Warfarin maintenance dose prediction for Chinese after heart valve replacement by a feedforward neural network with equal stratified sampling.

Patients requiring low-dose warfarin are more likely to suffer bleeding due to overdose. The goal of this work is to improve the feedforward neural network model's precision in predicting the low maintenance dose for Chinese in the aspect of training data construction. We built the model from a resampled dataset created by equal stratified sampling (maintaining the same sample number in three dose-groups with a total of 3639) and performed internal and external validations. Comparing to the model trained from the raw dataset of 19,060 eligible cases, we improved the low-dose group's ideal prediction percentage from 0.7 to 9.6% and maintained the overall performance (76.4% vs. 75.6%) in external validation. We further built neural network models on single-dose subsets to invest whether the subsets samples were sufficient and whether the selected factors were appropriate. The training set sizes were 1340 and 1478 for the low and high dose subsets; the corresponding ideal prediction percentages were 70.2% and 75.1%. The training set size for the intermediate dose varied and was 1553, 6214, and 12,429; the corresponding ideal prediction percentages were 95.6, 95.1%, and 95.3%. Our conclusion is that equal stratified sampling can be a considerable alternative approach in training data construction to build drug dosing models in the clinic.

The Prediction Model of Warfarin Individual Maintenance Dose for Patients Undergoing Heart Valve Replacement, Based on the Back Propagation Neural Network.

BACKGROUND AND OBJECTIVE: Because of the narrow therapeutic window and huge inter-individual variation, the individual precision on anticoagulant therapy of warfarin is challenging. In our study, we aimed to construct a Back Propagation Neural Network (BPNN) model to predict the individual warfarin maintenance dose among Chinese patients who have undergone heart valve replacement, and validate its prediction accuracy. METHODS: In this study, we analyzed 13,639 eligible patients extracted from the Chinese Low Intensity Anticoagulant Therapy after Heart Valve Replacement database, which collected data on patients using warfarin after heart valve replacement from 15 centers all over China. Ten percent of patients who were finally enrolled in the database were used as the external validation, while the remaining were randomly divided into the training and internal validation groups at a ratio of 3:1. Input variables were selected by univariate analysis of the general linear model; 2.0, the mean value of the international normalized ratio (INR) range 1.5-2.5, was used as the mandatory variable. The BPNN model and the multiple linear regression (MLR) model were constructed by the training group and validated through comparisons of the mean absolute error (MAE), mean squared error (MSE), root mean squared error (RMSE), and ideal predicted percentage. RESULTS: Finally, 10 input variables were selected and a three-layer BPNN model was constructed. In the BPNN model, the value of MAE (0.688 mg/day and 0.740 mg/day in internal and external validation, respectively), MSE (0.580 mg/day and 0.599 mg/day in internal and external validation, respectively), and RMSE (0.761 mg/day and 0.774 mg/day in internal and external validation, respectively) were achieved. Ideal predicted percentages were high in both internal (63.0%) and external validation (59.7%), respectively. Compared with the MLR model, the BPNN model showed a higher ideal prediction percentage in the external validation group (59.7% vs. 56.6%), and showed the best prediction accuracy in the intermediate-dose subgroup (internal validation group: 85.2%; external validation group: 84.7%) and a high predicted percentage in the high-dose subgroup (internal validation group: 36.2%; external validation group: 39.8%), but poor performance in the low-dose subgroup (internal validation group: 0%; external validation group: 0.3%). Meanwhile, the BPNN model showed better ideal prediction percentage in the high-dose group than the MLR model (internal validation: 36.2% vs. 31.6%; external validation: 42.8% vs. 37.8%). CONCLUSION: The BPNN model shows promise for predicting the warfarin maintenance dose after heart valve replacement.

Estimating benefit equity of government health subsidy in healthcare Services in Shandong Province, China: a cross-sectional study.

BACKGROUND: Government health subsidy (GHS) is an effective tool to improve population health in China. Ensuring an equitable allocation of GHS, particularly among the poorer socio-economic groups, is a major goal of China's healthcare reform. The paper aims to explore how GHS was allocated across different socioeconomic groups, and how well the overall health system was performing in terms of the allocation of subsidy for different types of health services. METHODS: Data from China's National Health Services Survey (NHSS) in 2013 were used. Benefit incidence analysis (BIA) was applied to examine if GHS was equally distributed across income quintile. Benefit incidence was presented as each quintile's percentage share of total benefits, and the concentration index (CI) and Kakwani index (KI) were calculated. Health benefits from three types of healthcare services (primary health care, outpatient and inpatient services) were analyzed, separated into urban and rural populations. In addition, the distribution of benefits was compared to the distribution of healthcare need (measured by self-reported illness and chronic disease) across income quintiles. RESULTS: In urban populations, the CI value of GHS for primary care was negative. (- 0.14), implying an allocation tendency toward poor region; the CI values of outpatient and inpatient services were both positive (0.174 and 0.194), indicating allocation tendencies toward rich region. Similar allocation pattern was observed in rural population, with pro-poor tendency of primary care service (CI = - 0.082), and pro-rich tendencies of outpatient (CI = 0.153) and inpatient services (CI = 0.203). All the KI values of three health services in urban and rural populations were negative (- 0.4991,-0.1851 and - 0.1651; - 0.482, - 0.247and - 0.197), indicating that government health subsidy was progressive and contributed to the narrowing of economic gap between the poor and rich. CONCLUSIONS: The inequitable distribution of GHS in China exited in different healthcare services; however, the GHS benefit is generally progressive. Future healthcare reforms in China should not only focus on expanding the coverage, but also on improving the equity of distribution of healthcare benefits.

Shengmai (a traditional Chinese herbal medicine) for heart failure.

BACKGROUND: Heart failure is a major public health problem worldwide. Shengmai, a traditional Chinese herbal medicine, has long been used as a complementary treatment for heart failure in China. This is an update of a Cochrane Review published in 2012. OBJECTIVES: To determine the effect (both benefits and harms) of Shengmai in treatment of people with heart failure. SEARCH METHODS: We searched CENTRAL on The Cochrane Library (Issue 5 of 12, April 2013); DARE on The Cochrane Library (Issue 2 of 4, April 2013); MEDLINE (1948 to June Week 1 2013); EMBASE (1980 to 2013 Week 23); AMED (1985 to August 2008); BIOSIS (1969 to 7 June 2013); CBM (1978 to June 2013); VIP (1989 to June 2013); and CNKI (1979 to June 2013). We also handsearched Chinese journals and did not apply any language restrictions. SELECTION CRITERIA: We included randomised controlled trials (RCTs) of Shengmai plus usual treatment for heart failure versus usual treatment alone, or Shengmai versus placebo, irrespective of blinding status. In this update we only included studies with a clear description of randomisation methods and classified as true RCTs. DATA COLLECTION AND ANALYSIS: Two authors independently selected trials, assessed methodological quality and extracted data. We calculated dichotomous data as risk ratios (RRs) and continuous data as mean differences (MDs) or standardized mean differences (SMDs) with corresponding 95% confidence intervals (CIs). We used a fixed-effect model to perform meta-analysis for outcomes without heterogeneity; and a random-effects model to perform meta-analysis for outcomes with heterogeneity. MAIN RESULTS: We included a total of 14 RCTs (858 patients) in this review update, four of which were new trials. Of these 14 RCTs, 11 trials compared Shengmai plus usual treatment with usual treatment alone, and three trials compared Shengmai with placebo. Improvement of NYHA functional classification was more common in patients taking Shengmai plus usual treatment than in those receiving usual treatment alone (RR 0.37; 95% CI 0.26 to 0.51; 10 trials, 672 participants; low quality evidence). Beneficial effects of Shengmai in treating heart failure were also observed in other outcomes, including exercise test, ejection fraction and cardiac output. The three RCTs (106 patients) comparing Shengmai with placebo reported improvement in NYHA functional classification and in stroke volume. Three of the 14 RCTs reported a total of six patients with mild adverse effects and two were withdrawn due to the adverse effects. The adverse events rate was 1.21%. AUTHORS' CONCLUSIONS: Shengmai may exert a positive effect on heart failure, especially for improving NYHA functional classification when Shengmai plus usual treatment is used. The review results should be interpreted with caution due to the high risk of bias of the included studies (particularly regarding allocation concealment and blinding), the small sample size of these studies, and the significant heterogeneity in outcomes such as ejection function, cardiac output and stroke volume. There was no evidence available concerning the effect of Shengmai on mortality, and more high quality studies with long-term follow-up are warranted.