RESEARCH ON PEDS Parents Evaluation of Developmental Status

RESEARCH ON PEDS: Parents’ Evaluation of Developmental Status Frances Page Glascoe Adjunct Professor of Pediatrics Vanderbilt University www. pedstest. com www. forepath. org

“Your teacher wishes me to delineate those watershed occasions in your life that have led you to become, slowly and inexorably, a loose cannon. ”

Developmental/Behavioral Tasks of Health Supervision Visits 4 developmental promotion 4 behavioral guidance and patient education 4 developmental/behavioral screening 4 observation/monitoring 4 reassurance 4 referral

Reasons for limited use of screening tests at well visits screening tests take too long many are difficult to administer children may not cooperate reimbursement is limited not all developmental/behavioral issues are addressed lack of familiarity with referral resources

Detection rates without screening tests 70% of children with developmental disabilities not identified (Palfrey et al. J PEDS. 1994; 111: 651 -655) 80% of children with mental health problems not identified (Lavigne et al. Pediatr. 1993; 91: 649=655)

“Looking Good”

Sample questions to parents that don’t work well Do you think he has any problems…. . ? Do you have any worries about her development?

First Question Please tell me any concerns about your child’s learning, development, and behavior.

Additional Questions Do you have concerns about how your child: 2…. is talking or making speech sounds? 3…. understands what you say? 4…uses his or her hands and fingers to do things? 5…uses his or her arms and legs? 6…behaves? 7…gets along with others? 8…is learning to do things for himself/herself? 9…. is learning preschool or school skills? 10. Do you have any other concerns?

Sample Parent Responses He can’t talk plain He’s mean She won’t get dressed or do nothing for herself He’s slow and behind and can’t do what other kids can She won’t mind me I used to be worried but I think he’s doing better

Reliability Test-retest: 88% consistency over time in parents’ concerns Inter-rater: 88% agreement in categorization of concerns given two different raters. Coding agreement 83%, weighted kappa =. 74

Subjects and Sites 4 cross-validation studies on 971 subjects nationally representative sample varied settings: public health, community practices, day care, schools, etc.

Subjects Child Subjects N = 771 Parent Subjects ages 0 - 8 80% mothers 64% white 18% < H. S. 22% African American education 14% Hispanic/Other 5% Spanish 26% low SES speaking 4% enrolled in S. E.

Sites Middle Tennessee Tampa, FL Plymouth, MA Denver, CO Carson City, NV N = 363 N = 112 N = 114 N = 68 N = 114 teaching hospitals private practices day care centers/public schools unenrolled N = 134 N = 123 N = 289 N = 229

Procedures Parents completed PEDS in writing or by interview Either a second psychological examiner blinded to parents’ concerns, or the same examiner, blinded to the potential significance of parents’ concerns, administered the concurrent battery

Procedures II 771 children were administered a dx battery (IQ, language, academic, motor) scores categorized into SE eligibility logistic regression used to identify concerns predictive of developmental status

Initial Results 130 were eligible for special education 24 had been previously identified 641 were not eligible and performed in the broad range of average on all measures

Predictive Concerns by Children’s Ages 0 - 1 1/2 yrs 1 1/2 - 3 yrs 3 - 4 1/2 yrs 4 1/2 - 8 yrs global/cognitive, expressive language, social, medical/other above (except social) + receptive language+ gross motor +fine motor +school

ACCURACY: ACROSS AGE RANGES AGE 0 - 1 1/2 yrs SENSITIVITY N 3/4 % 75 27/34 3 - 4 1/2 yrs 26/35 74 4 1/2 - 8 yrs 42/57 74 TOTAL 98/130 75 1 1/2 - 3 yrs SPECIFICITY N 66/82 79 % 80 117/149 79 118/165 72 172/245 70 473/641 74

Do Parental Characteristics Affect Their Concerns? No differences in accuracy on the basis of parents’ level of education or parenting experience Almost all parents’ derive concerns by comparing their children to others Educated parents, especially dads, 21 times as likely to raise concerns spontaneously. Children whose parents discussed concerns were 40 times more likely to be enrolled in special education

Evidence-Based Decisions when and where to refer when to screen and type of screen needed when to offer developmental promotion when to provide behavioral guidance when to observe vigilantly when reassurance and routine monitoring are sufficient

Multiple Predictive Concerns 52% Disabled OR = 11. 4 16% Below Average Single Predictive Nonpredictive Concerns 29% 11% 20% 23% 7% 17% OR = 7. 6 13% 3% 43% OR = 1. 3 No Concerns 5% 11% OR = 1. 0 No predictive concerns but communication barriers 19% 0 R = 4. 6 35%

Screening in Response to Multiple Predictive Concerns 11% Multiple concerns (N = 27) but passed Brigance/BDIST, or no significant concerns Multiple concerns and failed Brigance/BDIST (N = 36), or single significant concern Sensitivity Specificity DIAGNOSIS NO YES 335 29 100 47 435 76 47/76 = 62% 335/435 = 77% prior 74% 73%

Referral Accuracy: Multiple Predictive Concerns 11% 79% (19/24) of those needing speech-language evaluations had parents with two or more concerns about receptive language, self-help, school or social skills 71% (15/21) of those needing psychological/ educational testing had parents with one or fewer such concerns

Screening in the Presence of a Single Predictive Concern or 23% Communication Barrier 3% DIAGNOSIS NO YES no concerns or single concern or communication barrier and passed Brigance Screen multiple concerns, or single concern/communication barrier but child failed Brigance Screen Sensitivity Specificity 280 15 78 41 352 56 41/56 = 73% 280/352 = 80% prior 75% 74%

What of those who fail screening but aren’t eligible for special education? 23% 3% no concerns or single concern or communication barrier and passed Brigance Screen multiple concerns, or single concern/communication barrier but child failed Brigance Screen DIAGNOSIS NO YES 280 15 78 41 These children tended to perform below average in IQ, academics, and/or language—the better predictors of school success

Developmental Screening in Response to Non. Predictive Concerns or No Concerns 20% 43% no concerns or nonsignificant concerns and passed Brigance/BDIST Screens multiple concerns, single concern/communication barrier or no/nonsignificant concerns and failed Brigance/BDIST Sensitivity Specificity DIAGNOSIS NO YES 246 11 189 65 435 76 65/76 = 85% 246/435 = 56% prior 74% 73%

Behavioral Screening in the Presence of No or Nonpredictive Concerns: 20% 43% NO Nonpredictive Concerns YES Sensitivity Specificity >4 1/2 yrs. NO < 4 1/2 yrs 1 1/2 SDs above mean NO YES 91 6 98 2 46 13 26 13 137 19 124 15 13/19 = 68% 91/137 = 66% 13/15 = 87% 98/124 = 79%

Evidence-Based Decisions when and where to refer when to screen and type of screen needed when to offer developmental promotion when to provide behavioral guidance when to observe vigilantly when reassurance and routine monitoring are sufficient

Rationale for Electronic PEDS Empower parents Ensure consistency in scoring and administration Enhance test availability in multiple languages and locations Enable tailoring for local resources Create an electronic database for research and quality improvement Give immediate access to patient education information

Electronic PEDS www. forepath. org Web accessible PEDS for v Licensed PEDS users v Self-selected parents PEDS scoring Web service for EMR/EHR and other electronic systems

Directions for Future Research Do parents become more accurate when administered PEDS over time? Does the addition of professional judgment improve accuracy? Can PEDS detect school problems in older children, autism, CP, etc. ? Frances. P. Glascoe@Vanderbilt. edu