The DHS Program User Forum
Discussions regarding The DHS Program data and results
Home » Topics » Fertility » Informed choice and contraceptive discontinuation (Method information index and contraceptive discontinuation)
Informed choice and contraceptive discontinuation [message #23015] Fri, 25 June 2021 07:14 Go to next message
shafaq.mahmood93 is currently offline  shafaq.mahmood93
Messages: 4
Registered: March 2021
Hi. We are working on the following research question

What is the effect of Method Information Index (MII), as a measure of quality contraceptive counseling, on method discontinuation.

Currently I am working with DHS data set of Pakistan (2017-18). According to the definition of "informed choice" given in PDHS report it includes sample of only current users of modern methods age 15-49 who started the last episode of use within the 5 years preceding the survey.

Based on our understanding of the above definition, the MII variables (v3a02 - v3a06) should only be available for those females who are current modern contraceptive users. And if that is the case then probably we would not be able to look at contraceptive discontinuation among these females because all are current users. Although we were able to find one working paper of DHS (Myanmar 2015-16) that looked at informed choice as one of the predictor variables for 12-month contraceptive discontinuation rate and we are not sure how did they do this because apparently MII is only measured in current users, so there shouldn't be any discontinuation then. Please correct me if I am wrong. Is there any way we can still answer our research question?

Link to Myanmar study:

Second problem that we identified is that there is a slight discrepancy in the PDHS 2017-18 data set. As per the definition, MII variables should be recorded for only those females who are currently using a modern method. But when we look at v313 (current use by method type) and MII variables (v3a02-v3a06) in PDHS data set then some females who are currently using traditional methods have also been asked some of the MII questions (particularly v3a06). What is the reason for that?

In addition to that, after creating our event file and discontinuation indicator variable (according to the calendar tutorial) we again looked at v313, MII variables and discontinuation indicator, after merging it with the IR data set. And we noticed that for several cases, the discontinuation variable had a value of "1" (meaning that discontinuation was positive) against those females who are a part of informed choice subsample. Although, according to our understanding, all these females are current users and logically their discontinuation indicator variable should always have a value of "0" (meaning no discontinuation). Can you please clarify this.

I hope my queries are clear to you. Waiting anxiously for your response.

[Updated on: Fri, 25 June 2021 07:15]

Report message to a moderator

Re: Informed choice and contraceptive discontinuation [message #23034 is a reply to message #23015] Mon, 28 June 2021 11:45 Go to previous message
Bridgette-DHS is currently offline  Bridgette-DHS
Messages: 2804
Registered: February 2013
Senior Member

Following is a response from Senior DHS Specialist, Kerry MacQuarrie:

You are correct that the information for the MII is only collected of current users rather than past episodes of use, and so cannot be analyzed for its association with discontinuation. While it is commonly hypothesized that the MII (or quality of care more generally) and discontinuation are linked, this linkage is typically only analyzed with prospective data rather than retrospective data as found in the DHS contraceptive calendar.

The variables v3a02-v3a05, which go into the MII, are asked of users of modern methods (pills, IUD, injectables, female sterilization, implants, EC, other modern methods) except condoms and LAM--which do not require a health provider but can be purchased from a shop or used on one's own. The last variable which goes into the MII, v3a06--told about other FP methods, is asked of users of all modern and traditional methods, alike. For users of the methods in v3a02-v3a05, the question is, "At that time [that you got the method], were you told about other methods of family planning that you could use?" For the other methods, the question is, "Were you ever told by a health or family planning worker about other methods of family planning that you could use?"

After creating your event file and merging with the IR file, you would have the MII data for each woman and this information would be attached to the record for each episode of use. That is, if a woman contributes multiple episodes of use to the dataset, the MII would be attached to each episode. But it would pertain ONLY to the most recent, not discontinued episode/method. For example, consider a woman is a current pill user. Her contraceptive history in her calendar indicates that she had an earlier episode of pill use that she discontinued (episode 1), followed at some later point by an episode of condom use that she also discontinued (episode 2), followed at some later point by an episode of pill use that she used continuously until the time of the survey. This woman would contribute 3 episodes to the discontinuation analysis (2 discontinued, 1 not discontinued). The MII data would be present for all 3 episodes but only describes her experience when she obtained her current method of pills. Those data attached to the earlier 2 episodes should be ignored.

Previous Topic: coding for vcal_1 and vcal_2
Next Topic: DHS v302a-- Ever Use in Phase 6
Goto Forum:

Current Time: Fri Jun 2 15:24:17 Coordinated Universal Time 2023