• Improved developmental screening and surveillance through Health IT

    So besides this, The Upshot, and Healthcare Triage, I also spend a fair amount of time doing actual research. Much of it is in the use of information technology to improve pediatric care. In fact, we have a new paper out in JAMA Pediatrics, “Use of a Computerized Decision Aid for Developmental Surveillance and Screening: A Randomized Clinical Trial”:

    Importance: Developmental delays and disabilities are common in children. Research has indicated that intervention during the early years of a child’s life has a positive effect on cognitive development, social skills and behavior, and subsequent school performance.

    Objective: To determine whether a computerized clinical decision support system is an effective approach to improve standardized developmental surveillance and screening (DSS) within primary care practices.

    Design, Setting, and Participants: In this cluster randomized clinical trial performed in 4 pediatric clinics from June 1, 2010, through December 31, 2012, children younger than 66 months seen for primary care were studied.

    Interventions: We compared surveillance and screening practices after adding a DSS module to an existing computer decision support system.

    Main Outcomes and Measures: The rates at which children were screened for developmental delay.

    Developmental delay is a real issue in kids, and we do a rather crappy job of noticing it. The AAP and others recommend screening at 9, 18, and 30 months, and surveillance at other visits. When I say surveillance, I mean the use of a standardized screening tool, which can be many pages long.

    As you can imagine, this doesn’t happen.

    We have long been studying a system we created called Child Health Improvement through Computer Automation (CHICA). Many studies using it can be found here. For this study, we added a developmental screening module. At 9, 18, or 30 months (of close to it), the system provided automated and personalized ASQ questionnaires to parents in the waiting room to fill out. These were then scored (automatically or by physicians), and then the docs were coached through proper referral or follow-up if necessary. At other visits, parents were asked questions to determine if further screening should occur. If so, the system started the process again.

    We conducted a randomized controlled trial, by clinic, in four clinics. The main outcome of interest was the percentage of kids screened appropriately.

    We found that they system improved the rate at which kids were screened at 9, 18, and 30 months significantly (85.0% vs 24.4%, P < .001). If odds ratios are your thing, being in an intervention clinic improved your odds of being screened by 15.6. Interestingly, the rates of positive screens weren’t different in the clinics. This means that it’s not that screening only occurs in high risk kids in the control clinics. We’re likely just not doing enough of it.

    We did pick up kids earlier, however. The average age of a kid diagnosed in the intervention clinics was 17 months. In the control clinics it was 28 months. Since it’s believed that earlier intervention leads to better outcomes, this is also a big deal.

    Surveillance improved, too. In control clinics, about 42% of parents were surveilled, compared to 72% of parents in intervention clinics (P=0.04).

    It’s hard to write about your own research. I mean, can you really expect me to be totally objective? So take the following with a grain of salt:

    First, I think we spend too much time looking at health IT and focusing on its ability to improve spending or access through efficiency. I think that’s questionable. I think it really has the chance to make an impact in quality, and that’s where we’ve focused out efforts. They keep on bearing fruit. More focus should be put on quality.

    Second, I think CHICA is a remarkable system with amazing potential. It’s not an ADHD system, or a maternal depression system, or an iron deficiency anemia system, of an asthma system, or even a developmental screening system. It’s an everything system. It allows for target screening and monitoring of all pediatric conditions simultaneously, and its prioritization scheme allows it to figure out which things should be done when. It doesn’t require changes in practice when new guidelines come out. It does it all, and it realizes you can’t do everything at every visit.

    We know we’ve got to get CHICA on larger platforms, or get it all online, and we’re working on that. If I don’t write about CHICA more here, it’s only because I’m aware of the inherent conflict of interest of using the blog for that purpose. But maybe I’ve been selling you short. Let me know what you think. I’ll leave comments open here.


    • I will be citing this on my blog. Health IT has been on the receiving end of increasing carping lately (I give ample view to the critics, despite being a Health IT supporter). Nice to see something positive.

      Your blog has become a priority daily stop for me. Keep up the treaty work.

    • My area of research is in autism screening at well-child visits, especially among primary care providers and in rural and undeserved areas. Just yesterday some physicians were telling me that it’s just not going to happen formally and it’s just not practically possible in the clinic. There has to be a reasonable yet documented and validated bridge between the screening researchers, autism specialists, and primary care providers.

      • Lea… what is involved in autism screening?
        Can it be reduced to a simple list of questions and simple tests?
        If so, it could be done in a clinic setting.

        • There is a 23 question questionairre called the MCHAT that is used for initial screening. The screener is already built into CHICA along with followup questions.

    • Looks like an interesting system but the paywalls are preventing me from finding out much about it.
      It appears to be older tech with scanned forms. It could be easily re-coded for use on a tablet using something like opendatakit.org.
      I noticed that the 2008 version has already been translated to xforms (web based) so not too difficult to port to Open Data Kit to run on smartphones and tablets.

      • The patient questionairre is already on tablets and the physician form and other screeners should be electronic in the next year to year and a half.

    • How long does the assessment take, and is it covered for reimbursement under the ACA?