Inka Health

Inka Health is proud to partner with Loon in their advancement of evidence synthesis work. We’ve had an outstanding experience collaborating with Ghayath Janoudi, mara rada, and Tim Disher who passionately leverage technology and research expertise to address today’s critical questions. Check out their post below. If you are a PI, clinician, or representative of a patient group looking to accelerate your research, Loon is offering three fully funded AI-powered Systematic Literature Reviews (SLRs), including comprehensive methods and analysis support, through their community initiative #LoonChorus. This is a fantastic chance to enhance your research with cutting-edge tools and expert support. 🌟 👉 Interested or know someone who would be a great fit? Please check the details below or send them a direct message for more information. #InkaHealth #LoonChorus #SystematicReview #ResearchSupport #HealthcareResearch #Collaboration #AIinResearch #AcademicSupport

🚀 UPCOMING WEBINAR! AI IN SYSTEMATIC REVIEWS 🔥 Over the past year I find myself having frequent conversations with fellow researchers about the expanding use of AI in knowledge syntheses, and I often hear thoughts about both: 1️⃣ The excitement it holds for the future of HEOR and EBM; and 2️⃣ Apprehensions about adoption of such technologies. Both sentiments are highly understandable! Another key theme that often comes about in these discussions is the need to enhance our understanding related to AI tools and their potential roles and benefits related to knowledge synthesis. When it comes to the terminology used in this space and also the technology behind emerging tools, there remains a gap between what’s at the forefront of technology and what most researchers performing systematic reviews have had the chance to consume...it can be both exciting and daunting! 👉 I’ve had the chance to talk upon various occasions in the past year with Ghayath Janoudi, Tim Disher and mara rada from Loon about these topics, and it’s been a joy. Given the value these conversations held for me, I wanted to flag their upcoming webinar for others as a chance to further their background in this rapidly evolving space. They’re planning to cover a historical overview on AI in reviews, clarity on AI jargon and terminology that can be a struggle to grasp, how to interpret research in this emerging space, and other topics. It's also a great chance to get a sense of the plentiful materials and trends coming up at ISPOR Europe in November within this field. 🎉 For those that can attend, enjoy! 👉 Registration link: https://lu.ma/qsvocnu5   #ISPOREurope #ISPOR #HTA #MarketAccess #ComparativeEffectiveness #NetworkMetaAnalysis #EvidenceSynthesis #ResearchExcellence #EvidenceBasedMedicine #AI #ClinicalResearch #PublicHealth #SystematicReview #HEOR #AISystematicReview #EvidenceSynthesis

How can synthetic data change the future of medical imaging? What impact does this have on the development of AI models, especially in the critical field of X-ray image analysis? The recent research published in *Nature Machine Intelligence* proposes a significant shift in how we approach training AI models for medical imaging. The authors introduce SyntheX, a framework that utilizes synthetic data generated from CT scans to train AI models, eliminating the need for large-scale real-data collection. This approach not only matches the performance of models trained on real data but can even surpass it, particularly when real-world data is scarce or difficult to obtain. https://lnkd.in/e3unbBB7 For professionals working in AI and medical imaging, this study offers clear guidance: integrating synthetic data with domain generalization techniques can accelerate the development of robust and reliable AI systems. The researchers successfully applied this method to tasks like hip imaging, robotic surgical tool detection, and COVID-19 lesion segmentation, demonstrating that synthetic data can provide the necessary variability and accuracy for effective AI model training. As we push forward in the development of AI for clinical applications, the strategic use of synthetic data could be key to overcoming current limitations and improving patient outcomes. #ArtificialIntelligence #MedicalImaging #SyntheticData #XRayAnalysis #AIinHealthcare #HealthTech #MedicalInnovation #DataScience #MachineLearning #DigitalHealth #ClinicalResearch #Oncology #SurgicalRobotics #COVID19Research #HealthOutcomes #RealWorldData

We're biased against selection bias! The recent article by Liliana Paloma Rojas-Saunero et al. on selection bias in health disparities research is a must-read for anyone in the health research community. The authors highlight how selection bias can significantly distort our understanding of health disparities, using dementia research as a case study. https://lnkd.in/d3pJ2hR7 They emphasize the need for clearly defining target populations and estimands to avoid generalizability and collider-stratification biases. This meticulous approach ensures that research findings truly reflect the health disparities present in the broader population, rather than just a skewed subset. For patients, this means more accurate diagnoses, treatments, and interventions tailored to their real-world needs. The authors recommend specific strategies to mitigate selection bias, such as designing inclusive recruitment and retention strategies, and utilizing advanced analytic methods like inverse probability weighting. For example, in studies involving diverse racial/ethnic groups, employing consistent recruitment methods across all groups can prevent biases that arise from differential inclusion. These practices are not just academic exercises; they directly impact the quality of care and health equity, ensuring that all patients benefit from reliable and relevant health research findings. #HealthResearch #SelectionBias #HealthDisparities #DementiaResearch #InkaHealth #Epidemiology #PublicHealth #HealthcareAnalytics #PatientCare #ClinicalResearch #BiasMitigation #HealthEquity #ResearchMethods #PopulationHealth #DataScience

How can we ensure that risk prediction models for coronary heart disease (CHD) in diabetic patients are accurate and reliable? Are current models meeting the needs of clinicians and health technology assessment (HTA) stakeholders? A recent review by Li Jiu, an analyst at IQVIA et al. emphasizes critical gaps in the quality and applicability of these models, underscoring the need for improvement. The authors found that while many models demonstrate good predictive performance, most exhibit a high risk of bias, particularly in the analysis domain. This finding raises concerns about the reliability of these models for clinical and economic decision-making. https://lnkd.in/d3baxQ83 The authors recommend several specific actions to enhance model quality and applicability. Clinicians are advised to carefully examine the declared performance of prediction models and consider those with low risk of bias across all domains except analysis, provided there is validation in the same setting. Additionally, the development of new, adapted tools for assessing model applicability in HTA contexts is suggested. The study emphasizes the importance of providing transparent mathematical equations in model reports, allowing for accurate risk calculations over time. By addressing these recommendations, we can improve the predictive accuracy of CHD risk models, ultimately benefiting diabetic patients by enabling better-informed treatment decisions and resource allocation. #HealthcareAnalytics #HealthTech #DiabetesManagement #CoronaryHeartDisease #RiskPrediction #ClinicalDecisionMaking #HTA #MachineLearning #PredictiveModeling #HealthcareInnovation #HealthOutcomes #PatientCare #MedicalResearch #HealthEconomics #DataDrivenDecisions #Diabetes

Interested in BlueSky and #causalinference but don’t know where to start? Here’s a great causal inference starter pack!

Can we trust US based RWE to make healthcare decisions in Canada? It depends. For example, overall survival estimates for advanced non–small cell lung cancer (NSCLC) showed a mean absolute difference of less than 5% when transported from US to Canadian patients? A study published in JAMA Network Open by Sreeram Ramagopalan highlights the potential of using US clinical practice data to inform regulatory decision-making and health technology assessments in Canada. This is particularly crucial as it allows for timely and evidence-based healthcare decisions in regions where direct estimation is challenging due to small sample sizes or immature outcome data. https://lnkd.in/dGC6_Nzy The authors recommend leveraging real-world evidence from US clinical practice data to support regulatory and reimbursement decisions in other countries, provided that proper adjustments for baseline characteristics are made. This approach can significantly accelerate access to novel therapies for patients with advanced NSCLC by ensuring that evidence from established healthcare systems is utilized effectively. The study emphasizes the importance of rigorous transportability analysis to confirm the external validity of survival estimates, thus offering a reliable framework for global healthcare policy and practice. #HealthEconomics #OncologyResearch #RealWorldEvidence #NSCLC #CancerTreatment #ClinicalOutcomes #HealthcarePolicy #RegulatoryScience #HealthTechnologyAssessment #SurvivalAnalysis #MedicalResearch #DataAnalytics #PatientCare #GlobalHealth #EvidenceBasedMedicine #CDA #HealthCanada

How can we better predict treatment outcomes across diverse cancer types? What are the implications of tumor-specific decisions in the era of precision medicine? The recent study by Yilin Chen, a senior research associate at Curta et al. delves into the application of Bayesian hierarchical models to refine tumor-specific estimates using tumor-agnostic evidence from basket trials. By addressing the inherent heterogeneity in treatment effects across different tumor types, this research enhances the precision of outcome estimates, which is crucial for healthcare decision-making. https://lnkd.in/dhQpgrnU The authors recommend leveraging Bayesian hierarchical models to reduce variability and uncertainty in estimates, facilitating more robust policy decisions for tumor-agnostic drugs. For instance, their analysis of pembrolizumab across multiple cancer types highlighted significant variability in survival endpoints, underscoring the need for tailored confirmatory trials or post-marketing evidence generation. This approach can lead to more informed clinical and coverage decisions, ultimately improving patient outcomes by ensuring that treatments are appropriately targeted and effective. The insights provided by this study are essential for advancing the field of precision oncology and optimizing therapeutic strategies for cancer patients. #PrecisionMedicine #OncologyResearch #BayesianAnalysis #HealthOutcomes #BasketTrials #TumorAgnosticDrugs #ClinicalDecisionMaking #HealthcarePolicy #RealWorldEvidence #CancerTreatment #BayesianHierarchicalModels #PatientOutcomes #MedicalResearch #EvidenceBasedMedicine #InkaHealth

Bayesian Hierarchical Modelling (BHM) offers a powerful approach for addressing the growing challenges in comparing cancer therapies from single-arm basket trials, particularly for rare mutations like NTRK-fusion solid tumours. In their upcoming presentation at Bayes2024 (American Statistical Association - ASA), Emma Mackay and her co-authors Aaron Springford and Sofia Dias introduce two novel BHM approaches for performing Indirect Treatment Comparisons (ITC) between drugs trialled in different basket studies. These models provide a middle-ground solution by balancing the individual histology-level differences and overall treatment effects, allowing for more informed health technology assessments (HTA). The application of these methods to larotrectinib and entrectinib showcases their potential for estimating treatment efficacy while accounting for variability across tumour types. For patients, this research is crucial. As new drugs targeting rare mutations increase, having reliable comparisons between treatments trialled separately ensures that HTA can guide more accurate decision-making. The authors recommend adopting BHM-based ITC approaches to offer a more nuanced comparison of therapies, ensuring that the benefits across different histologies are properly accounted for. This not only helps ensure the right treatments reach the right patients, but also improves confidence in treatment recommendations.  Join her in Rockville, MD! https://lnkd.in/gbvKPGwa #BayesianMethods #HealthTechnologyAssessment #OncologyResearch #BasketTrials #NTRKFusion #CancerTreatment #BayesianHierarchicalModels #DrugComparison #RareMutations #Larotrectinib #Entrectinib #RealWorldEvidence #RWE #ClinicalDecisionSupport #AdvancedAnalytics #InkaHealth

𝐓𝐡𝐞 𝐛𝐞𝐬𝐭 𝐑𝐖𝐃 𝐲𝐨𝐮'𝐯𝐞 𝐧𝐞𝐯𝐞𝐫 𝐡𝐞𝐚𝐫𝐝 𝐨𝐟... Genotype data from 520,000 individuals combined with patient-level linkage to population-wide registry data including: 👉 Demographics 👉 Diagnoses 👉 Medications 👉 Procedures 👉 PROs 👉 Labs and more... Sound too good to be true? Not in Finland. 🚀 The Quantify Research team is here to introduce you to the world-class Finnish data ecosystem and guide you in how to capture its great potential Katja Nolvi Ville Perälä Emilie Toresson Grip