Insurance Coverage and Financial Burden for Families of Children with Special Health Care Needs

Abstract

Objective

To examine the role of insurance coverage in protecting families of children with special health care needs (CSHCN) from the financial burden associated with care.

Methods

Data from the 2001 National Survey of Children with Special Health Care Needs were analyzed. We built two multivariate regression models using “work loss/cut back” and “experiencing financial problems” as the dependent variables and insurance status as the primary independent variable of interest while adjusting for income, race/ethnicity, functional limitation/severity, and other sociodemographic predictors.

Results

Approximately 29.9 % of CSHCN live in families where their condition led parents to report cutting back on work or stopping work completely. Families of 20.9 % of CSHCN reported experiencing financial difficulties due to the child’s condition. Insurance coverage significantly reduced the likelihood of families experiencing financial problems at every income level. The proportion of families experiencing financial problems was reduced from 35.7 to 23.0 % for the poor and 44.9 to 24.5 % for low-income families with continuous insurance coverage (p<0.01 for both comparisons). Similarly, the proportion of parents having to cut back or stop work was reduced from 42.8 to 35.9 % for the poor (p<0.05) and 43.5 to 33.9 % for low-income families (p<0.01).

Conclusions

Continuous health insurance coverage provides protection from financial burden and hardship for families of CSHCN in all income groups. This evidence is supportive of policies designed to promote universal coverage for CSHCN. However, many poor and low-income families continue to experience work loss and financial problems despite insurance coverage. Hence, health insurance should not be viewed as a solution in itself, but instead as one element of a comprehensive strategy to provide financial safety for families with CSHCN.

Health insurance coverage, public or private, enables access to care and utilization of health services for children.^1–6 Because insurance coverage plays an important role in the health and well-being of children, policy makers have made significant strides in recent years to provide and expand health care coverage for all children in the United States.^7–8 Insurance coverage can be especially important for children with special health care needs (CSHCN) because of their greater needs compared to other children.^{5–6, 9–11}

CSHCN have increased use of health services and greater health care expenditures than the average pediatric population.^{6, 9–11, 30} Using data from the 2000 Medical Expenditure Panel Survey, Newacheck and Kim reported that CSHCN had three times higher annual health care expenditures than children without special health care needs.¹⁰ Research efforts have also consistently shown that insured CSHCN are more likely than those uninsured to have a usual source of care, improved utilization of medical, dental, and mental health services, fewer unmet needs, and higher rates of well-child care.^{5–6, 9–11} All of these findings underscore the importance of health insurance coverage for CSHCN.

However, for many families with CSHCN, health insurance coverage may not be adequate to meet their elevated needs. More so than other children, CSHCN require services that may not be covered under commercial health insurance plans, as well as out-of-network services that require higher co-payments.¹² Furthermore, CSHCN by definition utilize these services at greater frequencies than other children, thus incurring greater out-of-pocket expenses.^12–13 These cost-sharing responsibilities and out-of-pocket expenses can result in significant financial burden, particularly for poor families.¹³

In addition, previous studies have demonstrated that the added responsibilities of caring for a child with poor health is associated with reduced parental employment.^14–15 A parent may need to stay at home full time in order to care for a child with functional limitations, resulting in the loss of a wage-earner. Time off from work to take the child to seek care may also result in loss of wage, especially for those receiving hourly wages. Opportunity costs for these families can be far greater than the visible financial costs.^14–15

Several studies have shown that although insurance plays an important protective role for families with CSHCN, insurance coverage alone is often not enough.^17–19 More systematic efforts at reforming service delivery may be required to ensure that families receive the services they need without creating undue burdens. For example, a recent study by Kuhlthau et al. demonstrated that numerous families with CSHCN experience high levels of financial burdens and that well organized systems of care may alleviate some of the burdens experienced by these families.¹⁶

Our study complements the existing body of work in this area by using current nationally representative survey data to document the effects of insurance on financial burden and work loss by families with CSHCN. Our research questions include: 1) what is the independent contribution of insurance on attenuating financial burdens and employment disruptions for families of CSHCN; and 2) Does insurance provide differential levels of protection for families in different socioeconomic strata?

METHODS

Data Source

The 2001 National Survey of Children with Special Health Care Needs was conducted by the US Maternal and Child Health Bureau and the National Center of Health Statistics. It was a cross-sectional survey conducted using the State and Local Area Integrated Telephone Survey (SLAITS) platform.^20–21 From 196,888 households, the National Survey of CSHCN screened 373,055 children nationwide resulting in completed telephone interviews for 38,866 CSHCN (with a minimum of 750 CSHCN per state) between October 2000 and April 2002. Using data from the Census Bureau, sample weights were created to ensure that population estimates were representative of children under 18 years of age in each state and the nation.

The National Survey of CSHCN identifies children with special health care needs by using the CSHCN Screener developed by the Foundation for Accountability.^21–22 The survey also obtained information on health status, functional limitation, access to care, insurance coverage, satisfaction with care, and impact of the condition on the child and family; it is considered the largest and one of the most recent data sources available for CSHCN.^20–21 The survey respondent was the parent or guardian most knowledgeable about the child; the interview was conducted in English, Spanish, or other preferred languages.

Statistical Analysis

Using person-level data, we estimated two types of multivariate regression models using complementary measures of burden on the family as dependent variables: (1) whether or not the family experienced financial problems due to the child’s condition, categorized as having financial problems versus not having financial problem – based on the respondent’s report to “Has [child’s name]’s health conditions caused financial problems for your family?” and (2) whether or not the child’s condition impacted the work status of any family member, categorized as one or more family members having to stop or cut back work versus no impact at work – based on the respondent’s report to “Have you or other family members cut down on the hours you work to care for [child’s name]?” and “Have you or other family members stopped working because of [child’s name]’s health conditions?”. We categorized a “yes” response to either of these questions as an indication of employment impact. Since the financial problems and reduction in employment variables are both binary (yes/no) variables, we used logistic regression to model these variables.²³

We used the same set of explanatory variables for all regression models. The explanatory variables were derived using an adaptation of Andersen’s Behavioral Model of Health Services Use.²⁴ Other research studies focused on financial burden have used a similar sets of covariates, including child health characteristics and socio-demographic variables.^15–17 Bivariate analyses were conducted, using weighted data, on candidate variables and the two outcome variables in order to guide the selection of variables for the multivariate analyses (Table 2). Primary independent variables of interest were: (1) family income, categorized as poor (<100 % federal poverty line), low-income (100–199 % FPL), moderate-income (200–299 % FPL), middle-income (300–399 % FPL), and high-income (400+ % FPL); and (2) indicator variables for insurance coverage, categorized as continuous insurance coverage versus discontinuous coverage or no coverage; continuous coverage is defined as being covered by any private or public plan or program for the entire year prior to the interview date. Other explanatory variables included: age, categorized as 0–5 years of age, 6–11 years of age, and 12–17 years of age; child’s gender, categorized as female and male; race/ethnicity, categorized as non-Hispanic Black, Hispanic, White, multi-racial, and other; functional ability, categorized as affecting child’s activity usually/always or a great deal, and otherwise²⁵; family composition, categorized as one adult, and two or more adults in the family; and interview language, categorized as English, and other than English.

Table 2.

Percentage of Families with Financial Problem or Need to Cut Back on Work, by Explanatory Variables. Weighted Data.

Explanatory Variables	Categories	Financial Problem	Cut Back on Work
Demographic Factors
Age	0–5	21.9 %	40.1 %
	6–11	19.5 %	30.6 %
	12–17	21.7 %	24.5 %
Sex	Female	20.9 %	29.0 %
	Male	20.9 %	30.6 %
Race/ethnicity	Black	19.4 %	31.5 %
	Hispanic	24.8 %	42.4 %
	Multi-racial	24.6 %	33.0 %
	White	20.1 %	27.1 %
	Other	27.2 %	39.0 %
Enabling Factors
Family income	Poor	28.5 %	43.0 %
	Low-income	29.4 %	38.7 %
	Moderate-income	24.2 %	29.4 %
	Middle-income	18.5 %	24.0 %
	High-income	12.0 %	21.5 %
Health insurance	Insured for the entire year	18.3 %	28.2 %
	Uninsured or gaps in coverage	40.5 %	43.3 %
Family composition	Two adults or more	19.5 %	28.9 %
	One adult	27.5 %	35.2 %
Interview language	English	20.6 %	29.0 %
	Other than English	27.3 %	56.5 %
Need
Functional ability	Never or only sometimes affected	15.3 %	22.9 %
	Affected usually/always or greatly	39.5 %	53.3 %

To assess the protective effect of insurance coverage, we used the coefficient estimates from the logistic regression models to predict the proportion of families experiencing financial problems or reduction in employment after adjusting for other explanatory variables, stratified by income. More specifically, for each poverty category, we used the regression coefficients to estimate the probability of a family experiencing financial problems or reductions in employment by first assigning each observation (child) to continuous insurance coverage while leaving the other explanatory variables at their original values and then averaging the individual predictions across each poverty category. This simulation provides the predicted probability that families will experience financial problems or reductions in employment assuming that all children had continuous insurance coverage within each poverty category. Similarly, we repeated the procedure by assigning each observation to not having continuous insurance coverage while leaving the other explanatory variables at their original values and using regression coefficients to predict the proportion of families experiencing financial problems or reductions in employment. The difference between these predicted probabilities provides us with a quantitative estimate of the independent protective effect of insurance across poverty categories.

We conducted all analyses with STATA statistical package (version 8.1, Texas, 2003) using person-level weights provided by the survey to make data representative of the US population. We also accounted for the complex survey design (clustering) in our variance estimation using STATA. P-values of 0.05 or less were chosen as the criterion for statistical significance in all analyses.

RESULTS

Descriptive statistics of the sample are summarized in Table 1. The proportion of families with CSHCN who reported parents needing to stop work or cut back on work in order to care for their child was 29.9%. The overall proportion of families who reported having financial problems due to their child’s care was 20.9%. A large proportion of CSHCN were insured for the entire year (88.4 %). Nearly 97 % of the respondents used English as the interview language. Most CSHCN were from families with two or more adults. Other sociodemographic characteristics such as age, gender, race, and family income for CSHCN were similar to that of the U.S. pediatric population.

Table 1.

Characteristics of study sample.

Explanatory Variables	Categories	Unweighted N	Weighted Percent
Demographic Factors
Age (years)	0–5	6,964	19.4 %
	6–11	15,054	39.5 %
	12–17	16,821	41.1 %
Sex	Female	15,520	40.2 %
	Male	23,320	59.8 %
Race/ethnicity	Black	3,833	14.3 %
	Hispanic	3,253	11.1 %
	Multi-racial	1,366	3.1 %
	White	28,967	68.9 %
	Other	1,311	2.6 %
Enabling Factors
Family income	Poor	5,205	15.0 %
	Low-income	8,145	22.1 %
	Moderate-income	7,020	18.1 %
	Middle-income	5,549	16.5 %
	High-income	9,310	28.3 %
Health insurance	Insured for the entire year	34,666	88.4 %
	Uninsured or gaps in coverage	4,115	11.6 %
Family composition	Two adults or more	31,878	82.9 %
	One adult	6,759	17.1 %
Interview language	English	38,011	96.5 %
	Other than English	855	3.5 %
Need
Functional ability	Never or only sometimes affected	30,327	76.8 %
	Affected usually/always or greatly	8,323	23.2 %
Outcome Variables
Cut back on work	Yes	10,916	29.9 %
	No	27,844	70.1 %
Financial problem	Yes	7,876	20.9 %
	No	30,694	79.1 %

Bivariate results are presented in Table 2. The impact of insurance coverage on parental work status and financial difficulties for families in five income categories based on federal poverty thresholds are presented in Table 3, after adjusting for other explanatory variables. Similar to what was previously reported by the Maternal and Child Health Bureau, families with health insurance coverage were less likely than those without to report having financial problems, even after adjusting for known socioeconomic confounders.²⁵ Note that regardless of income level, continuous insurance coverage significantly reduced the likelihood of a family experiencing financial problems. Even among high-income families with CSHCN, lack of continuous insurance coverage predisposed them to financial difficulties.

Table 3.

The effect of insurance coverage on parent work and family finance, stratified by income (results adjusted for age, gender, race/ethnicity, and functional ability).

Income	Cut back/Stop work	Experiencing financial pb.
Poor (< 100 % FPL)
Insured for entire year	35.9 %	23.0 %
Uninsured or Interrupted coverage	42.8 %*	35.7 %**
Low-income (100–199 % FPL)
Insured for entire year	33.9 %	24.5 %
Uninsured or Interrupted coverage	43.5 %**	44.9 %**
Moderate-income (200–299 % FPL)
Insured for entire year	29.0 %	22.1 %
Uninsured or Interrupted coverage	30.3 %	40.0 %**
Middle-income (300–399 % FPL)
Insured for entire year	25.9 %	19.0 %
Uninsured or Interrupted coverage	33.7 %	33.8 %**
High-income (400+ % FPL)
Insured for entire year	23.9 %	13.1 %
Uninsured or Interrupted coverage	38.8 %**	29.4 %**

^*0.01< P ≤ 0.05 for test of difference with the comparison category.

^**P ≤ 0.01 for test of difference with the comparison category.

Table 4 summarizes other socioeconomic factors that were associated with parents’ cutting back or stopping work and families experiencing financial problems as a result of the child’s care. All else being equal, parents from lower income families were more likely to cut back or stop work. The income gradient is linear – as income increases, the proportion of parents cutting back or stopping work decreases. Causal direction for this association is ambiguous given that our data set is cross-sectional. Cutting back or stopping work also leads to loss of family income, which may result in families falling into a lower income category. Expectedly, all else equal, single adult families were more prone to reduced work and more likely to experience financial difficulties. Similarly, non-English speakers were also prone to reduced work; however, they were not statistically significantly different from English speakers with respect to experiencing financial problems.

Table 4.

Summary of selected socioeconomic variables associated with parents having to cut back/stop work and family experiencing financial problems (results adjusted for age, gender, race/ethnicity, insurance coverage, and functional ability).

Variables	Cut back/Stop work	Experiencing financial pb.
Income
Poor (< 100 % FPL)	35.7 %**	23.8 %**
Low-income (100–199 % FPL)	35.3 %**	26.4 %**
Moderate-income (200–299 % FPL)	30.0 %**	24.4 %**
Middle-income (300–399 % FPL)	26.7 %	20.5 %**
High-income† (400+ % FPL)	24.9 %	14.1 %
Family composition
One adult	32.2 %*	25.7 %**
Two or more adults†	29.9 %	20.7 %
Language
Other than English	39.5 %*	18.3 %
English†	30.0 %	21.8 %

^†Comparison category for the variable.

^*0.01< P ≤ 0.05 for test of difference with the comparison category.

^**P ≤ 0.01 for test of difference with the comparison category.

DISCUSSION

Our results demonstrate that large numbers of families with CSHCN experienced financial problems and many had to cut back or stop work in order to care for the child with special health care needs. Continuous health insurance coverage provided important protection from financial burden for families of CSHCN in all income groups. However, many poor and low-income families continue to experience work loss and financial problems despite continuous insurance coverage. Hence, insurance helps to equalize financial burden and work loss across income groups, but does not entirely eliminate disparities.

Our work is consistent with and builds on previous work by Kuhlthau and others which also demonstrated high levels of financial burden these families face as a result of their children’s health status and conditions.^{11–13, 16} Furthermore, studies on children’s health status and parental employment suggest that caregiving responsibilities may be directly associated with parental labor force involvement.^14–5 Caretakers spend a great deal of time caring for CSHCN; in addition to directly caring for CSHCN, they also arrange and coordinate for health and health-related services. According to the National Survey of CSHCN, about 20.1 % of parents or caretakers reported having to spend more than 5 hours per week directly related to CSHCN’s care.²⁵ Because of the increased health care needs among this population of children and the associated burden on their parents and families, these families are often faced with the reality of one or both parents cutting back work. Cutting back or stopping work can also lead to a significant decrease in family income level as the result of lost wages, thus placing these families in a downward financial spiral.

Classically, one protective mechanism for families facing excessive medical expenditures from financial ruin is adequate insurance coverage.^{2–4, 26–27} For families with CSHCN, continuous insurance coverage plays a critical role in protecting their financial well-being.^17–19 Our results underscored the importance of insurance coverage for CSHCN and their families. Even among high-income families, lack of continuous insurance coverage appeared to place them at greater financial risk than poor families with coverage.

Other solutions than providing adequate insurance coverage to families of CSHCN may alleviate burdens experienced by families. Kuhlthau et al. showed that the development of well designed systems of care that incorporate new mechanisms of care delivery, such as in medical home settings, and efficient organization of services may provide a mechanism for alleviating financial burdens for families of CSHCN.¹⁶ Insurance should be viewed as only one element, albeit an important one, of a multi-faceted strategy to provide financial protection and support for all families of CSHCN.

Another interesting finding from our analyses was that the most vulnerable families with CSHCN were not the poor, but the near-poor and other low-income families. CSHCN families between 100–199 % FPL were even more likely than those below 100 % FPL to experience financial problems (Table 4). Other researchers studying the broader pediatric population have also demonstrated the vulnerability of families with just high enough income not to qualify for Medicaid or other federal assistance programs.^28–29

Our study has several limitations. First, because we analyzed cross-sectional data, we cannot determine causal directions. A clear example of this is in Table 4: as family income level decreased, parents were more likely to cut back or stop work. Our study cannot distinguish whether parents cutting back on work led to loss in income and therefore a drop in family income level or the burden of being poor and unable to hire help forced parents to cut back on work in order to care for their children. Breadwinners in poor and near poor families often rely on “lower-end” jobs with limited pay and fringe benefits, thus making it necessary for them to care for their own children rather than hiring someone else or sending their children to care centers. Second, our study results were based on self-reports. Because of issues related to cultural and reporting preferences of various racial groups, we have refrained from commenting on racial/ethnic disparities in this paper. Our goal was to examine the protective role of insurance coverage for families of CSHCN from various income groups, after adjusting for racial and ethnic differences/preferences. However, differential response biases may exist across income groups in reporting financial problems. Third, the difference between income groups in work participation may be related to the type and quality of job as well as benefits associated with it. In other words, parents of CSHCN who hold jobs with generous insurance benefits (and likely “higher-end” jobs) may choose to keep working in order to maintain their insurance coverage. Because the data set lacked details related to employment and benefits, we could not disentangle confounding effect related to employment.

Both the Federal government and the states are considering ways to limit Medicaid spending. Our work demonstrates that one in four poor and near poor families with a CSHCN already experience financial difficulties. Since Medicaid insures many CSHCN in low income families, increases in Medicaid copayments may place even greater financial burden on these families. In addition, the State Children’s Health Insurance Program (SCHIP) will expire by 2007 unless renewed. SCHIP has a main focus on families we found most vulnerable in this study, families with incomes above Medicaid limits but still too low to purchase adequate private coverage, discontinuation of SCHIP could have catastrophic financial implications for many modest income families with CSHCN.

CONCLUSION

Our results demonstrate the importance of health insurance for protecting families against adverse financial impacts. Even when insurance status is held constant, a substantial income gradient remains in financial impact, such that poor and near poor families were significantly more likely to experience burdensome health care expenses and reductions in employment than their wealthier counterparts.