Page 309 - 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
Area = 0.92

Crofts DJ; Study Type: Mothers and their newborn babies born Phase 1: Inspection of stools, by Incidence of jaundice Report of a community
Non-diagnostic and resident of Sheffield and who were parents, from healthy babies and programme (non-diagnostic
Year: 1999 study (Project routinely visited by the health visitor at babies with cholestatic liver disease Related to breastfeeding study) Unselected population
report) 28 days of age. during the first 28 days of age to 3.4% (95% CI 2.9%, 4.1%)
Country: UK devise a stool colour chart using 20 No demographic details
Evidence Level: Phase 1: (n = 109 parent-baby pairs, total colours At 28 days in breastfed babies reported
62 III stool observations = 5053) 9.2% (95% CI 7.8%, 11.0%)
Mean BW: Not reported Phase 2: development of stool chart
Mean GA: Not reported – six most commonly selected stool % with abnormal LFT (n = 60)
Gender: Males = 56.9% colours from each of main colour
groups together with three pale Abnormal GGT and ALT
Ethnicity: colours used to develop a stool 38.3% (23/60)
Not reported chart.
Abnormal Alk. phosphate
Phase 3: (n = 3629 mother-baby pairs) Phase 3: Assess specificity of 70% (42/60)
colour chart – charts given to all
mothers at first health visitor visit Reasons for non-referral of babies with
(at 10–14 days), and information prolonged jaundice (n = 14)
collected at second visit of health
visitor (at 28 days). 9 = babies well and thriving
Babies with suspicion of jaundice 2 = confusion between midwife and health
or history of passing pale stools visitor
referred for further investigation 2 = family moving out
1 = refusal

Bilgen H; Study Type: Healthy term babies with jaundice aged 1) Ingram Icterometer on the nose Prevalence of TSB > Selected population
Diagnostic study more than 1 day but less than 5 days in a 220 micromol/litre = 18% (17/96) Test & Reference test not
Year: 1998 hospital. Threshold: reading ≥ 33 for best described in detail
Evidence Level: accuracy results Comparison 1: Test and reference test carried
Country: II n = 96 out within one hour
Turkey mean BW 3380 ± 419 g 2) TcB using Minolta JM-102 on Correlation of JM-102 with lab TSB levels Blinding – yes
the forehead (Pearson correlation coefficient, n = 96)
66 mean GA: 39.6 ± 1.4 weeks r = 0.83, P < 0.01
age at presentation: range 1 to 5 days Threshold: reading > 13 for best
Gender: Males = 58% accuracy results Diagnostic accuracy for detecting TSB
> 220 micromol/litre
Ethnicity: Reference standard: Laboratory Sensitivity: 100% (17/17)
Not reported TSB levels within 30 min Specificity: 55.7% (35/79)
PPV: 32.7% (17/52)
Exclusion: not received phototherapy NPV: 100% (44/44)

Comparison 2:

Correlation of Icterometer with lab TSB
levels (Pearson correlation coefficient,

227