The DHS Program User Forum
Discussions regarding The DHS Program data and results
Home » Topics » HIV » Risky & Condomless Sex (Calculaton)
Risky & Condomless Sex Tue, 05 May 2020 11:54
 Yawo Messages: 34Registered: February 2019 Member
Hello: I am attempting to create a variable that measures engagement in protective sex.

The DHS Statistics Manual 7 defines this variable as "Number of women (or men) age 15-49, or young women (or men) age 15-24, who used a condom the last time they had sexual intercourse with a person who was neither their spouse nor lived with them", and provides the following code:

((risk1 & v761 = 1) or  (not risk1 & risk2 & v761b = 1) or (not risk1 & not risk2 & risk3 & v761c = 1)),

where
1=Yes
0=No
(risk1) v767a: Relationship with most recent sex partner (women)
(risk2) v767b: Relationship with 2nd to most recent sex partner (women)
(risk3) v767c: Relationship with 3rd to most recent sex partner (women)
v761: Condom used during last sex with most recent partner (women)
v761b: Condom used during last sex with 2nd to most recent partner (women)
v761c: Condom used during last sex with 3rd to most recent partner (women)

But i do not understand the way the conditions are specified. here is my understanding of the various configurations of protective sex:

-- (risk1 & v761 = 1): women who used condom with their most recent partner
-- (not risk1 & risk2 & v761b = 1): women who used condom with their most recent partner, but I do not understand the NOT prefix for risk1 & risk2. What does the entire condition mean?
-- (not risk1 & not risk2 & risk3 & v761c = 1): same issue/question. What does the NOT prefix remain.

I thought a command such as this would suffice:

(risk1 & v761=1) OR (risk2 & v761b=1) OR (risk3 & v761c=1)

Please kindly let me know if I am misunderstanding the measurement.

Secondly, if i want to measure risky sex, what would be the approach? Just the converse, like:

(risk1 & v761=0) OR (risk2 & v761b=0) OR (risk3 & v761c=0)?

Thanks - cY
Re: Risky & Condomless Sex [message #19184 is a reply to message #19179] Wed, 06 May 2020 09:58
 Liz-DHS Messages: 1516Registered: February 2013 Senior Member
Dear User, a response by Courtney Allen:
Quote:

Thank you for catching this mistake in the Guide to DHS Statistics. DHS staff had already identified it but the correction has not yet been posted. The corrected code is as follows:

Number of women (or men) age 15-49, or young women (or men) age 15-24, who used a condom the last time they had sexual intercourse with a person who was neither their spouse nor lived with them:
Women:
```risk1 = (v527 in 100:251,300:311 & v767a in 2:6, 8, 96) (last partner)
risk2 = (v527 in 100:251,300:311 & v767b in 2:6, 8, 96) (next-to-last partner)
risk3 = (v527 in 100:251,300:311 & v767c in 2:6, 8, 96) (third-to-last partner)

((risk1 & v761 = 1) or
(not risk1 & risk2 & v761b = 1) or
(not risk1 & not risk2 & risk3 & v761c = 1))
```
When the Guide is next updated (soon, we expect), it will include this:

In earlier final reports, percentage of women or men who had intercourse in the past 12 months with a person who was neither their spouse nor lived with them and who used a condom as last sexual intercourse with such a partner, was based on using a condom with any such partner in the last 12 months rather than the most recent such partner. Previous calculation was as follows:
```Women: ((risk1 & v761 = 1) or
(risk2 & v761b = 1) or
(risk3 & v761c = 1))
```
Please let us know whether this takes care of your concerns, or you still have questions. Thanks again.

the GitHub code for these indicators available onhere
Re: Risky & Condomless Sex [message #19185 is a reply to message #19184] Wed, 06 May 2020 20:26
 Yawo Messages: 34Registered: February 2019 Member
thanks very much for the clarification. It makes sense now.

Just to clarify, does this code measure condom use by people who have at least one other spouse in addition to their spouse/partner?

```
//Have two or more sexual partners and used condom at last sex
gen hk_cond_2plus= (inrange(v527,100,251) | inrange(v527,300,311)) & inrange(v766b,2,99) & v761==1
replace hk_cond_2plus=. if v766b<2 // IMPORTANT TO EXCLUDE PEOPLE WHO HAVE NO PARTNERS OR ONE PARTNER
label values hk_cond_2plus yesno
label var hk_cond_2plus "Have two or more sexual partners in the past 12 months and used a condom at last sex"

```

I want to be sure since V766b includes spouse, so the variable 2 or more partners excludes those who have no partners and those who have only 1 partner.

Besides, don't we have to account for missing cases (i.e., excluding people who have not had sex within the past 12 months?)

Thanks - cY

[Updated on: Wed, 06 May 2020 22:41]

Report message to a moderator

 Previous Topic: Calculating DHS HIV Test Uptake Next Topic: Missing data for men's paid sex variable in DHS V recode
Goto Forum:

Current Time: Wed Jul 6 15:57:11 Coordinated Universal Time 2022