There is a broad misconception that scientists are sociopaths, torturing animals for sheer enjoyment and have free reign to inflict as much pain as is humanly possible. In actuality, nothing could be further from the truth. Any and all animal work conducted on vertebrate animals is heavily regulated, and rightly so. As an animal researcher one of my top priorities is to make sure that I treat the animals as compassionately as I can, and this is a shared priority among most members of the research community.
Any experiments to be conducted are written up for approval by the institution's Institutional Animal Care and Use Committee (IACUC). The IACUC protocol is then examined by an impartial board of volunteers, and is either passed so that the experiments can begin, or is returned asking for amendments. In many institutions, the IACUC protocol can go through several round of amendments before being approved.
Here are some of the things we have to address:
-Number of animals used. We must use enough animals for the study to be statistically significant, but also not use any more than are necessary.
-Minimization of pain and suffering. Any procedures conducted must be done so in a manner that minimizes discomfort, physical or emotional, that the animals may feel.
-Utilization of the least sentient animal possible. We are required to use the least intellectually, physically and emotionally developed species of animal as is possible for our studies. In most cases we will use mice or rats, although for some studies more advanced species are required.
-Humane end point. If at any time in the study an animal's suffering becomes too great, we must euthanize the animal.
-Relevance and dosage of any medications or experimental drugs, and their potential adverse effects on the animals.
-General significance of the study, and whether it will make a genuine contribution to the pre-existing body of knowledge.
After the study is approved, researchers are monitored by animal facility staff frequently, and periodic inspections are required. Any failure to adhere to the protocol or other forms of abuse are generally very quickly addressed, and depending upon the nature of the offense, can have a wide variety of consequences. In general, researchers don't want to enter that kind of territory, so strive to adhere to all regulations. Most breaches are accidental, although there are rare exceptions.
Although many anti-animal research activists say (and believe) that animal research is a free-for-all in which scientists can inflict any imaginable torture on lab animals, this is simply untrue. Early animal research was not well-regulated, and it's true that much undue suffering occurred. However, any activity with animals is now very regulated and is constantly monitored.
As scientists, we can sometimes grumble about the additional time it takes to file an IACUC protocol, and be subjected to inspections and testing. However, I believe that these things are relevant and necessary, and I am most certainly not alone. The scientific community as a whole is very cognizant of laboratory animals' needs, and strives to meet those needs as much as is possible.
