Hanoi Medical University (HMU, Vietnamese: Đại học Y Hà Nội) is the oldest university of Vietnam and is located in Hanoi. HMU was founded in 1902 by French during the French colonisation under the name Indochina Medical College. The first headmaster of HMU was Alexandre Yersin who was the co-discoverer of the bacillus responsible for the bubonic plague or pest, which was renamed in his honour (Yersinia pestis)..
People who inject drugs (PWID) living with HIV in Vietnam experience a high prevalence of common mental disorders (CMDs), which negatively impact their health, treatment adherence and overall quality of life. To address the mental health care needs of this vulnerable population, we adapted and piloted Friendship Bench, a problem-solving therapy-based mental health intervention, for people on methadone maintenance treatment (MMT) who live with HIV and have CMDs in 6 MMT clinics in Hanoi, Vietnam, in 2021–2023. This paper aims to explore the experience of implementing aFB and the possibility of aFB expansion from the perspectives of the intervention providers who were directly involved in the delivery of the aFB program. We conducted formal in-person interviews with aFB providers including doctors (n = 6), counselors (n = 10) and supervisors (n = 2)following aFB intervention completion. Collected data was coded using NVivo 12.0 and thematic analysis was applied. All respondents were satisfied with and praised the aFB. Most providers reported that the aFB was beneficial to both overall and mental health of the MMT patients with HIV. All counselors shared that problem-solving therapy in the aFB was new for them; nevertheless, their counseling skills improved significantly after delivering the aFB sessions. All respondents were optimistic about the expansion of the aFB in other clinics. Application of aFB as a mental health intervention for care and treatment for MMT and ART patients could be a promising approach for Vietnam to address mental health care needs for this vulnerable population. Clinical Trial Number: NCT04790201 registered 3/10/2021.
Background The use of 3D-printed Patient-Specific Instruments (PSI) has been investigated to enhance the postoperative functional results in total hip arthroplasty (THA) and has been recognized as an innovative approach for the optimal alignment of hip implant components. Point-of-care production is gradually becoming the norm for PSI manufacturing. The purpose of this article is to assess the accuracy and safety of PSI for total hip arthroplasty performed at the point-of-care in Vietnam. Methods 34 THA cases were assessed in this prospective study. A template for the size and orientation of the implant and the design of the PSI was generated using data from preoperative 3D computed tomography (CT) scanning of the lower limb. The principal surgeon determined the implants’ position and PSI design directly using the software. The PSI is then produced using a 3D-compatible resin printer in our manufacturing hospital. The PSI, consisting of an acetabulum and a femoral component placed press-fit on the bony surface, guided surgeons to precisely ream the acetabulum and cut the femoral neck according to the pre-planned plane. Postoperative CT scanning was obtained and superimposed onto the 3D model of the implant to evaluate the accuracy of the procedure by comparing the orientation values of the cup and the alignment of the stem between the planned and the actual results. Intra- and postoperative clinical parameters of surgery, including surgical time, intra-operative blood loss, complications, and the first ambulation, were also recorded to evaluate the safety of the surgery. Results The preparation for PSI required an average of 3 days. 94% of cup size and 91% of stem size were correctly selected. The mean values of postoperative inclination and anteversion were 44.2° ± 4.1° and 19.2° ± 5.6°, respectively. 64.7% of cases deviated from planned within the ± 5 0 range and 94.1% within the ± 10° range. There was no significant statistical difference between the planned and the achieved values of stem anteversion, osteotomy height, and leg length discrepancy (p > 0.05). The average surgical time was 82.5° ± 10.8 min, and the intraoperative blood loss was estimated at 317.7° ± 57.6 ml. 64.7% of patients could walk on the day of surgery. There were no complications reported. Conclusions The point-of-care manufactured PSI is a useful solution for improving the accuracy of total hip arthroplasty surgery, especially in restoring implant orientation and reducing leg length discrepancy. However, long-term clinical follow-up evaluation is needed to confirm the efficacy and safety of this approach.
Sarcopenia and depression are both common and impactful conditions in older adults. However, there has been limited research on their relationship, particularly in older patients with type 2 diabetes (T2D). To investigate the association between sarcopenia and depressive symptoms among older Vietnamese adults with T2D. From May to December 2024, 630 patients with T2D were recruited from the outpatient clinics of the National Geriatric Hospital in Vietnam, of whom 618 were included in the cross-sectional analysis. Sarcopenia was diagnosed based on the 2019 Asian Working Group for Sarcopenia criteria (AWGS 2019). Depressive symptoms were evaluated using the validated Vietnamese version of the 15-item Geriatric Depression Scale (GDS-15), with scores of 5 or higher indicating the presence of depressive symptoms. Demographic factors, lifestyle behaviours, diabetes-related indicators (HbA1c, duration, complications), and geriatric characteristics were also analysed as potential confounders. Logistic regression was used to assess the association between sarcopenia and depressive symptoms. Among the 618 participants (mean age 73.5 ± 7.1 years, 63.7
OBJECTIVE:Antimicrobial resistance (AMR) is a major challenge in complicated urinary tract infections (cUTIs), particularly among high-risk patients in resource-limited settings. METHODS:This retrospective study (January 2019-December 2023) analyses 622 urine culture-positive samples from cUTI patients at a rehabilitation centre to investigate pathogen distribution, AMR patterns, and associations with clinical outcomes using descriptive statistics and regression models. RESULTS:Escherichia coli (46%) was the predominant isolate, followed by Klebsiella pneumoniae (16%). Although Enterococcus faecalis and Pseudomonas aeruginosa were less common, their prevalence increased over time. High resistance rates to ciprofloxacin and ceftriaxone were observed in E. coli (69.5%, 74.7%) and K. pneumoniae (60.2%, 58.1%). Carbapenem resistance emerged in P. aeruginosa (51.4%) and K. pneumoniae (37.5%), with prior antibiotic exposure identified as a significant predictor (OR: 2.15; 95% CI:1.33-3.47; P = 0.002). In contrast, E. coli and E. faecalis retained high susceptibility to nitrofurantoin and fosfomycin. Multidrug-resistant (MDR) infections were independently associated with prolonged hospital stay (β = 1.889; P = 0.045). CONCLUSIONS:The high prevalence of resistance to first-line agents and emergence of carbapenem resistance underscore the urgent need for antimicrobial stewardship (AMS) interventions. The study suggests specific hospital-setting AMS should be prioritized the judicious use of agents such as nitrofurantoin and fosfomycin, which retain clinical effectiveness. Addressing MDR is crucial as it represents a key modifiable determinant of length of hospital stay in patients with cUTI.
Large vision-language models (VLMs) demonstrate strong performance in medical image understanding, but frequently generate clinically plausible yet incorrect statements, raising significant safety concerns. Existing medical hallucination benchmarks primarily focus on 2D imaging with one-shot diagnostic questions, offering limited insight into whether predictions are grounded in correct localization and abnormality identification, allowing critical reasoning errors to remain hidden behind seemingly correct diagnoses. We introduce Med-StepBench, the first large-scale benchmark for step-wise hallucination detection in 3D oncological PET/CT, comprising over 12,000 images and more than 1,000,000 image–statement pairs across volumetric and multi-view 2D data, which decomposes clinical reasoning into four expert-designed diagnostic stages. Using clinician-verified annotations, we perform the first step-level evaluation of general-purpose and medical VLMs, revealing systematic failure modes obscured by aggregate accuracy metrics. Furthermore, we show that current VLMs are highly susceptible to adversarial yet clinically plausible intermediate explanations, which significantly amplify hallucinations despite contradictory visual evidence. Together, our findings highlight fundamental limitations in grounding multi-step clinical reasoning and establish Med-StepBench as a rigorous benchmark for developing safer and more reliable medical VLMs.
