New Findings from the 2007 National Survey of Children’s Health (NSCH)
June 02, 2009, 2:00 – 3:00 p.m. ET (1:00 – 2:00 p.m. CT, 12:00 – 1:00 p.m. MT, 11:00 – 12:00 p.m. PT)
Stephen Blumberg, Ph.D., Senior Scientist, National Center for Health Statistics
Christina Bethell, Ph.D., Founding Director, The Child and Adolescent Health Measurement Initiative
Michael Kogan, Director, Office of Data and Program Development at Maternal and Child Health Bureau
Renee Schwalberg, M.P.H., Maternal and Child Health Information Resource Center
Gretchen Noonan, M.S., Moderator, DataSpeak
Michael Kogan – Office of Data and Program Development at Maternal and Child Health Bureau – Director
Good afternoon, everybody, and welcome to today’s DataSpeak Web Conference on the newly released findings of the 2007 National Survey of Children’s Health. I’m Michael Kogan, and I’m the Director of the Office of Data and Program Development at the Maternal and Child Health Bureau. The DataSpeak series is sponsored through the office’s Maternal and Child Health Information Resource Center.
Today we’re pleased to present the first DataSpeak Web conference of the 2009 calendar year. Our first presenter today will be Dr. Stephen Blumberg, a senior scientist at the National Center for Health Statistics and lead statistician for the State and Local Area Integrated Telephone Survey, or SLAITS, which is the random-digit-dial survey mechanism used to conduct the NSCH. He’ll begin our discussion by describing the content of the survey, including methods used to collect data, response rates, and differences in design between the 2003 and 2007 surveys. I’ll serve as the second presenter today, and I’ll be discussing the purpose and potential uses for the survey as well as highlight some national- and State-level findings. To round out the discussion, Dr. Christina Bethell, Founding Director of the Child and Adolescent Health Measurement Initiative, will discuss how to access results of the survey through the MCHB-supported Data Resource Center for Child and Adolescent Health. Dr. Bethell will also present selected NSCH results at the State level.
It’s now my pleasure to introduce Gretchen Noonan, the moderator for today’s program. Gretchen, I’ll now turn the floor over to you.
Gretchen Noonan – DataSpeak – Moderator
Thank you, Michael.
First I would like to welcome all of our participants today. We have a very large audience, and we’re excited that you all could join us. Before we begin our presentations, I have some very brief technical guidance to everyone who’s listening. First, please know that your phone line will be muted during the presentations today. After the presenters have finished, we’ll have a question and answer session. At that time, you’ll have the opportunity to ask questions through the telephone operator, who will come on to provide instructions. Although you have to wait until the end of the program to ask questions on the phone, you can post questions online at any time. On left of your screen, there’s a Q&A pod. Simply enter your question in the white area there and just hit “Enter.”
We’d also like you to know that right below the Q&A pod, there is one called “Download.” You can download each of the presentations from today—the slides, that is. Please just click on the presentation that you wish to download, and then the “Save to My Computer” button will highlight, and you can click that.
Also, we would like you to know that if you’re logged in through the Internet—(ahem) excuse me—a feedback form will be provided at the end of the presentation. We would greatly appreciate it if you could take a moment to complete this brief form.
If you encounter any technical problems during the presentation today, please feel free to call the MCHIRC helpline. I don’t think that number’s on your screen, so if you want to write it down, it is 202-842-2000.
Finally, I would just like to quickly let everyone know about some features of the DataSpeak Web site, which we welcome to you visit after today’s program. For everyone’s convenience, additional resources on today’s topics have been posted on the Web site, including some of those that the speakers will highlight today during their presentations. You might also like to know the archives of last year’s DataSpeak programs, along with all of the other programs we’ve held since 2000, can be found on the Web site. And the address is there on the slide that you’re going to see—those previous events.
Now, without any further ado, I would like to turn to our first presenter, Dr. Stephen Blumberg of the National Center for Health Statistics. Thank you for joining us today, Stephen.
Stephen Blumberg, Ph.D. – National Center for Health Statistics – Senior Scientist
Good afternoon, Gretchen.
Gretchen Noonan – DataSpeak – Moderator
I think the best way to begin today, Stephen, is by giving our audience an overview of the National Survey of Children’s Health. What is its purpose, and when was this first conducted?
Stephen Blumberg, Ph.D. – National Center for Health Statistics – Senior Scientist
Well, I suspect that many of the people on the phone are familiar with the National Survey of Children’s Health from the first time that we did it back in 2003. The National Survey of Children’s Health is a survey that is sponsored by the Maternal and Child Health Bureau at HRSA, conducted by the National Center for Health Statistics at CDC. And the goal is to produce national and State-based estimates on the health and well-being of children, their families, and their communities. Now, as I’ve said, we conducted it for the first time in 2003. The 2007 survey that we’ll be talking about for the next hour is the second time that the National Survey of Children’s Health was conducted. The goal here was to assess changes in health and well-being over the past 4 years, but also to improve the measurement of some of the components of the survey: health care service needs, service use, access to a medical home…. We also expanded the section on neighborhood and community influences.
Gretchen Noonan – DataSpeak – Moderator
Okay. I understand that the National Survey of Children’s Health is part of the State and Local Area Integrated Telephone Survey, which I think Michael mentioned, and that’s called SLAITS. Can you give us a brief overview of recent SLAITS surveys and the interview process that’s used for those?
Stephen Blumberg, Ph.D. – National Center for Health Statistics – Senior Scientist
Sure. SLAITS, as you indicated, is the State and Local Area Integrated Telephone Survey. This is a survey mechanism; it’s a way of doing surveys. And a number of surveys have been conducted using this mechanism. Several of these are listed on the slide here—certainly the National Survey of Children’s Health that we’re talking about today. The other one that many people are familiar with is the National Survey of Children With Special Healthcare Needs. But we conducted the National Survey of Early Childhood Health back in 2000, the National Asthma Survey in 2003, and the National Survey of Adoptive Parents, which I’ll mention a little later on in my time today.
Now, as for the interview process, it’s really—the interview process changes for each SLAITS service. Let me focus specifically on the NSCH. This is a random-digit-dial telephone survey. Our samples are independent in each of the 50 States plus the District of Columbia. Now, as I said, random-digit-dial—that means we’re dialing telephone numbers not knowing who’s on the other end. We try to screen for, first, whether or not that number reaches a household, then whether that household has any children under 18 years of age. If so, that household is eligible for our survey. And in those households where there’s more than one child, we randomly select one to be the target of the interview. The interview itself lasted about 28 minutes on average. The respondent was a parent or guardian who is knowledgeable about the health of the child. This is most commonly the mother but in other cases father or grandparent. Our interviews were conducted in English, Spanish, and four Asian languages. And as many people are familiar, completing interviews, whether in person or on telephone, has gotten more and more difficult, as people are busier and busier in their lives. So we offered cash incentives to household with children who had not completed the interview, and about a third of our interviews were completed after offering an incentive.
The data were collected from April 5 of 2007 through—until July 27 of 2008, but 79 percent of the interviews were completed in 2007, which is why we call this the 2007 National Survey of Children’s Health. Our goal was to complete at least 1,700 interviews in each State, and we achieved that goal for a total of more than 91,000 interviews nationwide. The overall response rate was 47 percent. Obviously, it varies by State, and I certainly thank our friends in North Dakota for their willingness to participate. But I think I should mention that when we calculate a response rate, we calculate a relatively conservative response rate that takes into account the idea that there are likely eligible households among all those various numbers that ring with no answer or are busy or get picked up and hung up before anybody really says anything. So one of the—perhaps a better estimate of how well—how cooperative people were once we contacted them is the interview completion rate of 66 percent. So that’s the percentage of completed interviews among all those eligible households that we were able to contact.
Those who choose to access the data file and run their own data will find that there are sampling weights on the file. These weights permit national and State-specific estimates of the health and well-being of children. They’re adjusted for potential nonresponse biases, adjusted to account for the noncoverage of nontelephone households, and are further ranked to match census bureau population totals for various demographic groups. This whole weighting process helps us to account for the noncoverage of cell phone-only households. The 2007 NSCH was a landline-only survey, but, you know, we’re certainly moving to a world where we’re going to have to start bringing in the cell phone-only households in our future surveys.
Gretchen Noonan – DataSpeak – Moderator
It sounds that way. I think that that basic break on our methodology is really important when we’re interpreting some of the results from the survey, which we’ll hear about from Michael and Christy a little later. But Stephen, moving on to the content of the survey, could you tell us about some of the major categories or—I think you called them “domains”—of data collection?
Stephen Blumberg, Ph.D. – National Center for Health Statistics – Senior Scientist
Sure. When we talk about the NSCH, it really was designed to get at the idea of healthy children, healthy families, and healthy communities all working together towards positive childhood outcomes. So some of the unique features of the National Survey of Children’s Health is this comprehensive detail around child, family, and neighborhood, but also the inclusion of many positive indicators to track youth development: family strengths, family relationships, household routines, and other family processes.
Gretchen Noonan – DataSpeak – Moderator
Okay. And under each of these domains, I understand you have specific topic areas. Could you tell us about what those are for each domain?
Stephen Blumberg, Ph.D. – National Center for Health Statistics – Senior Scientist
Well, I’m not sure that I could get into everything. So I’ll give you just a little bit of a flavor of them—
Gretchen Noonan – DataSpeak – Moderator
Okay.
Stephen Blumberg, Ph.D. – National Center for Health Statistics – Senior Scientist
—included, you know. Certainly, the National Survey of Children’s Health is a very broad survey in terms of the topics. So health and functional status—we get at health status; we get at height and weight for children 10 and up. We use the CSHCN screener to identify whether any children have special health care needs. We ask about selected health conditions—their severity—focus a little bit on oral health problems and depression, as well as using the PEDS to assess risks for developmental delay for young children. We get at health insurance coverage, both current and past-year coverage, coverage by public health insurance programs and whether that health insurance is adequate…. We also look at utilization of preventive health care, preventive dental care—also mental health care and specialty care. For adolescents, we get at some vaccinations, tetanus boosters, HPV, meningitis…. For younger children, we get at whether or not they’ve received developmental screening from their pediatricians.
As I said, we expanded the section on medical home. It now is much more similar to the medical home section from the National Survey of Children With Special Health Care Needs, for those people who are familiar with that. Medical home here gets at usual place for care, personal doctor or nurse, whether the family’s receiving family-centered care, whether they’re satisfied with communication among the providers and between providers and other service needs, whether they’ve had problems with referrals or problems with care coordination or unmet needs for interpreters. There are questions that are specific to age for younger children. We get at childcare, breastfeeding, reading, telling stories, singing, going out, playing… for the older children, school attendance, afterschool activities, sleep, exercise, volunteerism, how they spend some of their time, and both positive and negative social behaviors.
As I said, we also get at family and neighborhood. So families, we look at shared meals and closeness to the child and stress—coping. We ask about the parents’ general health and mental health status, whether they exercise, whether anyone smokes in the child’s household…. And when we look at neighborhood characteristics, we get at neighborhood and community cohesion. We get the negative side of neighborhoods—litter, vandalism, run-down housing—as well as the positive side: sidewalks, parks, playground, libraries, rec centers…. We also ask parents about their perceptions of the child’s safety in the community and at school.
And of course, around all of this are questions about demographics so that you can examine any of these various indicators by the education level of parents; whether the parents of the child were born in the United States; how often the child moves; whether the parents are employed, their income; and the household’s program participation in welfare programs, food stamps, free or reduced-cost lunches, and so forth.
Gretchen Noonan – DataSpeak – Moderator
I only wish we had more time. It sounds like these are all really interesting topics, and you said that wasn’t even an exhaustive list. So that’s really interesting. Finally, you mentioned earlier that you are going to talk about the National Survey of Adoptive Parents very briefly, another SLAITS survey that was recently released. Would you like to go ahead with that?
Stephen Blumberg, Ph.D. – National Center for Health Statistics – Senior Scientist
Sure. I did want to mention the Department of Health and Human Services Office of the Assistant Secretary for Planning and Evaluation and the Administration for Children and Families recognized that as part of the National Survey of Children’s Health, we were going to be able to identify a number of adopted children. In fact, out of that 91,000-plus completed interviews, we identified 2,000 children who had been adopted. Never before has there been such a large nationally representative dataset of adopted children. And so the National Survey of Adoptive Parents was born. This is a follow-back survey to the National Survey of Children’s Health to look at the pre- and post-adoption experiences of the adoptive family. All the data from the National Survey of Adoptive Parents can be linked to the health and well-being data from the National Survey of Children’s Health to create a very rich dataset for adopted children.
Gretchen Noonan – DataSpeak – Moderator
Wonderful. Thank you so much, Stephen. Where can our audience go for more information on what you’ve talked about today?
Stephen Blumberg, Ph.D. – National Center for Health Statistics – Senior Scientist
Well, on this final slide, I’ve listed the Web site where you can get all the information in great detail about the National Survey of Children’s Health as well as download the datasets. If people have questions about the National Survey of Children’s Health, they can always e-mail us at slaits@cdc.gov. It’s important that I mentioned that—you know, remember the name “SLAITS,” because if you come into the CDC Web site—or the MCHS Web site, I should say—you’re going to need to follow the SLAITS links to ultimately to get to the National Survey of Children’s Health.
Gretchen Noonan – DataSpeak – Moderator
Okay, excellent. Thank you, Stephen.
Now with that background information in mind, I’d like to reintroduce Dr. Michael Kogan from the Maternal and Child Health Bureau. He will continue our discussion today by providing us with some of the initial findings from the survey. Hi, Michael. Are you still there?
Michael Kogan – Office of Data and Program Development at Maternal and Child Health Bureau – Director
Hi, Gretchen.
Gretchen Noonan – DataSpeak – Moderator
All right. Michael, I think that Stephen mentioned this, but it’s really a basic, important point that’s, I think, worth repeating: What is the purpose of the National Survey of Children’s Health, and can you give us some examples to begin with of how the information can be used?
Michael Kogan – Office of Data and Program Development at Maternal and Child Health Bureau – Director
Sure. Thanks, Gretchen. Just to reiterate, the purpose of the—not just the National Survey of Children’s Health but also the National Survey of Children With Special Health Care Needs was to produce national and State-based estimates on children’s health. This is—you might have gone the flavor from Stephen’s presentation that this is one of the few areas—it’s one of the few surveys where we can get State-level estimates. And the reason we did that is it was—one of the original motivations for the survey was to help States obtain data for their Title V Needs Assessments that they need to work on every year. Since then, these surveys have been integral and used in planning and program development at the State level. One thing you can obviously use this survey for where data are comparable are to compare changes since the 2003 National Survey of Children’s Health. And I’ll be giving some examples of comparisons to 2003. And if the 2003 survey is any indication, there was—these surveys can be used for scientific research. A number of articles have come out of those surveys and—as well as chartbooks. There’s a special issue of Pediatrics devoted just to findings on the 2003 NSCH.
And so, let me move on to some of the findings. I’m going to try and touch on a number of areas in our limited time.
Now, first, looking at children’s overall health status, we see that most children are reported to be in either excellent or very good health—about 84 percent of kids. This is essentially unchanged from 2003. However, only about 70 percent of U.S. children were reported to be in excellent or very good oral health.
As Stephen mentioned, we asked about a number of chronic conditions. These are the most prevalent chronic conditions among U.S. children, with asthma being the most prevalent. And these are parents reporting that they had ever been told by a health care professional that their child had the condition and the parents reporting that the child currently had the condition as well. As you can see, looking down the list, almost 8 percent of kids were reported to have learning disabilities; over—about 6.5 percent of kids had attention deficit disorder or attention deficit hyperactivity disorder—on down the list.
Gretchen Noonan – DataSpeak – Moderator
Okay. And as Stephen mentioned, Michael, information was collected about health insurance status among children. And what were some of the findings in this area?
Michael Kogan – Office of Data and Program Development at Maternal and Child Health Bureau – Director
Well, what we’ve found is that about 9 percent of the kids at the time of the survey were currently uninsured. Now, if you just look at current underin– current uninsurance, it’s not a complete picture of the situation. So we also asked, “Was there a period within the prior year where the child was uninsured part or all the time?” When you take those two together, you found that about 15 percent of kids in the U.S. had a period without insurance coverage or were currently uninsured. And I should also mention that here we’re comparing 2003 to 2007. As you can see, a very slight increase in the number of uninsured kids—also a slight increase in the percent of kids who were receiving their coverage through public insurance. Let me mention that we did a similar comparison using the 2001 and 2005–2006 Children with Special Health Care Needs Survey. Now, when we looked at that population, we found there was a large shift towards public insurance. There was almost a 33 percent increase in that population over the 4- to 5-year period.
Further, in both Stephen’s talk and mine, we mentioned you can get State-level estimates. And here’s an example of that: If you look at current health insurance coverage, you’d see fairly wide variation among the States, all the way from Massachusetts, which recently instituted coverage on—health insurance coverage for—only about 2 ½ percent of kids are uninsured, compared to some States where about 20 percent of kids were uninsured at the time of the survey.
Now, for the first—unlike 2003, we added questions not just on whether the kids were insured or uninsured, but also we looked at adequacy of insurance, particularly among kids who were currently insured. To be considered adequately insured, the parent had to report that the child usually or always had benefits or services that met the child’s needs, that the child could see a provider if he or she needed, and that out-of-pocket costs were reasonable. If they met all those three criteria, the child was considered to have adequate insurance. As you can see from this slide, about a quarter of children would have inadequate insurance according to these criteria.
Gretchen Noonan – DataSpeak – Moderator
Okay. So it seems, Michael, that what might naturally follow a discussion of health insurance status is a discussion of children’s use of health services. What does the survey reveal about children’s health care utilization?
Michael Kogan – Office of Data and Program Development at Maternal and Child Health Bureau – Director
Okay. Well, when we look at preventive health and dental care, we see that almost 90 percent of kids in the U.S. were reported to receive preventive health care in the past year. Now, in terms of preventive dental care, the percentage is less: About 78 percent of kids were reported to receive preventive dental care. And when we compared 2003 to 2007, we see a large increase in their percent of kids who had received preventive medical visits in the past year, going from 79 percent to about 89 percent. This could be due to more emphasis on prevention. This is something that needs to be explored further. However, we also—when looking at mental care—mental health care, what we found is that among kids who were reported to have emotional developmental or behavioral problems, about 40 percent did not receive any mental health services in the past year.
We also looked at medical homes. Stephen has already mentioned that. The American Academy of Pediatrics defines “medical home” as medical care that’s accessible, continuous, comprehensive, family centered, coordinating, compassionate, and culturally effective. And here we see wide variations by race and ethnicity among children with the medical home, ranging from 68 percent of White children down to 38.5 percent of children of Hispanic ethnicity.
Gretchen Noonan – DataSpeak – Moderator
Okay. Now we’re going to switch gears just a little bit. And I think that Stephen mentioned very briefly in his long list that breastfeeding is one of the things that the survey covered. So what did the survey reveal about—I believe you measured initiation and duration of breastfeeding?
Michael Kogan – Office of Data and Program Development at Maternal and Child Health Bureau – Director
Yes. What we found is that 75 ½ percent of mothers with children in the U.S. were reported to have ever breastfed. This is a slight increase from 2003: In 2003, 72.3 percent of children were reported to have ever been breastfed.
Now, we added a new question for 2007 focusing on exclusive breastfeeding. And we did that because the American Academy of Pediatrics guidelines recommend 6 months of exclusive breastfeeding in the first 6 months of life. And when you look it that way, you see that we have a long way to go to meet that standard: Only about 12 ½ percent of children were exclusively breastfed in the first 6 months.
As another example of State variation in children ever breastfed, you see a very wide range here, ranging from almost 90 percent of kids in some States who were reported have ever been breastfed down to a little over 50 percent. Now, the good news is, we looked at this in 2003 as well, and among the States that had the lowest prevalence, this year there was a decent jump from 2003. For example, in Louisiana in 2003, only 45 percent of women were reported to—children were reported to have ever been breastfed. In West Virginia, it was 53 percent; now it’s 57. In Kentucky it went from 55 percent to 58 percent.
Gretchen Noonan – DataSpeak – Moderator
Well, that’s great. Stephen mentioned, Michael, that the survey also collects information or—excuse me—collected information on family-level influences. Can you discuss some of those?
Michael Kogan – Office of Data and Program Development at Maternal and Child Health Bureau – Director
Sure. Here we’re looking at screen time, and we found that among preschool children, over half of kids were reported to have more than 1 hour a day of screen time. Moving on to reading, singing, and telling stories to young children, we find that less than half of young children in the U.S. were read to everyday, but about 60 percent are sung to or told stories to every day.
We also look at parental health status and found that for U.S. children, 21 percent of their mothers were reported to be in neither excellent or very good physical or mental health. And for fathers, the figure was 16.1 percent.
We also looked at smoking in the household and found that over one-quarter of U.S. children lived in households with a smoker. As you can see also, there was a very strong socioeconomic gradient as to the likelihood that the child lived in a household—lived in households with a smoker. Again, looking at State variation in household smoking, we see almost a fourfold increase according to where the child lived, ranging from just over 10 percent all the way up to almost 40 percent.
We also looked at some educational experiences of the child. Overall, about—around 10 ½ percent of kids were reported to have ever repeated a grade. Again, you see a very strong SES gradient here, with the greatest likelihood of repeating a grade among those with less than 100 percent of the Federal poverty level.
We also looked at childcare experiences and found that, as you can see, about 30 percent of families experienced one or more childcare problems. When you break it down like family structure, as you can guess, those single-parent families, particularly single-mother families with no fathers present, were most likely to experience childcare problems.
Gretchen Noonan – DataSpeak – Moderator
Okay. And finally, to wrap up, perhaps you could talk about some of the neighborhood factors that Stephen had mentioned earlier.
Michael Kogan – Office of Data and Program Development at Maternal and Child Health Bureau – Director
Okay. First we look at neighborhood amenities and found that while there are only about 4 ½–5 percent of kids living in neighborhoods with none of these four amenities listed, there were a large percentage of children living in communities without at least one of these amenities. You see on this slide that only 65 percent of kids lived in a community or neighborhood where there was a recreational center or a community center, and only about 80 percent of kids lived in neighborhoods where there was a park or a playground.
We also looked at neighborhood conditions, as Stephen mentioned, and found that about 30 percent of U.S. kids live in neighborhoods with at least one of the conditions of either litter or garbage on the street or sidewalk; poorly kept or dilapidated housing; or vandalism, such as broken windows or graffiti.
When—we asked about safety of the child in the neighborhood and found that about half of U.S. children’s parents report that their neighborhood always felt safe for their children. And again, looking at the right-hand side of the slide, we see a very strong socioeconomic gradient, with the lower family income—the less likely they are to report safety in the neighborhood.
I’d like to acknowledge Renee Schwalberg and Joan Jordan at the Altarum Institute in this presentation. And this is my contact information if you have any questions about the survey.
Gretchen Noonan – DataSpeak – Moderator
Okay. Thank you so much, Michael. And if I could just interrupt for one moment, I’d like to just apologize to some folks on our teleph—in our telephone audience. We have, unfortunately, in an unprecedented move here, reached our online capacity or exceeded it, so some folks who have dialed in are not able to the presentations online. And we apologize for that, but they are going to be available on the DataSpeak Web site if you’d like to take a look at them after the program today.
Now, if I could, I’d like to introduce our final presenter, Dr. Chrina—Christina Bethell—excuse me—who is the Founding Director of the Child and Adolescent Health Measurement Initiative. Welcome, Christy.
Christina Bethell, Ph.D. – The Child and Adolescent Health Measurement Initiative – Founding Director
Hi.
Gretchen Noonan – DataSpeak – Moderator
Hi. Are you all set with your slides, Christy?
Christina Bethell, Ph.D. – The Child and Adolescent Health Measurement Initiative – Founding Director
I am.
Gretchen Noonan – DataSpeak – Moderator
Okay, great. I understand that the MCHB has supported you and your colleagues—you call it CAHMI, right?
Christina Bethell, Ph.D. – The Child and Adolescent Health Measurement Initiative – Founding Director
Yep.
Gretchen Noonan – DataSpeak – Moderator
Okay—that MCHB has supported CAHMI in developing and running a Data Resource Center for accessing data from the National Survey of Children’s Health. But before we discuss how to use this online tool, I was hoping you could give us some background on the Data Resource Center itself and its goals.
Christina Bethell, Ph.D. – The Child and Adolescent Health Measurement Initiative – Founding Director
Definitely. So the Data Resource Center came about, starting in about 2002–2003, in response to the need to expand and expedite as well as simplify access the findings from the National Survey of Children’s Health and other—and the National Survey of Children With Special Health Care Needs, which is actually the survey that the Data Resource Center began working with. The next slide shows of a picture of the homepage for the Data Resource Center, and the URL is childhealthdata.org.
The primary goal of the Data Resource Center is to provide a centralized resource for accessing data from the National Center—National Survey of Children’s Health as well as the National Survey of Children With Special Health Care Needs. And by providing this user-friendly and interactive Web-based resource and standardized indicators from these datasets, what we hope to do is build common knowledge and capacity for using the data to advance evidence-based program evaluation, policy, and advocacy and really fast-track the access to these data.
The Data Resource Center includes two portals. One is focused on the National Survey of Children’s Health, which we’re discussing today. Both the 2003 and the 2007 datasets are available for querying online there. And then the other portal is focused on the National Survey of Children With Special Health Care Needs, which also includes both years of that dataset as well.
Gretchen Noonan – DataSpeak – Moderator
Okay. It sounds like a really wonderful resource, and I have no doubt that everyone’s going to want to know more about this. So can you tell us about some of specific capabilities or features that are available through the Web site?
Christina Bethell, Ph.D. – The Child and Adolescent Health Measurement Initiative – Founding Director
Yeah, I’ll do that quickly, and then we’ll go into more detail later. But the basic features are to learn about the surveys using some of our quick glance guides and informational background about the surveys or contents, data collection, and so on; being able to search the data in that by subgroups and across States; getting resources or learning more about how people are using the information; and also getting help if there is additional information or assistance needed in using the Web site.
[Pause] All right, somebody just walk into my office. I apologize.
So in addition to the four primary features users can order and download datasets that have all of the variables constructed in there as well as codebooks associated with them. They can sign up for e-updates and access latest publications that we try to keep track of and link people to and that use the National Survey data.
And there are four types of user-generated out—data reports that people can get from here—from the Data Resource Center: the State profile tables, which provide a snapshot of a number of indicators across States; there’s the interactive all-State comparison feature and data report, which compare all States on a single indicator by subgroups or just overall; there’s State-ranking maps, which show visually how States rank, as well as tables that indicate whether their differences or significantly different from the national rates; and then the interactive graphics tables, which go into great depth on an indicator-by-indicator basis.
Gretchen Noonan – DataSpeak – Moderator
Great. I’m sure that our audience now would like to know more about how you can actually conduct the searches of the data. So can you tell us about the different ways that you can start navigating through the site?
Christina Bethell, Ph.D. – The Child and Adolescent Health Measurement Initiative – Founding Director
Yes. So on the next slide, you’ll see there are two different ways to began searching data; the first is to do a full search from the—go to the Child Health Data homepage or the NSCH portal page, and the second is to use the quick-search feature on the Data Resource Center homepage.
This next slide—you can—shows how you can do a step-by-step full search by first picking the National Survey of Children’s Health to select the search, and then you pick the survey you’re interested in—2007 or 2003—and then you pick a general search area that you’re interested in—Child Health Measures—or you can go right to the detailed survey section. In the next, you select from the list of topics and survey items and indicators that are available to search in that area.
If you already know the topic you want to find information about, you can conduct a faster quick search from the Data Resource Center homepage. And to do this, you would select the survey and the year in the little box that’s there in the middle; the geographic regions that you’re interested in, which could be as a nation, all States at one time, or a specific State; and then the topic of interest. And then what you’ll get through right there is fast track to the specific topics that are available in that search area.
Gretchen Noonan – DataSpeak – Moderator
Okay. Could you give us a quick walk through how the tool works? I understand you have some examples of how to get State-specific profiles of the data—or of the data findings.
Christina Bethell, Ph.D. – The Child and Adolescent Health Measurement Initiative – Founding Director
Yeah. So here we have the next slide on getting—how to get a State profile. So in addition to the full search and quick search, you can begin viewing your State summary profile. There are two steps in getting a State profile. First you click on the map of the U.S., and that map of the U.S. is on each of the portals as well as the homepage. Then you select the State that you’re interested in; here we’re selecting Alaska. And then it will give you a profile that shows the State-specific results on key indicators that are selected from the 2007 National Survey of Children’s Health. And this feature also allows users to select another State that they may want to compare their State to, up at the right. You can toggle comparing any two geographic areas. And they can also choose to compare their State to the Nation or the overall region. And finally, there is a State profile option to compare your State findings from the 2003 National Survey of Children’s Health to the 2007 National Survey of Children’s Health, especially for, of course, those indicators that stay the same across those years.
Gretchen Noonan – DataSpeak – Moderator
Okay. And you mentioned that the tool can be used to obtain data comparing States on a particular indicator. Can you tell what you mean by this and then just show us how that’s done?
Christina Bethell, Ph.D. – The Child and Adolescent Health Measurement Initiative – Founding Director
Yes. So you’ll see on the next slide, you begin by selecting the survey here on the “Search the Dataset” feature—that little middle puzzle piece on the homepage of the National Survey of Children’s Health. And then you select among the broad topical search areas, and we’re selecting “Child Health Measure.” Because we’re interested in comparing all States, as shown, select “All States” as the graphic area of interest. And then the topic of interest that I’m picking here is “Health Care Access and Quality.” So on the next page, you’ll see the specific questions for which you would like to view the data. And here I’m going to select “Medical Home,” the topic lots of people are interested in.
So using the “All States” as the geographic area and selecting “Medical Home,” you get something that looks a little bit this. It’s cropped to be able to fit on the slide. The slide illustrates the data output you receive, and here you see that States ranked—are ranked according to the proportion of children receiving health care that meet criteria for having a medical home. And then circled in red are findings sorted from highest to lowest. As you can see, the range across States is, for this survey right now, 45.5 percent for Nevada to 69.3 in New Hampshire. And it’s really important to recognize the power of this “All States” feature, because the data were sampled and collected in the same way, as Stephen described, across all States. It’s possible to take variations across States more seriously and really ask questions about why those variations exist.
Gretchen Noonan – DataSpeak – Moderator
Okay. And I realize we’re running out of time, but I’d like to move on and ask you about key performance measures. Is there a way you can obtain State rankings for those?
Christina Bethell, Ph.D. – The Child and Adolescent Health Measurement Initiative – Founding Director
Yeah, absolutely. First of all, you can rank States from the “All States” comparison tool to the above. In addition, you can get a bird’s-eye picture view of how States rank by selecting a State-ranking map from the indicator feature in the State profile section, such as “Developmental Screening,” “Smoking in the Household,” “Obesity and Weight,” “Adequacy of Insurance,” and so on. To view the—to view and download a map, select the State-ranking map from the highlight box on the special—or from the “Special Topics” menu. And it’s not on this slide, but there are two ways you can get it: either going under “Special Topics” or just going right to the map to highlight. And then you pick the map that you’re interested in—I’m picking “Smoking in the Household”—and then an interactive map will appear. This map allows you to select any State to view the State-specific results from—and a box will pop up to show you what the results are. And you can also download a PDF of the map, which includes a bonus data table showing State scores and ranks and their statistical significance from the national rate.
Gretchen Noonan – DataSpeak – Moderator
Great. I was interested to learn that the DRC tool allows users to develop tables and graphs based on their own questions or data needs. And could you walk us through those three steps that are involved and show us some examples?
Christina Bethell, Ph.D. – The Child and Adolescent Health Measurement Initiative – Founding Director
Yeah, absolutely. So perhaps the most powerful feature of the Data Resource Center is that it does allow users to do a customized and interactive search on any survey item or constructed indicator from the National Survey of Children’s Health. So in the same way that we access an “All States” comparison table, we begin by selecting the “Search the Dataset” feature just in the middle of the NSCH portal homepage. And then, as shown earlier, the broad topic of interest that we’re going to select here is “Child Health Measures.” From here, instead of selecting all States, you select the geographic area of interest—I’m just sticking with “Nationwide” as our example here—and the topic—again, I’m picking “Health Care Access and Quality.” So you’ll see that I’m looking at “Developmental Screening,” and the next two slides illustrate the type of tables and graphs you receive from this interactive data query search. So here the data table gives the estimate just weighted to rep—be representative of the State population as a Nation; it even explains. In this case, we have 19.5 percent having a standardized developmental screen. Confidence interval or margins of error are also provided, along with the actual sample size that was the data regenerated from and the sample size weighted to represent all children. From here, you can then interactively explore the findings for any State by selecting “Compare State Region” up at the right-hand side. You’ll see you can compare States with—compare whatever you—State or region you started with with any other State or region and you can also compare across a number of subgroups.
The next slide shows a graphical display of the data output. And again, you can right-click and copy and paste all data output from here as you need and put it into PowerPoint, Word, or other documents.
In terms of subgroup—a lot of people are interested in that—here’s a comprehensive list of the subgroups that you can currently select from—stratified survey item indicators from the 2007 NSCH. And it includes age and sex, race/ethnicity, primary household language, household income—two versions of that, family structure, special health care needs status, type of insurance, and consistency of insurance. We’ll be adding some others and welcome people to recommend those that they’d like to see added.
Gretchen Noonan – DataSpeak – Moderator
Okay. Great. And I know you’ll give some contact information later, so if they have ideas about that, they’re welcome to get a hold of that staff of the DRC. Michael and Stephen both discussed some ways that the data from NSCH can be applied, and I’m wondering if you have any suggestions for how folks can use the data retrieved through the DRC.
Christina Bethell, Ph.D. – The Child and Adolescent Health Measurement Initiative – Founding Director
I mean, the obvious way people use it is to collect findings, use them to gather people together in groups, stimulate partnerships and discussion and improvements—also a lot of use for writing grants, doing research, and so on. But in terms of any other tips I would give for looking for data, of course, var—we all know variation is information, so people should be looking for interesting variation by State, population subgroup, and now we can [inaudible] by surveyors. So here’s an illustration of interesting findings from the data using Indicator 1.3, which is children 0–5 who have been breastfed or fed breast milk for any length of time. Michael discussed that indicator earlier. So as we already saw, there’s almost a 37-point range across States, the highest prevalence occurring in the Pacific Coast States. We also see a variation across demographic subgroups, and in this case with a 40-point spread between the highest and lowest income groups, for example, in Mississippi. Further, since this question on breastfeeding was asked in the 2003 survey, we noticed variation by year. For—here’s an example: Delaware was 43rd out of that—States in breastfeeding in 2003 and have now jumped to 34th. So there’s a lot of different ways to start looking at variation.
And in terms of State variation in general, on key indicators that are highlighted in State profile from the National Survey of Children’s Health, we see a 6.6 percentage point difference across States for variables “In School” all the way up to 48.7 percentage point spread for “Attends Religious Services”; that’s the spread that we’re seeing. And this range will vary once again as soon as you look at any indicator via subgroup, such as race, ethnicity, household income, and so on. I really encourage people to sort of unpeel the onion and use the Data Resource Center to learn at a much more indepth level about the data findings.
Gretchen Noonan – DataSpeak – Moderator
Okay. And you have there a form for requesting technical assistance; is that—?
Christina Bethell, Ph.D. – The Child and Adolescent Health Measurement Initiative – Founding Director
Yeah. So we provide technical assistance for those who need more information about how to use the Data Resource Center or who want more help understanding these data. You can just fill out our “Ask a Question” feature right on the Data Resource Center Web site, and we respond pretty quickly. And I really encourage you to visit the homepage childhealthdata.org or write to the NSCH data portal and also to follow us on Twitter. We’re doing a lot of Twittering these days, where we discuss a lot of new updates and interact with our users in more detail on specific topics. So if you’re interested in following us on Twitter, that’d be great.
Gretchen Noonan – DataSpeak – Moderator
Great. And I know it can seem kind of overwhelming for folks, but I’ve used your Web site, and it’s really wonderful, and it’s very user friendly. So I would encourage people certainly to check it out for—on their own.
Christina Bethell, Ph.D. – The Child and Adolescent Health Measurement Initiative – Founding Director
Well, I just want to express my gratitude to the Maternal and Child Health Bureau for their vision and leadership in allowing us to develop this, as well as our collaborators at NCHS and, of course, the 17 people within in the CAHMI team who, at some level, have worked on this.
Gretchen Noonan – DataSpeak – Moderator
Great, all right. Christy, thank you, and I would like to go ahead and just thank all of our presenters again for their excellent presentations today and announce that we’re now in the question and answer portion of our program. We’re fortunate that all three of our presenters were able to remain with us. And as I mentioned at beginning, we’ll be taking questions both online and on the telephone. And I’d briefly like to just ask each of our presenters if they are available to stay for an extra 5 or 10 minutes today. Stephen and Michael and Christy, are you able to do that?
Michael Kogan – Office of Data and Program Development at Maternal and Child Health Bureau – Director
Not a problem, Gretchen.
Gretchen Noonan – DataSpeak – Moderator
Okay. Well, I’d just like to let our audience know we might be able to go a little over our 3 o’clock ending time if all of our presenters are able to stay with us and that we have gotten a large number of questions, and anything that’s submitted online and not answered today, we will get answers for you from the presenters, and we’ll be posting all of that on the DataSpeak Web site. So if your question isn’t answered, you will see it within the next few weeks on the Web site.
I would just like to ask our operator, Ryan, if he could come on and just let our telephone participants know how to ans—excuse me—ask a question on the phone.
Ryan – Operator
Sure. Ladies and gentlemen, if you would like to ask a question by phone, you can do so by pressing star-1 on your touchtone phone. A confirmation tone will indicate that your line is in the question queue. For participants using speaker equipment, it may be necessary to pick up the handset before pressing the star keys. So once again, that’s star-1 to ask a question by phone.
And we have a couple question from the phone coming in. Our first phone question comes from the line of Angela Mickalide with Home Safety Council.
Angela Mickalide, Ph.D., CHES – Home Safety Council
Thank you very much. Excellent presentations, all. I’m wondering if any of this data capture injuries to children. I see that there are some that address special health care needs, but are those are congenital, or are they acquired through unintentional injury?
Michael Kogan – Office of Data and Program Development at Maternal and Child Health Bureau – Director
This is Michael—it’s Michael Kogan. And yes, we do have some questions. We do have a couple questions about injuries. And this is focused on kids from 6 months to 5 years old. And we just—we ask if, during the last 12 months, the child has been injured and required medical attention. And then we ask if the injury occurred at home, at childcare, or some other place.
Angela Mickalide, Ph.D., CHES – Home Safety Council
Thank you.
Gretchen Noonan – DataSpeak – Moderator
Okay. Thank you. Ryan, do we have anyone else on the phone right now?
Ryan – Operator
Yes. Our next question comes from Devaiah Muccatira with North Dakota Department of Health.
Devaiah Muccatira – North Dakota Department of Health
Hi, Michael.
Michael Kogan – Office of Data and Program Development at Maternal and Child Health Bureau – Director
Hi, Devaiah.
Devaiah Muccatira – North Dakota Department of Health
I have a question. Is there is a component in the Survey of National Children’s Health where you can compare the component of special health care needs with the Special Health Care Needs Survey of 2006? There is a component in the survey when you compare between Children’s Health Survey and the National Children’s Health Survey for Special Health Care Needs. Can we compare those two between the surveys and within—between the special health care needs surveys?
Michael Kogan – Office of Data and Program Development at Maternal and Child Health Bureau – Director
Okay. Let me answer in a couple parts, and Stephen may want to add some things, too. You can—where questions are comparable, you can certainly compare between the two special health care needs surveys. And in this survey, yes, we do screen for kids with special health care needs, and you can compare findings on kids with special health care needs to kids without special health care needs. Now, in terms to your question, whether you can compare kids with special health care needs in this survey to kids with special health care needs in the Children With Special Health Care Needs Survey, remember a lot of the questions are going to be different. On some questions, you might be able to make a comparison, but remember, the sampling is different for the two surveys.
Devaiah Muccatira – North Dakota Department of Health
Okay. The numbers are different—the sample numbers itself.
Michael Kogan – Office of Data and Program Development at Maternal and Child Health Bureau – Director
Well, the sampling scheme is different.
Devaiah Muccatira – North Dakota Department of Health
Different, okay. Okay. Okay, it’s better to do it within rather than between.
Michael Kogan – Office of Data and Program Development at Maternal and Child Health Bureau – Director
Yes, in this case.
Devaiah Muccatira – North Dakota Department of Health
Okay. Thank you.
Gretchen Noonan – DataSpeak – Moderator
Okay. If we could go to the online questions, my colleague Renee Schwalberg of the MCHIRC has been monitoring these questions as they came through. Renee, do we have some questions from our audience?
Renee Schwalberg, M.P.H. – Maternal and Child Health Information Resource Center
Sure. Let’s start with a question from Sharon Hutchens, probably for Stephen. Is the child’s ZIP Code or city of residence available to make comparisons within States?
Stephen Blumberg, Ph.D. – National Center for Health Statistics – Senior Scientist
We do collect ZIP Code; however, due to confidentiality reasons, we don’t release ZIP Code on our publicly available datasets. So anybody who has an interest in doing such State analyses would have to go to a research data center in order to access the confidential data files.
Renee Schwalberg, M.P.H. – Maternal and Child Health Information Resource Center
Great. Thanks. Here’s one that actually might go to Christy. Elizabeth Mosel asks how to filter responses to try to get at the specific population of children with special health care needs.
Christina Bethell, Ph.D. – The Child and Adolescent Health Measurement Initiative – Founding Director
Okay. What you can do is, if you go into the—search the data and wind your way through selecting your geographic area and topic of interest, then variable that you want to look at—once you do that, you will see up at the right-hand side a “Compare Subgroups” button. You click on it, and you will see the opportunities for a dropdown menu that will allow you to select what subgroups you want to pick from, and special health care needs status is one of those subgroups. And select that, and your data will appear stratified by special health care needs status.
Renee Schwalberg, M.P.H. – Maternal and Child Health Information Resource Center
Thank you. Here’s another one for Stephen. A couple of people asked a version of this question. This one came from Francine Goodrich. As a public health nurse, she’s finding that a majority of families she sees no longer have landlines but rather just use cell phones. Did the phone calls use cell phone numbers, and if not, how does this change the survey sample?
Stephen Blumberg, Ph.D. – National Center for Health Statistics – Senior Scientist
Well, it’s a good question. In 2007, we know that about 12–13 percent of children were living in households that had only cell phone communications. This obviously has the potential to add bias to the survey; however, we make a number of weighting adjustments to census control totals. We make a number of weighting adjustments for nonresponse that, to large extent, reduce the potential for nonresponse and noncoverage bias as a result of the cell phone population. That said, there is always the possibility that you’re going to run into some variable that is remarkably different between kids who live in cell phone households and kids who live in landline households that is not accounted for by demographics. Now recognize how I say that. It’s not a matter of whether their parents are younger or so forth, because we’ve got adjustments for characteristics such as that, but it really requires something uniquely different about children in cell phone only households. And to date, we have yet to concretely identify what those are. That said, the prevalence has been growing dramatically. We’re now up to between 18 and 20 percent of children living in cell-only households, and so you’ll see, for instance, with the upcoming 2009–2010 National Survey of Children with Special Health Care Needs that were making efforts to start bringing cell phone-only children into the survey.
Renee Schwalberg, M.P.H. – Maternal and Child Health Information Resource Center
Great. Thank you. [Inaudible]—
Gretchen Noonan – DataSpeak – Moderator
Renee—oh, I’m sorry; no—do you mind if we turn to the phone for a moment?
Renee Schwalberg, M.P.H. – Maternal and Child Health Information Resource Center
That’s fine. I can go on forever.
Gretchen Noonan – DataSpeak – Moderator
Okay. I saw we have a very long list. So Ryan, do we have anyone else on the phone?
Ryan – Operator
Yes. Our next question comes from the line of Douglas Rittenhower with the State of Arizona
Gretchen Noonan – DataSpeak – Moderator
Douglas, are you there?
Douglas Rittenhower – State of Arizona
Can you hear me?
Gretchen Noonan – DataSpeak – Moderator
Yes, we can hear you. Go right ahead.
Douglas Rittenhower – State of Arizona
Okay. This—my question is for Dr. Kogan. Hi, Michael. I wanted to ask you about the use of any questions that got at the issue of medical debt as a barrier to you—health insurance status, coverage, or utilization of care? Are there questions that got at medical debt as a possible barrier?
Michael Kogan – Office of Data and Program Development at Maternal and Child Health Bureau – Director
That’s a wonderful question, Doug. Probably the closest question that would come to it is when I was talking about questions we asked about adequacy of health insurance. And we ask parents, “Does the insurance—are the—is the insurance adequate to meet your out-of-pocket costs, or are they unreasonable?” And remem—about 20 percent of parents reported that the costs were not usually or always reasonable.
Douglas Rittenhower – State of Arizona
Thank you.
Gretchen Noonan – DataSpeak – Moderator
Okay. Ryan, would you like to bring someone else on from the phone?
Ryan – Operator
Sure. Our next question comes from the line of Mandala Reisan with Insights Direct.
Mandala Reisan – Insights Direct
Yes, hi. You had mentioned that the PEDS is available in the survey, and I was—and I had a question to—I’ve also seen the—something called the PHDS, which is the Promoting Healthy Development… that was, you know… I was curious to find out what is it—the similarity or difference between those two.
Christina Bethell, Ph.D. – The Child and Adolescent Health Measurement Initiative – Founding Director
I probably can answer that. This is Christy. The CAHMI has developed the Promoting Healthy Development Survey and has collaborated with MCHB and NCHS on collecting some items from it that are relevant to include or be adopted for purposes of the national surveys. This main overlap here is asking about parental concerns, whether parents have concerns and whether those concerns have been addressed, as well as items about developmental screening—[inaudible] developmental screening. And then, it’s not from the PHDS, although included in it is the parents’ evaluation of developmental status, which actually is developed by Francis Glasgow. And that’s—there’s an overlap there with the PHDS as well.
Mandala Reisan – Insights Direct
Oh, there is. Okay. So there’s a pretty substantial overlap between the two.
Christina Bethell, Ph.D. – The Child and Adolescent Health Measurement Initiative – Founding Director
In terms of looking at developmental screening and developmental problems, yes.
Mandala Reisan – Insights Direct
And the parental concerns as well.
Christina Bethell, Ph.D. – The Child and Adolescent Health Measurement Initiative – Founding Director
Exactly, but not all the anticipatory guidance and family assessments.
Mandala Reisan – Insights Direct
Got it. Okay.
Gretchen Noonan – DataSpeak – Moderator
Great. Ryan, we’ll take one more person from our phone queue right now.
Ryan – Operator
Okay. Our next question comes from the line of Trina England with Maternal and Child Health.
Trina England – Maternal and Child Health Bureau
Hi, everybody. My question has to do with whether it’s going to be possible just to go around when the data are reported to be able to take a look at developmental indices. So for example, like, could there be an index so that—looking at vulnerable neighborhoods or strengths of neighborhoods ,so that the different variable in it, rather than reporting the frequencies individually, could be combined in some way, because clearly we have—these variables are very much related.
Stephen Blumberg, Ph.D. – National Center for Health Statistics – Senior Scientist
I mean, I can take that. We report all of the data from those questions. We would certainly encourage any data analysts to look at ways of combining those questions into indices of positive or negative neighborhood environments. But to my knowledge, nobody has yet, in the week or 2 weeks that the data has been available—has put such an index together.
Trina England – Maternal and Child Health Bureau
Because I know that you had done a really nice job in creating those indices when the 2003 survey first came out, but I don’t think that those—that way of combining the variables was ever posted in a public way. And I think it would have been really helpful to people to do so.
Christina Bethell, Ph.D. – The Child and Adolescent Health Measurement Initiative – Founding Director
I don’t know if it’s useful, but on the Data Resource Center, there are pop-up boxes for each indicator outlining how the variables are constructed. And then there’s codebooks, too, that describe it—describe how each variable and indicator is constructed.
Trina England – Maternal and Child Health Bureau
Well, for example, I mean, like, I think it would be really interesting, you know, to take a look, you know—this is off the top of head—parents’ perceived safety of the neighborhood for their children in the amount of graffiti and, you know—and other kind of, you know, refuse lying around in the neighborhood. I mean, those two things probably go together.
Michael Kogan – Office of Data and Program Development at Maternal and Child Health Bureau – Director
Uh-huh, they probably do.
Gretchen Noonan – DataSpeak – Moderator
Okay. If—may we move on? Renee, do we have any more online questions that you’d like to—?
Renee Schwalberg, M.P.H. – Maternal and Child Health Information Resource Center
We certainly do. Here’s one that should be pretty quickly answered. Christy, Geralyn Brennan in Oregon asks if the CAHMI Web sites have the ability to run cross-tabs.
Christina Bethell, Ph.D. – The Child and Adolescent Health Measurement Initiative – Founding Director
Yeah, that’s—one of the main purposes for the Data Resource Center is to allow you to look at your State’s four and then select subgroups that you’d like to view that four by. And I mentioned that there’s four set of demographic variables that you can look at, as well as some health status and insurance-type variables: insurance status and type of insurance and special health care needs status and so on. So when you go in to look for a data, you’ll see a button at the top and then query tool that says “Compare Subgroups.” And you can click that, and then you can get a dropdown menu where you can look at it in a stratified way.
Renee Schwalberg, M.P.H. – Maternal and Child Health Information Resource Center
Thank you. Michael, Karen Stiff asked for a little more information about the insurance adequacy variable. Is that measure new for 2007, and how is it derived?
Michael Kogan – Office of Data and Program Development at Maternal and Child Health Bureau – Director
Okay. Again, a very good question. It’s new on the National Survey of Children’s Health; however, it has been on the National Survey of Children With Special Health Care Needs, both in the 2001 survey and in the 2005–2006 survey. And there are a number of ways to try to define adequacy of insurance. I mean, some survey, such as Medical Care Expenditure Survey, might define it in terms of a percentage of income spent on health care in relation to total family income. We didn’t have that kind of detail. We focused more on the parents’ perceptions of what inadequate insurance was and different aspects of that.
Renee Schwalberg, M.P.H. – Maternal and Child Health Information Resource Center
Thanks. Here’s a question that’s, I guess, for Stephen. Maya in Washington State asks, “Would analysis of a subset of children age 10–14 within one State be appropriate, given the sample size and the sampling strategy?”
Stephen Blumberg, Ph.D. – National Center for Health Statistics – Senior Scientist
Yes, I would think that it could—that it certainly would be. You know, you’ve got 2,000 or about 1,700, at least, completed interviews per State. Ten to 14 is, well, what, about 30 percent of children. So you’ve got a sample of about 500 kids in that age range. You know, you may not be able to break that down by race and income looking at various indicators, but you certainly could look at indicators within that group.
Renee Schwalberg, M.P.H. – Maternal and Child Health Information Resource Center – Project Director
Great. Thanks. Aaron Hagan asks why some indicators provide information for children from—of all ages 0–17 while others are limited to children ages 10–17. How did data collection or analysis differ or the sample sizes different for these indicators?
Stephen Blumberg, Ph.D. – National Center for Health Statistics – Senior Scientist
Generally, the choice of age range for questions is based on what we felt was most appropriate for the topic of the questions. So, you know, we don’t ask about volunteerism from—for instance, for infants. But generally, analysis for kids 10–17 for a particular indicator that was asked only of that group will be as reliable and valid as any other indicator, recognizing that sample size is going to be a bit smaller. You know, it’ll be, you know, what, a little less than half of the total sample size. And so, therefore, you know, you may not be able to do as fine a detail on cross-tabs at the State level as one might hope, but you should be able to get good information from it nevertheless.
Renee Schwalberg, M.P.H. – Maternal and Child Health Information Resource Center
Thank you. Do we have time for one more?
Gretchen Noonan – DataSpeak – Moderator
I think we can do one more, Renee.
Renee Schwalberg, M.P.H. – Maternal and Child Health Information Resource Center
Okay. This is a particularly juicy one. How reliable are height and weight data for children, particularly for adolescents?
Michael Kogan – Office of Data and Program Development at Maternal and Child Health Bureau – Director
Well, we did some analysis from—somebody on our staff, Dr. Gopal Singh, did some analysis comparing findings in our surveys to the National Health and Examination Nutrition Survey, where they get actual weight and height measurements. And we found that parent report was pretty close to what you got on the National Health and Examination Nutrition Survey from the ages of 10–17 onward, which is why we limited that variable on the public use dataset to those age ranges.
Stephen Blumberg, Ph.D. – National Center for Health Statistics – Senior Scientist
Similar results were recently published by Lara Akinbami and Cynthia Ogden and in the journal Obesity—I think they were put online in February—demonstrating that for younger children less than 10 years of age, parents were unreliable in their reports of the children’s height, which obviously, you know, will affect their calculations of BMI, but that reports for adolescents were generally fairly reliable.
Gretchen Noonan – DataSpeak – Moderator
Great. Thank you. I’m afraid that that’s all the time we have for discussion today. As we mentioned, any questions that have come in online and not been answered will be posted to DataSpeak Web site in the next several weeks, I believe. And if you have any more questions that you didn’t submit yet, you can send them to the e-mail address on the screen there—it’s mchirc@altarum.org—and we’ll respond to the question along with the archives. Aand this program will be available on the Web site, so you access it at your inconvenience. That will go up in the next few weeks.
And before you go, we’d like to let you know that we’ll be broadcasting several more DataSpeak programs in the coming months, and you will receive announcement via e-mail about those. And these announcements will also be posted on the DataSpeak Web site, so you can check there for more information.
And finally, before you log out, we’d like to ask you to take a moment to provide us with some feedback on the program. You can do so by clicking on the “Program Evaluation” link on the screen that’s going to show up, and the short survey will open in a new window. And we’d like to thank you for joining us today, and the program is adjourned.