Nevin Manimala

Dig Dis Sci. 2024 Jun 28. doi: 10.1007/s10620-024-08538-y. Online ahead of print.

ABSTRACT

BACKGROUND AND AIMS: Drainage of pancreatic fluid collections (PFCs) is required in select cases including infected or symptomatic collections. In this network meta-analysis, we have compared lumen-apposing metal stents (LAMS), fully covered self-expandable metal stents (FCSEMS), and double-pigtail stents (DPS) to identify the most useful stent type in the management of PFCs.

METHODS: We reviewed several databases to identify studies that compared DPS or FCSEMS with LAMS and the ones which compared DPS with FCSEMS for the treatment of PFCs. Our outcomes of interest were clinical success, adverse events, technical success, recurrence of PFCs, and procedure duration. Random effects model and frequentist approach were used for statistical analysis.

RESULTS: We included 28 studies with 2974 patients. Rate of clinical success was significantly lower with DPS compared to LAMS, OR (95% CI): 0.43 (0.32, 0.59). Rate of recurrence was higher with DPS compared to LAMS, OR (95% CI): 2.06 (1.19, 3.57). We found no significant difference in rate of adverse events between groups. Rate of technical success was higher for FCSEMS compared to LAMS. Procedure duration was significantly shorter for LAMS compared to DPS and FCSEMS. Based on frequentist approach, LAMS was found to be superior to DPS and FCSEMS in achieving higher clinical success, lower rate of adverse events and recurrence, and shorter procedure time.

CONCLUSIONS: This network meta-analysis demonstrates the superiority of LAMS over DPS and FCSEMS in the management of PFCs in achieving a higher clinical success, shorter procedure time, and lower rate of recurrence. Some of the analyses are not adequately powered to make firm conclusions, and future large multicenter RCTs are required to further evaluate this issue.

PMID:38940974 | DOI:10.1007/s10620-024-08538-y

Dig Dis Sci. 2024 Jun 28. doi: 10.1007/s10620-024-08526-2. Online ahead of print.

ABSTRACT

BACKGROUND: Malnutrition in cirrhosis is associated with poor outcomes, leading to guidelines for a high protein, low sodium diet; however, there is no guidance regarding the implementation of diet education in clinical practice.

METHODS: A mixed methods study enrolled 21 patients with cirrhosis and their caregivers. Semi-structured interviews on barriers and facilitators of dietary education and adherence were conducted. Demographic and clinical data were obtained, along with quantitative measures of dietary adherence, including 24-h food recall and spot urine sodium. Combined deductive and inductive coding was used to identify qualitative themes, along with a quantitative assessment of interviews. Quantitative data was reported using descriptive statistics with frequencies, mean and confidence intervals.

RESULTS: Participants were mostly male (16/21) with a mean age 57.8 years (SE 2.8) and MELD-Na 9 (SE 1.2). 4 themes emerged: 1. More than 50% of participants and caregivers endorsed no or inadequate diet education 2. They reported mostly negative experiences with dietary adherence with largest impact on social life 3. Facilitators of adherence included the presence of household support and fear of complications of cirrhosis 4. Overwhelmingly desired non-generic handouts and information. Dietary adherence was poor with only one participant meeting protein and sodium requirements based on food recall. Four participants who adhered to < 2000 mg sodium had inadequate daily caloric intake.

CONCLUSIONS: Dietary education is inadequate, and adherence to dietary recommendations is poor in patients with cirrhosis. Future studies should use these barriers and facilitators for intervention development.

PMID:38940972 | DOI:10.1007/s10620-024-08526-2

Int Ophthalmol. 2024 Jun 28;44(1):292. doi: 10.1007/s10792-024-03176-9.

ABSTRACT

PURPOSE: Compared to Shack-Hartmann wavefront sensor (SHWS), the parameters of virtual SHWS (vSHWS) can be easily adjusted to obtain the optimal performance of aberration measurement. Its current optimal parameters are obtained with only a set of statistical aberrations and not statistically significant. Whether the above parameters are consistent with the statistical results of the optimal parameters corresponding to each set of aberrations, and which performance is better if not? The purpose of this study was to answer these questions.

METHODS: The optimal parameters to reconstruct 624 sets of clinical ocular aberrations in the highest accuracy, including the numbers of sub-apertures (NSAs) and the expansion ratios (ERs) of electric field zero-padding, were determined sequentially in this work. By using wavefront-reconstruction accuracy as an evaluation index, the statistical optimal parameter configuration was selected from some possible configurations determined by the optimal NSAs and ERs.

RESULTS: The statistical optimal parameters are consistent for normal and abnormal eyes. They are different from the optimal parameters obtained with a set of statistical aberrations from the same 624 sets of aberrations, and the performance using the former is better than that using the latter. The performance using a fixed set of statistical optimal parameters is even close to that using the respective optimal parameters corresponding to each set of aberrations.

CONCLUSION: The vSHWS configured with a fixed set of statistical optimal parameters can be used for high-precision aberration measurement of both normal and abnormal eyes. The statistical optimal parameters are more suitable for vSHWS than the parameters obtained with a set of statistical aberrations. These conclusions are significant for the designs of vSHWS and also SHWS.

PMID:38940969 | DOI:10.1007/s10792-024-03176-9

Int Ophthalmol. 2024 Jun 28;44(1):293. doi: 10.1007/s10792-024-03205-7.

ABSTRACT

BACKGROUND: This prospective clinical study evaluates the effect of a silicone stent tube (SST) on the success rate of endonasal-endoscopic dacryocystorhinostomy (EN-DCR) to treat primary acquired nasolacrimal duct obstruction.

METHODS: Patients were randomly assigned to receive EN-DCR with or without SST intubation over a period of 3 months. The surgery was performed using standardized techniques. Patients were assessed at three different timepoints: one day, 12 weeks and 24 weeks after the surgery. The results were compared in order to evaluate statistical differences. Surgical success was determined by means of positive irrigation procedures, as well as by the improvement of symptoms and a high level of patient satisfaction.

RESULTS: A total of 56 randomized cases completed 24 weeks of follow up. 1 Patient dropped out due to malignant genesis of the nasolacrimal duct obstruction. After 24 weeks of follow up no statistically significant differences in levels of epiphora (p > .10) or patency (p > .16) were revealed. Comparisons regarding changes in time did not show levels of significance (p > .28).

CONCLUSIONS: This study could not confirm a statistically significant benefit or disadvantage for SST Insertion in EN-DCR.

PMID:38940962 | DOI:10.1007/s10792-024-03205-7

Eur J Trauma Emerg Surg. 2024 Jun 28. doi: 10.1007/s00068-024-02577-w. Online ahead of print.

ABSTRACT

PURPOSE: The implementation of enhanced recovery after surgery programs (ERPs) has significantly improved outcomes within various surgical specialties. However, the suitability of ERPs in trauma surgery remains unclear. This study aimed to (1) design and implement an ERP for trauma laparotomy patients; (2) assess its safety, feasibility, and efficacy; and (3) compare the outcomes of the proposed ERP with conventional practices.

METHODS: This case-matched study prospectively enrolled hemodynamically stable patients undergoing emergency laparotomy after penetrating trauma. Patients receiving the proposed ERP were compared to historical controls who had received conventional treatment from two to eight years prior to protocol implementation. Cases were matched for age, sex, injury mechanism, extra-abdominal injuries, and trauma scores. Assessment of intervention effects were modelled using regression analysis for outcome measures, including length of hospital stay (LOS), postoperative complications, and functional recovery parameters.

RESULTS: Thirty-six consecutive patients were enrolled in the proposed ERP and matched to their 36 historical counterparts, totaling 72 participants. A statistically significant decrease in LOS, representing a 39% improvement in average LOS was observed. There was no difference in the incidence of postoperative complications. Opioid consumption was considerably lower in the ERP group (p < 0.010). Time to resumption of oral liquid and solid intake, as well as to the removal of nasogastric tubes, urinary catheters, and abdominal drains was significantly earlier among ERP patients (p < 0.001).

CONCLUSION: The implementation of a standardized ERP for the perioperative care of penetrating abdominal trauma patients yielded a significant reduction in LOS without increasing postoperative complications. These findings demonstrate that ERPs principles can be safely applied to selected trauma patients.

PMID:38940950 | DOI:10.1007/s00068-024-02577-w

Eur J Trauma Emerg Surg. 2024 Jun 28. doi: 10.1007/s00068-024-02587-8. Online ahead of print.

ABSTRACT

PURPOSE: The objective was to analyze the treatment and complications of the patients after a major amputation of the upper and lower extremities. Risk factors and predictors of a prolonged hospital stay should be outlined.

METHODS: This is a retrospective study of a national Level-1 Trauma center in Germany. In a 10-year period, patients were identified by major amputations in the upper and lower extremities. The medical reports were considered and the results were split into four main groups with analysis on basic-, clinical data, the course on intensive care unit and the outcome. A recovery index was established. The patients’ degree of recovery was summed up. Statistical analysis was performed.

RESULTS: 81 patients were included. A total of 39 (48.1%) major amputations were carried out on the lower leg and 34 (42.0%) involved the thigh. There were two instances (2.5%) of hip joint disarticulation. 6 major amputations were done on the upper extremities (n = 3 on the upper arm, n = 3 on the forearm). 13.83 ± 17.10 days elapsed between hospital admission and major amputation. The average length of hospital stay was 38.49 ± 26,75 days with 5.06 ± 11.27 days on intensive care unit. Most of the patients were discharged home followed by rehabilitation. A significant correlation was found between the hospital length of stay and the increasing number of operations performed (p = 0.001). The correlation between the hospital length of stay and the CRP level after amputation was significant (p = 0.003).

CONCLUSIONS: Major amputations in trauma patients lead to a prolonged stay in hospital due to severe diseases and complications. Especially infections and surgical revisions cause such lengthenings.

PMID:38940948 | DOI:10.1007/s00068-024-02587-8

Skeletal Radiol. 2024 Jun 28. doi: 10.1007/s00256-024-04735-y. Online ahead of print.

ABSTRACT

OBJECTIVE: Fatty infiltration of skeletal muscle (Myosteatosis) is associated with increased frailty, decreased muscle and mobility function, which seems fairly prevalent in multiple myeloma (MM) patients. This study aimed to determine the prognostic value of myosteatosis assessed by CT for progression-free survival (PFS) and overall survival (OS).

MATERIALS AND METHODS: This IRB-approved cohort study included patients with newly diagnosed MM who were treated at a single university hospital and received CT at baseline. Geriatric assessment was performed via International Myeloma Working Group frailty score and Revised Myeloma Comorbidity Index. Myosteatosis was determined through measurement of paravertebral muscle radiodensity. Statistical analyses included uni- and multivariable Cox proportional hazard models and the Kaplan-Meier-method.

RESULTS: A total of 226 newly diagnosed MM patients (median age: 65 years [range: 29-89], 63% males, mean BMI: 25 [14-42]) were analyzed. The prevalence of myosteatosis was 51%. Muscle radiodensity was significantly decreased in individuals with International Staging System stage III vs. I (p < 0.001), indicating higher fatty muscle infiltration in patients with advanced disease. Both PFS and OS were significantly decreased in patients with myosteatosis (PFS: median 32.0 months (95% CI 20.5.5-42.2) vs. 66.4 months without myosteatosis (95% CI 42.5-not reached), p < .001); OS: median 58.6 (95% CI 51.3-90.2) vs. not reached, p < .001). Myosteatosis remained an independent predictor of OS in multivariable analyses (HR: 1.98; 95%-CI: 1.20-3.27).

CONCLUSION: Myosteatosis seems fairly prevalent in patients with newly diagnosed MM and associated with impaired overall survival. Prospective clinical trials are required to better understand the role of myosteatosis in MM patients.

PMID:38940940 | DOI:10.1007/s00256-024-04735-y

Bioinformatics. 2024 Jun 28;40(Supplement_1):i501-i510. doi: 10.1093/bioinformatics/btae251.

ABSTRACT

MOTIVATION: In many biomedical applications, we are confronted with paired groups of samples, such as treated versus control. The aim is to detect discriminating features, i.e. biomarkers, based on high-dimensional (omics-) data. This problem can be phrased more generally as a two-sample problem requiring statistical significance testing to establish differences, and interpretations to identify distinguishing features. The multivariate maximum mean discrepancy (MMD) test quantifies group-level differences, whereas statistically significantly associated features are usually found by univariate feature selection. Currently, few general-purpose methods simultaneously perform multivariate feature selection and two-sample testing.

RESULTS: We introduce a sparse, interpretable, and optimized MMD test (SpInOpt-MMD) that enables two-sample testing and feature selection in the same experiment. SpInOpt-MMD is a versatile method and we demonstrate its application to a variety of synthetic and real-world data types including images, gene expression measurements, and text data. SpInOpt-MMD is effective in identifying relevant features in small sample sizes and outperforms other feature selection methods such as SHapley Additive exPlanations and univariate association analysis in several experiments.

AVAILABILITY AND IMPLEMENTATION: The code and links to our public data are available at https://github.com/BorgwardtLab/spinoptmmd.

PMID:38940158 | DOI:10.1093/bioinformatics/btae251

Bioinformatics. 2024 Jun 28;40(Supplement_1):i110-i118. doi: 10.1093/bioinformatics/btae239.

ABSTRACT

Artificial intelligence (AI) is increasingly used in genomics research and practice, and generative AI has garnered significant recent attention. In clinical applications of generative AI, aspects of the underlying datasets can impact results, and confounders should be studied and mitigated. One example involves the facial expressions of people with genetic conditions. Stereotypically, Williams (WS) and Angelman (AS) syndromes are associated with a “happy” demeanor, including a smiling expression. Clinical geneticists may be more likely to identify these conditions in images of smiling individuals. To study the impact of facial expression, we analyzed publicly available facial images of approximately 3500 individuals with genetic conditions. Using a deep learning (DL) image classifier, we found that WS and AS images with non-smiling expressions had significantly lower prediction probabilities for the correct syndrome labels than those with smiling expressions. This was not seen for 22q11.2 deletion and Noonan syndromes, which are not associated with a smiling expression. To further explore the effect of facial expressions, we computationally altered the facial expressions for these images. We trained HyperStyle, a GAN-inversion technique compatible with StyleGAN2, to determine the vector representations of our images. Then, following the concept of InterfaceGAN, we edited these vectors to recreate the original images in a phenotypically accurate way but with a different facial expression. Through online surveys and an eye-tracking experiment, we examined how altered facial expressions affect the performance of human experts. We overall found that facial expression is associated with diagnostic accuracy variably in different genetic conditions.

PMID:38940144 | DOI:10.1093/bioinformatics/btae239

Bioinformatics. 2024 Jun 28;40(Supplement_1):i266-i276. doi: 10.1093/bioinformatics/btae230.

ABSTRACT

SUMMARY: Pretrained large language models (LLMs) have significantly improved code generation. As these models scale up, there is an increasing need for the output to handle more intricate tasks and to be appropriately specialized to particular domains. Here, we target bioinformatics due to the amount of domain knowledge, algorithms, and data operations this discipline requires. We present BioCoder, a benchmark developed to evaluate LLMs in generating bioinformatics-specific code. BioCoder spans much of the field, covering cross-file dependencies, class declarations, and global variables. It incorporates 1026 Python functions and 1243 Java methods extracted from GitHub, along with 253 examples from the Rosalind Project, all pertaining to bioinformatics. Using topic modeling, we show that the overall coverage of the included code is representative of the full spectrum of bioinformatics calculations. BioCoder incorporates a fuzz-testing framework for evaluation. We have applied it to evaluate various models including InCoder, CodeGen, CodeGen2, SantaCoder, StarCoder, StarCoder+, InstructCodeT5+, GPT-3.5, and GPT-4. Furthermore, we fine-tuned one model (StarCoder), demonstrating that our training dataset can enhance the performance on our testing benchmark (by >15% in terms of Pass@K under certain prompt configurations and always >3%). The results highlight two key aspects of successful models: (i) Successful models accommodate a long prompt (>2600 tokens) with full context, including functional dependencies. (ii) They contain domain-specific knowledge of bioinformatics, beyond just general coding capability. This is evident from the performance gain of GPT-3.5/4 compared to the smaller models on our benchmark (50% versus up to 25%).

AVAILABILITY AND IMPLEMENTATION: All datasets, benchmark, Docker images, and scripts required for testing are available at: https://github.com/gersteinlab/biocoder and https://biocoder-benchmark.github.io/.

PMID:38940140 | DOI:10.1093/bioinformatics/btae230