Building infrastructure for data-driven decisioning

Beyond their intrinsic value as a showcase for new technology implementation, case studies provide additional benefit. By demonstrating the art-of the-possible, they help organizations envision how IT solutions might be applied in their own circumstance and to what end. The consumer of case studies can consider the ‘how to’ of implementation, absorbing lessons that may be used to develop business justification and smooth solution deployment. In the Grand River Hospital case study presented below, readers will come to understand the importance of planning and preparation in the building of information solutions — and how with expert help and the right technology, organizations can transform data into an effective decision support mechanism.


Waterloo, Wellington’s largest facility, the Grand River Hospital (GRH) is a 600-bed large community hospital delivering a full range of acute care, rehabilitation and complex continuing care services on two main campuses in Kitchener-Waterloo and Freeport, and through five regional satellite locations. With a staff of 3000 employees, GRH provides 15 clinical programs and services to 730,030 residents within Waterloo, Wellington and to patients referred from other parts of Ontario for neonatal intensive care, intensive care and other specialized services.

Business need:

Like many healthcare institutions, the Grand River Hospital has been under enormous pressure to speed service delivery in response to public demand and as a measure of accountability to the Ministry of Health. As Tina Mah, VP of planning and performance management at GRH has explained, the hospital has experienced increased demand to “go faster” from clinicians, and administratively from both planning and operational perspectives. But the institution has also faced demand for faster turnaround from external sources, such as public and regulatory reporting requirements, and due to the need to communicate with healthcare partners, including hospitals, home care and community care centres, on patient transfer. At the same time, new standards for data as a decision support tool have emerged: “ten years ago, we used to be quite comfortable with two-year-old data, we now have one-day-old data Ministry [of Health] requirements, which would have been unheard of before,” Mah noted.

Kathleen Lavoie, Chief Privacy Officer and director of health information management, Grand River Hospital
Kathleen Lavoie, Chief Privacy Officer and director of health information management, Grand River Hospital

Unfortunately, the hospital was equipped to respond to these changes with what Mah described as “1980s, 1990s mechanisms” that may have been capable of supporting some business demands, but from an efficiency standpoint “were not going to be sustainable because the pace was only going to pick up.” Geographic dispersion of GRH assets and the evolution of legacy systems to support different departmental specializations at Grand River produced a problem that is endemic to many organizations — information silos. At GRH, patient and operational data was stored in 68 different systems, some of which “talked to each other, and some of which did not,” according to Kathleen Lavoie, GRH Chief Privacy Officer and director of health information management : “we would have to pull information from 5 to 15 different Excel spreadsheets just to answer one question. It could take us up to a month to answer this properly, and often we could only see the question from one angle.”

The GRH was an early adherent to the Quality of Service care framework, and had created a program before it was required in Ontario healthcare institutions linking the quality framework with decision support and IT to provide the information needed for strategic decision-making. According to Lavoie, the hospital had a “scorecard” — a matrix with information on it, which formed the basis for the board, clinical teams, senior team and administration decisioning, but this was a very inefficient way to provide information to decision making groups. In addition to issues around access — data in different repositories managed by different individuals — in some cases, information in disparate systems did not match and processes around data collection and management were not as consistent as they could be. So while information was being generated by different departments in Excel sheets in multiple shared drives, retrieval was a very manual process, which did not provide the quick access to information that hospital workers needed to do their jobs, despite an existing BI tool that gave hospital workers the appearance of information automation.

To address this challenge, the GRH looked to new technology systems for the storage, protection and securing of information, but also, Mah explained, to a solution that would recycle and support multiple reuse of institutional data, “knowing we had put discipline into the management of that information.”

Building momentum:

Like many hospitals, the GRH operates in an environment of resource constraint, with many interests competing for investment priority. In establishing the case for investment in infrastructure as a means to improve patient care, the hospital was fortunate in having strong executive sponsorship for its transition to a data driven organization. According to Lavoie, the hospital’s CEO Malcolm Maxwell, who came on board in 2007, was a strong proponent of measurement as a means to performance improvement and to institutionalize information-based planning and strategic decision-making created Mah’s VP position within the organization in 2008. In her view, these staffing initiatives served as the beginning of a change in culture at the hospital, which was buttressed at a tactical level by focus in business case development on that fact that existing source systems were incapable of providing clinicians with the response time needed for effective patient management.

To ensure organizational buy-in for a central information repository, Mah and Lavoie worked over a two year period with various groups at all levels of the organization to build support for implementation of a data warehouse, soliciting input from HR, IT, decision support, finance as well as ad hoc members on who should be involved in the project. Over this period, Lavoie’s team also worked on reframing decision support to ensure consistent definition of information inputs (such as ‘alternate level of care’) and to introduce discipline around IT and other processes for data and information management — ex. the maintenance of tables for source data. This foundational work included Lavoie’s collaboration with IT in the creation of a Data Quality Accountability Framework covering items such as data dictionaries and metadata that fed naturally into the data warehouse project.

Data warehouse design and implementation:

As per the terms of the broader Accountability Act, the team generated an RFP, choosing ultimately to work with Information Builders (IB), the New York-based provider of analytics and information management platforms, which was able to bring its considerable experience to the table in creation of the GRH solution. GRH implementation of its data warehouse began two years ago, with the support of Information Builders professional services.

In strategizing around needs, Lavoie explained that multiple hospital teams considered “not only at the questions they needed to answer now, but also at the questions they would need to answer in future.” Part of this process entailed information consolidation. Tackling the onboard of the first system (Admissions, Discharge, Transfer), for example, the team examined over three thousand tables, moving through individual fields to determine which would be good candidates for porting to the data warehouse. Implementation was a joint effort, involving a huge amount of prep work on the part of GRH, as well as extensive discussion and collaboration with IB. Through this effort, Lavoie noted the vendor’s “openness and acceptance of challenge,” its willingness to engage in thinking outside the box about user roles, and its ability to work with the GRH culture. “They molded themselves to our culture,” Lavoie explained, working with many hospital teams to reconcile outstanding issues and to understand the unique requirements of the hospital. “We like to have multiple people involved in decision-making, and they embraced that way of working,” she added. Education was a key part of the process, with IB training the hospital team on the “five Ds of the data warehouse” and other subjects, while working with GRH to apply learnings from one iteration of the project to the next. IB worked closely with the hospital team throughout, supplying practical training which has ultimately allowed GRH to take on the next implementation stages independently. As Lavoie explained, “one thing that made me really happy with the partnership is that IB is not only in the business of data warehouses, but also in the business of teaching their clients how to use the data warehouse and how to do it themselves.”

The GRH solution involved implementation of iWay Data Migrator and adapters, WebFOCUS Report Server, Report Caster and Dev Studio. Depending on the technology and hospital systems involved, IB sent experts in database administration, integration or analytics to work onsite — and to effect knowledge transfer to GRH IT and decision support personnel. During roll out of the data warehouse, priority was also given to up front user training, and to ensuring that users had appropriate access tools. Developer tools, for example, might not be appropriate in all cases — the existing BI tool, which could establish linkages on the backend was considered better suited to some user needs. According to Lavoie, because many users had experience with access tools, transition to the data warehouse was fairly seamless.

Technology benefits and business outcomes:

Nine GRH administrative and patient care systems now reside within the data warehouse, and with IB’s help, the GRH team completed stage one of the implementation on time and on budget. While there was some initial hesitation from different hospital units around participation in the project, based on successful outcomes, Lavoie’s team has received several requests for inclusion, and is now working on a needs assessment in preparation for the move of the next 16 systems, beginning with HR, onto the data warehouse. According to Mah, the data warehouse has proven its ability to improve information delivery, having “reduced responsiveness from 100 percent down to ten percent of the time,” and due to automation capabilities which render processes reproducible and recyclable. For example, policies can be put in place to automatically generate a report over an extended period, without an individual staff member having to recreate the request, resulting is reduced demands on employee time. Prior to implementation, it was not possible for GRH systems to simultaneously process multiple jobs; however, as Mah explained, increased speed and efficiency in the data warehouse “has removed that bottleneck from the technology and human resources sides, as we only had one position with the knowledge to write those queries.” With 68 systems, the hospital potentially had requirement for knowledge of 68 languages — with the data warehouse, decision support and IT have provided new ways of accessing the data to enable corporate, as opposed to individual departmental discussions.

Increased speed, and reduced employee time input have translated into cost savings that the hospital has converted to increased productivity. Mah described this shift as transition “from technical work to knowledge work, where decision support and health records can find new efficiencies needed as we move into healthcare restructuring due to funding reform and into environments requiring greater integration with other organizations,” including satellite locations.

With the data warehouse, GRH information is also more reliable: Mah noted, “from a cultural point of view, it was really important for the organization to come to one source of truth, where we can demonstrate that we have put discipline to this source. Without this, all that you are left with is data verification and reconciliation, and you can never get to forecasting, smart analytics, support for future planning.” Populated with information that has been verified and linked through different identifiers, the data warehouse is now used by different departments within the hospital as a primary source data for more sophisticated queries. Cost case analysis, for example, a measure of the cost of care on an individual patient basis which is gaining popularity in the healthcare field, would entail extraction of data from a number of different systems. This was not possible prior to implementation, but with connected data, super users are now able to go to one location to extract multi-faceted information, and better able, Mah added, to ask questions that reflect the business. Standardization of electronic data has allowed users to get answers to the bigger questions more quickly, creating a sense of ownership by different units responsible for the quality of data collection and input. It has also provided continuity for a particular unit that will survive exit of an individual staff member. Discipline around data is a critical business requirement, Mah stressed, as data may serve as the foundation for multi-year, multi-million dollar decisions.

Through consultation with IB customers from other industries, Lavoie’s team is now working to better understand how IT and decision support can help individual users access and query information in the data warehouse. Since the system is agnostic, users can employ various interfaces such as SQL or Crystal Reports as needed, but Lavoie is developing a profile-based “triangle” of users as the basis for building a front end that will offer easier, more intuitive access to the data — relieving IT and decision support of the many questions that are coming with increased interest and use of the data warehouse, but also enabling new levels of data driven decision support for hospital workers.  “It’s not just about technology,” Mah concluded, “it’s about building human capacity.”





Please enter your comment!
Please enter your name here

This site uses Akismet to reduce spam. Learn how your comment data is processed.