Support #2289
open
Health condition (hcond) variables in Wave 13
50%
Description
Dear UKHLS Team,
I am using the hcond variables in Wave 13 to look at whether respondents have certain health conditions. Specifically, I would like to create a variable indicating if they currently have the condition (i.e. they have been diagnosed with the condition and still have the condition at the time of interview). I have followed the guidance in the PDFs you have posted in previous posts on the forum (e.g. #2220), however I believe that the way health conditions are recorded in the data has since changed, particularly for continuing respondents. Having looked at the data I think I have worked it out generally, however could you please let me know what the prevhcondno* variables record and how it is coded in the Wave 13 data? It appears to be different from in previous waves.
In the previous PDF guides (which I think were for up until Wave 12), you stated that conditions reported in previous waves were recorded in "PREVHCONDNO1-PREVHCONDNOn (where n was the maximum number of health conditions reported in that wave) where each variable simply recorded the number of the health condition from the list". However, in the Wave 13 data, m_prevhcondno1 appears to relate to asthma, m_prevhcondno3 appears to relate to congestive heart failure etc. The variables are coded 0 and 1 (without labels), but the vast majority of respondents are coded -8 (inapplicable). My guess is that 1 = previously reported the condition, and 0 = did not previously report the condition, however I am not sure why so many respondents are coded as -8.
Thank you very much for your help!
Best wishes,
Laurence Rowley-Abel
Files
Updated by Understanding Society User Support Team 2 months ago
- File FAQ_HCONDS - 20250214.pdf FAQ_HCONDS - 20250214.pdf added
- Category set to Data documentation
- Status changed from New to Feedback
- % Done changed from 0 to 50
- Private changed from Yes to No
Hello Laurence,
A value of 1 indicates respondents who still have the previously reported condition, while 0 refers to those who no longer have it. The (–8) code applies to respondents who did not report having the condition previously. The question was only asked to those who had reported the condition earlier; therefore, if it was not reported, they were not eligible for this follow-up question.
I’ve attached an updated version of the Health Conditions Guide, which now includes information up to Wave 14.
I hope this information is helpful
Best wishes,
Roberto Cavazos
Understanding Society User Support Team
Updated by Laurence Rowley-Abel 2 months ago
Dear Roberto,
Thank you for your reply and for the PDF with the detailed guidance. You said above that for prevhcondno* a value of 1 indicates respondents who still have the previously reported condition and 0 indicates those who do not still have the condition. However, is that not what the hcondp* variables indicate? In the PDF guide it says the prevhecondno* variables indicated whether reposndents previously mentioned the condition (1) or not (0).
Additionally, I find that many of the people who had previously reported a condition are coded as inapplicable (-8) for the relevant prevhcondno* variable. For example, the below table shows a cross-tab of the m_prevhcondno16 variable (for high blood pressure in Wave 13) against an indicator of whether the respondent had high blood pressure at the previous wave (Wave 12). Attached is the Stata code to reproduce this.
| m_prevhconno16 | Did not have high BP in Wave 12 | Had high BP in Wave 12 | |
|---|---|---|---|
| inapplicable | 19899 | 268 | |
| proxy | 48 | 1 | |
| 0 | 3193 | 49 | |
| 1 | 603 | 889 |
As you can see, there are 268 respondents who reported having high blood pressure at Wave 12, but who are coded as inapplicable on the m_prevhcondno16 variable in Wave 13. Since they had high blood presure at Wave 12, should they not be coded as 1 on the m_prevhcondno16 variable in Wave 13?
Many thanks for your help with this.
Best wishes,
Laurence Rowley-Abel
Updated by Understanding Society User Support Team about 2 months ago
- File clipboard-202511152134-iervo.png clipboard-202511152134-iervo.png added
- File clipboard-202511152134-bmw5m.png clipboard-202511152134-bmw5m.png added
Hello Laurence,
That’s correct, the variable prevhcondno is the routing variable that leads into the question hcondp. That is, if prevhcondno = 1, then hcondp is asked.
What I was referring to is the universe behind prevhcondno. In the questionnaire, it’s called FF_PREVHCONDSTIL, so my understanding is that respondents who reported having the condition in the previous wave are those who are fed forward to populate this variable. Given the variables available, these should correspond to: hconds16 – Still have health condition: High blood pressure; and hcondns16 – Still has newly diagnosed health condition: High blood pressure. The first captures whether respondents still have the condition reported in hcondcode16, while the second records whether they still have the new condition reported in hcondncode16.
However, I am still unsure about the 927 cases flagged as having had the condition previously but who are not captured by l_hconds16 or l_hcondns16. I’ll need to look into this in more detail, and I’ll let you know as soon as I have more information.
I hope this information is helpful
Best wishes,
Roberto Cavazos
Understanding Society User Support Team
Updated by Laurence Rowley-Abel about 2 months ago
Dear Roberto,
Thank you for your reply. That makes sense to me that the prevhcondno* variables are routing variables for hconp*. And thank you for looking into those 927 cases.
Additionally, you said that respondents who had the condition at the previous wave should be fed-forward to the m_prevhcondno16 variable and that this corresponds to respondents for whom l_hconds16 1 or l_hcondns16 1. However, what about those for whom l_hcondp16 == 1 (i.e., they had previously reported the condition before Wave 12 and then reported still having it in Wave 12)? Some of these individuals have values of 'inapplicable' or 0 for m_prevhcondno16 (see screenshot below), but my understanding is that they should have a value of 1?
Many thanks.
Best wishes,
Laurence