Please note: In 2003, the CTF updated its Grades of Recommendations to include an "I Recommendation" for situations where insufficient evidence exists to allow a recommendation to be made. (Formerly, these situations were captured under a "C Recommendation".) This change is not retroactive, and all "C Recommendations" made prior to 2003 have not been reevaluated in light of the new "I" recommendation grade. For a discussion of these recommendation grades, please link to the 2003 article in the Canadian Medical Association Journal here.
Preschool Screening for Developmental Problems
Prepared by John W. Feightner, MD, MSc, FCFP,
Deaprtment of Family Medicine, The University of Western Ontario
These recommendations were finalized by the Task Force in March 1993
Contents
Preschool screening is directed at children 3 to 5 years of age and entails early detection of cognitive and behavioural problems that could jeopardize school performance. We evaluated primarily those screening tests that can be used in the primary care physician’s office or in a relatively small system of care.<1> Remedial intervention for identified problems is usually carried out by professional educators, but often involves other health professionals such as speech and language therapists and occupational therapists. Early detection of developmental problems using the Denver Developmental Screening Test (DDST) has been shown not to improve school performance but to increase parental anxiety and is therefore not recommended. However, the benefits of other screening tools and remedial programs are controversial.
Burden
of Suffering ManeuverThe DDST is the most widely used test.<2> It is relatively easy to perform, takes little time and is inexpensive. It is effective among children with intelligence quotients (IQs) of less than 70, but its ability to identify less severe or specific developmental problems has been questioned. The reported sensitivity and specificity have varied considerably; in one study they were 29% and 89% respectively,<3> whereas in another they were 5% to 10% and 99%.<4> With an assumed prevalence rate of 20% the positive predictive value (proportion of true positive results) is 71% and the negative predictive value (proportion of true negative results) 81%; because of its low sensitivity, however, the DDST fails to identify 90% to 95% of children with developmental problems.
A revised version of the Denver Scale, the Denver-II, has added 20 new items primarily focusing on language expression and skills of articulation. There is, however, limited evidence to support its validity and indeed, work by Glascoe et al (1992),<5> raises serious concern regarding the sensitivity and specificity of the new version.
The Developmental Indicators for the Assessment of Learning<6> constitute a multi-dimensional instrument that reflects the school behaviour expected of children in a regular classroom setting. The instrument appears to have considerable potential (sensitivity 46% to 54% and specificity 93%<3>), but more research is required to establish its reliability and validity. Since a team of five to eight people and a moderate amount of equipment are required to administer the test its use is inappropriate for primary care physicians.<7>
The Early Screening Inventory<8> measures developmental abilities, is relatively easy to administer, takes 15 to 20 minutes to perform and has a high sensitivity (81% to 100%) although a lower specificity (67% to 72%).<9> Health care professionals could potentially administer the test, but the reliability of the results depends on the training of the examiner.<8>
The Minneapolis Preschool Screening Instrument<10> is educationally oriented, brief and economical to administer and has achieved a sensitivity of 60% to 63% and a specificity of 89% to 93%.<11> It shows promising reliability and validity.<10> Other instruments such as the McCarthy Screening Test<12> and the Jansky Screening Index (JSI)<13> need further validation. A study using the JSI assessed teachers’ ratings of overall reading skills in grade 1 against outcome in grade 2.<14> The teachers’ ratings had both a sensitivity and a specificity of 93%, as compared with 50% and 90% to 92% respectively for the JSI. Furthermore, the teachers’ assessments of children in grade 1 had a sensitivity of 61% and a specificity of 86% when compared with school performance in grade 6. The suggestion that the teacher is the best early identifier of future school problems has also been supported by other studies.<15>
Effectiveness
of Prevention and TreatmentThe children with positive results were assessed by various outcome measures such as the use of specialized educational services, academic achievement, cognitive and perceptual motor tests, and assessment of behavioural, social and emotional well-being. There were no statistically significant differences in outcome between the two screened groups; however, there was a statistically significant increase in worry about school work among the parents of the children in the intervention group. Since there was no benefit from the screening program and a potentially harmful labelling effect we caution against the widespread use of the DDST for screening purposes.
Large-scale community programs aimed at high-risk or disadvantaged groups have resulted in mixed and controversial findings. These programs have often been characterized by early optimism followed by disappointment.
Information on the Head Start programs in the U.S. is extensive and open to multiple interpretations. Meta-analysis showed that significant and immediate gains in cognitive test scores, socio-emotional test scores and health status were achieved but the improvements were not long-lasting.<18> These findings have been disputed by others.<19>
In the Perry Preschool Program<20> disadvantaged children in Michigan received early intervention. Although conclusions about the effectiveness of the program remain controversial, at least one reviewer has argued that, in the long term, treated children had better school attendance, needed fewer special education services, were more likely to graduate and had lower rates of school drop-out, delinquency and teenage pregnancy than untreated children.<19> (Other evidence on disadvantaged children is reviewed in Chapter 32.)
Recommendations
of Others Conclusions
and RecommendationsThere is insufficient evidence to support either the inclusion or exclusion of other screening instruments (C Recommendation). Caution is advised, however, since problems exist with all current assessment tools, and no interventions have been conclusively proven to be effective. Large-scale community programs to prevent poor school performance in high-risk or disadvantaged groups have also given mixed and controversial results.
Unanswered
Questions (Research Agenda) EvidenceThis review was initiated in January 1993 and updates a report published in December 1989.<1> Recommendations were finalized by the Task Force in March 1993.
Link to Structured Abstract of this review
Link to Summary Table of this review
Link to Selected References list of this review
Reprinted in modified format by the Canadian
Task Force on Preventive Health Care
with permission.
For any technical issues please contact: webmaster@ctfphc.org
Original Copyright
© 1994 Minister of Supply and Services Canada.
Last modified April 1, 1998.
