Page 313 - 16Neonatal Jaundice_compressed
P. 313

Appendix H: Evidence tables


 Bibliographic details   Study type &   Patient characteristics   Test, Reference Standard,   Results   Reviewers Comments
 Evidence level   Threshold for a positive test
 Rhodesia      GA: Not reported               not specified)
       Postnatal age: Range 2 to 24 days         Blinding – yes
 63    Gender: Not reported                   Data not extractable for
                                              calculating values of TP, FP,
    Ethnicity:                                TN & FN
 Black = 100%

 Exclusion: not defined

 Briscoe L;   Study Type:   Babies > 34 weeks who were having   TcB reading using Minolta JM-102   Correlation of JM-102 with lab TSB levels   Unselected population
    Diagnostic study   blood taken for any reason, mostly done   at the forehead   (Pearson correlation coefficient, n = 303)   Test & Reference test described
 Year: 2002      for clinical jaundice.   (mean of 3 readings used for      in detail
    Evidence Level:      analysis)   r = 0.76, P < 0.0001   Test and reference test carried
 Country:   II   n = 303                      out within one hour
 UK      median BW 3267 g - range 1800–5008   Reference standard: Laboratory   Diagnostic accuracy of JM-102 for detecting   Blinding – not specified
       median GA 39 weeks - range 34–42   TSB levels measured concurrently   TSB > 249 micromol/litre (n = 303)   Data not extractable for
 46    median age at presentation: 3 – range 0 to         calculating values of TP, FP,
    13 days   For diagnostic accuracy:   Area under ROC = 0.89   TN & FN
    Gender: Not reported   Area under ROC curve calculated
    for detecting TSB   Predictive accuracy of JM-102 value 19.9
 Ethnicity   > 249 micromol/litre   (highest accuracy from ROC curve)
 White: 94.7%      Sensitivity: 86% (81–89%)
               Specificity: 78% (73–83%)
 Prevalence of TSB >   PPV: Not reported
 300 micromol/litre = 3.3% (10/303)   NPV: Not reported

 Exclusion: babies who had previously
 received phototherapy

 Carbonell X;   Study Type:   Healthy term babies   Test:   Correlation of TcB levels with lab TSB   Unselected population but no
    Diagnostic study      1. Umbilical cord blood bilirubin   levels for Sternal vs Forehead site (Pearson   exclusion criterion
 Year: 2001      n = 2004 – 610 in phase one ± 1394 in   (UCB) measured at birth (threshold   correlation coefficient)   Test & Reference test described
    Evidence Level:   phase 2   value: ≥ 37 micromol/litre)      in detail
 Country:   II   ROC curve used to find the best   At < 24 hours (n = 120)   Reference test a standard one
 Spain      mean BW 3230 ± 491 g   cut-off value of UCB.   Sternum Forehead   Test and reference test carried
       mean GA 39 weeks      0.81 0.77        out within one hour
 26    Gender: Males = 50.7%   2. TSB (in phase 1 & 2) and TcB      Blinding – not specified
       (phase 1 only) measured at   At 24–48 hours (n = 126)
    Ethnicity   24 hours (threshold value for   Sternum Forehead
 Not reported   TSB = 102 micromol/litre and for   0.89 0.83
    TcB > 11)
 In first phase (n = 610), cord blood      At > 48 hours (n = 412)
 bilirubin (UCB) at birth and TcB with   3. TSB and TcB (in phase 1 & 2)   Sternum Forehead
 Minolta JM-102 measured at 24 hours,   measured at 48 hours (threshold   0.94 0.83
 48 hours & 60–96 hours of life.   value for TSB = 154 micromol/litre
 Additionally TSB was done for all at 60–  and for TcB > 13)   Diagnostic accuracy of TcB for detecting
 96 hours. On 169 babies TSB also      TSB > 222 micromol/litre
 measured at 24 & 48hours   TcB reading using Minolta JM-102


                                                                 229
   308   309   310   311   312   313   314   315   316   317   318