Table 6 Assessment of the OX-NDA against feasibility criteria for use of an early child developmental assessment in a low- and middle-income setting
Feasibility Criteria | Does OX-NDA fulfil the criteria? | Additional details |
|---|---|---|
World Bank Toolkit for Examining ECD1 | ||
Psychometrically adequate, valid and reliable | Yes | ICCs 0·63 and 0·68 (p < 0·001) between BSID-III and OX-NDA for cognition, and total score domains; and motor composite; ICC 0.30 (p < 0.04) for language composite. Internal consistency satisfactory. Sensitivity in predicting BSID-III composite scores < 85 (moderate delay) was 76, 73, and 43% for the OX-NDA cognition, motor and language domains at cut-off scores of <=60, 73, and 60 respectively. Specificity in predicting BSID-III composite scores < 85 (moderate delay) was 75, 80, and 33% for the OX-NDA cognition, motor, and language domains at cut-off scores of <=60, 73, and 60 respectively. Inter-rater reliability and test-rest reliability was k = 0.80-0.96, 95% CI: 0.78-0.97 and k = 0.85-0.94, 95% CI: 0.80-0.95 across all domains. |
Balanced in terms of number of items at the lower end to avoid children with low scores | Yes | Age range of items extend to 6 months |
Enjoyable for children to take (e.g. interactive, colourful materials) | Yes | |
Relatively easy to adapt to various cultures | Yes | Adapted via cultural customisation session during training and currently in use in Brazil, India, and Grenada |
Easy to use in low-resource settings, e.g. not requiring much material | Yes | Cost of kit GBP 100.0; no fee per use; manuals and forms available upon request, mobile phone/tablet based OX-NDA E-form available. |
Not too difficult to obtain or too expensive | Yes | See above |
Able to be used in a wide age range | No | Moderately narrow age range (10 to 14 months) |
Fischer et al’s feasibility criteria for use of developmental screening tools at primary healthcare level in low-middle income settings2 | ||
Results understood by health workers | Yes | Cut-offs for moderate-to-severe delay |
Reliable | Yes | High, inter-rater reliability and test-rest reliability of k = 0.80-0.96, 95% CI: 0.78-0.97 and k = 0.85-0.94, 95% CI: 0.80-0.95 across all subscales. |
Valid | Yes | See above |
Acceptable to caregivers | Yes | |
Provides information that is relevant to primary care providers | Yes | Cut-offs |
Information that can be used for referrals of early intervention | Yes | Cut-offs |
Information that is useful for anticipatory guidance | Unknown | |
Results understood by caregivers | Yes | |
Staff members have the expertise to answer questions | Yes | Session on maternal questions and responses included in training package. |
Access to application | Yes | Manuals, paper forms and E-form available upon request. |
Training involved | Yes | Time taken to train assessors in the OX-NDA: 1 day for ≤3 assessors, 2 days for 3-5 assessors, 3 days for 5-10 assessors |
How long it takes to administer the tool | Yes | 15-25 minutes |
Cover multiple areas of child development | Yes | Cognition, language, motor skills, and behaviour (positive and negative) |
Cost of the tool | Yes | Cost of kit GBP100.0; no fee per use; mobile phone/tablet based OX-NDA E-form available. |
Minimal adaptation needed | Yes | Sessions on cultural customisation and translation included in training |
Educational level of staff members | Yes | Primary education; non-specialist personnel |
How many staff members to administer the tool | Yes | 1 |
Local norms available | Yes | Cut-offs based on Brazilian sample. Research to develop international norms on-going. |
Space | Yes | Minimal space for storage of kit and forms. Mobile phone/tablet based OX-NDA E-form available. Home-based assessments possible. |
Boggs et al’s rating of early child development outcome measurement tools for routine health programme use3 | ||
Validity | Rating: 2 | Validity somewhat below widely accepted threshold (0.5 to 0.7) against another performance-based tool e.g. BSID-III |
Reliability | Rating: 3 | High, inter-rater reliability and test-rest reliability of k = 0.80-0.96, 95% CI: 0.78-0.97 and k = 0.85-0.94, 95% CI: 0.80-0.95 across all subscales. |
Cultural Adaptability | Rating: 3 | Easy modification of items, materials and procedures |
Accessibility | Rating: 2 | Tool administration, scoring and interpretation, all available online, but some intellectual property or other restrictions. Minimal cost to tool <US$ 10 per child App (mobile phone/tablet based OX-NDA E-form) available |
Training | Rating: 2 | Moderate (> 1 hour to < 1 day), requires standardization and training on direct assessment of children’s abilities, no certification requirement. |
Administration time | Rating: 2 | < 15 to 20 minutes, minimum to moderate scoring. |
Geographical uptake | Rating: 3 | Used in at least three continents (Asia, Europe, South America) |
Clinical relevance and utility | Rating: 2 | Sensitivity in predicting BSID-III composite scores < 85 (moderate delay) was 76, 73, and 43% for the OX-NDA cognition, motor and language domains at cut-off scores of <=60, 73 and 60 respectively. Specificity in predicting BSID-III composite scores < 85 (moderate delay) was 75, 80, and 33% for the OX-NDA cognition, motor and language domains at cut-off scores of <=60, 73, and 60 respectively. Further research to develop international norms, and contextually appropriate referral pathways underway. |