RESUMEN
The tumor microenvironment (TME) plays a fundamental role in tumorigenesis, tumor progression, and anti-cancer immunity potential of emerging cancer therapeutics. Understanding inter-patient TME heterogeneity, however, remains a challenge to efficient drug development. This article applies recent advances in machine learning (ML) for survival analysis to a retrospective study of NSCLC patients who received definitive surgical resection and immune pathology following surgery. ML methods are compared for their effectiveness in identifying prognostic subtypes. Six survival models, including Cox regression and five survival machine learning methods, were calibrated and applied to predict survival for NSCLC patients based on PD-L1 expression, CD3 expression, and ten baseline patient characteristics. Prognostic subregions of the biomarker space are delineated for each method using synthetic patient data augmentation and compared between models for overall survival concordance. A total of 423 NSCLC patients (46% female; median age [inter quantile range]: 67 [60-73]) treated with definite surgical resection were included in the study. And 219 (52%) patients experienced events during the observation period consisting of a maximum follow-up of 10 years and median follow up 78 months. The random survival forest (RSF) achieved the highest predictive accuracy, with a C-index of 0.84. The resultant biomarker subtypes demonstrate that patients with high PD-L1 expression combined with low CD3 counts experience higher risk of death within five-years of surgical resection.
Asunto(s)
Biomarcadores de Tumor , Carcinoma de Pulmón de Células no Pequeñas , Neoplasias Pulmonares , Aprendizaje Automático , Microambiente Tumoral , Humanos , Carcinoma de Pulmón de Células no Pequeñas/patología , Carcinoma de Pulmón de Células no Pequeñas/mortalidad , Carcinoma de Pulmón de Células no Pequeñas/cirugía , Femenino , Masculino , Anciano , Persona de Mediana Edad , Neoplasias Pulmonares/patología , Neoplasias Pulmonares/mortalidad , Neoplasias Pulmonares/cirugía , Pronóstico , Estudios Retrospectivos , Biomarcadores de Tumor/metabolismo , Antígeno B7-H1/metabolismo , Análisis de SupervivenciaRESUMEN
BACKGROUND: Variable data quality poses a challenge to using electronic health record (EHR) data to ascertain acute clinical outcomes in multi-site clinical trials. Differing EHR platforms and data comprehensiveness across clinical trial sites, especially if patients received care outside of the clinical site's network, can also affect validity of results. Overcoming these challenges requires a structured approach. METHODS: We propose a framework and create a checklist to assess the readiness of clinical sites to contribute EHR data to a clinical trial for the purpose of outcome ascertainment, based on our experience with the Strategies to Reduce Injuries and Develop Confidence in Elders (STRIDE) study, which enrolled 5451 participants in 86 primary care practices across 10 healthcare systems (sites). RESULTS: The site readiness checklist includes assessment of the infrastructure (i.e., size and structure of the site's healthcare system or clinical network), data procurement (i.e., quality of the data), and cost of obtaining study data. The checklist emphasizes the importance of understanding how data are captured and integrated across a site's catchment area and having a protocol in place for data procurement to ensure consistent and uniform extraction across each site. CONCLUSIONS: We suggest rigorous, prospective vetting of the data quality and infrastructure of each clinical site before launching a multi-site trial dependent on EHR data. The proposed checklist serves as a guiding tool to help investigators ensure robust and unbiased data capture for their clinical trials. ORIGINAL TRIAL REGISTRATION NUMBER: NCT02475850.
Asunto(s)
Lista de Verificación , Registros Electrónicos de Salud , Humanos , Exactitud de los Datos , Atención Primaria de Salud/organización & administración , Ensayos Clínicos como Asunto/métodos , Ensayos Clínicos como Asunto/organización & administración , Ensayos Clínicos como Asunto/normas , AncianoRESUMEN
Enrolling patients to the standard of care (SOC) arm in randomized clinical trials, especially for rare diseases, can be very challenging due to the lack of resources, restricted patient population availability, and ethical considerations. As the therapeutic effect for the SOC is often well documented in historical trials, we propose a Bayesian platform trial design with hybrid control based on the multisource exchangeability modelling (MEM) framework to harness historical control data. The MEM approach provides a computationally efficient method to formally evaluate the exchangeability of study outcomes between different data sources and allows us to make better informed data borrowing decisions based on the exchangeability between historical and concurrent data. We conduct extensive simulation studies to evaluate the proposed hybrid design. We demonstrate the proposed design leads to significant sample size reduction for the internal control arm and borrows more information compared to competing Bayesian approaches when historical and internal data are compatible.
Asunto(s)
Teorema de Bayes , Simulación por Computador , Modelos Estadísticos , Ensayos Clínicos Controlados Aleatorios como Asunto , Humanos , Ensayos Clínicos Controlados Aleatorios como Asunto/métodos , Tamaño de la Muestra , Proyectos de InvestigaciónRESUMEN
BACKGROUND: Diagnosis-code-based algorithms to identify fall injuries in Medicare data are useful for ascertaining outcomes in interventional and observational studies. However, these algorithms have not been validated against a fully external reference standard, in ICD-10-CM, or in Medicare Advantage (MA) data. METHODS: We linked self-reported fall injuries leading to medical attention (FIMA) from the Strategies to Reduce Injuries and Develop Confidence in Elders (STRIDE) trial (reference standard) to Medicare fee-for-service (FFS) and MA data from 2015-19. We measured the area under the receiver operating characteristic curve (AUC) based on sensitivity and specificity of a diagnosis-code-based algorithm against the reference standard for presence or absence of ≥1 FIMA within a specified window of dates, varying the window size to obtain points on the curve. We stratified results by source (FFS vs MA), trial arm (intervention vs control), and STRIDE's 10 participating health care systems. RESULTS: Both reference standard data and Medicare data were available for 4 941 (of 5 451) participants. The reference standard and algorithm identified 2 054 and 2 067 FIMA, respectively. The algorithm had 45% sensitivity (95% confidence interval [CI]: 43%-47%) and 99% specificity (95% CI: 99%-99%) to identify reference standard FIMA within the same calendar month. The AUC was 0.79 (95% CI: 0.78-0.81) and was similar by FFS or MA data source and by trial arm but showed variation among STRIDE health care systems (AUC range by health care system, 0.71 to 0.84). CONCLUSIONS: An ICD-10-CM algorithm to identify fall injuries demonstrated acceptable performance against an external reference standard, in both MA and FFS data.
Asunto(s)
Accidentes por Caídas , Algoritmos , Clasificación Internacional de Enfermedades , Medicare , Humanos , Estados Unidos , Accidentes por Caídas/estadística & datos numéricos , Anciano , Masculino , Femenino , Anciano de 80 o más Años , Sensibilidad y Especificidad , Heridas y Lesiones/diagnósticoRESUMEN
Good patient outcomes after treatment of the glabellar complex with botulinum toxin type A entail elimination of glabellar lines and maintenance of a natural eyebrow position. A precise injection technique that accurately targets the muscles that influence eyebrow position is required to reduce the risk of adverse aesthetic outcomes or unmasking an underlying eyelid ptosis. Here, we describe the glabellar lines optimization (GLO 3â +â 2) injection anatomy technique, a precise five-point injection pattern that is based on current understanding of facial functional anatomy and which aims to minimize the risk of affecting nontargeted muscles. Injection sites above the brow or that do not target the precise location of the muscles in the glabellar complex are likely to inadvertently expose the frontalis to botulinum toxin type A and result in undesirable aesthetic outcomes. Because the frontalis is a strong determinant of aesthetic outcomes, it is important to consider the overall effects of the interactions between the eyebrow depressors and the opposing forces of the frontalis on brow outcomes in both the resting brow position and during dynamic brow movement.
RESUMEN
Oral administration of peptide therapeutics faces challenges because of the distinct environment of the gastrointestinal tract. An oral formulation of semaglutide, a glucagon-like peptide 1 receptor agonist, was approved by the U.S. Food and Drug Administration in 2019 as a peptide therapy for the treatment of type 2 diabetes. Oral semaglutide uses sodium N-(8-[2-hydroxybenzoyl] amino) caprylate (SNAC) technology to enhance the absorption of semaglutide in the stomach and protect it from degradation by gastric enzymes. This article presents a summary of studies investigating SNAC technology as an absorption enhancer for a number of molecules and, in particular, explores how SNAC, once coformulated with oral semaglutide, facilitates increased absorption and bioavailability. Practical advice and dispensing information for pharmacists is also provided.
RESUMEN
The ability to sustain attention consistency is frequently assessed using either objective behavioral measures, such as reaction time (RT) variability, or subjective self-report measures, such as rates of task-unrelated thought (TUT). The current studies examined whether the individual-difference covariation in these measures provides a more construct valid assessment of attention consistency than does either alone. We argue that performance and self-report measures mutually validate each other; each measurement approach has its own sources of error, so their shared variance should best reflect the attention consistency construct. We reanalyzed two latent-variable studies where RT variability and TUTs were measured in multiple tasks (Kane et al. in J Exp Psychol Gen 145:1017-1048, 2016; Unsworth et al. in J Exp Psychol Gen 150:1303-1331, 2021), along with several nomological network constructs to test the convergent and discriminant validity of a general attention consistency factor. Confirmatory factor analyses assessing bifactor (preregistered) and hierarchical (non-preregistered) models suggested that attention consistency can be modeled as the shared variance among objective and subjective measures. This attention consistency factor was related to working memory capacity, attention (interference) control, processing speed, state motivation and alertness, and self-reported cognitive failures and positive schizotypy. Although bifactor models of general attention consistency provide the most compelling construct validity evidence for a specific ability to sustain attention, multiverse analyses of outlier decisions suggested they are less robust than hierarchical models. The results provide evidence for the general ability to sustain attention consistency and suggestions for improving its measurement.
Asunto(s)
Atención , Pensamiento , Humanos , Autoinforme , Tiempo de Reacción , Memoria a Corto PlazoRESUMEN
The ability to sustain attention is often measured with either objective performance indicators, like within-person RT variability, or subjective self-reports, like mind wandering propensity. A more construct valid approach, however, may be to assess the covariation in these performance and self-report measures, given that each of these is influenced by different sources of measurement error. If the correlation between performance-variability and self-report measures reflects the sustained attention construct, then task manipulations aimed at reducing the sustained attention demands of tasks should reduce the correlation between them (in addition to reducing mean levels of variability and mind wandering). The current study investigated this claim with a combined experimental-correlation approach. In two experiments (Ns ~ 1,500 each), participants completed tasks that either maximized or minimized the demand for sustained attention. Our demand manipulations successfully reduced the mean levels of sustained attention failures in both the objective and subjective measures, in both experiments. In neither experiment, however, did the covariation between these measures change as a function of the sustained attention demands of the tasks. We can therefore claim only minimal support for the construct validity of our measurement approach to sustained attention.
Asunto(s)
Atención , Humanos , AutoinformeRESUMEN
This paper presents novel datasets providing numerical representations of ICD-10-CM codes by generating description embeddings using a large language model followed by a dimension reduction via autoencoder. The embeddings serve as informative input features for machine learning models by capturing relationships among categories and preserving inherent context information. The model generating the data was validated in two ways. First, the dimension reduction was validated using an autoencoder, and secondly, a supervised model was created to estimate the ICD-10-CM hierarchical categories. Results show that the dimension of the data can be reduced to as few as 10 dimensions while maintaining the ability to reproduce the original embeddings, with the fidelity decreasing as the reduced-dimension representation decreases. Multiple compression levels are provided, allowing users to choose as per their requirements, download and use without any other setup. The readily available datasets of ICD-10-CM codes are anticipated to be highly valuable for researchers in biomedical informatics, enabling more advanced analyses in the field. This approach has the potential to significantly improve the utility of ICD-10-CM codes in the biomedical domain.
Asunto(s)
Registros Electrónicos de Salud , Clasificación Internacional de Enfermedades , Lenguaje , Aprendizaje Automático , Procesamiento de Lenguaje NaturalRESUMEN
Aims: Measuring the total and free concentrations of hormones is useful, but the technology to do this simultaneously is lacking. Methods: A new method offers the ability to measure these parameters concurrently for testosterone, thyroxine and triiodothyronine. Results: The free concentrations showed significant correlations with patients' vital statistics. Overall, 67% of correlations for total concentration showed that the new and classical methods had equal accuracy, or that comprehensive ultrafiltration was more accurate. The protein binding term was found to correlate significantly with the patients' luteinizing hormone, prostate-specific antigen and height. Conclusion: Comprehensive ultrafiltration for measuring the total concentration, free concentration and protein binding term uses less sample and is much faster than measuring these parameters with three separate methods.
Asunto(s)
Tiroxina , Triyodotironina , Masculino , Humanos , Unión Proteica , Testosterona , UltrafiltraciónRESUMEN
This paper presents novel datasets providing numerical representations of ICD-10-CM codes by generating description embeddings using a large language model followed by a dimension reduction via autoencoder. The embeddings serve as informative input features for machine learning models by capturing relationships among categories and preserving inherent context information. The model generating the data was validated in two ways. First, the dimension reduction was validated using an autoencoder, and secondly, a supervised model was created to estimate the ICD-10-CM hierarchical categories. Results show that the dimension of the data can be reduced to as few as 10 dimensions while maintaining the ability to reproduce the original embeddings, with the fidelity decreasing as the reduced-dimension representation decreases. Multiple compression levels are provided, allowing users to choose as per their requirements. The readily available datasets of ICD-10-CM codes are anticipated to be highly valuable for researchers in biomedical informatics, enabling more advanced analyses in the field. This approach has the potential to significantly improve the utility of ICD-10-CM codes in the biomedical domain.
RESUMEN
Background On June 24, 2022, the United States (U.S.) Supreme Court ruled in Dobbs v. Jackson Women's Health that the Constitution does not grant women the right to abortion. This new ruling may have a profound impact on not only the attitudes of medical trainees but the nature in which they are trained when it comes to abortion practices, indications, or procedures. Some clinics where medical schools provide first-hand abortion experience have closed. As a surge of extreme restrictions on abortion has been seen in certain states in the U.S., medical schools and residency programs need to address these issues to ensure future physicians are adequately prepared. The purpose of this study was to assess factors that influence medical students' attitudes toward abortion, specifically knowledge about abortion, religiosity, and philosophical group affiliation regarding abortion (i.e., "pro-choice vs. "pro-life"). Methodology This cross-sectional study collected data from a convenience sample of 413 medical students between October and December 2022 via an online, anonymous questionnaire. Major study variables as depicted in the published literature led to the development of the Abortion Attitudes Questionnaire (AAQ) for medical students. The AAQ contained validated scales to assess the contribution of levels of abortion knowledge and religiosity as well as sample characteristics on attitudes about abortion in medical students. Speakman rank correlation and linear multivariate regression were used for hypothesis testing to explore the contributions of the dependent variables to attitudes about abortion in medical students. Results The mean age of the participants was 25.8 years (SD = 2.96; range = 21-45 years). Linear regression modeling indicated that religiosity, abortion knowledge, being a woman, and group affiliation regarding abortion ("pro-choice" or "pro-life") explained a significant amount of the variance (60%) in abortion attitudes scores in medical students. A significant regression equation was found, F(6,373) = 83.570, p < 0.0001, R2 = 0.603, R2 adjusted = 0.611). Less religiosity, greater abortion knowledge, being a woman, and identifying as "pro-choice" significantly contributed to more positive attitudes toward abortion in this sample of medical students. Interestingly, while moderately correlated with abortion attitudes (r = 0.436,p < 0.01), the single item regarding the importance of religion in one's life did not contribute to the model. Conclusions The present study is the first, to our knowledge, to identify medical student characteristics (i.e., sex, "pro-choice" vs. "pro-life" group affiliation, level of knowledge about abortion, and religiosity) as indicators of abortion attitudes. With the reversal of Roe v. Wade, attention must be given to the possible change in medical students' attitudes toward abortion (as well as any newly developed constraints on clinical training) and to ensure the provision of comprehensive education as U.S. states will determine the limits of these practices and procedures. While further research in this area is needed, findings from this study can help assess students' attitudes about abortion and guide medical education efforts to better prepare tomorrow's obstetrics and gynecology physicians.
RESUMEN
Introduction: High-dose interleukin-2 (HD IL-2) and pembrolizumab are each approved as single agents by the U.S. F.D.A. for the treatment of metastatic melanoma. There is limited data using the agents concurrently. The objectives of this study were to characterize the safety profile of IL-2 in combination with pembrolizumab in patients with unresectable or metastatic melanoma. Methods: In this Phase Ib study, patients received pembrolizumab (200 mg IV every 3 weeks) and escalating doses of IL-2 (6,000 or 60,000 or 600,000 IU/kg IV bolus every 8 hours up to 14 doses per cycle) in cohorts of 3 patients. Prior treatment with a PD-1 blocking antibody was allowed. The primary endpoint was the maximum tolerated dose (MTD) of IL-2 when co-administered with pembrolizumab. Results: Ten participants were enrolled, and 9 participants were evaluable for safety and efficacy. The majority of the evaluable participants (8/9) had been treated with PD-1 blocking antibody prior to enrollment. Patients received a median of 42, 22, and 9 doses of IL-2 in the low, intermediate, and high dose cohorts, respectively. Adverse events were more frequent with increasing doses of IL-2. No dose limiting toxicities were observed. The MTD of IL-2 was not reached. One partial response occurred in 9 patients (11%). The responding patient, who had received treatment with an anti-PD-1 prior to study entry, was treated in the HD IL-2 cohort. Discussion: Although the sample size was small, HD IL-2 therapy in combination with pembrolizumab appears feasible and tolerable. Clinical trial registration: ClinicalTrials.gov, identifier NCT02748564.
RESUMEN
BACKGROUND: Botulinum toxin type A (BoNTA) injections for the treatment of facial lines may lead to pleasant or undesirable changes in eyebrow height and position. OBJECTIVES: The aim of this study was to evaluate the impact of glabellar injection of DaxibotulinumtoxinA for Injection (DAXI), a novel BoNTA formulation, on eyebrow position and frontalis activity. METHODS: This study involved the post hoc analysis of adult patients from the Phase 2a forehead lines (FHL, Nâ =â 60) and open-label safety (OLS, Nâ =â 175) studies who received a single dose of DAXI 40 U to the glabella and for whom facial photographs were taken at rest and at maximum eyebrow elevation. Median vertical and horizontal displacement of the brows and median forehead strain (an objective quantitative assessment of frontalis activity) from baseline to 2 weeks after glabellar DAXI injection were measured. RESULTS: Two weeks after glabellar DAXI injection, vertical eyebrow movement (at rest) of the lateral brow was observed. In both studies, vertical movement was greatest in the lateral (0.6-0.9 mm) and mid (0.5-0.7 mm) brow regions; movement in the medial brow was negligible (0-0.23 mm). In both studies, a graded reduction in forehead strain was observed 2 weeks post glabellar DAXI injection, with the greatest reduction being in the lower segment (FHL, -73%; OLS, -82%). Treatment with DAXI showed improvements in FHL wrinkle severity. CONCLUSIONS: Glabellar injection of DAXI showed a positive reduction in dynamic frontalis activity that maintained or had a positive effect on eyebrow position. Vertical movement of the brow was suggestive of an improved eyebrow shape with a mild lateral arch.
Asunto(s)
Toxinas Botulínicas Tipo A , Fármacos Neuromusculares , Envejecimiento de la Piel , Adulto , Humanos , Frente , Inyecciones , Músculos , Fármacos Neuromusculares/efectos adversos , Ensayos Clínicos Fase II como AsuntoRESUMEN
STUDY OBJECTIVE: Patients undergoing diagnostic imaging studies in the emergency department (ED) commonly have incidental findings, which may represent unrecognized serious medical conditions, including cancer. Recognition of incidental findings frequently relies on manual review of textual radiology reports and can be overlooked in a busy clinical environment. Our study aimed to develop and validate a supervised machine learning model using natural language processing to automate the recognition of incidental findings in radiology reports of patients discharged from the ED. METHODS: We performed a retrospective analysis of computed tomography (CT) reports from trauma patients discharged home across an integrated health system in 2019. Two independent annotators manually labeled CT reports for the presence of an incidental finding as a reference standard. We used regular expressions to derive and validate a random forest model using open-source and machine learning software. Final model performance was assessed across different ED types. RESULTS: The study CT reports were divided into derivation (690 reports) and validation (282 reports) sets, with a prevalence of incidental findings of 22.3%, and 22.7%, respectively. The random forest model had an area under the curve of 0.88 (95% confidence interval [CI], 0.84 to 0.92) on the derivation set and 0.92 (95% CI, 0.88 to 0.96) on the validation set. The final model was found to have a sensitivity of 92.2%, a specificity of 79.4%, and a negative predictive value of 97.2%. Similarly, strong model performance was found when stratified to a dedicated trauma center, high-volume, and low-volume community EDs. CONCLUSION: Machine learning and natural language processing can classify incidental findings in CT reports of ED patients with high sensitivity and high negative predictive value across a broad range of ED settings. These findings suggest the utility of natural language processing in automating the review of free-text reports to identify incidental findings and may facilitate interventions to improve timely follow-up.
Asunto(s)
Procesamiento de Lenguaje Natural , Radiología , Humanos , Estudios Retrospectivos , Alta del Paciente , Aprendizaje Automático , Servicio de Urgencia en Hospital , Hallazgos IncidentalesRESUMEN
Mind-wandering assessment relies heavily on the thought probe technique as a reliable and valid method to assess momentary task-unrelated thought (TUT), but there is little guidance available to help researchers decide how many probes to include within a task. Too few probes may lead to unreliable measurement, but too many probes might artificially disrupt normal thought flow and produce reactive effects. Is there a "Goldilocks zone" for how few thought probes can be used to reliably and validly assess individual differences in mind-wandering propensity? We address this question by reanalyzing two published datasets (Study 1, n = 541; Study 2, ns ≈ 260 per condition) in which thought probes were presented in multiple tasks. Our primary analyses randomly sampled probes in increments of two for each subject in each task. A series of confirmatory factor analyses for each probe "bin" size tested whether the latent correlations between TUT rate and theoretically relevant constructs like working memory capacity, attention-control ability, disorganized schizotypy, and retrospective self-reported mind wandering changed as more probes assessed the TUT rate. TUT rates were remarkably similar across increasing probe-bin sizes and zero-order correlations within and between tasks stabilized at 8-10 probes; moreover, TUT-rate correlations with other latent variables stabilized at about 8 thought probes. Our provisional recommendation (with caveats) is that researchers may use as few as 8 thought probes in prototypical cognitive tasks to gain reliable and valid information about individual differences in TUT rate.
Asunto(s)
Análisis de Datos Secundarios , Pensamiento , Humanos , Estudios Retrospectivos , Atención , Memoria a Corto PlazoRESUMEN
The Matsigenka people living traditional lifestyles in remote areas of the Amazon rely on a fish-based diet that exposes them to methylmercury (MeHg) at levels that have been associated with decreased IQ scores. In this study, the association between Hg levels and working memory was explored using the framework of the Multicomponent Model. Working memory tasks were modified to fit the culture and language of the Matsigenka when needed and included measures for verbal storage (Word Span) visuospatial storage (Corsi Block Task) and a measure of executive functions, the Self-Ordered Pointing Task (SOPT). An innovation of the Trail Making Tests A & B (TMT A & B) was pilot tested as another potential measure of executive functions. The mean hair Hg levels of 30 participants, ages 12 to 55 years, from three different communities (Maizal, Cacaotal and Yomibato) was 7.0 ppm (sd = 2.40), well above the World Health Organization (WHO) limit for hair of 2.0 ppm and ranged from 1.8 to 14.2 ppm, with 98% of a broader sample of 152 individuals exceeding the WHO limit. Hair Hg levels showed significant associations with cognitive performance, but the degree varied in magnitude according to the type of task. Hg levels were negatively associated with executive functioning performance (SOPT errors), while Hg levels and years of education predicted visuospatial performance (Corsi Block accuracy). Education was the only predictor of Word Span accuracy. The results show that Hg exposure is negatively associated with working memory performance when there is an increased reliance on executive functioning. Based on our findings and the review of the experimental research, we suggest that the SOPT and the Corsi Block have the potential to be alternatives to general intelligence tests when studying remote groups with extensive cultural differences.
Asunto(s)
Memoria a Corto Plazo , Mercurio , Animales , Función Ejecutiva , Humanos , Pueblos Indígenas , Mercurio/análisis , Pruebas Neuropsicológicas , PerúRESUMEN
Background There has been increased attention given to understanding the uses of medical cannabis (MC) for symptom management of various medical conditions. Physicians receive minimal training in medical school and rely mostly on anecdotal evidence; by proxy, medical students generally do receive formal training in MC. It is unknown how medical students perceive MC, including its efficacy, appropriateness in medicine, its possible adverse effects, and its value for patients. This study investigated medical students' perceived knowledge, beliefs, and attitudes toward MC to better understand their knowledge about and attitudes toward MC. Method Using a semi-structured interview guide, eight focus groups were conducted with 83 medical students via Zoom virtual meeting platform (Zoom Video Communications, Inc., San Jose, California, United States) in June 2022. The interviews were guided by the following content areas: (1) beliefs about cannabis' therapeutic utility, (2) perceived knowledge about MC, (3) the role of the physician regarding MC, (4) concern for cannabis' adverse effects, and (5) MC education in the school curriculum. Data were analyzed using thematic analysis, an iterative, systematic process of coding patterns, and emerged themes in the interview data to explore medical students' perceptions about MC. Themes were validated based on whether each theme captured distinct parts of the interview data and whether their content cohered meaningfully. Results Four themes emerged from the focus group interviews investigating medical students' perceptions of MC: (1) erroneous beliefs about MC, (2) unreliable sources of information, (3) mixed attitudes toward legalization, and (4) desire for MC education while in medical school. Attitudes regarding MC in general, including legalization, varied by United States state of origin of the student and exposure to MC (e.g., use by family member). Conclusion MC seems to be a significant issue for medical trainees who might be required to recommend it to patients and manage coexisting therapies. Cultivating new knowledge about students' perceptions and perceived knowledge about medicinal options and dosing of MC is critical for medical educators as they design undergraduate curricular initiatives for future physicians.