Objective
Women are a fast-growing segment of the incarcerated population, and those who enter prisons, jails, and detention centers often do so with undertreated mental health conditions, substance use disorders, chronic conditions, and infectious diseases. , About 4% of women who enter US prisons and jails will be pregnant at intake and will require prenatal care tailored to meet these needs. There are limited and inconsistent data on the risk for preterm delivery in this population and no data on the risk for other pregnancy complications such as severe maternal morbidity. Outcomes may be poorer because of limited or suboptimal care. We used a large administrative database to study pregnancy complications associated with incarceration.
Study Design
We conducted a serial cross-sectional analysis of delivery hospitalizations recorded in the National Inpatient Sample (NIS) from October 1, 2015 through December 31, 2018. The NIS approximates a stratified sample of 20% of United States discharges and, when weighted, can be used to derive national estimates. We included the cases of women aged 15 to 54 years recorded in the NIS who were hospitalized for delivery. To identify delivery hospitalizations, we applied an algorithm of the International Classification of Diseases Tenth Revision Clinical Modification (ICD-10-CM) codes that have been shown previously to capture >95% of deliveries in the NIS. We defined incarceration using the ICD-10-CM diagnosis code Z65.1. We calculated the standardized mean difference (SMD) for demographic comparisons based on the presence or absence of an incarceration diagnosis; for these comparisons, we considered ≥0.1 (10%) to be a meaningful difference. Using a weighted survey-adjusted logistic regression and adjusting for demographic and medical factors, we evaluated the association between incarceration and the risk for the following pregnancy complications: abruption and antepartum hemorrhage, preterm delivery, cesarean delivery, hypertensive disorders of pregnancy, postpartum hemorrhage, and severe maternal morbidity defined using a Centers for Disease Control and Prevention composite excluding transfusion. We reported the measures of association as unadjusted and adjusted odds ratios (aORs) with 95% confidence intervals (CIs). This analysis was granted an exemption from the institutional review board because the NIS is de-identified.
Results
Of an estimated 12 million delivery hospitalizations included in our analysis, 1580 (0.01%) (316 unweighted) had an incarceration diagnosis. Deliveries associated with incarceration had higher rates of mental health conditions (SMD, 52.2%), substance use disorders (SMD, 89.4%), viral hepatitis (SMD, 57.1%), sexually-transmitted infections (SMD, 24.4%), and tobacco use (SMD, 66.4%) than those who were not incarcerated; they also had higher rates of Medicaid insurance (SMD, 116.0% for payer) and were more likely to be in the lower income quartiles (SMD, for 36.5% ZIP code income quartile) ( Supplemental Table ). Women incarcerated at delivery had higher odds of abruption and antepartum hemorrhage (aOR, 2.08; 95% CI, 1.09–3.97) and preterm delivery (aOR, 2.13; 95% CI, 1.51–2.99) in adjusted analyses ( Figure ).
Conclusion
This study demonstrated a higher prevalence of preterm delivery among women who are incarcerated, a finding similar to other studies. , The finding of increased abruption among incarcerated pregnant women is new, but this is a population with a cluster of risk factors—smoking, substance use, and lower socioeconomic status—well known to be associated with abruption. Because these associations retained significance after adjustment it suggests that some additional contribution to maternal risk is conferred by incarceration. A limitation of this study is that the incarceration diagnosis code is not validated or designed to capture national estimates. There are no systematic data collection systems that capture data from all jails, prisons, and detention centers in the United States. Despite new efforts to collect data on pregnancy outcomes in prisons, data on complications are limited. Our findings suggest that there is a need for better monitoring of pregnancy complications in this high-risk population and for improvements in care quality and access.
Appendix
| Characteristics of study population | No incarceration diagnosis, n (%) | Incarceration diagnosis, n (%) | Absolute SMD |
|---|---|---|---|
| Demographics | |||
| Year of delivery | 16.1% | ||
| 2015 | 921,955 (7.66) | 95 (6.01) | |
| 2016 | 3,782,081 (31.41) | 405 (25.63) | |
| 2017 | 3,702,321 (30.75) | 510 (32.28) | |
| 2018 | 3,634,774 (30.19) | 570 (36.08) | |
| Maternal race | 32.4% | ||
| Non-Hispanic White | 6,028,260 (50.06) | 725 (45.89) | |
| Non-Hispanic Black | 1,715,930 (14.25) | 410 (25.95) | |
| Hispanic | 2,380,768 (19.77) | 260 (16.46) | |
| Other | 1,335,024 (11.09) | 105 (6.65) | |
| Unknown | 581,149 (4.83) | 80 (5.06) | |
| Maternal age (y) | 26.5% | ||
| 15–19 | 626,909 (5.21) | 95 (6.01) | |
| 20–24 | 2,401,888 (19.95) | 380 (24.05) | |
| 25–29 | 3,499,657 (29.06) | 565 (35.76) | |
| 30–34 | 3,405,058 (28.28) | 360 (22.78) | |
| 35–39 | 1,725,214 (14.33) | 145 (9.18) | |
| 40–54 | 382,405 (3.18) | 35 (2.22) | |
| Payer | 116.0% | ||
| Medicare | 87,470 (0.73) | a | |
| Medicaid | 5,157,556 (42.83) | 940 (59.49) | |
| Private Insurance | 6,145,250 (51.04) | 190 (12.03) | |
| Self-pay | 301,435 (2.50) | 145 (9.18) | |
| No Charge | 7315 (0.06) | a | |
| Other | 327,135 (2.72) | 285 (18.04) | |
| Unknown | 14,970 (0.12) | a | |
| Obstetrical and medical factors | |||
| Multiple gestation | 228,520 (1.90) | 15 (0.95) | 0.0801 (8.01%) |
| Any ≥1 mental health condition b | 746,605 (6.20) | 385 (24.37) | 0.5218 (52.18%) |
| Any ≥1 substance use disorder b | 290,200 (2.41) | 535 (33.86) | 0.8941 (89.41%) |
| Communicable diseases | |||
| HIV | 12,595 (0.10) | a | NA |
| Viral hepatitis | 84,885 (0.70) | 250 (15.82) | 0.5710 (57.10%) |
| Tuberculosis | 1250 (0.01) | a | NA |
| Any ≥1 sexually transmitted infection b | 236,725 (1.97) | 110 (6.96) | 0.2437 (24.37%) |
| Chronic conditions | |||
| Obesity | 1,147,684 (9.53) | 190 (12.03) | 0.0805 (8.05%) |
| Pregestational diabetes | 143,960 (1.20) | 35 (2.22) | 0.0788 (7.88%) |
| Chronic hypertension | 258,010 (2.14) | 45 (2.85) | 0.0452 (4.52%) |
| Asthma | 584,680 (4.86) | 200 (12.66) | 0.2787 (27.87%) |
| Tobacco use | 640,819 (5.32) | 460 (29.11) | 0.6640 (66.40%) |
| Hospital factors | |||
| Hospital location and teaching status | 0.2788 (27.88%) | ||
| Rural | 1,118,480 (9.29) | 125 (7.91) | |
| Urban nonteaching | 2,708,909 (22.50) | 195 (12.34) | |
| Urban teaching | 8,213,741 (68.21) | 1260 (79.75) | |
| Region | 0.2100 (21.00%) | ||
| Northeast | 1,913,859 (15.89) | 180 (11.39) | |
| Midwest | 2,541,384 (21.11) | 305 (19.30) | |
| South | 4,710,128 (39.12) | 605 (38.29) | |
| West | 2,875,760 (23.88) | 490 (31.01) | |
| Adverse pregnancy outcomes | |||
| Nontransfusion severe maternal morbidity | 93,500 (0.78) | 30 (1.90) | |
| Hypertensive disorders of pregnancy | 1,270,824 (10.55) | 195 (12.34) | |
| Preterm delivery | 544,750 (4.52) | 165 (10.44) | |
| Postpartum hemorrhage | 438,135 (3.64) | 65 (4.11) | |
| Cesarean delivery | 3,880,037 (32.22) | 550 (34.81) | |
| Abruption and antepartum hemorrhage | 182,550 (1.52) | 55 (3.48) |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
