动物模型 z-bo 2020-08-14 513

　　Animal experiments are usually conducted by studying the biological characteristics of a certain species, strain, or sex animal model to indirectly infer whether humans or other research objects have similar biological characteristics. Such philosophical extrapolation often brings certain errors, or even serious ones. Bias. In order to reduce this risk and enable the results of animal experiments to be reasonably extrapolated to humans, careful animal experiment design is particularly important. The correct selection of animal models for human diseases is a prerequisite to ensure the smooth progress of animal experiments. The next thing to consider is the principles that must be followed in the use of human disease animal models to carry out biomedical research, including the three principles of control, randomization, and repetition.

　　(1) the principle of comparison

　　"Control" is the primary principle of animal experiment design. Animal model related experiments also need to have controls, and comparisons can be used for identification. Controls are the basis for comparison. Except for the factors to be observed and processed, all other conditions that affect the effect indicators should be the same as possible in the experimental group and the control group. There must be a high degree of comparability in order to exclude the influence of confounding factors and make scientific conclusions about the experimental observation items. There are many types of comparisons, the following are commonly used:

　　1. Blank control The control group does not impose any treatment factors. This kind of control is often used to evaluate the success of animal models. This method is simple and easy to implement, but it may cause psychological differences between the experimental group and the control group, thereby affecting the determination of experimental effects. In some scientific experiments using mature animal models, sometimes a blank control group may not be set up according to specific circumstances.

　　2. Experimental condition control The control group does not apply treatment factors, but applies certain experimental conditions that are the same as the treatment factors. Experimental conditions include operating technology, solvent or capacity of the tested factors, etc. As in some animal experiments related to surgery. A sham operation group is often set up. This method should be used for all experimental conditions that affect the experimental effect.

　　3. Standard control is also called positive control. Use existing standard methods or conventional methods for comparison. For example, in pharmacological research, use known effective drugs, effective therapies or recognized standard therapies as controls.

　　(2) Randomization Principle

　　In animal experiments, in addition to processing factors, other non-processing factors that may produce confounding effects between each group are required to be as consistent as possible in each group (control and experimental groups) to maintain the balance of each group. Implementing the principle of randomization is an important means to improve the balance between groups, and it is also a prerequisite for statistical inference during statistical analysis of data. The purpose of randomized sampling is to make each research object in the population have the same chance of being selected and assigned to the experimental group or the control group. In animal experimental research, complete random allocation or stratified random allocation is generally used. Most of the small animal experiments are first matched or matched groups, and then randomly allocated within the "pair" or "wu", but most large animals are stratified first, and then randomly allocated within the strata. The basic methods of randomized sampling include random mathematics table, calculator random mathematics, and lottery, etc. The researcher may decide according to the specific situation.

　　Below, give an example of random stratified grouping to illustrate the importance of stratified grouping. In order to use rabbits to study the anti-atherosclerotic properties of a certain drug, it is planned to divide the rabbits into a control group and an experimental group. A high-cholesterol diet (0.3% high-cholesterol diet for 16 weeks) will be used to induce atherosclerosis. We selected 24 rabbits for the experiment, first fed the rabbits with a higher concentration of high-cholesterol feed (such as 0.5% high-cholesterol feed) for 1 week to detect the plasma cholesterol level of the rabbits, and rule out animals that are not sensitive to or too sensitive to high-cholesterol diets (According to specific experiments to determine the standard, such as eliminating rabbits with plasma cholesterol levels lower than 300mg/dl and higher than 600mg/dl). There are 16 rabbits that meet the criteria, sorted according to the cholesterol level from low (314mg/dl) to high (590mg/dl) (Table 4-1). In order to be more balanced, we divided 8 levels (block numbers) according to the level of plasma cholesterol. Then use the random number table to randomly group within the layer (for example, the smaller random number is the control group). After grouping, the results show that the plasma cholesterol level of the control group (439mg/dl±94mg/dl) is very close to the experimental group (438mg/dl±95mg/dl), the standard deviation is also the same, and the balance within the group is basically reached, and the individual variation is balanced. Assigned to the group.