Karen Eilbeck

Newborn screening (NBS), a mandated public health intervention, allows the identification of babies with potentially life-threatening disorders and facilitates disease diagnosis and management before the onset of symptoms. While NBS saves lives, the process can be fraught with anxiety and unanswered questions from parents or guardians of newborns, especially as they wait for an appointment with a clinician. This study aimed to describe the development and testing of an educational chatbot (NBSchat) to address the emotional support and information needs of parents of newborns identified with sickle cell trait via NBS. NBSchat, a fully scripted (ie, rule-based) chatbot, was developed by a multidisciplinary team and evaluated through a sequential multiple methods study, including interviews and a survey. To inform chatbot design, we conducted semistructured interviews with 11 adults-5 clinicians who work with parents of infants identified with sickle cell trait through NBS and 6 parents of infants aged 12 months or less-using the critical incident technique and think-aloud tasks while using a prototype of NBSchat. Transcripts underwent thematic analysis. In a survey, 250 parents of infants aged 12 months or less without abnormal NBS results were shown a mock newborn screening result letter and then interacted with NBSchat, after which they self-reported emotional and attitudinal outcomes before and after the simulated exposure. Feedback from interviews confirmed that parents are distressed by trait results and actively seek information and reassurance. Thematic analysis indicated that NBSchat provided reliable, accurate information that parents wanted and had the potential to reduce negative emotions (eg, provide relief and reduce stress). Key strengths included addressing an immediate health concern and offering reassurance. The results of the postintervention survey indicated that, compared to pre-exposure scores, participants reported significantly lower negative emotions (mean 7.0 [SD 3.2] vs 5.8 [SD 3.2] out of 12; mean difference -1.2, 95% CI -1.57 to -0.83; P<.001), improved positive emotions (reflected by a decrease in the reverse-coded positive emotion score; mean 8.6 [SD 4.2] vs 7.8 [SD 4] out of 16; mean difference -0.8, 95% CI -1.27 to -0.37; P<.001), and reduced uncertainty (mean 6.5 [SD 3] vs 5.5 [SD 3.4] out of 12; mean difference -1, 95% CI -1.42 to -0.58; P<.001). Parents noted that NBSchat provided immediate reassurance and was convenient to access. They further reported that the predefined, structured questions in the script helped guide their learning and understanding. Overall, participants who interacted with NBSchat found it to be acceptable, with improved emotional measures after its use. Future research will investigate the outcomes of using the chatbot and its implementation in a pragmatic randomized controlled trial.

BACKGROUND Newborn screening (NBS), a mandated public health intervention, allows the identification of babies with potentially life-threatening disorders and facilitates disease diagnosis and management before the onset of symptoms. While NBS saves lives, the process can be fraught with anxiety and unanswered questions from parents or guardians of newborns, especially while they wait for an appointment with a clinician. OBJECTIVE To describe the development and testing of an educational chatbot (NBSchat) to address the emotional support and information needs of parents of newborns identified with sickle cell trait (SCT) via NBS. METHODS NBSChat, a fully scripted (i.e, rule-based) chatbot, was developed by a multidisciplinary team and evaluated through a sequential multiple methods study, including interviews and a survey. To inform chatbot design, we conducted semi-structured interviews with 11 adults; five clinicians who work with parents of infants identified with SCT through NBS and six parents of infants ≤12 months—using the critical incident technique and think-aloud tasks while using a prototype of NBSchat. Transcripts underwent thematic analysis. 250 parents of infants ≤12 months without abnormal NBS results completed the survey. Parents were asked to self-report emotional and attitudinal outcomes before and after interacting with NBSchat. RESULTS Feedback from interviews confirmed that parents are distressed by trait results and actively seek information and reassurance. Thematic analysis indicated that NBSchat provided reliable, accurate information that parents wanted and had the potential to reduce negative emotions (e.g., provide relief and reduce stress). Key strengths included addressing an immediate health concern and offering reassurance. Results of the post-intervention survey indicated that, compared to pre-exposure scores, participants reported significantly lower negative emotions (mean 7.0 vs 5.8 out of 12; t-test, P<.001), higher positive emotions (mean 8.6 vs 7.8 out of 16; t-test, P<.001), and reduced uncertainty (mean 6.5 vs 5.5 out of 12; t-test, P<.001). Parents noted that NBSchat provided immediate reassurance and was convenient to access. They further reported that the predefined, structured questions in the script helped guide their learning and understanding. CONCLUSIONS Overall, participants who interacted with NBSchat found it to be acceptable, with improved emotional measures after its use. Future research will investigate outcomes of using the chatbot and its implementation in a pragmatic randomized controlled trial.

Communication of newborn screening (NBS) results often fails to provide clear explanations of NBS screen results to parents. Understanding existing NBS workflows is vital for improving NBS follow-up. We sought to describe a diverse range of state NBS programs as a starting point for designing tools to improve NBS follow-up, using the example of hemoglobinopathy traits. At a workshop of the 2023 Association of Public Health Laboratories NBS Symposium, participants filled out a survey and modeled their state workflows. Salient features were extracted and synthesized by state. A subset of models was member checked. Representatives from 19 U.S. states participated in the workflow analysis. Mail was overwhelmingly relied upon to convey the results. NBS programs differed by point of first contact with parents and degree of reliance on third parties. A participatory approach is useful for the rapid preliminary documentation of existing NBS program diversity and opportunities and challenges to improve patient education and follow-up. Future work should broaden the analysis to additional entities or individuals, particularly parents and caregivers.

Endometriosis is a debilitating, chronic disease that is estimated to affect 11% of reproductive-age women. Diagnosis of endometriosis is difficult with diagnostic delays of up to 12 years reported. These delays can negatively impact health and quality of life. Vague, nonspecific symptoms, like pain, with multiple differential diagnoses contribute to the difficulty of diagnosis. By investigating previously imprecise symptoms of pain, we sought to clarify distinct pain symptoms indicative of endometriosis, using an artificial intelligence-based approach. We used data from 473 women undergoing laparoscopy or laparotomy for a variety of surgical indications. Multiple anatomical pain locations were clustered based on the associations across samples to increase the power in the probability calculations. A Bayesian network was developed using pain-related features, subfertility, and diagnoses. Univariable and multivariable analyses were performed by querying the network for the relative risk of a postoperative diagnosis, given the presence of different symptoms. Performance and sensitivity analyses demonstrated the advantages of Bayesian network analysis over traditional statistical techniques. Clustering grouped the 155 anatomical sites of pain into 15 pain locations. After pruning, the final Bayesian network included 18 nodes. The presence of any pain-related feature increased the relative risk of endometriosis (p-value < 0.001). The constellation of chronic pelvic pain, subfertility, and dyspareunia resulted in the greatest increase in the relative risk of endometriosis. The performance and sensitivity analyses demonstrated that the Bayesian network could identify and analyze more significant associations with endometriosis than traditional statistical techniques. Pelvic pain, frequently associated with endometriosis, is a common and vague symptom. Our Bayesian network for the study of pain-related features of endometriosis revealed specific pain locations and pain types that potentially forecast the diagnosis of endometriosis.

International student exchange is a valuable opportunity for Biomedical and Health Informatics students to gain new perspectives and experiences. In the past, such exchanges have been made possible through international partnerships between universities. Unfortunately, numerous obstacles such as housing, financial concerns, and environmental implications related to travel, have made it difficult to continue international exchange. Experiences with hybrid and online education during covid-19 paved the way for a new approach that allows for short international exchange with a hybrid online-offline supervision model. This will be initiated with an exploration project between two international universities , each related to their respective institute's research focus.

Background: With the widespread adoption of electronic healthcare records (EHRs) by US hospitals, there is an opportunity to leverage this data for the development of predictive algorithms to improve clinical care. A key barrier in model development and implementation includes the external validation of model discrimination, which is rare and often results in worse performance. One reason why machine learning models are not externally generalizable is data heterogeneity. A potential solution to address the substantial data heterogeneity between health care systems is to use standard vocabularies to map EHR data elements. The advantage of these vocabularies is a hierarchical relationship between elements, which allows the aggregation of specific clinical features to more general grouped concepts. Objective: This study aimed to evaluate grouping EHR data using standard vocabularies to improve the transferability of machine learning models for the detection of postoperative health care–associated infections across institutions with different EHR systems. Methods: Patients who underwent surgery from the University of Utah Health and Intermountain Healthcare from July 2014 to August 2017 with complete follow-up data were included. The primary outcome was a health care–associated infection within 30 days of the procedure. EHR data from 0-30 days after the operation were mapped to standard vocabularies and grouped using the hierarchical relationships of the vocabularies. Model performance was measured using the area under the receiver operating characteristic curve (AUC) and F1-score in internal and external validations. To evaluate model transferability, a difference-in-difference metric was defined as the difference in performance drop between internal and external validations for the baseline and grouped models. Results: A total of 5775 patients from the University of Utah and 15,434 patients from Intermountain Healthcare were included. The prevalence of selected outcomes was from 4.9% (761/15,434) to 5% (291/5775) for surgical site infections, from 0.8% (44/5775) to 1.1% (171/15,434) for pneumonia, from 2.6% (400/15,434) to 3% (175/5775) for sepsis, and from 0.8% (125/15,434) to 0.9% (50/5775) for urinary tract infections. In all outcomes, the grouping of data using standard vocabularies resulted in a reduced drop in AUC and F1-score in external validation compared to baseline features (all P<.001, except urinary tract infection AUC: P=.002). The difference-in-difference metrics ranged from 0.005 to 0.248 for AUC and from 0.075 to 0.216 for F1-score. Conclusions: We demonstrated that grouping machine learning model features based on standard vocabularies improved model transferability between data sets across 2 institutions. Improving model transferability using standard vocabularies has the potential to improve the generalization of clinical prediction models across the health care system.

This study aims to explore and elucidate parents’ experience of newborn screening [NBS], with the overarching goal of identifying desiderata for the development of informatics-based educational and health management resources. We conducted four focus groups and four one-on-one qualitative interviews with a total of 35 participants between March and September 2020. Participants were grouped into three types: parents who had received true positive newborn screening results; parents who had received false positive results; and soon-to-be parents who had no direct experience of the screening process. Interview data were subjected to analysis using an inductive, constant comparison approach. Results are divided into five sections: (1) experiences related to the process of receiving NBS results and prior knowledge of the NBS program; (2) approaches to the management of a child’s medical data; (3) sources of additional informational and emotional support; (4) barriers faced by parents navigating the health system; and (5) recommendations and suggestions for new parents experiencing the NBS process. Our analysis revealed a wide range of experiences of, and attitudes towards the newborn screening program and the wider newborn screening system. While parents’ view of the screening process was – on the whole – positive, some participants reported experiencing substantial frustration, particularly related to how results are initially communicated and difficulties in accessing reliable, timely information. This frustration with current information management and education resources indicates a role for informatics-based approaches in addressing parents’ information needs.

Understanding the conditionally-dependent clinical variables that drive cardiovascular health outcomes is a major challenge for precision medicine. Here, we deploy a recently developed massively scalable comorbidity discovery method called Poisson Binomial based Comorbidity discovery (PBC), to analyze Electronic Health Records (EHRs) from the University of Utah and Primary Children’s Hospital (over 1.6 million patients and 77 million visits) for comorbid diagnoses, procedures, and medications. Using explainable Artificial Intelligence (AI) methodologies, we then tease apart the intertwined, conditionally-dependent impacts of comorbid conditions and demography upon cardiovascular health, focusing on the key areas of heart transplant, sinoatrial node dysfunction and various forms of congenital heart disease. The resulting multimorbidity networks make possible wide-ranging explorations of the comorbid and demographic landscapes surrounding these cardiovascular outcomes, and can be distributed as web-based tools for further community-based outcomes research. The ability to transform enormous collections of EHRs into compact, portable tools devoid of Protected Health Information solves many of the legal, technological, and data-scientific challenges associated with large-scale EHR analyses.