Maternal mortality in Europe

Maternal mortality in Europe

by | Sujit Rathod -
Number of replies: 21

From The Guardian

  1. What is the study design?
  2. What is the numerator for the maternal mortality rate calculation?
  3. What is the denominator for the maternal mortality rate calculation?
  4. Is the maternal mortality rate really a rate, in the epidemiologic sense?
  5. Is the maternal mortality rate an incidence or prevalence figure?

Mothers in the UK are three times more likely to die around the time of pregnancy compared with those in Norway, according to an international analysis of data.

  1. Explain how the ‘three times’ figure was calculated. What is the name of this epidemiologic measure?

FOR RETURNING STUDENTS

  1. What is the benefit/usefulness of presenting unstandardized maternal mortality rates?
  2. What is the benefit/usefulness of presenting standardized maternal mortality rates? What variable(s) should be used to standardize?

In reply to | Sujit Rathod

Re: Maternal mortality in Europe

by | Monica Mtei -

  1. Study design- Historical cohort study
  2. Numerator- number of deaths in pregnant women up to 42 days after birth
  3. Denominator-  live births
  4. The maternal mortality rate is not really a rate in an epidemiologic sense because it does not take account of changes in time at risk but rather takes all cases irrespective of the time during which they remain at risk
  5. Maternal mortality rate is an incidence figure since it includes only 'new cases' but not existing cases
  6. The 'three times' figure was calculated based on the ratio of the incidence of maternal death in the UK divided by the incidence of maternal deaths in Norway
  7. Benefits of unstandardized maternal mortality rates: a quick summary figure that can easily be compared in different settings
  8. Benefits of standardized: removing the effect of factor(s) that act as an alternative explanation to the outcome(confounder) and hence get the effect that can only be explained by the factor under consideration. Important factors to adjust for this case are age, gestation age, parity, and maternal complications


In reply to | Monica Mtei

Re: Maternal mortality in Europe

by | Sujit Rathod -
Welcome Monica! I'm totally with you for #2 to #8. I'm not convinced about the study design though.
In reply to | Sujit Rathod

Re: Maternal mortality in Europe

by | MAR ESTUPINAN FERNANDEZ DE MESA -

From The Guardian

  1. What is the study design?

Cross-sectional to assess maternal mortality rates among mothers in pregnancy and up to 42 days after the birth of the children between 2016-2018.

  1. What is the numerator for the maternal mortality rate calculation?

Numerator = number of deaths among mothers in pregnancy and up to 42 days after the birth of the children in a given time period (usually a year) (x100,000)     

  1. What is the denominator for the maternal mortality rate calculation?

Denominator = number of live births in the same given time (usually a year)

  1. Is the maternal mortality rate really a rate, in the epidemiologic sense?

I’d say MMR in the epidemiological sense would be a ratio (Ratio = n1/n2)

My understanding is that a rate is a measure of frequency with which a phenomenon occurs and equals to = n1 (subpopulation) / N (total pop) x T (person-years at risk)

  1. Is the maternal mortality rate an incidence or prevalence figure?

I think this is an incident figure reflecting all the new observations (maternity deaths) that happen during the follow up period.

Mothers in the UK are three times more likely to die around the time of pregnancy compared with those in Norway, according to an international analysis of data.

  1. Explain how the ‘three times’ figure was calculated. What is the name of this epidemiologic measure?

From the data in the article we know that in Norway the rates of death among mothers in pregnancy and up to 42 days after the birth of their children was 2.7 per 100,000 live births and 9.6 maternal deaths for every 100,000 babies born in the UK. Hence:

RR = RR(UK)/RR(Norway) = 9.6/2.7 = 3.6 death among mothers in pregnancy and up to 42 days after the birth of their children per 100,00 live births

FOR RETURNING STUDENTS

  1. What is the benefit/usefulness of presenting unstandardized maternal mortality rates?

To compare to the adjusted standardized MMR.

  1. What is the benefit/usefulness of presenting standardized maternal mortality rates? What variable(s) should be used to standardize?

Standardized MMR is useful when making comparison between populations of differing demographics structures (particularly different age-structures), where crude mortality and morbidity rate would be misleading.

For direct standardisation, age-specific rates from each of the population under study are applied to a standard population to obtain a standardized rates.

For indirect standardization, the age-specific rates from a standard population are applied to each of the study populations to obtain a set of standardized mortality rations (SMRs).


In reply to | Sujit Rathod

Re: Maternal mortality in Europe

by | Monica Mtei -

Thanks, Sujit, 

I agree that I have overlooked the study design. It is indeed a cross-sectional study

In reply to | Sujit Rathod

Re: Maternal mortality in Europe

by | ISHITA GUPTA -
1. What is the study design? Cross-sectional study
2. What is the numerator for the maternal mortality rate calculation? Number of deaths among mothers in pregnancy and up to 42 days after the birth of their children
3. What is the denominator for the maternal mortality rate calculation? live births (per 100,000)
4. Is the maternal mortality rate really a rate, in the epidemiologic sense? It is ratio with a numerator and denominator, not a rate as there is no time component.
5. Is the maternal mortality rate an incidence or prevalence figure? Incidence figure as it is the number of deaths (new cases).
6. Explain how the ‘three times’ figure was calculated. What is the name of this epidemiologic measure?
MMR in UK: 9.6 per 100,000 live births
MMR in Norway: 2.7 per 100,000 live births
Risk Ratio: 9.6/2.7 = 3.5
6. What is the benefit/usefulness of presenting unstandardized maternal mortality rates?
Provides an overview of mortality rates in a specific population at a given time, assess burden of condition and guiding the need of appropriate public health interventions
7. What is the benefit/usefulness of presenting standardized maternal mortality rates? What variable(s) should be used to standardize?
Takes into account age distribution, comparison across different populations.
Variable for standardization: Age distribution
Thanks.
In reply to | ISHITA GUPTA

Re: Maternal mortality in Europe

by | Sujit Rathod -
Hi Ishita - your response to #6 nails it, for me at least. To me, an unstandardized rate is a bit more helpful for public health, and a standardized rate a bit more helpful for epidemiology, though there isn't a clear dividing line between the two fields.
In reply to | Sujit Rathod

Re: Maternal mortality in Europe

by | FATHIMA MINISHA -
Hi everybody... let's have a go at this...

1) Study design is interested and always confusing for me when it comes to pregnancy related studies. I think I will probably think in terms of crosssectional study- looking at number of deaths between 2016 and 2018. Although they have mentioned this time only for the UK data, we have to assume they have looked at the maternal deaths in the other countries over this same time period.. so a 2-year cross sectional study

2) The numerator here is number of maternal deaths during pregnancy and up to 42 days after birth (regardless of if the cause of pregnancy is directly related to the pregnancy or not)

3) The denominator is the number of live births- although i can never understand why you would have a measure where the numerator are adults, and the denominator are live births or neonates. There are many problems with this- a mother could have multiple pregnancies, not all have live births and it's also possible that a maternal death added to the numerator does not have her child added to the denominator... really weird actually... I would think the better denominator would be the total number of pregnant women... I guess a more accurate measure would be live births as most countries have birth records.

4) Not a rate- thats probably the first thing we learnt in 101... :-DD.. there is no time element to it... it's actually a ratio (coz numerator is not part of the denominator - therefore cannot be called a risk as well)

5) All death related figures are incidence... you cannot die more than once.

6) The MMR in UK divided by the MMR in Norway- that will give us the ratio of the MMRs- basically a ratio of ratios...

7, 8) Well I am not sure if age is a good factor for standardization coz the reproductive age group is the same in the all populations 15-45 years... and there is higher risk of death in both extremes of age groups... so most likely standardizing on age might not be a requirement when it comes to MMR- what do you guys think? The advantage of standardized rates is that it allows easy comparison between different populations but i am not convinced that age is the best factor... Maybe it should be factors related to services available like hospital vs home birth...

I think I have asked more questions rather than answer your questions Sujit!! :-DD

Fathima
In reply to | FATHIMA MINISHA

Re: Maternal mortality in Europe

by | Kalkidan Yegele -
For your answer to number 3, which I completely agree with btw, what about twins and triplets? Are they counted as 1 live birth or 2 and 3? Or did were they ignored because their incidence is low and we thought it wouldn't affect our data? (Or I sound like a complete idiot because I'm missing something here?)
In reply to | FATHIMA MINISHA

Re: Maternal mortality in Europe

by | Sujit Rathod -
Hi Fathima -

RE #7: it is still possible to standardize on age even when the age _ranges_ are the same for the different comparison groups. This is because the _distribution_ of ages could be different. One country could have its live births concentrated among 15 years olds (which is very risky!) and another could be concentrated among 20-somethings (much safer!) and this difference in distributions could be a (partial) explanation for the difference in MMRs.
In reply to | Sujit Rathod

Re: Maternal mortality in Europe

by | Kalkidan Yegele -
I'm a huge fan of these topics, thank you for this.
1. Study design: cross-sectional study. (there is no follow-up from risk to outcome. They just picked a period and tabulated the numbers)
2. Numerator: Mothers who died during their pregnancy or with-in the 42 days after delivery during the time period.
3. Denominator: Live-births during the specified time period.
4. Maternal mortality rate: I would like to say no because it doesn't really take time into consideration. But also, can it be because, it calculates it per moment of reproductive age as opposed to pregnant women? I'm not really sure about this really.
5. Incidence.
6. The 3* figure came by dividing the UK numbers with the Norway numbers.
In reply to | Sujit Rathod

Re: Maternal mortality in Europe

by | Lina Abdulsamad -
Hello everyone,
I went down a rabbit hole trying to understand this lol.

1. What is the study design?

The study design is a cross-secitonal study from 2016-2018 (asusmingly the data years, as only mentioned for the UK).

2. What is the numerator for the maternal mortality rate calculation?

Numerator for maternal mortality rate=
recorded maternal deaths in the respective country (Denmark, Finland, France, Italy, the Netherlands, Norway, Slovakia, UK)
3. What is the denominator for the maternal mortality rate calculation?
demoninator= total deaths recorded live births in respective country
4.Is the maternal mortality rate really a rate, in the epidemiologic sense?
Since the time period is not considered, the reported maternal mortality rate is actually a proportion. So, this is actually a maternal mortality ratio. The maternal mortality rate would be = recorded maternal deaths in the respective country /number of women of reproductive age in woman-years ( in this case 42 days after birth) in respective country.

5. Maternal mortality is an incidence figure as it's measuring new, recorded cases of maternal deaths.

6. Ratio of MMR: MMR UK/MMR Noway= 9.6 per 100,000 live births/ 2.7 per 100,000 live births= 3.5 - I'm assuming it was easier to just say 3 than to round up?
In reply to | Lina Abdulsamad

Re: Maternal mortality in Europe

by | Sujit Rathod -
Welcome to the rabbit hole, Lina!

I like how you tried to propose a true maternal mortality "rate". What you've put together would be a maternal mortality risk, though. How can you make it a true "rate"?
In reply to | Sujit Rathod

Re: Maternal mortality in Europe

by | Lina Abdulsamad -
So, for a true rate- the person-time at risk would need to be counted for. [So, in this case from the beginning of pregnancy to the 42 days post-partum?] The maternal mortality rate would be = recorded maternal deaths in the respective country /number of pregnant women until 42 days post-patrum--- but now I'm thinking if it's actually possible to calculate the maternal mortlaity rate, not risk, based on what I defined here, since it wouldn't be possible to know the number of pregnant women in a country and during the beginning of pregnancy also.
In reply to | Sujit Rathod

Re: Maternal mortality in Europe

by | Sujit Rathod -
Interestingly enough, the researchers themselves call this a "Descriptive multicountry population based study" in their paper: https://www.bmj.com/content/379/bmj-2022-070621

Personally, I would go with ecologic, with geographic and temporal features. The underlying, country-level data are descriptive data, from routine records. And it definitely can't be cross-sectional, because it's not possible to measure death as an outcome in a cross-sectional study!
In reply to | Sujit Rathod

Re: Maternal mortality in Europe

by | FATHIMA MINISHA -
Hi Sujit…
Just to clarify my understanding… shouldnt the results of an ecological study be presented as correlation figures… ? So would we call this ecological just coz there are figures from different countries?
@Judith what do you think? Since you also agree that its ecological…:-)

Fathima
In reply to | FATHIMA MINISHA

Re: Maternal mortality in Europe

by | JUDITH MARGARET BURCHARDT -
Hi Fathima,

I thought it was ecological because the unit of analysis was a country's statistics. Is that correct?

I also note that I thought a MMR could be incidence or prevalence whereas everyone else said it was definitely incidence. Am I missing something here?

Many thanks

Judith
In reply to | JUDITH MARGARET BURCHARDT

Re: Maternal mortality in Europe

by | Sujit Rathod -
Hi Judith -

Try to talk through it: how would you calculate a prevalence figure when the outcome of interest is death? -s
In reply to | JUDITH MARGARET BURCHARDT

Re: Maternal mortality in Europe

by | FATHIMA MINISHA -

Hi Judith… I think sometime during our million discussions somebody said ‘you could not die more than once right?’… so doesnt it have to be incidence by default? It will be the first time somebody dies… looking at it technically

In reply to | FATHIMA MINISHA

Re: Maternal mortality in Europe

by | JUDITH MARGARET BURCHARDT -
Thank you Sujit and Fathima,

I have checked the definition of prevalence - the numerator is the number of people LIVING with the condition during a period of time, so I can see now why I'm wrong about there being a prevalence for fatal events.

That's very helpful for me

Best wishes

Judith
In reply to | Sujit Rathod

Re: Maternal mortality in Europe

by | Lina Abdulsamad -
oh, because death rates are incidence figures, whereas cross-sectional studies are prevalence-based?
In reply to | Sujit Rathod

Re: Maternal mortality in Europe

by | JUDITH MARGARET BURCHARDT -
What is the study design?
Ecological study

What is the numerator for the maternal mortality rate calculation?
Number of deaths of women within 42 days of them giving birth to a live baby in a given time period

What is the denominator for the maternal mortality rate calculation?
Number of live babies born in the same given time period

Is the maternal mortality rate really a rate, in the epidemiologic sense?
No. A rate is the number of events per unit time.

Is the maternal mortality rate an incidence or prevalence figure?
Tricky question. Incidence is the number of new events per unit time. Prevalence is the number of events per unit time. As mortality is always new it could be considered to be an incidence, but then it could also be considered to be a prevalence as it is an event. So either or both.

Mothers in the UK are three times more likely to die around the time of pregnancy compared with those in Norway, according to an international analysis of data.

Explain how the ‘three times’ figure was calculated. What is the name of this epidemiologic measure?
UK numerator/denominator as defined above is 3x higher than Norwegian numerator/denominator. Rate ratio

FOR RETURNING STUDENTS

What is the benefit/usefulness of presenting unstandardized maternal mortality rates?
Absolute numbers of maternal deaths matter very much. Useful for public health and planning.

What is the benefit/usefulness of presenting standardized maternal mortality rates? What variable(s) should be used to standardize?
Standardisation by age or ethnicity or both helps to understand the cause of the maternal deaths. If may be that it is confounded by age - the article mentions it is more common in younger, older and foreign mothers when compared to others. Knowing this can change policies so that more monitoring is offered to mothers in these groups.

Thank you Sujit for the question and everyone for the answers - it does seem strange that the denominator is number of live births - I wonder how that came about? I agree with Fathima that number of women who have had a live birth would seem a much more rational denominator. Perhaps the number of live births is easier to count in some countries?

Judith
Accessibility
Reset All

Background Colour

Font Face

Font Size

1

Text Colour