A predictive model for respiratory syncytial virus (RSV) hospitalisation of premature infants born at 33–35 weeks of gestational age, based on data from the Spanish FLIP study

Table 2 Analyses of the predictive accuracy of the various models

	True Positive	False Positive	False Negative	True Negative	Sensitivity	Specificity	PPV %	NPV %	LR	Diagnostic Accuracy %
FLIP 15 variable model ^§	130	102	53	265	0.71	0.72	56	83	2.56	72
FLIP Final 7 variable model ^¤	139	113	45	254	0.76	0.69	55	85	2.45	71
Munich 6 variable model ^†	14	106	4	247	0.78	0.70	12	98	2.59	70

^§ Records for 550 infants were included within the analysis. Seven records were dropped from the analysis due to missing data for one or more of the predictor variables
^¤ Records for 549 infants were included within the analysis. 8 records were dropped from the analysis due to missing data for one or more of the predictor variables
^† Records for 370 infants were included within the analysis. Three records were dropped from the analysis due to missing data for one or more of the predictor variables. Two records for hospitalised cases were removed from the analysis, as they each had one negative RSV test
PPV = positive predictive value
NPV = negative predictive value
LR = likelihood ratio of a positive test; for information about likelihood ratios see reference 25
Standardised canonical discriminant function coefficients for the FLIP final 7 variable model: birth ± 10 weeks of start of season = 0.678, birth weight, kg = 0.184, breast fed ≤ 2 months or not = 0.511, number of siblings ≥ 2 years = 0.489, number of family with atopy = 0.151, female sex = -0.113, number of family with wheeze = 0.125

ISSN: 1465-993X