Can Machine Learning Help Identify Patients at Risk for Recurrent Sexually Transmitted Infections?

BACKGROUND: A substantial fraction of sexually transmitted infections (STIs) occur in patients who have previously been treated for an STI. We assessed whether routine electronic health record (EHR) data can predict which patients presenting with an incident STI are at greatest risk for additional STIs in the next 1 to 2 years. METHODS: We used structured EHR data on patients 15 years or older who acquired an incident STI diagnosis in 2008 to 2015 in eastern Massachusetts. We applied machine learning algorithms to model risk of acquiring ≥1 or ≥2 additional STIs diagnoses within 365 or 730 days after the initial diagnosis using more than 180 different EHR variables. We performed sensitivity analysis incorporating state health department surveillance data to assess whether improving the accuracy of identifying STI cases improved algorithm performance. RESULTS: We identified 8723 incident episodes of laboratory-confirmed gonorrhea, chlamydia, or syphilis. Bayesian Additive Regression Trees, the best-performing algorithm of any single method, had a cross-validated area under the receiver operating curve of 0.75. Receiver operating curves for this algorithm showed a poor balance between sensitivity and positive predictive value (PPV). A predictive probability threshold with a sensitivity of 91.5% had a corresponding PPV of 3.9%. A higher threshold with a PPV of 29.5% had a sensitivity of 11.7%. Attempting to improve the classification of patients with and without repeat STIs diagnoses by incorporating health department surveillance data had minimal impact on cross-validated area under the receiver operating curve. CONCLUSIONS: Machine algorithms using structured EHR data did not differentiate well between patients with and without repeat STIs diagnosis. Alternative strategies, able to account for sociobehavioral characteristics, could be explored.

Characteristics of Cases With Repeated Sexually Transmitted Infections, Massachusetts, 2014-2016.

Background: Persons with prior sexually transmitted infections (STIs) are at high risk for reinfection. No recent studies have examined frequency with which persons are diagnosed and reported with multiple bacterial STIs over time. Methods: We conducted a retrospective, of confirmed syphilis, gonorrhea, and chlamydial infections reported to Massachusetts state surveillance system within a 2-year period, 28 July 2014-27 July 2016. Results: Among Massachusetts population aged 13-65 years (4847510), 49142 (1.0%) were reported with ≥1 STIs; 6999 (14.2% of those with ≥1 STI) had ≥2 STIs, accounting for 27.7% of STIs. Of cases with ≥5 or more STIs (high-volume repeaters), 118 (74%) were men and 42 (26%) were women. Men spanned the age spectrum and were predominantly non-Hispanic white; 87% reported same-sex contacts. Women were younger, predominantly nonwhite, and without known same-sex contacts. Women were reinfected with gonorrhea and chlamydia or chlamydia alone; none had syphilis or human immunodeficiency virus (HIV) infection. All men with syphilis also had gonorrhea and/or chlamydia; 35% were diagnosed with HIV before, during, or within 10 months after study period. The majority (56%) of high-volume repeaters were seen at more than 1 care site/system. Conclusions: In Massachusetts, a large proportion of bacterial STIs are reported from a small subpopulation, many of whom have repeated infections and are likely to have higher impact on STI and HIV rates. Public health can play a crucial role in reaching high-volume repeaters whose STI histories may be hidden from clinicians due to fragmented care.

Does Nonmetropolitan Residence Impact Timely Chlamydia Treatment in Massachusetts?

A mean of 4.5 days until treatment was documented in a subset of reported laboratory-confirmed Massachusetts chlamydia cases selected for active case report form completion. Treatment delay was associated with longer test result turnaround time, and absence of symptoms or contact to sexually transmitted disease. Nonmetropolitan versus metropolitan residence did not appear to impact treatment time.

Investigating the debate of home birth safety: A critical review of cohort studies focusing on selected infant outcomes.

AIM: There is a debate within the medical community regarding the safety of planned home births. The presumption of increased risk of maternal and infant morbidity and mortality at home due to limited access to life-saving interventions is not clearly supported by research. The aim of the present study was to assess strengths and limitations of the methodological approaches of cohort studies that compare home births with hospital births by focusing on selected infant outcomes. METHODS: Studies were identified that assess the risk for at least one of three infant outcomes (mortality, Apgar score, and admission to the neonatal intensive care unit [NICU]) of home births compared with hospital births. RESULTS: Fifteen cohort studies were included. Two studies of low-risk births and two including higher risk births found home births to be at an increased risk of neonatal mortality. However, mortality is rare in developed nations and may not be the best measure of safety. When studies focused on low-risk pregnancies, planned birth location, and well-trained birth attendants, there was no difference in neonatal morbidity (Apgar score and NICU admission). CONCLUSION: Many methodological challenges were identified among these studies. This review contributes to the home birth published work by identifying key strengths and limitations that need to be accounted for in the interpretation of study findings and the development of future studies. Based on this review, the key variables that would strengthen future studies are birth attendant identification, documented planned birth location, and specification of the birth risk level. Uniformity of data collection and minimizing missing data are also critical.