​Shocking maternal deaths in Colorado

​In this post, Gwen provides details supporting an analysis of maternal deaths in Colorado reported in the September issue of the CORCRC newsletter. Colorado maternal deaths are up in recent years, and there are no easy answers to solve this public health tragedy.
 
Introduction
Maternal deaths are on the rise around the country, in contrast to trends around that world that have seen a 44% drop from 1990-2015. A recent analysis showed that maternal deaths in eight states that did not adopt a 2003 revision of the US standard death certificate over the period studied (Alaska, Hawaii, Colorado, North Carolina, Massachusetts, Virginia, West Virginia, and Wisconsin), rose 23.1% in 14 years, from 8.0 deaths/100,000 live births in 2000, to 10.4 in 2014.2 This analysis of multiple states grouped together was intended to increase statistical power due to the rarity of maternal death in the US. Yet because of the many factors that could affect maternal deaths that are region-specific, a combined analysis may obscure trends that could be important for public health concerns in each state.
 
Methods
Birth and death data in this analysis was obtained through the publicly available Colorado Health Information Dataset (CoHID), maintained by the Colorado Department of Public Health and Environment. Maternal deaths were defined as all pregnancy, childbirth and puerperium deaths, including pregnancy with abortive outcome (n = 5 over the study period). Median, average, standard deviation, and linear regression analyses were calculated using Microsoft Excel for Mac 2011, Version 14.1.0.

Results and Discussion
In the table, raw data and calculated maternal deaths per 100,000 live births is given for 1999 to 2015, the last year for which data were available.

Maternal mortality rates in Colorado are presented by year in the figure below. A linear regression analysis of the data in years 1999-2015 shows a wide variance from year to year, as predicted based on the small event numbers. Based on linear regression analysis, the state saw a 37% increase in maternal mortality from 1999-2015.
 
A simple linear regression analysis does not appear to fit the data well. The data appear to be approximately constant from 1999-2012, with a precipitous increase beginning in 2013, as seen in an analysis of maternal mortality in Texas.
 
Joinpoint analysis is designed to fit different linear trends within data, and to test for statistical significance. Due to the small event rate of maternal deaths in Colorado, the dataset is not considered sufficiently robust to obtain meaningful results from Joinpoint analysis.
The median and average mortality rate from 1999-2012 is 8 ± 4 deaths per 100,000 live births. In comparison, the average mortality rate from 2013-2015 is 21 ± 5 deaths per 100,000 live births (median 23).
 
In addition to low event rates making robust statistical analysis problematic, the lack of physician adjudication of death certificate data is also cause for concern in this analysis. Lack of physician adjudication can mean reporting errors are not corrected, which can be an important factor with low event rate data.
 
Despite these drawbacks, the substantial increase in maternal mortality beginning in 2013 in Colorado is a flag for public health officials to further investigate these tragic deaths.
 
Please read my full article in the September issue of the CORCRC newsletter for more details on this important public health issue.
 
References
[1] WHO, UNICEF, UNFPA, World Bank Group and the United Nations Population Division. Trends in Maternal Mortality: 1990 to 2015 Estimates by WHO, UNICEF, UNFPA, World Bank Group and the United Nations Population Division Executive Summary. WHO /RHR/15.23. Last accessed September 9, 2016.

[2] MacDorman MF, Declercq E, Cabral H, Morton C. Recent Increases in the U.S. Maternal Mortality Rate: Disentangling Trends From Measurement Issues. Obstet Gynecol. 2016 Sep;128:447-55.

[3] Colorado Health Information Dataset. http://www.cohid.dphe.state.co.us/ Accessed September 15, 2016.
 
[4] Lakshminarayan K, Larson JC, Virnig B, et al. Comparison of Medicare claims versus physician adjudication for identifying stroke outcomes in the Women's Health Initiative. Stroke. 2014 Mar; 45(3): 815–821. doi:  10.1161/STROKEAHA.113.003408

---

Murphy Medical Writing, LLC

Contact us for more information on your medical writing project!