Skip to main content
. 2023 May 25;43(1):03. doi: 10.35946/arcr.v43.1.03

Appendix 3.

Predictive Characteristics* of Other Screening Tools for Fetal Alcohol Spectrum Disorders and Its Associated Symptoms

Author PAE/FASD Sample Comparison Sample Predictor Outcome Sensitivity Specificity PPV NPV Accuracy AUC from ROC
Aring et al., 202166 37 Children with FASD 65 Healthy children, 33 ADHD, 57 Moderate to late prematurity FASD Eye Code ≥ 10 FASD 43 94 0.78
65 Healthy children only 43 100 0.87
16 Silver-Russell syndrome 43 88 0.6
65 Healthy children 0.92
All comparison groups 0.76
33 ADHD 0.66
57 Moderate to late prematurity 0.75
65 Healthy children only FASD Eye Code ≥ 9 57 98 0.87
Astley & Clarren, 199658 42 FAS: 21 development sample and 21 validation sample 84 without FAS (including 4 with other genetic conditions) placed into 2 groups; 42 per group for development and then validation Facial features: 2D continuous measurements philtrum/lip Gestalt FAS 100 93
Likert scale rating philtrum & lip 100 100
Likert scale philtrum/lip continuous 100 100
Astley et al., 200259 Sampled 600 children in foster care screened Facial analysis software Facial features: 2D clinical FASD diagnosis Screened + Gestalt FAS 100 99.8 85.7 100 99.8
Moore et al., 200160 41 FAS & 59 pFAS 31 Controls Facial features: 6 2D craniofacial measurements Clinical FAS/pFAS 98 90 96
2 Craniofacial measurements 100 100 100
5 Craniofacial measurements 86 94 88
Widder et al., 202161 22 FASD 31 Controls; 15 ADHD; 20 AUD/OUD; 18 depression Facial features: 2D German BSI-FASD Clinical FASD 77 70–100
Facial analysis software 67 44–79
German BSI-FASD adapted scoring 86 70–100
Fang et al., 200862 50 Finnish FASD diagnosis 32 Finnish controls 3D facial coordinates FASD 88.2 100 100 83.3 92.6
36 FAS Cape-Colored 31 Finnish controls 91.7 90 91.7 90 90.9
86 Combined FASD 63 Combined controls 82.8 76.2 82.8 76.2 80
Mutsvangwa et al., 201064 4 FAS (age 5) 11 Controls (age 5) 3D facial coordinates Clinical FASD 80 100 100 90.9 93.3
13 FAS (age 12) 6 Controls (age 12) 90.9 62.5 76.9 83.3 79
Suttie et al., 201763 22 FAS and 75 heavy AE South Africans (ages 6–18) & 35 FAS and 73 heavy AE Caucasians from CIFASD (ages 3–18) 69 South Africans (ages 6–18) who were Cape-Colored & 141 Caucasians from CIFASD (ages 3–18) 3D facial curvature coordinates of face FAS or heavily AE 0.95–0.98
3D facial curvature coordinates of profile 0.82–0.96
3D facial curvature coordinates of eyes 0.92–0.95
3D facial curvature coordinates malar 0.90–0.95
3D facial curvature coordinates of mandible 0.85–0.93
3D facial curvature coordinates of nose 0.86–0.95
3D facial curvature coordinates of lip vermillion 0.69–0.84
3D facial curvature coordinates of philtrum 0.70–0.90
Valentine et al., 201765 36 FAS 50 Controls Facial dysmorphology novel analysis technology computer scoring FAS 78 92 88 85 0.95
31 pFAS pFAS 79 78 67 87 0.82
22 ARND ARND 50 92 70 83 0.84
89 FASD Any FASD 89 69 83 78 0.86
36 FAS 50 Controls Facial dysmorphology novel analysis technology manual scoring FAS 99 89 87 99 0.96
31 pFAS pFAS 76 89 81 86 0.89
22 ARND ARND 43 92 70 79 0.74
89 FASD Any FASD 87 77 87 77 0.88
Jacobson et al., 200868 12 FAS, 18 pFAS, 29 heavy PAE 20 Nonexposed controls; 4 nonexposed microcephalic Physiological neural response: % criteria for eye-blink conditioning FASD 70.2 75 87 51.4 71.6
10 FAS 100 75 62.5 100 82.4
Kable et al., 202169 26 PAE no diagnosis, 19 ARND, 5 FAS/pFAS 70 No PAE/no diagnosis Physiological neural response: cardiac orienting response auditory COR Deviation Index FASD 0.65
Visual COR Deviation Index 0.77
Mesa et al., 201770 Sample of Ukrainian infants with 26 having mild developmental delay Sample of Ukrainian infants with 98 within normal limit development Physiological neural response: cardiac orienting response Standard COR 12-month Bayley < 85 66 85 0.81
Key features COR 62 82 0.81
Maternal drinking 49 75 0.68
Maternal drinking + standard COR 65 87 0.84
Maternal drinking + key COR 62 80 0.8
Kaneko et al., 199671 14 FAS 14 Controls Physiological neural response: auditory event potentials FAS 78.6 42.9 57.9 66.7 60.7
14 Down syndrome Combination of P300 variables 78.6 85.7 84.6 80 82.1
Tseng et al., 201378 13 FASD 21 ADHD 19 Features of saccadic eye movements FASD 90.4
18 Controls 79.2
21 ADHD and 18 controls 73 91 77.3
Zhang et al., 201977 91 FASD 116 Controls Physiological neural response: eye tracking features of eye tracking, DTI, and neurobehavioral testing FASD 81.8 87.5 84.8
Prosaccade 69.6
Antisaccade 76.1
Mesasaccade 65.2
DTI (4 features) 67.4
Neurobehavioral (3 domains) 78.3
Bookstein et al., 200779 23 PAE 21 Unexposed or lightly exposed Neuroimage: MRI “hook” feature of corpus callosum AE 52.2 95.2 92.3 64.5 72.7
Little & Beaulieu, 202080 79 FASD 81 Controls Neuroimage: MRI 10 heavily weighted brain regions FASD 64 88 77
Burd et al., 199984 1,013 Screened in school system with 6 FAS 1,007 Screened in school/no FAS diagnosis Screening tool completed by trained staff > 20 Screened + FASD 100 94.1 9.2 100 94
Burd et al., 200385 152 FAS IOM FAS cohort FAS Diagnostic Checklist-total FAS 84.9 82.4 75.4 89.5
157 pFAS pFAS 54.3 83.3 66.1 74.1
87 PAE not FAS PAE Not FAS 77 90.8 70.5 93.2
Grant et al., 201386 25 FASD (FAS, ARND, FAE, static encephalopathy) 463 No PAE Self-report interview life history screen/addiction severity index FASD 80.8 65.5 67.6
Klug et al., 202181 76 FASD 76 Controls Caregiver Questionnaire: Checklist ARND-BC Parent All Questions FASD 91.9 95.8 95.2 92.5
ARND-BC parent questions positive from binary regression 90.8 92.8 92.4 91.3
ARND-BC parent questions positive from continuous regression 89.7 94.5 94 90.5
ARND-BC parent questions sum of positive domains 89.7 94.5 94 90.5
Nash et al., 200682 54 FASD (ages 6–16) 30 Controls Caregiver Responses to CBCL (7 items) FASD 96.5
30 Controls Caregiver responses to CBCL (6/7 items) 86 82
30 Controls Caregiver responses to CBCL (5/7 items) 80 70 80 90.1
30 ADHD Caregiver responses to CBCL (6/7 items) 86.3
30 ADHD Caregiver responses to CBCL (3 items) 81 72
30 ADHD Caregiver responses to CBCL (2 items) 70 80 84.9
Nguyen et al., 201483 79 PAE + ADHD; 36 PAE – ADHD 90 Controls + ADHD; 16 Controls – ADHD Caregiver responses to BRIEF AE 71.4
Bernes et al., 202190 177 Alcohol exposed CIFASD II 204 Controls CIFASD II Low cutoff > 1.5 neurobehavioral battery & dysmorphology exam AE 76.9 76.5 66 84.8 76.6
High cutoff > 2 neurobehavioral battery & dysmorphology exam AE 63.6 87.8 75.5 80.3 78.8
177 Alcohol exposed CIFASD III 346 Controls CIFASD III Low cutoff > 1.5 neurobehavioral battery & dysmorphology exam AE 83.1 59 50.9 87.2 67.1
High cutoff > 2 neurobehavioral battery & dysmorphology exam AE 66.1 77.5 60 81.7 73.6
Classification 3 AE vs. ADHD; latent profile analysis of complex neurobehavioral battery, dysmorphology, growth AE 59.8 75.8 87.4 40 64
Coles et al., 202091 82 High risk ARND CoFASP sample of 1st graders 80 No risk CoFASP sample of 1st graders Comprehensive neuropsychology battery ARND CoFASP 74.4 83.8 82.4 76 79
85 Low-risk ARND CoFASP sample of 1st graders 73 No risk CoFASP sample of 1st graders 90.6 89 90.6 89 89.9
Luca et al., 201687 21 FAS 86 No FASD Neurodevelopmental: quick screen neurological test-2 FAS 31.8 86.1 36.8 83.2 75
60 PAE 42 No PAE AE 18.3 87.5 64.7 46.2 49.1
Goh et al., 201688 146 CIFASD II AE 288 No AE Neurodevelopmental: complex neuropsychological battery AE 74.2 89.9 78.6 87.4 84.6
55 CIFASD III child AE 110 No AE Psychologist decision tree incorporating complex neurobehavioral battery & physical exam AE 70.7 93.5 87.9 82.9 84.5
98 CIFASD III adolescent AE 191 No AE AE 79.3 87.6 77.4 88.7 84.7
146 CIFASD II AE 288 No AE AE 79.2 80.6 70.7 86.7 80.1
55 CIFASD III child AE 110 No AE AE 63.8 93.4 85.7 80.7 82.1
98 CIFASD III adolescent AE 191 No AE AE 81.3 78.3 71.4 86.2 79.5
Johnston et al., 201992 43 FASD 20 PAE but no FASD Movement battery of tests (-2 SD, < 2nd percentile) FASD 2–38 80–100
Movement battery of tests 5th percentile 9–75 68–100
Movement battery of tests 9th percentile 35–85 60–85
Movement battery of tests 16th percentile 44–83 45–70
Mattson et al., 201093 CIFASD children ages 8–17; 41 AE/FAS 46 CIFASD controls Latent profile analysis of a complex battery AE/FASD 87.8 95.7 94.7 89.8 92
CIFASD children ages 8–17; 41 alcohol-exposed/deferred not FAS 60 Controls Latent profile analysis of a complex battery AE/deferred FAS 68.4 95 89.7 82.6 84.7
Mattson et al., 201389 CIFASD children (ages 8–17) 209 AE (79 were FAS) 185 Controls; 74 ADHD Neurodevelopmental: complex neuropsychological battery Latent profile analysis of complex neurobehavioral battery Dysmorphia, growth AE/FASD 77.2 75.7 57.6 88.6 76.1
Classification 2 AE/non-FAS vs. controls AE/FASD 70.1 72.4 61.7 79.3 71.5
Thorne & Coggins, 2008;94 Thorne et al., 200795 16 FASD 16 Normal controls Narrative speech samples (NSS)-ANRTW FASD 87.5 75 81.3 0.86
NSS-PR 81.3 62.5 71.9 0.77
NSS-ANR 81.3 81.3 81.3 0.76
NSS-ANR Cutoff 1.7% 100 1
NSS-AR 0.76
NSS-Nominal Reference Errors (rNRE) 88 0.9
NSS-Nominal Reference Errors (rNRE) 2% 97 0.98
*

Prediction characteristics evaluated in each study included sensitivity, specificity, NPV, PPV, accuracy, and AUC derived from ROC curves. Sensitivity refers to the probability that the test is positive when the condition is present. Specificity refers to the probability that the test is negative when the condition is not present. PPV refers to the probability that the condition, is present when the test is positive. NPV refers to the probability that the condition is not present when the test is negative. Accuracy refers to the overall probability that the case is correctly classified from the test. Finally, AUC is derived from creating receiver operating curves by plotting the true positive rate (sensitivity) relative to the false positive rate (1-specificity). The AUC references the area on the graph created by the regression line relative to the chance rate of prediction. Values of 1 would indicate perfect condition, and values of 0.50 would indicate chance prediction using a yes/no model. Predictive validity values are presented as percentages with the exception of AUC values, which are reported in proportions of accurate diagnostic classification with values of 0 to 1.00. The different categories of predictive data (facial, neurophysiological, neuroimaging, questionnaire, and psychological performance measures are shaded from white to dark blue.

Note: 2D, two-dimensional; 3D, three-dimensional; ADHD, attention-deficit/hyperactivity disorder; AE, alcohol-exposed; ANRTW, Ambiguous Normal Reference Total Word; ARND, alcohol-related neurodevelopmental disorder; ARND-BC, Alcohol-Related Neurodevelopmental Disorder Behavior Checklist; AUC, area under the curve; AUD, alcohol use disorder; BRIEF, Behavior Rating Inventory of Executive Function; BSI-FASD, biographic screening interview for fetal alcohol spectrum disorders; CBCL, Child Behavior Checklist; CIFASD, Collaborative Initiative on Fetal Alcohol Spectrum Disorders; CIFASD-II, CIFASD, Phase II; CIFASD-III, CIFASD, Phase III; CoFASP, Collaboration on FASD Prevalence; COR, cardiac orienting response; DTI, diffusion tensor imaging; FAE, fetal alcohol effect; FAS, fetal alcohol syndrome; FASD, fetal alcohol syndrome disorders; IOM, Institute of Medicine; MRI, magnetic resonance imaging; NPV, negative predictive validity; NSS, narrative speech samples; NSS-ANR, narrative speech sample–ambiguous normal reference; NSS-AR, narrative speech sample–ambiguity rate; NSS-PR, narrative speech sample–pronoun reference; OUD, opioid use disorder; PAE, prenatal alcohol exposure; pFAS, partial fetal alcohol syndrome; PPV, positive predictive value; rNRE, rate of nominal reference errors; ROC, receiver operating characteristic; SD, standard deviation.