A retrospective cohort study led by rheumatologist, of the Hospital for Special Surgery in New York, and colleagues looked at data from the National Inpatient Sample database involving 93,820 pregnant women with SLE and 78,045,054 without SLE who were hospitalized in the United States between 1998 and 2015.
The results showed that over the 18-year study period, in-hospital maternal deaths per 100,000 admissions declined among patients with as well as those without SLE (442 vs. 13 for 1998-2000 and less than 50 vs. 10 for 2013-2015), although the decrease was greater in women with SLE (difference in trends, P less than .002).
Fetal mortality declined both for patients with SLE (268 deaths per 10,000 deliveries in 1998-2000 vs. 153 in 2013-2015) and those without SLE (72 deaths per 10,000 deliveries in 1998-2000 vs. 66 in 2013-2015).
“Although the decrease in fetal mortality seems somewhat greater in patients with SLE than those without it, the difference in trends is not statistically significant (P = .064),” the study authors noted in their paper, published in.
Although patients with SLE in their study showed greater progress in rates of preeclampsia or eclampsia and length of stay, they still had worse outcomes in all measures when compared against women without SLE, the authors noted.
“Our study provides nationwide evidence that SLE pregnancy outcomes have become markedly better in the past 2 decades and continue to improve. However, SLE pregnancy risks remain high, and more work is needed to ensure good pregnancy outcomes among women with SLE,” they concluded.
Writing in an accompanying editorial, Megan E.B. Clowse, MD, of Duke University, Durham, N.C., noted that although the study findings of a progressive reduction in maternal mortality seemed very likely, the absolute decrease seen – from 140 maternal deaths per 100,000 births to fewer than 50 in 2013-2015 – warranted some reflection (Ann Intern Med. 2019;171:212-3.).
“SLE pregnancy management has not advanced within the past 5 years to an extent great enough to explain such a large drop in mortality,” she wrote.
There were also question marks around whether the NIS data should be used to analyze very rare events, such as maternal deaths in women with SLE. The SLE diagnosis in the database may also have included pregnancies in women who did not have SLE but instead had an elevated level of antinuclear antibody or lupus anticoagulant, which “may have diluted the frequency of poor outcomes,” Dr. Clowse wrote.
“For these reasons, I am concerned that the observed decline in maternal mortality may be an artifact, underestimating the ongoing risk of pregnancy for women with SLE,” Dr. Clowse wrote.
She added that recent analyses demonstrated that the use of hydroxychloroquine and aspirin in SLE pregnancy was not widespread.
“The inaugural reproductive health guidelines soon to be published by the American College of Rheumatology will have the potential to help expand state-of-the-art approaches to the management of pregnant women with SLE seen in everyday practice,” she concluded.
The study had no primary funding source and no conflicts of interest were declared. Dr. Mehta is supported by the C. Ronald MacKenzie Young Scientist Endowment Award. Dr. Clowse reported grants from GlaxoSmithKline and personal fees from UCB, both outside the submitted work.
SOURCE: Mehta B et al. Ann Intern Med. 2019;171:164-71.