Severe Maternal Morbidity and Maternal Mortality Chris Glantz, MD, MPH
Severe Maternal Morbidity (SMM) Severe/multisystem organ dysfunction or need for major intervention Defined by CDC as one or more of 18 conditions 1. Acute MI/aneurysm 10. Pulmonary edema 2. Acute renal failure 11. Severe anesthesia comps 3. ARDS 12. Sepsis 4. AFE 13. Shock 5. Cardiac arrest/V fib 14. Sickle cell crisis 6. DIC 15. Air or thrombotic embolism 7. Eclampsia 16. Blood transfusion 8. CHF 17. Hysterectomy 9. Stroke 18. Temporary trach/vent 50- 100x more common than death; Incidence ≈1.5%
Severe Maternal Morbidity in the US More common in Black women Increasing rates of PPH due to atony The lesser rise is mostly due to increases in hysterectomy and assisted ventilation https://www.cdc.gov/reproductivehealth/maternalinfanthealth/severematernalmorbidity.html
SPDS: Birth Certificate Database • Has fields for transfusion, unplanned hysterectomy, eclampsia, and ICU admission • Does not have specific fields for the other 15 CDC conditions, but most of these would be managed in an ICU • Collectively, ICU admission should capture these 15 • Thus, SMM ≈ Transfusion | Eclampsia | Hyst | ICU
Morbidities from Birth Certificate 80% 70% 60% 50% 40% 30% 20% 10% 0% Transfusion ICU Eclampsia Hyst Combination Alone Total (in combination)
SMM by Monroe County Hospital 1.4% 1.2% 1.0% 0.8% 0.6% 0.4% 0.2% 0.0% HH UH RGH SMH SMH: more trans, hyst, & ICU RGH: fewer hyst & trans
Rochester’s Northern Crescent Crescent = ZIP 14605, 14609, and 14621 High rates of poverty and unemployment, and low rates of education and home ownership
Monroe County 2009 - 2018 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% Crescent NonCrescent
Severe Maternal Morbidity By Year 2.0% 1.8% 1.6% 1.4% 1.2% 1.0% 0.8% 0.6% 0.4% 0.2% 0.0% 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 NonCrescent Crescent TOTAL Trend: Not Statistically Significant
Severe Morbidity by Race 1.6% 1.4% 1.2% 1.0% 0.8% 0.6% 0.4% 0.2% 0.0% Asian Black WhiteHisp WhiteNonHisp Other Total Crescent 1.24% NonCrescent 0.97% Blacks had the highest rates of transfusions, ICU admissions, and eclampsia
Controlling for Possible Confounders Using Logistic Regression • Nulliparity • Not Fin HS • Diabetes • Poor PNC (<3 visits) • HTN (chronic or PIH) • Smoking • Cong Anom • Drugs • Mult Gest • EtOH • Medicaid • Obesity • AMA • Race • Teen • Crescent • College Deg
Adjusted SMM: Odds Ratios 4.5 4.0 3.5 3.0 2.5 2.0 1.5 ? 1.0 0.5 0.0 The other 8 variables were not statistically significant (including Medicaid)
Morbidity Conclusions • Transfusion accounts for the majority of SMM • After adjusting for multiple demographic and medical factors: • Black women have a higher rate of severe maternal morbidity • Hispanics do not differ significantly from non-Hispanic Caucasians • Medicaid, smoking, and living in the Crescent were not independently associated with SMM • Caveat: there may be unmeasured factors influencing the association
Maternal Mortality • Various definitions (often confusing*) • Early (within 42 days of del) or late (within 1 year) • Pregnancy-related vs nonrelated • Related = due to complication of pregnancy or pregnancy- related deterioration of an existing condition • Mat Mort Ratio (MMR) = deaths per 100,000 liveborns • ≈700 maternal deaths per year in US *Early v late, preg-related v non-related, preg-associated v non-assoc, direct v indirect
Worldwide Maternal Mortality Indirect causes* Hemorrhage HTN Infection Direct causes† Abortion VTE 0% 5% 10% 15% 20% 25% 30% *e.g., HIV, other med dx Global causes of maternal death: †e.g., obstructed labor a WHO systematic analysis, 2014
USA: Causes of Maternal Death (CDC) Other CV Other non-CV Infection Hemorrhage Cardiomyopathy VTE Stroke Hypertension Unknown AFE 0.0% 2.0% 4.0% 6.0% 8.0% 10.0% 12.0% 14.0% 16.0% Top 3: Other CV, Other non-CV, Infection MMWR / May 10, 2019 / Vol. 68 / No. 18
USA: Causes of Maternal Death (MMRIA) From a nine-state review of 680 deaths: Top 3: Hem, CV, Infection Building U.S. Capacity to Review and Prevent Maternal Deaths, 2018. http://reviewtoaction.org/Report_from_Nine_MMRCs
Causes of Maternal Death (NYS) Hemorrhage HTN Other non-CV VTE Other CV Stroke Infection Cardiomyopathy 0% 5% 10% 15% 20% 25% 30% 2006-8 2012-14 Top 3 2006-8: Hem, HTN, Other nonCV Number of cases ranged from 4 to 29 Top 3 2012-14: VTE, Hem, infection
Timing of Death 31% 57% 12% CDC, 2019 MMRIA, 2018 Percentages Vary
Maternal Mortality in the USA Up to 50% of deaths are potentially preventable https://www.cdc.gov/reproductivehealth/maternalinfant health/pregnancy-mortality-surveillance-system.htm
Making America Not-So-Great Again Population heterogeneity? Lack of universal coverage? Large geographic areas underserved? Disparities in care? Sicker/heavier women? Too many unnecessary interventions? Not enough midwives? Tracking differently/more accurately?
NYS: Maternal Mortality by Year ROS = “Rest of State”...
USA: Racial Disparity in Maternal Mortality 45 40 35 30 25 20 15 10 5 0 White Black Alaskan/indian Asian Hispanic Maternal Deaths per 100,000 liveborns
NYS: Maternal Mortality by Race
Difficult to Track Local Rates • Fortunately, death is very uncommon • But small changes in # of deaths large swings in rates • Difficult to ascertain deaths occurring up to 12 months after delivery • Local stats either not available or out of date • Attribution: prenatal care office vs. delivery hospital vs. hospital where death occurred
Maternal Mortality: Women Delivered at Strong 2013-18 • 12 deaths • 2 Antepartum • 1 ruptured ectopic and 1 heroin overdose at 23 wks • 2 during delivery admission • 1 AFE and 1 influenza • 8 Postpartum • 4 from cancer (1-11 mo pp) • 1 complications from lupus (3 mo pp) • 1 heroin overdose (5 mo pp) • 1 complications from neurosurgery (7 mo pp) • 1 complications from cystic fibrosis (11 mo pp)
Two Examples • SMH: 2800 del/year, mean 2 deaths (inc. late) • MMR = (2/2800)*100000 = 71 • If ±1 death in a given year, MMR = 36 or 107 • If only early deaths tracked (conception to 42 days pp), SMH’s MMR falls by half • If only pregnancy- related deaths tracked, SMH’s MMR = 18 • Level 1 Hosp X: 1000 del/year, and no deaths in 10 years • MMR = 0 = perfect! • If 1 death one year, MMR for that year = 100 = terrible
Mortality Conclusions • Although rare, maternal deaths do occur • US rate is rising (various possible explanations) • It’s declining in other countries • Stats on rates, timing, and cause of death vary by study and by definitions used • Accurate local rates require very long time periods • Racial disparities persist after adjusting for other factors • Reason unclear — biases? unmeasured factors?
Recommend
More recommend