Swine have become popular animal models for preclinical trials for medical research because of their size and anatomical and physiological similarity to humans. For this reason, these animals are widely used for research on physiopathology and new surgical techniques. Over the last 20 years, swine have replaced dogs as the general surgical model for both training and research [1
A scientific and moral argument is that if pigs are used to ‘model’ human beings undergoing surgery, then they should receive the equivalent standards of perioperative care humans would; however, most bioscience journals provide little or no guidance on what information to report when describing animal research, and many details are omitted [2
]. Unfortunately, this might be a contributing reason why researchers have been increasingly unable to replicate the positive results from animal studies in the clinical trials that have followed [3
]. Ideally, scientific publications should present enough information to allow a knowledgeable reader to understand what, why, and how experiments were done and to assess the reliability and validity of findings [4
]. Omitting essential information might lead to scientific and ethical concerns and does not facilitate the reproducibility of the experimental conditions [5
The National Centre for the Replacement, Refinement and Reduction of Animals in Research (NC3Rs), a UK government-sponsored scientific organization, has led an initiative to produce guidelines for reporting animal research. In 2010, the Animal Research: Reporting of In Vivo Experiments (ARRIVE) guidelines were published [5
] to address the growing concerns with poor experimental design and lack of transparent reporting of in vivo experiments in published literature. These guidelines consist of a checklist of 20 categories that provide all the information that researchers should include in scientific publications using animals [6
Recent publications focused on the quality of reporting revealed very little improvement in reporting standards since the guidelines were introduced [7
]. Research with pigs has also presented minimum information published on topics about perioperative care (anesthetic protocols and pain alleviation) [2
]. Bradbury et al. (2016) found that reporting postoperative pain management in studies was remarkably low, reflecting either under-reporting or under-use of analgesics. For this reason, a systematic review was performed to evaluate the quality of reports on research in surgeries where swine were used as biomodels. We decided to analyze only the materials and methods section of each article, with a special emphasis on anesthetic and analgesic procedures, because this section describes the procedures that have a direct impact on animal welfare.
2. Materials and Methods
2.1. Search Strategy
A literature review was performed of articles that were published from January 2013 to December 2018. An internet search was performed using PubMed, Google Scholar, Scopus, and Science Direct as electronic databases. The keywords used were: Swine OR Pig OR Minipig AND Surgery.
2.2. Inclusion and Exclusion Criteria
Articles included had these characteristics:
Original articles published in English
Use of swine models in vivo
Studies that included surgery as an experimental procedure
Studies published between 2013 and 2018. This period was selected to sample the recent biomedical literature (for the last 5 years), considering a suitable period for the ARRIVE guidelines to be implemented
Studies that described painful experimental surgical procedures like skin incision, craniotomies, thoracotomies, laparotomies, laparoscopies, dental surgeries, and orthopedic surgeries
Review articles, commentaries, or communications were excluded. Studies without in vivo experiments were also excluded.
2.3. Evaluation of Publication Quality
We used the ARRIVE guidelines to analyze the articles, focusing on the “Material and Methods” section to evaluate the degree of compliance of publications with these guidelines.
We established a score based on 3 levels for the evaluation of the categories of the guidelines. They were defined as follows: Score 0: not mentioned, total absence of any type of information, Score 1: unclear/not complete, items not mentioned completely in the category assess, Score 2: adequate/clear, complete information for all items corresponding to the category evaluated.
Data were extracted from the articles and presented into tables, and the information of each ARRIVE’s category, subcategory, and evaluated item was calculated and expressed as percentage.
Forty-three subcategories or items were evaluated as the levels of greater detail indicated for the ARRIVE guidelines. The frequency for each year was calculated and expressed in percentage according to the maximum and minimum percentage presented for each subcategory or item.
The “Materials and Methods” section of research papers should provide basic information about how research was performed. Comprehensive reporting is essential to understand how investigations were undertaken, to interpret findings properly [7
], and to allow the reproduction of results if the same methodology will be applied in other studies. The information analyzed in this systematic review shows that many of these studies would not be reproducible, considering that important details were omitted.
After examining articles from 2013 (three years after the creation and implementation of the ARRIVE guide), we found that the report of all data describing the used methodology according to the ARRIVE guidelines did not improve through the years, even though 27 of the 81 journals consulted request the authors to use these guidelines. The outcomes obtained resemble those found by Barker [10
] et al. in 2014. These authors made an analysis of papers published in PLOS
journals and observed very little improvement in reporting standards since the ARRIVE guidelines were published in 2010. In fact, they observed in this review a better report of the methods in the years 2014 and 2015 (although they were incomplete in most cases), but then the quality of the reports decreased (2016 was the year with less reported data, followed by 2018 and 2013). The reproducibility of these studies in similar studies can thus be affected [2
The high percentage of reports of ethical review permissions is an indication that projects were previously reviewed by ethical institutional committees to ensure the welfare of the animals; however, many articles did not indicate the protocol numbers or the guides on which their protocols were based. These data should allow tracking the approved protocols and the guides that justify the care provided to the animals used in each investigation [12
The report of study design showed a relative increase in the number of subcategories of experimental and control groups and experimental units; however, a detailed description of the methods to avoid bias was undermentioned (it was present only in 27% of articles). The lack of randomization and blinding can affect the scientific validity because biased—both conscious and unconscious—factors with no relation to the biological action can influence the results; therefore, mentioning the way an experiment was done is critical [13
Considering the experimental procedures, the journals did not require important data such as the time at which the experiments were performed, in all the articles. Probably, because of the scope of these studies focused on surgical procedures’ description, this factor may not have been considered. The circadian cycle might change animal physiology, leading to different results, which makes it important to provide the time of day the experiments were done to ensure experiment reproducibility [14
]. Other little detailed data were the choice of specific medications, doses, and administration routes in swine undergoing surgery. According to its pharmacokinetics and pharmacodynamics, each medication presents a different effect on animal physiology, which may alter the results, so justifying their use is essential. Similarly, different routes of administration may lead to different absorption levels, and the effect generated may produce differences in the results obtained, so their choice should be well justified as well [15
The least mentioned data in the category of experimental animals were the source, sex, and race of the animals. These data should be mandatory in any report of experimental procedures using animals because genetic and/or hormonal characteristics of the animals might influence the results. Most articles also inadequately detailed information about housing and husbandry. Although the studies were not specifically on animal welfare issues, these data must be mentioned, so that other researchers can reproduce the same conditions. The experimental animals should not be unnecessarily stressed and should be kept under appropriately controlled conditions. Poor animal welfare is likely to result in poor science [7
Most of the studies reported the number of animals; however, almost none reported the statistical method used to calculate this number. Determining the sample size by power size or simple calculations helps to design animal research with an appropriate number of animals to detect important biological effects. Omission in reporting means potentially flaws on a research [16
The statistical methods were not mentioned in 24% of the studies. This section should not be omitted because it shows the way data were analyzed. Likewise, much of this information reported in journals was not detailed, and thus, its lack raises doubts on the validity of the statistical methods chosen to evaluate the results. While focusing on technically challenging research and on generating innovative science, many journals fail to ensure compliance with the basic standards of experimental design and data analysis. One solution to this problem is to have an additional statistical review of submitted manuscripts (as it is often done by journals in health sciences). In addition, learned societies might suggest methods of analysis of standard outcomes and data reporting to their members [10
All the surgical procedures generate pain at different intensities, according to tissue invasiveness [17
]. In this review, the reported surgeries mostly generated moderate to severe pain. Thoracotomy and orthopedic surgeries were the most invasive. On the contrary, dentistry was considered slightly less invasive, and skin incisions and craniotomies the least invasive procedures [18
]. That is why the use of effective anesthetic and analgesic protocols is essential to ensure the welfare of the animals, decreasing pain and stress.
In this review, the drugs used to anesthetize the animals were generally reported at a high level of (81%). The great variety of anesthetic protocols and administered doses showed the absence of standardization for these protocols. This may be due to the type of procedure developed and the local availability of products. However, it was reported that ketamine, which is classified as a N-methyl D-Aspartate (NMDA) antagonist that causes a dissociative anesthesia, [19
], was the most used drug in combination with other medications for both induction and anesthetic maintenance. Isoflurane was the second anesthetic administered alone or in combination with other drugs, probably because it allows the maintenance of the animal’s unconsciousness in a simple and long-lasting manner.
On the other hand, the under reporting of analgesia is a concern because it reveals an inadequate management of pain in pigs, which has not being improved over the years (2014 was the year with the highest number of mentions of the drugs used, which appeared only in 56% of the articles). These results are like those of the Bradbury study, in which the authors also conducted a review of pain management in swine in articles from 2012 to 2014. The need to control pain, particularly in animals used in research, is not only for ethical reasons, but also because pain side effects may occur in the used animals [18
]. Along with surgical stress, pain leads to an endocrine response, which could generate massive physiological changes that could alter the quality of the results [20
]. In addition, all cases were biomedical studies that sought to extrapolate results to humans. Conditions should be the same as those used in a human patient, otherwise, the results cannot be compared. Coulter and colleagues [21
] found that papers reporting ethical approvals were also more likely to report about systemic analgesic administration than those that did not. Furthermore, standards of ethical review differ widely between countries. The lack of reporting data may be due to the fact that the objectives of these articles were not related to animal pain. Failure to provide adequate postoperative analgesia undermines the three Rs principles, not complying with refinement—which seeks to improve experimental procedures to avoid suffering and pain of the animal—, replacement—individual animals should be replaced by alternative methods—, or reduction—less animals should be employed, using of a more powerful study design [2
Another factor is that the selected drugs should be the best choice for a good pain management in the experiments. The few articles examined in this review that reported analgesics looked for adequate options for pain management. Fentanyl (the most reported intraoperative analgesic in the articles examined) is a strong opioid that can generate excellent analgesia. Lidocaine, on the other hand, is a local anesthetic widely used to generate local anesthesia in dental and orthopedic surgeries. For postoperative analgesia, buprenorphine (an opioid agonist drug that has been shown to have a longer duration of action in pigs compared to other opioid drugs) [22
] was the most commonly used drug, and meloxicam (a nonsteroidal anti-inflammatory drug (NSAID) that preferentially inhibits cyclooxygenase-2 and has demonstrated potent analgesic and anti-inflammatory activities) was the second most used drug. However, few articles mentioned how postoperative pain was evaluated, which does not allow the reader to verify if the protocol used was the most appropriate for the type of surgery performed. The development of a pain-scoring system in pigs, together with the mandatory description of pain management in submitted articles, would contribute to improve laboratory pig welfare [2