Dataset Information

Identification of risk factors for patients with diabetes: diabetic polyneuropathy case study.

ABSTRACT: BACKGROUND:Methods of data mining and analytics can be efficiently applied in medicine to develop models that use patient-specific data to predict the development of diabetic polyneuropathy. However, there is room for improvement in the accuracy of predictive models. Existing studies of diabetes polyneuropathy considered a limited number of predictors in one study to enable a comparison of efficiency of different machine learning methods with different predictors to find the most efficient one. The purpose of this study is the implementation of machine learning methods for identifying the risk of diabetes polyneuropathy based on structured electronic medical records collected in databases of medical information systems. METHODS:For the purposes of our study, we developed a structured procedure for predictive modelling, which includes data extraction and preprocessing, model adjustment and performance assessment, selection of the best models and interpretation of results. The dataset contained a total number of 238,590 laboratory records. Each record 27 laboratory tests, age, gender and presence of retinopathy or nephropathy). The records included information about 5846 patients with diabetes. Diagnosis served as a source of information about the target class values for classification. RESULTS:It was discovered that inclusion of two expressions, namely "nephropathy" and "retinopathy" allows to increase the performance, achieving up to 79.82% precision, 81.52% recall, 80.64% F1 score, 82.61% accuracy, and 89.88% AUC using the neural network classifier. Additionally, different models showed different results in terms of interpretation significance: random forest confirmed that the most important risk factor for polyneuropathy is the increased neutrophil level, meaning the presence of inflammation in the body. Linear models showed linear dependencies of the presence of polyneuropathy on blood glucose levels, which is confirmed by the clinical interpretation of the importance of blood glucose control. CONCLUSION:Depending on whether one needs to identify pathophysiological mechanisms for one's prospective study or identify early or late predictors, the choice of model will vary. In comparison with the previous studies, our research makes a comprehensive comparison of different decisions using a large and well-structured dataset applied to different decision support tasks.

SUBMITTER: Metsker O

PROVIDER: S-EPMC7444272 | biostudies-literature | 2020 Aug

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Identification of risk factors for patients with diabetes: diabetic polyneuropathy case study.

Metsker Oleg O Magoev Kirill K Yakovlev Alexey A Yanishevskiy Stanislav S Kopanitsa Georgy G Kovalchuk Sergey S Krzhizhanovskaya Valeria V VV

BMC medical informatics and decision making 20200824 1

<h4>Background</h4>Methods of data mining and analytics can be efficiently applied in medicine to develop models that use patient-specific data to predict the development of diabetic polyneuropathy. However, there is room for improvement in the accuracy of predictive models. Existing studies of diabetes polyneuropathy considered a limited number of predictors in one study to enable a comparison of efficiency of different machine learning methods with different predictors to find the most efficie ...[more]

PMID: 32831065

Similar Datasets

Project description:Diabetic foot is a serious condition in patients with a long lasting diabetes mellitus. Diabetic foot treated improperly may lead not only to delayed ulceration healing, generalized inflammation, unnecessary surgical intervention, but also to the lower limb amputation. The aim of this study was to compare diabetic foot risk factors in population with type 2 diabetes and risk factors for diabetes in healthy subjects.The study included 900 subjects: 145 with diabetic foot, 293 with type 2 diabetes without diabetic foot and 462 healthy controls matched in terms of mean age, gender structure and cardiovascular diseases absence. Study was conducted in Gastroenterology and Metabolic Diseases Department, Medical University of Warsaw, Poland. In statistical analysis a logistic regression model, U Mann-Whitney's and t-Student test were used.The binomial logit models analysis showed that the risk of diabetic foot in patients with type 2 diabetes was decreased by patient's age (odds ratio [OR] = 0.94; 95% confidence interval [CI]: 0.92-0.96; p = 0.00001) and hyperlipidaemia (OR = 0.54; 95% CI: 0.36-0.81; p = 0.01). In contrast, male gender (OR = 2.83; 95% CI: 1.86-4.28; p = 0.00001) diabetes duration (OR = 1.04; 95% CI: 1.03-1.06; p = 0.0003), weight (OR = 1.04; 95% CI: 1.03-1.06; p = 0.00001), height (OR = 1.08; 95% CI: 1.05-1.11; p = 0.00001) and waist circumference (OR = 1.028; 95% CI: 1.007-1.050; p = 0.006) increase the risk of diabetic foot. The onset of type 2 diabetes in healthy subjects was increased by weight (OR = 1.035; 95% CI: 1.024-1.046; p = 0.00001), WC (OR = 1.075; 95% CI: 1.055-1.096; p = 00001), hip circumference (OR = 1.03; 95% CI: 1.01-1.05; p = 0.005), overweight defined with body mass index (BMI) above 24,9 kg/m(2) (OR = 2.49; 95% CI: 1.77-3.51; p = 0.00001) and hyperlipidaemia (OR = 3.53; 95% CI: 2.57-4.84; p = 0.00001).Risk factors for Type 2 diabetes and diabetic foot are only partially common. Study proved that patients who are prone to developing diabetic foot experience different risk factors than patients who are at risk of diabetes. Identification of relationship between diabetic foot and diabetes risk factors in appropriate groups may help clinicians to focus on certain factors in diabetic foot prevention.

Project description:AimsThe aim of the present study was to investigate the major determinants of diabetic peripheral neuropathy (DPN) in patients with type 2 diabetes (T2D), considering the traditional and newly discovered risk factors, including hypoglycaemia and glycemic variability.MethodsThis retrospective case-control study was conducted in a tertiary care hospital in Taiwan. A total of 2,837 patients with T2D were recruited, medical history and biochemical data were obtained, and patients were screened for DPN using the Michigan Neuropathy Screening Instrument (MNSI). DPN was defined as an MNSI exam score > 2. A stepwise selection of variables was used based on the Akaike Information Criterion (AIC) and the Schwarz Criterion (SC). Multivariate analysis was performed using the identified variables obtained from the stepwise selection.ResultsAmong the recruited patients, 604 (21.3%) were found to have DPN. 275 patients with DPN were selected because of longer follow up period before enrollment and complete data of glycemic parameters, and paired with 351 patients with T2D without DPN and matched for age, gender, and diabetes duration. The results of the stepwise selection showed that the presence of moderately and severely increased albuminuria yielded the lowest values of AIC and SC, which indicate the best predictive performance. Multivariate analysis demonstrated that moderately and severely increased albuminuria and greater long-term glycemic variability significantly increased the risk of DPN, with a corresponding odds ratio of 1.85 and 1.61 (95%confidence intervals of 1.25-2.73and1.02-2.55, respectively), after adjusted for hypoglycaemia and types of diabetes treatment.ConclusionsAlbuminuria is a potent predictor of DPN, and greater long-term glycemic variabilityis clearly associated with DPN in adults with T2D. These findings indicate that, in addition to achieve average blood glucose control, screening for albuminuria and reducing blood glucose fluctuations might be useful for improving diabetic microvascular complications.

Project description:BackgroundDespite the high prevalence of depression and anxiety in chronic pain conditions, current knowledge concerning emotional distress among painful diabetic polyneuropathy (pDSPN) and other diabetes mellitus (DM) sufferers is limited.MethodsThis observational multicentre cohort study employed the Hospital Anxiety and Depression Scale, the Beck Depression Inventory II and the State-Trait Anxiety Inventory to assess symptoms of depression and anxiety in several groups with diabetes, as well as in a control group. The study cohort included 347 pDSPN patients aged 63.4 years (median), 55.9% males; 311 pain-free diabetic polyneuropathy (nDSPN) patients aged 63.7 years, 57.9% males; 50 diabetes mellitus (DM) patients without polyneuropathy aged 61.5 years, 44.0% males; and 71 healthy controls (HC) aged 63.0 years, 42.3% males. The roles played in emotional distress were explored in terms of the biological, the clinical (diabetes-, neuropathy- and pain-related), the socio-economic and the cognitive factors (catastrophizing).ResultsThe study disclosed a significantly higher prevalence of the symptoms of depression and anxiety not only in pDSPN (46.7% and 60.7%, respectively), but also in patients with nDSPN (24.4% and 44.4%) and DM without polyneuropathy (22.0% and 30.0%) compared with HCs (7.0% and 14.1%, p < 0.001). Multiple regression analysis demonstrated the severity of pain and neuropathy, catastrophic thinking, type 2 DM, lower age and female sex as independent contributors to depression and anxiety.ConclusionsIn addition to the severity of neuropathic pain and its cognitive processing, the severity of diabetic polyneuropathy and demographic factors are key independent contributors to emotional distress in diabetic individuals.SignificanceIn large cohorts of well-defined painless and painful diabetic polyneuropathy patients and diabetic subjects without polyneuropathy, we found a high prevalence of the symptoms of depression and anxiety, mainly in painful individuals. We have confirmed neuropathic pain, its severity and cognitive processing (pain catastrophizing) as dominant risk factors for depression and anxiety. Furthermore, some demographic factors (lower age, female sex), type 2 diabetes mellitus and severity of diabetic polyneuropathy were newly identified as important contributors to emotional distress independent of pain.

Project description:BackgroundThis study aimed to determine the prevalence of genetic and environmental vascular risk factors in non diabetic patients with premature peripheral arterial disease, either peripheral arterial occlusive disease or thromboangiitis obliterans, the two main entities of peripheral arterial disease, and to established whether some of them are specifically associated with one or another of the premature peripheral arterial disease subgroups.Methods and resultsThis study included 113 non diabetic patients with premature peripheral arterial disease (diagnosis <45-year old) presenting either a peripheral arterial occlusive disease (N?=?64) or a thromboangiitis obliterans (N?=?49), and 241 controls matched for age and gender. Both patient groups demonstrated common traits including cigarette smoking, low physical activity, decreased levels of HDL-cholesterol, apolipoprotein A-I, pyridoxal 5'-phosphate (active form of B6 vitamin) and zinc. Premature peripheral arterial occlusive disease was characterized by the presence of a family history of peripheral arterial and carotid artery diseases (OR 2.3 and 5.8 respectively, 95% CI), high lipoprotein (a) levels above 300 mg/L (OR 2.3, 95% CI), the presence of the factor V Leiden (OR 5.1, 95% CI) and the glycoprotein Ia(807T,837T,873A) allele (OR 2.3, 95% CI). In thromboangiitis obliterans group, more patients were regular consumers of cannabis (OR 3.5, 95% CI) and higher levels in plasma copper has been shown (OR 6.5, 95% CI).ConclusionsAccording to our results from a non exhaustive list of study parameters, we might hypothesize for 1) a genetic basis for premature peripheral arterial occlusive disease development and 2) the prevalence of environmental factors in the development of thromboangiitis obliterans (tobacco and cannabis). Moreover, for the first time, we demonstrated that the 807T/837T/873A allele of platelet glycoprotein Ia may confer an additional risk for development of peripheral atherosclerosis in premature peripheral arterial occlusive disease.

Dataset Information

Identification of risk factors for patients with diabetes: diabetic polyneuropathy case study.

Publications

Identification of risk factors for patients with diabetes: diabetic polyneuropathy case study.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets