Healthcare and Science thomsonreuters.com

The State of California is second only to the Federal Government as the largest purchaser of health insurance in the United States. CalPERS purchases health insurance for State employees, their dependents, and approximately 1,100 local government and school district subscribers and dependents.  As such, we purchase all healthcare services on behalf of over 1.3 million people. 

When I started with CalPERS in 2002, my job was to supervise the CalPERS team and actuarial consultants who negotiate the premium rates with our health plans year on year. Each year the rate team negotiates with very large health plans to provide insurance for our members. That current purchasing bill is about $6 billion. Everybody from elected officials to civil servants and their families then have access to medical and pharmacy services. Currently, our health plan providers are Blue Shield of California and Kaiser Permanente for health maintenance coverage (HMO) and Anthem Blue Cross for Preferred Provider Organization (PPO) coverage.  Each plan provider enrolls approximately 400,000 subscribers and dependents each.

We can cover people for their whole lives as they dedicate themselves to public service until they retire. Our basic health plan coverage can start for employees from the age of 18 until they are 64, at which point they go into the Medicare system which is a federal insurance program.

My job requires me to take all of the information and data that we receive from each health plan for each rate setting year and translate it into what we expect to pay for health coverage across future years. Every year we send data requests for quotes from our health plans, how much they are going to charge us and we then negotiate with those health plans on how much we are actually going to pay.

Our rate negotiation process is based on actuarial science. The underwriting and actuarial staff and consultants debate amongst themselves about the assumptions involved in the rate setting process, we compare forecasts, and then calculate what we expext the next year's health insurance bill is going to be. Rate negotiations with our health plans is, in part, about arguing the future costs with our health plans and coming to agreement with our health plans on what we expect to be the best estimate of that cost. In June we finalize rate negotiations and the CalPERS Board of Administration sets our rates for the following year. Of the $6 billion, the State pays about 80% for healthcare coverage for its employees and then the member makes up the last 20%. Local government employer contributions vary so that those employees contribute different amounts.

Our aging work force has an impact on our health plan rate negotiations. This is not unique to CalPERS or other employers, however. Other western industrial countries are facing the same financial difficulties as their aging workforce and dependent populations consume more healthcare services. How CalPERS manages the health needs of our population is key to how we purchase affordable, quality health care through our health plans, and how those health plan negotiate with their hospital and doctor providers for high quality, reasonable cost health care services delivered to our CalPERS members. 

When I joined CalPERS we didn't have the data warehouse; we were just starting to build our healthcare decision support system (HCDSS). Prior to implementing the HCDSS, we could only manipulate underwriting and actuarial data in the form of flat file Excel spreadsheets and Access databases. These analyses required extensive manual data manipulation, were time-consuming, and required extensive scrubbing of health plan data to detect error. With the HCDSS, the CalPERS rate team had the opportunity to view the health plan data independent of what the health plans were providing during the annual rate setting process. In short, with the HCDSS implementation we could analyze our costs the same way as the health plans in order to validate or refute their assumptions for setting future rates. Prior to implementing the HCDSS, my team of researchers (who are still the same team now) had the capability and expertise, but not the tools. Because the health plans had this enormous amount of data, they held the cards in terms of the negotiations so it was imbalanced.

On January 1, 2004, we went live with our HCDSS and immediately started using the data for the 2005 health plan contract negotiations. Suddenly, we had the same information as the health plans with whom we were negotiating. Before we had the data warehouse, the plans would come in with very high rates; in 2003 the proposed health plan rate increases ranged from 29% to over 40% and we didn't really have the computing power to argue or contest the rates they were giving us. Once we were armed with our data from the HCDSS in 2004, the playing field leveled, and we engaged that enormous computing power and Decision Analysis knowledge base to go back and have real negotiations with our health plans. In those first years, it gave us great insight into how we are spending our money.

The first year we went live with the HCDSS, we avoided $37 million in health plan rate increases that could be traced to use of the HCDSS. Each year since then, we've tracked how much we've saved in health plan rate negotiations using the HCDSS. We calculate the costs that were avoided as a result of using the HCDSS were $19 million in 2006, $25 million in 2007, $32 million in 2008.. In the last six years, CalPERS rate staff, in partnership with our health plans, have collectively analyzed our data to reduce our health plan rate increases year over year.  Our 2010 health plan premium increases have been the lowest in 14 years and CalPERS saved $600 million dollars from what we received in our initial health plan 2010 quotes to what was achieved for the final 2010 health plan rates.

The data-driven reporting we do is key to understanding our health program cost drivers, validating member costs, and building performance-based administrative and clinical measures for compliance. For example, as a result of the data warehouse, not only do we measure health plan performance, but we have built performance measures and targets right into the health plan contracts. These guarantees lock in solid data supply requirements, quality and timeliness of the data, as well as set health plan performance targets for quality and cost. This was all a direct result of the reporting made possible by Thomson Reuters Health Care Decision Support System.

When you have a $6 billion program and you can squeeze out even a quarter of a percent in a health plan rate negotiation, that's a lot of money. Thomson Reuters data gave us a big hand at the table, so much so that we now tell our health plans what we the purchaser wants, versus our health plans telling us what they will do and how much it will cost.

We believe that people get a better standard of care as a result of our efforts. Because quality measurement capabilities are built into our decision system, we can actually measure the quality for our members to determine the value receive for dollars spent. So members are paying less and getting a better standard of care as a result. Moving forward will involve the new generation HCDSS to give us greater ability to manipulate data and the ability to do more comprehensive deep dives into the data

When we began building the data warehouse we were pioneers. We just said 'this is our charge' and we are going to do it. And we did. We had a sense of what it could do for us, but it actually met our wildest dreams.