Page 297 - 16Neonatal Jaundice_compressed
P. 297

Appendix H: Evidence tables


 Bibliographic details   Study type &   Patient characteristics   Test, Reference Standard,   Results   Reviewers Comments
 Evidence level   Threshold for a positive test
               Each observer separately
               P < 0.001 for both the mean values and
               absolute values

               Diagnostic accuracy of BiliEye in detecting
               hyperbilirubinaemia

               Area ROC = 0.82

               Best AROC
               0.93 for observations at > 60 hours in babies
               ≥ 37 weeks GA

               Worst AROC
               0.64 for observations at < 36 hours
               0.61 for babies < 37 weeks


 Moyer VA;   Study Type:   Full-term healthy babies (BW > 2000 g   Visual observation by two   Unselected population
    Diagnostic study   and GA> 36 weeks) in well-newborn   experienced staff (paediatric   Agreement between observers on   Reference test not described
 Year: 2000      nursery of an urban public hospital, in   residents, paediatric nurse   presence/absence of icterus at different sites   adequately
    Evidence Level:   whom TSB was measured because of   practitioners, paediatric physicians)   (Weighted κ with 95% CI)   Test and reference test carried
 Country: USA   II   clinical jaundice, Rh-negative mother or   regarding      out within one hour
       positive maternal Coomb’s test.   a) Subjective assessment of   Face & neck:   Blinding – yes
 60          presence/absence of icterus at   0.16 (-0.02 to 0.34)
    n = 122,   different sites   Neck to nipple line:   Funding: Not reported
    GA: > 36 weeks   b) Estimated TSB levels   0.15 (0.01 to 0.29)
 BW > 2,000 g      Nipple line to umbilicus:
 mean age = 2 days (range 8 to 168 hours)   Reference standard: Laboratory   0.23 (0.09 to 0.38)
 Gender: Males = 54.1%   TSB levels within 1 hr   Umbilicus to groin:
 Ethnicity      0.19 (0.05 to 0.34)
 Not reported      Upper legs:
               0.20 (0.06 to 0.35)

               Weighted K not statistically significant for
 Exclusion: babies having previous TSB   other sites – Lower legs, Soles, Arms,
 determination and under phototherapy   Palms, Tip of nose and palate

               Correlation of estimated TSB levels with lab
               TSB (Pearson correlation coefficient)

               Observer 1: r = 0.43
               Observer 2: r = 0.54

               Accuracy of clinical icterus in lower chest
               (nipple line to umbilicus) in detecting TSB >
               205 micromol/litre (n = 243 observations)




                                                                 221
   292   293   294   295   296   297   298   299   300   301   302