Page 333 - 16Neonatal Jaundice_compressed

Appendix H: Evidence tables

Bibliographic details Study type & Patient characteristics Test, Reference Standard, Results Reviewers Comments
Evidence level Threshold for a positive test
TcB measurement r = 0.82, P < 0.001 Altman analysis for TSB - TcB
n = 261 Knee
mean BW 2521 g - range 680 to 4645 g, Diagnostic accuracy of TcB from r = 0.77, P < 0.001
mean GA 34.6 weeks - range 25 to forehead (threshold ≥ 0.70 of Foot
st
43 weeks postnatal age at 1 TcB: 98.4 - phototherapy limit) estimated for r = 0.51, P < 0.001
range 48 – 840 predicting TSB levels ≥
Gender: Males = 60.1% phototherapy limits as suggested by On comparing correlation coefficient of
the Danish Pediatric Society forehead with that for sternum, knee and
Ethnicity: foot, P < 0.001 for each of the comparison

Non-northern European descent = 9%
Group 2:
Forehead
Group 2: Healthy term and near-term r = 0.87, p > 0.05
infants with GA ≥ 35 weeks in the Sternum
r = 0.90, P < 0.05
maternity ward
Knee
n = 227 r = 0.83, P < 0.05
mean BW 3362 g - ange 2170 to 5000 g Foot
mean GA 38.6 weeks - range 35 to r = 0.67, P < 0.001
43 weeks On comparing correlation coefficient of
st
postnatal age at 1 TcB: 74.4 - range 48 forehead with that for sternum, knee and
– 360 Gender: Males = 55.5% foot, P < 0.05 for comparison with knee and
foot only
Ethnicity:
Non-northern European descent = 7% Diagnostic accuracy of TcB (threshold value
> 0.70 times the phototherapy limit) from
Exclusion: forehead in detecting TSB > phototherapy
babies already receiving phototherapy or limit
who received phototherapy 6 hours
before TSB measurement, Group 1 (n = 504 observations):
with skin infection, Sensitivity: 108/109 (99.1%)
purpura, Specificity: 177/395 (44.8%)
bruising PPV: 108/326 (33.1%)
NPV: 177/178 (99.4%)

Group 2 (n = 317 observations):
Sensitivity: 3/3 (100%)
Specificity: 254/314 (80.9%)
PPV: 3/63 (4.8%)
NPV: 254/254 (100%)

Samanta S; Study Type: All babies > 33 weeks in the postnatal TcB using BiliChek (site not Correlation of TcB levels with lab TSB Unselected population
Diagnostic study ward of a regional teaching hospital who specified) – single measurement levels (Pearson correlation coefficient, Test & Reference test described
Year: 2004 were due to have blood taken for TSB taken. n = 300) adequately
Evidence Level: estimation Test and reference test carried
Country: II Reference standard: Laboratory r = 0.77, P < 0.0001 out within one hour
UK n = 300 TSB levels taken concurrently with Blinding – not specified

239