Page 249 - 16Neonatal Jaundice_compressed
P. 249

Appendix H: Evidence tables


 Bibliographic details   Study type &   Patient characteristics   Test, Reference Standard,   Results   Reviewers Comments
 Evidence level   Threshold for a positive test
                PPV: 11/67 (16.4%)
                NPV: 101/102 (100%)

                For TSB in phase 2
                (threshold = 154 micromol/litre)
                Sensitivity: 45/46 (97.8%)
                Specificity: 348/774 (45%)
                PPV: 45/471 (9.5%)
                NPV: 348/349 (99.7%)

 Agarwal R;   Study Type:   All infants with GA > 35 weeks with   Test:   Diagnostic accuracy of TSB (threshold   Unselected population
    Diagnostic study   no significant illness requiring NICU   TSB at 24 ± 6 hours after birth –   value > 102 micromol/litre) for predicting   Test & Reference test described
 Year: 2002      admission for > 12 hours, absence of   TSB = 290 micromol/litre (n = 213)   in detail
    Evidence Level: 1b   any major congenital malformations   three samples taken and mean of      Reference test a standard one
 Country:      and residing near hospital whose   two closest values taken for   Prevalence: 22/213 (10.3%)   Blinding – yes
 analysis
 India      parents agreed to come for follow-up.      Sensitivity: 21/22 (95.4%)
                Specificity: 135/191 (70.7%)
 27       n = 220   Threshold value:   PPV: 21/77 (27,3%)
    mean GA 38 ± 1.4 weeks   > 102 micromol/litre   NPV: 135/136 (99.3%)

    mean BW 2827 ± 459 g   Reference standard: Laboratory
 TSB measured on Day 5 when
 Gender: Males = 53.3%   clinical jaundice
 Ethnicity: Not reported
    > 171 micromol/litre
 Exclusion:
 babies requiring NICU admission,   TSB ≥ 290 micromol/litre taken as
 indicative of hyperbilirubinaemia
 Rh hemolysis.


 Alpay F;   Study Type:   All healthy full term newborn babies   Test:   Diagnostic accuracy of TSB for predicting   Unselected population
    Diagnostic study   with GA = 38 weeks.   TSB within first 24 hours (mean   TSB = 290 micromol/litre (n = 498)   Test & Reference test described
 Year: 2000         17.1 hours)                in detail
    Evidence Level: II   n = 498      Threshold value = 102 micromol/litre   Reference test a standard one
 Country:      mean GA Not reported   ROC curve used for threshold   Prevalence: 60/498 (12.0%)   Blinding – Not reported
 Turkey      mean BW Not reported   value with highest sensitivity for   Sensitivity: 54/60 (90%)
       Gender: Not reported   predicting hyperbilirubinaemia   Specificity: 286/438 (65.3%)
 28       Ethnicity: Not reported   (threshold   PPV: 54/206 (26.2%)
       value: = 102 micromol/litre)   NPV: 286/292 (97.9%)
    Exclusion:
    babies with blood groups A, AB, B   Results also given for threshold   Threshold value = 120 micromol/litre
 and O/Rhesus blood factor   values = 120 micromol/litre   Sensitivity: 36/60 (60%)
 incompatibility and a positive direct   and = 137 micromol/litre
 antiglobulin test result      Specificity: 363/438 (82.9%)
 G6PD deficiency   Reference standard: Laboratory   PPV: 36/111 (32.4%)
                NPV: 363/387 (97.8%)
    TSB measured at 24 hours interval
 for next 4 days   Threshold value = 137 micromol/litre

                Sensitivity: 21/60 (35%)


                                                                 197
   244   245   246   247   248   249   250   251   252   253   254