Fluoridation Statistics — Population Methodology Changes
To maintain data integrity, CDC conducted an extensive review of data quality and the methodology that community water systems (CWS) used to estimate population data. CDC took time to investigate, evaluate, derive solutions, and implement new methods to address issues related to population controlling.
State-reported CWS populations are usually based on the number of accounts or connections, with an estimated number of people associated with each account. Because the population estimation methods used by states may under- or over-estimate the CWS population, CDC adjusts state-reported CWS population estimates to more closely reflect the population on CWS in that state.
Through the process of evaluating data quality and methodology, CDC determined that it was necessary to update the population controlling methods to create more accurate estimates. Before 2016, state-reported data were only adjusted if the state-reported CWS population exceeded the US Census Bureau state population estimate. This method had two limitations. The first was that the correction it produced was still inflated (no state CWS population is 100% of the total population of that state). The second was that there was no adjustment upward if a state underreported the CWS population.
The most noticeable impact of the methodology change is that the estimated population on CWS is now lower both nationally and in many states. The impact on specific estimates include:
- US population on CWS, people – This estimate is now lower, as a number of reported state CWS populations were previously overestimated.
- US population on fluoridated drinking water systems, people – This estimate is now lower, as a number of reported state CWS populations were previously overestimated, and populations served by each CWS were controlled proportionately in each state.
- Percentage of US population receiving fluoridated water – This estimate is now lower as “US population on fluoridated drinking water systems” (numerator) is lower as a result of the methodology change, but “total US population” (denominator) remains the same./li>
- Percentage of US population on CWS receiving fluoridated water – This estimate should be similar to numbers produced using the prior controlling method. The related variables, “US population on fluoridated drinking water systems” (numerator) and “US population on CWS” (denominator), were similarly affected by the changes to the methodology. However, this number may not be exactly the same as with the previous method since CWS numbers are controlled at the state level and added together to provide a national estimate.
Population served by CWS with naturally occurring fluoride at or above optimal levels – This estimate is now lower, as a number of reported state CWS populations were overestimated, and populations served by each CWS were controlled proportionately in each state.
State and water system estimates:
Population estimates for states may also change, if the estimated CWS populations reported to WFRS were higher or lower than the estimates from the Census Bureau State Population Estimates and the USGS estimate for the Percentage of State Population served by Public Supply. A state’s percentage of population on CWS receiving fluoridated water will not change because the population receiving fluoridated water (numerator) and the population on CWS (denominator) are adjusted by the same method at the state level. The water system population estimates in each state will be proportionally controlled.
The estimates from 2000 to 2014 were calculated using the best data available and methods chosen at that time. While the new methods are believed to improve the accuracy of these estimates, that does not necessarily mean that the older estimates are invalid and should not be used. The population estimates using the current methodology are not directly comparable to previous years’ estimates, but the percentage of state populations on CWS receiving fluoridated water will remain comparable across the old and new methodologies.