Request
I would like data on reported cases of Legionnaire's Disease in the City. A 2014 study by the Department of Health and Mental Hygiene references such data including information on: (1) sex, (2) age, (3) race/ethnicity, (4) pre-existing medical conditions, (5) occupation, (6) nights away from home, (7) recreational water exposures, (8) other risk factors for acquiring Legionella, (9) death. In addition to the information mentioned in the study, it would be helpful if, for each case, information on the (a) date and time of legionellosis positive test result, (b) source of Legionella, (c) any linked cases of Legionella, and (d) location information about where the patients lived at the lowest aggregate level deemed necessary to protect the patients' privacy.
I am working on a project to explain some of the recently developed concepts in fair machine learning. One of our goals is to help the public understand some of the fairness issues the City is grappling with as the Automated Decision System Task Force works on its report. The requested data can help answer the family of questions: How do the different fair machine learning techniques compare to each other in predicting risk of Legionnaires’ disease and to the traditional fairness-blind approaches? And in particular, how do their predictions compare when disaggregated into groups based on race and risk factors associated with Legionnaires’ disease such as age and occupation. Thanks to the City’s impressive open data ecosystem, we can conduct several similar and complementary analyses. The Legionnaires’ disease data would allow us to add another layer of nuance adding further value for our audience. Thank you!
Response
Thank you for your inquiry about NYC DOHMH communicable disease data, which is not public data pursuant to New York State law. In order to access the data, therefore, you would need to go through the DOHMH data use agreement and IRB processes, which can take time and may not meet the timeline you have set for your work. To begin this process, you may contact Dr. Robert Fitzhenry of the Bureau of Communicable Disease at rfitzhenry@health.nyc.gov to discuss your data needs and the study question(s)you wish to address.
Thank you
DOHMH