The DHS Program User Forum
Discussions regarding The DHS Program data and results
Home » Countries » Uganda » Causes of maternal mortality
Causes of maternal mortality [message #20067] Tue, 22 September 2020 17:59 Go to next message
maternalresearch is currently offline  maternalresearch
Messages: 3
Registered: September 2020
Member
Hi,
My co-researcher and I are looking for some data mentioned in this Ugandan government paper (http://npa.go.ug/wp-content/uploads/NDPII-Final.pdf). On pages 30-31 it says that "The major causes of maternal mortality include; haemorrhage (42 percent), obstructed or prolonged labour (22 percent) and complications from abortion (11 percent)". It does not cite a source for this claim but the entire preceding paragraph cites various DHS surveys.
However, we looked at the various DHS survey datasets cited and cannot find data about cause of maternal death in any of them in Stata. The entries under mm11 are all completely empty. Are we missing something, or does this data simply not exist for the Ugandan datasets?
Many thanks.

[Updated on: Tue, 22 September 2020 17:59]

Report message to a moderator

Re: Causes of maternal mortality [message #20083 is a reply to message #20067] Wed, 23 September 2020 16:13 Go to previous messageGo to next message
Bridgette-DHS is currently offline  Bridgette-DHS
Messages: 2025
Registered: February 2013
Senior Member

Following is a response from DHS Research & Data Analysis Director, Tom Pullum:

The paper you describe must have been referring to clinical maternal mortality data from Uganda or to what is known internationally. DHS surveys do not include questions that could possibly lead to such estimates. You are right to question the source. Thanks for letting us know about this potential misinterpretation of what the DHS data include.
Re: Causes of maternal mortality [message #20084 is a reply to message #20083] Wed, 23 September 2020 17:39 Go to previous messageGo to next message
maternalresearch is currently offline  maternalresearch
Messages: 3
Registered: September 2020
Member
Many thanks for your response, I am very grateful.
I eventually tracked down the clinical data from which this estimate was taken (it was a different, uncited source). However, I still have queries regarding your response:
1) Another paper I have seen recently cites the Kenya 1998 DHS for an estimate of causes of maternal death.
2) The Pakistan 2006-2007 DHS report itself explicitly gives some data for causes of maternal death (by percentage): https://dhsprogram.com/pubs/pdf/FR200/FR200.pdf on pages 179-180.
3) The recode manuals do mention variable 'mm11' which is the cause of maternal death. So it does seem in principle possible to give estimates of the proportion of maternal deaths for each cause. The DHS guide says that because it is a relatively rare event, MMR calculated from the DHS is subject to large sampling errors. But it does still seem in principle that you could get very bad estimates of the proportion of maternal deaths by each cause, even if these estimates are virtually worthless.

Given 2) and 3), it definitely seems like some data is collected re: cause of maternal deaths. I'd be really grateful to know whether this data exists specifically for Kenya 1998 DHS (and has been used irresponsibly), or whether the data is completely non-existent (and therefore completely made up).

Thanks so much again.
Re: Causes of maternal mortality [message #20092 is a reply to message #20084] Thu, 24 September 2020 08:30 Go to previous message
Bridgette-DHS is currently offline  Bridgette-DHS
Messages: 2025
Registered: February 2013
Senior Member

Following is a response from DHS Research & Data Analysis Director, Tom Pullum:

I can't completely answer this question but can add to the previous response. Yes, the mm variables always include mm11, but I don't recall this variable ever being coded. I just checked all the Kenya surveys, including 1998, and mm11 was always NA. In the 1998 main report I searched for "obstruct" and found this:

"It is important for the health of both the mother and child that trained medical personnel are available in cases of prolonged labour or obstructed delivery, which are major causes of maternal morbidity and mortality."

I expect that many other survey reports have referred to obstructed delivery or hemorrhage or eclampsia/pre-eclampsia, etc., because these are generally known factors in maternal deaths. But there is no basis for giving a percentage breakdown and saying they come from DHS.

Pakistan has never included sibling histories. However, they recently did a different type of special survey on maternal mortality. There may have been earlier special surveys.

I would question any breakdown that uses medical terms such as hemorrhage or obstruction unless the data are clinically-based. But I would not go so far as to say the data were made up. There may be an origin story that we just don't know.
Previous Topic: URGENT QUESTION: Missing child nutrition values in Uganda DHS2016
Goto Forum:
  


Current Time: Thu Nov 26 22:03:33 Coordinated Universal Time 2020