Andrew Evans remembers when Lyme disease first began to emerge in the local consciousness.
In the early 1990s, Evans served as a Dutchess County public health adviser whose job it was to investigate how many cases were afflicting county residents.
“The first year I started doing surveillance, I remember getting 600 (Lyme) rash reports in one month,” said Evans, who now oversees the county’s disease-count investigations. “It was really intense.”
Two things were happening: Ticks with the Lyme pathogen were increasing at alarming rates, and so was local awareness about the disease. As a result, Hudson Valley counties became known for having the highest Lyme infection rates in the nation.
That is still the case, though you would never know it from federal statistics. That’s because in 2009, the state health department began allowing counties to review just a portion of laboratory tests for Lyme and then estimate the total number of cases.
As a result, counties like Dutchess and Ulster have fallen far down in the national rankings of case counts and per-capita rates of infection, according to a Poughkeepsie Journal investigation. That trend has alarmed advocates who count on federal and state funds to pay for research and resources.
“We are going to lose those dollars,” said Dutchess County Legislator Donna Bolner, chair of the Dutchess County Tick Task Force.
It is unclear whether that is happening. Just where grant money goes depends on each agency’s guidelines and myriad factors, not just the pervasiveness of a disease.
Federal Lyme research grants to New York have held steady in recent years, even as a greater number of counties in the state have been allowed to estimate cases. From 2012 through 2015, New York averaged $2 million a year in National Institutes of Health grants for Lyme, third-best behind Connecticut and Texas, according to a Journal analysis of NIH data.
Just what gets counted is determined by the Council of State and Territorial Epidemiologists, or CSTE. Epidemiologists are public health professionals who investigate the patterns and causes of disease and injury.
By better understanding where and how frequently a disease is affecting people, state and federal agencies can target interventions more strategically.
The definitions are not meant to inform doctors’ treatment of patients, only how disease patterns should be tracked and counted. More than 70 diseases must be reported by physicians.
CSTE’s executive director, Dr. Jeffrey Engel, said estimated cases of Lyme are not counted in order to maintain consistency within the national reporting system.
“In New York state, they are only investigating 20 percent of their (lab results) and then they extrapolate to a full number,” Engel said. “But the national system needs every single case. It’s just not reconcilable.”
The impact has been stark.
Dutchess falls in national rankings
From 2000 to 2008, Dutchess County averaged 1,145 Lyme disease cases a year, tops in the nation by a wide margin, according to CDC data analyzed by the Journal. By comparison, Chester County in Pennsylvania was No. 2 during that period with an average of 699 annual cases.
That all changed in 2009, when Dutchess became one of the first New York counties estimating its cases. In the national rankings, Dutchess went from 1,141 cases in 2008 to just 413 in 2009.
More than half of the 979 cases Dutchess reported to the state health department that year were not included in the national counts because they were derived from estimates. Between 2009 and 2014, Dutchess’ national average was 226 cases a year, dropping the county from No.1 to 36th overall.
“You have a national map that does not reflect true numbers for select counties in New York state,” Engel admitted. “It’s almost like you need an asterisk.”
Ulster County began estimating cases in 2010. The year before, it ranked 10th nationally with 582 CDC-accepted cases. In 2010, it fell to 22nd with 269.
It’s worth noting that in 2008, the CDC began counting probable cases as well as confirmed ones. So while most counties saw an increase after 2008, the opposite happened here.
Concerns over funding for research
Rick Ostfeld, a disease ecologist at the Cary Institute of Ecoystem Studies in Millbrook, said he routinely includes national-level statistics in his grant applications to emphasize the importance of the Lyme problem. He also includes county or other more local statistics to underscore that Dutchess is an appropriate location for the proposed work.
“If official statistics seriously underestimated the true extent of the public health problem in our region, then this could definitely reduce the probability that I’d convince peer-reviewers that this is an appropriate site in which to conduct our research,” Ostfeld said.
He said he is not aware of any grant request that has been undermined by the lower, local statistics.
But since fewer than 10 percent of proposals submitted to some divisions of the National Science Foundation are funded, “almost any perceived weakness can push an otherwise competitive proposal into the ‘do not fund’ category,” Ostfeld said.
How cases are counted
The CSTE traces its roots back to 1955, when people with fevers often were being diagnosed with either typhoid or malaria. In many situations, doctors would describe the condition as typho-malaria.
“That was totally unhelpful because malaria is a mosquito-borne infection and typhoid is a waterborne infection,” said Engel, the CSTE executive director. “So the states didn’t know what to do.”
CSTE panels work with the CDC to craft case definitions. The CSTE votes on the final case definition and sends it for review to the CDC.
“Usually it is accepted,” Engel said, “but sometimes it is not and there is a negotiation and it is revised.”
For a Lyme case to count as “confirmed,” one of three things must happen:
A patient must present with the erythema migrans, or bull’s-eye, rash and have visited a wooded, brushy or grassy area in the past 30 days.
A patient who did not visit those areas must have the rash and laboratory evidence of infection.
A patient must have some of the later Lyme disease symptoms such as muscle pain and aching joints along with laboratory evidence of infection.
A case is counted as “probable” whenever a doctor diagnoses Lyme disease and there is laboratory evidence of infection.
When a county health department learns of a Lyme case, either through a lab report or a doctor’s diagnosis, staffers investigate whether all the pieces are present to count it either as confirmed or probable.
State adopts new method
In places where there is a lot of Lyme, that can mean a lot of follow up calls with physicians, especially when you consider there are dozens of reportable diseases.
From 2003 to 2005, Dutchess averaged more than 5,200 Lyme investigations each year, according to the state health department.
“It was really intense,” said Evans, the Dutchess public health advisor.
In 2009, New York began allowing four counties — Dutchess, Orange, Suffolk and Westchester — to investigate 20 percent of the lab results and then estimate a total count based on the smaller sample. (All physician-reported cases are investigated.) Now, 25 counties use the sampling method.
Engel said New York’s lower numbers in the national rankings do not compromise the primary mission of epidemiologists — to gain a strategic understanding of where problem areas are.
“A surveillance system really is in place not just to count, it’s in place to protect,” he said. “If the 20 percent estimation is doing the job to protect New York, then it’s good enough for them. Those in the know, who read the national counts, will know to correct for New York.”
But he acknowledged the statistics could affect some funding sources, something local advocates say is already happening.
Bolner, the chair of the county Tick Task Force, said she diverts some of her legislator’s salary to pay for 2,000 tick-removal kits each year that are handed out at various events.
“That,” she said, “is where under-reporting hurts us.”