You are on page 1of 6

Freddie Daw

Literary Review Essay

Meta Analysis of Welfare Studies For my project I will be investigating the usage of meta-analysis and determining its effectiveness in assessing welfare studies. To do this I need to research how to perform a meta-analysis as well as collect data for an example meta-analysis I will be performing. I have chosen to use pig tail biting as my example topic due to its large scale welfare impact, the availability of literature on the subject and a personal interest in pig welfare issues. The review is split into two sections, dealing with the literature about meta-analysis and the literature related to pig tail biting, the information gathered from both will be used to form the experimental plan outlined in the appendix. Meta Analysis Practical Meta-Analysis (Lipsey and Wilson, 2001) defines a meta-analysis as a form of survey research in which research reports, rather than people, are surveyed while What is a metaanalysis (Crombie and Davies, 2001) establish that meta-analyses provide a quantitative (statistical) estimate of net benefit aggregated over all the included studies. Summarised, meta-analysis is a method where a reviewer takes the data collected in multiple previous studies on the same topic and uses advanced statistical methods to get an overall effect size for the topic. This systematic review of research means that a much wider range of results can be found than in any lone experiment, therefore making it a much more powerful statistical method; however it is also far more time consuming and difficult to perform. Quantitative data must be collected from each paper involved in the study; this means any studies that do not publish their figures cannot be used in the meta-analysis. Those that are used have to have data that can be suitably compared to each other, even research done on the same topic uses different measures and in some case this will be different enough that a meta-analysis will not be possible. These limitations mean that it is important to gather as much of the literature as possible on a subject to ensure that a meta-analysis will contain enough information to make it worth attempting. Once collected the data is given different weighting, usually this is due to variation in sample sizes of studies giving each of them a different statistical significance which must be accounted for in the meta-analysis. Another problem that must be overcome is that of publication bias, a term coined by Robert Rosenthal (1979), it is possible that papers which are funded by or carried out by people with a particular aim could be weighted in favour of that aim (Thornton and Lee, 2000). In these cases a meta-analysis is useful, as combining with other papers without this bias produces a more accurate investigation of the situation. Publication bias can usually be detected by plotting a funnel plot of the data (Light and Pillemer, 1984) this compares the scatter of effect size against sample size. Since studies with smaller samples should have greater variability than larger studies, the plot should form the shape of a funnel; due to this publication bias is indicated by the shape of the plot being skewed. However in small meta-analyses with limited data points it may be hard to notice the effect of bias (Lipsey and Wilson, 2001). The meta-analysis will produce data representing the different effect sizes of each papers results and how statistically significant the topic is overall. This can be represented visually by a forest plot, a visual representation that has been used for many years, though the origin of the term is unclear (Lewis and Clarke, 2001). This graph has a data point for each study, offset either to the left or right from a line indicating a basal treatments expected effect; left indicates the study has a more positive effect while right indicates that the study had less effect. Each data point will be of varying size based on its precision (usually related to its sample size), larger markers indicating larger

Freddie Daw

Literary Review Essay

studies. A line indicating the confidence interval of the study will also be plotted to indicate the possible effect area of that study. Using all of the data collected from each study an overall effectiveness of what is being looked at can be plotted, usually by a diamond at the bottom of the graph, this will show whether or not there is a positive or negative effect compared to the basal line (Crombie and Davies, 2001) When reviewing the papers to extract the data for my meta-analysis it is important to know what should be present, the ARRIVE guidelines (Animal Research: Reporting In Vivo Experiments, 2010) give an overview of what should ideally be in research papers, papers that follow these guidelines will contain information easily extractable for the systematic review to perform a meta-analysis. When looking at how to present a meta-analysis it is important to see previous examples, the main source for this is the Cochrane Collaboration. This organisation set up in 1993 collects and publishes Cochrane Reviews, healthcare meta-analyses covering a wide range of human medical issues. Though the organisation is set up for human medicine it provides many examples of papers that will be useful to show how a meta-analysis can be presented, therefore I have reviewed a selection of studies, not related to the topic I will be using (Glynn et al, 2010; Hemil et al, 2007; Smyth et al, 2007). The Cochrane library has over 4000 reviews available online, the papers I have chosen papers that are from their top 50 papers to prevent having to choose from the full selection. The Cochrane Collaboration also supplies a handbook for people who intend on producing a systematic review, this highlights the sections that should be contained in a review and how each of these sections should be approached (Higgins and Green, 2009). The handbook covers not just how to write a paper but also covers the selection of papers and data, highlighting key factors about this important stage for a meta-analysis. Pig Tail Biting Pigs are a large industry in both the UK and worldwide, in 2009 the pig industry was worth 1015 million (ukagriculture website) in the UK and many more millions globally. NADIS (National Animal Diseases Information Services) places the potential costs for a farm due to tail biting at up to 10,000 due to deaths attributed to it, lost growth from pain and stress and treatments for affected pigs (NADIS, 2007). Pig tail biting is a common problem on many farms, especially larger more intensive ones. Because of tail bitings impact on the welfare of many animals and its economic importance it is an excellent candidate for further analysis to assess its effects and the factors that cause it. By bringing together the results from as many papers as possible it would be possible to get a more conclusive result as to why it may happen so often. This can lead to improvements in the future for the welfare of pigs, opposed to using individual papers which may lead to conflicting answers or incomplete information on the topic. It has been highlighted that no one predictive behaviour is seen in every case of tail biting, showing that though there are easily identifiable risk factors all must be taken into account when trying to predict the outbreak of tail biting (Statham et al, 2009). An initial attempt at combining factors to create a predictive model has been carried out with some success, using a system that looks at multiple variables affecting pigs instead of individual factors alone (Bracke et al, 2004 a, b) Tail biting is identified as having multiple stages in The European Food Safety Authority (ESFA) report, (2007), pre-injury where there is some manipulation of the tail with no visible wounding and the injury stage where a clear wound with associated bleeding is made. The injury stage is most identifiable and likely to be measured in the literature more commonly. The report highlights the unpredictable nature of tail biting as well as its multi-factorial origins; due to this multi-factorial

Freddie Daw

Literary Review Essay

nature it is hard to fully explain causes behind tail biting. Most studies have looked at identifying one factor in particular but have had to take others into account in their experimental design. One problem with reviewing the literature is the difference in how tail biting can be defined by different authors, though many use a system similar to the EFSA one based on Fraser and Broom (1990) (cited in Taylor et al, 2009) where a pre-injury and injury stage are identified. Taylor et al propose a new system of identifying tail biting to try and overcome the differences between observers using three forms of behaviour two-stage, sudden-forceful and obsessive, proposing that tail biting from different motivations falls under each of these three categories. A study on general risk factors was performed by Moinard et al, (2002), in this study 92 farms were studied over an 11 month period to assess what risk factors there are on pig farms leading to tail biting. The factors identified in the study for increasing the risk were having slatted flooring, more than 5 farms in a pig unit, high stocking density and feeding density, diseases were also linked to changes in rate of tail biting. Adding straw was found to decrease tail biting. All of these factors have been quantified and their significance compared to a control farm. The data is very useful for the meta-analysis due to the large sample size used and range of factors. Breuer et al, (2005), looks at the effects of breed and genetics on tail biting, this paper looks at whether some pigs are more inclined to tail biting and whether this trait is passed on genetically, the paper used 9,018 pigs over 12 months looking at large whites and landacres. The results indicate that there is a weak genetic component to tail biting prevalence with landacres being more prone. Posture of pigs tails has been looked at to see any links with the likelihood of being bitten, in both Statham et al, (2009), and Zonderland et al, (2008), it is shown that pigs with tails tucked are more likely to suffer from being bitten, this is shown with data collected on tail position compared to that animal being bitten. The papers both take their results from observational data on large scale pig units, however only Zonderland et al were looking specifically at tail posture, while Statham et al looked at a larger range of behavioural indicators pulling out tail posture as significant in their analysis. Another factor that is addressed in a number of papers is the topic of whether tail biting can be linked to nutritional issues in pigs, the main concept being that possible nutritional deficiencies in a pigs diet will lead to it performing aversive behaviours to try and get what it is lacking. This is often assumed to be for salt, though this is rebuked by Jankevicius and Widowski, (2004), who found through experiments that saltiness did not seem to be the driving factor for pigs attraction to blood. It is shown in Hunter et al, (2001) that there is a difference in pigs motivation to chew on rubber hosing when it contains more nutrition, but due to the similarities of nutrition in the experimental pigs diets no effect could be observed on tail biting. A common related issue is the idea that the taste of blood attracts the pigs so they will attack wounded tails (Fraser, 1987), causing further wounding, generally this is why bitten pigs or biters are removed from pens when tail biting occurs. Zonderland et al, (2008), use the removal of tail biters as one of their experimental curative factors to assess its usefulness, finding a positive result but concluded that it did not deal with why tail biting occurred nor could it completely prevent it occurring again. Management techniques are highlighted as being important in affecting tail biting. Hunter et al, (2001), collected questionnaire results from 450 units using 6 abattoirs, representing 27,870 pigs; this data gave information about use of straw, sex mixing and feeding technique. This data was then compared to the docking procedure (docked, tipped, undocked) and the prevalence of tail biting on the farms. The main conclusions they found were that docking had the highest preventative effect of any management technique on long-tailed pigs while straw had a positive effect reducing tail biting

Freddie Daw

Literary Review Essay

on both undocked and docked pig tails. Artificial ventilation was linked to an increase in tail biting, though it is speculated that this is due to natural ventilation usually being linked to provision of straw on a farm. No link between sex mixing or feed type was found, though multi-space feeders did have a dramatic effect on reducing tail biting. Another study comparing slatted systems to straw systems found that the effect of bedding did affect welfare for the pigs, with both systems having benefits not found in the other, slatted systems had a higher chance of tail biting though, this highlights that though reducing tail biting will improve welfare the methods to do this must be weighed against other potential welfare issues (Scott et al, 2006). Straw provision is one of the most commonly highlighted issues in the literature, Day et al, (2002), did an experiment using pigs to see the effects prior experience of straw had on their behaviour as well as the effect of different levels of straw. The study used a design where it had two levels of prior experience against four levels of provision of straw. Over 32 groups, of ten pigs each, results were found that showed pigs with a prior experience of straw bit more when given no straw, other results were analysed and found to be not statistically significant. All the data is summarised showing whether it is significant to P<0.05or P<0.01, interactions between the different effects are explored, overall the data shows a positive link between providing straw and pig welfare. This is also shown in other papers dealing with straws effect on tail biting (Day et al, 2008; Hunter et al, 2001; Moinard et al, 2003). In another paper small quantities of straw were found to reduce tail biting, however the results also indicated that the manner in which straw is provided could have an impact, loose straw being much more effective at reducing tail biting than straw from a rack (Zonderland et al, 2008). Arey and Franklin, (1995), found that straw did not reduce fighting between newly mixed pigs, however in this study the aggression measured was not tail biting and so will not be relatable to other studies on tail biting for further analysis; it does however suggest that tail biting has other motivations than usual aggression in pigs. Not all studies show straw to have an effect, Statham et al, (2009), conducted research that showed different levels of straw to have no effect on their experiment, their results were analysed with this variable removed to look at other factors they had measured relating to pig tail biting. Conclusion The review has shown that there is a wide range of information available about pig tail biting that can be used for the meta-analysis, highlighting a range of factors that may be impacting its prevalence. Meta-analysis is widely written about with specific information available for individual aspects as well as texts that give a generalised outline; this will be useful when it comes to performing the analysis myself, as well as for extracting the data sets from the pig tail biting literature. Cochrane reviews will provide me with the main point of reference when it comes to writing a report following the meta-analysis as there is no current standardised animal welfare system for these reports.

Freddie Daw References

Literary Review Essay

Bracke, M., Hulsegge, B., Keeling, L., Blokhuis, H., Decision support system with semantic model to assess the risk of tail biting in pigs 1. Modelling, 2004, Applied Animal Behaviour Science 87, 31-44 Bracke, M., Hulsegge, B., Keeling, L., Blokhuis, H., Decision support system with semantic model to assess the risk of tail biting in pigs 2. Validation, 2004, Applied Animal Behaviour Science 87, 4554 Breuer, K., Sutcliffe, M., Mercer, J., Rance, K., OConnell, N., Sneddon, I., Edwards, S., Heritability of clinical tail-biting and its relation to performance traits, 2005, Livestock Production Science 93, 8794 Day, J., Burfoot, A., Docking, C., Whittaker, X., Spoolder, H., Edwards, S., The effects of prior experience of straw and the level of straw provision on the behaviour of growing pigs, 2002, Applied Animal Behaviour Science 76, 189202 Day, J., Van de Weerd, H., Edwards, S., The effect of varying lengths of straw bedding on the behaviour of growing pigs, 2008, Applied Animal Behaviour Science 109, 249-260 Fraser, D., Attraction to blood as a factor in tail-biting by pigs, 1987, Applied Animal Behaviour Science 17, 61 68. Fraser, D., Broom, D.M., 1990. Farm Animal Behaviour and Welfare, third ed. Baillire Tindall, London. pp. 327328. Glynn, L., Murphy, A., Smith, S., Schroeder K, Fahey T., Interventions used to improve control of blood pressure in patients with hypertension, 2010, Cochrane Database of Systematic Reviews, Issue 3. Art. No.: CD005182. DOI: 10.1002/14651858.CD005182.pub4 Hemil, H., Chalker, E., Douglas, B., Vitamin C for preventing and treating the common cold, 2007, Cochrane Database of Systematic Reviews, Issue 3. Art. No.: CD000980. DOI: 10.1002/14651858.CD000980.pub3 Higgins JPT, Green S (editors)., Cochrane Handbook for Systematic Reviews of Interventions Version 5.0.2 [updated September 2009]. The Cochrane Collaboration, 2009. Available from www.cochranehandbook.org. http://www.ukagriculture.com/livestock/pig_industry.cfm Hunter, E., Jones, T., Guise, H., Penny, H., Hoste, S., The Relationship Between Tail Biting in Pigs, Docking Procedure and Other Management Practices, 2001, The Veterinary Journal 161, 72-79 Jankevicius, M., Widowski, T., The effect of ACTH on pigs attraction to salt or blood-flavored tailmodels, 2004, Applied Animal Behaviour Science 87, 55-68 Kilkenny, C., Browne, W., Cuthill, I.C., Emerson, M. and Altman, D.G., Animal research: reporting in vivo experiments: the ARRIVE guidelines, 2010, British Journal of Pharmacology 160, 1577-1579 Lewis, S., Clarke, M., Forest plots: trying to see the wood and the trees, 2001, British Medical Journal 322, 14791480

Freddie Daw

Literary Review Essay

Light, R., Pillemer, D., Summing up: The Science of Reviewing Research, 1984, Cambridge, Massachusetts.: Harvard University Press. Moinard, C., Mendl, M., Nicol, C., Green, L., A case control study of on-farm risk factors for tail biting in pigs, 2002, Applied Animal Behaviour Science 81, 333-355 Practical Meta Analysis, Lipsey, M. and Wilson, D., 2001, Sage Publications, Inc. Rosenthal, R., The file drawer problem and tolerance for null results, 1979, Psychological Bulletin 86, 638641 Scientific Report on the risks associated with tail biting in pigs and possible means to reduce the need for tail docking considering the different housing and husbandry systems, 2007, Annex to the EFSA Journal 611, 1-13 Scott, K., Chennels, D., Campbell, F., Hunt, B., Armstrong, D., Taylor, L., Gill, B., Edwards, S., The welfare of finishing pigs in two contrasting housing systems: Fully-slatted versus straw-bedded accommodation, 2006, Livestock Science 103, 104 115 Smyth, R., Alldred, S., Markham, C., Amniotomy for shortening spontaneous labour, 2007, Cochrane Database of Systematic Reviews, Issue 4. Art. No.: CD006167. DOI: 10.1002/14651858.CD006167.pub2 Statham, P., Green, L., Bichard, M., Mendl, M., Predicting tail-biting from behaviour of pigs prior to outbreaks, 2009, Applied Animal Behaviour Science 121, 157164 Taylor, N.R., et al. Tail-biting: A new perspective, 2009, The Veterinary Journal, doi:10.1016/j.tvjl.2009.08.028 Thornton, A., Lee, P., Publication bias in meta-analysis: its causes and consequences, 2000, Journal of Clinical Epidemiology, 53, 207-216 What is Meta Analysis?, Crombie, I. and Davies, H., 2001, what is...? Series www.cochrane.org Cochrane Collaboration website www.nadis.org.uk NADIS pig health, 2007 Zonderland, J., van Riel, J., Bracke, M., Kemp, B., Hartog, L., Spoolder, H., Tail posture predicts tail damage among weaned piglets, 2009, Applied Animal Behaviour Science 121, 165170 Zonderland, J., Wolthuis-Fillerup, M., van-Reenen, C., Bracke, M., Kemp, B., Prevention and treatment of tail biting in weaned piglets, 2008, Applied Animal Behaviour Science 110, 269281

You might also like