Good practice includes engagement in principles such as the NC3Rs[1], and the ARRIVE guidlines[2] and is important in animal research not only in terms of ethical treatment of the animals but also to provide a scientifically valid study. This is because the minimum amount of stress posed on the animal as possible, will reduce the effect on results. Good practice It is necessary to critique the study design of papers, not only to ensure correct ethical treatment of animals but also to ensure the research is valid. “Bile secretion and liver regeneration in partially hepatectomized rats”, 1993, aimed to show that there was a correlation between bile secretion and liver regeneration. However, the validity of these results is questionable due to the multiple improvements that could have been made.
The article begins with a short introduction stating that there was a lack of understanding, due to the scant previous literature on the subject, of the correlation between bile secretion and liver regeneration. It was then also clearly pointed out that the purpose was to investigate this correlation as well as physiological changes in bile salt kinetics. This introduction would have benefited from the addition of previous relevant data or scientific background, which would have aided in providing a null and an alternative hypothesis, neither of which were specified. Although there was a preconceived notion that a significant alteration in bile secretion was expected after hepatectomy, the lack of a written hypothesis prevents the results from validating or falsifying a theory, which decreases the reliability of a claim that any relationship or pattern in the results exists. The scientific background and hypothesis would also have been useful in justifying the experimental approach, which was not done, therefore making it unclear how the model addressed the scientific objective and how it was relevant to human biology. Overall the lack of scientific evidence and rationale behind the rat model could mean that replacement (of the 3 R’s) had not been put into practice because it was unstated why the animal model was necessary over a non-animal design.
The materials and methods section begins with information on the animals used in the study, though an opening ethical statement was absent. This is needed to indicate the nature of research permissions from an ethics committee, relevant licenses and which guidelines were used throughout the research. The animal species, sex, source of the animals, and body weight range were all indicated, though it was not said if and why any of these traits were necessary for the experiment, for example it should have been stated why only male rats were used, and why a body weight over 250g was chosen. Age range was not stated at all. Without a rationale for these characteristics, it is not known (and has not been stated) whether the sample accurately reflects the population. If in fact the sample was a poor representation of the population, the validity of the results is compromised, increasing the need for repetition of the study and the need more animals to be used. Therefore, refinement was not being applied where possible in this instance.
Animals were kept in identical housing units with the same room temperature, food and water (to which they had free access to) and a controlled light cycle. Although this is important to mention, there aren’t any specifics actually talked about here. For example, they should have specified if each rat was kept in a single unit or if they were kept in groups, specified the temperature, mentioned which rat chow was used and where it was sourced, and the specifics of the light cycle. There was no identification of the type of cage or bedding material, both of which can impact the anxiety of the animals. The aim should be to make the living conditions as comfortable as possible not only to uphold animal welfare but also because increased anxiety could impact the results of the experiment. The fact that the animals had free access to rat chow is also an issue because amount of food eaten greatly impacts bile secretion; the same amount should have been given to each rat in order to control this variable. Again, these factors are reducing validity of the results and therefore reducing the amount of refinement of the study. The lack of detail here, once more, prevents the study from being repeated by others.
The study design reveals that 8 rats were designated to each of the 3 groups, but there was no evident sample size calculation used to show why 8 rats in particular were needed. Without evidence to prove otherwise, it is possible that more animals than necessary were used here (reduction was not being applied). It was also indicated that more rats were used in 3 sub groups, to measure liver weight on days 1, 3, 5 and 7, and 5 rats were killed on each day. This is a very vague statement with multiple details lacking; there is no indication of how many animals were used all together or in each group. It is also unclear whether the three subgroups mirror the original three sample groups in terms of the hepatectomies performed and there is also no reasoning for why 5 animals were killed each day. Without any of this information, the results from the ‘sub-experiment’ cannot be well understood and again, there cannot be repetition of the study. It also appears no randomisation procedure was used to minimise subjective bias when allocating animals to their sample groups; a completely random model would have been sufficient for this study, considering the fact that there were no female rats used, in which case a block design would have been useful. Again, without details described here and evidence for a randomisation technique, the validity of the results is forfeited.
Experimental procedures should be explained as precisely as possible for similar reasons as mentioned previously; for study repetition and refinement of the study design. The sample collection included details of where the animals underwent the procedure, and also the specific timings of bile collection, however, no details were given as to where or at which time the hepatectomies were carried out, which is necessary for repetition of the experiment. For the drainage technique, it was claimed that samples were collected in 30-minute intervals according to the standard biliary drainage technique described by Myant and Eder. However, the original procedure claimed to collect in 2.5 hour intervals[6] which is a considerable difference and therefore requires a rationale as well as evidence that it was not more harmful or painful for the animals. If there was no valid evidence for this, then it should not have been performed. Bile return after drainage was also reported, and although it appears to be returning the animal to a normal state, suggesting a reduction of stress, the duodenal cannula is a very invasive technique and it is highly possible that the stress caused had an effect on the results regardless. Furthermore, the animals were given no analgesics to reduce pain after surgery and during bile drainage, increasing stress further. This not only affects the results but is also a disregard for animal welfare.
Another very important detail was that ether anaesthesia was used, although there was no rationale behind it described. Being an explosive,[3] the location of the ether anaesthetic procedure (ideally a fume cabinet) should have been made evident to show that the risk of fires/explosions was considered. The dosage was also not stated, which is an issue for multiple reasons in regard to refinement of the experiment; it was performed on non-fasting rats which require a higher dose than fasting, ether is extremely irritable to the airways, and also has a risk of liver necrosis.[4] Therefore, it should have been explained why it was chosen over less risk-associated anaesthetics, and if there was no valid reason for this then other anaesthetics should have been used. There was also the fact that there was no monitoring of the rats whilst they were anaesthetised, which again is important in terms of animal welfare (refinement) due to the potential for myocardial depression and hence death.[4] Monitoring is also important for maintenance of the anaesthesia; the animal should have been checked for a response to aversive stimuli such as a tail pinch, for example.[5]
Due to the lack of rationale behind much of the study design and a disregard for ethical treatment of the animals in multiple instances, it can be concluded that the results are somewhat invalid in this paper. In the case of unreliable results, a repetition of the study would be needed, which would require the use of more animals. That being said, there are multiple details of the methodology, which are required for repetition, that have not been given in the report. Given that the 3Rs were constructed in 1959[7], and the report was written in 1961, there is no reason that they should not have been implemented, regardless of the lack of ARRIVE guidelines until 2010.[8] Therefore, with possibly void results, and methodology which cannot be repeated, the value of the study itself is debatable.