NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.

Ip S, Glicken S, Kulig J, et al. Management of Neonatal Hyperbilirubinemia. Rockville (MD): Agency for Healthcare Research and Quality (US); 2003 Jan. (Evidence Reports/Technology Assessments, No. 65.)

  • This publication is provided for historical reference only and the information may be out of date.

This publication is provided for historical reference only and the information may be out of date.

Cover of Management of Neonatal Hyperbilirubinemia

Management of Neonatal Hyperbilirubinemia.

Show details

Meta-Analyses

Results of the meta-analyses of diagnostic test performance are presented in this chapter. We used the summary ROC method (Figure S1, S2) and a random effects model to combine independently the sensitivity and specificity values across studies (Table S1S3). The methodologies for performing meta-analyses were described in Chapter 2.

SROC and Combined Sensitivity and Specificity Analyses

Figure S1. SROC curve of AirShields Minolta bilirubinometer Measuring at Forehead as a Screening Tool for Serum Bilirubin Level ≥ 13 mg/dL in Healthy Infants with GA ≥ 34 weeks not on Phototherapy or Exchange Transfusion

Image er-neonatalf10.jpg

Table S1. Pooled Estimates of the Test Performance of AirShields Minolta TcB measurements to Serum Bilirubin Level of 13 mg/dL at Forehead in Healthy Infants with GA ≥ 34 weeks not on Phototherapy or Exchange Transfusion

Study Characteristics and Pooled Results
#StudyTP/FNFP/TNSens95% CISpec95% CI1/Var.Threshold
1Lin12/657/2300.670.41–0.860.800.75–0.843.9427Var
2Bilgen17/035/441.000.80–1.000.560.44–0.670.856013
3Knudsen, 199330/028/151.000.88–1.000.350.21–0.510.857715
4Tsai19/216/1410.900.68–0.980.900.84–0.942.137316
5Bhutta35/528/350.880.72–0.950.560.43–0.683.808717
6Knudsen, 199024/466/1130.860.67–0.950.630.56–0.703.671117
7Harish24/114/210.960.78–1.000.600.42–0.761.477318
8Schumacher16/120/690.940.69–1.000.780.67–0.851.528420
9Karrar36/1311/950.730.59–0.850.900.82–0.945.125621
10Taha11/158/960.690.42–0.880.920.85–0.962.575922.5
11Maisels7/05/1451.000.59–1.000.970.92–0.990.684724
Total (Range)26812920.67–1.000.35–0.97
REM Pooled0.850.77–0.910.770.66–0.85

Var = TcB ≥ 11, ≥ 16, or ≥ 20 on the 1st, 2nd, and following days of life

Table S2. Pooled Estimates of Test Performance of AirShields Minolta TcB measurements to Serum Bilirubin Level of 11 mg/dL at Forehead in Healthy Infants with GA ≥ 34 weeks not on Phototherapy or Exchange Transfusion

Study Characteristics and Pooled Results
#StudyTP/FNFP/TNSens95% CISpec95% CI1/Var.Threshold
1Dai8/011/211.000.63–1.000.660.47–0.810.723217
2Sheridan-Per4/013/401.000.40–1.000.750.61–0.860.625920
3Fok16/064/1221.000.79–1.000.660.58–0.720.870622
Total(Range)282711.00–1.000.66–0.75
REM Pooled0.950.77–0.990.670.61–0.73

Figure S2. Summary ROC of AirShields Minolta bilirubinometer to Serum Bilirubin Level of 11 mg/dL at Forehead in Healthy, Term or Near-Term Infants not on Phototherapy or Exchange Transfusion

Image er-neonatalf11.jpg

Table S3. Pooled Estimates of Test Performance of AirShields Minolta TcB measurements to Serum Bilirubin Level of 15 mg/dL at Forehead in Healthy Infants with GA ≥ 34 weeks not on Phototherapy or Exchange Transfusion

Study Characteristics and Pooled Results
#StudyTP/FNFP/TNSens95% CISpec95% CI1/Var.Threshold
1Knudsen 199221/738/720.750.55–0.890.650.56–0.744.62459
2Knudsen 199068/2721/910.720.61–0.800.810.73–0.889.296910.2
3Maisels20/214/1210.910.70–0.980.900.83–1.942.111120
Total (Range)1453570.72–0.910.65–0.90
REM Pooled0.760.64–0.850.800.63–0.91

Meta-analysis of Correlation Coefficients

The results of the meta-analysis of correlation coefficients should be interpreted carefully due to several limitations:

  • The correlation coefficient does not provide any information about the clinical utility of the diagnostic test.
  • Although correlation coefficients (r) measure the association between transcutaneous bilirubin and “standard” serum bilirubin measurements, the correlation coefficient is highly dependent on the distribution of serum bilirubin in the study population selected.
  • Correlation measures ignore bias and measure relative rather than absolute agreement.

The analytic issues, which arose regarding potential duplication of information, were described in Chapter 2, Methodology.

The meta-analysis of correlation coefficients figures on the following pages indicate:

  • Abbreviated name of the measurement site/metric being analyzed (e.g. forehead)
  • Article citation (author and year)
  • Number of subjects used in the comparison
  • Graphic description of the individual effects and 95% confidence intervals
  • Combined random effects model estimate and its confidence bounds

Figure S3. Overall Meta-Analysis of Correlation Coefficients for Minolta AirShields bilirubinometer

Image er-neonatalf12.jpg

Figure S4. Meta-analysis of Correlation Coefficients for Minolta AirShields Bilirubin Measurements on Different Measurement Sites

Image er-neonatalf13.jpg

Figure S5. Effect of Gestational Age or Birth Weight

Image er-neonatalf14.jpg

Figure S6. Effect of Race or Skin Color Measured at Forehead

Image er-neonatalf15.jpg

Figure S7. Effect of Phototherapy

Image er-neonatalf16.jpg

Figure S8. Overall Meta-Analysis of Correlation Coefficients for BiliCheck™ Measurements

Image er-neonatalf17.jpg

Figure S9. Overall Meta-Analysis of Correlation Coefficients for Ingram Icterometer Bilirubin Measurements

Image er-neonatalf18.jpg

Views

  • PubReader
  • Print View
  • Cite this Page

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...