Why an Information System? --- Despite substantial expenditures in the California Workers' Compensation system, remarkably little is known about workplace injuries, the costs of the system, and its ultimate effectiveness in encouraging prevention, providing indemnity, and facilitating return to work. To remedy this problem, states are increasingly concerned about developing better workers' compensation information systems that provide data and knowledge about the compensation system. This report describes a Workers' Compensation Information System (WCIS) for California. The design is based upon extensive interviews and consultations with insurers, self-insurers, employers, representatives of injured workers, legislators, and administrators in the Division of Workers' Compensation other state agencies, and upon the deliberations of a working group convened at UC DATA (University of California Data Archive and Technical Assistance). This working group included Division of Workers' Compensation (DWC) staff members, UC DATA staff, and academics from the University of California.
This is a propitious time to develop such an information system. Changes in 1989 and 1993 in the laws governing workers' compensation in California have laid the groundwork for significant alterations in the system. The public, the legislature, insurers, employees, employers, the Division of Workers' Compensation, and others have an interest in monitoring and evaluating these changes. In addition, the ongoing Electronic Data Interchange (EDI) project of the International Association of Industrial Accident Boards and Commissions (IAIABC) provides a basis for thinking about information reporting and analysis throughout the workers' compensation system. Finally, based upon a desire to assess the impact of the 1989 and 1993 reforms in the workers' compensation system, a concern with the lack of reliable information about system, and the need to improve its administration, the California legislature has provided the DWC with a legislative mandate for developing a WCIS which is described in California Labor Code ¤138.6.
The Goals of an Information System --- UC DATA's design of a WCIS began by considering the concerns of those affected by the system: workers, employers, claims administrators, the state legislature, the governor, the DWC, and others. These concerns were then translated into specific questions about workers' compensation (see Table 1 and Appendix 1 in the report). The WCIS described in this document is designed to:
It is also designed to:
These goals are met by selecting only the minimum necessary data for reporting on all cases, reporting more detailed information on only a sample of cases, limiting sample sizes to those necessary to satisfy minimum requirements for generalizability, using administratively simple sampling techniques, keying reporting to the requirements of the Benefit Notices to the injured worker, standardizing the reporting of data wherever possible on IAIABC formats, and limiting, as much as possible, the data reported from claims administrators to information that they will have as a standard part of carrying out their business functions. As much as possible, data that is incidental to a standard business purpose is collected through surveys conducted by DWC.
The Basic Strategy --- The basic strategy we propose has three features. First, it recognizes that objective information about injured workers' claims is the basic building block of any worker's compensation information system. Once we have information on injured workers' claims along with information about the employer and claims administrator involved with each claim, it is possible to answer questions not only about workers, but also about types of employers and types of claims administrators. We can track injuries, medical treatments, workers' benefits, claims processing, and return to work outcomes by types of employers and claims administrators. These are some of the most important things we want to know about the system.
The second feature is an optimal use of various methods of data collection. In the report we discuss in detail how different methods of data collection such as surveys and administrative data have different strengths and weaknesses and complement one another in important ways. The approach we propose can be described as a pyramidal information system in which the base consists of information on small number of items gathered on a large number of cases. Administrative sources are well-suited for this purpose. This information is then used to identify cases for which more detailed information is gathered at the second step of the pyramid. In turn, information gathered at the second step is used to identify cases for which even more information is obtained at the third and higher steps. As we work up the pyramid, more and more use is made of surveys which are better able to collect substantial amounts of information on a relatively small number of cases. At each step, the data collection method is designed to meet the goals listed earlier including providing reliable, accurate, and useful data, minimizing problems for claims administrator and employer reporting, protecting the privacy of injured workers and others, and minimizing complexity.
The third feature of the proposal is the linking of data to other data sources. Additional strength can be gained by linking the various steps of the pyramid to internal DWC databases and to external governmental databases, such as the wage and employment data available at the California Employment Development Department. These linkages will not impose any additional burdens upon carriers or employers, and the costs of creating the databases will be borne by DWC. Moreover, these linkages have the potential for providing many key variables that will allow for even more detailed answers to the questions posed in this report.
The Steps of the Pyramid --- The essential feature of the proposed system is that it makes use of the fact that any administrative process must have ongoing administrative information systems for tracking cases. The basic information about claims in these systems is used as the base of the WCIS---the first step of the information pyramid. Claims administrators will be asked to gather and report a small number of pieces of information on all cases. The information provided by claims administrators will be used to produce statistics on the number of claims by basic categories in each year. This information will also be used to improve the efficiency of DWC case make-up and to reduce future reporting requirements on claims administrators.
At the second step of the pyramid, data will be collected on a ten percent sample of the entire claims population. Information on the various transitions in benefit status, health status, and work status will be collected on this sample. Claims administrators will be asked to provide this information so that it will be possible to answer basic questions about the number of injured workers receiving benefits, the benefits they receive, length of indemnity payments, and the total indemnity benefits received. By requiring only a ten percent sample, we will reduce the burden on claims administrators. At the same time, a ten percent sample will allow for satisfactory inferences about statistics for the entire population and it will provide enough cases for the analysis of important subpopulations. This last fact is especially important as we go to the final steps of the pyramid.
At the third step of the pyramid, medical cost and utilization data will be collected on a one percent sample of cases. This will make it possible to track medical costs over time and across cases. In the fourth step of the pyramid, telephone surveys will be undertaken of a subsample of the ten percent sample. These surveys will provide information on the basic characteristics of workers, employers, and claims administrators that could not be obtained in any other way. The final step of the pyramid includes special studies based upon samples from previous steps.
The approach we propose decreases the number of cases while increasing the amount of information on each case as we go from the administrative database to the survey samples. This makes sense, however, because as the required information becomes more detailed, sample surveys make it possible to reduce the burdens on administrative systems while obtaining high quality information. By linking the data from these different sources, the approach we propose provides an extraordinarily useful and versatile database for answering questions about workers' compensation in California. Moreover, it minimizes the burdens on carriers, it provides for the possibility of great flexibility through changes in the survey designs, and it makes most of the costs of data collection visible and controllable.
Implementing the Project --- The report recommends the creation of an overall WCIS Advisory Committee chaired by the Director which should be composed of members of all groups involved the workers' compensation system. The report also recommends the creation of at least three task forces on "Computer System Design and Development," "Administrative Database and Ten Percent Sample," and "Survey and Special Studies" to implement the proposal. The Advisory Committee should periodically review the plans and progress of each task force for timeliness, integration with other task forces, and adherence to the plan outlined in this document and to all legislative mandates.
Reports that Can be Produced --- The system envisioned in this report should provide the basis for a number of useful reports on the state of workers' compensation in California. These include routine statistical reports pertaining to incidence of injury and work related illness and associated costs as well as annual reports on consumer satisfaction with different aspects of the operation of the system (e.g., adequacy of benefits, regulatory burden) and with the Division's performance (see Section V of the report). In order to obtain the maximum benefit of the information system, much of the data produced should be available in easily accessible public use data bases (see Section III H of this report for a discussion of confidentiality issues). A Standing Committee on Confidentiality and Public Use Data that includes representatives from the major participants of the workers' compensation system, staff of agencies that report statistics on occupational health, legislative representatives, and university-based experts on data collection and archiving would set priorities for the production, publication and release of statistical reports and ground rules and timetables for the release of public use databases.