Medical practice and scientific research is an increasingly data-centric endeavor. Managing and leveraging that data through advanced information technologies and analytical methods is now a critical competency for academic medical centers (AMCs). To continually improve medical care, innovate and to remain competitive, AMCs find they must now develop and maintain robust Biomedical Informatics (BMI) resources.[1]
Giles & Associates Consultancy (GAC) was recently engaged by a leading AMC to design and facilitate the planning process for a dedicated BMI unit. The AMC had previously assembled a steering committee that met over six months prior to seeking external help. The request for external help was driven by high demand for computational sciences expertise, market competition, a desire to understand best practices, and the need to ensure internal alignment.
GAC designed and conducted a multi-stage process that included:
- executive advisory committee working sessions
- stakeholder interviews to identify existing BMI strengths and gaps
- primary research to understand best practices, organizational structure options and personnel skill set models of other leading medical centers
- development of implementation scenarios and financial forecasts
- iterative steering committee reviews and recommendations
A comprehensive business plan was delivered, outlining the scope and focus of a new BMI Department that will foster an innovative, productive research and support community across the enterprise. The plan included organizational structure, mission, strategic goals, hiring and transfer strategy, financial pro forma, implementation timeline, and performance metrics.
Current Inventory and Skill Assessment
Interviews with researchers and clinical thought leaders throughout the institution consistently indicated that BMI support for scientific research was suboptimal. Current and future needs assessments were developed and these needs were specifically addressed in the new departmental design.
The immediate need was to augment BMI resources available to support scientific research. In addition, research productivity and grant funding needed computational scientists and supporting expertise. The AMC’s centralized information technology function had previously allocated limited resources to scientific research support, but had not considered the need for expanded storage, and software and hardware infrastructure.
Additional concerns were raised about the organization’s ability to hire and retain highly qualified BMI personnel. Existing BMI resources were siloed within individual medical departments or research institutes. The new BMI Department would draw a “critical mass” of resources together in a community-like structure to support and sustain their unique competencies, and interests.
External Benchmarking/Organizational Structure
GAC interviewed BMI program leaders at Cleveland Clinic, Columbia University, Harvard, Mayo Clinic, M.D. Anderson, The Broad Institute, UCLA, UCSD, University of Pittsburgh, Vanderbilt University and others. Common themes and individual innovations emerged. These organizations had either already centralized, or were in the process of centralizing, BMI resources to eliminate siloed data and personnel, and achieve the aforementioned community. Two elements — service to the larger research community and a vibrant research program within BMI itself — were considered essential to a robust, sustainable, and productive program.
Organization size ranged from 60 to 130 depending on the scope of services offered, and they comprised a broad range of skill sets, either as part of the direct reporting structure or connected through the community interface. The most frequently mentioned were computer science, mathematics, statistics, machine learning, machine vision, biology, chemistry, biochemistry, and omics.
Other interesting approaches included rotation of BMI resources to different assignments, automated or self-service data tools for researchers, and a “concierge” or single point of contact for researchers seeking assistance (whether basic or complex) from BMI.
Based on these insights, and in working sessions with the steering committee, a set of key capabilities were identified as critical to the BMI department’s ability to support research and clinical BMI needs while also cultivating future talent. (Figure 1)
For organizations desiring rapid growth and highest impact, the highest priority is development of collaborative analytic capabilities. However, to be fully productive, computational scientists require the other capabilities as well. When brought together with the common goal of using technology, data, and communication to reach an actionable conclusion, groups can share information and available assets, and develop greater insights than they would as individuals.
Biomedical Informatics Organization Design
After considering current capabilities and gaps, and best practice recommendations, GAC defined four alternative BMI department structures for consideration. These ranged from minimal change and supplementation of existing BMI resources only, up to an extensive build out of BMI capabilities and BMI-focused research. (See Figure 2)
The Steering Committee’s first choice was Alternative D (≈70%), with Alternative C as second choice (≈30%). These scenarios were refined to yield the recommended organization shown in Figure 3.
Personnel Planning
With an established organizational structure, GAC re-interviewed stakeholders to determine which existing BMI resources/groups should transfer into the new department and how their skill sets compared to the key capabilities identified in Figure 1. Guidance on investment upper limits was sought to ensure an appropriately sized organization with realistic projections for staffing and recruiting.
Based on skill sets and headcount projections ranging from 100 to 130 people, a 3 to 5-year plan for onboarding department personnel was developed, along with a financial model that included salary and direct/indirect expense projections based on existing institutional cost data. Faculty to staff ratio is approximately 1:2, with faculty primarily in the Research & Education and Services Divisions.
Research IT is a separate organization staffed and managed independently of the new department. However, as part of this engagement Research IT’s resource requirements were evaluated and included as part of the total skill set to be hired and developed. Therefore it was included in the financial forecast.
Financial Forecast Model
GAC’s 5-year financial model forecasted the range of investment required for the new department at $60 to $80 million. Using benchmarks and actuals to validate assumptions, model inputs included:
- Recruitment and compensation costs by position/title
- Salary increases for transferrees
- Indirect expenses
- Primary/secondary grant and charge-back revenues
While there is opportunity for additional revenue through industry partnerships and IP, it is not expected to occur within the forecasted 5-year period.
Total headcount and the proportion of new hires to transfers had the greatest impact on required investment, highlighting the need to train and grow talent from within the organization rather than filling the group exclusively with external hires (which may not even be feasible in some highly competitive markets).
Integrated Business Plan Documentation and Transfer
In conclusion, the strategy, organization design with divisional details, head count and skill hiring plan, financial projections, and preliminary performance metrics were integrated in a business plan document and executive presentation. Re-evaluation over time and modification in response to changing circumstances is expected.
Through this business planning process and deliverable, GAC adapted externally-derived best practices for leading AMC’s unique pre-existing research enterprise. This “fix it forward” approach solved existing shortcomings while simultaneously investing in de novo BMI research and service capability, effectively propelling the institution’s BMI status from lagging to leading.
1: For purposes of this case study, biomedical informatics is defined as an interdisciplinary field that studies and pursues the effective uses of biomedical data and knowledge for scientific inquiry, problem solving, and decision making, motivated by efforts to improve human health.