A critical report in the Tampa Bay Times Saturday, which is essentially saying the Florida Department of Health (DOH) is “cooking the books” concerning HIV data in the state, was quickly responded to the state health officials the same day.
Health officials say the claims by the newspaper are completely false and accuse the media organization for knowingly compared data that was not suitable for comparison to reach its conclusions.
“Preventing transmission of HIV in Florida is extremely important to the department which is why we actively work with communities to conduct targeted outreach and testing and to link patients to treatment quickly,” said Interim State Surgeon General Dr. Celeste Philip. “Unfortunately this article does not highlight any of the efforts of the department and our numerous community partners and rather presents cherry-picked data to fit a predetermined conclusion. We welcome the opportunity to discuss with the Tampa Bay Times the entire data gathering and deduplication process to clear up the misconceptions presented in the article.”
DOH responds to each claim individually:
“State lawmakers blasted the state surgeon general in January for cutting staff and spending at a time when new HIV cases were spiking in Florida. A month later, the Florida Department of Health quietly revised its figures.”
- The Monthly Surveillance reports dating back to 2008 have been, and remain, publically available on the department’s website here.
- Florida follows national protocols to deduplicate data sets.
- The Centers for Disease Control (CDC) distributes a Routine Interstate Duplicate Review every six months (January and July) to every state for cases diagnosed with HIV in the previous six months that appear to be duplicates.
- States work together to investigate potential duplicates to determine which jurisdiction and year of diagnosis they are in.
- This process is the same across all states – all states numbers are adjusted through this process.
“The department’s division of disease control lowered the number of new HIV cases logged in 2014 from 6,147 to 4,613 — erasing one in four new infections from the rolls that year, state records show.”
- Cases are never erased – they are shifted to the appropriate year of diagnosis or to the state where the individual resides.
- There was a 20 percent adjustment made to the cases of HIV/AIDS in Florida in 2014.
- This is consistent with adjustments made in previous years.
- Typically the number of names the department receives to analyze for deduplication has been about 4,000 names per year, but due to increased use of Electronic Lab Reporting and a change in the way CDC managed the process, Florida received about 10,000 names for the 2014 data.
“But experts say an adjustment that large raises questions, even accounting for duplicate cases.”
“But the most recent revision was the largest of any dating back to 2009, according to a Tampa Bay Times analysis of available reports.”
- The department informed the Times that the Monthly Surveillance Reports should not be used to analyze trends.
- Monthly surveillance reports are a snapshot in time and because we know they include duplicates – but not necessarily how many – they should not be relied upon for trend analysis or to be compared to each other.
- The number of duplicates can vary from month to month and year to year.
- HIV data should be viewed as living data and is constantly being updated.
- It is not accurate to compare previous year’s data from older surveillance reports. You should always compare the most current data to the year-end report to analyze trends from previous years.
“The changes weren’t limited to the new cases from 2014. State health officials revised the counts dating back to 2005, reducing the overall number of cases diagnosed over the past decade by more than 7,300.”
- The department does not stop analyzing data simply because a calendar year ends.
- Any necessary adjustments are made to data regardless of year.
- It is critical for the national data set to be unduplicated to ensure resources are properly allocated. States work together, along with CDC, to ensure that cases are appropriately assigned to the correct state.
“Gambineri said her department found more duplicates than usual this year because of changes in how data was collected across the country.”
- The Times misrepresented the department’s position.
- The department does not find duplicates.
- Like every state, Florida receives a list of potential duplicates to investigate from the CDC.
- There are thousands of potential duplicate cases to sort out and verify each year. The number of duplicate cases to sort out has increased over time as technology has enhanced reporting practices.
“In addition, she noted that earlier this year, Florida made a change to the way it reports new HIV cases. Instead of counting cases by the year they were reported, the state now tallies cases by the year they were diagnosed, she said.”
- The data provided to the CDC has not changed.
- All states use the Enhanced HIV/AIDS Reporting System (eHARS), a browser-based, CDC-developed application that assists health departments with reporting, data management, analysis, and transfer of data to CDC.
- Information in eHARS is reported by ‘year of report’ – which can be different from ‘year of diagnosis’
- The data by ‘year of report’ that is found in Table 1 of the Monthly Surveillance Report each month is useful to gauge the activity of HIV cases initially reported in Florida for the first time. It is marked as “provisional” as they are not the final year-end numbers. Once those cases are de-duplicated, data by ‘year of diagnosis’ should be used as it more accurately reflects the trends of the epidemic.
- Beginning January 2016, the provisional notations were added to the Monthly Surveillance Report to help users better understand the limitations of the data by year of report.