Skip to main content

Table 2 Misclassification bias and mean squared error (MSE) for self-reported and MIME-corrected hypertension prevalence

From: Combining self-reported and objectively measured survey data to improve hypertension prevalence estimates: Portuguese experience

  INS2014
Bias,
self-reported
(pp)
INS2014
Bias,
MIME-corrected (pp)
INS2014
MSE,
self-reported
INS2014
MSE,
MIME-corrected
Overall 11.5 0.7 0.0132 0.0002
Sex
 Female 5.9 −0.5 0.0036 0.0002
 Male 17.6 2.0 0.0309 0.0007
Age group
 25–44 years 5.0 −1.0 0.0026 0.0003
 45–54 years 13.3 − 1.5 0.0179 0.0007
 55–64 years 18.6 5.7 0.0347 0.0038
 65–74 years 17.0 3.1 0.0291 0.0014
Educational level
 ISCED 2011 levels 0–1 16.1 2.0 0.0261 0.0007
 ISCED 2011 level 2 11.1 −0.1 0.0124 0.0006
 ISCED 2011 levels 3–4 11.1 2.4 0.0125 0.0011
 ISCED 2011 levels 5–8 5.2 0.0 0.0028 0.0003
  1. MSE = variance + bias2
  2. Bias is defined as the absolute difference between the INS2014 and INSEF objectively measured prevalence estimates in percentage points (pp)
  3. INSEF – Portuguese Health Examination Survey, INS2014 – National Health Interview Survey 2014, MIME – Misclassification error correction