Rocky Mountain Data Science Network

Rocky Mountain Data Science Network

Overview

Biomedical research is becoming more complex with an ever-increasing volume of data (for example clinical data repositories, genomic or tissue data banks), the need to link data from disparate sources, and the increasing complexity of research design. Biostatistics and biostatisticians play a key role within biomedical research as a field of academic study concerned with development of the theory and methodology of data analysis and data science, and their sound application in medicine and health. This was recognized by the journal Science, which recently established a Statistics Board of Reviewing Editors in recognition of the basic role statistics and biostatistics play in scientific research, especially in medicine and health, and to ensure that published manuscripts meet rigorous statistical standards. A number of human and animal health journals, including the British Medical Journal, Epidemiology, and the Journal of Addiction Medicine, among many others, also count statistics editors among their editorial board members to help them “do a better job of demonstrating ‘how useful information can be extracted from difficult data’.” Most high-impact health and human research journals require authors to obtain the input of biostatisticians as part of the submission process. Accordingly, biostatisticians must have a comprehensive understanding of the clinical/scientific context of the research in addition to their mastery of statistical methods and sound data science. A recent US study reported that the number of biostatistics faculty of a university is positively associated with the amount of NIH awards. As a result, universities with strong biostatistics departments/groups such as the University of Washington and the Harvard School of Public Health account for larger and larger portions of NIH funds/grants. Indeed, many granting agencies require a biostatistician as a co-applicant, or at least to have collaborated, in the submission of a proposal. This training network will play a key role here and will facilitate statistical leadership in health science collaborations.

Benefits to Partners

This training network aims to bring benefit community partners. For example, Alberta Health Services (AHS) requires biostatisticians with advanced skillsets in the design of medical and health studies and the analysis of the resulting data. The establishment of a centre for biostatistics will allow data scientists at the University of Calgary and AHS to work together in advancing data science across the province. AHS also provides an ideal venue for internships and job opportunities for graduates in biostatistics, given our mutual interest in highly qualified biostatisticians. Further, Alberta’s Health Research and Innovation Strategy includes the training of highly skilled people in disciplines who will contribute to platforms such as clinical trials programs, biobanks, an “Alberta Genomics for Health” plan, among others, and will build Alberta’s health informatics capability. These government initiatives, adopted by funding agencies such as Alberta Innovates: Health Solutions, will require investment in and development of biostatistical support to ensure that the University of Calgary has world-class research-intensive capability in biomedical research.

The collaboration with University of British Columbia – Okanagan (UBC) extends these benefits to additional community partners in British Columbia (B.C.) as well as strengthens our ties. The BC Cancer Agency is establishing closer research ties with UBC-O, and biostatistics is a key area that has been identified as the agency attempts to improve health outcomes using the large amounts of data it is confronted with. Leveraging statistical resources on the UBC-O and Calgary campuses leads to improvements in how these data are analyzed. Furthermore, because of the nature of the UBC-O statistics group and its close connections with medical physics, new areas for joint research along the interface of biostatistics and medical physics are envisioned. A number of projects involving quality assurance related to imaging have been initiated. Training programs in R, quality assurance, clinical trials, longitudinal modelling and other areas of biostatistics will be important as this research progresses.

Career Prospects

Biostatisticians, with their expertise in the development and application of sophisticated statistical methods, play critical roles in the design of health and medical studies and the analysis of complex data arising from such studies. Additionally, the human and animal health industries employ a large number of biostatisticians. The demand for biostatistical expertise is so great that Fortune magazine declared a PhD in Statistics and an MSc in Biostatistics as the best two graduate degrees for jobs in 2015.

The Training Network

In recognition of the clear need and opportunity, we have established the Rocky Mountain Data Science Training Network (RMDSTN). The RMDSTN brings together researchers, clinicians, health scientists and members from a number of health organizations that will stimulate innovation through collaboration and the cross-pollination of ideas. These interdisciplinary collaborations will advance health research through the application of the most appropriate analytical methods as well as advance the development of methodological methods to meet the research needs of the health community.

Biostatistics graduate students who participate in the RMDSTN internship training program will gain hands-on experience in the application of statistical methods and study designs that build on and complement their formal training. They will learn both the art and science of biostatistics and develop important communication, project and time management skills. Doctoral students will also learn how to identify methodological problems suitable for their own research program; solutions that will impact health research by addressing limitations in existing approaches.

The RMDSTN is a natural progression of the collective activities of all biostatisticians at the University of Calgary – the University of Calgary Biostatistics Centre ( UCBC ) as well as the statistical and medical physics activities at UBC-O. Since the summer of 2014, this UCBC group of over 25 biostatisticians has been successful in achieving a number of important research and educational goals: (1) cross-listing of graduate courses in applied biostatistics in the faculties of Veterinary Medicine, Nursing, Kinesiology, Science and Cumming School of Medicine; (2) creation of thesis-based degree programs in Biostatistics (Master’s, Doctoral) to be offered in the Departments of Community Health Sciences and Mathematics and Statistics in the Fall of 2017; and (3) organization of three one-day biostatistics workshops on the topics of missing and incomplete data, causal inference and longitudinal data analysis. Current activities of this group are: (1) creation of a non-thesis Master’s program in Biostatistics (new graduate program); (2) summer school in Biostatistics; and (3) networking group for consulting and collaborating UCBC members. On the other side of the Rockies, statisticians have been collaborating actively with medical physicists and computer scientists and are pursuing an increasing number of projects at the BC Cancer Agency.

The RMDSTN is a critical component in our commitment to provide educational programs of excellence in Biostatistics and to substantially expand our collaborations in health research. These integrated goals will result in methodological innovations, improvements and even transformations in health care and development of the next generation of biostatisticians who are research-ready.