Ninth Annual Maternal and Child Health Epidemiology Conference / December 10-12, 2003
JODI DRISKO: Thank you. Good morning everyone. I think Marion did such a great job giving the overview on what they’ve done in California that I have almost nothing to talk about now. But we used some very similar methods and I think we went about doing the survey a little differently. And what I want to make the disclaimer upfront before we get too far into the presentation is that this is really a development process and I’m going to really focus on how we did it, who we involved, why we did it, how we plan to use the data, because at this point we’ve pilot tested it for one month and then we repiloted some problem questions which I’ll also address. But I don’t have a lot of data to show.
I will show you some pilot data, but it’s just one month of data, so it’s obviously not generalizable. But I’ll talk a little bit about that. But this is really more about how you can do quite a large effort with no money. What I think we all need in these budget cutting times. And to tell you just a little bit about Colorado is we are one of those states who have not enjoyed big budgets at this time. In the health department we have received cuts in our general fund money. I am really lucky that in our division, in the health statistics section we do not have any general fund money, so we haven’t been cut, but at the same time, it’s not like we can ask for money either. So we had to be a little creative about how we tried to fund this effort which is a pretty large effort. why did we need a child health survey anyway? We have a vision of data across the whole life cycle of cradle to grave, other people have called womb to tomb and like with women’s health there are no current surveillance systems to address the health of children under 14 in Colorado. We have PRAMS, that’s perinatal data, moms and newborns. And then we have YRBS for 14 to 18 year olds. We have BRFSS for 18 years and older for the adult population and we have our birth data and our death data, but we have this huge gap and really don’t know what happens with kids.
And we just found out a few months ago that we are the worst state in immunization rates, so there’s a real need to know what’s going on with families and what’s going on with children. And additionally we’re really gearing up into the National Public Health Performance Standards kind of movement and the essential public health service functions. So primarily this survey, to address those, will go under that monitoring area to monitor health status and identify community health problems. And secondarily it will evaluate--has the potential to evaluate effectiveness, accessibility and quality of personal and population based health services. We want to be able to in the future do research and hopefully come up with some innovative solutions because we might actually have prevalence estimates on some of these health problems that right now we really don’t have good data on. And then potentially to plan policies and plans for individual and community health efforts overall. So what is this child health survey? I think other states have child health surveys, SLATES has a big child health survey as well, but basically it’s a surveillance system to monitor the health and risk behaviors of children. And our target age group is one to 14 year olds and we got a group of interested parties, some of the likely suspects, some not so likely and the advisory group decided on this age range.
We didn’t necessarily want this age range, we wanted a little more concentrated. We were trying to push for the five to 12 year olds, shorter band, we’d have better power with the data. But they really wanted this one to 14 year age group and since they were paying the money, we kind of had to do what they wanted. So we came up with some core content areas about the major predictors of death and disability in children and we set up some optional modules and questions. We came up with a whole list of questions that we thought were kind of good questions and then let the people that we invited to the table decide if they wanted those questions, if they had better questions, just exactly how we could meet everybody’s needs, but still be very scientifically valid and at the same time have other states or national comparisons for at least a core of the data. So thinking from a planning perspective as you’re developing something like this, I always love when people plan, you know, I need all of this data and then--I do a lot of program evaluation, so it’s like, well, how are you going to use the data? So that was the first question that I’m asking the group. How are we going to use these data besides this essential services piece, which are really bigger topic areas, but exactly what are we going to do?
What do you want to see happen in the future with the data? So we decided that we would monitor trends in child health, I mean most of this stuff is pretty basic, kind of no brainer stuff, we’d be able to estimate the prevalence of specific health conditions and risk factors associated with these health conditions. And then the one thing that we’re really in to, we have a turning point initiative and they’re really into looking at health disparities and differences between subgroups in the population, so we really want to be able to look at that and that health disparities issue. And then of course to monitor progress towards the Healthy People 2010 goals, have our MCH section be able to better answer some of their grant requirement indicators, because for a lot of the indicators we don’t have really good data. And SLATES we know is coming and it’s going to be great, but we aren’t going to be able to have state level data as often as we want it.
So I went through a pretty comprehensive literature review and instrument review as well. So I looked into this CHQ, child health questionnaire, the Iowa Hawkeye child health survey. If any of you have a child health survey and you’re not on this list it’s just ‘cause I couldn’t easily find it. But if you do have one I would really love to hear about it and your experiences with it. New South Wales in Australia, their child health survey is amazing, it’s really pretty long and they get about 14,000 completes a year in a state that’s probably the size of Colorado, population-wise at least. So I looked up new federalism, child health survey from the Urban Institute, SLATES of course, the health survey for England is also another big one, but they have an adult and a child questionnaire and children fill out a lot of the questions themselves. And then the Utah Child Health Survey since a lot of times in Colorado we compare ourselves a little bit to Utah. We don’t have quite exactly the same ratio ethnic breakdown, but geographically we’re similar. We’re right next to each other, 80 percent of our population lives in this Front Range corridor, right by the mountains so we have mountain and frontier area so we kind of use them and they had one, which is great and we know their people pretty well. And then I looked at the American Academy of Pediatrics and HRQ for recommendations on what they think needs to be looked at in the areas of child health. So I came up with the core content areas that we started with. And demographics we looked at, including income, who they’re living with, who’s their primary care giver, race ethnicity, gender; of course, and a few others. And then access to healthcare and dental.
Basic health status and health behavior questions. Safety and injury. Mental health and behavioral health questions and then the identification of children with special healthcare needs with that special healthcare needs screener that Rohini spoke about, we used the same one. So we decided that since we didn’t have any money and we were basically having to beg all our partners to fund this effort, that we would do it as an add on to BRFSS. We have our own in-house survey research unit, so we do PRAMS and BRFSS and a lot of other surveys for internal partners at the health department already and we thought that being an add on to BRFSS would save us money because BRFSS--what cut the largest expense in survey research over the phone is getting the respondent on the phone and actually answer the phone and agree to talk to you. So by doing this add on we would save all that expense so we could charge less money for each question. So basically, after we get a respondent on the phone who has graciously agreed to do the BRFSS, we ask them if they have children in the home that are one to 14 years old. One question that we already know through BRFSS is if they have anyone in the house under 18, so we know they have kids or not, but we don’t know how old they are because we don’t ask that.
So then if they do have a child of that age range living in the house, we ask if we can call them back. And amazingly enough people actually agree to this, I mean, they’ve just completed about a 20--22 minute BRFSS, and then they submit to more survey interviews about a week later to the tune of about another 16 minutes. And if they do have children in the house of 1 to 14 years old, and if they have more than one, then we randomly select a child so we aren’t--we’re trying to minimize bias with all of this. So then one of our telephone interviewers call back about a week or two later. So this has been a collaborative effort with the health department but with other people as well. So our health department programs are obviously the MCH section and they have actually just given us a chunk of money and say, you know, sponsor, I think 20 questions. And then oral health and the injury, they’ve been really involved. And then a lot of the chronic disease programs. And I’ve got to tell you, if you have a lot of chronic disease programs and you’re interested in an effort like this, get them involved because they’re the people with all the money right now. And tobacco, well, in our state they used to have money, but they don’t really anymore. And the immunization program, HCP has sponsored all the special healthcare needs screeners.
And then we have Colorado State University involved, which is about 60 miles north and they want some injury epi-type questions. And the health science center, which is in Denver, their child development folks were interested in some behavioral health questions. And then the preventive medicine department is just kind of interested in some ongoing stuff and also through that preventive medicine department there is somebody in environment health who has been coming and we’re trying to figure out how we can add that piece because I think someone over here mentioned that it’s often overlooked and we still haven’t come up with any great questions, but it’s always in the back of our mind as what we can do. We have a question, I think on lead right now, which really isn’t a huge issue in Colorado, but we want to do something and the lead program was interested. So we have a prevention research center and they are helping out with some of the physical activity and nutrition questions as well as chronic disease, they’re sharing that because we have about 14 physical activity and nutrition questions.
And the Colorado department of Ed has just come on board, we just got the health department and the Colorado Department of Education just received a comprehensive school health grant, so I’ve been trying to put a little bug in their ear about maybe you want to ask some questions about what parents think about health education and what they think about vending machines in schools and things like that that will help you drive policy if you have data instead of saying we need to get vending machines out of school, let’s back it up with some data about X percent of parents think that we should do this or change nutrition in the schools so they have bought into that. And then our (inaudible) at the state has some really interesting questions about driving with your child in the car after you’ve had a drink or five drinks or using illegal drugs. So those are kind of interesting. And then the Colorado Children’s Campaign is a local foundation. They don’t have a lot of money, but they do a lot of policy, they do the kids count for Colorado each year, so they’ve been involved in helping out. And then Kaiser, in Colorado they just made too much money this year and their non profit status may be in jeopardy, so they need to give out a lot of money and we were a lucky beneficiary of some of that money.
So we can help to pay for some of our analysis, which we were a little worried about how we were going to pay for everything. So what we--the scheme we devised to get this done, since it was unfunded, is that we had to charge people for questions like they did in California. Since we were tacking it on to BRFSS and let BRFSS do a lot of the sample management for us, we could charge a little less than the $2,000. When you said $2,000 I heard a lot of uh behind me, and around in the room. But we charge about that for people to add questions to BRFSS every year, depending on how much CDC gives us, we have to increase or decrease the funding, the cost per question. So we decided that we would charge $1,200 for the first question, because we don’t think it’s fair for people to have to pay for the demographic questions because everybody’s going to want the demographic data and everyone’s going to want different cross tabs by different demographics. So we thought everyone could help defray the cost of that and then $1,000 for each additional question. And we didn’t just make this up off the top of our head, we knew it would cost less than BRFSS, and we do a lot of other survey research, so we know approximately how much it’s going to cost. So we figured the whole effort would probably cost between $80,000 and $100,000 a year at first just to do the survey and do very basic analysis. We are offering all the data to our partners. We will in our section do some of the analysis, but we’re not going to do really in depth analysis.
So then we held multiple meetings of all of our partners and they, as I said before, they decided on the target age range and then the group decided how, it was really interesting because when the question came up as, well, you know, a lot of us are interested in physical activity and nutrition. How? It’s not fair for all of us to have to pay for those questions. And we were kind of like, hum, yeah, you’re right, but we didn’t really know how to answer that question because we were just hoping that people would come up and be interested and want to give us the money to do this. So they decided among themselves and we just kind of stayed out of it, that they would all try and partner people who had like interests. So we helped say, okay, you and you and you are interested in injury. You four are interested in that. You meet, decide exactly what you want and get back to us. So they formed their own subgroups and either did conference calls or email or met, but most--it was kind of interesting because they didn’t necessarily already know each other and they were all interested in the same issues. So the reason why we thought it would cost so much is because this is everything we’ve had to do simplistically to get this going. So for the survey development, lit review, and question development, and our section has done most of this kind of in kind, pilot testing, the programming of our Caddy system which takes a while and training our interviewers, collecting the data, cleaning it, doing analysis and then just in the pilot phase then we have to do--we ended up having to repilot and revise the questionnaire so then we have to revise our Caddy programming once again. And then as we go into January in just a few weeks, which is a little nerve racking, we’ll start our data collection phase.
So at this point we need to translate it into Spanish, retrain the interviewers because the instrument’s changed a little. And then we need to do more quality assurance monitoring, which we do anyway. We monitor our interviewers about 10 percent of their completes over the phone to make sure that they’re reading the questions as they’re supposed to be read and not kind of going out on little tangents. And just hearing what people’s responses are and how they code, some real basic stuff. Then dating planing, we’re going to have to weight the data and then analysis a report generation and also one thing I forgot to mention is when we have that BRFSS respondent on the phone, when we ask if we can call back, we also ask if we can link their BRFSS data to the data we get from the child health survey. And then when we call back, we ask again because we may be talking to a different person. But we want to make sure that our bases are covered and we wrote that into our IRB, which we did get this approved by our IRB, so hopefully we can link parent behavior and child behavior with this enriched BRFSS child health survey data.
So we piloted it in July of this year and we just finished piloting a much, much shorter version, more of our problem questions. And then we’re going live in January and we hope to collect our data annually. All of our partners have really bought in and have agreed to just write in their grant every year X thousand dollars for child health survey questions. So it’s worked out really well, a lot of our chronic disease programs just recently had CDC site visits, so I spoke to all of them about the child health survey and they all seemed to be pretty happy that we were going to be able to collect this data. So now, just to switch gears, here’s our pilot results. So we had 98 eligible respondents. And each month in BRFSS we get about 330 completed interviews. So out of those 330 we had 98 eligible respondents, which means they had at least one child in their house between the ages of 1 and 14. So we had 10 refusals and they were all kind of interesting.
Some of them, well, I didn’t want my wife to complete that first survey in the first place and I don’t want to talk to you people ever again. And we didn’t have the pilot instrument in Spanish because it was going to take a really long time to translate it and just for one month of data collection. Our main questions when we piloted it was, were people even going to want to do this after completing a 20 minute BRFSS? How much can you really push people to get them to talk on the phone about their health or their children’s health? So we had 10 people that we could never contact them again. We call people up to 15 times and vary day, weekend, and evening attempts. And then we had 72 completed surveys. And no one refused to have their data linked to BRFSS. So here is just the age breakdown of the selected child. So when we just randomly selected one child. And it’s just one month of data so I’m not too worried about it at this point in time. It looks a little low on the five to six year olds and maybe a little high on the 14--13 year olds. But I think after collecting months and months of data that will even out.
Our race ethnicity and gender, shockingly, the gender looks almost exactly like our state gender does for kids in this age group, which we were really happy about that. And it shows that random digit dialing really does work. And the race and ethnicity, the Hispanic is exactly our state percentage for kids this age and also African American is the same. And the only differences here is that white non Hispanic is slightly under represented in child health survey and Asian, American Indian and other is slightly over represented, which is fine with us. That’s great. So just for a little sneak peak at what some of the data looks like, 18 percent of kids drink pop seven or more times a week, and 24 percent eat fast food three or more times a week. And a third eat less than seven servings of vegetables per week. And this is per week. But this is one of the problems. These are a few questions that we’re redoing in the pilot as well.
And then 29 percent of kids watch three or more hours of TV everyday. Yes?
UNIDENTIFIED SPEAKER: People who are answering this survey are all adults or (inaudible)?
JODI DRISKO: Yes, the adults are answering about their children. So, yeah?
UNIDENTIFIED SPEAKER: (Inaudible)?
JODI DRISKO: No, the adult is answering for their kid who is between the ages of 1 and 14. And depending on how old the kid is, we ask different questions for different age groups. We’re not going to ask a parent of a one year old how often their child bikes to school or anything like that. But yeah, so the parents are answering for their child. Or actually, it’s not--the question we ask is who knows the most about the health of the child in the household and that’s who we select to answer the survey. So only 10 percent of kids have a health status of good, fair or poor, which means that 90 percent have excellent or very good self reported health status. Seven percent have no healthcare coverage, which is a little surprising, we thought it’d be a little higher, but we have about 13 percent of people in this state without healthcare coverage. Fifteen percent have asthma and 19 percent are using some sort of prescription medication.
Fifty-four percent have had a sunburn in the last year, although, 61 percent always, or nearly always, use sunscreen. And then we ask this one question, this was on the UK survey, do you think a tan makes a child look healthy and 21 percent think a tan does make a child look healthy. And you may think that’s kind of a stupid question, but it’s a barrier to get people to use sunscreen if they think a tan makes a child look healthy. So I think that’s it for our pilot data. As I mentioned, we are repiloting a few questions and I can tell you the one that we’ve had the hardest time with is height and weight. We are trying to get BMI for kids, I don’t know how many of you know, but in Colorado we’re supposedly, through BRFSS, we are the leanest state in the country, but we really want to look at, but our trend is ever so increasing, so we wanted to look at what’s going on with kids and this height and weight is just really hard. The weight seems like it might be okay, but I can’t tell you how many two foot four-year olds there are in Colorado. So we’re repiloting the height and weight and trying to ask those in a series of five different questions to see which one or which two or three in combination we can get a little better data. Kaiser is thinking that they may give us money and growth charts to send to parents and we could get a more accurate measure, but we don’t know these people’s names or their addresses, so asking them that is a whole different barrier because this is an anonymous survey and we don’t have those identifiers.
And we’re also changing all of, or we are at least piloting to see if it made a difference, a lot of these food, and these different kinds of foods, how many fruits do you eat, vegetables, soda pop, fast food, with the fruit and vegetable there were clumps around 7 and 14. So we aren’t sure if people--we want to stay away from serving size, how many servings did they take because kids of different age there’s a different serving size and we don’t want our interviewers to have to explain, for a kid your age, a serving is this. So we’re changing it to a BRFSS question where they can say four times a day, three times a week, twice a month, once a year, whatever. So we’ve changed those questions and we’re seeing how the pilot data comes back from that. But overall we asked a couple questions at the end of the survey during the pilot, if there were any questions that people thought were too sensitive or really bothered them or if they didn’t understand and all of those went pretty well.
We had one question that people had a bit of a hard time, that was one of the injury questions about how old do you think a child can be to swim in a pool without an adult or lifeguard present? So a lot of people said never. And then we got, you know, just a couple people commented that that question didn’t really work for them, which I can see why. But it’s a national question from some other survey, so we’re just going to try and explain it a little better. And then, I think, that’s about it. But the parents liked it. They commented that they thought it was a great idea, we need to know what’s going on with kids. And, you know, most adults love to talk about their kids, so it was really quite easy to get them to participate in this. And that’s all I have, thank you.