Acne has several effects on physical symptoms, but the main impacts are on the quality of life, which can be improved by treatment. There are several acne treatments but less evidence comparing their relative efficacy. Thus, we assessed the comparative efficacy of pharmacological and nonpharmacological interventions for acne. We searched PubMed, Embase, and the Cochrane Central Register of Controlled Trials from inception to April 2019, to include randomized controlled trials for acne that compared topical antibiotics (TA), benzoyl peroxide (BPO), topical retinoids (TR), oral antibiotics (OA), lasers, light devices including LED device (LED), photodynamic therapy (PDT), and intense pulsed light, chemical peels (CP), miscellaneous therapies or complementary and alternative medicine (MTCAM), or their combinations. We performed Bayesian network meta-analysis with random effects for all treatments compared with placebo and each other. Mean differences (MDs) of lesions count and risk ratios of adverse events with their 95% credible intervals (CrIs) were calculated, and all interventions were ranked by the Surface Under the Cumulative Ranking (SUCRA) values. Additional frequentist additive network meta-analysis was performed to detect the robustness of results and potential interaction effects. Sensitivity analyses were carried out with different priors, and metaregression was to adjust for nine potential effect modifiers. In the result, seventy-three randomized controlled trials (27,745 patients with mild to moderate acne), comparing 30 grouped intervention categories, were included with low to moderate risk of bias. For adverse effects, OA had more risk in combination treatment with others. For noninflammatory lesions reduction, seventeen interventions had significant differences comparing with placebo and three interventions (TR+BPO: MD = −21.89, 95%CrI [−28.97, −14.76]; TR+BPO+MTCAM: −22.48 [−34.13, −10.70]; TA+BPO+CP: −20.63 [−33.97, −7.13]) were superior to others with 94, 94, and 91% SUCRA values, respectively. For inflammatory lesions reduction, nineteen interventions were significantly better than placebo, and three interventions (TR+BPO: MD = −12.13, 95%CrI [−18.41, −5.80]; TR+BPO+MTCAM: −13.21 [−.39, −3.04]; LED: −11.30 [−18.34, −4.42]) were superior to others (SUCRA: 81, 81, and 77%, respectively). In summary of noninflammatory and inflammatory lesions results, TR+BPO and TA+BPO were the best options compared to others. The frequentist model showed similar results as above. In summary, current evidence supports the suggestion that TR+BPO and TA+BPO are the best options for mild to moderate acne. LED is another option for inflammatory lesions when drug resistance occurs. All the combinations involved with OA showed more risk of adverse events than others. However, the evidence of this study should be cautiously used due to the limitations.

Objective: This study aimed to construct an intelligent prescription-generating (IPG) model based on deep-learning natural language processing (NLP) technology for multiple prescriptions in Chinese medicine. Materials and Methods: We selected the Treatise on Febrile Diseases and the Synopsis of Golden Chamber as basic datasets with EDA data augmentation, and the Yellow Emperor's Canon of Internal Medicine, the Classic of the Miraculous Pivot, and the Classic on Medical Problems as supplementary datasets for fine-tuning. We selected the word-embedding model based on the Imperial Collection of Four, the bidirectional encoder representations from transformers (BERT) model based on the Chinese Wikipedia, and the robustly optimized BERT approach (RoBERTa) model based on the Chinese Wikipedia and a general database. In addition, the BERT model was fine-tuned using the supplementary datasets to generate a Traditional Chinese Medicine-BERT model. Multiple IPG models were constructed based on the pretraining strategy and experiments were performed. Metrics of precision, recall, and F1-score were used to assess the model performance. Based on the trained models, we extracted and visualized the semantic features of some typical texts from treatise on febrile diseases and investigated the patterns. Results: Among all the trained models, the RoBERTa-large model performed the best, with a test set precision of 92.22%, recall of 86.71%, and F1-score of 89.38% and 10-fold cross-validation precision of 94.5% ± 2.5%, recall of 90.47% ± 4.1%, and F1-score of 92.38% ± 2.8%. The semantic feature extraction results based on this model showed that the model was intelligently stratified based on different meanings such that the within-layer's patterns showed the associations of symptom–symptoms, disease–symptoms, and symptom–punctuations, while the between-layer's patterns showed a progressive or dynamic symptom and disease transformation. Conclusions: Deep-learning-based NLP technology significantly improves the performance of IPG model. In addition, NLP-based semantic feature extraction may be vital to further investigate the ancient Chinese medicine texts.

Context The role of traditional Chinese medicine injections (TCMIs) in diabetic foot (DF) has not been well estimated. Objective To evaluate the clinical effective rate, safety, and the financial cost of TCMIs in treating DF and ulcer wound healing. Methods We searched PubMed, Embase, CENTRAL, China National Knowledge Infrastructure (CNKI), VIP database, and Wanfang database from inception to May 2018 to find all randomized control trials (RCTs) related to TCMIs in DF treatment. The search items were “Traditional Chinese Medicine Injection” AND “Diabetic foot or Diabetic foot ulcer” AND “random”. Study Selection and Synthesis Only RCTs of TCMIs combined conventional therapies versus conventional therapies and that can be quantitatively synthesized were included. Finally, 17 studies and 1294 participants were included after extraction. Two investigators independently extracted and analyzed the data using RevMan5.3 software. Results The overall clinical effective rate of TCMI groups is higher than that of control groups [RR=1.27, 95CI % (1.20, 1.34), P<0.00001] based on fixed effect model analysis. Regarding motor nerve conduction velocity of median nerve and peroneal nerve, TCMI group showed a significant improvement (MD=3.84[2.28, 5.41], P<0.00001; MD=2.89[0.63, 5.15], P=0.01). Regarding plasma viscosity TCMI group showed a statistically difference (MD=0.27[0.04, 0.49], P=0.02). In terms of blood viscosity at high shear rate, there was an improvement of TCMI group (MD=0.36[0.05, 0.67], P=0.02). However, sensory nerve conduction velocity of peroneal nerve and median nerve showed a contradiction to motor nerve conduction velocity, respectively (MD=2.59[-1.69, 6.87], p=0.24; MD=2.73[-0.96, 6.43], P=0.15). Conclusion The data of this study shows that TCMIs can bring benefits to patients with diabetic foot. However, due to low methodological quality of included RCTs, more rigorous designed RCTs with large sample size are recommended to provide more high-quality evidence.

BackgroundTo provide evidence for medical management of chronic idiopathic constipation (CIC) in China based on comparisons of all clinical practical interventions using Bayesian network meta-analysis.MethodsWe conducted a systematic literature review by searching PubMed, Embase, Cochrane Central, the China National Knowledge Infrastructure (CNKI), and the Wanfang Database (inception to May 2019) for randomized controlled trials (RCTs) for CIC in Chinese people. Only RCTs that recruited participants aged over 18 and diagnosed with CIC by the Rome II, III or IV criteria were included. We used three outcomes to examine efficacy. The risk ratio (RR) of the responder rate, based on ≥3 spontaneous bowel movements (SBMs) per week after treatment, was the primary outcome, and the SBM count per week and the Bristol score (BS) were secondary outcomes. In addition, adverse effects (AEs) were also considered a secondary outcome to evaluate safety. We conducted Bayesian network meta-analysis with random effects, and the RR or mean difference with its 95% credible interval was calculated. In addition, we ranked all treatments via their cumulative curves (SUCRA) and assessed the quality of evidence according to the GRADE criteria.ResultsWe included a total of 42 trials (6820 participants) of 20 grouped interventions that included pharmacological and nonpharmacological treatments. For the primary outcome, fourteen interventions were significantly better than placebo, and Probiotics plus Mosapride (PB + MP) appeared superior to others (GRADE quality of evidence: Moderate to Low), followed by Prucalopride (PP) (High to Low) and Electroacupuncture (EA) (High to Low). For SBM, Compound sodium bicarbonate suppository (CSBS) appeared to be best, with an SUCRA value of 90% (High to Low). For BS, Lactulose plus Probiotics (LT + PB) was superior to others (Moderate to Low), followed by Polyethylene glycol (PEG) (High to Moderate). Although all interventions appeared non-significant when compared with placebo in terms of adverse effects, Lactulose plus Mosapride showed greater risk than others on ranking probability.ConclusionsGiven the GRADE assessment, PB + MP, PP and EA may be the priory options with moderate certainty in the quality of evidence for the primary outcome. For SBM, a CSBS may be the best option with moderate certainty in the quality of evidence. For BS, PEG may be the priory option with high certainty in the quality of evidence. However, due to a lack of high certainty in the quality of evidence, caution is needed when recommending the interventions. Because of the limitations, an increased number of trials are required for more accurate results.

Recently, there is an increasing number of clinical trials on Traditional Chinese medicine (TCM) published, but the implementation of Clinical Trial Registration (CTR), Ethical Review (ER), and Informed Consent (IC) in clinical trials of TCM is unclear. This study aims to investigate the status of CTR, ER, and IC in clinical trials of TCM. Clinical trials of TCM published in 10 high-quality Chinese journals in 2016 were selected as a sample. Information of clinical trial registration, ethical review, and informed consent of clinical trials was extracted for analysis. Two authors independently screened the literature and extracted the relevant information. A total of 659 clinical trials met the criteria and were included for analysis. Only 9 clinical trials reported information of clinical trial registration (1.4%). The number for ethical review and informed consent were 156 (23.7%) and 502 (76.2%). Trial registration, protocol approval, and informed consent were not well executed. Especially registration and ethical review of clinical trials in TCM should be carefully concerned by researchers, clinicians, and journal editors. Training on methodology of clinical trial should be strengthened.