Shah, Raghav

ABSTRACT Background Acute kidney injury (AKI) in hospitalised children is a major complication associated with significant morbidity and mortality. The integration of artificial intelligence (AI)/machine learning (ML) models may enable early detection and risk stratification. This systematic review evaluates the performance of AI/ML models for predicting paediatric AKI across clinical settings. Methods We systematically searched PubMed, Embase, and Web of Science for studies applying AI/ML models to predict AKI in paediatric populations. Studies reporting performance metrics such as the area under the curve (AUC), sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), accuracy, and F1 score were included. Results Among 470 records identified, 11 studies met the inclusion criteria, with 14 AI/ML models used. The overall sample size included 33 949 paediatric patients with an AKI proportion of 12.5%. Meta-analyses of the AUC were conducted on neural network, gradient boosting, and logistic regression. Gradient boosting had the highest pooled AUC of 0.873 (95% confidence interval 0.836–0.909). Random forest demonstrated the highest median sensitivity (0.821), specificity (0.942), PPV (0.860), NPV (0.935), and accuracy (0.821); however, these metrics could not be pooled due to inconsistent reporting and limited validation. Conclusion Gradient boosting, random forest, and logistic regression demonstrated reasonable predictive performance for paediatric AKI prediction within specific clinical contexts. However, small sample size, heterogeneity, lack of testing/validation cohorts, insufficient data, and inconsistent patient populations and AKI diagnostic criteria restrict generalisability.

Background: The use of spinal cord stimulation (SCS) for managing severe refractory chronic pain has expanded considerably due to positive statistical evidence regarding its use; however, the statistical robustness of the underlying randomized controlled trials (RCTs) requires further scrutiny. One such tool that can be used for this purpose is the fragility index, which quantifies how many individual outcome events must be altered for an outcome to lose statistical significance. Thus the index can be used quantitatively to assess the stability and robustness of an RCT’s conclusions, with higher values indicating increased trial stability. This study assesses the fragility of pain outcomes across RCTs investigating SCS for chronic pain to better understand the quality and robustness of evidence. Methods: A systematic search was conducted for RCTs assessing SCS for any chronic pain indication. The primary outcome was an evaluation of the trial-specific fragility index for the prespecified pain primary outcomes of RCTs. Secondary outcomes included an evaluation of fragility for (1) specified indications for SCS therapy, (2) reported pain outcomes appearing in three or more RCTs, (3) the presence/absence of a conflict of interest, and (4) comparisons of SCS to conservative management or different SCS waveform modalities. Results: A total of 30 RCTs were included. The median (interquartile range [IQR]) fragility index across the primary outcome of all trials was 5.45 (3.00 to 11.45). There was no statistical difference between the (1) types of outcomes (dichotomous vs. continuous; P = 0.710), (2) primary versus secondary pain outcomes ( P = 0.771), or (3) presence versus absence of trial conflict of interest ( P = 0.753). Indications with a median (IQR) fragility score greater than three included persistent spinal pain syndrome type 2 with a score of 8.00 (2.80 to 12.60), painful diabetic neuropathy with a score of 6.00 (3.00 to 11.80), complex regional pain syndrome with a score of 7.20 (4.62 to 81.50), mixed etiologies with a score of 9.00 (3.00 to 38.00), and other etiologies with a score of 3.00 (1.00 to 8.55). Conclusions: This study suggests that the majority of RCTs investigating primary pain outcomes after SCS therapy are robust with relatively high fragility scores. Reporting the fragility of outcomes in trials can provide a more comprehensive assessment of trial robustness and can further aid clinicians in interpreting trial results and making informed treatment decisions.

AbstractBackgroundThe SARS‐CoV‐2 pandemic and corresponding acute respiratory syndrome have affected all populations and led to millions of deaths worldwide. The pandemic disproportionately affected immunocompromised and immunosuppressed adult patients who had received solid organ transplants (SOTs). With the onset of the pandemic, transplant societies across the world recommended reducing SOT activities to avoid exposing immunosuppressed recipients. Due to the risk of COVID‐19‐related outcomes, SOT providers adapted the way they deliver care to their patients, leading to a reliance on telehealth. Telehealth has helped organ transplant programs continue treatment regimens while protecting patients and physicians from COVID‐19 transmission. This review highlights the adverse effects of COVID‐19 on transplant activities and summarizes the increased role of telehealth in the management of solid organ transplant recipients (SOTRs) in both pediatric and adult populations.MethodsA comprehensive systematic review and meta‐analysis were conducted to accentuate the outcomes of COVID‐19 and analyze the efficacy of telehealth on transplant activities. This in‐depth examination summarizes extensive data on the clinical detriments of COVID‐19 in transplant recipients, advantages, disadvantages, patient/physician perspectives, and effectiveness in transplant treatment plans via telehealth.ResultsCOVID‐19 has caused an increase in mortality, morbidity, hospitalization, and ICU admission in SOTRs. Telehealth efficacy and benefits to both patients and physicians have increasingly been reported.ConclusionsDeveloping effective systems of telehealth delivery has become a top priority for healthcare providers during the COVID‐19 pandemic. Further research is necessary to validate the effectiveness of telehealth in other settings.