'tanitrd ?rms ?rm WASHINGTON, DC 20510 March 29, 2019 Secretary Azar US. Department of Health Human Services 200 Independence Avenue, SW. Washington, DC. 20201 Dear Secretary Azar: We write regarding a recent article published in The Trace and iveThim?yEight highlighting the unreliable nature of the Centers for Disease Control and Prevention?s (CDC) ?rearm injury data.1 Researchers, journalists, lawmakers, law enforcement, and the American public at large rely on the data: ?[s]ince 2010, at least 50 academic articles have cited the gun injury estimates.?2 Yet, the CDC acknowledges that its national estimate of gun injuries is ?unstable and potentially unreliable?.3 Given the wide scale reliance on the numbers, it is imperative that the agency ensure its data is accurate and up to date. While we appreciate that the CDC is looking ?into various ways to strengthen the estimates for nonfatal ?rearm injuries?,4 we respectfully request information regarding actions the agency is taking to remedy its data shortcomings. According to The Trace, the gun injury confidence interval for 2017 is four times wider than the interval from 2001.5 Moreover, the coefficient variable, ?a measure of an estimate?s uncertainty in which higher values indicate larger potential errors? rose from 22.1 percent in 2001 to 39.1 percent in 2017.6 A possible explanation for these increases could be due to the CDC sourcing its data from a small number of hospitals known as the National Electronic Injury Surveillance System a database maintained by the Consumer Product Safety Commission.7 The NEISS collects records from approximately 100 hospitals, and only 66 of those 100 hospitals collect data regarding gunshot injuries.8 In contrast, the Department of Health and Human Services uses the Healthcare Cost and Utilization Project database containing data from more than 950 hospitals. Moreover, the sample of hospitals fails to take into datel', ?3 anyway/ 3 4 .thetraceorgXZO 5 Id. 6 Id. 71m date/; 3 anyway/ data from more than 950 hospitals. Moreover, the NEISS sample of hospitals fails to take into account regional differences in gunshot injuries.9 There appears to be no rational reason that the CDC and HHS use different databases. Given that the CDC is not currently conducting gun violence research,10 the very least the agency can do is to ensure that its gun injury numbers are accurate. To better understand the issue, we request responses to the following questions: 1. What is the CDC doing to more accurately track non-fatal ?rearm injuries? Please be speci?c. 2. What additional resources, if any, does the CDC need in order to better ensure more accurate ?rearm injury data? 3. Why does the CDC use the NEISS database rather than the Healthcare Cost and Utilization Project database in order to track non-fatal ?rearm injuries? 4. What accounts for the four?fold increase in the gun injury con?dence interval between 2011 and 2017? 5. As the con?dence interval increased over the course of 16 years, did the CDC take any steps to ameliorate the problem? If so, please describe those steps. If not, please explain why no action was taken. 6. What role has the Dickey Amendment played in the decision to continue using NEISS even as the data proved less and less reliable? 7. Has the CDC received instructions from any executive branch of?cials pursuant to its non?fatal ?rearm injury data? If so, please describe any such instructions and communications. Please direct any questions to Shelby Boxenbaum with Senator Menendez. or 202-224-4744). Thank you in advance for your cooperation. We look forward to receiving your responses by April 20, 2019. Sincerely, a do 5 I. I Kirsten Gillibrand United States Senator obert Menendez United States Senator 9 Id. l0h violence-I'esearchf; Angus King Jr. United States Senator Richard Blumenthal United States Senator Tina Smith United States Senator Chris Van Hollen United States Senator Mag. Mazie K. Hirono United States Senator Cory A. Booker United States Senator D. Harris United States Senator Christopher S. Murphy United States Senator