Debate focuses on if estimates or hospital visits better reflect contamination effects
Local health official Dr. Rahul Gupta and environmental engineer Andrew Whelton are confident in their estimates that close to 100,000 people suffered some negative health effect following the chemical leak and water contamination.
Their conclusions are estimates though, and vary substantially from the data collected by the state, said Department of Health and Human Resources spokeswoman Allison Adler.
“The data that the West Virginia Bureau for Public Health gathered during the water crisis were actual medical charts on patients who visited Emergency Departments of hospitals in the affected area,” Adler said.
“The information is actual numbers, not estimates.”
The state started collecting health information in the days following the discovery of thousands of gallons of chemicals leaking from a faulty storage tank into the Elk River. The chemicals overwhelmed filters at the local water plant, contaminating tap water for 300,000 people across a nine-county region over the course of several days.
When DHHR stopped collecting data in late January, 26 people had been admitted to hospitals and 533 treated in hospitals complaining of symptoms they believe were connected to the chemical leak.
However, Gupta and Whelton argue there were many people who suffered from symptoms that they considered bothersome but not threatening to the degree they warranted a trip to the hospital.
“Simply relying on hospital records to predict the public health impacts may be extremely misleading,” said Whelton, who’s also leading a team of scientists researching other effects of the spill at the behest of Gov. Earl Ray Tomblin.
Adler said “it is possible” people other than those who went to the hospital suffered from symptoms they believe were related to the spill.
“This is one of the reasons our Bureau for Public Health sent out letters to more than 800+ medical providers to collect additional data in the event they visited their physician with symptoms from the spill,” Adler said in an email.
The bureau doesn’t expect to have the results of that request for information until June, and is waiting on other health survey data as well, Adler said.
Gupta and Whelton each explained Tuesday how they arrived at their estimates. Gupta used a type of survey that looked at 10 different physicians’ offices in the affected area, he said during an online seminar hosted by the National Association of County and City Health Officials.
He took the number of patients who complained of symptoms they believed connected to the spill and compared it to the total number of patients treated at the 10 clinics over the same time frame. He then extrapolated the data to determine the number of people he expected who went to a doctor, about 37,000.
Gupta then took the ratio of total patients who visited the 10 clinics compared to those visiting with reported chemical leak symptoms (about 2.5) and multiplied that number by the 37,000. The resulting number, about 93,000, is the number of people Gupta estimates experienced adverse health effects from the leak.
Whelton used a similar method, but relied on data he collected during a substantially smaller survey that looked at 16 homes. He arrived at an estimate of about 108,000 people possibly affected by negative health effects from contaminated water.
Gupta said he’s confident his data used only unique patient visits — as opposed to double counting one patient who possibly sought medical treatment at several of the 10 clinics examined.
Both Gupta and Whelton acknowledged their numbers are estimates, but said they’re still important measures that need to be considered to understand the complete health effect from the leak and contamination on the region.
“Clearly it is an estimate,” Whelton said in an email Tuesday.
“As more data become available from reputable sources the bracketing of the population affected will become clearer. More work is needed.”
In addition to Gupta, other local health officials presented information about their response to the leak.
Dr. Elizabeth Scharman, director of the West Virginia Poison Control Center, said she thought the center carried out its emergency plan very well. Phone lines were flooded, but that’s to be expected, she said.
The control center — and other health experts — had problems chasing down accurate and relevant medical data about MCHM, the main chemical contained in the faulty tank. While some state and federal agencies relied on the material safety data sheet — a form each chemical maker is required to produce for its products that outline safety and chemical component data — Scharman repeated her views about the dangers of relying on that information.
“People use them because (the sheets) are required,” Scharman posted in the online seminar.
“Many do not know that they are not regulated which is the problem (they believe since they are required they are required to be accurate).”
She also lamented about the challenges of conducting long-term medical monitoring. The bureau is tasked with examining how to conduct such surveillance, but Adler and other state health officials have said they’re waiting on all relevant health data they can find before considering how best to move forward.
A key factor in that decision is the ability to definitively connect the chemical leak with any or all of the reported conditions recorded at the hospitals. After the spill the U.S. Center for Disease Control and Prevention collected information from the medical charts of the applicable patients to determine if there was a connection.
The state is still waiting on that report.
CDC spokeswoman Bernadette Burden said Monday the analysis is still ongoing. She acknowledged receiving a question about the 100,000 people estimated to have been affected by the leak, but did not respond with a comment.
Contact writer Dave Boucher at 304-348-4843 or firstname.lastname@example.org. Follow him at www.Twitter.com/Dave_Boucher1.