Jean-Robert Makulo

Sickle cell disease (SCD) is the most common monogenic hemoglobinopathy worldwide, affecting approximately 300,000 newborns annually, with the highest burden in sub-Saharan Africa, India, the Middl...

Sickle cell disease (SCD) is one of the most common haemoglobinopathies worldwide, with nearly 300,000 newborns affected each year, of which about 80% are born in sub-Saharan Africa [1,2]. In the D...

Background and Objectives: On-line hemodiafiltration (OL-HDF) has been proposed as an alternative to conventional hemodialysis (HD) for patients with end-stage chronic kidney disease (CKD). Randomized controlled trials suggest that OL-HDF may reduce mortality, particularly when the convection volume (CV) exceeds 23 L/1.73 m² per session. However, achieving this target depends on local practices and may be limited to selected populations. The CONVINCE trial reported a 97% success rate using a structured optimization protocol, but its applicability to unselected real-world populations remains uncertain. This study aimed to evaluate the incidence of high CV in OL-HDF among unselected patients managed under routine conditions with a standardized optimization protocol. Methods and Materials: This prospective cohort study (May–October 2024) included 67 unselected incident and prevalent patients undergoing HD or HDF in a hospital-based dialysis center. All patients were switched to post-dilution OL-HDF following the CONVINCE optimization protocol, which involved stepwise increases in blood flow, adjustment of filtration fraction, and optimization of session duration. Results: The mean age was 68.8 ± 14.9 years; 56.7% were male. Blood flow increased from 283 to 338 mL/min (p < 0.001), and the use of dialyzers > 2 m² increased from 36% to 68% (p < 0.003). Kt/V improved from 1.22 to 1.6 (p < 0.01). CV increased by ~2 L from M1 onward and was sustained through M6, correlating positively with blood flow, session duration, and Kt/V (all p < 0.01). Conclusions: Stepwise optimization protocol enabled sustained achievement of high CV (23.5 L/session) in 62.3% of patients, improving dialysis adequacy.

Background: Kidney transplantation activity at the University Hospital of Guadeloupe was briefly interrupted at the onset of the COVID-19 pandemic, reflecting the global impact of this health crisis on organ transplantation. This study assessed patient and graft recovery in 335 recipients transplanted between 2013 and 2023, comparing those transplanted before 2020 and after the resumption of activity. The objective was to evaluate changes in recipient profiles, surgical parameters, and post-transplant outcomes following this disruption. Methods: This retrospective cohort included all kidney transplants performed at the University Hospital of Guadeloupe over a ten-year period. Most patients (70%) received transplants before 2020, with 30% afterward. All grafts were ABO-compatible, and 98.2% were from deceased donors. Trends in transplant activity were analyzed to identify variations over time, with a peak observed in 2018, followed by a decline until 2021 and a progressive recovery from 2022. Comparative analyses were performed to examine disparities in donor and recipient characteristics, ischemia durations, and outcomes between the two periods. Results: After 2020, recipients were more likely to be elderly (≥70 years), immunized, obese, have heterozygous sickle cell disease, or have polycystic kidney disease (p < 0.05). Mean cold ischemia time decreased (p = 0.009), while warm ischemia time increased (p < 0.001), reflecting procedural and logistical adaptations. Graft survival remained stable, with 97.5% at 6 months and 89.8% at 4 years for transplants before 2020, versus 100% and 96.9%, respectively, after 2020 (p = 0.160). Patient survival did not differ significantly between periods (p = 0.199). Independent factors associated with mortality included recipient age ≥ 60 years, diabetes, graft failure, transplantation before 2020, cold ischemia time ≥ 1200 min, and graft pyelonephritis. Conclusions: Despite the temporary suspension of activity and an increased proportion of transplants with expanded criteria after 2020, graft recovery and patient survival were not adversely affected. These findings suggest that kidney transplantation in Guadeloupe demonstrated strong resilience and capacity for adaptation during and after the COVID-19 crisis, maintaining outcomes comparable to the pre-pandemic period.

Kidney transplantation activity at the University Hospital of Guadeloupe was briefly interrupted at the onset of the COVID-19 pandemic. This study assessed patient and graft survival in 335 recipients transplanted between 2013 and 2023, comparing those transplanted before 2020 and after the resumption of activity. Most patients (70%) received transplants before 2020, with 30% after-ward. ABO compatibility was observed in 98.2% of grafts from brain-dead do-nors. Peak transplant activity occurred in 2018, followed by a decline until 2021 and recovery from 2022. After 2020, recipients were more likely to be el-derly (≥70 years), immunized, obese, have heterozygous sickle cell disease, or polycystic kidney disease (p < 0.05). Mean cold ischemia time decreased (p = 0.009), while warm ischemia time increased (p < 0.001). Graft survival re-mained stable, with 97.5% at 6 months and 89.8% at 4 years for transplants before 2020, versus 100% and 96.9%, respectively, after 2020 (p = 0.160). Patient survival did not differ between periods (p = 0.199). Independent factors associ-ated with mortality included recipient age ≥ 60 years, diabetes, graft failure, transplantation before 2020, cold ischemia time ≥ 1200 minutes, and graft py-elonephritis. Despite an increased proportion of transplants with expanded criteria after 2020, graft and patient survival were not adversely affected.

Sentiment analysis is vital for evaluating user feedback in drug reviews because understanding patient experiences leads to more personalized treatment recommendations by providing insights into the real-world effectiveness and tolerability of medications, which are often overlooked in clinical trials. This study evaluates the effectiveness of word-level and sentence-level embeddings for feature extraction in sentiment analysis. These embeddings are used in sequential models (Bi-LSTM, CNN) and non-sequential models (Random Forest, DNN, ExtraTreesClassifier). The Random Forest model with LLM2Vec achieves the best performance, with 0.93 accuracy, F1-scores of 0.95 (positive) and 0.88 (negative), and precision scores of 0.93 (positive) and 0.94 (negative). This approach detects subtle negative feedback often missed by standard models. To capture social consensus in patient feedback, we introduce Adaptive Confidence-Weighted Scoring. This method leverages social validation as an implicit confidence signal, enabling sentiment scores to reflect both individual experiences and community agreement. It enhances trust and interpretability while standardizing sentiment polarity on a $[-5, +5]$ scale. Clinically validated drug-related information, scraped from trusted sources, is encoded using the Llama-3.2-3B-Instruct model to extract context-aware representations. Features derived from this external medical knowledge improve semantic understanding, ensure grounding and safety, and enhance robustness even when reviews are sparse or noisy. The final vectors are employed within a cosine similarity framework to recommend relevant drugs, aligning recommendations with user satisfaction. This work demonstrates how LLM-based feature engineering can advance clinically valid, patient-aligned healthcare recommender systems informed by real-world feedback.

Background and Aim: Several non-invasive tests are used to assess liver fibrosis and cirrhosis in patients with liver disease. However, most validation studies have not included populations in sub-Saharan Africa. This study aimed to evaluate the diagnostic performance of the APRI and FIB-4 scores in a Congolese cohort. Patients and Methods: A cohort of patients in Kinshasa underwent FibroScan and laboratory testing to calculate APRI and FIB-4 scores. Pearson correlation, sensitivity, specificity, and ROC curve analyses were used to evaluate the performance of these non-invasive scores against FibroScan. Cirrhosis was defined as liver stiffness ≥ 14 kPa by FibroScan. Thresholds for APRI and FIB-4 scores predicting cirrhosis were set at ≥ 1.5 and ≥ 2.67, respectively. Results: The study included 316 patients (mean ± SD age: 48.1 ± 14.1 years; 60.8% male; 10.1% with diabetes; 37.1% obese; 14.2% with hepatitis B; 6.7% with hepatitis C; 25.6% with a history of alcohol use). The Pearson correlation between APRI and FibroScan was r = 0.210 (p < 0.001), while the correlation between FIB-4 and FibroScan was better (r = 0.478, p < 0.001). In subgroup analyses, FIB-4 correlated with FibroScan only among patients with alcohol use or hepatitis B or C, APRI only correlated with FibroScan in alcohol dependent patients. The sensitivity and specificity of APRI were 29.7% and 97.9% respectively, compared to 60.0% and 93.3% for FibroScan. The areas under the ROC curve were 0.8462 for APRI and 0.8312 for FIB-4, with thresholds lower than those reported in the literature: 0.422 for APRI and 1.285 for FIB-4, but these varied according to the subgroup. Conclusion: APRI and FIB-4 scores demonstrate high specificity but low sensitivity for diagnosing cirrhosis in this population. Their diagnostic performance is notably better in patients with alcohol-related liver disease or viral hepatitis, but poor among those with diabetes or obesity.

Background and Aims: The prevalence of overweight/obesity increases worldwide, including among students. BMI is a commonly used indicator to assess excess fat mass (FM). Our objective was to determine the prevalence of malnutrition (overweight, obesity and underweight) and correlation with FM among Congolese students, and to assess the correlation between BMI and FM measured by bioimpedance. Methods: Cross-sectional study among students applying for registration at the University of Kinshasa during the 2024– 2025 academic year. Underweight, overweight and obesity were defined according to WHO criteria. FM values ≥ percentile 90 were considered as pathological. Student’s T test, Mann Whitney U, Pearson correlation, ROC curves and Youden index were the statistical analyses used, with p < 0.05. Results: We included 13,381 students (59.7% male; mean age 19.8 ± 3.3 years). The prevalence of overweight and obesity in women (13.4%, 95% CI [12.5– 14.3] and 4.6% [4.1– 5.2]) was higher than in men (9.0% [8.4– 9.6] and 2.2% [1.9– 2.5]), p < 0.001. The underweight was also more frequent in women (10.2[9.4– 11.0]) than in men (9.0[8.4– 9.7]), p = 0.021. Women had a higher percentage of FM (median 23.6%, IQ 25: 15.1% and IQ 75: 30.1%) compared to men (13.8%; 10.4% and 19.4%), p < 0.01. The correlation between FM and BMI (r = 0.414; p < 0.001) was better in women compared to the men (r = 0.318, p < 0.001). The areas under the ROC curves plotted from BMI values and FM cut-offs ≥ percentile 90 (women = 35.8% and men = 26.0%), were 0.9038 in women (youden index = 0.73) versus 0.5899 in men (youden index = 0.14). Conclusion: Overweight/obesity and underweight coexist among candidates particularly in women for student recruitment, at the University of Kinshasa, which illustrate the double burden of malnutrition in the academic young population. Using bioimpedance to assess body composition, BMI correlates better with FM in women; the usual thresholds defining overweight and obesity were not valid in men.

Background and Objectives. The chest computed tomography (chest CT) has played an important role in the management of COVID-19. Few data on its use in sub-Saharan Africa (SSA) are reported. The objectives of this study conducted in Kinshasa, DR Congo, were to describe the lung lesions on day 1 of hospitalization in patients admitted for suspected COVID-19 and to identify those that were most associated with SARS-CoV-2 infection/RT-PCR and the determinants of chest CT associated with death. Methods. We included all patients with respiratory symptoms (dyspnea, fever, and cough) and/or respiratory failure admitted to the SOS Médecins de nuit SARL hospital, DR Congo, during the 2nd and 3rd waves of the COVID-19 pandemic. The diagnosis of COVID-19 was established based on RT-PCR anti-SARS-CoV-2 tests (G1 (RT-PCR positive) vs. G2 (RT-PCR negative)), and all patients had a chest CT on the day of admission. We retrieved the digital files of patients, precisely the clinical, biological, and chest CT parameters of the day of admission as well as the vital outcome (survival or death). Chest CT were read by a very high-definition console using Advantage Windows software and exported to the hospital network using the RadiAnt DICOM viewer. To determine the threshold for the percentage of lung lesions associated with all-cause mortality, we used ROC curves. Factors associated with death, including chest CT parameters, were investigated using logistic regression analysis. Results. The study included 200 patients (average age years; 19% diabetics and 4.5% obese), and COVID-19 was confirmed among 56% of them (G1). Chest CT showed that ground glass (72.3 vs. 39.8%), crazy paving (69.6 vs. 17.0%), and consolidation (83.9 vs. 22.7%), with bilateral and peripheral locations (68.8 vs. 30.7%), were more frequent in G1 vs. G2 ( ). No case of pulmonary embolism and fibrosis had been documented. The lung lesions affecting 30% of the parenchyma were informative in predicting death (area under the ROC curve at 0.705, ). In multivariate analysis, a percentage of lesions affecting 50% of the lung parenchyma increased the risk of dying by 7.194 (1.656-31.250). Conclusion. The chest CT demonstrated certain characteristic lesions more frequently in patients in whom the diagnosis of COVID-19 was confirmed. The extent of lesions affecting at least half of the lung parenchyma from the first day of admission to hospital increases the risk of death by a factor of 7.

Background: Data on 25 [OH] vitamin D and intact parathyroid hormone [iPTH] in hemodialysis patients are very limited in sub-Saharan African countries. The present study aimed to assess the magnitude of hypovitaminosis D, and to evaluate the achievement of iPTH KDIGO 2017 targets among chronic hemodialysis patients followed in Kinshasa. Methods: We conducted a multicenter cross-sectional study in 6 hospitals in Kinshasa. All patients followed on hemodialysis for more than 3 months were included. Hypovitaminosis D was defined as < 30 ng/mL (insufficiency = 20– 29 ng/mL; deficiency if < 20 ng/mL) and the targets for iPTH values were based on the 2017 KDIGO guidelines. The determinants for hypovitaminosis D were evaluated by logistic regression. Results: 251 patients [mean age 56 ± 14 years, 72.5% men, 63% hypertensive, 31% diabetic, 100% supplemented with native 25 [OH] vitamin D + CaCO3 were included. Hypovitaminosis D was found in 79.7% (deficiency 47.4%) and was associated with the male gender aOR 2.7 [1.4– 5.2], p = 0.004, the low-permeability dialyzer 2.2 [1.1– 4.2], p = 0.025 and anemia 3.9 [1.2– 12.7], p = 0.022. Only 40% of patients with 25 [OH] vitamin D deficiency had iPTH according to KDIGO targets vs 6% of patients with severe hyperparathyroidism (iPTH > 600 pg/mL), 45% with levels between 16 and 150 pg/mL and 9% a iPTH ≤ 15 pg/mL. Conclusion: Despite a sunny environment, a large proportion of Congolese hemodialysis patients have hypovitaminosis D, in particular a deficiency. Among them, less than half have target iPTH values. These results show the benefit of regular monitoring of these parameters in order to optimize treatment.