Portal de Periódicos da CAPES

New modelled estimates of maternal mortality

2010; Elsevier BV; Volume: 375; Issue: 9730 Linguagem: Inglês

10.1016/s0140-6736(10)60924-5

1474-547X

Affette McCaw‐Binns, Karen Lewis‐Bell,

Global Health and Epidemiology

Monitoring progress towards the fifth Millennium Development Goal (MDG) to reduce maternal mortality by 75% is challenging when developing countries lack reliable vital registration systems; consequently estimates from statistical models have been used.1Hogan MC Foreman KJ Naghavi M et al.Maternal mortality for 181 countries, 1980–2008: a systematic analysis of progress towards Millennium Development Goal 5.Lancet. 2010; 375: 1609-1623Summary Full Text Full Text PDF PubMed Scopus (1493) Google Scholar, 2Hill K Thomas K AbouZahr C et al.Estimates of maternal mortality worldwide between 1990 and 2005: an assessment of available data.Lancet. 2007; 370: 1311-1319Summary Full Text Full Text PDF PubMed Scopus (363) Google Scholar After three reproductive-age mortality studies in 1981–83, 1986–87, and 1993–95, Jamaica began routine surveillance in 1998 by making maternal death a class I notifiable event. Maternal deaths are reported as they occur, backed up by active surveillance of deaths among women aged 10–50 years to determine if they are pregnancy-related. With more than 95% of births taking place in hospital, we capture most events. Findings inform staff training and system-improvement strategies. We compared our findings with those of models by Margaret Hogan and colleagues1Hogan MC Foreman KJ Naghavi M et al.Maternal mortality for 181 countries, 1980–2008: a systematic analysis of progress towards Millennium Development Goal 5.Lancet. 2010; 375: 1609-1623Summary Full Text Full Text PDF PubMed Scopus (1493) Google Scholar and by Hill and colleagues (WHO).2Hill K Thomas K AbouZahr C et al.Estimates of maternal mortality worldwide between 1990 and 2005: an assessment of available data.Lancet. 2007; 370: 1311-1319Summary Full Text Full Text PDF PubMed Scopus (363) Google Scholar The Hogan model relies on vital registration for 82% of 181 countries, including Jamaica, and on surveillance systems in less than 1%. Except for 2005, WHO figures track our data; however, the Hogan model grossly underestimates our experience and we question the validity of their model assumptions. In 1998, of 49 maternal deaths identified by surveillance, only 36 had medical certificates. Others were lost in an inefficient Coroner's system which may result in death registration years after occurrence. Just 13 of 36 were correctly classified as maternal deaths, yielding a maternal mortality ratio of only 29 maternal deaths per 100 000 livebirths, which is remarkably similar to the 34 per 100 000 reported by Hogan and colleagues for 2008. The 36 registered deaths, if correctly classified, would have yielded a ratio of 81 per 100 000—ie, still substantially less than the actual 1998 ratio of 110 per 100 000. Similar wide discrepancies are seen for Guyana (4701Hogan MC Foreman KJ Naghavi M et al.Maternal mortality for 181 countries, 1980–2008: a systematic analysis of progress towards Millennium Development Goal 5.Lancet. 2010; 375: 1609-1623Summary Full Text Full Text PDF PubMed Scopus (1493) Google Scholar vs 1432Hill K Thomas K AbouZahr C et al.Estimates of maternal mortality worldwide between 1990 and 2005: an assessment of available data.Lancet. 2007; 370: 1311-1319Summary Full Text Full Text PDF PubMed Scopus (363) Google Scholar), supporting our view that vital data, at least for these countries, are unreliable. Systems are needed to validate the quality of vital data before they are used in models to estimate maternal mortality. Over 25 years, the epidemiology of our maternal deaths has changed remarkably.3McCaw-Binns A Alexander SF Lindo JLM et al.Epidemiologic transition in maternal mortality and morbidity: new challenges for Jamaica.Int J Gynecol Obstet. 2007; 96: 226-232Summary Full Text Full Text PDF PubMed Scopus (21) Google Scholar In 1981–83, 47% of deaths were from hypertension and haemorrhage, whereas indirect deaths only accounted for 12%; by 2004–06, both direct causes were down to 20%, with indirect deaths climbing to 44%. Our successes in reducing direct deaths4McCaw-Binns AM Ashley DE Knight L et al.Strategies to prevent eclampsia in a developing country: I, re-organisation of maternity services.Int J Gynecol Obstet. 2004; 87: 286-294Summary Full Text Full Text PDF PubMed Scopus (32) Google Scholar were mostly offset by increased indirect deaths, which are more likely to be misclassified if heart disease or HIV deaths were certified without noting the pregnancy. From 120 deaths per 100 000 livebirths in 1986–87, Jamaica's MDG of 30 per 100 000 had only reached 88 per 100 000 by 2004–06. We are evaluating a new, more expensive, model of high-risk care to improve care for pregnant women with medical and obstetric complications. Now, would a Jamaican legislator, asked to invest more in maternal health, support this request in an era of economic retrenchment when international “experts” report Jamaica's practical achievement of MDG5? Such experts seem to be oblivious of the implications of their estimations at a country level. We declare that we have no conflicts of interest. New modelled estimates of maternal mortality – Authors' replyWe appreciate the rich set of letters in response to our paper on maternal mortality. The authors of the letters raise many important points, but we focus our short response on four larger themes that have been raised. Full-Text PDF