Correspondence between

Thank you. {will await confirmation, before analyzing the data.

Importance: High i talked with Jennifer Eisner and here is what we need: Only the 2014 data (not 2015 since we don’t typically release a year’s data before the year is completed), by zip code for all of Michigan with smail cells suppressed. And only confirmed venous tests as CLPPP has done for previous years: we'll explain that we’re including capillary tests in the Flint reperts in an effort to capture all possible EBLL children. xxxEND_PAGE:dhhs03_b0758_2644_2650_4 002648 Could | please have an estimate cf when this can be done?

Until Junaicl has access to the warehouse, Pattl with Yan's additional support would be updating the blood leed test counts. Once Junaid has acress to the warehouse data he should be able to take this function an. {My understanding was that Patti has already demonstrated to you [Martha, Linda, Junaid] haw she pulis the data from the warehouse and then can provide SAS code for the analysis to Junaid.)

Importance: High Ok, | deleted all the chatter in-between. i believe the request is for annual 2014 and 2015 (not by quarter) for all Michigan zip codes (suppressing cells less than 6). The data provided should match what we’re reporting to the Directer/Governor, so retain the highest test whether venous or capillary. Jennifer /Geralyn: is that accurate?

To clarify you want 2014 by zip code and quarter with retaining highest test? Or only highest venous test? Im still ina meeting and may not get this done today. Will also talk to Yan to see if she is available. Patti Sent from my iPhone xxxEND_PAGE:dhhs03_b0758_2644_2650_5 002649

The other reporter was given counts by county. But only for verious, as | understand it. So this needs to be rerun and then suppression applied. And we probably should give the same information to the other reporter. Patti — are you OK with this?

Is this what Bob has already done? | will try to run this before my 1:00 meeting. Sent from my iPhone

But it seems to me that we should not release counts less than 6 — those cells should be suppressed. Your thoughts?

Do we need zip code from the file we deduplicated by year? Either way we can provide, just need to clarify. Sent from my iPhone

Flint is based on zip cod, not city cade. Do any of the zipcodes overlap county Hnes? Also it ts always possible that county, city or gipcade could be miscategorized, Geooding these data would be useful. Ori may have transnased colurnns when | enteredin Excel. Esent the SAS output ta Corinne and Sarah on Friday. {will check on what ican during the lunch break here, but will be facilitating a workgroup after lunch.

Attached is the draft renort as requested by the Governor’s office. Please do not share with anyone as it is not yet final. The time for these discussions, at least for this version of the report, was last week and we do not have any additional time to make changes. Maybe on the next version or when the epi analysis has transitioned to DEH.

Hi all, Bob Scott updated me on the upload of child lead data into the warehouse “As of next Monday, we'll be two weeks behind on Lead records in the Warehouse. Jessica’s computer (and everyone’s around here) was upgraded to Office 2013 last Friday. ....but Access 2013 will not handle DBF files. We have used Access for a few years as a pass-through to get the DBF format (dates, | think) exactly right for loading to the Warehouse. When | get back, I’ll work with the folks from Optum to see what we can do about getting files in Warehouse-ready format.” Bob will be back next Monday. ican pull data from the warehouse next Monday, but be aware that it may not be complete. Patricia McKane, DVM MPH - Manager, Maternal & Child Health Epiderniology Section | Michigan Department of Health and Human Services | Lifecourse Epidemiology and Genomics Division | PO Box 30195 / 201 Townsend St 4" Floor - Lansing, MI 48909 | Cell Work Hours: M_F 7:30am — 4:00 pm Michigan has.a public records law. Most written communications to or from state officials regarding state business are public records available to the public and media upon request. Your e-mail communications and any attachments to them may be subject to public disclosure. xxxEND_PAGE:dhhs03_b0401_1522_1524_2 001524 This message, inclsttiag any attainments is intedred solely for the use af the RAHA reciprent(s) anh may eerttitin confistentia! and/or priviteged informtrtion. Any uirrtrehorzed review, use, dactesure, or distribation of any confitential endfor prviteged informenion eomtofded in this emi! is exprasty protibited. if you are net the irtterrfed recipient, please ewrtdct the sender by reply exraill art! destroy any td all copies of fe original messoge. xxxEND_PAGE:dhhs03_b0401_1522_1524_3

We have it broken down by age and zip. Under 6, 6-17 and 18+. We are verifying code and may need to tweak wording in the tables, but we will have tables broken out by age and zip and we may have something for Eden by tomorrow afternoons meeting. I can't stress enough that we need to be accurate first, fast second. Yan has done an amazing amount of work in a short period time. She is very efficient and very accurate. We are documenting the process and the issues or questions that we are finding as we get to know the data set. We will be double checking the code,as we are finding few repeat tests. It may be code issue or data issue, we are working to figure it out. Ideally you would link to education data to analyze by school. That would be more accurate than zip. Sent from my iPhone Shoot me now, but first answer Eden’s question. i'm guessing they'd like that broken down by age and zip code to go with the school drinking water data...

xxxEND_PAGE:dhhs02_b0517_1995_1997_2 001997 We are using the highest test and following AAP guidelines as this has clinical irnplications. Yan and |reviewed outout and modified our plan. We are using specimen date and not report date, which means the tables will be dynamic. When we used report date to define the cohort we found some records with specimen dates fram as far back as 2012 and 2013, the mean and sd were astronomical, which is another quality improvement analysis for a later time. Only 22 records reported on 10-30-14 were fram the week ending 10/30/14. Interestingly all 22 records were <5 meg/dl. Ry changing te specimen date to select the cohort ,we find mean interval between specimen and report is 12.4 days, so we are examining other reporting intervals such that we have sufficient n, and feel that we have most of the tests. Also the latest record that we found for the week reported 10/30 was on 10/27, will need to comrnunicate some of the fags in reporting. We have table shells, we have the datasets created to do what we need te do, and are new analyzing the data te determine the suitable reporting interval. Then we will fill the tables and review everything. I'm working on a ‘technical specs’ to explain what we did. anticipate that we will have something for the group te look at later this week, hopefully within a day or two, if all goes as planned. Patti xxxEND_PAGE:dhhs02_b0517_1995_1997_3

All I'm out of the office today for a family medical issue. Could we schedule some time on Tuesday morning to discuss what data HHS management would like to see in the reports and if there's anything we could be prepared to release later this week? It would be helpful if someone privy to the discussion taking place between DEQ and HHS communications could provide some insight into the plan (if there is one) for releasing the school water testing results and what coordination has been agreed to by HHS. Linda

Au, linda stopped by and | gave her a quick update- and few items since she checked in. 1. Bob, Yan and f met this morning to discuss the data and the reporting. Bob shared the “AAP Recornmendations on Medical Management of Childhood Lead Exposure and Poisoning”. 3. lextracted the 2015 file to date from the data warehouse and imported it ta SAS. We found some incompatibility with SAS versions. | run 9.4 and any dataset that | create Yan can’t read using 5AS 9.2. There is aiso known bug in 9.4 for exporting data and | couldn’t export data in a format that she could use. So we basically tricked SAS to give us an excel file, 3. Yan and met to discuss how to manipulate the data to provide the desired metrics. itis complicated! Yan has the data and is cleaning it and creating flat files. 4. Martha and Junaid sent me the specs fram the tracking program. | reviewed thern and forwarded them to Yan. tthought the flow chart was helpful to assign the child lead classification (confirmed venous, confirmed capillary or unknown, unconfirmed etc... For now we are using new { capillary or venous} or repeat (capillary or venous}, but we will certainly consider the tracking method. iwould prefer to use the lead levels as outlined in the AAP suicielines as they align with clinical recommendations. 5. lerwated quick and ugly table shells, primarily to organize my brain. Yan is very efficient, but | anticipate that we will be doing quite a bit of reviewing and revising aur rnethods before we carn provide you with draft tables. xxxEND_PAGE:dhhs02_b1147_4316_4317_1 004317 Patti

Hi Patti, I’m out on Monday, but would like to chat on Tuesday about how to best go about compiling the Flint lead reports, not the manipulation of the data itself, but rather just the logistics of working with CLPPP, getting the data into the report, etc. I’m hoping you and Bob Scott will connect ASAP to get the ball rolling on the data work, so please let me know if you have meetings planned. | don’t need to be part of the meeting: | just need to know that the work is getting done as | have to answer to HHS management and various outside entities. Thanks Linda D. Dykema, PRD. Environmental Public Health Director Division of Environmental Health Michigan Department of Health & Human Services 517.335.8566 xxxEND_PAGE:dhhs02_b1147_4316_4317_2