RESUMEN
Social Determinants of Health (SDoH) are an important part of the exposome and are known to have a large impact on variation in health outcomes. In particular, housing stability is known to be intricately linked to a patient's health status, and pregnant women experiencing housing instability (HI) are known to have worse health outcomes. Most SDoH information is stored in electronic health records (EHRs) as free text (unstructured) clinical notes, which traditionally required natural language processing (NLP) for automatic identification of relevant text or keywords. A patient's housing status can be ambiguous or subjective, and can change from note to note or within the same note, making it difficult to use existing NLP solutions. New developments in NLP allow researchers to prompt LLMs to perform complex, subjective annotation tasks that require reasoning that previously could only be attempted by human annotators. For example, large language models (LLMs) such as GPT (Generative Pre-trained Transformer) enable researchers to analyze complex, unstructured data using simple prompts. We used a secure platform within a large healthcare system to compare the ability of GPT-3.5 and GPT-4 to identify instances of both current and past housing instability, as well as general housing status, from 25,217 notes from 795 pregnant women. Results from these LLMs were compared with results from manual annotation, a named entity recognition (NER) model, and regular expressions (RegEx). We developed a chain-of-thought prompt requiring evidence and justification for each note from the LLMs, to help maximize the chances of finding relevant text related to HI while minimizing hallucinations and false positives. Compared with GPT-3.5 and the NER model, GPT-4 had the highest performance and had a much higher recall (0.924) than human annotators (0.702) in identifying patients experiencing current or past housing instability, although precision was lower (0.850) compared with human annotators (0.971). In most cases, the evidence output by GPT-4 was similar or identical to that of human annotators, and there was no evidence of hallucinations in any of the outputs from GPT-4. Most cases where the annotators and GPT-4 differed were ambiguous or subjective, such as "living in an apartment with too many people". We also looked at GPT-4 performance on de-identified versions of the same notes and found that precision improved slightly (0.936 original, 0.939 de-identified), while recall dropped (0.781 original, 0.704 de-identified). This work demonstrates that, while manual annotation is likely to yield slightly more accurate results overall, LLMs, when compared with manual annotation, provide a scalable, cost-effective solution with the advantage of greater recall. At the same time, further evaluation is needed to address the risk of missed cases and bias in the initial selection of housing-related notes. Additionally, while it was possible to reduce confabulation, signs of unusual justifications remained. Given these factors, together with changes in both LLMs and charting over time, this approach is not yet appropriate for use as a fully-automated process. However, these results demonstrate the potential for using LLMs for computer-assisted annotation with human review, reducing cost and increasing recall. More efficient methods for obtaining structured SDoH data can help accelerate inclusion of exposome variables in biomedical research, and support healthcare systems in identifying patients who could benefit from proactive outreach.
RESUMEN
Background: Immune-mediated inflammatory diseases (IMIDs) are likely to complicate maternal health. However, literature on patients with IMIDs undergoing pregnancy is scarce and often overlooks the presence of comorbidities. We aimed to evaluate the impact of IMIDs on adverse pregnancy outcomes after assessing and addressing any discrepancies in the distribution of covariates associated with adverse pregnancy outcomes between patients with and without IMIDs. Methods: We conducted a retrospective cohort study using data from an integrated U.S. community healthcare system that provides care across Alaska, California, Montana, Oregon, New Mexico, Texas, and Washington. We used a database containing all structured data from electronic health record (EHRs) and analyzed the cohort of pregnant people who had live births from January 1, 2013, through December 31, 2022. We investigated 12 selected IMIDs: psoriasis, inflammatory bowel disease, rheumatoid arthritis, spondyloarthritis, multiple sclerosis, systemic lupus erythematosus, psoriatic arthritis, antiphospholipid syndrome, Sjögren's syndrome, vasculitides, sarcoidosis, and systemic sclerosis. We characterized patients with IMIDs prior to pregnancy (IMIDs group) based on pregnancy/maternal characteristics, comorbidities, and pre-pregnancy/prenatal immunomodulatory medications (IMMs) prescription patterns. We 1:1 propensity score matched the IMIDs cohort with people who had no IMID diagnoses prior to pregnancy (non-IMIDs cohort). Outcome measures were preterm birth (PTB), low birth weight (LBW), small for gestational age (SGA), and caesarean section. Findings: Our analytic cohort had 365,075 people, of which 5784 were in the IMIDs group and 359,291 were in the non-IMIDs group. The prevalence rate of pregnancy of at least 20 weeks duration in people with a previous IMID diagnosis has doubled in the past ten years. 17% of the IMIDs group had at least one prenatal IMM prescription. Depending on the type of IMM, 48%-70% of the patients taking IMMs before pregnancy continued them throughout pregnancy. Overall, patients with one or more of these 12 IMIDs had increased risk of PTB (Relative risk (RR) = 1.1 [1.0, 1.3]; p = 0.08), LBW (RR = 1.2 [1.0, 1.4]; p = 0.02), SGA (RR = 1.1 [1.0, 1.2]; p = 0.03), and caesarean section (RR = 1.1 [1.1, 1.2], p < 0.0001) compared to a matched cohort of people without IMIDs. When adjusted for comorbidities, patients with rheumatoid arthritis (PTB RR = 1.2, p = 0.5; LBW RR = 1.1, p = 0.6) and/or inflammatory bowel disease (PTB RR = 1.2, p = 0.3; LBW RR = 1.0, p = 0.8) did not have significantly increased risk for PTB and LBW. Interpretation: For patients who have been pregnant for 20 weeks or greater, the association between IMIDs and adverse pregnancy outcomes depends on both the nature of the IMID and the presence of comorbidities. Because this study was limited to pregnancies resulting in live births, results must be interpreted together with other studies on early pregnancy loss and stillbirth in patient with IMIDs. Funding: National Institutes of Health.
RESUMEN
OBJECTIVE: There is uncertainty around the safety of SSRIs for treating depression during pregnancy. Nevertheless, the use of SSRIs has been gradually increasing, especially during the COVID-19 pandemic period. We aimed to (1) characterize maternal depression rate and use of SSRIs in a recent 10-year period, (2) address confounding by indication, as well as socioeconomic and environmental factors, and (3) evaluate associations of the timing of SSRI exposure in pregnancy with risk for preterm birth (PTB), low birthweight (LBW), and small for gestational age (SGA) infants among women with depression before pregnancy. METHODS: We conducted propensity score-adjusted regression to calculate odds ratios (ORs) of PTB, LBW, and SGA. We accounted for maternal/pregnancy characteristics, comorbidity, depression severity, time of delivery, social vulnerability, and rural residence. RESULTS: There were 50.3% and 40.3% increases in the prevalence rate of prenatal depression and prenatal SSRI prescription rate during the pandemic. We identified women with depression ≤180 days before pregnancy (n = 8406). Women with no SSRI order during pregnancy (n = 3760) constituted the unexposed group. The late SSRI exposure group consisted of women with an SSRI order after the first trimester (n = 3759). The early-only SSRI exposure group consisted of women with SSRI orders only in the first trimester (n = 887). The late SSRI exposure group had an increased risk of PTB of OR = 1.5 ([1.2,1.8]) and LBW of OR = 1.5 ([1.2,2.0]), relative to the unexposed group. Associations between late SSRI exposure and risk of PTB/LBW were similar among a subsample of patients who delivered during the pandemic. CONCLUSIONS: These findings suggest an association between PTB/LBW and SSRI exposure is dependent on exposure timing during pregnancy. Small for gestational age is not associated with SSRI exposure.
