CHARLESTON, W.Va. -- Each day, federal and state officials and West Virginia American Water are telling West Virginians that water testing shows levels of the toxic chemical Crude MCHM are dropping -- in many cases to levels they aren't able to detect.On Wednesday, for example, West Virginia American issued a news release that touted "results showing non-detectable or extremely low levels" of the chemical "in water samples gathered systematically through the" distribution system."Data points collected by our interagency team over the past few days indicate decreasing levels of MCHM," stated water company President Jeff McIntyre. "The majority of samples are reading non-detectable. In areas where sample results show levels above the non-detectable limit, they are still extremely low and only a fraction of the CDC-established 1 ppm health-protective limit."However, outside experts have a variety of questions about the testing being done in the water system that serves 300,000 West Virginians in Kanawha and surrounding counties.For one thing, West Virginia officials for many days did not list with the test results any information about what concentration of the chemical the state's process was able to detect. Officials were posting results that listed "non-detect," but without knowing how low the state's test could go, that designation meant nothing, the outside experts said.On Saturday, Gov. Earl Ray Tomblin issued a statement that identified the state's detection limit: 10 parts per billion.That means that state testing, coordinated by the National Guard, could detect down to levels that were one one-hundredth of the 1-part-per-million public-health screening level devised by the U.S. Centers for Disease Control and Prevention, experts say.While the CDC has said levels below 1 part per million are "not likely to be associated with any adverse health effects," the agency also said on its website that, "there should be no MCHM in drinking water." Also, an advisory that warned pregnant women to drink only bottled water remains in effect in the nine-county area affected by the leak.Officials from the Louisville Water Company, though, have said they used a method that allowed them to detect lower levels -- down to about 1 part per billion -- as the MCHM plume passed through their city in the Ohio River.Glenys Webster, a Simon Fraser University epidemiologist, said differing detection limits are important in understanding how much MCHM is really in the water and what the potential health impacts could be."Ideally, the water would be tested using the method with the lowest detection limits," Webster said. "Saying that levels are below the detection limit means that they are un-measurable using a specific method. It does not mean the concentration is zero or that the level is low enough to be 'safe.' Using another method with a lower detection limit might find measurable levels in the same sample."If the only goal is to determine if the chemical levels are greater or less than 1 part per million, the West Virginia method can do that," Webster said. "But, if people want to know how much chemical is actually in their water, the Louisville method would be the better choice."Federal and state officials have not explained why they are using the testing method they are using or how it's different from Louisville Water's method.Pressed about the matter earlier this week, Tomblin said, "I'm not a scientist," adding that whether the detection limit is 10 parts per billion or 1 part per billion, the concentration involved "is still miniscule."Dr. Vikas Kapil, chief medical officer for the CDC's National Center for Environmental Health, said he wasn't familiar with the different sampling methods being used and added a remark similar to Tomblin's."There are always these differences between methods," Kapil said. "When you get down to numbers that low, to some extent, I really don't think it's an issue."The Morgantown-based consulting firm Downstream Strategies has been following the chemical leak closely, and is one of several companies that will perform home water testing, for a fee, to residents concerned about the leak.Downstream Strategies is doing sampling that goes down to the same 10 parts per billion as the state, and similar to what other firms have reported for private home testing done in the wake of the leak.Marc Glass of Downstream Strategies said his firm is working with laboratories that use gas-chromatograph and mass-spectrometer technologies to concentrate samples prior to analysis. Glass said he understands that Louisville Water is using a different method, called direct purge and trap.Both methods use U.S. Environmental Protection Agency-approved protocols, Glass said, but chemists performing such analysis "have differing opinions" about which procedures are more accurate for MCHM.Ben Gilmer, a project manager with Downstream Strategies, said the limited information about state sampling that's been released has made it hard to analyze the results and draw many firm conclusions.For example, Gilmer said, complete information about the locations of sampling and timing of sampling related to "flushing" the system have not been made available. And, many samples appear to have been taken only at publicly available locations, such as fire hydrants, which might not necessarily represent the same water quality as that coming from a kitchen sink. It's also not clear if tests are being done only for Crude MCHM or also for any of that chemical's ingredients."We are really trusting the water company and the National Guard at this point," Gilmer said. "Until we receive more detailed descriptions of their sampling approach, and until we are able to map all existing test results to specific locations, and compare those tests with the region's water infrastructure and water-use advisory zones, it will remain difficult for outside experts to judge how well chemicals are being pushed out of the system."Sen. Jay Rockefeller, D-W.Va., has asked West Virginia American Water to provide more detail about its test results."Many West Virginians are asking for maximum transparency throughout this process and would like to have details about sampling, testing and chemical levels in a timely manner," Rockefeller wrote in a letter to the company. "Please publicly post to your company's website, in real time, the sampling times, coordinates where samples are being taken, test results in [parts per billion], and the substance for which you are testing."Reach Ken Ward Jr. at email@example.com or 304-348-1702.