Suman Lama

Machine Learning Engineer

Suman Lama

Suman Lama is a dedicated Senior Machine Learning Engineer at the Renal Research Institute (RRI) and a PhD candidate at the Pontifical Catholic University of Paraná (PUCPR). With a master’s degree in data science from Worcester Polytechnic Institute (WPI), Suman has cultivated a strong foundation in applying advanced machine learning techniques to solve real-world challenges in the healthcare industry.

Since transitioning into healthcare, Suman has focused on developing innovative machine learning models to enhance patient care and outcomes. Among these is the AVF Failure Prediction Model, a project currently in the pilot phase across several clinics, aimed at improving the management of arteriovenous fistulas in patients undergoing dialysis. Some of Suman’s work has also been published, reflecting a commitment to contributing to the scientific community and advancing the state of research in healthcare and artificial intelligence. Suman remains dedicated to making a lasting impact on patient outcomes through data-driven solutions.

Outside of work, Suman enjoys exploring new ideas, staying updated on the latest advancements in machine learning, and fostering a lifelong curiosity for knowledge.

Recent Articles by Suman Lama

  • Kidney medicine
    July 17, 2026
    Real-world Evidence for Improvements in Inflammation and Anemia Biomarkers After the Initiation of GLP-1RA in Patients Receiving Hemodialysis
    Suman Lama, Sheetal Chaudhuri, Derek Blankenship, Andrea Nandorine Ban, Len Usvyat, Roberto Pecoits-Filho, Benjamin E Hippen
    RESULTSCompared with matched controls, GLP-1RA users demonstrated reductions in systemic inflammation, characterized by lower neutrophil-to-lymphocyte ratio (3.97 vs 4.64, with a mean difference of -0.67 [95% CI, -0.89 to -0.45]) and white blood cell counts at 12 months, alongside greater improvements in serum albumin level (3.92 vs 3.86 g/dL, with a mean difference of 0.06 [95% CI, 0.04-0.08]). Although hemoglobin levels remained comparable between groups (10.9 vs 10.84 g/dL, with a mean difference of 0.06 [95% CI, -0.004 to 0.12]), GLP-1RA users had lower cumulative ESA exposure (1,881.9 vs 2,005.8 μg; mean difference of -123.9 μg, [95% CI, -201.7 to -46.1] μg) and a modestly reduced rate of erythropoietin resistance index compared with nonusers. Additionally, iron stores (ferritin level and transferrin saturation value) were consistently higher in the GLP-1RA group.Patients receiving hemodialysis often experience inflammation and anemia, which can make treatment more difficult and affect outcomes. Glucagon-like peptide-1 receptor agonists (GLP-1RAs), commonly used for diabetes and weight management, may also have anti-inflammatory effects. We studied whether starting GLP-1RA therapy in patients on hemodialysis was associated with changes in markers of inflammation, nutrition, and anemia over 1 year, compared with similar patients not receiving these medications. We found that patients using GLP-1RAs showed consistent improvements in inflammatory markers and indicators of iron use, along with lower requirements for anemia medications, while maintaining similar hemoglobin levels. These findings suggest that GLP-1RAs may have benefits beyond glucose control, though further studies are needed to confirm their clinical impact.RATIONALE & OBJECTIVEGlucagon-like peptide-1 receptor agonists (GLP-1RAs) are increasingly used for diabetes and obesity but have not been studied in patients with kidney failure with replacement therapy receiving hemodialysis in prospective trials. GLP-1RAs may influence inflammatory pathways relevant to anemia, although their effects in hemodialysis patients remain unclear.CONCLUSIONSIn patients receiving maintenance hemodialysis, GLP-1RA agonist use was associated with reductions in inflammatory markers and lower ESA requirements; however, as an observational study, causality cannot be inferred. These findings suggest GLP-1RAs may have potential benefits in managing inflammation and inflammation-mediated anemia in kidney failure.STUDY DESIGNRetrospective matched cohort study.ANALYTICAL APPROACHLinear mixed models were used to compare longitudinal trajectories between matched groups.SETTING & PARTICIPANTSData were derived from a large US dialysis provider (Fresenius Kidney Care). We identified 2,468 adult patients who initiated GLP-1RA therapy between January 1, 2023, and November 1, 2024, and matched 1:1 to nonusers using propensity scores based on demographic characteristics, comorbidities, and baseline laboratory values.OUTCOMESLongitudinal changes over 12 months in inflammatory markers (neutrophil-to-lymphocyte ratio, white blood cell count, and albumin level) and anemia parameters (hemoglobin level, ferritin level, transferrin saturation value, erythropoiesis-stimulating agent (ESA) dose, and erythropoietin resistance index).EXPOSUREInitiation of GLP-1RA therapy.

Working at RRI inspires me daily to push the boundaries of research and innovation in renal care. Together, we are transforming challenges into opportunities for better patient outcomes.

Suman Lama
Machine Learning Engineer