A study is designed to investigate the rates of central line–associated blood stream infections (CLABSI) among pediatric hematology/oncology patients. Investigators wish to compare the length of stay (LOS) between subjects receiving three common central line types (totally implanted catheter [port], peripherally inserted central catheter [PICC], and tunneled externalized catheter [TEC]). It is discovered that LOS is not normally distributed. What is the appropriate test for comparing the LOS between patients receiving the three central line types?
Correct Answer: D
Rationale: The correct answer is D: Kruskal-Wallis test. Since the LOS data is not normally distributed, parametric tests like Student's t test and ANOVA are not appropriate. The Kruskal-Wallis test is a non-parametric test used to compare three or more independent groups when the data is not normally distributed. It ranks the LOS values across the three central line types and tests whether there are differences in medians. The Wilcoxon-Mann-Whitney test (choice C) is used for comparing two independent groups, not three. Therefore, it is not the appropriate test in this scenario.