The Effect of Mobile Text Message on Behavioral Risks of Cancer Among College Students, Northeast Ethiopia: A Randomized Controlled Trial

The study will be a single blind randomized controlled trial, which utilize quantitative method of data collection. College students who will receive a mobile text message about the behavioral risks of cancer and prevention mechanism will be considered as intervention groups while controls will be those who receive a text message related with general health once a week. A total of 160 (80 interventions and 80 controls) students will be included in the study using simple random sampling technique. After the baseline assessment, participants will be randomized to either usual care (control) or the text message intervention group in a uniform 1:1 (control: intervention) allocation ratio, using computer-based randomization service.

Self-administered questionnaire will be used to assess baseline and end line data of behavioral risk factors and other variables. Participants in the control group will receive an initial text message welcoming them to the study, but they will not receive the text message support program. The intervention group will receive a text message support program, where they will receive seven messages per week at random times and days for two months. The text message will be prepared and delivered based on health belief model (HBM), the items of this subscale will be measured on a Likert scale ranging from 1= "Strongly disagree" to 5 = "Strongly agree". Tobacco use, harmful use of alcohol, unhealthy diet and physical inactivity with a series of items will be used to measure the behavioral outcomes the items in this subscale will be measured on a Likert scale.

Validity of the items will be measured by an expert panel of specialists in health education and nutrition. They will judge about the necessity and relevance of the scale items. To control the data quality all data collectors will be trained for two days. Supervisors will monitor data collection process and the principal investigator will visit frequently. The questionnaire will be first prepared in English language and translated to Amharic and again re-translated in to English by another person to check for consistency. About 10% of the questionnaire will be pretested and internal consistency will be checked. Completeness of the collected data will be checked during data collection.

The data will be checked for completeness and consistency, then categorized, coded and entered using Epidata version 3.1 to minimize error. Then, the clean data will be exported to Stata version 14 for analysis. Normality of the data will be checked through Kolmogorov-Smirnov test. Student's paired samples t-test will be used to test within-group changes in terms of health belief model (HBM) variables and behavioral outcomes. The analysis of Covariance will be used to make over-group comparisons. Data will be reported as mean ± standard deviation. The significance level for all of the results will be presented at the P<0.05 level.