This document discusses business intelligence (BI) within the context of the UK's National Health Service (NHS). It identifies the key drivers for BI as focusing on patients, processes, departments, and organizational goals. There is also a need to control spending and understand costs. BI can help by creating a single version of truth from detailed data and providing consistent reporting, trends, and analytics. Successful BI requires executive commitment, clear goals, and involvement from key players like sponsors, champions, and IT. The first six weeks should establish governance, analyze needs and technology, prioritize projects, and plan workstreams. Ongoing costs and measuring return on investment are also discussed.
Introducing . . . Donna Kelly twenty years business intelligence experience . . . Provided best practices to National Programme in Leeds (NHS Spine/Secondary Uses Services).She created a greenfield technical data warehouse and business intelligence architecture for an Acute Trust (WWL).Programme Manager and Enterprise Architect for greenfield Commissioning Support Service in support of 30 London Primary Care Trusts; she brought the organisation from an empty office to fully staffed and operational business intelligence status; she worked at all levels from the infrastructure to the boardroom.Business Intelligence Programme Manager at Queen Elizabeth Hospital NHS Trust in King’s Lynn; she created the programme framework including infrastructure, organisation design and staffing, security , and methodology, in a total greenfield setting; then, she delivered Redwing’s ORIOLE Operating Room Intelligence as their first BI product. http://www.donnapkelly.com/http://www.redwing-bi.com/donna@redwing-bi.com 0(781) 380-0181
There’s a need for . . . Financial Intelligence: Must gain greater visibility into the full scope of financial operations including detailed analysis of cost and revenues, Need projections for total spend against budget. Analyze current care practice patterns, identify unnecessary or underutilized services and create reports to meet mandated reporting. Patient Care Intelligence: Need to be able to analyze patient data for clinical patterns and treatment protocols.Share more information with patients; Connect patient data across all settings e.g. hospital, GPMonitor and better forecast patient diagnoses and use of services .Why? To improve patient care, reduce wait times and administer more effective treatments.
A few words on SLR and PLC: Why the initiative to move to SLR and PLC.?SLR is a stepping stone to PLC. Why PLC?Brings finance and activity into line- allows a single, integrated set of activity- allows drill down into the detail , even the individual records, behind service deliveryThis is because we start at the bottom with the most detailed data, and then aggregate upWhy? What’s the prize?Service Line Reports–That are flexible–That are sensitive to changes in-year–That support dialogue with cliniciansYou get reference cost production a by-productP&L analysis at any level of intereste.g. Service line, consultant, commissioner, point of delivery, period, HRG, etc.
When discussing the last slide, I mentioned that we start at the bottom, then aggregate up. In Business Intelligence, we look for business processes of interest. For example, waiting list management. These processes generate business events of interest, for example, an outpatient appointment. Each of these events has a data record. Each of these records are captured. Business intelligence consists of adding up these records – the lowest level of detail.We store all these records in a single, coherent structure. We call that the data warehouse. All reporting comes from this one source, thus guaranteeing that everything is consistent.We hold details, but present aggregations in the first instance. You can drill down to the details if you need to, but otherwise they are masked. Given this base, we can delivery some very smart reporting indeed . . .
Absolutely the most important success factor is the attitude and the involvement of the Executive. Directives and communications must flow from the top.Number two is a recognition by everybody that BI is an enterprise-wide strategic programme. It will be implemented as a series of interlinked projects, but is, at heart, one single and unified programme.BI doesn’t come free. Management must devote time away from fire-fighting when needed, to make the programme work.The programme direction must be based on clear business needs. Must avoid the ‘field of dreams’ pitfall (if we build it, they will come).Goals must be measurable. For example, time spent per month generating performance reports to be reduced by 100% from 120 hours to zero.
The Project Sponsor. This is the formal head of the programme, the person one goes to when all else fails. Often called something like the Chief Information Officer or Director of Performance and Informatics. This person is much more than an administrative head; she is the ultimate driving force and motor of the programme. Without her committed and active involvement, the programme has little chance of success.The Project Champion. This is the sparkplug of the engine, the life and soul of the programme. This is the person who says, ‘we want, we need . . . ‘. He is typically an executive line manager, responsible for the line of business for which BI is being provided. He might be the Director of Operations, or perhaps Head of Surgical Directorate. The Project Champion will change, as the programme moves forward.The Steering group members have a vital role to play. Involvement is perhaps a couple of hours per month, but programme governance emanates from this group. Without this group, the programme will lack enterprise-wide authority and will lose enterprise-wide direction. There is a likelihood of the programme disappearing into obscurity.Involvement of the top IT people is critical to getting the programme off the ground. Defining the infrastructure is straightforward, but it looks nothing like that put in place traditional transaction-oriented systems. IT Executive needs to understand BI infrastructure requirements and progress them rapidly, to avoid the programme bogging down and suffering unnecessary delays.
A business intelligence initiative begins for many reasons; primarily to address a real business need. This business need may not be well-defined or understood.That’s the purpose of these first six weeks.These steps are well understood, and can be completed within the six week timebox.However, it’s extremely easy for this to stretch over several months. The way this happens is key players don’t perceive that they absolutely must make time for the process, as and when needed.We can remove one part of the problem in advance: the detailed six-week-step model identifies who is needed, when. The second part of the problem, key players deferring or cancelling meetings, lays in the hands of Executive management. With (going back two slides) Executive commitment in place, and directives flowing from the top, the schedule can be maintained.Note that there’s a Six Steps briefing presentation which is given when the programme starts. It goes through the steps in detail, and has an accompanying project plan and checklists.
Business intelligence is an iterative, evolving never-ending world. New questions arise, new answers demanded. Indeed, answering one set of questions usually generates an entirely new set. Our way of delivering business intelligence uses an approach called Evolutionary Development Methodology 2. It’s a tried and tested methodology,; It was created in the late 1980s for agile systems development using prototyping, and version 2 is adapted specifically for business intelligence. Delivery is accomplished though a program of work, broken into small projects called Releases. Each Release represents a unit of business functionality.Before going into that in more depth, a few words on establishing the environment. Amongst other things, this means infrastructure and organisation. Expect to pay on the order of £100K for infrastructure. This and subsequent costing figures, are for an outfit the size of an NHS Trust.The first Release will take about 20 weeks. Why? We’re not sure, but it always does. And the size of the organisation doesn’t matter. It’s always 20 weeks.You will need three to four people, plus the infrastructure team for this period. Personnel costs will be around £250K. Expect to then replace contractors with permanent staff.Bottom line: expect to pay £600K to £750K for year 1. Expect ongoing costs to be around 50-70% of Year 1. And always measure ROI.
When an NHS Trust, say, decides to commit to a full-tilt business intelligence programme, there’s usually a Performance and Informatics group already in place.Typically, this group works with Access and Excel to produce ‘spreadmarts’ which are collections of data to support a specific reporting requirement. The name comes from a combination of spreadsheet and data mart. Sometimes there’s a bit of SQL Server thrown into the mix. There’s an enormous amount of manual effort involved, and there’s little to no governance, documentation, or procedures.These are the people who are often called something like ‘information analysts’ and who should be (but usually aren’t) working hand-in-glove with the line business users. Their jobs will change out of all recognition. Retraining and redeployment will be required.Once the infrastructure’s done, then the BI people live in a separate bubble from the regular IT staff.Can’t stress often enough the need for Executive to give (and periodically reinforce) their blessing on the programme. The last point reflects the profound cultural shift that will need to occur.
BI is not just another IT cost centre. It’s a business change programme, strategic to how the business is to operate into the future. Technology is just the enabler. The BI programme must be driven by, owned by, the business.For example, it’s been traditional for IT to prepare the business cases for projects. In the case of BI, it’s the business which should be making the business cases.We’ve said that technology is just an enabler. True. But it is a big enabler. A sound infrastructure is essential for the programme to succeed. BI infrastructure is architected very differently from traditional transaction systems. It also needs to be put in place quickly. If IT want to do things the same old way, the programme can grind to a halt. We’ve developed menus of technical architectures to avoid this.Data problems are to do with sourcing the data. There are two issues here. The first is to create an understanding amongst all stakeholders, that if the source data isn’t present, there’s no possibility of any BI reporting. If a hospital doesn’t do a good job of collecting raw numbers on infection rates (for example) then there’s going to be no analysis of MRSA status trends. The second is a little more subtle. If the data is of insufficiently good quality, then any BI reporting is going to be meaningless, or worse, actively misleading. That said, everyone has data problems, and there are things that can be done.
Choosing to implement a business intelligence programme requires Executive to be committed and steadfast for the long haul. The quest then, is what is the organisation going to look like in say, three years?First and foremost, is why we’re here. The reason is patient care. Most organisations have systems that aren’t joined up. Each offers only a fragment of the total patient picture. As more and more data feeds are brought in, the more we get a complete picture of the patient. Ultimately, we’re in a position to offer a clinical portal, a one-stop place where we can see the total patient picture. And that’s the holy grail.There are major benefits along the way. Quickest win is via staff saving because we’ve automated much manual effort in performance reporting. Because the tools we use mean that people in the service delivery divisions get direct access to the information they need, the requirement for informatics staff is sharply reduced.As people get used to self-service bread-and-butter BI, new tools can be introduced.As performance management becomes pervasive, a cultural change occurs. The organisation becomes more effective. In the long-term, it can become driven by Executive strategy.. BI, for example, can be the foundation for the move to Service Line Management and ultimately, Patient Level Costing.
The keynote here is focus on the frontline.The goal is pervasive business intelligence.The change is cultural, such that the easy availability of actionable information becomes not just the norm, but an unnoticed part of day-to-day business. Improvements in efficiency and effectiveness are natural spin-offs.The ultimate big win is better patient care.Questions?
You are about to be wheeled into the ORIf you have any questions, now is the time . . .