APPENDIX 8

Positive Predictive Values of Screening Tests

Study (Year)Screening TestAge of EnrolleesNProportion With ConditionPositive Predictive Value (95% CI)Negative Predictive Value (95% CI)
Barry and Konig28 (2001)Retinomax autorefractor3 y404Amblyopia: 2.5% (10/404)0.05 (0.02–0.09)0.99 (0.97–1.0)
Barry and Konig29 (2003)Visual inspection, cover-uncover test, eye motility and head posture examination, Lea symbols visual acuity test3 y1180Amblyopia or amblyogenic risk factors: 2.3% (26/1114)0.25 (0.16–0.36)1.0 (0.9 9–1.0)
Berry et al30 (2001)MTI photoscreenerPreschool (subgroup)51Amblyogenic risk factors: 45% (23/51)0.68 (0.48–0.84)0.83 (0.61–0.95)
Bertuzzi et al31 (2006)Lea symbols visual acuity test38–54 mo149Amblyogenic risk factors: 16% (23/143)A, 0.52 (0.36–0.68); B, 0.69 (0.48–0.86)A, 0.99 (0.95–1.0); B, 0.96 (0.90–0.99)
Chang et al32 (2007)A, Distance visual acuity; B, near visual acuity; C, NTU random-dot stereogramPreschool5232Amblyopia: 2.20% (115/5232)A1, 0.12a; A2, 0.04a; B, 0.13a; C, 0.17aA1, 0.995a; A2, 0.996a; B, 0.988a; C, 0.986a
Chui et al33 (2004)Lea symbols visual acuity test, Frisby stereoacuity test, and external visual inspection35–58 mo178 (141 completed gold-standard evaluation)Amblyogenic risk factors: 13% (18/141)0.41 (0.24–0.61); <41 mo: 0.41 (0.21–0.64); ≥41 mo: 0.43 (0.10–0.82)0.95 (0.89–0.98); <41 mo: 0.90 (0.74–0.98); ≥41 mo: 0.96 (0.90–0.99)
Cogen and Ottemiller34 (1992)Visiscreen 100 photoscreener6 mo to 6 y127Any visual condition: 12% (13/113); refractive error: 5% (6/113); strabismus: 4% (5/113); refractive error + strabismus: 1% (1/113); media opacity: 1% (1/113)0.65 (0.38–0.86)0.98 (0.93–1.0)
Cooper et al35 (1999)A, Fortune Optical VRB-100 photoscreener; B, MTI photoscreener12–44 mo10561 cases (amblyopia), 44 controlsA (reader 1): 0.76 (0.61–0.87); A (reader 2): 0.86 (0.72–0.95); B (reader 1): 0.78 (0.62–0.89); B (reader 2): 0.88 (0.74–0.96)A (reader 1): 0.60 (0.46–0.72); A (reader 2): 0.69 (0.54–0.80); B (reader 1): 0.59 (0.46–0.72); B (reader 2): 0.65 (0.50–0.78)
Ehrt et al38 (2007)Vision Screener video refractor0–7 y161Amblyogenic risk factors: 43% (70/161)0.71 (0.59–0.82)0.78 (0.68–0.86)
Guo et al39 (2000)A, Computer-photorefractor; B, noncycloplegic retinoscopy9–50 mo300Amblyogenic risk factors: 56% (168/300)A, 0.92 (0.87–0.96); B, 0.85 (0.79–0.90)A, 0.93 (0.87–0.97); B, 0.82 (0.74–0.88)
Hope and Maslin40 (1990)Random dot E stereogram3–4 y176Refractive error or strabismus: 5% (9/168); refractive error: 5% (9/168); strabismus: 0.6% (1/168)0.17 (0.08–0.31)0.99 (0.96–1.0)
Kennedy and Sheps42 (1989)A, Otago-type photoscreener (noncommercial); B, Off-axis-type photoscreener (noncommercial)≤6 y236Any amblyogenic risk factor: 42% (98/236); strabismus only: 14% (33/236); strabismus + refractive error or anisometropia: 18% (42/236); refractive error or anisometropia: 8% (18/236); anisocoria or lid tumor: 2% (5/236)Any condition; A, 0.92 (0.85–0.96); B, 0.82 (0.73–0.89)Any condition; A, 0.96 (0.91–0.98); B, 0.89 (0.82–0.94)
Kennedy et al43 (1995)A, Otago-type photoscreener (noncommercial); B, Snellen E or Stycar graded balls visual acuity test and Titmus stereotestNot reported264Any visual condition: 8% (21/264); strabismus: 1.1% (3/264); refractive error: 4.2% (11/264); strabismus and refractive error: 0.8% (2/264); structural: 0.4% (1/264)A, 0.77 (0.60–0.95); B, 0.54 (0.28–0.81)A, 0.98 (0.91–1.00); B, 0.94 (0.91–0.97)
Kennedy and Thomas44 (2000)iScreen photoscreener45% ≤6 y449Amblyogenic risk factors: 64% (273/423)0.94 (0.90–0.96); ≤3 y 0.97a; 4–6 y 0.97a0.86 (0.80–0.91)
Miller et al46 (1999)A, Lea symbols visual acuity test; B, Retinomax K-plus autorefractor3–5 y245Significant refractive error: 31% (76/245); all had astigmatismA, 0.42 (0.35–0.50); B, 0.75 (0.65–0.83)A, 0.92 (0.83–0.96); B, 0.95 (0.901–0.98)
Miller et al47 (2001)A, Lea symbols visual acuity test; B, MTI photoscreener; C, Nidek KM-500 Keratometry Screener; D, Retinomax K-Plus autorefractor3–5 y379Astigmatism ≥1.00 D: 48% (182/379)A, 0.48 (0.41–0.54); B, 0.68 (0.60–0.75)b; C, 0.79 (0.73–0.84); D, 0.94 (0.90–0.97)A, 0.93 (0.88–0.97); B, 0.70 (0.63–0.76)b; C, 0.94 (0.90–0.97); D, 0.94 (0.89–0.96)
Morgan and Johnson49 (1987)Visiscreen 100 photoscreener3 mo to 8 y63Any visual condition: 60% (34/57)0.84 (0.68–0.94)0.85 (0.62–0.97)
Newman and East50 (1999)Sheridan-Gardiner visual acuity; cover-uncover test; ocular movements and convergence; prism test; TNO screening plate; Snellen visual acuity3.5 y and at 5–6 yCohort of 936 children; data reported on 597Amblyopia: 2.5% (15/597)0.27 (0.16–0.41)1.0 (0.99–1.0)
Ottar et al51 (1995) and Donahue et al59 (2002)MTI photoscreener6–59 mo949Amblyogenic risk factors: 20% (192/949)A, 0.69 (0.62–0.75); B, 0.77 (0.64–0.87)cA, 0.95 (0.93–0.97); B, 0.95 (0.93–0.96)c
Rogers et al52 (2008)MTI photoscreener SureSight autorefractor1–6 y100Clinically significant amblyopia: 58% (58/100)A, 0.68 (0.57–0.78); B, 0.75 (0.63–0.86); C, 0.75 (0.61–0.86); D, 0.77 (0.62–0.88); E, 0.92 (0.82–0.97)A, 0.89 (0.65–0.99); B, 0.69 (0.52–0.83); C, 0.60 (0.45–0.74); D, 0.58 (0.44–0.72); E, 0.92 (0.80–0.98)
Shallo-Hoffmann et al53 (2004)Lea symbol and HOTV charts and random dot E stereoacuity test2–6 y269Any vision condition: 6% (5/81)0.24 (0.08–0.47)1.00 (0.94–1.0)
Tong et al56 (2000)MTI photoscreener<4 y old387Strabismus: 49% (190/387); refractive error: 55% (211/387)All photographs: 0.95 (0.90–0.98); informative subset of 313 photographs: 0.95 (0.90–0.98)All photographs: 0.43 (0.36–0.50); informative subset of 313 photographs: 0.41 (0.33–0.49)
VIP Study Group55 (2004)Crowded linear Lea symbols visual acuity test3, 4, or 5 y old3121Any vision condition: 29% (755/2588); “very important to detect and treat early” conditions: 5.4% (135/2588); amblyopia: 2.9% (75/2588); reduced visual acuity: 5.1% (132/2588); strabismus: 1.9% (48/2588); refractive error: 9.3% (240/2588)Any condition; A, 0.73 (0.67–0.78); B, 0.78 (0.72–0.83)Any condition; A, 0.84 (0.82–0.86); B, 0.81 (0.78–0.83)
Crowded linear HOTV visual acuity testAny condition; A, 0.68 (0.62–0.74); B, 0.69 (0.62–0.76)Any condition; A, 0.82 (0.79–0.84); B, 0.77 (0.74–0.80)
Random dot E stereoacuity testAny condition; A, 0.64 (0.58–0.71); B, 0.54 (0.46–0.63)Any condition; A, 0.78 (0.75–0.81); B, 0.80 (0.78–0.83)
Stereo Smile II stereoacuity testAny condition; A, 0.66 (0.60–0.72); B, 0.68 (0.62–0.75)Any condition; A, 0.73 (0.70–0.76); B, 0.78 (0.76–0.80)
Retinomax autorefractorAny condition; : 0.71 (0.68–0.75); B, 0.78 (0.74–0.82)Any condition; A, 0.86 (0.84–0.87); B, 0.83 (0.81–0.84)
SureSight autorefractorAny condition; A1, 0.47 (0.43–0.51); A2, 0.71 (0.66–0.76); B, 0.77 (0.72–0.82)Any condition; A1, 0.91 (0.89–0.93); A2, 0.86 (0.84–0.88); B, 0.83 (0.81–0.85)
iScreen photoscreenerAny condition; 0.71 (0.64–0.77)Any condition; 0.79 (0.77–0.81)
MTI photoscreenerAny condition; 0.71 (0.64–0.77)Any condition; 0.79 (0.77–0.81)
Power Refractor IIAny condition; A, 0.68 (0.65–0.73); B, 0.70 (0.64–0.76)Any condition; A, 0.83 (0.81–0.85); B, 0.79 (0.76–0.81)
Cover-uncover testAny condition; 0.78 (0.66–0.86)Any condition; 0.73 (0.70–0.76)
Weinand et al57 (1998)MTI photoscreener6–48 mo112Any abnormality: 81% (83/102); refractive error: 41% (41/102); strabismus without refractive ; error: 7% (7/102); strabismus with refractive error: 21% (21/102); organic anomaly: 13% (13/102)A (pediatrician interpreter): 0.88 (0.79–0.94); B (orthoptist interpreter): 0.93 (0.84–0.98); C (ophthalmologist 1 interpreter): 0.92 (0.83–0.98); D (ophthalmologist 2 interpreter): 0.90 (0.81–0.96)A (pediatrician interpreter): 0.62 (0.32–0.86); B (orthoptist interpreter): 0.45 (0.27–0.64); C (ophthalmologist 1 interpreter): 0.38 (0.22–0.55); D (ophthalmologist 2 interpreter): 0.48 (0.27–0.69)
Williams et al58 (2000)Topcon PR2000 autorefractor12.5–68.7 mo222A, spherical error >3.75 D: 19% (36/189); B, anisometropia >1.25 D: 12% (23/189); C, astigmatism >1.25 D: 16% (30/189)A, 0.69 (0.48–0.86); B, 0.68 (0.46–0.85); C, 0.70 (0.46–0.88)A, 0.89 (0.83–0.93); B, 0.96 (0.92–0.99); C, 0.91 (0.85–0.94)
  • TNO indicates a Dutch stereoacuity test.

  • a Raw data not provided; unable to calculate CIs.

  • b Calculation based on n = 379, median sensitivity and specificity.

  • c Based on reported sensitivity and specificity; do not match values reported in article.