Vaccine - (Page 5) 6160 Table 1 Quality measures for the different adjusted models Residual degree Model BN0 Model BN1b Model BN2c Model BN3d Model BN3e Model BN3f Model BN4g a b c d e f g a M. Oviedo et al. / Vaccine 26 (2008) 6157–6164 Residual deviance 61.35 60.97 58.33 54.74 55.67 54.82 54.17 AIC 522.61 510.68 479.79 477.52 485.13 481.16 493.58 p – <0.01 <0.01 0.03 – – <0.01 Overdispersion parameter ( ) 4.16 5.96 11.18 11.36 9.72 10.16 7.12 Standard error S.E. ( ) 1.11 1.09 1.33 1.21 1.30 1.13 1.17 51 49 48 47 47 47 49 Incidence adjusted for: year, age group and population as offset parameter. Incidence adjusted for: variables model BN0 and sinusoidal terms (amplitude: 6 years). Incidence adjusted for: variables model BN0, sinusoidal terms (amplitude: 6 years) and coverage vaccination. Incidence adjusted for: variables model BN0, sinusoidal terms (amplitude: 6 years), coverage vaccination and interaction year with coverage vaccination. Incidence adjusted for: variables model BN0, sinusoidal terms (amplitude: 5 years), coverage vaccination and interaction year with coverage vaccination. Incidence adjusted for: variables model BN0, sinusoidal terms (amplitude: 7 years), coverage vaccination and interaction year with coverage vaccination. Incidence adjusted for: variables model BN0, coverage vaccination and interaction year with coverage vaccination. from to 1.1 person-years). In the 19–39 years age group, the reduction was from 8.2 to 5.1. Table 1 shows the statistical information required to decide which of the models best adjusts the HVA incidence. The first model (model BN0) included the covariates year and age group, and both were statistically significant (p-values < 0.01). Model BN1 included two sinusoidal variables which allowed the statistical significance of the natural cycle of incidence to be estimated (p-value < 0.01). In order to determine whether the variable vaccination coverage (vac) was necessary, model BN2 was adjusted and the likelihood tests were calculated; vaccination coverage was statistically significant with a p-value < 0.01. Model BN3 also included the term interaction (year × vac). The likelihood ratio test showed that there was a significant interaction between the year of report and vaccination coverage (p-value < 0.01). The other interactions were tested one by one, and none were significant (data not shown). The value of the dispersion parameter for each adjusted model is shown in Table 1 and the sign was positive, indicating overdispersion in the model (p-value < 0.001). The residual deviance and the AIC of the adjusted models was reduced when adjustment was made for vaccination coverage (vac) and the cycling component adjusted by amplitude of the cycle to 6 years, indicating that including a new variable was not overadjusting the model and that the reduction was significant. Fig. 2 shows the values observed for the 12–18 years age group and compares them with the curves of the models with and without the effect of vaccination (models BN3 and BN1, respectively). Model BN3, which included the estimate of the vaccination that interacts with the year, correctly adjusted the nonlinear trend in the incidence rate in the 12–18 year age group. When the estimates of the two models were compared, the introduction of vaccination in the model reduced the estimated incidence in the 12–18 years age group by 62.75% with respect to the model without the estimate of vaccination (BN1), but this reduction was different according to the year; in 1999 the incidence with respect to the model without coverage was reduced by 12.87%. The reduction increased to 46% in 2000, 63% in 2001, 71.9% in 2002, 75.4% in 2003 and decreased smoothly to 74.5% in 2004 and 68.6% in 2005. Table 2 shows the values of the estimated parameters with their corresponding p-values. All variables introduced in the model were statistically significant with p-values < 0.05. The negative signs of Fig. 2. Temporal evolution of the reported incidence of hepatitis A in Catalonia in 12–18-year-old children and estimated incidence with and without the effect of vaccination coverage.
Table of Contents Feed for the Digital Edition of Vaccine Vaccine Vaccine - (Page Cover1) Vaccine - (Page 2) Vaccine - (Page 3) Vaccine - (Page 4) Vaccine - (Page 5) Vaccine - (Page 6) Vaccine - (Page 7) Vaccine - (Page 8) Vaccine - (Page 9)
For optimal viewing of this digital publication, please enable JavaScript and then refresh the page. If you would like to try to load the digital publication without using Flash Player detection, please click here.