This column is written for statisticians with master’s degrees and highlights areas of employment that will benefit statisticians at the master’s level. Comments and suggestions should be sent to Keith Crank, the ASA’s research and graduate education manager, at keith@amstat.org.

Contributing Editors
Karen Valentine earned a BS in mathematics and an MStat in biostatistics from the University of Utah. At Intermountain Healthcare, she helped clinical investigators design clinical trials, before moving to work within a clinical program team in 2005 so she could have a greater impact on patient care.

Roberta James has an undergraduate degree in chemistry and, after working at a health services research company while in graduate school, she took a position at Intermountain Healthcare to gain experience in combining clinical operations with outcomes monitoring and research.

Cempaka Martial earned a BS in statistics from Brigham Young University and an MStat from the University of Utah. She joined the Intermountain Healthcare team in 2004 as part of the Intensive Medicine Clinical Program; however, she recently joined the Women and Newborns Clinical Program on a part-time basis.

With persistent spiraling health care costs and the passage of the health care reform bill (Affordable Healthcare for America Act), the need to improve the effectiveness and efficiency of clinical care delivery is greater than ever. One of the models often mentioned by the Obama administration as a high-performing organization that reduces health care costs is Intermountain Healthcare’s integrated system. An integrated health care system provides patients with coordinated service delivery across the continuum of care, from physician office visits to hospitalizations and back to outpatient care.

As a fully integrated delivery system, Intermountain also owns and operates a health plan with multiple insurance products. One of the cornerstones of Intermountain’s success is their data-driven approach to providing the highest quality of care at the lowest cost using evidence-based medicine and sophisticated process improvement strategies supported by a state-of-the-art electronic health record.

Intermountain Healthcare’s integrative approach is achieved by first categorizing care into service line categories. Each service line category is managed by an organizational entity called a “clinical program.” The clinical program concept was conceived by Drs. David Burton and Brent C. James in the 1990s based on a key process analysis. Currently, Intermountain has a clinical program for each of the following service categories:

Behavioral health
Cardiovascular
Emergency and intensive medicine
Oncology
Pediatric specialty
Primary care
Surgical services
Women and newborns

Each clinical program has a staffing infrastructure that, at minimum, includes a medical director, an operations director, a data manager, and a statistician (with a job title of outcomes analyst). Statisticians are instrumental to the Intermountain clinical program infrastructure, working as core members of the clinical program teams. They are honed in the Intermountain Institute for Health Care Delivery Research, creating a community of outcomes analysts who dedicate two-thirds of their time to clinical programs responsibilities and the remainder to institute activities around research, consulting, and education/training.

We support three of the eight clinical programs—pediatric specialty, cardiovascular, and women and newborns—with more than 20 years of combined Intermountain experience. Individually, our career paths started in different industries, but converged at Intermountain, given a shared interest in applying statistical tools and concepts to health care delivery.

An integrated health care system consists of many teams within and across clinical programs. As core members of these teams, statisticians’ key functions are to retrieve data, analyze data, and summarize results. Specific applications of these include the following:

• Routine feedback reporting to clinical teams for monitoring care processes and board goal progress
• Applying a clinical epidemiology perspective to population management
• Assessing tests of change from continuous quality improvement initiatives
• Building the evidence base through clinical and outcomes research

Although the eight clinical programs are unique, their organizational structure is similar. Clinical program statisticians need to have the following knowledge base/skills in fulfilling key functions:

• Statistical methods and research design
• Quality improvement science, concepts, and tools
• Health care delivery and reimbursement
• Strong analytical skills
• Effective communication skills
• Statistical software applications programs (e.g., SAS and Statit)
• Structured query language (SQL) programming for large data sets
• Ability to use and/or learn a variety of software applications for reporting or presenting results
• Ability to summarize and present results in table, graphical, or presentation formats
• Ability to mentor or teach quality improvement, statistical, or research methods to various types of health care professionals

For instance, to improve the overall quality of care within a health care system or clinical service line, processes for caring for patients need to be identified and prioritized according to volume, cost, and variation. A health care system can then focus quality improvement efforts in the area with the greatest need. Statisticians are typically involved with the data retrieval, analysis, and summarization of this process.

Similarly, statisticians are essential to all the identified quality improvement projects. However, for a statistician to be an effective team member, he or she needs to communicate well in a multidisciplinary setting. Statisticians should understand the details of the care processes that need improving to be most helpful. Similar to the initial identification and prioritization process, the team will need help identifying the best opportunities for improvement.

The need for data retrieval, analysis, and summarization continues throughout the quality improvement effort. Most quality improvement projects can be tied to a measurable outcome measure. Statisticians are typically the creators of statistical process control or run charts used to monitor quality improvement efforts. Thus, statisticians are essential for identifying, implementing, and evaluating quality improvement efforts within an integrated health care system.

At times, there exist different, but acceptable, methods for care delivery, and the optimal method is unknown. When this occurs, a research experiment could be incorporated into the quality improvement effort, thereby improving the known methods while comparing the effectiveness among them. An integrated health care system is the ideal environment for conducting this type of comparative effectiveness research. In addition to the usual assistance for quality improvement efforts, statisticians can collaborate with research investigators in designing research experiments and contribute to writing research proposals and preparing manuscripts. Best practice in health care can be refined and identified by conducting this type of peer-reviewed research.

Finally, statisticians who are knowledgeable and trained in both quality improvement and outcomes research can serve as mentors or consultants to clinicians. This requires that statisticians stay abreast of emerging design and methodological advances such as comparative effectiveness. The clinical program statisticians at Intermountain Healthcare, in addition to helping their assigned clinical program, serve as consultants to health care professionals enrolled in Intermountain Healthcare’s Institutes for Healthcare Delivery Research training courses in quality improvement, called Advance Training Program. Course participants are typically from other health care institutions.