Development of Better Information and More Effective Information Tools
Information is fundamental to all of the immunization issues that have been discussed in this report. A variety of systems collect some immunization-related information, but none of them is comprehensive. In general many of the steps that the committee believes will increase timely immunization among preschool children require better information and information systems. Information needs begin with the very basic task of assessing immunization rates for children of all ages and for every recommended vaccine. Those organizations and individuals with responsibility for ensuring that children are immunized cannot be held accountable unless information on whether they are fulfilling that responsibility is available. With disease surveillance, outbreaks of vaccine-preventable diseases can be identified and targeted for special attention. A reporting system for adverse events following immunizations provides essential information for assessing vaccine-related risks. Providers need records that allow them to identify children who should be immunized, and they may also need to track a complicated vaccine inventory. With information from their own records or other sources, providers can assess their performance: Are their patients receiving appropriate immunizations? Are their practices consistent with the Standards for Pediatric Immunization Practices (CDC, 1993b)? Families need information about the immunization status of their children and, more generally, about the importance of timely immunization. Information must also be available to evaluate the effectiveness of interventions.
AN IMMUNIZATION INFORMATION SYSTEM
The information resources that are currently available cannot answer critical questions about the immunization status of individual children or overall immunization levels among groups of children. The committee observed that data collection at the national, state, and local levels often is not coordinated. National surveys, for example, can rarely provide state-level results. The population-based data, which are essential for public health assessments and public accountability for program performance, are often unavailable. Furthermore, data collection and analysis remain impractical for many providers because they must rely on time-consuming manual procedures.
The committee felt that the success of the United Kingdom's immunization program clearly demonstrated the value of having good information and using it effectively. Immunization coordinators, providers, and families all need to know about children's immunization status. By monitoring the performance of health districts and individual providers, the United Kingdom's immunization program has been able to identify problem areas and, by comparing districts and providers, promote improvements. Providers receive a complete list of children for whom they are responsible and information on whether those children have had the appropriate immunizations. The United Kingdom has also made careful use of a somewhat different kind of information—market research—to guide the development of their public information messages and to target the delivery of those messages. All of these pieces make essential contributions to the program's success.
In the United States, there is renewed interest in immunization information and information systems.1 Immunization questions added to the NHIS in 1991 have produced the first national data since the last round of the U.S. Immunization Survey in 1985. CDC's random-digit dialing surveys, mentioned earlier, are expected to begin producing data on immunization levels for the IAP areas (states and 24 cities and counties) by the end of 1994. Even though telephone surveys are not ideal (e.g., they cannot include families that do not have a telephone and may have difficulty including respondents who do not speak English), this new program will produce the country's most comprehensive information on immunization coverage. Planned validation
In the 1960s, birth records were used for immunization surveillance and tracking. The limitations of the technology of the time made the process time-consuming and inefficient, and it was eliminated in the 1970s. From 1959 to 1985, the U.S. Immunization Survey provided annual estimates of national immunization rates. Apparent success in controlling vaccine-preventable diseases led to discontinuation of the survey. See Johnson (1992) for a more detailed history.
studies using NHIS data should provide analytic tools that can compensate for some of the limitations of telephone surveys.
The other activity that promises major improvements in immunization information is the development of immunization registries and tracking systems. Systems are being developed independently by commercial firms, individual providers, HMOs, communities, and some states. In addition, Congress is considering legislation that would provide federal funds to support the development of state registries, which would enhance the ability of states to define and respond to immunization needs among their populations.
The greatest value will be derived from systems that encompass all children and that are used by all providers. The committee sees merit in aiming for a national system of state-based registries. Registering children when they are born (or arrive in this country), as is done in the United Kingdom, ensures that the complete population of children will be known and that a child's immunization records can follow the child to any provider in any state.
In addition to computer-based registries and tracking systems, information tools such as family-held immunization records remain useful. They are an information resource for the family and can be used by providers to encourage families to take responsibility for having their children immunized. In some cases, registries can be used to generate records for a family's children.
MOVING TOWARD A BETTER SYSTEM
At the same time that advances in technology are creating new opportunities to obtain, store, and use information on immunization services for children, changes in the ways in which vaccines are purchased are creating an incentive to have accurate information on the demand for vaccines. Because the Vaccines for Children Program will provide states with enough federally purchased vaccine to immunize Medicaid-eligible and uninsured children, states want to know the size of that population and the immunization status of those children. In the past, states played a larger role in purchasing vaccines for these groups, but financial constraints gave states little incentive to make accurate estimates of any increase in demand.
Clearly, immunization registries and tracking systems can help states with their vaccine accounting and public health assessments. These systems will also benefit children and their families by facilitating individualized outreach and follow-up. They can help providers identify those children who are due for immunizations and assess compliance with immunization guidelines. It must always be clear, however, that registries and other information systems are tools for improving immunization levels and not ends in themselves.
THE POWER OF POLITICAL WILL
Implementing a state immunization registry makes various technical demands such as maintaining a central record system, establishing a unique and identifiable record for each child, and giving providers throughout the state access to those records so that they can read and add to them. To some, this process may seem too difficult and time-consuming, but when funding and political will exist, it seems possible to overcome such hurdles.
The Texas Health Commissioner made this point to the state legislature by reminding legislators that within only 6 months Texas was able to implement a statewide lottery system. The state installed a computer system for ticket purchases in many large and small stores in cities and rural communities throughout the state, tied each one to a central computer in the capital, and developed the software to process millions of 13-digit numbers each day.
Registry and Tracking Projects
One of the more widely known registry and tracking projects is the All Kids Count program funded by the Robert Wood Johnson Foundation. In 1992, 1-year demonstration grants were awarded to 23 cities, counties, and states to work with providers and the broader community to develop computerized immunization registry and tracking systems. In 1993, with additional funding from the Annie E. Casey Foundation, 16 projects received 4-year implementation grants.
Jeffrey Goldhaggen described the Cleveland project, which aims to improve the delivery of primary care to children as well as increase immunization rates. It integrates an appointment reminder and health information component, which are accessed by Touch-Tone telephone, with a regional computerized medical record system, which accepts record updates submitted by fax. The telephone-based portion of the system is used to register children, track immunization appointments, and generate appointment reminders and follow-up messages.
Providers transmit some information about the immunizations that they administer through by telephone and more detailed medical record information by fax. They can access information in the system to check a child's
immunization status, for example, or to generate reports, official documents, or bills to third-party payers. Identification codes and other mechanisms are used to control access to information in the system.
In 1990, the Northern California Kaiser Permanente HMO implemented an immunization tracking system. Steven Black explained that the central database can be accessed by the pediatric clinics, advice nurses, and emergency departments at the 29 centers throughout the Northern California region. Because children receive even routine care at more than one site, the centralized data system is essential. Kaiser uses the system to generate immunization reminders for families and information for providers seeing patients. Kaiser also is beginning to use data produced by the system for evaluation of service delivery and research. Appropriate methods for identifying missed opportunities for immunization for Kaiser's clinics and individual providers still must be developed.
Registry development is being supported by others as well. Private vendors are beginning to offer immunization tracking modules as part of clinic registration and management systems. Thomas Vernon noted that Merck is involved in such projects in Chicago and San Antonio. A subcommittee of the National Vaccine Advisory Committee (NVAC) has been one forum for developing policy on immunization registry systems.
Current Limitations and Concerns
There is much enthusiasm for immunization registries and tracking systems, but the workshop discussion pointed to some limitations that must be overcome. Currently, many providers (including some clinics) use manual record-keeping systems. Comprehensive registries, however, must inevitably rely on computer-based systems to manage the volume of information that they will contain, and participation by all providers will be essential if such registries are to fulfill their promise.
Some providers may be reluctant to adopt computer-based systems to improve immunization tracking in their practices because of the cost of equipment, software, and training or apprehensions that such systems will create additional administrative burdens. Some information tools can be used without computer systems. For example, the automated dialing device described by Eugene Dini can be programmed manually to place telephone calls and deliver taped reminder messages (e.g., Stehr-Green et al., 1993a), but it also can be linked to computerized registry systems. The telephone and fax interface used by Cleveland 's All Kids Count registry also minimizes the technology burden on individual providers. Alan Kohrt, a pediatrician in private practice, encouraged linking immunization information systems with providers ' computerized billing systems to avoid the need for duplicate data entry.
Because existing registry systems cover only portions of the population, accounting for in- and outmigration is a problem. In community-based systems, these untracked changes in the population create uncertainty as to the true size of the denominator for immunization rates. For HMOs, enrollment records establish the size of the member population, but turnover can be substantial and members may receive care from other sources. Accurate assessments of immunization levels among member children require that HMO records include information on immunizations received from all other sources.
Expansion of registries will eventually require unique identifiers for patient records. Some advocate using the Social Security number, but others feel that its use poses serious risks to privacy (Gostin et al., 1993). All registry systems must ensure that they protect the confidentiality of their records (IOM, 1994b).
Although the relationship between immunization registries and the information systems that health care reform is expected to establish is not yet clear, the issue is being studied by the NVAC subcommittee. Under any system, population-based, individual-level data, such as those available from immunization registries, will remain essential for public health purposes. The committee underscores the concern expressed at the workshop that public health data needs are not well understood and must receive adequate attention in health care reform plans.
BIBLIOGRAPHY OF STUDIES ON IMMUNIZATION PRACTICES
The committee identified another specific immunization information need: a comprehensive and annotated bibliography of studies on immunization practices. Although further research remains necessary, work that has been done over many years should be applied to current efforts to improve immunization rates among preschool children. In fact, workshop participants observed that many of the current discussions echoed those of several years earlier when the concern was immunization of school-age children.
The committee felt that such a bibliography should focus on two broad categories of studies: (1) those on barriers to immunization that examine who is not immunized and why and (2) those on interventions that assess what approaches are succeeding in improving immunization rates and in what setting they are successful. It should include not only published reports but also unpublished reports, which can be especially difficult to locate. Information on these studies would be valuable to a variety of audiences including clinicians, researchers, public health officials, and policymakers.
IMPROVING INFORMATION FOR IMMUNIZATION
Steps to Take in the Short Term
Surveys on immunization coverage. CDC's quarterly telephone surveys of IAP areas will provide valuable state- and community-level information on the immunization coverage of preschool children and on progress toward the 1996 goal of 90 percent coverage for 2-year-olds. Further analysis linked to NHIS results should overcome some of the limitations of telephone surveys.
Available communication and information tools. Individual providers and public health departments can make more use of available communication and information tools. Automated dialing technology, for example, is well-established and can be applied in many settings to improve patient outreach and follow-up. Some providers, especially those without computer-based tracking capabilities, may also be able to make more use of family-held records that the family brings for every visit. If accurately maintained, such a record allows any provider to assess a child's immunization status.
Registry and tracking systems. States and communities can endorse existing demonstration projects for registry and tracking systems and can encourage the initiation of new projects with currently available public and private resources. Legislation pending in Congress would provide federal grants to states to develop and operate statewide registries. These projects must be responsive to the needs of individual providers and the community. If they create demands without providing a recognizable benefit, they may be difficult to sustain.
Immunization tracking in HMOs. HMOs can implement immunization tracking in their patient record systems. Every location within an HMO system should be able to retrieve a child's complete record and enter new information about immunizations and other care that was provided. Because immunizations are almost always a covered benefit in HMOs, children are likely to receive most of their immunizations through the HMO. For these systems to be effective, however, they must include information about any immunizations received outside the HMO.
Methods to audit delivery of services. CDC, researchers, or private vendors can develop methodologies, and perhaps software, that private providers, particularly those in small office-based practices, can use to audit their delivery of immunization services. Some office-based systems can be linked to larger registries under development.
Steps to Take in the Longer Term
Expansion of registry and tracking systems. States can promote the coordination and expansion of registry and tracking projects with the aim of developing a statewide system. Federal leadership in Congress and the Public Health Service can encourage the further development of a national system of state-based registries.
Population-based data after health care reform. Parties concerned about immunization (e.g., federal agencies, state and local health departments, providers, and community organizations) can monitor the effects that health care reform measures have on the organization of health services and information systems to ensure that the population-based data necessary to assess the immunization status of preschool children are available.