Thanks so much for flagging this—I had seen this on BBC but this has pushed me to follow up! Here’s a bunch of not so ordered thoughts. As background I’m a doctor who manages rural health centers, some of which deliver babies.
The benefit claimed in this study is insane. A 50% drop in bloodloss related maternal mortality is beyond the pale of almost any recent medical intervention. It’s important to remember though that this effect cannot be all be attributed t the nterventno is not only not a RCT, but there is no control group. Having a control group here was impossible for ethical reasonsbecause the intervention was already proven to have benefit
There will have been other developments during the 5 years of the study that contributed to the 50% drop in maternal mortality, apart from their intervetions. As one concrete example according to the worldbank fertiility rates have reduced in NIger from 7.2 too 6.8 during the study period, which of itself will have reduced maternal mortality by a small percentage. Also their statement “Niger has not seen changes in its population’s socioeconomic status or infrastructure that could affect the results.” seems inaccurate as Niger’s GDP has increased by about 15% in that time.
Also they claim that surrounding countrys’ maternal mortality rate has not decreased in that time, which seems almost impossible given the trend of maternal mortality reducing (if at a far slower rate). For example below is modelling estimating that bordering country Niger’s mortality rate reportedly dropped by around 15% during the study period. Of course data are limited.
It however seems likely their intervention had a big effect, but I feel like they could have put a bit more effort in achknowledging these confounders. EAs would conservatively and heavily discounted for these kind of things, but they barel acknowledged them in this study.
2. This intervention is probably very cost effective, while keeping in mind that they haven’t accounted for other potential causes or discounted at all. From the study “the financial benefits to the population have been roughly estimated to be 5·7–7·8 times the annual running cost of the intervention, costing about US$37·94–27·73 per disability-adjusted life-year prevented (appendix pp 7, 17–24), more than Expanded Program on Immunization vaccination, but less than treating diarrhoea or acute respiratory illness in children younger than 5 years.”. Seems pretty good. Also these kind of country AID funded programs usually have massively bloated costs. I’m sure many orgs could do it much cheaper, perhaps even at half the cost or less which leaves room for increased cost-effectveness.
3.This study will immediately change my practice. The main intervention, giving misoprostol after every delivery isn’t standard practise in Uganda at the moment but is super easy and cheap to implement. Just making sure everyone uses misoprostl could potentially be an excitng intervention of EA level cost-effectiveness. I don’t have the time but someone SHOULD look int ths. This study will prompt me to start doing this within the next couple of months at the 5 health centers we manage which deliver (low numbers) of babies.
Just flagging that (in NZ at least), misoprostol is not given routinely for prevention of PPH. That’s probably because there’s access to IV oxytocin, so if you already have that in your practice then it’s probably worse for patient outcomes to change to misoprostol.[1] If there’s no access to IV oxytocin though, then misoprostol is recommended as second line recommendation by other bodies (FIGO 2022).
Alongside access to ecbolics, fairly simple things like active management of third stage of labour is probably pretty important in reducing PPH (my guess is the NNT of active management is probably comparable or lower than administering misoprostol for PPH prevention, though I’m less sure about the cost-effectiveness of scalable interventions in these areas).
Thanks so much for this Bruce—I had completely missed the fact that oxytocin was still clearly better—our faciities which deliver babies (apart from one) have fridges and already use that, although we don’t have standard practice of using it for every delivery which we should! (We use for most). I know that’s a bad miss by me.
So it seems like if you have a fridge—Oxytocin (it’s cheap anyway), and if not misoprostol.
Thanks so much again. In my defence I would have talked to the midwives and would have read guidelines before making that mistake anyway.
Thanks so much for flagging this—I had seen this on BBC but this has pushed me to follow up! Here’s a bunch of not so ordered thoughts. As background I’m a doctor who manages rural health centers, some of which deliver babies.
The benefit claimed in this study is insane. A 50% drop in bloodloss related maternal mortality is beyond the pale of almost any recent medical intervention. It’s important to remember though that this effect cannot be all be attributed t the nterventno is not only not a RCT, but there is no control group. Having a control group here was impossible for ethical reasonsbecause the intervention was already proven to have benefit
There will have been other developments during the 5 years of the study that contributed to the 50% drop in maternal mortality, apart from their intervetions. As one concrete example according to the worldbank fertiility rates have reduced in NIger from 7.2 too 6.8 during the study period, which of itself will have reduced maternal mortality by a small percentage. Also their statement “Niger has not seen changes in its population’s socioeconomic status or infrastructure that could affect the results.” seems inaccurate as Niger’s GDP has increased by about 15% in that time.
Also they claim that surrounding countrys’ maternal mortality rate has not decreased in that time, which seems almost impossible given the trend of maternal mortality reducing (if at a far slower rate). For example below is modelling estimating that bordering country Niger’s mortality rate reportedly dropped by around 15% during the study period. Of course data are limited.
https://www.researchgate.net/figure/Modeled-maternal-mortality-ratio-from-2014-to-2023-in-Mali-based-on-three-different-LiST_fig5_320914797
It however seems likely their intervention had a big effect, but I feel like they could have put a bit more effort in achknowledging these confounders. EAs would conservatively and heavily discounted for these kind of things, but they barel acknowledged them in this study.
2. This intervention is probably very cost effective, while keeping in mind that they haven’t accounted for other potential causes or discounted at all. From the study “the financial benefits to the population have been roughly estimated to be 5·7–7·8 times the annual running cost of the intervention, costing about US$37·94–27·73 per disability-adjusted life-year prevented (appendix pp 7, 17–24), more than Expanded Program on Immunization vaccination, but less than treating diarrhoea or acute respiratory illness in children younger than 5 years.”. Seems pretty good. Also these kind of country AID funded programs usually have massively bloated costs. I’m sure many orgs could do it much cheaper, perhaps even at half the cost or less which leaves room for increased cost-effectveness.
3.This study will immediately change my practice. The main intervention, giving misoprostol after every delivery isn’t standard practise in Uganda at the moment but is super easy and cheap to implement. Just making sure everyone uses misoprostl could potentially be an excitng intervention of EA level cost-effectiveness. I don’t have the time but someone SHOULD look int ths. This study will prompt me to start doing this within the next couple of months at the 5 health centers we manage which deliver (low numbers) of babies.
Interested to hear more comments about this.
Just flagging that (in NZ at least), misoprostol is not given routinely for prevention of PPH. That’s probably because there’s access to IV oxytocin, so if you already have that in your practice then it’s probably worse for patient outcomes to change to misoprostol.[1] If there’s no access to IV oxytocin though, then misoprostol is recommended as second line recommendation by other bodies (FIGO 2022).
Alongside access to ecbolics, fairly simple things like active management of third stage of labour is probably pretty important in reducing PPH (my guess is the NNT of active management is probably comparable or lower than administering misoprostol for PPH prevention, though I’m less sure about the cost-effectiveness of scalable interventions in these areas).
Note this also seems to be true for management of PPH.
Thanks so much for this Bruce—I had completely missed the fact that oxytocin was still clearly better—our faciities which deliver babies (apart from one) have fridges and already use that, although we don’t have standard practice of using it for every delivery which we should! (We use for most). I know that’s a bad miss by me.
So it seems like if you have a fridge—Oxytocin (it’s cheap anyway), and if not misoprostol.
Thanks so much again. In my defence I would have talked to the midwives and would have read guidelines before making that mistake anyway.