Converting exposure proportions calculated from surveys based on one reference period length to another.

Abstract

Commonly used components of health risk assessments and outbreak investigations are the proportion of the general population exposed to potential sources, often coming from national surveys collected with one reference period, such as 7 days. However, risk assessors may need this value for a different period of time (e.g. 1 day, a full year, etc.) and outbreak investigators may need a reference period that matches their case questionnaire. Use of these proportions is biased if the reference period does not match what is needed. Three surveys that collected exposure information with more than one reference group were used to fit six models that were developed to estimate the relationship between proportion and reference period length, and thus, convert proportions collected with one reference period to any desired reference period. A practical implementation of the best model is provided in a spreadsheet for health practitioners, such as those working in risk assessment or outbreaks who need to convert proportions to a new reference period. The conversion method can be used in any subject matter across the sciences, where surveys collect the proportion of respondents who experienced an event in a defined reference period.