Responses to the following questions provided by Bonnie Strickland of the MCHB:
Q: In between the years that the SLAITS is done nationally, is it expected that the Title V programs will collect the same or similar data on their program enrollees?
A: I believe that Dr. van Dyck responded to this on the call. His response was that there is no expectation that states should collect similar data in between the years that the Survey is conducted, although they can if they wish. The in-between years should be used to implement change strategies for impacting the status of the measures by the next Survey date.
Q: According to Dr. van Dyck's presentation on slide #9, 17.4% of CSHCN need specialized therapies. Was there any indication in the data regarding the number that are using and/or are dependent on technology. Any age breakdowns related to this issue?
A: Among all CSHCN,
the weighted number and percent of CSHCN needing mobility devices = 407,506 (4.4%);
the weighted number and percent of CSHCN needing communication aids = 182,267 (2.0%);
the weighted number and percent of CSHCN needing other medical equipment = 988,025 (10.6%).Among CSHCN needing specialized therapies,
the weighted number and percent of CSHCN needing mobility devices = 322,704 (6.8%);
the weighted number and percent of CSHCN needing communication aids = 146,661 (3.1%);
the weighted number and percent of CSHCN needing other medical equipment = 663,342 (14.0%).Among CSHCN needing specialized therapies, ages 0-5
the weighted number and percent of CSHCN needing mobility devices = 36,234 (3.6%);
the weighted number and percent of CSHCN needing communication aids = 23,446 (2.3%);
the weighted number and percent of CSHCN needing other medical equipment = 253,609 (25.0%).
Among CSHCN needing specialized therapies, ages 6-11
the weighted number and percent of CSHCN needing mobility devices = 102,546 (5.8%);
the weighted number and percent of CSHCN needing communication aids = 60,733 (3.4%);
the weighted number and percent of CSHCN needing other medical equipment = 216,764 (12.2%).Among CSHCN needing specialized therapies, ages 12-17
the weighted number and percent of CSHCN needing mobility devices = 183,924 (9.4%);
the weighted number and percent of CSHCN needing communication aids = 62,482 (3.2%);
the weighted number and percent of CSHCN needing other medical equipment = 191,814 (9.8%).
Q: When the survey information differs from data we have for our actual Title V/CSHCN enrollees, should we just address this in the narrative of our Title V Block Grant application/report? For example, we have a strong transition program so that the outcome for our enrollees would be considerably better than the outcome from the SLAITS sample.
A: Dr. van Dyck responded to this question on the call. If states have different or additional data they should report/discuss it in the narrative.
Q: Can you discuss how child care and out of school time care fit into the Core Outcomes for coordinating community systems of care and if questioning regarding these needs will fit into futrue CSHCN surveys?
A: Child care is part of measure #5, community-based services will be organized so that families can use them easily. The Survey did not include a question on access to/unmet need for child care, so this aspect of measure #5 will have to be monitored using other measures at this time. The new National Survey of Children’s Health will provide information on this aspect of the community system for CSHCN, and other sources of information will be used to monitor this component.
Q: Will specific questions about access to and quality of child care and out of school time care be included in next two surveys under composite questions related to Core Outcome 5 (related to community systems of care being organized so families can easily use them)?
A: This issue has not been discussed, but will be addressed along with other suggestions when the survey revisions are planned.
Q: Do you have a list of state contacts for this survey?
A: The performance measures were sent to the State MCH and CSHCN Directors by Cassie Lauver, Director of the Division of State and Community Health (DSCH). Stephanie McDaniel at AMCHP has a list of state data contacts.
Q: Given the definition of CSHCN (chronic condition), how would the condition "never" affect a child's activities?
A: CSHCN are defined as children with a chronic condition who have a health care service need beyond that required by children generally. The chronic condition and/or the service need do not necessarily affect the child’s ability to do things that other children his or her age can do.
Q: Any plans to integrate FACCT screener into the Youth Risk Behavior Surveillance efforts?
A: No plans as yet. But the screener has been integrated into several other surveys including the new National Survey of Children’s Health, the Medical Expenditure Panel Survey (MEPS), and the Consumer Assessment of Health Plans Survey (CAHPS), as well as several state administered surveys.
Q: With regard to the "annual preventive dental visits," when is it hoped that these will begin?
A: Dr. Kogan identified unmet dental needs as the most prevalent unmet needs identified by the National Survey of Children with Special Health Care Needs. However, there is not yet a specific MCHB initiative to begin “annual preventive dental visits for CSHCN.
Q: Is it a problem when 2 different core outcome measures use the identical subscale items - specifically, outcomes #1 and #2 both use "Doctors usually or always make the family feel like a partner"?
A: Using the same subscale items to monitor 2 different core outcomes is not a problem from a monitoring perspective. The selected subscale items are only indicators. The Survey contains multiple additional items that can be used to explain and differentiate between the two. Additional measures will also be used to monitor each performance measure.
Q: Goal #4 appears to lack the specificity one would want for screening for CSHCN. Wouldn’t it be more helpful to ask about developmental screening, since most children with SHCN have developmental disabilities? A behavioral screen could also be promoted to identify those with behavioral disorders.
A: Goal #4 is intended to include developmental/behavioral screening, two very major aspects of early and continuous screening for which we have little data. The CSHCN Survey does not address screening, therefore we will use other sources, such as the new National Survey of Children’s Health, and other national data sources to monitor this goal.
Q: Could Dr. Newacheck please address the limitations of the monitoring approach?
A: To the extent that the monitoring approach relies on data from the National Survey of Children with Special Health Care Needs, limitations are evident in our ability to monitor Goal #4 (screening) and Goal #6 (youth transition). The survey, as currently constructed, does not lend itself to gathering accurate screening information from parents of CSHCN, so other measures will be used to monitor this goal. With regard to adolescent transition, the survey provides information only for youth under 18 years of age, and not from youth themselves. In addition, this core outcome was restricted to CSHCN aged 14 years and older. Finally, all of the survey data are family reports. Although this approach makes an important contribution to our thinking about how to monitor measures focused on improving the service system for families, MCHB realizes that that consumers are not the only source of information. Therefore MCHB will assess progress from other perspectives as well.
Q: Do you have any ideas as to why are we doing so poorly as a nation in achieving Goal #6 for teens with SHCN?
A: The data from the National Survey of Children with Special Health Care Needs on youth transition confirms existing data and reflects an ongoing weakness in health care and other components of the community system for youth with special needs. Reasons may include: (1) lack of adult health care providers adequately trained to provide ongoing care for youth and adults with special health care needs; (2) low expectations and/or lack of training on the part of pediatric providers, educators, employers; and (3) our poor performance in health care for youth/adolescents in general.
Responses to the following questions provided by Michael Kogan:
Q: Can you please repeat the information on how to obtain a chart book?
A: Copies of the CSHCN Databook will be available in early July 2003. Please call the
HRSA Information Center at 1-888-275-4772 to order a copy when available.
Q: Will MCHB or NCHS make detailed information on replicating the graphs in Dr. van Dyck's presentation? Will MCHB or NCHS provide an avenue for researchers at the state level to easily share their work?
A: All of the graphs presented will be in the chartbook. There are various avenues for researchers at the state level to share their work. First, they can probably use the MCH Epi listserv. Second, there are sessions planned at the MCH epi conference. Third, we are working on having a journal supplement for the state-level findings. At MCHB’s request, NCHS can prepare and release a SAS program to enable other analysts to replicate the chartbook estimates.
Q: When can we in the states expect assistance in developing state-specific data?
A: We are working towards having a web site that can be queried at the Foundation for Accountability’s web site and are also discussing the possibility of offering further training in SUDAAN at the MCH Epi conference. We have also provided the states with the performance measures for their states, and there will be state-specific data in the chartbook being published. We (MCHB and NCHS) are also available to answer specific questions.
Q: For states to report the MCH report measures, will the components be aggregated into the measure at the national level? Will states have the ability or need to do this reporting?
A: We have generated each performance measure and its component parts for every state. We will also populate that information into their block grant reports.
Q: If this data is used for the Title V Block Grant performance measures, will the data be provided to the states annually for reporting for the Title V Block Grant?
A: We plan to conduct this survey every four years. In the intervening years, we don’t expect states to generate new data.
Q: If families with ONLY private insurance have fewer unmet needs, is this a reflection of: (1) They have higher incomes so they can purchase care out of pocket if it isn't covered, thereby meeting needs better, or (2) Their children have fewer needs, or both -- Rather than indicating that private insurance ONLY has the best range of coverage of the needs?
A: This is a good question that can only be answered with further analysis. I would suspect that it’s both reasons.
To best answer this question, one would use a Poisson regression model to predict number of unmet needs from type of insurance coverage, income, and out-of-pocket expenses, including the total number of needs as an offset variable. Such an analysis was beyond the scope of these presentations.
Q: How does one get the state-level estimates for core outcomes data for the states, or at least for Michigan and the US?
A: They were sent to every State MCH and CSHCN director. Please check with them. If they don’t have it, then contact your regional representative in our Division of State and Community Health.
State-level estimates for the core outcomes are now available on the NCHS web site for the National Survey of CSHCN. Go to:
and click on the link for “Progress toward implementing community-based systems of services for children with special health care needs: Summary tables from the National Survey of Children with Special Health Care Needs, 2001.”
Q: Is MCHB going to use these data for analysis of health disparities (e.g. rural/urban, by region of country, by race/ethnicity)?
A: Yes. You will already find some of that work when the chartbook is published.
Responses to the following questions provided by NCHS:
Q: What denominator did you use to define unmet need: all CSHCN or children whose parent expressed a need?
A: The denominator used in the chartbook was all CSHCN.
Q: In slide #10 of Dr. van Dyck's presentation, how many children had 2, 3, 4, or 5 of the special needs shown?
A: The unweighted distribution of CSHCN by number of special needs:
1 SHCN = 20,073 (51.7%)
2 SHCN = 8,932 (23.0%)
3 SHCN = 5,798 (14.9%)
4 SHCN = 2,785 (7.2%)
5 SHCN = 1,278 (3.3%)The weighted distribution of CSHCN by number of special needs:
1 SHCN = 4,874,298 (52.1%)
2 SHCN = 2,092,761 (22.4%)
3 SHCN = 1,393,756 (14.9%)
4 SHCN = 690,183 (7.4%)
5 SHCN = 309,356 (3.3%)
Q: If a very small number of people was actually asked a question, how do we handle that in using the data? The transition information, particularly, had very small numbers.
A: Certain questions were asked for only a small number of children because (a) the question was added to the questionnaire late in the data collection period, and/or (b) the question applied to only a subpopulation of CSHCN. When the former was true, all children for whom interviews were completed prior to the addition of the question were dropped from all analyses of the item. When the latter was true, analyses of that question were restricted to the subpopulation of interest. Furthermore, standard errors were carefully considered. Small sample sizes often yield large standard errors, and relative standard errors greater than 30% were considered unreliable.
Q: Where can state-level data from the survey be accessed?
A: The raw microdata sets are available on-line at:
Summary statistics at the state level will be available in the chartbook that will be released by MCHB. Summary tables concerning the core outcomes at the state level are available on the NCHS website for the National Survey of CSHCN. Go to:
and click on the link for “Progress toward implementing community-based systems of services for children with special health care needs: Summary tables from the National Survey of Children with Special Health Care Needs, 2001.”
Q: If the incidence of CSHCN is 12.8% but 20% of households have a child with SHCN, can we infer that families with CSHCN have fewer children?
A: This inference is inappropriate. For example, if we assume that special health care needs are not clustered in households, the probability that one child has special needs is independent of whether or not his/her brother or sister has special needs. If this assumption were true, then households with more children in total would have a greater likelihood of having at least one child with special needs. In fact, an analysis of the total number of children in households with children showed that households that have a child with SHCN have more children (mean = 2.14) than households that do not have a child with SHCN (mean = 1.82).
Q: Regarding families who indicated in the survey that they had to reduce work hours or quit work, holding severity categories constant, do you know how that varied by income or by type of insurance coverage (i.e., for Medicaid-enrolled families)?
A: To best answer this question, one would use a regression model to predict the impact on work from severity, income, and type of insurance coverage. Such an analysis was beyond the scope of these presentations.
Q: It seems odd that Hispanic children's prevalence of special health care needs was so low. Would you explain what may contribute to this? Could this be due to sampling? Would you consider oversampling some groups in the future surveys?
A: Hispanic children were not oversampled. Within each state, Hispanic children were sampled at the same rate that they are present in the population. Oversampling would not have changed the prevalence estimates. Reasons why the prevalence for Hispanic children is lower than the prevalence for non-Hispanic children merit further investigation. It should be noted that the impact of the special needs identified among Hispanic children were slightly greater than the impact of the special needs identified among non-Hispanic White and non-Hispanic Black children (28.3% of Hispanic CSHCN were affected usually, always, or a great deal, compared with 21.4% of non-Hispanic white CSHCN and 25.5% of non-Hispanic black CSHCN). One hypothesis that requires further investigation is that the CSHCN Screener is not as sensitive to less bothersome special needs among the Hispanic population. Remember, however, that an alternative hypothesis is that Hispanic children do in fact have a lower prevalence of special needs.
Q: Have you finished the data analysis for the SLAIT survey?
A: Yes, the information presented by Dr. van Dyck discusses these findings.
Microdata sets are available on the NCHS website so that all interested researchers may continue to analyze the data from the National Survey of CSHCN. The raw microdata sets are available on-line at:
In addition, the health insurance analysis file and related analyses have not yet been released. We expect that this file and these analyses will be released later this summer.