When 2019 NHIS data will be available on IPUMS

Hey there, I’m new to the forum and I’d like to ask about the 2019 NHIS data:

  1. When will it be available on IPUMS?
  2. Would it be appropriate to use the 2019 data that is available on the CDC website (NHIS - 2019 NHIS)?

Any feedback would be highly appreciated!

Thanks,
XC

Thanks for your questions. I will answer each in turn. Please let me know if I can provide any additional information.

  1. I don’t have a definitive timeline for the release of the 2019 NHIS data via IPUMS. We are in the process of updating our standard procedures for integrating data quickly to accommodate a major survey redesign in 2019. However, I can assure you that the IPUMS NHIS team is prioritizing releasing the data quickly. We will announce the release promptly via the IPUMS NHIS revision history page and the IPUMS Twitter account (and send an email to our users shortly after the release as well).

  2. It is definitely appropriate to use the data directly from NCHS/CDC. They are the original data providers; these data are the starting point for IPUMS NHIS. If you are only using 2019 data, it may be more efficient to get them directly from NCHS/CDC. However, if you are interested in pooling multiple years of data, you may be interested in waiting for the IPUMS NHIS release as this will save you the effort of comparing variable names that change over time, merging the different annual files (e.g., sample adult, sample child) together over time, and constructing the correct weight variable based on variable availability (to name a few).

Hi Kari,

Thanks for your detailed response! I’m interested in pooling multiple years of data, so I would probably just wait for the IPUMS NHIS release. I also noticed the survey redesign in 2019, and I was wondering if 2019 data is comparable with previous years. For example, if we’d like to visualize the trend of a variable using 2010-2019 NHIS data, is there any additional data management steps we need to prepare for 2019 data? Or IPUMS NHIS release will capture the change so that we can use the data directly (or with some weight variables)?

Really appreciate your help!
Xi

The redesign definitely poses some comparability issues with earlier years. It is up to each individual researcher to decide if they think pooling is appropriate for their specific application. That being said, I would encourage you to 1) read the IPUMS (forthcoming) and NCHS (linked in previous post) documentation closely, 2) be cautious when interpreting time trends in 2019, and 3) run analyses without 2019 to see if it drastically changes your results. Given that the redesign only retains the sample adult and sample child aspect, I would also recommend a close examination of estimates that use variables previously available on the household/family/person file that move to the sample adult/sample child records in 2019. While eliminating household/family/person files is the major deviation, I will also note that in earlier years the data include one sample adult/child per family; in the 2019 design it is one sample adult/child per household.

You will be able to pool 2019 data with earlier years via the IPUMS NHIS extract system, so you shouldn’t need to do any additional data management from that perspective. As always, we recommend consulting the weights tab for each variable for information on weighting, as well as relevant documentation as part of our user guide.