Search

BackgroundObesity is one of today’s most visible public health problems worldwide. Although modern bariatric surgery is ostensibly considered safe, serious complications and mortality still occur in some patients. ObjectiveThis study aimed to explore whether serious postoperative complications of bariatric surgery recorded in a national quality registry can be predicted preoperatively using deep learning methods. MethodsPatients who were registered in the Scandinavian Obesity Surgery Registry (SOReg) between 2010 and 2015 were included in this study. The patients who underwent a bariatric procedure between 2010 and 2014 were used as training data, and those who underwent a bariatric procedure in 2015 were used as test data. Postoperative complications were graded according to the Clavien-Dindo classification, and complications requiring intervention under general anesthesia or resulting in organ failure or death were considered serious. Three supervised deep learning neural networks were applied and compared in our study: multilayer perceptron (MLP), convolutional neural network (CNN), and recurrent neural network (RNN). The synthetic minority oversampling technique (SMOTE) was used to artificially augment the patients with serious complications. The performances of the neural networks were evaluated using accuracy, sensitivity, specificity, Matthews correlation coefficient, and area under the receiver operating characteristic curve. ResultsIn total, 37,811 and 6250 patients were used as the training data and test data, with incidence rates of serious complication of 3.2% (1220/37,811) and 3.0% (188/6250), respectively. When trained using the SMOTE data, the MLP appeared to have a desirable performance, with an area under curve (AUC) of 0.84 (95% CI 0.83-0.85). However, its performance was low for the test data, with an AUC of 0.54 (95% CI 0.53-0.55). The performance of CNN was similar to that of MLP. It generated AUCs of 0.79 (95% CI 0.78-0.80) and 0.57 (95% CI 0.59-0.61) for the SMOTE data and test data, respectively. Compared with the MLP and CNN, the RNN showed worse performance, with AUCs of 0.65 (95% CI 0.64-0.66) and 0.55 (95% CI 0.53-0.57) for the SMOTE data and test data, respectively. ConclusionsMLP and CNN showed improved, but limited, ability for predicting the postoperative serious complications after bariatric surgery in the Scandinavian Obesity Surgery Registry data. However, the overfitting issue is still apparent and needs to be overcome by incorporating intra- and perioperative information.

Background The World Health Organisation (WHO) 2013 diagnostic criteria for gestational diabetes mellitus (GDM) has been criticised due to the limited evidence of benefits on pregnancy outcomes in different populations when switching from previously higher glycemic thresholds to the lower WHO-2013 diagnostic criteria. The aim of this study was to determine whether the switch from previous Swedish (SWE-GDM) to the WHO-2013 GDM criteria in Sweden following risk factor-based screening improves pregnancy outcomes. Methods and findings A stepped wedge cluster randomised trial was performed between January 1 and December 31, 2018 in 11 clusters (17 delivery units) across Sweden, including all pregnancies under care and excluding preexisting diabetes, gastric bypass surgery, or multifetal pregnancies from the analysis. After implementation of uniform clinical and laboratory guidelines, a number of clusters were randomised to intervention (switch to WHO-2013 GDM criteria) each month from February to November 2018. The primary outcome was large for gestational age (LGA, defined as birth weight >90th percentile). Other secondary and prespecified outcomes included maternal and neonatal birth complications. Primary analysis was by modified intention to treat (mITT), excluding 3 clusters that were randomised before study start but were unable to implement the intervention. Prespecified subgroup analysis was undertaken among those discordant for the definition of GDM. Multilevel mixed regression models were used to compare outcome LGA between WHO-2013 and SWE-GDM groups adjusted for clusters, time periods, and potential confounders. Multiple imputation was used for missing potential confounding variables. In the mITT analysis, 47 080 pregnancies were included with 6 882 (14.6%) oral glucose tolerance tests (OGTTs) performed. The GDM prevalence increased from 595/22 797 (2.6%) to 1 591/24 283 (6.6%) after the intervention. In the mITT population, the switch was associated with no change in primary outcome LGA (2 790/24 209 (11.5%) versus 2 584/22 707 (11.4%)) producing an adjusted risk ratio (aRR) of 0.97 (95% confidence interval 0.91 to 1.02, p = 0.26). In the subgroup, the prevalence of LGA was 273/956 (28.8%) before and 278/1 239 (22.5%) after the switch, aRR 0.87 (95% CI 0.75 to 1.01, p = 0.076). No serious events were reported. Potential limitations of this trial are mainly due to the trial design, including failure to adhere to guidelines within and between the clusters and influences of unidentified temporal variations. Conclusions In this study, implementing the WHO-2013 criteria in Sweden with risk factor-based screening did not significantly reduce LGA prevalence defined as birth weight >90th percentile, in the total population, or in the subgroup discordant for the definition of GDM. Future studies are needed to evaluate the effects of treating different glucose thresholds during pregnancy in different populations, with different screening strategies and clinical management guidelines, to optimise women's and children's health in the short and long term. Trial registration The trial is registered with ISRCTN (41918550)., Author(s): Maryam de Brun 1, Anders Magnuson 2, Scott Montgomery 2,3,4, Snehal Patil 2, David Simmons 5,6, Kerstin Berntorp 7, Stefan Jansson 8, Ulla-Britt Wennerholm 9, Anna-Karin Wikström 10, Helen [...]

Purpose: To study risk factors for uncontrolled asthma and insufficient quality of life (QoL) in patients with mild asthma, ie those without preventer treatment. Patients and Methods: Patients aged 18– 75 years with a doctor's diagnosis of asthma randomly selected from primary and secondary care in Sweden. Mild asthma was defined as self-reported current asthma and no preventer treatment. Data were collected from self-completed questionnaires in 2012 and 2015. Well-controlled asthma was defined as Asthma Control Test (ACT) ≥ 20 points and no exacerbation and uncontrolled asthma as ACT< 20 points and/or at least one exacerbation in the previous six months. QoL was measured by the Mini Asthma Quality of Life Questionnaire (Mini-AQLQ), where a total mean score of ≥ 6 indicated sufficient and < 6 insufficient QoL. Multivariate logistic regression analyses were performed using asthma control and Mini-AQLQ as dependent variables. Asthma control was dichotomized as controlled and uncontrolled asthma and the Mini-AQLQ as sufficient QoL (mean score ≥ 6) and insufficient QoL (mean score < 6). Results: Among 298 patients, 26% had uncontrolled asthma, 40% insufficient QoL and 20% both uncontrolled asthma and insufficient QoL. Age ≥ 60 years, obesity, daily smoking, rhinitis and inadequate knowledge of asthma self-management were independently associated with poor asthma control. Factors independently associated with insufficient QoL were age ≥ 60 years, overweight, obesity, rhinitis, sinusitis and inadequate knowledge of asthma self-management. Age ≥ 60 years, obesity, rhinitis and inadequate knowledge of asthma self-management were independently associated with both uncontrolled asthma and insufficient QoL. Conclusion: Among asthma patients without preventer medication, 26% had uncontrolled asthma and 40% had insufficient asthma-related QoL. Older age, obesity, and rhinitis were risk factors for both poor asthma control and a reduced QoL, but having good knowledge of asthma self-management reduced this risk. Our findings suggest that this group of patients requires further attention and follow-up. Plain Language Summary: Many patients with little symptoms of asthma do not take asthma-preventer medication as their asthma is recognized as mild. Still, it is well-known that in this group there are patients with frequent and severe symptoms and acute attacks of asthma, defined as uncontrolled asthma. Quality of life (QoL) is less studied in these patients. Our aim was to study patient characteristics and factors with a higher risk for uncontrolled asthma and insufficient QoL in patients with mild asthma. We studied patients with asthma diagnosis 18– 75 of age in Sweden who reported asthma and no asthma preventer medication. They answered questionnaires about characteristics, knowledge of asthma, asthma symptoms, acute asthma attacks and QoL. A test for asthma symptom control, the Asthma Control Test (ACT), was answered, where 20 points or more meant good asthma symptom control. Patients were regarded as having uncontrolled asthma if they had less than 20 points and/or at least one acute asthma attack the last six months. QoL was measured by the Mini Asthma Quality of Life Questionnaire. A mean score of 6 or more meant sufficient and less than 6 insufficient QoL. Of the 298 patients, 26% had uncontrolled asthma, 40% had insufficient QoL and 20% had both. Patients with risk for both uncontrolled asthma and insufficient QoL were 60 years and over, with obesity, rhinitis and those who reported insufficient knowledge of how to handle asthma. Our results suggest that patients with mild asthma need more attention, better management and follow-ups. [ABSTRACT FROM AUTHOR]

Despite access to effective therapies many asthma patients still do not have well-controlled disease. This is possibly related to underuse of inhaled corticosteroids (ICS) and overuse of short-acting β2-agonists (SABA). Our aim was to investigate longitudinal trends and associated factors in asthma treatment. Two separate cohorts of adults with physician-diagnosed asthma were randomly selected from 14 hospitals and 56 primary health centers in Sweden in 2005 (n = 1182) and 2015 (n = 1225). Information about symptoms, maintenance treatment, and use of rescue medication was collected by questionnaires. Associations between treatment and sex, age, smoking, education, body mass index (BMI), physical activity, allergic asthma, and symptom control were analyzed using Pearson's chi2-test. Odds ratios (ORs) were calculated using logistic regression. Maintenance treatment with ICS together with long-acting β2-agonists (LABA) and/or montelukast increased from 39.2% to 44.2% (p = 0.012). The use of ICS + LABA as-needed increased (11.1–18.9%, p < 0.001), while SABA use decreased (46.4– 41.8%, p = 0.023). Regular treatment with ICS did not change notably (54.2–57.2%, p = 0.14). Older age, former smoking, and poor symptom control were related to treatment with ICS + LABA/montelukast. In 2015, 22.7% reported daily use of SABA. A higher step of maintenance treatment, older age, obesity, shorter education, current smoking, allergic asthma, low or very high physical activity, and a history of exacerbations were associated with daily SABA use. The use of ICS + LABA both for maintenance treatment and symptom relief has increased over time. Despite this, the problem of low use of ICS and high use of SABA remains. [ABSTRACT FROM AUTHOR]

Purpose: Viable cartilage allograft (VCA) is a cartilage tissue matrix that contains cryopreserved viable allogeneic cartilage fibres. This study aimed to assess safety and benefits in treating focal knee cartilage defects with VCA. We hypothesized that VCA is a safe single‐stage procedure in isolated chondral defects. Method: In vitro analysis, in vivo studies and a prospective case series were performed. VCA was evaluated in a goat cartilage repair model. Symptomatic International Cartilage Repair Society grade 3/4A lesions of the femoral condyle or patella were implanted with VCA. International Knee Documentation Committee (IKDC), Knee injury and Osteoarthritis Outcome (KOOS) subscales, Lysholm, Short Form‐12, Visual Analog Scale and pain frequency levels were assessed. Radiographic and magnetic resonance imaging (MRI) was performed at regular intervals postoperatively. Data were analysed by statisticians to determine the power and significance of the results. Results: The goat study confirmed that VCA is effective for cartilage repair. Twenty patients were implanted; the mean age was 28.1 (16–56), the mean body mass index (BMI) was 27.9 ± 5.6 and the mean follow‐up was 24.1 months (range = 12.0–36.0 months). Lesions were in either the femoral condyle (7) or patella (13). Lesion sizes ranged from 1.5 to 6.0 cm2 (mean = 4.58 cm2). Outcome scores improved from preoperative baseline (POB): IKDC (78.2), Lysholm (89.0), KOOS: Pain (95.8), Symptoms (86.3), ADL (87.8), Sports (85.0) and QOL (75.0). MRI imaging demonstrated excellent osteochondral allograft assimilation. Second‐look arthroscopy (two patients) demonstrated complete fill and incorporation (Brittberg scores 11/12). Functional scores were maintained at 24 (M): IKDC (86.24 ± 17.2), Lysholm (87.23 ± 15.0), KOOS: Pain (91.72 ± 17.3), Symptoms (84.92 ± 16.1), ADLs (93.80 ± 16.1), Sports (84.45 ± 27.7), QOL (81.30 ± 20.8). Conclusion: VCA is an off‐the‐shelf, single‐stage, conformable allogeneic graft that treats chondral defects with no additional fixation. Preclinical and short‐term prospective clinical studies show that VCA can safely treat chondral defects with potential advantages to existing options. Level of Evidence: Level IV study. [ABSTRACT FROM AUTHOR]