Experimental Design and Statistical Analysis for Pharmacology and the Biomedical Sciences
Whatever the type of data our experiment has generated, we need to present our results in some form or other such as text, tables, or figures. The form that is chosen depends on the type of data you have and the message you wish to convey.
Text
Typically, data presented as text in the body of a manuscript (e.g. your laboratory report or that paper you wish to submit to the British Journal of Pharmacology). The following is a description of the phenylephrine data provided in the previous chapter (see Figure 1.3).
Figure 1.3 summarizes the ability of increasing concentrations of the α1‐adrenoceptor agonists, phenylephrine, to induce contractions of the isolated anococcygeus muscle of male Wistar rats in vitro (N = 70). Data were analysed by Repeated Measures ANOVA. However, initial Mauchly's test indicated that the assumption of sphericity had been violated, χ2 (90) = 1903.64, p < 0.001; therefore the degrees of freedom in the subsequent Repeated Measures ANOVA test were corrected using Greenhouse–Geisser estimates of sphericity (ε = 0.156). The Repeated Measures ANOVA revealed significant variation between the contractile responses according to the final organ bath concentration of phenylephrine [F(2.027, 139.862) = 1618.36, p < 0.001]. Post hoc analysis (repeated Paired t‐test with Bonferroni correction) revealed significant differences between successive concentrations in all pairwise comparisons (p < 0.01) except between the highest two concentrations examined (p > 0.05). These data clearly show that the response of the anococcygeus muscle is dependent on the concentration of phenylephrine to which it is exposed. The threshold concentration to significantly induce a response was 2 × 10−8 M with a maximum response achieved at 4.1 × 10−5 M. The mean EC50 for phenylephrine was 3.6 (2.8, 4.7) × 10−7 M, pEC50 = 6.44.
Note that the manuscript description of the data contains the same factual information as the legend for the figure (see Figure 1.3) but allows the opportunity to provide some extra detail, and the language used does not have to be so concise.
Tables
Generally, tables are used by experimental pharmacologists for two principal purposes in reporting experimental data; either as a means of summarizing their own data from a series of experiments or to summarize a substantial collection of data or information from a series of previously published studies already in the public domain.
Table 2.1 summarises the potency of atropine, a muscarinic cholinergic antagonist, and mepyramine, a histamine H1 receptor antagonist, to block the contractile responses to acetylcholine and histamine in the isolated guinea pig ileum in vitro.
Table 2.1 The relative potency of atropine and mepyramine on muscarinic and histaminic responses in the isolated guinea pig ileum.
| Atropine | Mepyramine | |
|---|---|---|
| Acetylcholine | 9.0 | >5.0 |
| Histamine | 5.8 | 9.3 |
Summary of pA2 values for atropine and mepyramine for acetylcholine‐ and histamine‐induced contractile responses in isolated guinea pig ileum in vitro. Data on file.
The table summarizes the pA2 values (−Log10 of the antagonist concentration, expressed in M values, estimated to double the EC50 concentration of the respective agonist); the magnitude of the pA2 value reflects antagonist potency on the receptor system stimulated by each agonist. The data indicate that atropine is about 1000 × more potent on muscarinic M3 receptor than on histaminic H1 receptors, while mepyramine may be up to 10 000 × times more selective for H1 receptors. This table demonstrates the phenomenon of differential antagonism expressed by atropine and mepyramine for these two receptor systems, and the relationships between these values are far easier to see in the table compared with when the values are buried in a paragraph of text.
In contrast, Table 2.2 summarizes the association of neuropathic pain with various disease states in terms of its classification and aetiology according to NICE clinical guidelines. This table simply shows how a wealth of information may be summarized efficiently; to describe all the relevant clinical studies within the body of a manuscript would be laborious and time‐consuming to prepare and, most likely, extremely boring to read.
Table 2.2 Disease associated with neuropathic pain.
| Disease | Classification | Aetiology |
|---|---|---|
| Painful diabetic neuropathy | Peripheral | Metabolic |
| Cancer pain due to surgery or chemotherapy | Peripheral | Paraneoplastic |
| HIV‐related neuropathy | Peripheral | Infection |
| Post‐herpetic neuralgia | Peripheral | Infection |
| Radiculopathy (nerve compression) | Peripheral | Trauma |
| Spinal cord injury | Central | Trauma |
| Multiple sclerosis | Central | Neurodegeneration |
| Post‐stroke pain | Central | Neurotoxic |
| Phantom limb | Peripheral/central | Trauma |
| Cancer | Peripheral/central | Paraneoplastic |
Classification is based on originating lesion. NICE clinical guideline 173 (2013).
Figures
In most cases, experimental data may be most efficiently communicated by the use of figures.
Line charts are useful to show trends in categories. Care must be taken to ensure that the use of line charts is not confused with scatter plots. While the magnitude of the data (shown on the Y‐axis) may be a continuous variable, the values on the X‐axis in line charts are not, and the data are plotted at set intervals or individual categories along the X‐axis; line charts are therefore wholly inappropriate for plotting information where the X‐axis reflects the magnitude of a continuous variable.
X‐Y Scatter plots should