Scalability gaps in Medicine

Representative open questions

Showing 30 of 113 — one per source paper, highest-quality first.

• Effectiveness of screening modalities for early detection of diabetic retinopathy: a systematic review and meta-analysis of tele-ophthalmology, AI-based tools, and conventional methods (2026) · doi

  Inter-grader variability in AI-assisted and physician-based DR assessment has been identified in single urban healthcare systems; provider-level variability assessment and quality control mechanisms for deep learning DR algorithms across geographically diverse settings remain underdeveloped.

• Management of depression utilizing Traditional Chinese Medicine (2026) · doi

  The paper calls for standardization of herbal preparations, rigorous experimental design, and shared data platforms to enable reproducible science in integrated TCM-Western medicine therapy for depression, but does not specify standards for bioactive compound quantification, extraction protocols, or data repository infrastructure needed.

• Cardiovascular Risk Prediction Using Machine Learning: Advances and Clinical Translation (2026) · doi

  Continuous learning systems for cardiovascular risk prediction pose unresolved regulatory challenges due to their dynamic model update nature, but the paper does not specify what post-deployment monitoring mechanisms, performance drift detection thresholds, or adaptive regulatory approval processes FDA and EMA should implement for evolving machine learning medical devices.

• Echocardiographic Assessment of Left Ventricular Mass Regression and Functional Changes after Bariatric Surgery (2026) · doi

  The research involved a small population size, which restricts the generalizability of echocardiographic findings regarding structural and functional cardiac changes after bariatric surgery. Future studies should recruit larger, more heterogeneous patient cohorts to validate whether E/A ratio improvements and LV mass regression are consistent across diverse obesity phenotypes.

• Inaugural Lifestyle Medicine Editorial (2020) · doi

  While the paper emphasizes international collaboration for vaccine development and evidence-based treatments in COVID-19, it does not address how lifestyle medicine programs can be standardized, scaled, or adapted across different healthcare systems, regulatory frameworks, and resource settings globally to achieve universal health coverage.

• City planning as preventive medicine (2015) · doi

  The paper proposes that city planners and clinicians collaborate on community health needs assessments required by the ACA and IRS, but does not address how to operationalize or scale this collaboration across municipalities with varying governance structures, planning capacities, and healthcare system configurations.

• Personalized Medicine and the Practice of Medicine in the 21st Century (2020) · doi

  Development and implementation of high-quality, integrated pharmacogenomics databases and IT-based technologies for clinical decision support requires standardization specifications and interoperability standards. The current state of infrastructure for translating pharmacogenomic findings into actionable clinical recommendations across diverse healthcare systems is inadequately defined.

• Principles of Social Emergency Medicine (2019) · doi

  The authors emphasize that emergency medicine researchers must collaborate with departments of public health, law enforcement, and case managers, but provide no concrete framework for integrating data collection, outcome tracking, or intervention evaluation across these traditionally siloed systems.

• Alarplasty and its effects on respiratory function: a cross-sectional study (2026) · doi

  The relatively small sample size (n=22) with multiple statistical comparisons across pulmonary and respiratory muscle strength variables increases risk of Type I error. Future studies must employ larger sample sizes with appropriate statistical corrections (e.g., Bonferroni adjustment) to increase generalizability of findings regarding alarplasty's effects on MIP, MEP, and spirometric parameters.

• Quantitative susceptibility mapping and MRS-based multimodal machine learning for early Parkinson’s disease classification (2026) · doi

  The four classification algorithms (Random Forest, SVM, XGBoost, LightGBM) were evaluated on an independent test cohort, but the study did not assess model robustness across external validation cohorts from different imaging centers or patient populations, which is critical for determining generalizability of the multimodal QSM-MRS approach for early PD detection.

• Clinical outcomes of the tunnelized-facial artery myomucosal island flaps in the oral cavity and comparison with cutaneous flaps in an animal model (2026) · doi

  The animal model comparing t-FAMMIF and submental skin island flap (SMIF) was exploratory with limited scale and did not investigate the mechanism underlying nerve sprouting and salivary gland preservation. Larger-scale animal experiments with temporal analysis of nerve regeneration markers (e.g., growth-associated protein 43, neurofilament) and fibrosis quantification are needed to mechanistically validate the proposed permissive microenvironment for nerve growth.

• Application of deep learning-clinical baseline feature fusion model to predict postoperative mortality in elderly patients with hip fracture: a multicenter study (2026) · doi

  The ratio of 1-year mortality outcome events to effective predictors in the final deep learning-clinical baseline fusion model falls below ideal clinical research criteria, creating overfitting risk; expanding the multicenter sample size to achieve optimal event-predictor ratios is essential for validating model generalizability and robustness across diverse healthcare settings.

• Global trajectories of polygenic risk score research: a systematic bibliometric review of precision medicine, equity, and clinical translation (2026) · doi

  While the paper documents that expanding collaborative networks and diversifying funding streams are needed to support inclusive PRS innovation, it does not specify which lower-income regions currently lack research infrastructure for PRS development or identify concrete capacity-building mechanisms (e.g., training programs, technology transfer agreements, federated learning frameworks) necessary to broaden global participation in polygenic risk prediction research.

• Cerebrospinal fluid cytokine-driven immune responses in HIV-negative cryptococcal meningitis (2026) · doi

  The study included only 15 untreated cryptococcal meningitis (UCM) patients with variability in principal component scores, limiting statistical power and generalizability. Larger cohorts are needed to validate the observed correlations between CSF parameters (protein concentration, leucocyte counts, glucose ratio) and specific cytokine levels (IL-18, MCP-3, MIG, MIP-1α) in HIV-negative cryptococcal meningitis populations.

• The expanding role of artificial intelligence in personalised medicine: from innovation to individualized care (2026) · doi

  Integration of foundation models for AI-assisted drug repurposing with real-time clinical phenotyping using Eye2Gene and multimodal imaging for inherited retinal diseases lacks standardized data sharing frameworks and interoperability protocols across institutional genomic databases.

• Radiomics and artificial intelligence in precision radiotherapy for cervical cancer: a narrative review (2026) · doi

  Integration of radiomics-AI models with treatment planning systems in cervical cancer radiotherapy, inference time requirements within routine clinical workflows, regulatory requirements, data governance, model interpretability, and clinician trust mechanisms are rarely addressed in current literature.

• Traditional Korean medicine treatment patterns in patients with low back pain: A cross-sectional study based on the 2017 Korean Medicine Utilization and Herbal Medicine Consumption Survey (2020) · doi

  Treatment patterns with fewer respondents (physiotherapy alone group, specific acupuncture-cupping-pharmaco-acupuncture combinations) showed extreme results that are difficult to generalize; increasing patient numbers within each distinct treatment pattern would determine whether findings remain consistent.

• The adjunctive diagnostic value of arterial spin labeling and depth of invasion in grading tongue cancers (2026) · doi

  The conclusion calls for validation with larger sample sizes from multiple centers and at higher magnetic field strengths (implying the current study used lower field strength MRI). ASL protocols for tongue cancer grading should be optimized and prospectively validated across 3T and 7T magnetic field strengths at multicenter sites.

• Ultrasound-enhanced fine-needle aspiration biopsy improves yield of solid benign parotid gland tumor tissue: a pilot study (2026) · doi

  The study involved only one experienced radiologist in a general anesthesia setting with a small sample size; generalizability of USeFNAB yield improvement requires multi-center validation with multiple operators performing biopsies in typical outpatient clinic conditions.

• Short-Term Clinical Effects of Standardized Syzygium polyanthum (Bay Leaf) Tea Infusion on Serum Uric Acid Modulation in Hyperuricemia: A Pilot Trial in Primary Care (2026) · doi

  Integration of the standardized botanical tea intervention into Indonesian primary care frameworks (specifically the Integrated Guidance Post for Non-Communicable Diseases) and village-level healthcare facilitator distribution programs requires implementation studies assessing feasibility, adherence, cost-effectiveness, and clinical outcomes when deployed at scale across rural and urban Southeast Asian populations.

• Continuous-flow ventilation with VENTIJET in moderate ARDS: a pilot safety and feasibility study (2026) · doi

  The small sample size and pilot safety feasibility design restrict statistical power and generalizability; a larger prospective study is needed to determine whether the hemodynamic tolerability observed (absence of hypotension and vasoactive escalation despite increased mean airway pressure) holds across diverse ARDS patient populations with varying baseline intrathoracic pressure and venous return characteristics.

• A novel transcutaneous auricular neurostimulation electrode configuration for treatment of heavy menstrual bleeding: an open-label trial (2026) · doi

  Treatment duration was limited to two menstrual cycles post-baseline; extended evaluation across multiple consecutive menstruations is needed to assess durability, tolerance, and long-term efficacy of the transcutaneous auricular neurostimulation electrode configuration for sustained reduction in menstrual blood loss.

• Radiopharmaceutical Therapy in Oncology: Perspectives from Transatlantic Experts (2026) · doi

  The paper identifies the need to establish reliable distribution networks capable of safe and timely radiopharmaceutical delivery with ability for 'upscaling' to meet rapid increases in demand, but does not specify logistics models, distribution infrastructure requirements, or scalability thresholds needed to ensure access across countries with varying nuclear medicine facility availability.

• Global Co-occurrence Patterns of Cancer and Cardiovascular Disease: A Comprehensive Analysis Based on the Global Burden of Disease Study 2021 (2026) · doi

  The PAF and SHAP methodology for integrated cancer-cardiovascular disease prevention prioritization has not been applied beyond this specific disease pairing. Validation of this precision population health framework's broader applicability to other non-communicable disease categories and prevention intervention domains requires testing across diverse disease combinations.

• Integrated Advances in Minimally Invasive Surgery: Ergonomics, Visualization, and Artificial Intelligence in Modern Laparoscopy (2026) · doi

  Advanced visualization systems (HD, 3D laparoscopy) show higher penetration than robotic surgery and AI in Latin America, but the paper does not address how visualization quality improvements should be systematically integrated with AI-based phase recognition and safety detection to create cost-effective intelligent surgical ecosystems in resource-limited settings.

• Elevated carbonic anhydrase-1 in the aqueous humor in diabetic macular edema: associations between inflammatory cytokines and retinal vascular dysfunction (2026) · doi

  The sample was recruited exclusively from a tertiary hospital in southwest China with severe and complex conditions, and the relatively small sample size may have insufficient statistical power to detect weaker associations between CA-1 and structural biomarkers (CRT, SRF, CME, EZ disruption, HRF); larger multicentre studies across diverse geographic regions and disease severity ranges are needed to validate the CA-1 correlation patterns.

• Effectiveness comparison of pembrolizumab versus other immune checkpoint inhibitors in neoadjuvant chemoimmunotherapy for esophageal squamous cell carcinoma (2026) · doi

  The pembrolizumab sample size was relatively small in this study; future multi-center prospective research must increase the sample size of pembrolizumab cohorts and involve more medical centers to capture potential prognostic differences between pembrolizumab and domestic checkpoint inhibitors in neoadjuvant chemoimmunotherapy for ESCC.

• Marburg virus disease: the next threat in the making? (2026) · doi

  Safe delivery of intensive care for Marburg virus disease in Rwanda is documented, but the scalability of these clinical management protocols to other endemic regions with varying healthcare infrastructure and resource availability is not addressed.

• Hypoxic burden calculation in patients with obstructive sleep apnea diagnosed by peripheral arterial tonometry: diagnostic accuracy and clinical implications (2026) · doi

  The study used an institutional algorithm for HB calculation based on ≥4% desaturation events from WatchPAT® platform; direct extrapolation of absolute HB values to other pulse oximetry platforms with differing sampling dynamics and technical specifications requires comparative validation studies.

• Device-based therapies in cardio-renal syndrome: a pathophysiology-driven approach to a complex bidirectional disease (2026) · doi

  Operator experience and procedural learning curves significantly influence outcomes for innovative cardio-renal syndrome devices, yet standardized training protocols and competency benchmarks for device insertion and hemodynamic adjustment are not established. The scalability of device-based CRS management beyond specialized centers requires structured operator training frameworks and protocol refinement documentation.