Water results meet the child health records
Published 13 October 2026
By September 8, Virginia Tech had received 252 of the three hundred kits sent to residents. Its update reported that 101 first-draw samples contained more than five parts per billion of lead, with a ninetieth percentile of twenty-five parts per billion. Several measurements exceeded one hundred, and a sample collected after forty-five seconds of flushing exceeded one thousand. Those are results of the university's participating-home survey at that stage. They are not the state's percentile, and they do not imply that all Flint taps had the same concentration. The team said the remaining forty-eight results could not erase its conclusion that there was a serious problem. The argument rested on a growing collection, not only on the extreme result in one home, and it directly challenged the confidence conveyed by the official monitoring rounds. F025, flintwaterstudy.org, Our sampling of 252 homes demonstrates a high lead in water risk: Flint should be..., 2015-09-08 F090, michigan.gov, Flint Water Advisory Task Force final report, March 2016
The researchers' update moved from reporting measurements to advising precautions. It recommended certified lead-removal filters or its stated flushing procedure before using tap water for drinking and cooking. These were the team's instructions in September 2015, not present-day instructions for Flint residents or a universal guide for every plumbing arrangement. The team also called for an audit of the official sampling rounds. It argued that regulatory sampling focused on the highest-risk homes should have found worse results than the university survey, rather than better ones. Its challenge was specific: examine whether the city's site selection and collection process had followed the law. The problem was not resolved by choosing whichever percentile looked most reassuring. F025, flintwaterstudy.org, Our sampling of 252 homes demonstrates a high lead in water risk: Flint should be..., 2015-09-08
The health department began developing educational material about reducing children's lead exposure in September, responding to coverage of the university water tests. On September 22, Mona Hanna-Attisha, director of the pediatric residency programme at Hurley Medical Center, asked for access to state childhood lead records. The task force says the state shared no data, as had also been the case with an earlier request from Edwards. The next day, staff discussed rerunning the July analysis and obtaining formal epidemiological help. At the same time, an internal message said the governor's-office briefing still did not regard Flint water as an imminent public-health problem. The chronology records MDHHS preparing educational materials while a Governor's office briefing said Flint water was not an imminent public health problem. Data existed, but access and interpretation had not become a single prompt response. F090, michigan.gov, Flint Water Advisory Task Force final report, March 2016
Hanna-Attisha presented findings publicly on September 24. The Detroit Free Press described her comparison of children's blood results before and after the change, including an increase from 2.1 percent to 4 percent in the early presentation and a larger change in some ZIP codes. It reported 1,746 Flint test results compared with 1,640 elsewhere in Genesee County, where no significant increase appeared. Those were the figures reported at the announcement. They should not be silently replaced by percentages from the later peer-reviewed analysis, which used its own stated group and periods. The immediate significance was that a doctor had examined children's blood lead results from before and after the source change and presented them publicly. It supplied another form of evidence, distinct from the compliance programme and from a laboratory corrosion test. F072, freep.com, Doctor: Lead seen in more Flint kids since water switch, 2015-09-24 F069, freep.com, Flint doctor makes state see light about lead in water, 2015-10-11
The state did not immediately accept the doctor's interpretation. In the same news report, Wurfel pointed to lead results of six and eleven parts per billion and described the change as consistent with earlier fluctuations. A health-department spokeswoman attributed increases to predictable seasonal spikes and said concern could not be assigned to the water. The article also printed the state's different comparison of children tested across two periods. These positions are part of the contemporaneous dispute, with their spokespeople attached. They cannot be turned into a later finding that the doctor's work was wrong. Nor should the doctor's preliminary percentages be merged with the state's figures into a single trend line. The groups, periods and analyses differed, and the argument was about which comparison best captured the change in exposure. F072, freep.com, Doctor: Lead seen in more Flint kids since water switch, 2015-09-24
The task force records an internal shift on the same day as the press conference. Robert Scott wrote that he saw patterns similar to those Hanna-Attisha reported. Four days later, health director Nick Lyon asked for an analysis that would demonstrate blood lead was not out of the ordinary; the task force says no such analysis was produced. On September 29, the Free Press published its own analysis supporting the doctor's findings, the county issued a health advisory and the governor's office contacted the departments about emergency responses. On October 1, the state health department confirmed Hanna-Attisha's analysis. The change in position was therefore a sequence, not an instantaneous announcement. It followed the public presentation, internal recognition, further reporting and a county advisory. F090, michigan.gov, Flint Water Advisory Task Force final report, March 2016
Flint returned to Detroit-supplied water on October 16, 2015. In the task force's chronology, that date closes the period of the interim river source. It does not close the harm or the practical work of making the distribution network reliable. A source can be changed more quickly than the condition of pipes affected by earlier water can be restored. The report's later recommendations about testing, public health and system repair show that the return was treated as an important response, not an erasure of the preceding exposure. The physical switch came after almost a year and a half of exposure to improperly treated water, and lead levels did not fall below the federal action level until late 2016. F090, michigan.gov, Flint Water Advisory Task Force final report, March 2016 F091, epa.gov, EPA Office of Inspector General, Management Weaknesses Delayed Response to Flint Water Crisis, Report 18-P-0221, July 19, 2018
The EPA inspector general's 2018 account explains why the return did not immediately produce low lead results. The replacement water already included corrosion control, but almost a year and a half of improperly treated water had damaged the infrastructure. The audit says lead concentrations continued to rise after the source change. In December, the state reported that Flint was supplementing the purchased water with additional treatment. Lead levels did not fall below the federal action level until late 2016, according to that report. They distinguish ending use of the river source from repairing its consequences within the network, and they explain why a ceremonial or administrative announcement of a return was not enough to certify every resident's tap at once. F091, epa.gov, EPA Office of Inspector General, Management Weaknesses Delayed Response to Flint Water Crisis, Report 18-P-0221, July 19, 2018
In November, the ACLU of Michigan examined the final federal report against Del Toral's June memo. Its commentary described warnings about lead particles loosened by road work or hydrant flushing, but criticised the omission of two earlier concerns: the missing corrosion control and sampling methods that could reduce detected lead. The EPA spokeswoman said the new technical task force would address corrosion control, and that group would also examine pre-flushing. This is an advocacy organisation's comparison, not an impartial verdict by this account. It belongs in the record because it names the issues that remained contested after the river source ended. The argument was no longer only whether a high home sample was real; it concerned how clearly the federal agency's public document explained the treatment and testing failures behind it. F088, aclumich.org, Revised EPA Memo Raises Troubling Questions Amid Flint Water Crisis - ACLU of Michigan, 2015-11-11
The childhood study, accepted in November 2015 and prepared for the 2016 American Journal of Public Health, used a defined comparison. It included children under five whose blood lead was processed through Hurley's laboratory, comparing January 1 to September 15 in 2013 with the same dates in 2015. Its Flint group comprised 1,473 children, divided almost evenly between the two periods. That acceptance followed the public presentation, but preceded the journal issue in which the study appeared. These dates and selection rules explain why the published figures are not automatically identical to the numbers reported at the September press conference. The authors were analysing a stated retrospective group, not all Flint children. F067, uvm.edu, Hanna-Attisha and colleagues, Elevated Blood Lead Levels in Children Associated With the Flint Drinking Water Crisis, AJPH F069, freep.com, Flint doctor makes state see light about lead in water, 2015-10-11
For the published comparison, the proportion of Flint children with elevated blood lead rose from 2.4 percent to 4.9 percent. The paper reports a statistically significant change. Outside the Flint water area, the increase from 0.7 percent to 1.2 percent was not statistically significant. In the high-water-lead part of Flint, the proportion increased from 4 percent to 10.6 percent. Those are group proportions defined by the study, not a count of all exposed children and not a measure of the eventual severity of every child's injury. They also do not describe an increase of 6.6 percent relative to the earlier value; the high-water-lead comparison spans 6.6 percentage points. Keeping the base, later value and geography together avoids converting a spatial finding into one undifferentiated city statistic. F067, uvm.edu, Hanna-Attisha and colleagues, Elevated Blood Lead Levels in Children Associated With the Flint Drinking Water Crisis, AJPH
The authors used geography to connect water and health evidence. They mapped blood results to addresses and used a statistical interpolation method to visualise the variation across the area. The paper reports clustering inside Flint, with higher water-lead areas corresponding to higher predicted blood-lead risk. A predicted surface is not a blood measurement taken at every address on the map. It uses observed results and spatial relationships to estimate variation between them. In the study, geography helped prioritise response in areas where measured and predicted risk aligned, rather than treating every neighbourhood as though it had the same plumbing, water conditions or starting disadvantages. F067, uvm.edu, Hanna-Attisha and colleagues, Elevated Blood Lead Levels in Children Associated With the Flint Drinking Water Crisis, AJPH
The paper's discussion argued that the source change was the probable cause of the increase. The authors described reviewing other possible changes in exposure, including demolition, new lead-producing facilities and indoor remediation programmes. They found no known alternative that explained the pattern during the comparison period. Their inference combined the geographic correspondence, river-water chemistry and absence of corrosion control. This is stronger than a statement that two dates merely coincided, but it still belongs to the study and its design. It is not a retrospective measurement of each child's dose from the water system. The discussion also notes that Flint's elevated-lead proportion was higher than the surrounding area's before the switch. The new hazard widened an existing disparity rather than arriving in a community with no earlier lead burden. F067, uvm.edu, Hanna-Attisha and colleagues, Elevated Blood Lead Levels in Children Associated With the Flint Drinking Water Crisis, AJPH
The university team made its full collection available in December. The page says the release responded to requests for the lead-in-water analysis from the citizen-science effort and offered a comparison for selected homes sampled by both the team and the state. The ledger records the page and its description, not the contents of the linked spreadsheet. This account therefore does not derive a fresh percentile or calculate a new total from that file. What the page establishes is the team's move to release the dataset and the checks credited to named researchers. It allows others to examine the underlying records, but a description of the download is not itself evidence of every number the download might contain. F029, flintwaterstudy.org, [Complete Dataset] Lead results from Tap Water Sampling in Flint, MI, 2015-12-02