T H E U S E O F A P R E D I C T I O N M O D E L T O P R E V E N T A B N O R M A L W E I G H T I N C H I L D H O O D I N T H E Y O U T H H E A L T H C A R E P R A C T I C E To compare our model with other methods (using the weight for length growth chart and calculation of PI) to estimate the growth of children, we calculated the sensitivity and specificity of these methods, all applied at the age of approximately 1 year and using + 2SD as cut-‐off for overweight, shown in Table 1-‐3. The sensitivity of our model is 46% (50/109). In other words, 46% of children with a PIm > PIe + 2SD at the age of 1 year developed overweight at the age of 5 years. The specificity is 84% (93/111). In other words 84% of the children with a PIm < PIe + 2SD at the age of 1 year did not develop overweight at the age of 5 years. The likelihood ratio for positive outcomes (sensitivity / 1-‐specificity) is 2.8 and for neg-‐ ative outcomes (1-‐sensitivity / specificity) 0.6. The diagnostic odds ratio (true posi-‐ tive * true negative / false positive * false negative) is 4.4. The accuracy to predict overweight at the age of 5 years, based on the prediction model applied at the age of 1 year, is 65% (143/220). Table 1: Sensitivity (true positive) and specificity (true negative) of the prediction model (applied at 1 year, for prediction of overweight at the age of 5 years). Overweight Not overweight TOTAL PIm > PIe + 2SD 50 (46%) True positive 18 (16%) False positive 68 PIm < PIe + 2SD 59 (54%) False negative 93 (84%) True negative 152 TOTAL 109 111 220 Se: 46%; Sp: 84%; LR +: 2.88; LR -‐: 0.64; DOR: 4.4 PIe= Expected Ponderal Index, PIm = Measured Ponderal Index, SD = Standard deviation. Se= Sensitivity, Sp= Specificity, LR += Positive likelihood ratio, LR -‐= Negative likelihood ratio, DOR= Diagnostic odds ratio. The sensitivity of the weight for length growth chart at the age of 1 year and the PI calculation at the age of 1 year is 16% and 28% respectively. The specificity values are 95% and 89% respectively. Likelihood ratios for positive outcomes are 3.2 re-‐ spectively 2.5; those for negative outcomes are 0.8 for both models. The diagnostic ratios are 3.6 respectively 3.1. The accuracy of the weight for length growth chart at the age of 1 year and PI calcu-‐ lation at the age of 1 year is 56% and 58% respectively. 95
Proefschrift binnenwerk Manon Ernst_DEF.indd
To see the actual publication please follow the link above