 county population >1 million county population Household income quartile by zip code 1 (lowest) 2 3 4 (highest) Missing Primary expected payer Medicaid Private Other Transfer status† Transfer in Transfer out LRTI diagnosis Pneumonia Bronchiolitis Asthma diagnosis Respiratory failure

White (n = 98 254) discharges

n*

% (95% CI)

n*

% (95% CI)

1694 1747 1943

49.2 (45.9-52.6) 50.8 (47.4-54.1) 56.5 (54.2-58.7)

49 781 48 473 55 967

50.7 (49.8-51.5) 49.3 (48.5-50.2) 57.0 (56.6-57.4)

112 455 988 1886

3.2 (1.8-4.7) 13.2 (6.0-20.5) 28.7 (17.7-39.7) 54.8 (42.1-67.5)

15 977 24 891 40 667 16 719

16.3 (13.5-19.0) 25.3 (21.6-29.0) 41.4 (37.7-45.1) 17.0 (14.1-20.0)

1794 1074 573

52.1 (44.4-59.9) 31.2 (22.7-39.8) 16.6 (9.0-24.3)

27 719 32 403 37 988

28.2 (26.3-30.1) 33.0 (30.0-36.0) 38.7 (35.2-42.2)

1804 599 442 233 362

52.4 (46.2-58.7) 17.4 (13.9-20.9) 12.9 (10.0-15.7) 6.8 (3.3-10.2) 10.5 (5.6-15.4)

29 480 26 125 22 520 18 700 1428

30.0 (28.2-31.8) 26.6 (25.4-27.8) 22.9 (21.9-24.0) 19.0 (17.3-20.7) 1.4 (1.3-1.6)

2626 417 398

76.3 (70.8-81.9) 12.1 (9.6-14.6) 11.6 (6.0-17.1)

48 257 43 163 6835

49.1 (47.6-50.6) 43.9 (42.4-45.5) 7.0 (6.3-7.6)

684 94

19.9 (13.2-26.5) 2.7 (1.8-3.6)

9559 2312

1775 2085 747 310

51.6 (47.5-55.6) 60.6 (56.8-64.4) 21.7 (18.8-24.6) 9.0 (5.7-12.3)

46 645 61 079 20 246 6845

9.7 (8.6-10.8) 2.4 (2.2-2.5) 47.5 (46.5-48.4) 62.2 (61.2-63.1) 20.6 (20.0-21.2) 7.0 (6.1-7.8)

*Weighted estimate of number of discharges. †Percent missing <0.5%.

hospitalization only in the West census region (aOR, 2.5; 95% CI, 1.8-3.4). The Figure shows aORs for racial disparity in LRTI-associated hospitalizations by census region. After accounting for race in final models, male sex, and Medicaid insurance remained statistically significant risk factors in each census region (Table VIII; available at www.jpeds.com). Although low household income and rural residence approached statistical significance in all regions, these factors were only significant in the South census region. However, the “missing” category for income, a proxy for highly rural areas of residence, also was significant in final models for the West.

Discussion Our study showed persistent disparities in LRTI-associated hospitalization rates for AI/AN children and extended earlier work by using a current and nationally representative sample of children discharged from nonfederal hospitals. Our study showed that these racial disparities persisted despite declines in LRTI and pneumonia rates during 2012. These disparities were most prominent in the West and for infants. We found that AI/AN patients were more likely to have markers of lower socioeconomic status, including Medicaid insurance,

rural residence, and low household income. Yet, to the extent that we could control for socioeconomic factors in our regression models, AI/AN race remained a significant predictor of LRTI-associated hospitalization in the West. Since the introduction of the pneumococcal conjugate vaccines, studies have shown declining rates of hospitalization of children for pneumonia.6,7,19 Yet, racial disparities in rates for all LRTIs of all types have persisted, suggesting that, even though the vaccines are important, there are other factors contributing to the higher rates of LRTI-associated hospitalization in AI/AN children. We speculate that the higher rates of LRTIs, especially in the West census region, are attributable to modifiable risk factors such as living conditions and health care access.5-7 Living conditions that predispose AI/AN children to LRTIs include overcrowded housing, smoke exposure, poor indoor air quality, and lack of adequate plumbing, which are of particular concern on reservations in the West census region.5-7,28,29 Lower socioeconomic status and barriers to accessing health care also may play a role in the higher rates of LRTI-associated hospitalization for AI/AN children.6 This pattern has been seen in other populations; a recent study in Ohio found significant variation in LRTI hospitalizations according to census tract.30 LRTI hospitalization “hotspots” were located in urban areas of poverty.30

Hospitalizations at Nonfederal Facilities for Lower Respiratory Tract Infection in American Indian and Alaska Native Children Younger than 5 Years of Age, 1997-2012

5

THE JOURNAL OF PEDIATRICS



An analysis of discharge data on LRTI-associated hospitalizations of IHS enrollees (AI/AN children using the IHS health care system) by Foote et al7 showed that rates (per 1000) fell from 31.9 in 1998-1999 to 20.7 by 2009-2011 for children <5 years old and from 121.5 to 75.1 in infants <1 year of age.7 Using an entirely different method, based on discharges from nonfederal hospitals in a nationally representative sample, we found that rates in 2012 were comparable (22.5 [95% CI, 17.5-27.6] in children ages <5 years and 55.5 [95% CI, 41.9-69.0] in infants). In our study, the 2012 LRTI-associated hospitalization rate for AI/AN infants in the West census region (72.8 [95% CI, 45.8-99.8]) was similar to the rate in the Southwest (90.1) found by Foote et al.7 We are aware of one study that analyzed LRTIassociated discharges of AI/AN children from nonfederal facilities.8 In this study, based on information in the HCUP State Inpatient Database for Arizona, the rate for AI/AN infants was 51.2 during 2005-2008.8 Our study is the first to analyze LRTI disparities for AI/AN children at nonfederal hospitals nationwide, and our results are consistent with previously reported rates in IHS enrollees, especially in the West. Although trend analysis failed to find a significant decrease in AI/AN LRTI-associated hospitalization rates in our study, there was a noticeable decline in the point estimates for rates over time, especially in 2012 (Table II). One limitation of our study is that the LRTI-associated hospitalization rates were calculated using hospital births to derive proxy population denominators. However, results through this method were consistent with recent rate estimates,7 and we designed our study to circumvent known problems with misclassification of race in administrative databases. Another limitation is that our data sample likely included some IHS contract health hospital discharges, and some of these discharges could have occurred in children who had been born at IHS hospitals. Based on information about primary expected payer, however, a large majority (88.4%) of discharges in AI/AN children had Medicaid or private insurance rather than other forms of coverage (which may have included IHS contracted health care in some cases; Table VII). A third limitation is that our trend analysis may have been impacted by several problems inherent in the 1997 KID sample, including a high frequency of missing information about race/ethnicity and under-representation of several states with large AI/AN populations. Comparison of rates across the study period may have been biased if states (such as Oklahoma, New Mexico, and Alaska) that were added later included AI/AN populations with higher than average rates of LRTI-associated discharges. Although we failed to find a significant decrease in LRTI-associated discharge rates over time as reported in recent studies of rates in IHS enrollees,5,7 we did find that point estimates for national rates in 2012 were noticeably lower than 1997 rates. We theorize that we failed to find a significant decrease because of the imprecision of 1997 rates and the later addition of states with potentially higher LRTIassociated hospitalization rates. A final limitation is that information related to socioeconomic status available 6

Volume -

www.jpeds.com

through KID is likely inadequate to fully describe the living conditions and socioeconomic status of AI/AN children, especially in the West. Although the rate of missing information on race has been improving over time in the KID, it still occurs for approximately 8% of discharges in the most recent (2012) sample. Misclassification likely contributes to an undercounting of LRTI-associated hospitalizations in AI/AN children in any given year. Misclassification of AI/AN status has been noted on death certificates and administrative records, as well as various cancer registries and HIV/AIDS databases.11,31-37 One pediatric study found that information on race and ethnicity in an administrative database correlated poorly with birth certificate data and that AI/AN children were more likely to have missing race information in administrative data compared with other children (22% vs 15%).34 Of note, one of the goals of the 2010 Affordable Care Act is to improve research and data collection on health disparities.38 Standardization of data collection on race and ethnicity will be critical to better understand and then address health disparities. Our study analyzed rates of LRTI-associated hospitalization of AI/AN children at nonfederal hospitals based on a nationally representative sample. We found persistent disparities in these hospitalization rates overall and for pneumonia and pneumococcal pneumonia specifically. LRTIassociated discharge rates are especially high for AI/AN infants in the West, and are comparable with rates recently reported for IHS enrollees.5,7 There are important next steps that this study highlights, including: (1) the need for more consistent and accurate race and ethnicity data collection for health research, especially for small minority populations; (2) a need for better information about small-area variation in these rates; and (3) the need for further research on modifiable risk factors for LRTI in AI/AN children including living conditions and health care access. Despite the widespread use of effective pneumococcal conjugate vaccines, disparities in LRTI-associated hospitalization persist, especially in the West census region, suggesting that other modifiable risk factors will need to be addressed to reduce disparities for AI/AN children. n We acknowledge Jens C. Eickhoff, PhD (Department of Biostatistics and Medical Informatics, School of Medicine and Public Health, University of Wisconsin-Madison), for assistance with statistical analysis. Submitted for publication Dec 17, 2015; last revision received Feb 25, 2016; accepted Mar 4, 2016. Reprint requests: Bethany A. Weinert, MD, MPH, Department of Pediatrics, University of Wisconsin School of Medicine and Public Health, 2870 University Avenue, Madison, WI 53705. E-mail: [email protected]

References 1. Chang AB, Chang CC, O’Grady K, Torzillo PJ. Lower respiratory tract infections. Pediatr Clin North Am 2009;56:1303-21. 2. Holman RC, Curns AT, Cheek JE, Bresee JS, Singleton RJ, Carver K, et al. Respiratory syncytial virus hospitalizations among American Indian and

Weinert and Edmonson

- 2016

3.

4.

5.

6.

7.

8.

9.

10.

11.

12.

13.

14.

15.

16.

17.

18.

19.

Alaska Native infants and the general United States infant population. Pediatrics 2004;114:e437-44. Nair H, Simoes EA, Rudan I, Gessner BD, Azziz-Baumgartner E, Zhang JS, et al. Global and regional burden of hospital admissions for severe acute lower respiratory infections in young children in 2010: a systematic analysis. Lancet 2013;381:1380-90. Peck AJ, Holman RC, Curns AT, Lingappa JR, Cheek JE, Singleton RJ, et al. Lower respiratory tract infections among American Indian and Alaska Native children and the general population of U.S. Children. Pediatr Infect Dis J 2005;24:342-51. Singleton RJ, Holman RC, Folkema AM, Wenger JD, Steiner CA, Redd JT. Trends in lower respiratory tract infection hospitalizations among American Indian/Alaska Native children and the general US child population. J Pediatr 2012;161:296-302.e2. Groom AV, Hennessy TW, Singleton RJ, Butler JC, Holve S, Cheek JE. Pneumonia and influenza mortality among American Indian and Alaska Native people, 1990-2009. Am J Public Health 2014;104:S460-9. Foote EM, Singleton RJ, Holman RC, Seeman SM, Steiner CA, Bartholomew M, et al. Lower respiratory tract infection hospitalizations among American Indian/Alaska Native children and the general United States child population. Int J Circumpolar Health 2015;74:29256. Callinan LS, Holman RC, Esposito DH, McDonald M. Racial/ethnic disparities in infectious disease hospitalizations in Arizona. J Health Dispar Res Pract 2013;6:49-71. Singleton RJ, Wirsing EA, Haberling DL, Christensen KY, Paddock CD, Hilinski JA, et al. Risk factors for lower respiratory tract infection death among infants in the United States, 1999-2004. Pediatrics 2009;124: e768-76. Norris T, Vines PL, Hoeffel EM. The American Indian and Alaska Native population: 2010. Washington, DC: US Department of Commerce, Economics and Statistics Administration, US Census Bureau; 2012. www. census.gov/prod/cen2010/briefs/c2010br-10.pdf. Accessed February 2, 2012. Indian Health Service (IHS). Trends in Indian Health: 2014 edition. Rockville (MD): Indian Health Service, US Department of Health and Human Services; 2014. www.ihs.gov/dps/index.cfm/publications/ trends2014/. Accessed June 6, 2015. Indian Health Service (IHS). Find health care. Rockville (MD): Indian Health Service, US Department of Health and Human Services; 2015. www.ihs.gov/findhealthcare/. Accessed July 16, 2015. Healthcare Cost and Utilization Project (HCUP). Introduction to the HCUP kids’ inpatient database (KID) 2012. Rockville (MD): US Agency for Healthcare Research and Quality; 2014. www.hcup-us.ahrq.gov/db/ nation/kid/kid_2012_introduction.jsp. Accessed June 2, 2015. HCUP Technical Assistance. Healthcare Cost and Utilization Project (HCUP). HCUP sample design: national databases. Common errors. Rockville (MD): US Agency for Healthcare Research and Quality; 2015. www.hcup-us.ahrq.gov/tech_assist/tutorials.jsp. Accessed January 14, 2016. Healthcare Cost and Utilization Project (HCUP). Calculating Kids’ Inpatient Database (KID) variances, HCUP methods series report # 2005-5. Rockville (MD): US Agency for Healthcare Research and Quality; 2005. www.hcup-us.ahrq.gov/db/nation/kid/reports/CalculatingKIDVariances. pdf. Accessed June 2, 2015. Indian Health Service (IHS). Contract health services. Rockville (MD): Indian Health Service, US Department of Health and Human Services; 2015. www.ihs.gov/chs/. Accessed July 16, 2015. Centers for Disease Control and Prevention. ICD-9-CM official guidelines for coding and reporting. National Center for Health Statistics; 2011. www.cdc.gov/nchs/data/icd/icd9cm_guidelines_2011.pdf. Accessed June 2, 2015. Rothberg MB, Pekow PS, Priya A, Lindenauer PK. Variation in diagnostic coding of patients with pneumonia and its association with hospital risk-standardized mortality rates: a cross-sectional analysis. Ann Intern Med 2014;160:380-8. Grijalva CG, Nuorti JP, Arbogast PG, Martin SW, Edwards KM, Griffin MR. Decline in pneumonia admissions after routine childhood immunisation with pneumococcal conjugate vaccine in the USA: a time-series analysis. Lancet 2007;369:1179-86.

ORIGINAL ARTICLES 20. Guevara RE, Butler JC, Marston BJ, Plouffe JF, File TM Jr, Breiman RF. Accuracy of ICD-9-CM codes in detecting community-acquired pneumococcal pneumonia for incidence and vaccine efficacy studies. Am J Epidemiol 1999;149:282-9. 21. Healthcare Cost and Utilization Project (HCUP) Houchens RWD, Elixhauser A. Comparative analysis of the HCUP Kids’ Inpatient Database (KID), 1997. HCUP methods series report #2001-1. Rockville (MD): US Agency for Healthcare Research and Quality; 2001. www. hcup-us.ahrq.gov/reports/methods/KIDComp1997Final.pdf. Accessed June 8, 2015. 22. Cochran WG. Some methods for strengthening the common c2 tests. Biometrics 1954;10:417-51. 23. Armitage P. Tests for linear trends in proportions and frequencies. Biometrics 1955;11:375-86. 24. Wacholder S, Boivin JF. External comparisons with the case-cohort design. Am J Epidemiol 1987;126:1198-209. 25. Prentice RL. A case-cohort design for epidemiologic cohort studies and disease prevention trials. Biometrika 1986;73:1-11. 26. Healthcare Cost and Utilization Project (HCUP). HCUP NIS description of data elements. Rockville (MD): Agency for Healthcare Research and Quality; 2015. www.hcup-us.ahrq.gov/db/nation/nis/nisdde.jsp. Accessed June 9, 2015. 27. Healthcare Cost and Utilization Project (HCUP) Houchens R, Elixhauser A. Final report on calculating Nationwide Inpatient Sample (NIS) variances for data years 2011 and earlier. 2015. HCUP Method Series Report # 2003-02. Rockville (MD): US Agency for Healthcare Research and Quality; 2015. www.hcup-us.ahrq.gov/reports/methods/ methods.jsp. Accessed January 14, 2016. 28. Robin LF, Less PS, Winget M, Steinhoff M, Moulton LH, Santosham M, et al. Wood-burning stoves and lower respiratory illnesses in Navajo children. Pediatr Infect Dis J 1996;15:859-65. 29. Hennessy TW, Ritter T, Holman RC, Bruden DL, Yorita KL, Bulkow L, et al. The relationship between in-home water service and the risk of respiratory tract, skin, and gastrointestinal tract infections among rural Alaska natives. Am J Public Health 2008;98:2072-8. 30. Beck AF, Florin TA, Campanella S, Shah SS. Geographic variation in hospitalization for lower respiratory tract infections across one county. JAMA Pediatr 2015;169:846-54. 31. Espey DK, Jim MA, Richards TB, Begay C, Haverkamp D, Roberts D. Methods for improving the quality and completeness of mortality data for American Indians and Alaska Natives. Am J Public Health 2014; 104:S286-94. 32. Jim MA, Arias E, Seneca DS, Hoopes MJ, Jim CC, Johnson NJ, et al. Racial misclassification of American Indians and Alaska Natives by Indian Health Service contract health service delivery area. Am J Public Health 2014;104:S295-302. 33. Espey DK, Wiggins CL, Jim MA, Miller BA, Johnson CJ, Becker TM. Methods for improving cancer surveillance data in American Indian and Alaska Native populations. Cancer 2008;113:1120-30. 34. Smith N, Iyer RL, Langer-Gould A, Getahun DT, Strickland D, Jacobsen SJ, et al. Health plan administrative records versus birth certificate records: quality of race and ethnicity information in children. BMC Health Serv Res 2010;10:316. 35. Bertolli J, Lee LM, Sullivan PS , AI/AN Race/Ethnicity Data Validation Workgroup. Racial misidentification of American Indians/ Alaska Natives in the HIV/AIDS Reporting Systems of five states and one urban health jurisdiction, U.S., 1984-2002. Public Health Rep 2007;122:382-92. 36. Rhoades DA. Racial misclassification and disparities in cardiovascular disease among American Indians and Alaska Natives. Circulation 2005;111:1250-6. 37. Harwell TS, Hansen D, Moore KR, Jeanotte D, Gohdes D, Helgerson SD. Accuracy of race coding on American Indian death certificates, Montana 1996-1998. Public Health Rep 2002;117:44-9. 38. The Henry J Kaiser Family Foundation. Focus on health care disparities: Key facts. Menlo Park (CA): The Henry J Kaiser Family Foundation; 2012. https://kaiserfamilyfoundation.files.wordpress.com/2013/01/8396. pdf. Accessed June 8, 2015.

Hospitalizations at Nonfederal Facilities for Lower Respiratory Tract Infection in American Indian and Alaska Native Children Younger than 5 Years of Age, 1997-2012

7

THE JOURNAL OF PEDIATRICS



Volume -

www.jpeds.com

Table III. Pneumonia hospitalization rates per 1000 children* by race, US census region, and year, 1997-2012 AI/AN rate (95% CI) Year Age (y) 2012 <1 1-4 2009 <1 1-4 2006 <1 1-4 2003 <1 1-4 2000 <1 1-4 1997 <1 1-4

Northeast

Midwest

4.5 (0.2-8.8) 25.7 (11.3-40.1) 7.0 (4.1-10.0) 8.0 (3.8-12.2) 10.2 (4.4-16.0) 34.6 (22.7-46.6) 6.2 (3.9-8.6) 9.5 (6.7-12.4) 8.1 (4.0-12.1) 55.1 (30.4-79.9) 5.6 (3.1-8.1) 14.8 (8.1-21.4) 9.8 (2.3-17.2) 103.1 (53.6-152.7) 8.2 (5.5-10.9) 23.0 (14.6-31.5) 10.8 (6.6-14.9) 39.2 (14.7-63.7) 6.0 (3.9-8.1) 8.1 (4.1-12.2) 7.1 (0.0-14.2) 29.7 (12.1-47.4) 2.9 (0.1-5.7) 6.6 (2.4-10.8)

South 16.0 (7.9-24.0) 10.4 (6.8-14.0) 16.6 (9.9-23.3) 15.4 (9.4-21.4) 15.6 (9.4-21.9) 14.8 (9.1-20.6) 16.1 (8.0-24.2) 8.4 (5.7-11.1) 11.1 (4.8-17.4) 4.6 (2.9-6.4) 45.3 (0.0-112.4) 7.2 (2.0-12.4)

White rate (95% CI) West

Northeast

25.4 (15.5-35.2) 6.5 (5.3-7.7) 9.3 (5.9-12.7) 4.0 (3.3-4.8) 42.2 (30.7-53.6) 8.2 (6.8-9.5) 15.7 (11.1-20.4) 5.1 (4.3-5.9) 44.8 (22.8-66.8) 7.7 (6.6-8.8) 12.5 (6.6-18.4) 4.7 (4.0-5.3) 65.7 (18.0-113.3) 8.0 (6.7-9.3) 17.7 (8.4-26.9) 4.7 (4.0-5.4) 50.4 (22.6-78.3) 9.0 (7.4-10.6) 12.9 (5.3-20.5) 4.3 (3.5-5.0) 65.6 (12.9-118.4) 10.6 (9.0-12.1) 14.6 (6.2-23.1) 4.6 (3.9-5.3)

Midwest 9.8 (8.3-11.3) 5.0 (4.2-5.8) 12.1 (10.5-13.6) 6.2 (5.4-7.0) 15.1 (13.0-17.2) 6.8 (5.8-7.7) 16.6 (14.5-18.8) 7.5 (6.5-8.4) 15.6 (13.2-18.1) 5.4 (4.6-6.1) 20.6 (17.0-24.2) 7.0 (5.9-8.1)

South

West

8.7 (7.6-9.7) 5.6 (4.7-6.6) 5.0 (4.4-5.6) 3.6 (2.9-4.4) 12.2 (10.8-13.6) 8.2 (6.9-9.4) 6.8 (6.1-7.6) 4.6 (3.8-5.4) 11.2 (9.9-12.5) 9.6 (7.9-11.4) 6.2 (5.5-6.9) 4.3 (3.5-5.1) 13.4 (11.8-15.1) 10.6 (8.6-12.5) 7.2 (6.4-8.0) 4.8 (3.9-5.7) 14.3 (12.4-16.2) 12.0 (10.0-14.1) 6.9 (6.0-7.7) 4.9 (4.0-5.7) 16.0 (13.1-18.9) 14.2 (11.8-16.7) 7.0 (5.9-8.0) 5.5 (4.5-6.4)

*Population denominators estimated from race-specific hospital birth discharges (see Methods).

Table IV. Pneumococcal pneumonia hospitalization rates per 1000 children* by race, US census region, and year, 19972012 AI/AN rate (95% CI) Year

Age (y)

2012

<1 1-4 <1 1-4 <1 1-4 <1 1-4 <1 1-4 <1 1-4

2009 2006 2003 2000 1997

White rate (95% CI)

Northeast

Midwest

South

West

Northeast

Midwest

South

West

1.9 (0.0-4.1) 4.8 (2.5-7.0) 7.1 (2.4-11.7) 5.2 (3.2-7.2) 6.1 (2.6-9.6) 4.8 (2.6-7.0) 10.4 (3.9-16.8) 8.1 (5.4-10.8) 10.3 (5.9-14.7) 5.8 (3.8-7.9) 6.2 (0.3-12.2) 2.7 (0.1-5.3)

18.8 (7.6-29.9) 6.2 (3.0-9.4) 24.8 (16.1-33.4) 7.9 (5.5-10.4) 40.1 (22.0-58.1) 11.8 (6.3-17.4) 62.2 (35.8-88.6) 15.2 (9.1-21.3) 26.5 (9.3-43.6) 7.6 (3.4-11.8) 24.8 (10.6-39.1) 4.4 (1.4-7.3)

13.1 (7.7-18.6) 7.9 (5.4-10.5) 14.0 (7.8-20.2) 12.2 (7.4-17.0) 14.2 (9.1-19.4) 12.3 (7.1-17.6) 15.4 (7.5-23.2) 7.4 (4.8-9.9) 9.9 (4.4-15.4) 3.8 (2.3-5.4) 36.5 (0.0-90.6) 7.5 (2.1-13.0)

21.2 (12.5-30.0) 8.1 (5.0-11.2) 37.1 (26.5-47.7) 12.8 (9.0-16.6) 36.7 (17.4-55.9) 9.1 (4.3-13.9) 45.6 (12.7-78.4) 12.5 (4.8-20.1) 37.5 (7.5-67.6) 10.9 (2.8-19.1) 41.4 (4.6-78.2) 9.8 (4.7-14.9)

6.1 (4.7-7.4) 3.4 (2.7-4.0) 7.3 (5.8-8.8) 4.4 (3.6-5.1) 6.8 (5.8-7.8) 4.2 (3.6-4.8) 6.9 (5.7-8.1) 4.2 (3.5-4.8) 7.4 (6.0-8.9) 3.8 (3.1-4.4) 8.5 (7.1-9.9) 4.1 (3.4-4.7)

9.0 (7.2-10.8) 4.2 (3.5-4.9) 10.2 (8.6-11.8) 5.2 (4.5-5.9) 12.2 (10.1-14.3) 5.6 (4.8-6.4) 13.1 (10.9-15.2) 6.0 (5.2-6.8) 10.0 (8.2-11.8) 4.2 (3.6-4.9) 12.8 (10.4-15.3) 5.3 (4.5-6.1)

7.6 (6.5-8.8) 4.1 (3.6-4.6) 10.3 (8.9-11.7) 5.5 (4.9-6.2) 9.6 (8.2-11.0) 5.2 (4.6-5.8) 10.8 (9.4-12.3) 5.9 (5.2-6.6) 11.2 (9.6-12.8) 5.8 (5.1-6.6) 11.2 (9.5-12.9) 5.5 (4.8-6.3)

5.2 (4.1-6.4) 2.9 (2.4-3.5) 6.7 (5.5-7.9) 3.6 (3.0-4.2) 7.3 (5.8-8.8) 3.4 (2.8-4.1) 7.9 (6.2-9.5) 3.7 (3.1-4.4) 8.2 (6.6-9.7) 4.0 (3.2-4.6) 8.9 (7.2-10.6) 4.2 (3.5-5.0)

*Population denominators estimated from race-specific hospital birth discharges (see Methods).

7.e1

Weinert and Edmonson

- 2016

ORIGINAL ARTICLES

Table V. Incidence rate ratios with 95% CIs for pneumonia* in AI/AN vs white children ages 0-4 years, by US census region, 2012 <1 year 1-4 years Total

Northeast

Midwest

South

West

0.7 (0.0-1.4) 1.8 (1.0-2.6) 1.4 (0.8-2.1)

2.6 (1.1-4.1) 1.6 (0.7-2.5) 1.9 (0.9-3.0)

1.8 (0.9-2.8) 2.1 (1.3-2.8) 2.0 (1.3-2.8)

4.5 (2.6-6.4) 2.6 (1.5-3.6) 3.1 (1.8-4.4)

*Based on case definition used by Grijalva et al19; n = 1775 discharges for AI/AN children and n = 46 654 discharges for white children.

Table VI. Incidence rate ratios with 95% CIs for pneumococcal pneumonia* in AI/AN vs white children ages 0-4 years, by US census region, 2012 <1 year 1-4 years Total

Northeast

Midwest

South

West

0.3 (0.0-0.7) 1.4 (0.7-2.1) 1.1 (0.6-1.6)

2.1 (0.8-3.4) 1.5 (0.7-2.3) 1.7 (0.8-2.6)

1.7 (1.0-2.5) 1.9 (1.3-2.6) 1.9 (1.2-2.5)

4.1 (2.9-5.3) 2.8 (1.6-4.0) 3.2 (1.8-4.6)

*Based on case definition used by Guevara et al20; n = 1339 discharges for AI/AN children and n = 35 595 discharges for white children.

Hospitalizations at Nonfederal Facilities for Lower Respiratory Tract Infection in American Indian and Alaska Native Children Younger than 5 Years of Age, 1997-2012

7.e2

THE JOURNAL OF PEDIATRICS



Volume -

www.jpeds.com

Table VIII. Logistic regression analysis of risk for LRTI-associated hospitalization in infants, 2012, stratified by census region

Northeast Race AI/AN White Sex Male Female Rural-urban residence <50 000 county population 50 000-999 999 county population >1 million county population Household income quartile by zip code Missing 1 (lowest) 2 3 4 (highest) Primary expected payer Medicaid Other Private insurance Midwest Race AI/AN White Sex Male Female Rural-urban residence <50 000 county population 50 000-999 999 county population >1 million county population Household income quartile by zip code Missing 1 (lowest) 2 3 4 (highest) Primary expected payer Medicaid Other Private insurance South Race AI/AN White Sex Male Female Rural-urban residence <50 000 county population 50 000-999 999 county population >1 million county population Household income quartile by zip code Missing 1 (lowest) 2 3 4 (highest)

7.e3

LRTI cases*

Birth cohort*

Weighted n = 8274 N (SE)

Weighted n = 328 106 N (SE)

Unadjusted OR (95% CI)

aOR (95% CI)

43 (11) 8231 (844)

2057 (312) 326 049 (17 000)

0.8 (0.5-1.3) 1 (ref)

0.7 (0.5-1.2) 1 (ref)

4945 (505) 3328 (351)

168 393 (8703) 159 713 (8435)

1.4 (1.3-1.5) 1 (ref)

1.4 (1.3-1.5) 1 (ref)

1228 (150) 2244 (319) 4801 (709)

43 031 (3087) 87 880 (9055) 197 195 (16 652)

1.2 (0.9-1.6) 1.1 (0.8-1.4) 1 (ref)

1.0 (0.7-1.5) 1.0 (0.7-1.4) 1 (ref)

134 (30) 1476 (203) 1807 (252) 2038 (248) 2819 (356)

4589 (691) 49 748 (4374) 66 884 (5372) 85 657 (5905) 121 228 (9640)

1.3 (0.9-1.8) 1.3 (1.0-1.6) 1.2 (0.9-1.5) 1.0 (0.9-1.2) 1 (ref)

0.9 (0.6-1.2) 0.9 (0.7-1.3) 0.9 (0.6-1.2) 0.9 (0.7-1.1) 1 (ref)

3670 (394) 425 (69) 4178 (473)

93 890 (5692) 16 691 (1546) 217 525 (13 039)

2.0 (1.8-2.4) 1.3 (1.0-1.7) 1 (ref)

2.1 (1.8-2.4) 1.4 (1.0-1.8) 1 (ref)

Weighted n = 13 315 N (SE)

Weighted n = 41 4019 N (SE)

261 (78) 13 054 (1281)

4701 (1055) 409 318 (16 916)

1.7 (1.0-3.0) 1 (ref)

1.4 (0.8-2.3) 1 (ref)

7667 (756) 5648 (547)

212 069 (8784) 201 950 (8357)

1.3 (1.2-1.4) 1 (ref)

1.3 (1.2-1.4) 1 (ref)

4426 (388) 4185 (538) 4704 (863)

118 978 (5393) 133 934 (10 787) 161 107 (14 168)

1.3 (0.9-1.8) 1.1 (0.7-1.6) 1 (ref)

1.1 (0.8-1.6) 1.0 (0.6-1.5) 1 (ref)

104 (19) 3642 (439) 4294 (400) 3401 (380) 1875 (274)

2537 (226) 87 609 (5027) 133 976 (5708) 116 375 (6024) 73 521 (6809)

1.6 (1.1-2.4) 1.6 (1.2-2.1) 1.3 (1.0-1.6) 1.2 (1.0-1.4) 1 (ref)

1.2 (0.8-1.8) 1.2 (0.9-1.5) 1.0 (0.8-1.2) 1.0 (0.9-1.2) 1 (ref)

6948 (765) 811 (144) 5556 (539)

147 323 (5756) 33 643 (2929) 233 053 (12 467)

2.0 (1.7-2.3) 1.0 (0.8-1.3) 1 (ref)

1.9 (1.7-2.1) 1.0 (0.7-1.3) 1 (ref)

Weighted n = 20 656 N (SE)

Weighted n = 694 187 N (SE)

362 (88) 20 294 (1264)

9617 (1473) 684 570 (24 016)

1.3 (0.9-1.9) 1 (ref)

1.0 (0.7-1.5) 1 (ref)

12 084 (746) 8572 (538)

357 915 (12 354) 336 272 (11 892)

1.3 (1.3-1.4) 1 (ref)

1.3 (1.3-1.4) 1 (ref)

7028 (459) 7711 (712) 5917 (756)

160 106 (6939) 233 113 (15 189) 300 969 (22 263)

2.2 (1.7-2.9) 1.7 (1.2-2.3) 1 (ref)

1.8 (1.3-2.4) 1.5 (1.1-2.1) 1 (ref)

297 (29) 8721 (561) 5541 (405) 3797 (346) 2299 (267)

6650 (374) 219 198 (9170) 177 336 (7657) 160 475 (8305) 130 529 (10 770)

2.5 (2.0-3.3) 2.3 (1.8-2.8) 1.8 (1.4-2.2) 1.3 (1.1-1.6) 1 (ref)

1.5 (1.2-1.9) 1.2 (1.0-1.5) 1.1 (0.9-1.4) 1.1 (0.9-1.3) 1 (ref) (continued )

Weinert and Edmonson

- 2016

ORIGINAL ARTICLES

Table VIII. Continued South Primary expected payer Medicaid Other Private insurance West Race AI/AN White Sex Male Female Rural-urban residence <50 000 county population 50 000-999 999 county population >1 million county population Household income quartile by zip code Missing 1 (lowest) 2 3 4 (highest) Primary expected payer Medicaid Other Private insurance

Weighted n = 20 656 N (SE)

Weighted n = 694 187 N (SE)

12 118 (770) 1314 (125) 7224 (524)

284 812 (11 639) 57 365 (2973) 352 010 (15 990)

2.1 (1.8-2.3) 1.1 (1.0-1.3) 1 (ref)

1.8 (1.6-2.0) 1.0 (0.9-1.2) 1 (ref)

Weighted n = 9181 N (SE)

Weighted n = 379 779 N (SE)

1028 (215) 8153 (793)

14 087 (2512) 365 691 (15 286)

3.3 (2.3-4.7) 1 (ref)

2.5 (1.8-3.4) 1 (ref)

5469 (511) 3711 (376)

194 372 (8162) 185 406 (7636)

1.4 (1.3-1.5) 1 (ref)

1.4 (1.3-1.5) 1 (ref)

1710 (168) 3194 (415) 4277 (649)

55 286 (3182) 119 317 (10 764) 205 176 (13 965)

1.5 (1.1-2.0) 1.3 (0.9-1.8) 1 (ref)

1.0 (0.8-1.4) 1.1 (0.8-1.5) 1 (ref)

330 (69) 1728 (203) 2261 (248) 2518 (271) 2344 (296)

7654 (700) 53 319 (3491) 82 421 (4556) 112 438 (5806) 123 947 (8656)

2.3 (1.6-3.3) 1.7 (1.3-2.2) 1.4 (1.2-1.8) 1.2 (1.0-1.4) 1 (ref)

1.4 (1.0-1.9) 1.1 (0.9-1.4) 1.1 (0.9-1.4) 1.0 (0.9-1.2) 1 (ref)

4610 (454) 829 (134) 3742 (408)

121 840 (5656) 27 528 (1843) 230 411 (11 851)

2.3 (2.0-2.7) 1.8 (1.4-2.4) 1 (ref)

2.1 (1.8-2.3) 1.6 (1.3-2.1) 1 (ref)

*Total of 50 cases and 5315 cohort members excluded owing to missing values.

Hospitalizations at Nonfederal Facilities for Lower Respiratory Tract Infection in American Indian and Alaska Native Children Younger than 5 Years of Age, 1997-2012

7.e4