Changes in Preterm Birth Phenotypes and Stillbirth at 2 Philadelphia

Letters

Table. Count and Proportion of Deliveries With a Recorded Stillbirth and Rate Ratio of Periods Before Lockdown and in Lockdown Adjusted by Calendar Montha

Prelockdown comparison period Lockdown period (April 2019 to June 2019) (April 2020 to June 2020) Incidence No. of Total No. with a Stillbirth rate, % No. of Total No. with a Stillbirth rate, % rate ratio P stillbirths delivery outcome (95% CI) stillbirths delivery outcome (95% CI) (95% CI) valueb All England 565 139 745 0.40 (0.37-0.44) 543 131 218 0.41 (0.38-0.45) 1.02 (0.91-1.15) .69 North of England 150 38 165 0.39 (0.33-0.46) 170 35 400 0.48 (0.41-0.56) 1.22 (0.98-1.52) .07 Midlands and East 150 42 220 0.36 (0.30-0.42) 150 39 295 0.38 (0.32-0.45) 1.07 (0.86-1.35) .53 of England London 130 25 960 0.50 (0.42-0.59) 125 24 640 0.51 (0.42-0.60) 1.02 (0.79-1.30) .90 South of England 135 33 390 0.40 (0.34-0.48) 110 31 880 0.35 (0.28-0.42) 0.85 (0.66-1.10) .21 a Subnational figures have been rounded to the nearest 5; percentages Secondary Uses Service, NHS Digital. Copyright 2020, reused with the calculated using rounded counts. Sources: April 2019-March 2020: Hospital permission of NHS Digital. All rights reserved. Episode Statistics, NHS Digital. Copyright 2020, reused with the permission of b Significance test at the .05 level. NHS Digital. All rights reserved (provisional data). April 2020-June 2020:

cohort. It will be important to continue to monitor pregnancy 4. Khalil A, von Dedelszan P, Draycott T, Ugwumadu A, O’Brien P, Magee L. outcomes in the future. Change in the incidence of stillbirth and preterm delivery during the COVID-19 pandemic. JAMA. 2020;324(7):705-706. doi:10.1001/jama.2020.12746 5. Coronavirus (COVID-19) in the UK. Public Health England. Accessed Julia Stowe, PhD September 17, 2020. https://coronavirus.data.gov.uk/cases Helen Smith, BSc(Hons) 6. Coronavirus and pregnancy—preserving maternal health across the Kate Thurland, MPH European Region. World Health Organisation Europe. Published June 30, 2020. Mary E. Ramsay, MBBS Accessed August 14, 2020. https://www.euro.who.int/en/health-topics/Life- Nick Andrews, PhD stages/maternal-and-newborn-health/news/news/2020/6/coronavirus-and- pregnancy-preserving-maternal-health-across-the-european-region Shamez N. Ladhani, PhD

Author Affiliations: Public Health England, Immunisation and Changes in Preterm Birth Phenotypes and Stillbirth Countermeasures Division, London, England (Stowe, Ramsay, Andrews, at 2 Philadelphia Hospitals During the SARS-CoV-2 Ladhani); Public Health England, Health Intelligence Division, London, England (Smith, Thurland). Pandemic, March-June 2020 The severe acute respiratory syndrome coronavirus 2 (SARS- Corresponding Author: Julia Stowe, PhD, Immunisation and Countermeasures, Public Health England, 61 Colindale Ave, London NW9 5EQ, England CoV-2) pandemic has had far-reaching implications, includ- ([email protected]). ing changes in societal stressors and health care delivery, Accepted for Publication: October 9, 2020. which may alter preterm birth risk. Previous studies in the Published Online: December 7, 2020. doi:10.1001/jama.2020.21369 US regarding SARS-CoV-2

Author Contributions: Dr Stowe had full access to all of the data in the study Related article page 86 in pregnancy focused on and takes responsibility for the integrity of the data and the accuracy of the data associations of SARS-CoV-2 analysis. infection with cesarean delivery, neonatal transmission, pre- Concept and design: Stowe, Andrews, Ladhani. 1 Acquisition, analysis, or interpretation of data: All authors. term birth, and stillbirth. In a relatively homogeneous Dan- Drafting of the manuscript: Stowe, Smith, Thurland. ish population, Hedermann et al2 reported a decrease in pre- Critical revision of the manuscript for important intellectual content: Stowe, term birth during the pandemic among uninfected patients. Ramsay, Andrews, Ladhani. 3 Statistical analysis: Stowe, Smith, Thurland, Andrews. Given differences in preterm birth across populations, we Administrative, technical, or material support: Stowe. examined a diverse urban cohort in the US to determine if Supervision: Ramsay. preterm birth, spontaneous preterm birth, medically indi- Conflict of Interest Disclosures: None reported. cated preterm birth, and stillbirth rates have changed during Funding/Support: This work was funded by Public Health England. the SARS-CoV-2 pandemic. Role of the Funder/Sponsor: Public Health England participated in the design and conduct of the study; collection, management, analysis, and interpretation Methods | GeoBirth is a curated pregnancy cohort of all births of the data; preparation, review, or approval of the manuscript; and decision to in 2 Penn Medicine hospitals in Philadelphia ongoing since submit the manuscript for publication. 2008 (approximately 9000 births per year), in which each 1. Kourtis AP, Read JS, Jamieson DJ. Pregnancy and infection. N Engl J Med. 2014;370(23):2211-2218. doi:10.1056/NEJMra1213566 preterm birth (<37 weeks’ gestation) is manually classified by 2. Pregnancy and coronavirus, 2020. NHS England. Published July 22, 2020. 2 independent blinded reviewers, with further adjudication Accessed September 4, 2020. https://www.nhs.uk/conditions/coronavirus- by a third reviewer when there is nonconcordance. Preterm covid-19/people-at-higher-risk/pregnancy-and-coronavirus/ birth phenotypes are categorized as spontaneous preterm 3. Knight M, Bunch K, Vousden N, et al; UK Obstetric Surveillance System birth (eg, preterm labor, spontaneous rupture of membranes) SARS-CoV-2 Infection in Pregnancy Collaborative Group. Characteristics and or medically indicated preterm birth (eg, clinician initiated outcomes of pregnant women admitted to hospital with confirmed SARS-CoV-2 infection in UK: national population based cohort study. BMJ. 2020;369:m2107. due to a maternal or fetal health condition, such as pre- doi:10.1136/bmj.m2107 eclampsia or intrauterine growth restriction). Stillbirth is

jama.com (Reprinted) JAMA January 5, 2021 Volume 325, Number 1 87

Downloaded From: https://jamanetwork.com/ on 09/26/2021 Letters

Table. Birth Outcomes by Race/Ethnicity Before (March-June 2018 and 2019) and During (March-June 2020) the Severe Acute Respiratory Syndrome Coronavirus 2 Pandemic in 2 Philadelphia Hospitalsa

No. (%) Adjusted absolute Prepandemic epoch Pandemic epoch Unadjusted risk difference a Race/ethnicity was based on patient Birth outcome (n = 5907) (n = 3007) P valueb (95% CI), %c self-identification during patient Preterm birthd 617 (10.5) 283 (9.5) .12 −1.1 (−2.4 to 0.2) registration. Non-Hispanic Black 323 (13.1) 157 (12.4) .57 −0.7 (−3.0 to 1.5) b Calculated using a 2-tailed Fisher exact test. Non-Hispanic White 177 (7.9) 73 (6.8) .26 −1.0 (−2.8 to 0.9) c Adjusted for month of birth, age, Other race/ethnicity 117 (9.9) 53 (8.2) .24 −1.7 (−4.4 to 1.0) parity, body mass index, Spontaneous preterm 315 (5.7) 135 (4.7) .09 −0.8 (−1.8 to 0.2) race/ethnicity (except in stratified e birth models), marital status, smoking, Non-Hispanic Black 150 (6.6) 77 (6.5) .99 0.1 (−1.6 to 1.9) and insurance status, calculated Non-Hispanic White 96 (4.5) 30 (2.9) .04 −1.4 (−2.8 to −0.1) using marginal effects models. Presented as percentages for all Other race/ethnicity 69 (6.1) 28 (4.5) .16 −1.6 (−3.7 to 0.6) birth outcomes except for stillbirth. Medically indicated 302 (5.4) 148 (5.2) .65 −0.3 (−1.4 to 0.6) d preterm birthf Preterm birth calculations exclude stillbirths. Non-Hispanic Black 173 (7.5) 80 (6.7) .45 −1.0 (−2.7 to 0.8) e Spontaneous preterm birth Non-Hispanic White 81 (3.8) 43 (4.1) .70 0.4 (−1.1 to 1.9) calculations exclude medically Other race/ethnicity 48 (4.3) 25 (4.0) .80 −0.3 (−2.3 to 1.7) indicated preterm births and Stillbirth (per 1000 32 (0.54) 15 (0.50) .88 −0.03 (−0.34 to 0.29) stillbirths. births) f Medically indicated preterm birth Non-Hispanic Black 25 (1.01) 9 (0.71) .47 −0.29 (−0.90 to 0.31) calculations exclude spontaneous Non-Hispanic Whiteg 4 (0.18) 2 (0.19) .99 preterm births and stillbirths. g Adjusted models did not converge Other race/ethnicityg 3 (0.25) 4 (0.61) .26 due to small numbers of events.

defined as intrauterine fetal demise at 20 weeks’ gestation or racial/ethnic groups had significant changes in outcomes. How- greater. We compared preterm birth, spontaneous preterm ever, no significant interaction was detected between race/ birth, medically indicated preterm birth, and stillbirth rates ethnicity and epoch with spontaneous preterm birth (P =.09 among singleton pregnancies during the pandemic period for interaction). (March-June 2020) with the same months in 2018 and 2019 In the 2 hospitals, universal SARS-CoV-2 testing began on (prepandemic) to account for seasonality using a 2-tailed April 1, 2020, and April 13, 2020. Among 86 patients with test Fisher exact test with a significance threshold of P <.05 results positive for SARS-CoV-2, the preterm birth rate was using R, version 4.0.2. We used marginal effects models to 11.6% (n = 10; 6 spontaneous and 4 medically indicated pre- calculate absolute risk differences between the 2 epochs term births) and there was 1 stillbirth. adjusting for birth month, age, parity, body mass index, race/ ethnicity, marital status, smoking, and insurance status. We Discussion | This study did not detect significant changes in pre- also performed analyses stratified by race/ethnicity because term or stillbirth rates during the SARS-CoV-2 pandemic in a of persistent preterm birth disparities.3 This study was racially diverse urban cohort from 2 Philadelphia hospitals. Al- approved by the University of Pennsylvania Institutional though these data allow for disaggregation of spontaneous and Review Board with a waiver of informed consent. medically indicated preterm births, no differences in overall rates of these phenotypes were detected. Results | There were a total of 8867 singleton, live-born deliv- These findings differ from a Danish report of decreasing pre- eries in March through June of 2018, 2019, and 2020 (42% non- term birth rates2 and higher stillbirth rates in a UK hospital4 dur- Hispanic Black, 37% non-Hispanic White, and 21% other race/ ing the pandemic. The differences between studies may be due ethnicity); 2992 deliveries occurred during the pandemic to differences in enforcement of lockdown orders, population period, including 283 preterm births (135 spontaneous and 148 heterogeneity, access to health care, or societal stressors. medically indicated) and 15 stillbirths. Prepandemic and pan- Study limitations include examination of a single health demic birth outcomes were as follows: 10.5% vs 9.5% of de- system, short epochs, limited representation of other races/ liveries were preterm births (adjusted difference, −1.1% [95% ethnicities, few stillbirths, and potential for change in deliv- CI, −2.4% to 0.2%]), 5.7% vs 4.7% were spontaneous preterm ery hospital choice during the pandemic. births (adjusted difference, −0.8% [95% CI, −1.8% to 0.2%]), 5.4% vs 5.2% were medically indicated preterm births (ad- Sara C. Handley, MD, MSCE justed difference, −0.3% [95% CI, −1.4% to 0.6%]), and 5.4 per Anne M. Mullin, BS 1000 births vs 5.0 per 1000 births were stillbirths (adjusted Michal A. Elovitz, MD difference, −0.03 per 1000 births [95% CI, −0.34 to 0.29]) Kristin D. Gerson, MD, PhD (Table). Spontaneous preterm birth among non-Hispanic White Diana Montoya-Williams, MD patients declined during the pandemic (4.5% vs 2.9%; ad- Scott A. Lorch, MD, MSCE justed difference, −1.4% [95% CI, −2.8% to −0.1%]); no other Heather H. Burris, MD, MPH

88 JAMA January 5, 2021 Volume 325, Number 1 (Reprinted) jama.com

Downloaded From: https://jamanetwork.com/ on 09/26/2021 Letters

Author Affiliations: Division of Neonatology, The Children’s Hospital of cline of more than 20% were 52% and 56%, respectively. A post- Philadelphia, Philadelphia, Pennsylvania (Handley, Montoya-Williams, Lorch, exercise ABI less than 0.90 exhibited the best sensitivity (70%) Burris); Maternal and Child Health Research Center, University of Pennsylvania 4 Perelman School of Medicine, Philadelphia (Mullin, Elovitz, Gerson). and specificity (38%). In 2019, the best postexercise criteria to detect an arterial stenosis greater than 50% using com- Corresponding Author: Heather H. Burris, MD, MPH, Biomedical Research 5 Bldg II/III, Room 1352, 421 Curie Blvd, Philadelphia, PA 19104-6160 puted tomographic angiography as a gold standard was a post- ([email protected]). exercise ABI decrease of 18.5% or greater, with sensitivity and Accepted for Publication: October 6, 2020. specificity of 71% and 64%, respectively, whereas the sensi- Published Online: December 7, 2020. doi:10.1001/jama.2020.20991 tivity and specificity of a postexercise ABI less than 0.90 were Author Contributions: Ms Mullin and Dr Burris had full access to all of the data 71% and 62%.5 In our study, the cutoff value for postexercise in the study and take responsibility for the integrity of the data and the accuracy ankle pressure was a decrease of 20 mm Hg or greater, but the of the data analysis. Dr Handley and Ms Mullin contributed equally. sensitivity (52%) and specificity (72%) remained poor.5 There- Concept and design: Handley, Mullin, Elovitz, Montoya-Williams, Lorch, Burris. Acquisition, analysis, or interpretation of data: All authors. fore, based on the literature since 2012, use of a postexercise Drafting of the manuscript: Handley, Mullin, Burris. ankle pressure decrease of more than 30 mm Hg to diagnose Critical revision of the manuscript for important intellectual content: Mullin, PAD should not be proposed. Elovitz, Gerson, Montoya-Williams, Lorch, Burris. Statistical analysis: Handley, Mullin, Burris. Exercise transcutaneous oxygen pressure (TcPO2) mea- Administrative, technical, or material support: Elovitz, Gerson. surement could be useful to diagnose PAD when ankle pres- Supervision: Elovitz, Lorch, Burris. sures do not suggest PAD, even if this test is more compli- 5 Conflict of Interest Disclosures: Dr Lorch reported receiving grants from the cated to perform routinely. Exercise TcPO2 had a sensitivity National Institutes of Health outside the submitted work. Dr Burris reported of 48% and a specificity of 85%. receiving grants from March of Dimes during the conduct of the study. No other disclosures were reported. Funding/Support: This work was supported by the Department of Pediatrics of Guillaume Mahé, MD, PhD Children’s Hospital of Philadelphia. Damien Lanéelle, MD Role of the Funder/Sponsor: The funder had no role in the design and conduct Alexis Le Faucheur, PhD of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the Author Affiliations: Vascular Medicine, CHU Rennes, Rennes, France (Mahé); manuscript for publication. Vascular Medicine Unit, CHU Caen Normandie, Caen, France (Lanéelle); 1. Khoury R, Bernstein PS, Debolt C, et al. Characteristics and outcomes of 241 University of Rennes 2, Rennes, France (Le Faucheur). births to women with severe acute respiratory syndrome coronavirus 2 Corresponding Author: Guillaume Mahé, MD, PhD, CHU Pontchaillou, 2 Rue (SARS-CoV-2) infection at five New York City medical centers. Obstet Gynecol. Henri le guilloux, 35033 Rennes, France ([email protected]). 2020;136(2):273-282. doi:10.1097/AOG.0000000000004025 Conflict of Interest Disclosures: None reported. 2. Hedermann G, Hedley PL, Bækvad-Hansen M, et al. Danish premature birth 1. Mehta A, Sperling LS, Wells BJ. Postexercise ankle-brachial index testing. JAMA. rates during the COVID-19 lockdown. Arch Dis Child Fetal Neonatal Ed. Published 2020;324(8):796-797. doi:10.1001/jama.2020.10164 online August 11, 2020. doi:10.1136/archdischild-2020-319990 2. Aboyans V, Criqui MH, Abraham P, et al; American Heart Association Council 3. Martin JA, Hamilton BE, Osterman MJK, Driscoll AK. Births: final data for on Peripheral Vascular Disease, Council on Epidemiology and Prevention, 2018. Natl Vital Stat Rep. 2019;68(13):1-47. Council on Clinical Cardiology, Council on Cardiovascular Nursing, Council on 4. Khalil A, von Dadelszen P, Draycott T, Ugwumadu A, O’Brien P, Magee L. Cardiovascular Radiology and Intervention, and Council on Cardiovascular Change in the incidence of stillbirth and preterm delivery during the COVID-19 Surgery and Anesthesia. Measurement and interpretation of the ankle-brachial pandemic. JAMA. 2020;324(7):705-706. doi:10.1001/jama.2020.12746 index: a scientific statement from the American Heart Association. Circulation. 2012;126(24):2890-2909. doi:10.1161/CIR.0b013e318276fbcb COMMENT & RESPONSE 3. Mahe G, Pollak AW, Liedl DA, et al. Discordant diagnosis of lower extremity peripheral artery disease using American Heart Association postexercise guidelines. Medicine (Baltimore). 2015;94(31):e1277. doi:10.1097/MD. Postexercise Ankle-Brachial Index Testing 0000000000001277 to Diagnose Peripheral Artery Disease 4. Aday AW, Kinlay S, Gerhard-Herman MD. Comparison of different exercise To the Editor Dr Mehta and colleagues presented the advan- ankle pressure indices in the diagnosis of peripheral artery disease. Vasc Med. tages and drawbacks of postexercise ankle-brachial index (ABI) 2018;23(6):541-548. doi:10.1177/1358863X18781723 testing based on a clinical case.1 However, several concerns 5. Stivalet O, Paisant A, Belabbas D, et al. Exercise testing criteria to diagnose about the postexercise criteria were not mentioned in the ar- lower extremity peripheral artery disease assessed by computed-tomography angiography. PLoS One. 2019;14(6):e0219082. doi:10.1371/journal.pone.0219082 ticle. Postexercise ABI criteria to diagnose peripheral artery disease (PAD) are debated. In 2012, Aboyans et al2 proposed using either a postexer- In Reply Dr Mahé and colleagues are concerned about the sen- cise decline of more than 20% or a postexercise ankle pres- sitivity and specificity of the thresholds used for diagnosing sure decrease of more than 30 mm Hg to establish the diag- PAD in patients with normal (1.00-1.40) or borderline (0.91- nosis of PAD. In 2015, among 7995 consecutive patients with 0.99) resting ABI and exertional non–joint-related leg pain.1,2 claudication and ABI greater than 0.90, 19% of patients pre- The diagnostic threshold of either a postexercise ABI de- sented with a mismatch between the criteria.3 In 2018, to decrease of more than 20% or a postexercise ankle systolic pres- tect arterial stenosis greater than 50% in patients with a rest- sure decrease of more than 30 mm Hg were suggested by ing ABI of 0.90 or greater, Aday et al4 found that the sensitivity Aboyans et al3 as a reasonable diagnostic criterion for PAD (rec- and specificity of a decrease in the postexercise ankle pres- ommendation class IIa; level of evidence A) in the 2012 Ameri- sure greater than 30 mm Hg were 3% and 94%, respectively, can Heart Association scientific statement on measurement and whereas the sensitivity and specificity of a postexercise de- interpretation of the ABI.

jama.com (Reprinted) JAMA January 5, 2021 Volume 325, Number 1 89

Downloaded From: https://jamanetwork.com/ on 09/26/2021