Exploratory analysis of machine learning approaches for surveillance of Zika-associated birth defects.

In 2016, Centers for Disease Control and Prevention (CDC) established surveillance of pregnant women with Zika virus infection and their infants in the U.S. states, territories, and freely associated states. To identify cases of Zika-associated birth defects, subject matter experts review data reported from medical records of completed pregnancies to identify findings that meet surveillance case criteria (manual review). The volume of reported data increased over the course of the Zika virus outbreak in the Americas, challenging the resources of the surveillance system to conduct manual review. Machine learning was explored as a possible method for predicting case status. Ensemble models (using machine learning algorithms including support vector machines, logistic regression, random forests, k-nearest neighbors, gradient boosted trees, and decision trees) were developed and trained using data collected from January 2016-October 2017. Models were developed separately, on data from the U.S. states, non-Puerto Rico territories, and freely associated states (referred to as the U.S. Zika Pregnancy and Infant Registry [USZPIR]) and data from Puerto Rico (referred to as the Zika Active Pregnancy Surveillance System [ZAPSS]) due to differences in data collection and storage methods. The machine learning models demonstrated high sensitivity for identifying cases while potentially reducing volume of data for manual review (USZPIR: 96% sensitivity, 25% reduction in review volume; ZAPSS: 97% sensitivity, 50% reduction in review volume). Machine learning models show potential for identifying cases of Zika-associated birth defects and for reducing volume of data for manual review, a potential benefit in other public health emergency response settings.

Vital Signs: Zika-Associated Birth Defects and Neurodevelopmental Abnormalities Possibly Associated with Congenital Zika Virus Infection - U.S. Territories and Freely Associated States, 2018.

INTRODUCTION: Zika virus infection during pregnancy causes serious birth defects and might be associated with neurodevelopmental abnormalities in children. Early identification of and intervention for neurodevelopmental problems can improve cognitive, social, and behavioral functioning. METHODS: Pregnancies with laboratory evidence of confirmed or possible Zika virus infection and infants resulting from these pregnancies are included in the U.S. Zika Pregnancy and Infant Registry (USZPIR) and followed through active surveillance methods. This report includes data on children aged ≥1 year born in U.S. territories and freely associated states. Receipt of reported follow-up care was assessed, and data were reviewed to identify Zika-associated birth defects and neurodevelopmental abnormalities possibly associated with congenital Zika virus infection. RESULTS: Among 1,450 children of mothers with laboratory evidence of confirmed or possible Zika virus infection during pregnancy and with reported follow-up care, 76% had developmental screening or evaluation, 60% had postnatal neuroimaging, 48% had automated auditory brainstem response-based hearing screen or evaluation, and 36% had an ophthalmologic evaluation. Among evaluated children, 6% had at least one Zika-associated birth defect identified, 9% had at least one neurodevelopmental abnormality possibly associated with congenital Zika virus infection identified, and 1% had both. CONCLUSION: One in seven evaluated children had a Zika-associated birth defect, a neurodevelopmental abnormality possibly associated with congenital Zika virus infection, or both reported to the USZPIR. Given that most children did not have evidence of all recommended evaluations, additional anomalies might not have been identified. Careful monitoring and evaluation of children born to mothers with evidence of Zika virus infection during pregnancy is essential for ensuring early detection of possible disabilities and early referral to intervention services.

Pregnancy Outcomes After Maternal Zika Virus Infection During Pregnancy - U.S. Territories, January 1, 2016-April 25, 2017.

Pregnant women living in or traveling to areas with local mosquito-borne Zika virus transmission are at risk for Zika virus infection, which can lead to severe fetal and infant brain abnormalities and microcephaly (1). In February 2016, CDC recommended 1) routine testing for Zika virus infection of asymptomatic pregnant women living in areas with ongoing local Zika virus transmission at the first prenatal care visit, 2) retesting during the second trimester for women who initially test negative, and 3) testing of pregnant women with signs or symptoms consistent with Zika virus disease (e.g., fever, rash, arthralgia, or conjunctivitis) at any time during pregnancy (2). To collect information about pregnant women with laboratory evidence of recent possible Zika virus infection* and outcomes in their fetuses and infants, CDC established pregnancy and infant registries (3). During January 1, 2016-April 25, 2017, U.S. territories† with local transmission of Zika virus reported 2,549 completed pregnancies§ (live births and pregnancy losses at any gestational age) with laboratory evidence of recent possible Zika virus infection; 5% of fetuses or infants resulting from these pregnancies had birth defects potentially associated with Zika virus infection¶ (4,5). Among completed pregnancies with positive nucleic acid tests confirming Zika infection identified in the first, second, and third trimesters, the percentage of fetuses or infants with possible Zika-associated birth defects was 8%, 5%, and 4%, respectively. Among liveborn infants, 59% had Zika laboratory testing results reported to the pregnancy and infant registries. Identification and follow-up of infants born to women with laboratory evidence of recent possible Zika virus infection during pregnancy permits timely and appropriate clinical intervention services (6).

Evaluation of Placental and Fetal Tissue Specimens for Zika Virus Infection - 50 States and District of Columbia, January-December, 2016.

Zika virus infection during pregnancy can cause congenital microcephaly and brain abnormalities (1), and detection of Zika virus RNA in clinical and tissue specimens can provide definitive laboratory evidence of recent Zika virus infection. Whereas duration of viremia is typically short, prolonged detection of Zika virus RNA in placental, fetal, and neonatal brain tissue has been reported and can provide key diagnostic information by confirming recent Zika virus infection (2). In accordance with recent guidance (3,4), CDC provides Zika virus testing of placental and fetal tissues in clinical situations where this information could add diagnostic value. This report describes the evaluation of formalin-fixed paraffin-embedded (FFPE) tissue specimens tested for Zika virus infection in 2016 and the contribution of this testing to the public health response. Among 546 live births with possible maternal Zika virus exposure, for which placental tissues were submitted by the 50 states and District of Columbia (DC), 60 (11%) were positive by Zika virus reverse transcription-polymerase chain reaction (RT-PCR). Among 81 pregnancy losses for which placental and/or fetal tissues were submitted, 18 (22%) were positive by Zika virus RT-PCR. Zika virus RT-PCR was positive on placental tissues from 38/363 (10%) live births with maternal serologic evidence of recent unspecified flavivirus infection and from 9/86 (10%) with negative maternal Zika virus immunoglobulin M (IgM) where possible maternal exposure occurred >12 weeks before serum collection. These results demonstrate that Zika virus RT-PCR testing of tissue specimens can provide a confirmed diagnosis of recent maternal Zika virus infection.

Voices from the street: exploring the realities of family homelessness.

Homelessness threatens the health and well-being of thousands of families in the United States, yet little is known about their specific needs and how current services address them. To fill this knowledge gap, we explored the experiences of homelessness families in Detroit, Michigan. We targeted homeless mothers and their caseworkers for study to see if the perceptions of needs and services were in alignment. Using focus groups and content analysis, we identified four overarching themes that illustrate homeless mothers' experience with homelessness. We then analyzed data from caseworkers to look specifically for similarities and differences in their perceptions. Key findings included reports of family histories of violence, poverty, social isolation, and a lack of informal support as contributing to homelessness. The differing perspectives of mothers and their caseworkers regarding how best to move forward highlight how current programs and services may not be meeting the needs of this growing and vulnerable cohort.