Ranking states' immunization coverage: an example from the National Immunization Survey.

Area of Science:

• Public Health
• Biostatistics
• Epidemiology

Background:

• The National Immunization Survey (NIS) provides state-level preschool immunization coverage estimates.
• Ranked lists of these coverage estimates are common but often over-interpreted.
• Point estimates can obscure the uncertainty inherent in rank comparisons.

Purpose of the Study:

• To highlight the challenges in interpreting ranked point estimates from immunization surveys.
• To demonstrate the uncertainty associated with state-level vaccination coverage ranks.
• To advocate for improved methods for presenting and interpreting survey data.

Main Methods:

• Utilized parametric bootstrap methods to generate 90% confidence intervals for state vaccination coverage ranks.
• Employed graphical methods to compare individual states against a reference state.
• Focused on preschool children's vaccination coverage levels.

Main Results:

• Confidence intervals for ranks demonstrate significant uncertainty in state-level comparisons.
• Graphical comparisons reveal the potential for misinterpretation of simple ranked lists.
• The variability in ranks suggests that precise ordering may not be statistically supported.

Conclusions:

• Ranked lists of immunization coverage should be interpreted with caution due to inherent uncertainty.
• Presenting confidence intervals for ranks is crucial for accurate data interpretation.
• Graphical comparisons enhance understanding when comparing state vaccination coverage levels.