Skip to main content

Table 3 Random effects of covariates from models on Table 2 at the GP and PHC team level on the odds of being screened

From: Variability in screening prevention activities in primary care in Spain: a multilevel analysis

  Hypertension screening Dyslipidaemia screening Obesity screening Tobacco use screening Alcohol screening
  Variance (SE) Variance (SE) Variance (SE) Variance (SE) Variance (SE)
Between PHC team level      
Intercepta 0.052 (0.030) 0.112 (0.037) 0.139 (0.059) 0.330 (0.092) 0.136 (0.061)
Age slope 0.063 (0.012) 0.019 (0.004) 0.055 (0.010) 0.110 (0.019) 0.085 (0.015)
Sex slope   0.011 (0.003)    
Number of visits slope 0.011 (0.002) 0.018 (0.003) 0.013 (0.002)   0.021 (0.004)
Number of health problems slope 0.011 (0.002) 0.028 (0.005)    0.026 (0.004)
Hypertension slope -   0.052 (0.013)   0.050 (0.014)
Dyslipidaemia slope   -    
Obesity slope 0.035 (0.011) 0.014 (0.005) -   0.011 (0.004)
Smoker slope 0.020 (0.005) 0.017 (0.004) 0.095 (0.017) - 0.020 (0.005)
Between GP level      
Intercepta 0.125 (0.018) 0.088 (0.012) 0.371 (0.036) 0.196 (0.043) 0.400 (0.041)
Age slope 0.056 (0.004) 0.018 (0.003) 0.062 (0.005) 0.108 (0.008) 0.079 (0.006)
Sex slope 0.019 (0.002)   0.015 (0.002)   
Number of visits slope 0.019 (0.002)   0.021 (0.002) 0.064 (0.004) 0.036 (0.003)
Number of health problems slope   0.013 (0.002) 0.021 (0.002)   0.015 (0.002)
Hypertension slope - 0.067 (0.008) 0.054 (0.007) 0.160 (0.014) 0.074 (0.008)
Dyslipidaemia slope 0.029 (0.004) - 0.042 (0.005) 0.053 (0.006) 0.030 (0.004)
Obesity slope 0.048 (0.007) 0.014 (0.004) - 0.063 (0.007) 0.038 (0.005)
Smoker slope 0.027 (0.003)   0.047 (0.005) - 0.058 (0.005)
Drinker slope     0.093 (0.028) -
Variances of the empty model b      
Between PHC teams 0.380 (0.088) 0.165 (0.039) 0.319 (0.078) 0.659 (0.151) 0.918 (0.206)
Between GPs 0.229 (0.017) 0.129 (0.010) 0.359 (0.026) 0.357 (0.026) 0.384 (0.028)
PCV a,c      
Between PHC teams 86.3% 32.2% 56.4% 49.9% 85.2%
Between GPs 45.4% 31.8% −3.3% 45.1% −4.2%
  1. Variances estimated in logits.
  2. Abbreviations: GP general practitioner, PCV proportional change in variance, PHC team Primary health care team, SE standard error.
  3. Dash indicates covariate not considered in that analysis.
  4. The models included the variance of the random intercepts and slopes that significantly varied between GP and PHC team results.
  5. aAs the variance is a function of individual characteristics that vary randomly, the values in the table are for the intercepts (for individuals with characteristics at their reference values).
  6. bEmpty model: model with no covariates at either level, i.e. only the intercepts and the random parameters.
  7. cPCV: The proportional change in variance expresses the change in the PHC team or GP level variance between the empty model and the final model.