The following represents disclosure information provided by authors of this manuscript. All relationships are considered compensated. Relationships are self-held unless noted. I 5 Immediate Family Member, Inst 5 My Institution. Relationships may not relate to the subject matter of this manuscript.

228 Background: Use of anti-cancer therapies in the last 14-30 days of life may worsen patient outcomes and increase cost; accordingly, rate of chemotherapy use near EOL is an important quality measure. Contemporary benchmarks are needed, with transparent methods describing the cohort in which the measure is assessed, and criteria for calculation. Methods: Data on chemotherapy use, mortality, and cancer diagnosis was sourced from electronic health records (EHRs) of >8,000 patients seen between 2014-2016 at two large US academic centers, for whom dates of death were available. Death dates were sourced from the EHR and public records (e.g., obituaries). Patients were grouped by diagnosis using ICD-10 codes. Rates of infusional chemotherapy receipt within 14 or 30 days of death were calculated. Results: Across 10 tumor types, 3-7% of patients received chemotherapy within 14d of death, and 6-16% received it within 30d. Rates were stable from 2014-2016 and did not differ by cancer center. Rates were highest in diseases where patients may experience rapid clinical decline near EOL: in pancreatic and rectal cancer, 30d rates were 16% and 13%. The 30d rate was lowest (6%) in kidney cancer. When the cohort was restricted to only treated patients (who received >=5 chemotherapy administrations at the center), rates of chemotherapy use at EOL increased to 6-12% (14d) and 17-28% (30d). Conclusions: This study provides baseline estimates of current rates of EOL chemotherapy use at academic centers. Transparency in methodology is critical; for example, when the whole population of cancer patients seen at a center is considered, rates are low, but when the analysis is limited to patients who received chemotherapy there, rates nearly double. Further studies should focus on whether this quality measure is a meaningful driver of patient and health system outcomes. This work also demonstrates that it is possible to assess this metric across multiple centers; this approach could be easily scaled to all oncology practices integrated in a data sharing network. [Table: see text]

BackgroundAs artificial intelligence (AI) continues to advance with breakthroughs in natural language processing (NLP) and machine learning (ML), such as the development of models like OpenAI’s ChatGPT, new opportunities are emerging for efficient curation of electronic health records (EHR) into real-world data (RWD) for evidence generation in oncology. Our objective is to describe the research and development of industry methods to promote transparency and explainability.MethodsWe applied NLP with ML techniques to train, validate, and test the extraction of information from unstructured documents (eg, clinician notes, radiology reports, lab reports, etc.) to output a set of structured variables required for RWD analysis. This research used a nationwide electronic health record (EHR)-derived database. Models were selected based on performance. Variables curated with an approach using ML extraction are those where the value is determined solely based on an ML model (ie, not confirmed by abstraction), which identifies key information from visit notes and documents. These models do not predict future events or infer missing information.ResultsWe developed an approach using NLP and ML for extraction of clinically meaningful information from unstructured EHR documents and found high performance of output variables compared with variables curated by manually abstracted data. These extraction methods resulted in research-ready variables including initial cancer diagnosis with date, advanced/metastatic diagnosis with date, disease stage, histology, smoking status, surgery status with date, biomarker test results with dates, and oral treatments with dates.ConclusionsNLP and ML enable the extraction of retrospective clinical data in EHR with speed and scalability to help researchers learn from the experience of every person with cancer.

ObjectiveTo identify the prevalence and predictors of malnutrition among 2-year old children in the Western Highlands of Guatemala.MethodsProspective cohort of 852 Guatemalan children in San Lucas Toliman, Guatemala followed from birth to age 2 from May 2008 to December 2013. Socio-demographic, anthropometric, and health data of children was collected at 2 month intervals.ResultsAmong the 402 males and 450 females in the cohort, mean weight-for-age Z-score (WAZ) declined from -0.67 ± 1.01 at 1 year to -1.07 ± 0.87 at 2 years, while mean height-for-age Z-score (HAZ) declined from -1.88 ± 1.19 at 1 year to -2.37 ± 0.99 at 2 years. Using multiple linear regression modeling, number of children <5 years old, vomiting in the past week, fever in the past week, and WAZ at 1 year were significant predictors of WAZ at 2 years. Significant predictors of HAZ at 2 years included household size, number of children <5 years old, diarrhea in the past week, WAZ at 1 year, and HAZ at 1 year. Vomiting in the past week and WAZ at 1 year were significant predictors of weight-for-height z-score (WHZ) at 2 years.ConclusionsNumber of children <5 years old, symptoms such as vomiting or diarrhea in the previous week, and prior nutritional status were the most significant predictors of malnutrition in this cohort. Future research may focus on the application of models to develop predictive algorithms for mobile device technology, as well as the identification of other predictors of malnutrition that are not well characterized such as the interaction of environmental exposures with protein consumption and epigenetics.

No abstract