A time series driven model for early sepsis prediction based on transformer module

Sepsis remains a critical concern in intensive care units due to its high mortality rate. Early identification and intervention are paramount to improving patient outcomes. In this study, we have proposed predictive models for early sepsis prediction based on time-series data, utilizing both CNN-Transformer and LSTM-Transformer architectures. By collecting time-series data from patients at 4, 8, and 12 h prior to sepsis diagnosis and subjecting it to various network models for analysis and comparison. In contrast to traditional recurrent neural networks, our model exhibited a substantial improvement of approximately 20%. On average, our model demonstrated an accuracy of 0.964 (± 0.018), a precision of 0.956 (± 0.012), a recall of 0.967 (± 0.012), and an F1 score of 0.959 (± 0.014). Furthermore, by adjusting the time window, it was observed that the Transformer-based model demonstrated exceptional predictive capabilities, particularly within the earlier time window (i.e., 12 h before onset), thus holding significant promise for early clinical diagnosis and intervention. Besides, we employed the SHAP algorithm to visualize the weight distribution of different features, enhancing the interpretability of our model and facilitating early clinical diagnosis and intervention.

Specific serum autoantibodies predict the development and progression of Alzheimer's disease with high accuracy

Autoimmunity plays a key role in the pathogenesis of Alzheimer's disease (AD). However, whether autoantibodies in peripheral blood can be used as biomarkers for AD has been elusive. Serum samples were obtained from 1,686 participants, including 767 with AD, 146 with mild cognitive impairment (MCI), 255 with other neurodegenerative diseases, and 518 healthy controls. Specific autoantibodies were measured using a custom-made immunoassay. Multivariate support vector machine models were employed to investigate the correlation between serum autoantibody levels and disease states. As a result, seven candidate AD-specific autoantibodies were identified, including MAPT, DNAJC8, KDM4D, SERF1A, CDKN1A, AGER, and ASXL1. A classification model with high accuracy (area under the curve (AUC) = 0.94) was established. Importantly, these autoantibodies could distinguish AD from other neurodegenerative diseases and out-performed amyloid and tau protein concentrations in cerebrospinal fluid in predicting cognitive decline (P < 0.001). This study indicated that AD onset and progression are possibly accompanied by an unappreciated serum autoantibody response. Therefore, future studies could optimize its application as a convenient biomarker for the early detection of AD.

An accurate prediction of negative lymph node metastasis with consideration of glucose metabolism in early-stage non-small cell lung cancer

OBJECTIVE

We aimed to identify risk factors in lymph node metastasis in early-stage non-small cell lung cancer (NSCLC) and predict lymph node metastasis.

METHODS

A total of 416 patients with clinical stage IA2-3 NSCLC who underwent lobectomy and lymph node dissection between July 2016 and December 2020 at National Cancer Center Hospital East were included. Multivariable logistic regression was performed to develop a model for predicting lymph node metastasis. Leave-one-out cross-validation was performed to evaluate the developing prediction model, and sensitivity, specificity, and concordance statistics were calculated to evaluate its diagnostic performance.

RESULTS

The formula for calculating the probability of pathological lymph node metastasis included SUVmax of the primary tumor and serum CEA level. The concordance statistics was 0.7452. When the cutoff value associated with the risk of incorrectly predicting pathological lymph node metastasis was 7.2%, the diagnostic sensitivity and specificity for predicting metastasis were 96.4% and 38.6%, respectively.

CONCLUSIONS

We created a prediction model for lymph node metastasis in NSCLC by combining the SUVmax of the primary tumor and serum CEA levels, which showed a particularly strong association. This model is clinically useful as it successfully predicts negative lymph node metastasis in patients with clinical stage IA2-3 NSCLC.

A Short-Form Measure of Loneliness to Predict Depression Symptoms Among Adolescents

The purpose of this study was to produce a short-form measure of loneliness and assesses its prediction of depressive symptoms relative to a comprehensive measure. Western Australian adolescents completed the Friendship Related Loneliness and Isolation subscales of the Perth Aloneness Scale (PALs) three times over 18 months (T 1 n = 1538; T 2, n = 1683; T 3, n = 1406). Items were reduced while preserving predictability. Follow-up confirmatory factor analyses and predictive models with the reduced and full PALs were then tested. A reduced six-item scale (PALs-6) preserved the two-factor structure of the PALs and showed strong prediction of very elevated depressive symptoms (Sensitivity = 0.70, Specificity = 0.78, AUC = 0.81); it was less successful in predicting future symptoms (Sensitivity = 0.67, Specificity = 0.64, AUC = 0.74). The PALs-6 provides a brief measure of adolescent loneliness for clinicians and researchers that also predicts very elevated levels of depression.

Nephrometry scores to predict oncological outcomes following partial nephrectomy (UroCCR Study 70)

PURPOSE

Partial nephrectomy (PN) for large or complex renal tumors can be difficult and associated with a higher risk of recurrence than radical nephrectomy. We aim to evaluate the clinical useful of nephrometry scores for predicting oncological outcomes in a large cohort of patients who underwent PN for renal cell carcinomas.

METHODS

Our analysis included patients who underwent PN for renal cell carcinoma in 21 French academic centers (2010-2020). RENAL, PADUA, and SPARE scores were calculated based on preoperative imaging. Uni- and multivariate cox models were performed to identify predictors of recurrence-free survival and overall survival. The area under the curve (AUC) was used to identify models with the highest discrimination. Decision curve analyses (DCAs) determined the net benefit associated with their use.

RESULTS

A total of 1927 patients were analyzed with a median follow-up of 32 months (14-45). RENAL score (p = 0.01), age (p = 0.002), histological type (p = 0.001), high nuclear grade (p = 0.001), necrotic component (p < 0.001), and positive margins (p = 0.005) were significantly related to recurrence in multivariate analyses. The discriminative performance of the 3 radiological scores was modest (65, 63, and 63%, respectively). All 3 scores showed good calibration, which, however, deteriorated with time. Decision curve analysis of the three models for the prediction of overall and recurrence-free survival was similar for all three scores and of limited clinical relevance.

CONCLUSION

The association between nephrometry scores and oncological outcomes after NP is very weak. The use of these scores for predicting oncological outcomes in routine practice is therefore of limited clinical value.

Predicting the formation of mixed pattern hemorrhages in ruptured middle cerebral artery aneurysms based on a decision tree model: A multicenter study

OBJECTIVE

Mixed-pattern hemorrhages (MPH) commonly occur in ruptured middle cerebral artery (MCA) aneurysms and are associated with poor clinical outcomes. This study aimed to predict the formation of MPH in a multicenter database of MCA aneurysms using a decision tree model.

METHODS

We retrospectively reviewed patients with ruptured MCA aneurysms between January 2009 and June 2020. The MPH was defined as subarachnoid hemorrhages with intracranial hematomas and/or intraventricular hemorrhages and/or subdural hematomas. Univariate and multivariate logistic regression analyses were used to explore the prediction factors of the formation of MPH. Based on these prediction factors, a decision tree model was developed to predict the formation of MPH. Additional independent datasets were used for external validation.

RESULTS

We enrolled 436 patients with ruptured MCA aneurysms detected by computed tomography angiography; 285 patients had MPH (65.4%). A multivariate logistic regression analysis showed that age, aneurysm size, multiple aneurysms, and the presence of a daughter dome were the independent prediction factors of the formation of MPH. The areas under receiver operating characteristic curves of the decision tree model in the training, internal, and external validation cohorts were 0.951, 0.927, and 0.901, respectively.

CONCLUSION

Age, aneurysm size, the presence of a daughter dome, and multiple aneurysms were the independent prediction factors of the formation of MPH. The decision tree model is a useful visual triage tool to predict the formation of MPH that could facilitate the management of unruptured aneurysms in routine clinical work.

Competing endogenous RNA network characterization of lymph node metastases in Leuran gastric cancer subtypes

Gastric cancer is a kind of tumor with strong heterogeneity. Long noncoding RNAs (lncRNAs) acting as competing endogenous RNAs (ceRNAs) play significant roles in the development of tumors. In this study, we divided all TCGA gastric cancer patients into the whole, intestinal and diffuse cohorts for further analysis, and constructed competitive endogenous RNA network and evaluated immune cells using CIBERSORTx. The support vector machines recursive feature elimination (SVM-RFE) was used for screening significant signatures and the support vector machines (SVM) for establishing model predicting the lymph node metastasis. The performance of SVM model was good in the intestinal and diffuse cohort, while the model in the whole cohort was relatively poor. Some important co-expression patterns between immune cells and ceRNAs network indicated significant correlation CD70 with dendritic cells and so on. Our research inferred competing endogenous RNA network of lymph node metastasis and built an excellent predicting model.

Study on the prediction effect of a combined model of SARIMA and LSTM based on SSA for influenza in Shanxi Province, China

BACKGROUND

Influenza is an acute respiratory infectious disease that is highly infectious and seriously damages human health. Reasonable prediction is of great significance to control the epidemic of influenza.

METHODS

Our Influenza data were extracted from Shanxi Provincial Center for Disease Control and Prevention. Seasonal-trend decomposition using Loess (STL) was adopted to analyze the season characteristics of the influenza in Shanxi Province, China, from the 1st week in 2010 to the 52nd week in 2019. To handle the insufficient prediction performance of the seasonal autoregressive integrated moving average (SARIMA) model in predicting the nonlinear parts and the poor accuracy of directly predicting the original sequence, this study established the SARIMA model, the combination model of SARIMA and Long-Short Term Memory neural network (SARIMA-LSTM) and the combination model of SARIMA-LSTM based on Singular spectrum analysis (SSA-SARIMA-LSTM) to make predictions and identify the best model. Additionally, the Mean Squared Error (MSE), Mean Absolute Error (MAE) and Root Mean Squared Error (RMSE) were used to evaluate the performance of the models.

RESULTS

The influenza time series in Shanxi Province from the 1st week in 2010 to the 52nd week in 2019 showed a year-by-year decrease with obvious seasonal characteristics. The peak period of the disease mainly concentrated from the end of the year to the beginning of the next year. The best fitting and prediction performance was the SSA-SARIMA-LSTM model. Compared with the SARIMA model, the MSE, MAE and RMSE of the SSA-SARIMA-LSTM model decreased by 38.12, 17.39 and 21.34%, respectively, in fitting performance; the MSE, MAE and RMSE decreased by 42.41, 18.69 and 24.11%, respectively, in prediction performances. Furthermore, compared with the SARIMA-LSTM model, the MSE, MAE and RMSE of the SSA-SARIMA-LSTM model decreased by 28.26, 14.61 and 15.30%, respectively, in fitting performance; the MSE, MAE and RMSE decreased by 36.99, 7.22 and 20.62%, respectively, in prediction performances.

CONCLUSIONS

The fitting and prediction performances of the SSA-SARIMA-LSTM model were better than those of the SARIMA and the SARIMA-LSTM models. Generally speaking, we can apply the SSA-SARIMA-LSTM model to the prediction of influenza, and offer a leg-up for public policy.

Three-dimensional amide proton transfer-weighted and intravoxel incoherent motion imaging for predicting bone metastasis in patients with prostate cancer: A pilot study

PURPOSE

To explore the value of 3-dimensional amide proton transfer-weighted (APTw) and intravoxel incoherent motion (IVIM) imaging in predicting bone metastasis (BM) of prostate cancer (PCa) in addition to routine diffusion-weighted imaging (DWI).

METHODS

The clinical and imaging data of 39 PCa patients who were pathologically confirmed in our hospital from March 2019 to February 2022 were retrospectively analyzed, and they were divided into BM-negative (27 patients) and BM-positive (12 patients) groups. MR examination included APTw, DWI and IVIM imaging. The IVIM data was fitted by single-exponential IVIM model (IVIM_mono) and double-exponential IVIM model (IVIM_bi), respectively. The APTw, ADC, IVIM_mono (D_mono, D*_mono, and f_mono), and IVIM_bi (D_bi, D*_bi, and f_bi) parameters were independently measured by two radiologists. The synthetic minority oversampling technique (SMOTE) was conducted to balance the minority group. Mann-Whitney U test or Student's t-test was used to compare above values between the BM-negative and BM-positive groups. The diagnostic performance was evaluated with receiver operating characteristic (ROC) analysis of each parameter and their combination. The Delong test was used for ROC curve comparison.The relationship between APTw and IVIM was explored through Spearman's rank correlation analysis.

RESULTS

The APTw and D*_mono values were higher, and the ADC, f_mono, and f_bi values were lower in the BM-positive group than in the BM-negative group (all P < 0.05). Among the individual parameters, the AUC of f_mono was the highest (AUC = 0.865), and AUC (f_mono) was significantly higher than AUC (f_bi), AUC (D*_mono), and AUC (ADC) (all P < 0.05). The AUC (IVIM_mono) was higher than the AUC (IVIM_bi) (P = 0.0068). The combination of APTw and IVIM_mono further improved diagnostic capability, and the AUC of APTw+IVIM_mono was significantly higher than those of APTw and DWI (all P < 0.05). No correlation was found between IVIM-derived parameters and APTw value.

CONCLUSION

Both 3D APTw and IVIM imaging could predict BM of PCa. IVIM showed better performance than APTw and DWI, and the single-exponential IVIM model was superior to the double-exponential IVIM model. The combination of APTw and IVIM could further improve diagnostic performance.

Adopting effective hierarchal IoMTs computing with K-efficient clustering to control and forecast COVID-19 cases

The Internet of Medical Things (IoMTs) based on fog/cloud computing has been effectively proven to improve the controlling, monitoring, and care quality of Coronavirus disease 2019 (COVID-19) patients. One of the convenient approaches to assess symptomatic patients is to group patients with comparable symptoms and provide an overview of the required level of care to patients with similar conditions. Therefore, this study adopts an effective hierarchal IoMTs computing with K-Efficient clustering to control and forecast COVID-19 cases. The proposed system integrates the K-Means and K-Medoids clusterings to monitor the health status of patients, early detection of COVID-19 cases, and process data in real-time with ultra-low latency. In addition, the data analysis takes into account the primary requirements of the network to assist in understanding the nature of COVID-19. Based on the findings, the K-Efficient clustering with fog computing is a more effective approach to analyse the status of patients compared to that of K-Means and K-Medoids in terms of intra-class, inter-class, running time, the latency of network, and RAM consumption. In summary, the outcome of this study provides a novel approach for remote monitoring and handling of infected COVID-19 patients through real-time personalised treatment services.

Clinical significance of neutrophil gelatinase-associated lipocalin and sdLDL-C for coronary artery disease in patients with type 2 diabetes mellitus aged ≥ 65 years

BACKGROUND AND AIMS

Although type 2 diabetes mellitus (T2DM) and coronary artery disease (CAD) share many common pathological and physiological characteristics, there are few studies assessing the predictive capacity of novel biomarkers in occurrence and development of CAD in T2DM patients aged ≥ 65 years. In addition, T2DM patients aged ≥ 65 years are prone to CAD. Therefore, it is of great significance to find novel biomarkers for the development CAD in T2DM.

METHODS

In this retrospective cohort study, 579 T2DM patients aged ≥ 65 years were consecutively enrolled in this work, and 177 of whom had major adverse cardiovascular and cerebrovascular events (MACCE: cardiovascular or cerebrovascular death, acute coronary syndrome, coronary stent implantation, and stroke) during the follow up. Univariate and multivariate factors were employed to analyze the correlation between each variable and the occurrence of MACCE, and the Spearman's rank correlation analysis was performed to assess the relationships between Neutrophil gelatinase-associated lipocalin (NGAL) and small dense low-density lipoprotein-cholesterol (LDL-C) (sdLDL-C). The receiver operating characteristic (ROC) curve was adopted to determine the predictive value of NGAL and sdLDL-C elevation for MACCE in T2DM patients aged ≥ 65 years.

RESULTS

After a median 48 months follow-up [19, (10 ~ 32) ], the levels of NGAL, sdLDL-C, hemoglobin A1c (HbA1c), LDL-C, and apolipoprotein B (ApoB) were significantly higher while those of high-density lipoprotein cholesterol (HDL-C) and apolipoprotein A I (ApoA-I) were lower in MACCE positive group. NGAL correlated to body mass index (BMI) (r = 0.391, P = 0.001) and triglyceride (TG) (r = 0.228, P = 0.032), and high-sensitivity CRP (hsCRP) (r = 0.251, P = 0.007), and neutrophils (r = 0.454, P = 0.001), sdlDL-C level was found to be positively correlated with LDL-C (r = 0.413, P = 0.001), TG (r = 0.432, P = 0.001), and ApoB (r = 0.232, P = 0.002); and it was negatively correlated with HDL-C (r = -0.362, P = 0.031) and ApoA-I (r = -0.402, P = 0.001). Age-adjusted Cox regression analysis showed that NGAL (HR = 1.006, 95% confidence interval (CI): 1.005-1.008, P < 0.001) and sdLDL-C (HR = 1.052, 95% CI: 1.037-1.066, P < 0.001) were independently associated with occurrence of MACCE. ROC curve analysis showed that NGAL (area under ROC (AUC) = 0.79, 95% CI: 0.75-0.84, P < 0.001) and sdlDL-C (AUC = 0.76, 95% CI: 0.72-0.80, P < 0.001) could predict the occurrence of MACCE (area under ROC. NGAL combined with sdlDL-C could predict the occurrence of MACCE well (AUC = 0.87, 95% CI: 0.84-0.90, P < 0.001).

CONCLUSION

The higher NGAL and sdLDL-C in T2DM patients aged ≥ 65 years were significantly and independently associated with the risk of MACCE, and showed higher clinical values than other lipid biomarkers or other chronic inflammation, so they were expected to be the most effective predictors of MACCE assessment.

Prototypes virus of hand, foot and mouth disease infections and severe cases in Gansu, China: a spatial and temporal analysis

Background

Little research has been conducted on the spatio-temporal relationship between the severe cases and the enteroviruses infections of hand, foot and mouth disease (HFMD). This study aimed to investigate epidemic features and spatial clusters of HFMD incidence rates and assess the relationship between Enterovirus 71 (EV71) and Coxsackievirus A16 (CoxA16) and severe cases of HMFD in Gansu province, China.

Methods

Weekly county-specific data on HFMD between 1st January and 31st December 2018 were collected from the China Infectious Disease Information System (CIDIS), including enterovirus type (EV71 and CoxA16), severe and non-severe cases in Gansu province, China. Temporal risk [frequency index (α), duration index (β) and intensity index (γ)] and spatial cluster analysis were used to assess epidemic features and identify high-risk areas for HFMD. Time-series cross-correlation function and regression model were used to explore the relationship between the ratios of two types of viruses (i.e. EV71/Cox16) (EC) and severe cases index (i.e. severe cases/non-severe cases) (SI) of HFMD.

Results

Some counties in Dingxi City, Gansu were identified as a hot spot for the temporal risk indices. Time-series cross-correlation analysis showed that SI was significantly associated with EC (r = 0.417, P < 0.05) over a 4-week time lag. The regression analysis showed that SI was positively associated with EC (β = 0.04, 95% confidence interval (CI) 0.02–0.06).

Conclusion

The spatial patterns of HFMD incidence were associated with enteroviruses in Gansu. The research suggested that the EC could be considered a potential early warning sign for predicting severe cases of HFMD in Gansu province.

The AGH score is a predictor of disease-free survival and targeted therapy efficacy after liver transplantation in patients with hepatocellular carcinoma

BACKGROUND

Liver transplantation (LT) is the "cure" therapy for patients with hepatocellular carcinoma (HCC). However, some patients encounter HCC recurrence after LT. Unfortunately, there is no effective methods to identify the LT patients who have high risk of HCC recurrence and would benefit from adjuvant targeted therapy. The present study aimed to establish a scoring system to predict HCC recurrence of HCC patients after LT among the Chinese population, and to evaluate whether these patients are suitable for adjuvant targeted therapy.

METHODS

Clinical data of HCC patients who underwent LT from March 2015 to June 2019 were retrospectively collected and analyzed.

RESULTS

A total of 201 patients were included in the study. The multivariate Cox analysis suggested that preoperative alpha fetoprotein (AFP) > 200 µg/L (HR = 2.666, 95% CI: 1.515-4.690; P = 0.001), glutamyl transferase (GGT) > 96 U/L (HR = 1.807, 95% CI: 1.012-3.224; P = 0.045), and exceeding the Hangzhou criteria (HR = 2.129, 95% CI: 1.158-3.914; P = 0.015) were independent risk factors for poor disease-free survival (DFS) in patients with HCC who underwent LT. We established an AFP-GGT-Hangzhou (AGH) scoring system based on these factors, and divided cases into high-, moderate-, and low-risk groups. The differences in overall survival (OS) and disease-free survival (DFS) rates among the three groups were significant (P < 0.05). The efficacy of the AGH scoring system to predict DFS was better than that of the Hangzhou criteria, UCSF criteria, Milan criteria, and TNM stage. Only in the high-risk group, we found that lenvatinib significantly improved prognosis compared with that of the control group (P < 0.05).

CONCLUSIONS

The AGH scoring system provides a convenient and effective way to predict HCC recurrence after LT in HCC patients in China. Patients with a high-risk AGH score may benefit from lenvatinib adjuvant therapy after LT.

Inclusion of quantitative high-density plaque in coronary computed tomographic score system to predict the time of guidewire crossing chronic total occlusion

OBJECTIVE

This study aimed to establish a new scoring system that includes histological quantitative features derived from coronary computed tomographic angiography (CCTA) to predict the efficiency of chronic total occlusion percutaneous coronary intervention (CTO-PCI).

METHODS

This study analyzed clinical, morphological, and histological characteristics of 207 CTO lesions in 201 patients (mean age 60.0 [52.0-65.0] years, 85% male), which were recruited from two centers. The primary endpoint was a guidewire successfully crossing the lesions within 30 m. The new predictive model was generated by factors that were determined by multivariate analysis. The CCTA plaque (CTAP) score that included a quantitative plaque characteristic was developed by assigning an appropriate integer score to each independent predictor, then summing all points. In addition, the CTAP score was compared with other predictive scores based on CCTA.

RESULTS

The endpoint was achieved in 63% of the lesions. The independent predictors included previous CTO-PCI failure, the proximal blunt stump, proximal side branch, distal side branch, occluded segment bending > 45°, and high-density plaque volume (fibrous volume + calcified volume) ≥ 19.9 mm³. As the score increased from 0 to 5, the success rate of the guidewire crossing within 30 m decreased from 96 to 0%. Comparing the CTAP score with other predictive scores, the CTAP score showed the highest discriminant power (c-statistic = 0.81 versus 0.73-0.77, p value 0.02-0.07). The CTAP score showed similar results for procedural success.

CONCLUSION

The CTAP score efficiently predicted the guidewire crossing efficiency and procedural success.

KEY POINTS

• An increase in high-density plaque volume (fibrous + dense calcium) was more probable to reduce the efficiency of crossing and lead to procedural failure. • The new prediction scoring system with the addition of the quantitative characteristics of plaques had an improved predictive ability compared with the traditional prediction scoring system.

Predicting the incidence of COVID-19 using data mining

BACKGROUND

The high prevalence of COVID-19 has made it a new pandemic. Predicting both its prevalence and incidence throughout the world is crucial to help health professionals make key decisions. In this study, we aim to predict the incidence of COVID-19 within a two-week period to better manage the disease.

METHODS

The COVID-19 datasets provided by Johns Hopkins University, contain information on COVID-19 cases in different geographic regions since January 22, 2020 and are updated daily. Data from 252 such regions were analyzed as of March 29, 2020, with 17,136 records and 4 variables, namely latitude, longitude, date, and records. In order to design the incidence pattern for each geographic region, the information was utilized on the region and its neighboring areas gathered 2 weeks prior to the designing. Then, a model was developed to predict the incidence rate for the coming 2 weeks via a Least-Square Boosting Classification algorithm.

RESULTS

The model was presented for three groups based on the incidence rate: less than 200, between 200 and 1000, and above 1000. The mean absolute error of model evaluation were 4.71, 8.54, and 6.13%, respectively. Also, comparing the forecast results with the actual values in the period in question showed that the proposed model predicted the number of globally confirmed cases of COVID-19 with a very high accuracy of 98.45%.

CONCLUSION

Using data from different geographical regions within a country and discovering the pattern of prevalence in a region and its neighboring areas, our boosting-based model was able to accurately predict the incidence of COVID-19 within a two-week period.

The clinical role of combined serum C1q and hsCRP in predicting coronary artery disease

OBJECTIVE

C1q has been shown to be associated with coronary heart disease (CAD) and can co-deposit with C-reactive protein (CRP) in atherosclerotic plaques. However, few studies have been conducted between C1q, CRP parameters and CAD. The aim of this study is to explore the relationship between C1q and CRP parameters and assess their clinical significance in CAD.

METHODS

238 total patients who underwent coronary artery angiography were enrolled and divided into control group (n = 65), stable CAD group (n = 47) and unstable angina group (UA group, n = 126). Patients' data were collected from self-administered questionnaires and electrical medical records. The severity of coronary stenosis was presented by Gensini score. The relationship between C1q, CRP parameters and CAD were evaluated by multivariate regression analysis and their predicting performance were assessed by ROC analysis and odds ratio analysis.

RESULTS

Compared with control group, C1q was showed significantly lower in stable CAD (P = 0.004) and UA groups (P = 0.008), while hsCRP was higher in UA group (P = 0.024). Serum C1q was weakly positively associated with hsCRP (r = 0.24, P < 0.001) but not correlated with Gensini score. Logistic regression identified C1q (OR: 0.87 per 10 mg/L, 95% CI: 0.79-0.95, P = 0.001) and hsCRP (OR: 1.08 mg/L, 95% CI: 1.01-1.15, P = 0.032) as independent determinants of CAD. Furthermore, combined C1q and hsCRP level showed higher discriminatory accuracy in predicting CAD than C1q (AUC: 0.676 vs 0.585, P = 0.101; NRI: 10.4%, P = 0.049; IDI: 3.9%, P < 0.001) or hsCRP (AUC: 0.676 vs 0.585, P = 0.101; NRI: 16.7%, P = 0.006; IDI: 5.8%, P < 0.001).

CONCLUSIONS

Reduced serum C1q and increased hsCRP are independently associated with CAD and could be potential predictors for CAD diagnosis. Furthermore, combined C1q and hsCRP showed better performance in predicting CAD than using single one.

Evaluation of the value of fasting plasma glucose in the first trimester for the prediction of adverse pregnancy outcomes

AIMS

To evaluate the importance and usefulness of fasting plasma glucose (FPG) in the first trimester in predicting adverse pregnancy outcomes.

METHODS

A retrospective study of 22,398 singleton pregnancies was conducted. Participants were divided into subgroups according to first-trimester FPG (low FPG, FPG < 5.1 mmol/L; medium FPG, 5.1 mmol/L ≤ FPG < 5.6 mmol/L; high FPG, 5.6 ≤ FPG < 7.0 mmol/L) and oral glucose tolerance test(OGTT) results (normal and abnormal) during pregnancy. Patient characteristics and risk of adverse pregnancy outcomes were compared. Then, the whole population of women with abnormal OGTT served as a reference, and the relative risks of maternal and neonatal complications in normal OGTT women were analyzed by categorical analyses and logistic regression. Subgroup analyses were performed according to pre-pregnancy body mass index (BMI).

RESULTS

The frequency of adverse pregnancy outcomes increased with increasing FPG levels during the first trimester, regardless of OGTT results. High FPG + Abnormal OGTT had the worst outcome. Compared to the whole population of women with abnormal OGTT, Normal OGTT + Medium FPG showed the same risk of PIH and macrosomia. Normal OGTT + High FPG showed the same risk of PIH, macrosomia as well as LGA and preterm birth. Additionally, Normal OGTT + Medium FPG + BMI ≥ 24 kg/m² showed significantly higher risk of PIH (OR = 1.867, 1.245-2.800), macrosomia (OR = 1.748, 1.304-2.344) and LGA (OR = 1.274, 1.019-1.593). Furthermore, the OR value for PIH was 3.759 (1.680-8.412) in Normal OGTT + High FPG + BMI ≥ 24 kg/m² compared to women with abnormal OGTT.

CONCLUSIONS

First-trimester FPG values can help identify women at increased risk for adverse pregnancy outcomes. Increased attention and management should be given to women with early pregnancy FPG ≥ 5.10 mmol/L despite a normal OGTT, especially if their BMI ≥ 24 kg/m².

Comparing the performance of time series models with or without meteorological factors in predicting incident pulmonary tuberculosis in eastern China

Background

Many studies have compared the performance of time series models in predicting pulmonary tuberculosis (PTB), but few have considered the role of meteorological factors in their prediction models. This study aims to explore whether incorporating meteorological factors can improve the performance of time series models in predicting PTB.

Methods

We collected the monthly reported number of PTB cases and records of six meteorological factors in three cities of China from 2005 to 2018. Based on this data, we constructed three time series models, including an autoregressive integrated moving average (ARIMA) model, the ARIMA with exogenous variables (ARIMAX) model, and a recurrent neural network (RNN) model. The ARIMAX and RNN models incorporated meteorological factors, while the ARIMA model did not. The mean absolute percentage error (MAPE) and root mean square error (RMSE) were used to evaluate the performance of the models in predicting PTB cases in 2018.

Results

Both the cross-correlation analysis and Spearman rank correlation test showed that PTB cases reported in the study areas were related to meteorological factors. The predictive performance of both the ARIMA and RNN models was improved after incorporating meteorological factors. The MAPEs of the ARIMA, ARIMAX, and RNN models were 12.54%, 11.96%, and 12.36% in Xuzhou, 15.57%, 11.16%, and 14.09% in Nantong, and 9.70%, 9.66%, and 12.50% in Wuxi, respectively. The RMSEs of the three models were 36.194, 33.956, and 34.785 in Xuzhou, 34.073, 25.884, and 31.828 in Nantong, and 19.545, 19.026, and 26.019 in Wuxi, respectively.

Conclusions

Our study revealed a possible link between PTB and meteorological factors. Taking meteorological factors into consideration increased the accuracy of time series models in predicting PTB, and the ARIMAX model was superior to the ARIMA and RNN models in study settings.

Severe or critical hypotension during post cardiac arrest care is associated with factors available on admission - a post hoc analysis of the TTH48 trial

PURPOSE

We explored whether severe or critical hypotension can be predicted, based on patient and resuscitation characteristics in out-of-hospital cardiac arrest (OHCA) patients. We also explored the association of hypotension with mortality and neurological outcome.

MATERIALS AND METHODS

We conducted a post hoc analysis of the TTH48 study (NCT01689077), where 355 out-of-hospital cardiac arrest (OHCA) patients were randomized to targeted temperature management (TTM) treatment at 33 °C for either 24 or 48 h. We recorded hypotension, according to four severity categories, within four days from admission. We used multivariable logistic regression analysis to test association of admission data with severe or critical hypotension.

RESULTS

Diabetes mellitus (OR 3.715, 95% CI 1.180-11.692), longer ROSC delay (OR 1.064, 95% CI 1.022-1.108), admission MAP (OR 0.960, 95% CI 0.929-0.991) and non-shockable rhythm (OR 5.307, 95% CI 1.604-17.557) were associated with severe or critical hypotension. Severe or critical hypotension was associated with increased mortality and poor neurological outcome at 6 months.

CONCLUSIONS

Diabetes, non-shockable rhythm, longer delay to ROSC and lower admission MAP were predictors of severe or critical hypotension. Severe or critical hypotension was associated with poor outcome.

Predicting Inpatient Length of Stay in Iranian Hospital: Conceptualization and Validation

Objective:

The length of stay is an important indicator of hospital performance and efficiency. Regarding the importance of the length of stay, this study aimed to design a structural model of the inpatients’ length of stay in the educational and therapeutic health care facilities of Iran in order to identify the influencing dimensions.

Methods:

The present study was an analytical and applied study. The face validity of the data gathering tool was investigated by the expert judgment and the construct validity was examined by using the exploratory factor analysis. In order to verify the reliability of the tool, the internal consistency was also trialed by using the Cronbach’s alpha. For ranking the influencing dimensions and factors and also in order to examine the causal relationships between the variables in a coherent manner and presenting the final model, the structural equation modeling technique was used in AMOS software at a significant level of 0.05.

Results:

The mentioned structural model consists of 4 dimensions and 29 factors influencing the length of stay of hospitalized patients. The independent variables are based on priority and importance as follows: patients’ conditions, the underlying factors, the clinical staff performance, and hospitals’ service delivery, which were examined by second-order factor analysis in order to study the relationship between them and the inpatients’ length of stay.

Conclusion:

Considering the importance of each one of the proposed dimensions from the point of view of service providers in some therapeutic centers of the country by paying attention to the role of each one of them in preventing prolonged hospitalization can be essential in the effectiveness of the treatment and cost reduction.

Excision of pediatric craniopharyngioma: pattern of recurrence in 35 patients at a tertiary care hospital in Saudi Arabia

BACKGROUND AND PURPOSE

Craniopharyngiomas are benign tumors of central nervous system which are known to affect both adults and children. Despite their benign origin, the recurrence is still one of the main postoperative challenges. The aim of this study was to investigate in retrospect factors related to recurrence of craniopharyngioma in a tertiary center in Riyadh, Saudi Arabia.

PATIENTS AND METHODS

We conducted a review of charts of all craniopharyngioma patients operated in neurosurgery department at King Faisal Specialist Hospital & Research Center in Riyadh (KFSH-RC). Age at surgery, gender, body mass index, symptoms at presentation, hormonal data, tumor characteristics and location, presence of hydrocephalus, previous treatments, neuroimaging features, surgical results, and recurrence were abstracted from the medical charts of the patients retrospectively.

RESULTS

In all, 70.6% of patients had gross total resection (GTR). The recurrence after GTR in our series was 25% which considered low when compared to most surgical series. From all above studied variables, VP shunt insertion at presentation was constantly significant in both uni- and multi-variable analysis.

CONCLUSION

In this study, we analyzed several factors to determine if they had any significant correlation with recurrence. Only VP shunt insertion was found significant. Further researches are needed to verify these factors and to discover others.

ACCBN: ant-Colony-clustering-based bipartite network method for predicting long non-coding RNA-protein interactions

BACKGROUND

Long non-coding RNA (lncRNA) studies play an important role in the development, invasion, and metastasis of the tumor. The analysis and screening of the differential expression of lncRNAs in cancer and corresponding paracancerous tissues provides new clues for finding new cancer diagnostic indicators and improving the treatment. Predicting lncRNA-protein interactions is very important in the analysis of lncRNAs. This article proposes an Ant-Colony-Clustering-Based Bipartite Network (ACCBN) method and predicts lncRNA-protein interactions. The ACCBN method combines ant colony clustering and bipartite network inference to predict lncRNA-protein interactions.

RESULTS

A five-fold cross-validation method was used in the experimental test. The results show that the values of the evaluation indicators of ACCBN on the test set are significantly better after comparing the predictive ability of ACCBN with RWR, ProCF, LPIHN, and LPBNI method.

CONCLUSIONS

With the continuous development of biology, besides the research on the cellular process, the research on the interaction function between proteins becomes a new key topic of biology. The studies on protein-protein interactions had important implications for bioinformatics, clinical medicine, and pharmacology. However, there are many kinds of proteins, and their functions of interactions are complicated. Moreover, the experimental methods require time to be confirmed because it is difficult to estimate. Therefore, a viable solution is to predict protein-protein interactions efficiently with computers. The ACCBN method has a good effect on the prediction of protein-protein interactions in terms of sensitivity, precision, accuracy, and F1-score.

First-trimester fasting plasma glucose as a predictor of gestational diabetes mellitus and the association with adverse pregnancy outcomes

Objective:

To evaluate the usefulness of a fasting plasma glucose (FPG) at the first trimester in predicting gestational diabetes mellitus (GDM) and the association between FPG and adverse pregnancy outcomes.

Methods:

The levels of FPG in women with singleton pregnancies were measured at 9-13⁺⁶ weeks. A two hour 75-g oral glucose tolerance test (OGTT) was completed at 24-28 weeks and the International Association of Diabetes and Pregnancy Study Groups (IADPSG) criteria was used. Adverse pregnancy outcomes were assessed and recorded.

Results:

Among 2112 pregnant women enrolled in the study, 224 (10.6%) subjects were diagnosed with GDM. The AUC for FPG in predicting GDM was 0.63 (95% CI 0.61- 0.65) and the optimal cutoff value was 4.5 mmol/L (sensitivity 64.29% and specificity 56.45%). Higher first-trimester FPG increased the prevalence of GDM, large for gestational age (LGA) and assisted vaginal delivery and/or cesarean section (all P < 0.05).

Conclusion:

FPG at first trimester could be used to predict GDM and higher first-trimester FPG was associated with adverse pregnancy outcomes.

Predicting Surgical Complications in Adult Patients Undergoing Anterior Cervical Discectomy and Fusion Using Machine Learning

Objective

Machine learning algorithms excel at leveraging big data to identify complex patterns that can be used to aid in clinical decision-making. The objective of this study is to demonstrate the performance of machine learning models in predicting postoperative complications following anterior cervical discectomy and fusion (ACDF).

Methods

Artificial neural network (ANN), logistic regression (LR), support vector machine (SVM), and random forest decision tree (RF) models were trained on a multicenter data set of patients undergoing ACDF to predict surgical complications based on readily available patient data. Following training, these models were compared to the predictive capability of American Society of Anesthesiologists (ASA) physical status classification.

Results

A total of 20,879 patients were identified as having undergone ACDF. Following exclusion criteria, patients were divided into 14,615 patients for training and 6,264 for testing data sets. ANN and LR consistently outperformed ASA physical status classification in predicting every complication (p < 0.05). The ANN outperformed LR in predicting venous thromboembolism, wound complication, and mortality (p < 0.05). The SVM and RF models were no better than random chance at predicting any of the postoperative complications (p < 0.05).

Conclusion

ANN and LR algorithms outperform ASA physical status classification for predicting individual postoperative complications. Additionally, neural networks have greater sensitivity than LR when predicting mortality and wound complications. With the growing size of medical data, the training of machine learning on these large datasets promises to improve risk prognostication, with the ability of continuously learning making them excellent tools in complex clinical scenarios.

Cut-off Value for Body Mass Index in Predicting Surgical Success in Patients with Lumbar Spinal Canal Stenosis

STUDY DESIGN

Case-control.

PURPOSE

To determine optimal cut-off value for body mass index (BMI) in predicting surgical success in patients with lumbar spinal canal stenosis (LSCS).

OVERVIEW OF LITERATURE

BMI is an essential variable in the assessment of patients with LSCS.

METHODS

We conducted a prospective study with obese and non-obese LSCS surgical patients and analyzed data on age, sex, duration of symptoms, walking distance, morphologic grade of stenosis, BMI, postoperative complications, and functional disability. Obesity was defined as BMI of ≥30 kg/m2. Patients completed the Oswestry Disability Index (ODI) questionnaire before surgery and 2 years after surgery. Surgical success was defined as ≥30% improvement from the baseline ODI score. Receiver operating characteristic (ROC) analysis was used to estimate the optimal cut-off values of BMI to predict surgical success. In addition, correlation was assessed between BMI and stenosis grade based on morphology as defined by Schizas and colleague in total, 189 patients were eligible to enter the study.

RESULTS

Mean age of patients was 61.5±9.6 years. Mean follow-up was 36±12 months. Most patients (88.4%) were classified with grades C (severe stenosis) and D (extreme stenosis). Post-surgical success was 85.7% at the 2-year follow-up. A weak correlation was observed between morphologic grade of stenosis and BMI. Rates of postoperative complications were similar between patients who were obese and those who were non-obese. Both cohorts had similar degree of improvement in the ODI at the 2-year followup. However, patients who were non-obese presented significantly higher surgical success than those who were obese. In ROC curve analysis, a cut-off value of ≤29.1 kg/m2 for BMI in patients with LSCS was suggestive of surgical success, with 81.1% sensitivity and 82.2% specificity (area under the curve, 0.857; 95% confidence interval, 0.788-0.927).

CONCLUSION

This study showed that the BMI can be considered a parameter for predicting surgical success in patients with LSCS and can be useful in clinical practice.

Association of increased microvessel density with skeletal extramedullary disease relapse in multiple myeloma patients who have skeletal extramedullary disease at diagnosis

The aim of the study was to investigate whether microvessel density (MVD) could be associated with skeletal extramedullary disease relapse (skeletal-EMDR) in patients with multiple myeloma (MM) who have skeletal-EMD at diagnosis. Seventy-nine newly diagnosed MM patients who have skeletal-EMD were retrospectively enrolled in this study. The 4-year cumulative incidence of skeletal-EMDR was 35.0%±8.3%. The 4-year probability of overall survival (OS) was 54.0%±7.6%. Multivariate analysis showed that skeletal-EMDR (HR = 4.144; 95% CI: 1.608-10.685; P = 0.003) was independently associated with inferior OS for the MM patients who have skeletal-EMD at diagnosis. The factors associated with skeletal-EMDR were MVD (HR = 3.990, 95%CI:1.136-14.018; P = 0.031), white blood cell (WBC) (HR = 0.262, 95% CI:0.090-0.769; P = 0.015), and the EMD sites involved at onset (HR = 0.263, 95% CI: 0.074-0.937; P = 0.039). The MVD in patients with thoracic and lumbar vertebrae as the involved sites at diagnosis was significantly lower than those with other sites involved (41.59 ± 14.39 vs. 60.82 ± 35.14, P=0.001). Our data suggest that increased MVD could be used to predict skeletal-EMDR, which is associated with inferior survival in patients with MM who have skeletal-EMD at diagnosis.

Identification of a low risk population for parametrial invasion in patients with early-stage cervical cancer

Background

Recent studies have challenged radical procedures for less extensive surgery in selected patients with early-stage cervical cancer at low risk of parametrial invasion. Our objective was to identify a subgroup of patients at low risk of parametrial invasion among women having undergone surgical treatment.

Methods

Data of 1447 patients with cervical cancer treated between 1996 and 2016 were extracted from maintained databases of 10 French University hospitals. Patients with early-stage (IA2–IIA) disease treated by radical surgery including hysterectomy and trachelectomy, were selected for further analysis. The Kaplan–Meier method was used to estimate the survival distribution. A Cox proportional hazards model including all the parameters statistically significant in univariate analysis, was used to account for the influence of multiple variables.

Results

Out of the 263 patients included for analysis, on final pathology analysis 28 (10.6%) had parametrial invasion and 235 (89.4%) did not. Factors significantly associated with parametrial invasion on multivariate analysis were: age > 65 years, tumor > 30 mm in diameter measured by MRI, lymphovascular space invasion (LVSI) on pathologic analysis. Among the 235 patients with negative pelvic lymph nodes, parametrial disease was seen in only 7.6% compared with 30.8% of those with positive pelvic nodes (p < 0.001). In a subgroup of patients presenting tumors < 30 mm, negative pelvic status and no LVSI, the risk of parametrial invasion fell to 0.6% (1/173 patients).

Conclusion

Our analysis suggests that there is a subgroup of patients at very low risk of parametrial invasion, potentially eligible for less radical procedures.

Electronic supplementary material

The online version of this article (10.1186/s12967-018-1531-6) contains supplementary material, which is available to authorized users.

The role of propionic acid at diagnosis predicts mortality in patients with septic shock

PURPOSE

This study aims to assess the diagnostic and prognostic value of propionic acid in patients with septic shock on a medical intensive care unit (ICU).

METHODS

Serum propionic acid and clinical common cytokines levels were measured within 24h after the diagnosis of sepsis, and the Acute Physiology and Chronic Health Evaluation II (APACHE II) score, Sequential Organ Failure Assessment (SOFA) score, and Mortality were recorded in ICU. A 28-day and 90-day follow-up was performed for all patients.

RESULTS

A total of 118 septic patients were enrolled in this study. The propionic acid was higher in patients with septic shock compared with sepsis. Multivariate logistic regression analysis showed that propionic acid was independent predictor of sepsis (odds ratio: 1.279; 95% confidence interval: 1.069-1.530; P=0.007) and septic shock (odds ratio: 1.859; 95% confidence interval: 1.342-2.576; P<0.001) and ICU mortality (odds ratio: 1.331; 95% confidence interval: 1.107-1.600; P=0.002), 28-day mortality (odds ratio: 1.259; 95% confidence interval: 1.046-1.514; P=0.015) and 90-day mortality (odds ratio: 1.304; 95% confidence interval: 1.092-1.558; P=0.003). The receiver operating characteristic curve (AUC) analysis showed the areas under of propionic acid on ICU admission day for predicting sepsis and septic shock were 0.773 and 0.85 respectively, the areas under of propionic acid for predicting ICU mortality, 28-d and 90-d mortality were 0.779, 0.739 and 0.809 respectively. Using a PA cutoff of 0.053 and 0.095 for predicting sepsis and septic shock respectively, the sensitivity was 97.62% and 85.5%, and the specificity was 58% and 83.5%, respectively. Using a PA cutoff of 0.139 for predicting ICU mortality, 28- and 90-day mortality, the sensitivity was 69.39%, 67.44% and 69.09% respectively, and the specificity was 78.26%, 73.33% and 82.54% respectively.

CONCLUSIONS

Propionic acid showed diagnostic capacity to diagnose septic shock and revealed prognostic information for mortality.

Predicting the incidence of portosplenomesenteric vein thrombosis in patients with acute pancreatitis using classification and regression tree algorithm

BACKGROUND AND OBJECTIVE

The accurate prediction of portosplenomesenteric vein thrombosis (PVT) in patients with acute pancreatitis(AP) is very important but may also be difficult because of our insufficient understanding of the characteristics of AP-induced PVT. The purpose of this study is to design a decision tree model that provides critical factors associated with PVT using an approach that makes use of classification and regression tree (CART) algorithm.

METHODS

The analysis included 353 patients with AP who were admitted between January 2011 and December 2015. CART model and logistic regression model were each applied to the same 50% of the sample to develop the predictive training models, and these models were tested on the remaining 50%. Statistical indexes were used to evaluate the value of the prediction in the 2 models.

RESULTS

The predicted sensitivity, specificity, positive predictive value, negative predictive value, and accuracy by CART for PVT were 78.0%, 87.2%, 64.0%, 93.2%, and 85.2%, respectively. Significant differences could be found between the CART model and the logistic regression model in these parameters. There were significant differences between the CART and logistic regression models in these parameters (P<.05). When the CART model was used to identify PVT, the area under receiver operating characteristic curve was 0.803, which demonstrated better overall properties than the logistic regression model (area under the curve=0.696) (95% confidence interval, 0.603-0.812).

CONCLUSION

The CART model based on serum amylase, d-dimer, Acute Physiology and Chronic Health Evaluation II, and prothrombin time is more likely to predict the occurrence of PVT induced by AP.

Predicting stillbirth in a low resource setting

Background

Stillbirth is a major contributor to perinatal mortality and it is particularly common in low- and middle-income countries, where annually about three million stillbirths occur in the third trimester. This study aims to develop a prediction model for early detection of pregnancies at high risk of stillbirth.

Methods

This retrospective cohort study examined 6,573 pregnant women who delivered at Federal Medical Centre Bida, a tertiary level of healthcare in Nigeria from January 2010 to December 2013. Descriptive statistics were performed and missing data imputed. Multivariable logistic regression was applied to examine the associations between selected candidate predictors and stillbirth. Discrimination and calibration were used to assess the model’s performance. The prediction model was validated internally and over-optimism was corrected.

Results

We developed a prediction model for stillbirth that comprised maternal comorbidity, place of residence, maternal occupation, parity, bleeding in pregnancy, and fetal presentation. As a secondary analysis, we extended the model by including fetal growth rate as a predictor, to examine how beneficial ultrasound parameters would be for the predictive performance of the model. After internal validation, both calibration and discriminative performance of both the basic and extended model were excellent (i.e. C-statistic basic model = 0.80 (95 % CI 0.78–0.83) and extended model = 0.82 (95 % CI 0.80–0.83)).

Conclusion

We developed a simple but informative prediction model for early detection of pregnancies with a high risk of stillbirth for early intervention in a low resource setting. Future research should focus on external validation of the performance of this promising model.

Electronic supplementary material

The online version of this article (doi:10.1186/s12884-016-1061-2) contains supplementary material, which is available to authorized users.

The development of an automated ward independent delirium risk prediction model

Background A delirium is common in hospital settings resulting in increased mortality and costs. Prevention of a delirium is clearly preferred over treatment. A delirium risk prediction model can be helpful to identify patients at risk of a delirium, allowing the start of preventive treatment. Current risk prediction models rely on manual calculation of the individual patient risk. Objective The aim of this study was to develop an automated ward independent delirium riskprediction model. To show that such a model can be constructed exclusively from electronically available risk factors and thereby implemented into a clinical decision support system (CDSS) to optimally support the physician to initiate preventive treatment. Setting A Dutch teaching hospital. Methods A retrospective cohort study in which patients, 60 years or older, were selected when admitted to the hospital, with no delirium diagnosis when presenting, or during the first day of admission. We used logistic regression analysis to develop a delirium predictive model out of the electronically available predictive variables. Main outcome measure A delirium risk prediction model. Results A delirium risk prediction model was developed using predictive variables that were significant in the univariable regression analyses. The area under the receiver operating characteristics curve of the "medication model" model was 0.76 after internal validation. Conclusions CDSSs can be used to automatically predict the risk of a delirium in individual hospitalised patients' by exclusively using electronically available predictive variables. To increase the use and improve the quality of predictive models, clinical risk factors should be documented ready for automated use.

Utility of Pregestational Body Mass Index and Initial Fasting Plasma Glucose in Predicting Gestational Diabetes Mellitus

OBJECTIVE

The present study aimed to evaluate the pregestational body mass index (preBMI) and initial fasting plasma glucose (FPG) in predicting gestational diabetes mellitus (GDM) in southern Chinese women.

STUDY DESIGN

A total of 327 pregnant women were recruited from the third affiliated hospital of Sun Yat-Sen University, Guangzhou, China. The preBMI and initial FPG at 16-18 weeks' gestation were measured. Oral glucose tolerance test was performed at 24-28 weeks' gestation. The sensitivity and specificity of preBMI and initial FPG as predictors for GDM were evaluated by receiver-operator characteristic curve analysis.

RESULTS

Both preBMI and initial FPG correlated with the 0-hour, 1-hour and 2-hour plasma glucose during oral glucose tolerance test (P < 0.05). The area under receiver-operator characteristic curve was 0.63 (95% CI: 0.57-0.68) for preBMI and 0.68 (95% CI: 0.61-0.72) for initial FPG in diagnosing GDM. The optimal cutoff for preBMI was 21.5 kg/m(2) (sensitivity 52.1% and specificity 69.2%) and 4.6 mmol/L (sensitivity 64.6% and specificity 65.2%) for initial FPG. Interestingly, the initial FPG had a better sensitivity compared to preBMI when the specificity was the same. Multivariate logistic regression analysis showed that initial FPG but not preBMI was the independent risk factor for the later development of GDM. After adjustment for the preBMI and the maternal age, the odds ratios of initial FPG and parity were 3.57 (95% CI: 1.72-7.45) and 2.11 (95% CI: 1.20-3.72).

CONCLUSIONS

Although both preBMI and initial FPG could be used as indicators for GDM, the initial FPG may be more suitable for predicting GDM in southern Chinese women.

Can neonatal sepsis be predicted in late preterm premature rupture of membranes? Development of a prediction model

OBJECTIVE

Women with late preterm premature rupture of membranes (PROM) have an increased risk that their child will develop neonatal sepsis. We evaluated whether neonatal sepsis can be predicted from antepartum parameters in these women.

STUDY DESIGN

We used multivariable logistic regression to develop a prediction model. Data were obtained from two recent randomized controlled trials on induction of labor versus expectant management in late preterm PROM (PPROMEXIL trials, (ISRCTN29313500 and ISRCTN05689407). Data from randomized as well as non-randomized women, who consented to the use of their medical data, were used. We evaluated 13 potential antepartum predictors for neonatal sepsis. Missing data were imputed. Discriminative ability of the model was expressed as the area under the receiver operating characteristic (ROC) curve and a calibration with both a calibration plot and the Hosmer and Lemeshow goodness-of-fit test. Overall performance of the prediction model was quantified as the scaled Brier score.

RESULTS

We studied 970 women. Thirty-three (3.4%) neonates suffered neonatal sepsis. Maternal age (OR 1.09 per year), maternal CRP level (OR 1.01 per mmol/l), maternal temperature (OR 1.80 per °C) and positive GBS culture (OR 2.20) were associated with an increased risk of neonatal sepsis. The model had an area under the ROC-curve of 0.71. The model had both a good calibration and accuracy.

CONCLUSIONS

Antepartum parameters aid in the more precise prediction of the risk of neonatal sepsis in women with late preterm PPROM.

Effect of weather on medical patient volume at Kansas Speedway mass gatherings

INTRODUCTION

Provision for the safety and health care of persons attending mass-gathering events presents unique challenges to organizers. This study was designed to determine the factors that contribute to patients seeking medical care during these events.

METHODS

We performed a retrospective review of patient care records for visits that occurred during race weekends at the Kansas Speedway from April 2007 to October 2010. Data were collected regarding the overall gathering size of each event to calculate the number of patient encounters per 10,000 attendees. Patients' final disposition was determined to calculate the transfer-to-hospital rate per 10,000 attendees. Weather data, including temperature, humidity, and precipitation, were documented for each event. Negative binomial regression was used to test the relationship between weather factors and the rate of patient encounters.

RESULTS

Twenty-two event days over 6 race weekends were evaluated, with a total of 1305 patients (58% male; mean age: 37 years), a mean patient encounter rate of 13 per 10,000 attendees, and a mean transfer-to-hospital rate of 0.24 per 10,000 attendees. Our regression model demonstrated that each 0.55°C (1°F) increase in daily mean temperature was associated with a 4% increase in the rate of total complaints (P = .03) and a 6% increase in major trauma presentations (P = .019). Major trauma events were 2.4 times more frequent at ambient temperatures >17.2°C (63°F) (P = .03). Each inch of precipitation was associated with a 61% decrease in total patient volume (P = .05).

CONCLUSION

Weather factors significantly and predictably affect the use of medical services at the Kansas Speedway. Such data regarding mass-gathering events can be used for resource planning.