health

Risk of maternal mortality in women with severe anaemia during pregnancy and post partum: a multilevel analysis


In our analysis of a dataset of 312 281 pregnancies in 29 countries from the WHO Multicountry Survey, the odds of maternal death were twice as high in those with severe anaemia compared with those without severe anaemia. The association seemed to be moderately strong, temporal, and consistent and was reproducible in both multilevel and propensity score regression analyses. Previously published estimates of the relation between anaemia and maternal mortality were limited by low event rates, low geographical variability,4x4Rukuni, R, Bhattacharya, S, Murphy, MF, Roberts, D, Stanworth, SJ, and Knight, M. Maternal and neonatal outcomes of antenatal anemia in a Scottish population: a retrospective cohort study. Acta Obstet Gynecol Scand. 2016; 95: 555–564
Crossref | PubMed | Scopus (4)
| Google ScholarSee all References

Our finding of a link between severe maternal anaemia and mortality challenges the assumption that haematological indices such as haemoglobin are not suitable outcome measures.6x6Cantor, AG, Bougatsos, C, Dana, T, Blazina, I, and McDonagh, M. Routine iron supplementation and screening for iron deficiency anemia in pregnancy: a systematic review for the US Preventive Services Task Force. Ann Intern Med. 2015; 162: 566–576
Crossref | PubMed | Scopus (24)
| Google ScholarSee all References

Our study had several strengths. The dataset was large and diverse (both geographically and ethnically), with ample numbers of events and many available confounding variables, allowing for adjustment of statistical models without compromising the model’s goodness-of-fit.22x22Cheung, GW and Rensvold, RB. Evaluating goodness-of-fit indexes for testing measurement invariance. Struct Equ Modeling. 2002; 9: 233–255
Crossref | Scopus (3979)
| Google ScholarSee all References

The association of anaemia with poor maternal outcomes, including increased morbidity related to post-partum haemorrhage,23x23Say, L, Chou, D, Gemmill, A et al. Global causes of maternal death: a WHO systematic analysis. Lancet Glob Health. 2014; 2: e323–e333
Summary | Full Text | Full Text PDF | PubMed | Scopus (604)
| Google ScholarSee all References

Our study also had several limitations. Anaemia data were collected according to the WHO definition of severe adverse outcome.12x12Souza, JP, Gulmezoglu, AM, Carroli, G, Lumbiganon, P, and Qureshi, Z. The world health organization multicountry survey on maternal and newborn health: study protocol. BMC Health Service Res. 2011; 11: 286
Crossref | PubMed | Scopus (26)
| Google ScholarSee all References

More than 4000 cases were excluded from our analyses because of missing data. This issue was due to how data were gathered in the original survey12x12Souza, JP, Gulmezoglu, AM, Carroli, G, Lumbiganon, P, and Qureshi, Z. The world health organization multicountry survey on maternal and newborn health: study protocol. BMC Health Service Res. 2011; 11: 286
Crossref | PubMed | Scopus (26)
| Google ScholarSee all References

We used propensity score regression analysis in addition to multivariable logistic models, because such analyses allow for comparisons of outcomes of interest from observational data via a scoring system for groups with similar characteristics, mimicking a randomised study.27x27Austin, PC. An Introduction to propensity score methods for reducing the effects of confounding in observational studies. Multivariate Behav Res. 2011; 46: 399–424
Crossref | PubMed | Scopus (1629)
| Google ScholarSee all References

The availability of care varied between the individual facilities contributing data to the dataset. An analysis stratified by individual facilities was unfeasible and therefore was not done. We controlled for variables that could predispose women to postnatal anaemia in our multilevel and propensity regression analyses and did sensitivity analyses to establish whether the relation between severe anaemia and maternal death was solely related to post-partum haemorrhage within these data.

We are confident that our findings are a robust demonstration of an independent link between severe anaemia and maternal death. Although our data cannot explain a direct causal relation between severe anaemia and maternal death, examination of the extent to which causal criteria have been met by our work is worthwhile.28x28Hill, AB. The environment and disease: association or causation?. Proc Royal Soc Med. 1965; 58: 295–300
PubMed
| Google ScholarSee all References

Our sensitivity analyses showed that, when cases of post-partum haemorrhage were removed from the multilevel regression model, the strength of the association between severe anaemia and maternal death increased. Thus, inclusion of post-partum haemorrhage in our models could have resulted in underestimation of the strength of the relation. Furthermore, our sensitivity analyses emphasise that severe anaemia in pregnant and postnatal women strongly and independently contributes to adverse outcomes, irrespective of the cause. Anaemia treatment and prevention in pregnancy should remain a global priority.



Source link قالب وردپرس