Data and Sample

Two sources of data are used in this study: the 2001 Rhode Island Health Interview Survey (Rl HIS) and the 2000 U.S. census. The Rl HIS is a random-digit-dial telephone survey administered every 3-5 years. The 2001 survey collected health-related information on children and adults in sampled Rhode Island households. Within each household, the adult with the most knowledge of the health and healthcare utilization of all household members reported for each household member. In 2001, the interviewed sample consisted of 2,600 respondents, which yielded information on 3,263 persons in family units and 6,877 household members. The response rate was 55%, similar to the median response rate of 51.3% for the 2000 Behavioral Risk Factor Surveillance System (BRFSS) survey. For the current study, we excluded children who were Asian (n=46), Hawaiian (n=2), Native American (n=30) or whose race could not be determined (n=20) because sample sizes for these race groups were too small to produce reliable estimates (n=98).

Outcome Variable: Asthma

A single item in the Rl HIS assessed whether a child in the family had asthma. A child was considered to have asthma if the survey respondent answered “yes” to the question: “Did a doctor say that [name of child] has asthma?”

Independent Variables

Neighborhood SEP. 2000 census geography codes were used to link children to their neighborhood characteristics using five-digit census ZIP code tabulation areas (ZCTAs) from the 2000 U.S. census. Census ZCTAs were matched to the reported ZIP code of each household. All 70 ZCTAs for Rhode Island based on the 2000 census were included in this study. ZIP codes were classified as a poverty area if >25% households had incomes <200% of the federal poverty level (FPL). We used the 200% rather than the 100% cutoff to include those individuals who qualify for state and federal means-tested entitlement programs (e.g., Special Supplemental Nutrition Program for Women, Infants and Children; Children’s Health Insurance Program). In addition, we created a variable that combined a child’s race/ethnicity and area of residence: 1) Hispanic children living in poverty neighborhoods, 2) black children living in poverty neighborhoods, 3) white children living in poverty neighborhoods, and 4) children in non-poor neighborhoods. Rhode Island has the highest percentage of Hispanic children living in poverty (47%) compared to the national rate (28%) and the highest percentage of black children living in poverty in New Eng­land (38%). Three-fourths (75%) of children in poverty are concentrated in poor urban communities. When you need your medication buy canadian pharmacy viagra

Family-level SEP. Information on type of health insurance was available for each household member, which was categorized as none, publicly funded or private. Although health insurance is a proxy for access to care, it is also an economic indicator. Respondents were asked their total family income before taxes from all sources during the past 12 months. For respondents who did not want to report a specific dollar amount, a series of questions determined if the income fell within a specific range. Income was classified according to 2001 poverty thresholds adjusted for household size from household income tables available from the U.S. Census Bureau. Other SEP variables included the highest educational level attained by a child’s parents, but the variable was highly correlated with family income and was not included in the multivariable models. cialis canada online pharmacy

Covariates. Other variables included (* indicates reference group): race/ethnicity (Hispanic, black, white*); age (0-5,* 6-11, 12-17 years); sex (boys, girls*); routine medical visit in past year (yes, no*); whether the child had a parent with asthma (yes, no*); and whether the home had asthma triggers, including exposure to passive smoking in the home (yes, no*); and mildew, mold or excess dampness in the home in the past 12 months (yes, no*). The question on whether a parent had asthma was identical to the question asked of children. Doctor-diagnosed asthma in a parent served as a proxy for a genetic predisposition to asthma since a family history of allergy is strongly associated with persistent asthma during childhood.