The paper presents a statistical analysis of Brisbois, Kock, Watanabe, Mirhosseini, Lamoureux, Chasen, MacDonald, Baracos, and Wisner (2011) article “delta-9 tetrahydrocannabinol may palliate altered chemosensory perception in cancer patients: results of a randomized, double blind, placebo controlled pilot trial”. The research aimed at determining whether Tetrahydrocannabinol (THC) therapy on cancer patients can improve smell and taste perception. This is well outlined in the article. The research was triggered by the effect of anorexia and weight loss on patients with advanced cancer; this results to functional loss, low quality life and low survival chances. Therefore, the research investigated the hypothesis that THC can alter chemosensory perception, consequently, enhancing food intake for cancer patients with reported chemosensory problems. The research hypothesis is clearly stated in the article. However, there are no clear statements of the research questions being investigated, although, it can be inferred from the hypothesis. This assumption is quite inconsistent with quantitative research methods requirement.
The study is an experimental study. The investigators applied THC or placebo treatments to the selected sample. The sampling process used a randomized and double blinded technique. In this regard, it involved assigning patients by a third party pharmacist or a computer random generated scheme. In double blinding subjects do not know whether they belong to the control group or the experimental group. This is critical in avoiding biases. The Sample adult patients with advanced cancer were recruited for the study. The sampling process is well outlined in the article defining the eligibility and exclusion criteria for those to be included in the study. Moreover, the sampling process is defined step by step. A researcher intending to conduct a similar study follows the guidelines with ease. The sample consisted of 46 patients who were randomized for the study. The study allocated 24 patients to receive THC (Marinol capsules), while 22 patients allocated to receive a placebo. Good sampling techniques should be consistent with the objectives of the study. In this case, double blinded technique is suited for testing medical treatment as the subjects cannot interfere with results of the experiments. A diagrammatic presentation of the elimination of patients from the study is outlined in the article which clearly shows that the final number of patients completed the study; 11 those who received THC and 10 those who receive placebo. In this case, the sample for experimental results analysis had been reduced to 21. This shows a strong variation from the initial 46 patients.
The main statistical parameters analyzed in the data were the descriptive statistics (mean and the standard deviation) for each desired variable. This was obtained through the use of SPSS statistical software. In addition, the analysis of the data was done per protocol. The primary outcome was treated to be the scores for the taste and smell. For the baseline survey, the main statistical parameters were the mean and standard deviation. In this case, THC standard deviation was used as an appropriate parameter for differentiation. Moreover, the study used the chi-square and fisher’s exact test to evaluate patient characteristics. The fisher’s exact test is an inferential parametric test for significance of the data and was appropriate for the analysis, while the chi-square is an inferential parameter for non-parametric test for significance. This was an appropriate tool. However, the analysis did not consider variation of the research sample resulting from elimination of patients from the study as a result of adverse effects or negative side effects. Inclusion of these subjects in the analysis could possible increase the standard deviation of the variables, consequently, negating the results of the experiment. Lastly, the data obtained for the placebo, THC and baseline can be analyzed to find out if there is a correlation in the changes.
Data is presented in the article using tables. For example, the questions used for the baseline survey have been presented in table 1of the article. The presentation of the questions in the article serves to validate the results of the survey as they can give information on what was investigated. In addition, the results of the experiment were presented in tables showing the standard deviation and the mean for various variables with regard to THC and the placebo. Moreover, the p values were given for each variable in the tables. However, data presentation is not appropriate for comparison of the THC results and the placebo results. Alternatively, the data could be presented using bar double bar graph for easy comparison between the baseline and post treatment data.
Analysis of the results is presented for all the variables. This includes results analysis for variables such as taste and smell, appetite, food preference and caloric intake, and quality of life among others. The analysis for the data is convenient as each variable is analyzed in the data. A comparison between post treatment and baseline data is done. Also, changes resulting from the experimental treatment of THC and placebo are compared with respect to baseline data. The analysis involves interpreting the main statistical parameters (mean and standard deviation) in the limits of the confidence intervals for each variable statistics.
The study concluded that THC improves chemosensory perception, macro nutrient preference, appetite, and quality of life for patients with advanced cancer. The conclusion has been justified by the fact that the THC treatment compared to placebo treatment enhanced the poor conditions of the patients.
Brisbois,T. H., De Kock, H. I., Watanabe, S.M., Mirhosseini, M., Lamoureux, D. C., Chasen, M.,
MacDonald, N., Baracos, V. E., and Wismer, W. V. 2011. Delta-9-tetrahydrocannabinol
may palliate altered chemosensory perception in cancer patients: results of a randomized,
double blind, placebo controlled pilot trial. Annuals Of Ocology. Oxford University press.