Differences in coronavirus disease 2019 screening, diagnosis, and hospitalization rates among U.S. pregnant women with and without a substance use disorder.

OBJECTIVE: Pregnant women are at increased risk for severe morbidity and mortality due to respiratory infections like SARS-CoV-2 (COVID-19) during pregnancy1. Those with substance use disorders (SUD) may be especially vulnerable due to high rates of smoked tobacco, cannabis and methamphetamine use which adversely affect pulmonary function.2 However, little is known about differential effects of COVID-19 on pregnant women with and without SUD. STUDY
DESIGN: This was a retrospective cohort study of commercially insured pregnant women ages 15-44 using national administrative healthcare data from Optum's deidentified Clinformatics® Data Mart Database version 8.1 (2007-2020). Women with a delivery hospitalization between January 1, 2020-August 19, 2020 continuously enrolled in insurance for ≥8 weeks during pregnancy and 6 weeks postpartum were included. SUD was defined using ICD-10 diagnoses for ≥ 1 SUD during pregnancy including alcohol, amphetamine, cannabis, cocaine, opioid or other substance use disorder.3 COVID-19 screening and diagnosis were defined using ICD-10 coding and reporting guidelines published by the CDC.4, 5 We used T-tests, Fisher's exact tests, and quantile regression (percentiles) to examine statistical differences. This study was exempt by the University of Pittsburgh IRB because deidentified healthcare data were used.
RESULTS: Among 65,009 pregnancies, 2,616 (4.0%) had ≥1 SUD diagnosis. Almost half of the pregnant women in the cohort were non-Hispanic White (48%) and most lived in the South (41%) or the Midwest (28%) region of the United States (Table). Pregnant women with an SUD diagnosis were significantly more likely to be younger (30±6 vs 32±5; p<.05). Compared with the overall sample, non-Hispanic Black (10% vs 7%; p,.05) and non-Hispanic White (54% vs 48%; p<.05) women were overrepresented among those with SUD as opposed to Hispanic and non-Hispanic Asian women. Overall, 27% of pregnant women were screened and evaluated for COVID-19 during pregnancy or postpartum. Women without an SUD diagnosis were significantly more likely to be screened for COVID-19 during pregnancy (27% vs 24%; p<0.001) than those with an SUD. Overall, 3.4% of pregnant women were diagnosed with COVID-19, mostly (86%) in the third trimester. The prevalence of a COVID-19 diagnosis was higher among those with a SUD compared to those without an SUD (5% vs 3%; p<0.05). Further, there was a higher percentage of COVID-19 related hospitalizations among pregnant women with an SUD (36%) than those without an SUD (8%) (p<0.001). The median length of stay was one day longer among those with a SUD (3 vs 2 days; p<0.01) compared to those without an SUD. There were no COVID-19 related deaths reported during pregnancy or postpartum.
CONCLUSION: In this national cohort, pregnant women with a SUD had a higher COVID-19 diagnosis rate than those without a SUD, despite lower screening rates. Among those with a COVID-19 diagnosis, pregnant women with a SUD had a significantly higher rate of COVID-19 related hospitalizations and longer median length of stay than those without a SUD. Efforts to improve COVID-19 screening rates and decrease morbidity associated with COVID-19 diagnosis among pregnant women with SUDs may be warranted.

Objective

Pregnant women are at an increased risk of severe morbidity and mortality due to respiratory infections like SARS-CoV-2 (COVID-19) during pregnancy. Those with substance use disorders (SUDs) may be especially vulnerable due to high rates of smoked tobacco, cannabis, and methamphetamine use which adversely affect pulmonary function. However, little is known about the differential effects of COVID-19 on pregnant women with and without SUD.

Study Design

This was a retrospective cohort study of commercially insured pregnant women ages 15 to 44 years using national administrative healthcare data from Optum’s (Eden Prairie, MN) deidentified Clinformatics Data Mart Database version 8.1 (2007–2020). Women with a delivery hospitalization between January 1, 2020, and August 19, 2020, continuously enrolled in insurance for ≥8 weeks during pregnancy and 6 weeks after delivery were included. SUD was defined using the tenth revision of the International Classification of Diseases (ICD-10) diagnoses for ≥1 SUD during pregnancy, including alcohol, amphetamine, cannabis, cocaine, opioid, or other SUD. COVID-19 screening and diagnosis were defined using ICD-10 coding and reporting guidelines published by the Centers for Disease Control and Prevention. , We used t tests, Fisher exact tests, and quantile regression (percentiles) to examine statistical differences. This study was exempt by the University of Pittsburgh Institutional Review Board because deidentified healthcare data were used.

Results

Among 65,009 pregnancies, 2616 (4.0%) had ≥1 SUD diagnosis. Almost half of the pregnant women in the cohort were non-Hispanic White (48%), and most lived in the South (41%) or the Midwest (28%) region of the United States (Table ). Pregnant women with an SUD diagnosis were significantly more likely to be younger (30±6 vs 32±5; P<.05). Compared with the overall sample, non-Hispanic Black (10% vs 7%; P<.05) and non-Hispanic White women (54% vs 48%; P<.05) were overrepresented among those with SUD as opposed to Hispanic and non-Hispanic Asian women. Overall, 27% of pregnant women were screened and evaluated for COVID-19 during pregnancy or postpartum. Women without an SUD diagnosis were significantly more likely to be screened for COVID-19 during pregnancy (27% vs 24%; P<.001) than those with an SUD. Overall, 3.4% of pregnant women were diagnosed as having COVID-19, mostly (86%) in the third trimester. The prevalence of a COVID-19 diagnosis was higher among those with an SUD than those without an SUD (5% vs 3%; P<.05). Furthermore, there was a higher percentage of COVID-19–related hospitalizations among pregnant women with an SUD (36%) than those without an SUD (8%) (P<.001). The median length of stay was 1 day longer among those with an SUD (3 vs 2 days; P<.01) than those without an SUD. There were no COVID-19–related deaths reported during pregnancy or postpartum.

Table

COVID-19 screening, diagnosis, and related hospitalization rates among pregnant persons in the United States, overall and by SUD diagnosis

Characteristic	Overall (n=65,009)	SUD diagnosisa (n=2616)	No SUD diagnosis (n=62,393)
Demographic characteristics
Age at delivery, yb	32±5	30±6c	32±5c
Race and ethnicityd
Non-Hispanic White	31,215 (48)	1400 (53.5)c	29,815 (47.8)c
Hispanic or Latino	6805 (10.5)	169 (6.5)c	6636 (10.6)c
Non-Hispanic Black	4583 (7.1)	251 (9.6)c	4332 (6.9)c
Non-Hispanic Asian	3467 (5.3)	44 (1.7)c	3423 (5.5)c
United States region
Northeast	6501 (10)	198 (7.6)c	6303 (10.1)c
South	26,624 (41)	1052 (40.2)	25,572 (41)
Midwest	17,950 (27.6)	955 (36.5)c	16,995 (27.2)c
West	13,789 (21.2)	410 (15.7)c	13,379 (21.4)c
COVID-19 screening, diagnosis, and related hospitalization rates
COVID-19 screening test performede	17,231 (26.5)	632 (24.2)f	16,599 (26.6)f
COVID-19 diagnosis	2241 (3.4)	119 (4.5)f	2122 (3.4)f
First trimester	55 (2.5)	3 (2.5)	52 (2.5)
Second trimester	252 (11.2)	14 (11.8)	238 (11.2)
Third trimester	1934 (86.3)	102 (85.7)f	1832 (86.3)f
COVID-19–related hospitalizationsg	215 (9.6)	43 (36.1)c	172 (8.1)c
Antepartum	149 (69.3)	34 (79.1)f	115 (66.9)f
Postpartum	66 (31.7)	9 (21.9)f	57 (33.1)f
Hospital length of stay, dh
Median (IQR)	2 (3)	3 (3)f	2 (3)f
Min–max range	1–35	0–29	0–35
25%–75% percentile	1–4	2–5f	1–4f

IQR, interquartile range; SUD, substance use disorder.

Gao. Differences in COVID-19 screening, diagnosis, and hospitalization rates. Am J Obstet Gynecol 2021.

Includes any diagnosis of the following SUDs in pregnancy (calculated as 280 days before delivery): tobacco, opioid, alcohol, cannabis, cocaine, amphetamine, sedative, hallucinogen, inhalants, and other psychoactive and nonpsychoactive substances

Mean±standard deviation, statistical significance assessed with 2-sample t test

P<.001

Approximately 29% of race and ethnicity data were missing owing to a lack of collection or patient disclosure

Screening includes encounters in which individuals had a charge for a COVID-19 test with negative or unknown test results

P<.05

Excludes delivery hospitalization

Statistical significance assessed using quantile regression, excludes the delivery hospitalization.

Conclusion

In this national cohort, pregnant women with an SUD had a higher COVID-19 diagnosis rate than those without an SUD, despite lower screening rates. Among those with a COVID-19 diagnosis, pregnant women with an SUD had a significantly higher rate of COVID-19–related hospitalizations (36% vs 8%, P<.001) and longer median length of stay (3 vs 2 days, P<.01) than those without an SUD. Efforts to improve COVID-19 screening rates and decrease morbidity associated with COVID-19 diagnosis among pregnant women with SUDs may be warranted.