Search

Random allocation is commonly used in medical researches, and has become an essential part of designing clinical trials. It produces comparable groups with regard to known or unknown prognostic factors, and prevents the selection bias which occurs due to the arbitrary assignment of subjects to groups. It also provides the background for statistical testing. Depending on the change in allocation probability, random allocation is divided into two categories: fixed allocation randomization and dynamic allocation randomization. In this paper, the author briefly introduces both the theory and practice of randomization. The definition, necessity, principal, significance, and classification of randomization are also explained. Advantages and disadvantages of each randomization technique are further discussed. Dynamic allocation randomization (Adaptive randomization), which is as yet unfamiliar with the anesthesiologist, is also introduced. Lastly, the methods and procedures for random sequence generation using Microsoft Excel is provided. [ABSTRACT FROM AUTHOR]

Propensity score matching method (PSM) is widely used in observational study to reduce selection bias. Observational study lacks randomization, hence, statistical inferences without bias adjustments usually include observed or unobserved effects of covariates. If a subject with specific characteristics has a higher chance to be selected for a specific treatment, the characteristics have a possible effect on statistical results. PSM is the method for controlling covariate imbalance that produces the selection bias. In this paper, we introduce the basic concepts of PSM and simplified methods of PSM process. However, PSM is a rapidly developing statistical area with many limitations and some disadvantages. These points are described in the concluding section to emphasize the importance of considering the various features of PSM in the study design. [ABSTRACT FROM AUTHOR]

Background: This study aimed to systematically review the literature investigating the effect of intravenous (IV) tranexamic acid (TXA) on the perioperative blood loss and amount of transfused blood in patients who underwent spine surgery under general anesthesia. Methods: MEDLINE, EMBASE, and CENTRAL databases were searched for studies published up to June 2017. Randomized controlled trials comparing the intervention of tranexamic acid use with a placebo in corrective spine fusion surgery were included. Outcome measures included intraoperative and postoperative blood loss and perioperative blood transfusion volume. We also conducted subgroup analyses according to the dose of administration of intervention drug. Results: In total, 12 studies involving 875 patients were included in the final analysis. Blood loss decreased in patients treated with perioperative IV TXA by a mean volume of 189.58 ml intraoperatively (95% confidence interval [CI]: 135.82 to 243.34 ml, P < 0.001; I2 = 74%), 121.04 ml postoperatively (95% CI: 84.92 to 157.17 ml, P < 0.001; I2 = 0%), and 285.97 ml in total (95% CI: 190.70 to 381.23 ml, P < 0.001; I2 = 0%). Perioperative transfusion volume also decreased by 162.10 ml in patients who received IV TXA (95% CI: 31.77 to 292.44 ml, P = 0.010; I2 = 47%). Conclusions: Blood loss was reduced in patients who were administered TXA during spine surgery. Additionally, the perioperative volume of blood transfusion decreased in patients who received TXA. Further research is required to confirm the optimal dosage and timing of TXA usage. [ABSTRACT FROM AUTHOR]

A noteworthy change in recent medical research is the rapid increase of research using big data obtained from electrical medical records (EMR), order communication systems (OCS), and picture archiving and communication systems (PACS). It is often difficult to apply traditional statistical techniques to research using big data because of the vastness of the data and complexity of the relationships. Therefore, the application of artificial intelligence (AI) techniques which can handle such problems is becoming popular. Classical machine learning techniques, such as k-means clustering, support vector machine, and decision tree are still efficient and useful for some research problems. The deep learning techniques, such as multi-layer perceptron, convolutional neural network, and recurrent neural network have been spotlighted by the success of deep belief networks and convolutional neural networks in solving various problems that are difficult to solve by conventional methods. The results of recent research using artificial intelligence techniques are comparable to human experts. This article introduces technologies that help researchers conduct medical research and understand previous literature in the era of AI. [ABSTRACT FROM AUTHOR]