HHS Logo
Office of the Assistant Secretary for Planning and Evaluation

Policy Information Center

Chapter III

PHS Agency Evaluation Activities

The eight Agencies of the Public Health Service (PHS) and the Office of the Assistant Secretary for Health (OASH) maintain their own evaluation programs--including the functions of evaluation planning and policy review; quality assurance through technical review; project coordination and management; dissemination of reports; and utilization of results. This chapter provides an agency-by-agency summary of the evaluation activities. Each section begins with an overview of the Agency's evaluation program, describing its philosophy, policies, and procedures. Next are summaries of the major evaluations completed in fiscal year (FY) 1994 and the evaluations currently in progress. Finally, there is a discussion of new directions for the Agency's evaluations, including priorities for future program and policy evaluation projects. Abstracts and contact persons for all evaluations completed in FY 1994 are presented at appendix A, organized by Agency. Appendix B lists all evaluations in progress, by Agency. Review criteria used by the special panel of senior editorial advisors for assessing evaluations can be found in appendix C.

Agency for Health Care Policy and Research MISSION: To generate and disseminate information that improves the health care system.

AHCPR Evaluation Program The Agency for Health Care Policy and Research (AHCPR) has designed a portfolio that responds to three types of evaluation needs:

To address these evaluation needs, an evaluation component is built into each AHCPR project. The evaluation mechanisms, which vary depending on the project, include special evaluation studies; peer review of grant applications and subsequent review by the National Advisory Council for Health Care Policy, Research, and Evaluation; the User Liaison Program (which provides information on the value of our research to State policymakers); and other evaluation efforts such as focus groups to provide feedback on AHCPR products and to provide baseline information to assist in the design of future evaluation projects.

AHCPR received delegated authority to review evaluation projects in 1992 and, as part of delegated review, established a formal structure to review proposed evaluation projects that coincides with the review of concepts for all other Agency projects. Thus, review of proposed evaluation contracts is integrated into the Agency's formal planning and budget process.

AHCPR has implemented a two-tier process for reviewing evaluation projects eligible for 1 percent set-aside funds. The process begins with an executive evaluation review to assess the policy relevance and relative priority of proposals submitted. This review is conducted by the Administrator and senior staff. The second, or technical, review assesses policy-relevant project proposals for feasibility, soundness of design, costs, potential importance of the findings, and relation to ongoing evaluation activity. This second review is conducted by the Task Force on AHCPR Evaluation Projects, a group consisting of one individual with evaluation expertise from each Office and Center.

Summary of FY 1994 Evaluations During fiscal year (FY) 1994, AHCPR completed six evaluations on two major topics: examining the process of developing clinical practice guidelines, and monitoring and evaluating health care delivery.

One of the statutory responsibilities of AHCPR is to arrange for the development, review, and revision of clinical practice guidelines. AHCPR-supported guidelines may be used by physicians, educators, and health care practitioners to help determine how diseases, disorders, and other health conditions can most effectively and appropriately be prevented, diagnosed, treated, and managed. The development of guidelines relates to AHCPR's strategic plan goal to determine what works best in clinical practice and to translate that knowledge into clinical decisionmaking. The first three AHCPR-supported evaluation studies completed in FY 1994 focused on strengthening the methodology underlying the guidelines by examining the strengths and limitations of current methods, and by identifying improved approaches and how they might be implemented.

The first project examined the methods used to conduct peer and pilot reviews for six of the guidelines. Peer and pilot review methods are used before a guideline is released to ensure its methodological soundness and scientific accuracy as well as to obtain information on its clarity, organization, and format. The study culminated in several recommendations for improving the quality of peer and pilot review.

Two other studies focused on methods for incorporating treatment cost analyses into the guideline development process. One examined the methodologies and data sources used in cost studies for eight of the guidelines and examined some of the difficulties in making cost estimates about various guideline treatment options. The other focused on identifying databases that could be used to provide information to generate cost estimates. This study found that no single database can be used to estimate the costs of recommendations in clinical practice guidelines. It is necessary to assemble data from several sources to estimate costs because claims databases sometimes lack information on certain services, and there are significant questions about the accuracy of data in various databases.

Information from these three studies has been used to inform two AHCPR reports to Congress on methods for developing guidelines and methods for including cost information in the guidelines. The findings have also been used to formulate a more uniform approach to guideline development based on what works best from the various approaches used to date.

The second area of emphasis among the AHCPR evaluations completed in FY 1994 relates to the Agency's strategic plan goal to monitor and evaluate health care delivery. Critical to fulfilling this goal is research that can be useful to policymakers by providing a greater understanding of the effects of emerging delivery systems and changes in the health care market structure. One such project focused on methods and issues related to the implementation and integration of advanced clinical information systems. The report presented a design for studying integrated health care information systems, including issues such as study feasibility, site selection, research design, and a data collection plan.

Another study examined inappropriate extension of hospitalization and the role of discharge planning, and identified research needs and potential existing data sources. Major findings include the following: 5 percent of extended hospital stays are not medically necessary; the cost of these stays was conservatively estimated at $420 million in 1987 dollars. Reasons for extended hospital stay include lack of sufficient nursing home beds and other alternate placement settings, delays in qualifying patients for posthospital benefits, and delays in appointing conservators for incompetent patients. Discharge planning appears to have only a small impact on extended hospital stays.

The final evaluation completed in FY 1994 assessed the availability of research on the cost-effectiveness of managed care health plans. The study found managed care plans had lower hospital utilization, greater use of less costly alternatives to expensive procedures and tests, greater use of preventive measures, comparable quality of care, and somewhat lower enrollee satisfaction generally, but greater satisfaction with cost. The study recommended additional research on managed care performance, including more comprehensive research projects to determine what works in managed care, rather than just whether managed care works.

Evaluations in Progress AHCPR currently supports seven evaluations that focus on its mission to generate and disseminate information that improves the health care system. AHCPR's ongoing projects provide information that will facilitate improvements in health care quality by advancing the science of quality measurement. In addition, AHCPR continues its commitment to improving the quality of AHCPR-supported clinical practice guidelines.

Two prominent FY 1995 projects relate to improving the science of quality measurement and improvement. The first, the Measurement Typology Project, was designed to collect and summarize information on clinical quality measures. These are technical quality measures used to estimate the extent to which health care providers deliver services that are appropriate for each patient's condition; whether services are provided safely, competently, and in an appropriate timeframe; and what the outcomes are from the services provided. The project summarizes 40 clinical performance measure sets--which contain a total of 1,287 performance measures--and provides a framework for evaluating the usefulness of various measures. The next phase of this project will expand and refine the typology, which ultimately will provide the foundation for an ongoing national resource of validated clinical quality measures that will assist in quality measurement and improvement.

The second project, to be completed in FY 1995, was designed to focus on consumer information needs. It developed a model consumer survey to examine how consumers perceive the quality of health care they receive, including their attitudes toward their access to care, use of specific health services, and health outcomes. This project and the Measurement Typology Project are complementary efforts to facilitate evaluation of quality from the perspectives of clinical quality assessment and consumer satisfaction.

AHCPR is also continuing its commitment to improving its clinical practice guidelines. Three projects focus on methods for improving the efficiency and effectiveness of guideline development, and three large-scale targeted evaluation projects will be completed during FY 1995. These projects will provide information on the usefulness of the guidelines for quality measurement and improvement.

An evaluation design study conducted by the George Washington University Center for Health Policy is examining methods for studying the process of guideline development. A follow-on study being conducted by the American Institutes for Research (AIR) is applying these evaluation design principles to examine factors that influence providers' views of the quality of various guidelines. Some key findings of the George Washington University study compare AHCPR's guidelines to those sponsored by other organizations. AHCPR's guidelines are found to address a broader audience (including a wider range of clinicians and clinical settings and a wider range of nonclinicians, including patients, families, payers, regulators, and purchasers); to have very detailed and comprehensive written policies and procedures for their development; to have more specialized expert resources on which to draw; and to use multiple methods for external review.

These differences have important implications for evaluating the credibility, usability, and potential impact of the AHCPR-supported guidelines on provider practice. Recommendations from this study, as well as those from the AIR follow-on project, are expected to be used to improve guideline development. AHCPR is also undertaking a project with the Institute of Medicine to examine optimal methods for selecting guideline topics.

In addition to projects to improve the guideline development process, AHCPR has sponsored a number of targeted evaluation projects. Through the use of evaluation contracts, cooperative agreements, and intramural research projects, AHCPR has supported evaluative efforts for guidelines on acute postoperative pain; urinary incontinence; benign prostatic hyperplasia; cataracts in adults; prediction and prevention of pressure ulcers; and depression in primary care.

Two targeted evaluation projects, to be completed in FY 1995, will provide important information on the feasibility of using the guidelines for quality measurement and improvement, as well as for insight into the difficulties of implementing guidelines. The first project develops, implements, and evaluates quality and utilization review criteria and educational outreach based on AHCPR-supported clinical practice guidelines. The project uses five Medicare Peer Review Organizations (PROs) to develop criteria based on three guidelines (urinary incontinence, acute postoperative pain, and benign prostatic hyperplasia [BPH]); develop and test training materials to use the criteria for case review; pilot test the criteria to assess intra- and inter-rater reliability; apply guideline-based review criteria; and compare guideline-based review with the review systems currently used by PROs. The PROs are playing an integral role in developing, implementing, and evaluating alternative educational outreach strategies based on the BPH guideline.

The second project is building on AHCPR's existing efforts to evaluate clinical practice guidelines by translating AHCPR-supported guidelines into medical review criteria for use in a variety of care settings including hospitals, health maintenance organizations (HMOs), ambulatory clinics, and physicians' offices. This project, like the project described in the previous paragraph, will assess the usefulness of alternative educational interventions for disseminating the guidelines and review criteria and for changing provider practice behavior. One component of the project focuses on cataract in adults, and the other on pressure ulcers in adults. The criteria for this project will be developed with the Department of Veterans Affairs (VA) and will be tested in a number of VA facilities. In a follow-on project, AHCPR and the VA will test alternative methods for disseminating the guideline and using the criteria as tools to assess the effect of guidelines on practice patterns.

New Directions for Evaluation AHCPR is in the process of developing a new strategic plan that reexamines current activities in relation to a rapidly changing health care marketplace. This new direction recognizes that AHCPR's programs must generate the information and tools needed to improve health care delivery and its outcomes. In addition to generating new knowledge, AHCPR's portfolio will focus on translating research findings into forms of information that actively assist consumers, practitioners, payers, and others in making effective health care decisions.

AHCPR is also responding to new challenges to government to become more efficient, flexible, creative, and effective. The Agency recognizes the need for evaluation activities to be linked to the planning process and to yield information that can inform policymaking, budget planning, and program management.

In the coming years AHCPR will continue to support research in the area of quality measurement and improvement. A Request for Applications, titled Consumer Assessments of Health Plans Study (CAHPS), was announced in the NIH Guide in May 1995. The proposed study will build on the consumer survey project described above. The project will demonstrate and evaluate the use of consumer surveys to assess health plans. AHCPR will also build on work in the area of clinical performance measurement by expanding the Measurement Typology Project to incorporate guideline-based clinical performance measures into the measures inventory.

Centers for Disease Control and Prevention MISSION: To promote health and quality of life by preventing and controlling disease, injury, and disability.

and

Agency for Toxic Substances and Disease Registry MISSION: To prevent exposure and adverse human health effects and diminished quality of life associated with exposure to hazardous substances from waste sites, unplanned releases, and other sources of pollution in the environment.

CDC Evaluation Program The Centers for Disease Control and Prevention (CDC) place a high priority on evaluations seeking to answer policy, program, and strategic planning questions. Evaluation studies are developed and selected on the basis of eight strategies to achieve its mission. These strategies are to--

The CDC Director provides annual guidance to the various Center, Institute, and Office (CIO) Directors on 1 percent set-aside evaluation activities. This memorandum generally includes information about the types of studies to be carried out with 1 percent evaluation funds. Each proposal undergoes multiple levels of review. Initial review is conducted by the Office of Program Planning and Evaluation. Subsequent reviews are completed by CDC analysts in the Office of the Assistant Secretary for Health (OASH) and the Office of the Assistant Secretary for Planning and Evaluation (OASPE). Study authors are provided with comments, questions, and recommendations made by reviewers. In addition to providing their responses, authors are given the opportunity to revise their proposals at this time.

A panel of CDC evaluators, scientists, and program managers are convened to review and rank proposals. Review criteria include (1) relevance to prevention effectiveness; (2) relative importance of the public health problem being addressed; (3) probability that the proposed project will accomplish its objectives; and (4) extent to which other CDC programs will benefit from the project. Results from this panel review are converted into a comprehensive ranking that is provided to the Director of CDC. Final funding decisions are made at this time.

Finally, staff in the Office of Program Planning and Evaluation work closely with program staff to ensure development of a clear statement of work for selected projects. Before initiation of procurements, a final ad hoc review of the project statement of work is completed.

ATSDR Evaluation Program Agency for Toxic Substances and Disease Registry (ATSDR) receives its funds from Environmental Protection Agency/Superfund appropriations rather than Public Health appropriations; therefore, ATSDR does not receive a 1 percent evalu-ation set-aside. Nevertheless, the Agency is responding to the changes mandated in its program planning and evaluation efforts by the National Performance Review and the Government Performance and Results Act (GPRA) of 1993. To meet those requirements, ATSDR staff members modified the Agency's planning process, incorporating implementation strategies and outcome/performance measures.

Prominent issues addressed in the new planning system emphasize ATSDR's commitment to improve the health of people affected by hazardous substances polluting the environment. Improvements include using exposure assessments and demographic data to identify people at risk and, more directly, assessing/addressing the concerns of customers. The new planning system provides the basis for measuring ATSDR performance and making systematic improvements as part of its internal evaluation activities.

Summary of FY 1994 CDC Evaluations CDC completed 12 evaluations in fiscal year (FY) 1994. These evaluations covered training and information dissemination, surveillance, program effectiveness, prevention, and costs of disease.

Training and information dissemination was the focus of several evaluations, two of which were highlighted in chapter II. The first was an evaluation of CDC and ATSDR training activities that assessed the training needs of State and local health departments and inventoried CDC's current training activities. It provided an example of using an evaluation to document current practice to help generate a new agenda for program action. The other highlighted evaluation was a survey of readers of the Morbidity and Mortality Weekly Report, a CDC publication of interest to public health professionals around the Nation. The survey found that this publication generally met the needs of its readers and was valued for its accuracy, relevance, and concise reporting format. Study recommendations are expected to help fine-tune this publication in response to reader suggestions.

Information dissemination activities also were addressed in an evaluation sponsored by the Office on Smoking and Health (OSH). Recommendations for key management and operational aspects of information dissemination in OSH were made based on interviews with key officials in CDC and outside the organization. The evaluation also made suggestions for strategic planning in view of OSH's evolving leadership role in the tobacco control community.

With respect to surveillance, several evaluations focused on gathering statistics for policy analysis and decisionmaking. One evaluation addressed the resiliency of the Model State Vital Statistics Act and Regulations to accommodate changes in social customs and technology of registering vital events and statistics. The results are being used to advise States about the need for revisions in the collection of vital statistics. Two evaluations are concerned with medical records in relation to national survey data. Both studies are intended to determine whether medical records confirm survey respondents' reports about selected conditions and impairments. In the first study, selected data elements of the 1988 National Maternal and Infant Health Survey (NMIHS) were compared to records maintained by the States. This evaluation assessed the quality and completeness of the information reported, identified discrepancies, and examined the nature and frequency of discrepancies. The second study evaluated the accuracy of diagnostic reporting in the National Health Interview Survey. Information collected from users of health care services was compared with infor-mation collected from the source of the health care services.

A third category of studies sought to evaluate health programs to determine their effectiveness. One study assessed the effect of burgeoning patient loads on sexually transmitted disease (STD) clinics, the impact of changing funding levels, and the increase in more elaborate patient testing. The study examined the effectiveness of services and identified factors that contributed to overburdening the delivery system. A similar type of study assessed the effectiveness and identified exemplary practices of State-based diabetes control programs (DCPs) in providing services with the potential for reducing diabetes-related mortality and morbidity. The outcomes examined were (1) the number of people reached by the programs; (2) the improved coverage provided by clinics; (3) the level of integration into ongoing medical service delivery; and (4) the effect of leveraging resources for diabetes programs. Findings showed that DCPs have had a measurable effect on diabetes services. Recommendations for program expansion were provided as part of this evaluation.

"Assessing Prevention Effectiveness: A Collaborative Effort With Selected Health Maintenance Organizations" is the first part of a two-phase study. A key component of this study includes the development of a framework and process to assess prevention effectiveness in health maintenance organizations (HMOs), types of services provided, and their potential to work with CDC. Data collected in Phase I provide the information required to move to Phase II of the study.

The purpose of the evaluation study by the Division of Cancer Prevention and Control is to collect data pertinent to program-related decisions. The evaluation focuses on components of a comprehensive breast and cervical cancer early detection and control program. In addition to these components (public and provider education, quality assurance, surveillance, screening, and followup), the combined effect of components is assessed.

The final category of evaluations sponsored by CDC developed methods to estimate direct medical costs for various diseases. One study estimated direct medical costs of chronic hepatitis B, concentrating on acute care costs. Costs from both Medicaid and private sector data were collected. Another study estimates the direct medical costs of congenital syphilis, using 1990 figures. The estimate included medical care costs for the first year, special education costs required by children with congenital syphilis, and lifetime custodial care costs, all of which were categorized by severity.

CDC Evaluations in Progress CDC has a total of 32 evaluations in progress. They fall into four general categories: surveillance/data collection studies; program evaluations; community/intervention effectiveness studies; and evaluation methodology studies. Performance improvement is a major focus of each of these studies.

Surveillance/data collection is the focus of the largest number of evaluations. For example, a study of the effectiveness of CDC surveillance for drug-resistant pneumococcal infections addresses drug resistance, which was identified as a major challenge in CDC's Emerging Infections Plan. This project will evaluate the validity of antimicrobial resistance data collected from sentinel hospitals, using CDC's sentinel hospital surveillance program. The hospital surveillance program, located in 13 hospitals in 12 States, is designed to determine the magnitude of drug-resistant pneumococcal disease and to provide clinicians with the ability to select optimal regimens of empiric therapy. By using population-based surveillance for invasive pneumococcal disease in two geographically distinct areas, this project will evaluate the extent to which CDC's sentinel surveillance program is capturing drug-resistant pneumococcal disease.

Another study that falls within the category of surveillance/data collection entails evaluation of STDs in the United States. The objectives of this study are to (1) determine the accuracy of CDC's surveillance data on STDs by comparing them with independently collected data from a survey of providers; and (2) to determine, also from a survey of providers, the extent of adherence to CDC's diagnostic and treatment guidelines for STDs and to identify ways of increasing compliance with those guidelines.

Program evaluations, also an important focus of numerous current evaluations, are being undertaken for grant programs, including the Lead Poisoning Prevention Program and the Injury Prevention and Control Program. Other studies in this group involve evaluations of the National Laboratory Training Network, the San Juan Laboratory's Dengue Hemorrhagic Fever Program, and the Fatality Assessment and Control Program.

Community-based interventions are the subject of several other evaluations. Four studies address the prevention of violence. These include projects focused on youth violence prevention, suicide in Native American communities, domestic violence medical education programs, and support systems for battered women.

Evaluation methodology is the focus of several ongoing projects. For example, one project is developing a comprehensive evaluation strategy that can be incorporated into planning, budget, and legislation for the National Center for Chronic Disease Prevention and Health Promotion.

New Directions for CDC Evaluation Evaluation studies focusing on program performance and effectiveness will continue to be of primary importance to CDC. As CDC moves toward a comprehensive performance monitoring system, focused studies in this area will be of utmost importance. Evaluations will be conducted to provide data for decisionmaking regarding the need for broader program implementation. Similarly, as programs develop and implement performance indicators, projects designed to provide data for performance measurement and to assess the effectiveness and efficacy of such indicators will be initiated.

Food and Drug Administration MISSION: To protect and promote public health through food, drug, medical device, and cosmetic regulation.

FDA Evaluation Program Systemic changes in the government management environment are strongly influencing the setting, conduct, and use of evaluation activities in the Food and Drug Administration (FDA). Three forces--all related to the mandates of the Government Performance and Results Act (GPRA) of 1993--are reshaping the evaluation function in the FDA.

1. Performance management. The reorientation of all government managers toward performance management has shifted the responsibility for program evaluation from specialized staff offices and contractor studies to day-to-day line managers. Picking appropriate program goals, establishing valid measures toward those goals, and collecting management information to record measured progress are now integral parts of the government manager's responsibility. Relearning the role of management with regard to these shifted responsibilities is a key priority. The old order--evaluation as a proprietary domain of staff offices--is a bygone luxury to FDA programs in the era of streamlined government.

2. Customer participation. Most of FDA's management measures of its own performance are also measures of the performance of the regulated industries. Thus, while the beneficiary of FDA's performance is ultimately the general public, the Agency operates in such a manner that it supplies industry with an essential component of commercial success. FDA's approval of a new drug, for example, not only satisfies a legal requirement but also assures the public of the safety and efficacy of the drug. Collaboration between the FDA and its regulated customers regarding the design and coordination of the joint responsibilities to ensure effective, high-quality products has been a revolutionary concept, but it is becoming the norm under the customer-conscious GPRA directives.

3. Increasing rigor of the rulemaking process. The FDA establishes standards of safety and efficacy through rules published in the Federal Register. Many safety and efficacy standards are also performance standards that industry is obliged to meet. Examples are the Good Manufacturing Practices regulations. Today, virtually every FDA final rule of significant magnitude includes elements found in classic program evaluations: a critique of the existing system, alternatives for better performance, performance and cost tradeoffs, reactions and suggestions of customers, and conclusions, with an action timetable for implementation. New directives from the Administration as well as possible legislative action by Congress will make the rulemaking process of the future even more like the classic evaluation process.

In sum, FDA's evaluation efforts are driven by the mandates of GPRA and its corollaries, are carried out by line managers rather than specialized evaluation staffs, and are focused in the areas of performance management, customer participation, and more rigorous rulemaking. Because it receives its funds from Agriculture appropriations rather than Public Health appropriations, the FDA does not manage a 1 percent evaluation set-aside.

Summary of FY 1994 Evaluations Fiscal year 1994 was an important test period for the new evaluation paradigm. The following examples show how evaluation has been integrated into the line manager responsibilities and interlocked with the affected customer.

Implementation of the Prescription Drug User Fee Act (PDUFA). In FY 1994, the more than 1,000 full-time employees involved in FDA's drug and biologic review process completed their first full year using the performance management goals jointly established by the FDA and industry. Line managers developed measures for evaluating progress toward these performance measures, reported the results to Congress and the several hundred industry customers of this program, and used the resulting management information to self-manage the program toward higher FY 1995 performance goals. Line managers have also focused on streamlining, achieving other review efficiencies, and reducing the regulatory burden while maintaining the high quality and standards associated with FDA decisionmaking.

The negotiation process toward medical device user fees. During FY 1994, the FDA and representatives of the medical device industry engaged in a series of meetings that constituted a joint evaluation of each other's performance in the device development process. Current performance of both parties was assessed quantitatively; new alternatives were modeled, critiqued, and cost-audited by independent accountants; and the specifications for a new program based on shared performance expectations were translated into a framework for multiyear performance goals. While the culminating legislative step was not realized in 1994, the joint effort stands as a model of a customer-oriented, performance-based program evaluation.

Harmonization of international regulatory requirements. While benchmarking has long been an important evaluation technique in the private sector, it has not been widely used in the Federal sector--and especially in regulatory agencies--because of the perceived uniqueness in function and process of the Government Agencies. At the international level, of course, benchmarking--especially for regulators--is not only possible but highly desirable. Global forums such as the International Committee on Harmonization invite such benchmarking. Their efforts are maturing and reaching the point at which specific components of the regulatory process in different countries have been compared and assessed by the managers of regulatory programs. The initial successes suggest that regulatory benchmarking in the context of international forums will be an increasingly important way for regulators to evaluate and improve their performance.

Evaluations in Progress The following evaluation objectives will influence line managers during FY 1995:

Performance management. Managers of all FDA programs are evaluating their performance measures in light of GPRA standards. Broad-based training of managers to enable them to evaluate the performance design of their responsibilities is under way. The first phase of this multiyear process toward a new standard of performance measure will end with the formulation of FDA's FY 1997 budget.

Additional user fee programs. FDA managers will finalize measures for two performance--oriented user fee programs identified in the President's budget. One is a culmination of efforts by FDA managers and industry representatives for a medical device user fee act; the other involves developing with import brokers shared expectations and performance goals for a user fee-supported electronic processing system for imports.

Commencement of negotiated rulemaking. In FY 1995, the FDA will add a new customer-sensitive dimension to its increasingly rigorous rulemaking function by beginning to implement the President's directive to promote negotiated, consensual rulemaking. This marriage of customer participation and rigorous rulemaking will receive serious attention from FDA managers.

Two major projects are currently under way:

PDUFA management. During FY 1995--the middle year of the 5-year user fee program--the FDA will use the performance data from this GPRA-style program to generate the resource alignment needed to achieve the high performance goals for FY 1997--the final year of the current user fee legislation.

Mammography Quality Standards Act of 1992 (MQSA) assessment. The FDA's second major user fee program, MQSA, is also a performance-oriented, GPRA-style program. Although it is similar in design to PDUFA, much of the data on its performance are more external to the agency. The FDA is therefore using a contract study to gather data from affected facility sites to gauge the effects of the standards on the availability of mammography facilities.

New Directions for Evaluation Changes in government management are creating systemic changes in the FDA's evaluation function. A new paradigm driven by line managers' performance responsibilities, by the imperatives for involving customers, and by directives for analytically rigorous rulemaking has replaced the traditional practice of evaluation studies guided by specialized evaluation staffs or third party consultants.

Health Resources and Services Administration MISSION: To improve health by assuring quality health care to underserved, vulnerable populations and by promoting a primary care and public health workforce.

HRSA Evaluation Program The objectives of the Health Resources and Services Administration (HRSA) evaluation program are to improve program management and policy development and to provide information that will enhance strategic planning, budget decisions, and legislative planning. Consequently, a high priority is given to studies that will (1) assess program effectiveness or outcomes, or (2) enhance the Agency's capacity to measure performance by identifying meaningful performance indicators, creating or refining databases, and developing study designs or approaches. Over the past several years, HRSA has made considerable progress in expanding such capacity.

The evaluation program in HRSA is designed to ensure that the four Bureaus and the Office of Rural Health Policy (ORHP) identify their particular information needs within a framework established by the Administrator each spring through a memorandum on evaluation planning guidance. Other staff offices may also propose studies. After study proposals are reviewed for clarity by staff of the Office of Planning, Evaluation, and Legislation (OPEL), proposals are reviewed for technical merit by a committee of senior evaluators representing each Bureau, ORHP, the Agency for Health Care Policy and Research, the National Center for Health Statistics (NCHS), the Office of the Assistant Secretary for Health, and the Office of the Assistant Secretary for Planning and Evaluation. Authors receive written comments on methodological or other design issues. Proposals are then reviewed by a committee of the Bureau Directors and other senior HRSA executives, whose role is to consider the proposals for relevance to important policy, budgetary, or legislative issues; potential to answer questions about program effectiveness or impact; and degree of attention to crosscutting topics. Proposals and committee recommendations are submitted to the Administrator for decision. OPEL staff work closely with program staff to develop a strong scope of work and other required documentation.

HRSA places major emphasis on the dissemination of information about findings and utilization of completed studies. Developing articles for publication in the professional literature and presentation at professional meetings is strongly encouraged, and draft manuscripts are typically called for as a contractor product. Briefings on completed studies of broad interest are scheduled periodically for the Administrator and other senior officials. An annual report on completed evaluation studies describes study purposes and findings, and tells how study results have been disseminated and used. Another annual publication provides brief summaries that describe the purpose and proposed methodology for studies initiated in the preceding fiscal year.

Summary of FY 1994 Evaluations Two significant studies were completed that examined aspects of the community and migrant health centers. "Evaluation Design: Evaluation of the Effectiveness of Community Health Centers" developed a study methodology to examine the efficacy and effectiveness of the centers. This work has provided a way to assess the impact these providers have on the health status of their users. "Linkage Programs: Survey of Mental Health Services" examined a joint National Institute on Drug Abuse/HRSA demonstration program designed to test community-based approaches for linking substance abuse and primary care services. The study found that the projects had been successful in creating supportive environments for the assessment, referral, and delivery of mental health services. Findings from this study will help other providers of primary care services establish the necessary alliances and treatment networks to reduce delivery system fragmentation and develop a continuum of mental health services.

"HIV Service Networks in Four Rural Areas," highlighted in chapter II, is a series of case studies undertaken to help States identify ways to improve the quality, availability, and organization of services for rural residents with HIV. The case studies described two distinct approaches to providing medical care to rural residents with HIV: (1) cooperative working relationships between HIV clinics and physicians in private practice, and (2) linkages between physicians and a medical school or rural-based regional care facility for training and backup consultation. Findings are being used to formulate HRSA policies on HIV care in rural areas and to provide technical assistance to States with large rural populations. The findings are also being used to design a rural component, funded by HRSA, for a large national study of AIDS services sponsored by the Agency for Health Care Policy and Research.

Three other AIDS-related studies were completed by HRSA in FY 1994. Two of these address the priority area of community infrastructure building. "Implementation of Title I of the Ryan White CARE Act of 1990" established a preimplementation baseline in Baltimore, Maryland, and Oakland, California, covering the organization and funding of HIV-related services, the functioning of planning councils, and the perceptions of HIV patients regarding service delivery. The Henry J. Kaiser Family Foundation subsequently funded followup surveys on these issues. "The Participation of People With HIV in Title I HIV Health Services Planning Councils" studied planning councils in Atlanta, Philadelphia, San Diego, and Seattle, and identified factors supporting the initial and sustained involvement of HIV-positive individuals in these councils. It also identified alternative methods employed by the councils to gain input from and provide feedback to HIV-positive populations in the absence of their direct participation on the councils. "Evaluation of Methods for Estimating Unit Costs of HIV Health and Support Services" developed common definitions for service, cost, and expenditure categories, and identified potential indicators of cost savings attributable to coordinated, comprehensive, community-based care systems.

"The Demographic and Treatment Characteristics of the Hill-Burton Population" obtained data on a sample of patients receiving services under the Hill-Burton Uncompensated Services Program and compared these with data on the general population. The study found that, compared with the total U.S. population, the Hill-Burton population was more likely to be female, young adult, and unmarried, with a very low household income.

Three studies dealt with health professions programs. "Survey of Beneficiaries of Nursing Education Projects" concluded that Federal support of these projects has enhanced the supply of advanced practice nurses. "The Training and Practice of Preventive Medicine Specialists: An Underutilized Resource" studied the area and, based on its findings, recommended incorporating preventive medicine into the medical school curriculum. In light of the increased attention being given to improved management practices, "Evaluation of Bureau of Health Professions Strategic Directions--Phase I" developed a set of outcome indicators and identified potential data sources and gaps relating to programs under Titles VII and VIII of the Public Health Service Act.

Evaluations in Progress The 41 studies that were ongoing as of September 1994 include four types: (1) measurement of program outcomes, effectiveness, or impact; (2) development of performance indicators, and/or data systems or databases; (3) design and/or testing of study approaches; and (4) provision of other types of information for program management or policy development.

The largest study of program performance and outcomes is the National Evaluation of the Healthy Start initiative. Begun in September 1993, this 5-year prospective study includes both process and outcomes components. The process evaluation is assessing the development and implementation of comprehensive, coordinated systems of perinatal care in the 15 initially funded Healthy Start sites. The outcomes component concerns changes in the health status of pregnant women and infants across the sites. Client-specific data from a standard data set will be used to examine the relationship among clients' sociodemographic characteristics, project-specific service use, and maternal and infant health outcomes. Secondary data, including linked birth and death records and Medicaid claims data, will be used to compare outcomes of clients residing in Healthy Start project areas with those of women living in matched comparison areas. The study also will employ a series of more qualitative methodologies to gain information about the unique features of each of the 15 projects. Four questions are being addressed. Did the Healthy Start initiative succeed? If so, why? If not, why not? And, what would be required for a similar intervention to succeed in another setting?

Two other studies are assessing the effectiveness of the community and migrant health centers programs, using national samples of grantees and including both quantitative and qualitative techniques. Additional studies concern effectiveness and outcomes of the National Practitioner Data Bank, the Healthy Schools/Healthy Communities program, and the Injury Prevention Implementation Incentive Grants program. One further study examines the effects of Ryan White CARE Act Title I funding on services for active and recovering HIV-infected drug users.

Among studies of the second type are projects to assess data systems for the Ryan White CARE Act Titles I and II programs, and to develop national integrated requirements estimates for physician assistants, nurse practitioners, certified nurse-midwives, and physicians.

The Agency also has several ongoing studies to design and test study approaches. One effort is intended to design a method to evaluate the validity of current estimates of the reported mortality rate among infants of Mexican descent residing in the United States and examine factors contributing to the reported low rate. The study includes nontraditional sources of data concerning births that occur outside the confines of the infant tracking systems. Another study is developing a self-assessment tool for use by HIV health planning councils in Title I cities and by care consortia in Title II states. Grantees and their planning councils and care consortia have expressed interest in methodologies and resources that will help them know what they are accomplishing. The tool will be sensitive to the limited resources available for planning and evaluation, providing a cost-effective method for evaluating performance. An additional study is developing a design for assessing variations in the procurement rates of Organ Procurement Organizations (OPOs). Identification of the factors associated with low procurement rates should help increase these rates.

Finally, several studies are intended to provide other types of information for program management or policy development. One exploratory project on the nature of telemedicine systems serving rural populations is providing baseline information in connection with the new HRSA Rural Telemedicine Grant program. An additional study is examining the impact of certain Medicaid waivers on Federally Qualified Health Centers. Another study is examining the relationship of medical school characteristics to graduates who choose primary care and ultimately provide care to the underserved. Still another study is exploring the impact of the Ryan White CARE Act on strategies for serving African Americans. One of the study products will be a report providing detailed examples of effective and ineffective processes and strategies for enhancing access for African Americans. The report will also identify performance indicators being used and approaches for gathering data to measure performance. Finally, a 2-year study by the Institute of Medicine is defining the place of primary care in the health care environment of the future and developing normative goals for primary care.

New Directions for Evaluation The major priority of HRSA evaluations in FY 1995 and 1996 will be studies to continue the development of meaningful measures and data systems to facilitate performance assessment. During FY 1995, HRSA is conducting a project to review the status of performance indicators for all line organizations and major programs and determine the degree to which data sources are currently available for applying the appropriate indicators. The impetus for this effort is the implementation of the GPRA, which will begin with the FY 1997 budget.

In FY 1995 and 1996, HRSA's other evaluation studies will continue to reflect the Agency's mission of expanding access to care for the underserved and such crosscutting priority areas as academic/community partnerships in health professions education, community infrastructure building, managed care, expanded roles for States, information technology, and HIV/AIDS. HRSA also expects to develop multiyear evaluation goals for selected program areas in connection with new authorizing legislation and to initiate or continue studies of effectiveness, outcomes, and/or impact in selected program areas.

Indian Health Service MISSION: To provide a comprehensive health services delivery system for American Indians and Alaska Natives, allowing for maximum tribal involvement in developing and managing programs to meet their health needs.

IHS Evaluation Program The evaluation program of the Indian Health Service (IHS) serves the Agency's programmatic and policy objectives. These objectives, developed in consultation with tribal communities, were most recently articulated in a "vision statement" of the IHS Director in November 1994. The statement calls for the IHS to continue to be the best primary care rural health system in the world. This goal is predicated, in part, on additional program emphasis in areas in which demographic and disease indicators point to a strong need for service delivery: elderly care, child abuse, women's health care, and substance abuse. Substance abuse, for example, is considered to be one of the most pressing problems facing American Indians and Alaska Natives[1] because of its toll in morbidity, mortality, and social turmoil.

The IHS evaluation program provides valid and reliable information to help the Director promote this vision for the Agency, and the evaluation program is also expected to help guide its implementation. The program traditionally has addressed three general areas--service delivery, health status, and management. The areas are interrelated, especially as the IHS faces times of fiscal austerity. If the IHS is to improve service delivery and maintain quality care, it must increase efficiency, effectiveness, and accountability. Management practices are continually being reviewed to ensure that IHS programs operate efficiently to ensure quality patient care and preventive care--the cornerstones of the IHS.

The IHS conducts technical and policy reviews of proposed evaluations--much like other PHS evaluation programs--but the program also has several distinguishing features. The most noteworthy feature is the active participation of tribes in the development and execution of evaluation studies. American Indians are recognized as primary stakeholders whose needs and concerns must be addressed, especially if the IHS is to help tribes develop and provide health care to their communities. For example, the IHS sponsors roundtable discussions with tribal leaders and Indian health program officials to elicit advice and consultation about directions for evaluations and policies. This approach to evaluation--a responsiveness to stakeholders and an orientation toward a pluralism of cultural values--is referred to by professional evaluators as a "fourth-generation" evaluation. Earlier types of evaluations, which began after World War II, were more technical, descriptive, and marked by the judgments and values of the evaluator.

Another distinguishing feature is that most evaluations are performed by contracting firms owned and operated by American Indians. This policy is an outgrowth of regulations and guidelines intended to increase the participation of American Indians and tribes in program management and service delivery.

Summary of FY 1994 Evaluations The IHS completed nine evaluations in FY 1994, focusing on service delivery, health status, and management. These areas are pivotal to the Director's vision of ensuring the delivery of high-quality primary care. Three of these evaluations deserve special notice because they address areas that are slated for renewed programmatic emphasis in the coming years.

One evaluation assessed the availability, utilization, and quality of data sets describing the health status of women and children in the Navajo area. Maternal and child health (MCH) data sets play a vital role in monitoring the health status of American Indian populations, shaping programs and assessing their effectiveness. The first phase of the project sought, through interviews and site visits, to determine whether IHS service unit staff were able to fulfill existing data requirements use the data that were generated. The project found staff to be inundated with data requests that did not completely meet their needs for the care of patients and for the promotion of maternal and child health. The second phase examined the quality of a specific automated data set on childbirth outcomes by comparing it with original chart entries. The comparisons were favorable, thus reinforcing the quality of the automated data set. The findings and recommendations of the first phase of the report are leading to a better understanding of MCH data needs and to refinements in the types of MCH data required of service area staff, along with improved vehicles for obtaining them.

Another evaluation, highlighted in chapter II, analyzed the health status of American Indians in California. It documented the health status of, and access to health care services for, those in both federally recognized and non-federally recognized tribes. These and other comparisons showed that the overall health status of both groups was poorer than that of the California population as a whole. The findings underscore the significance of maintaining and expanding coverage to the entire American Indian population of California.

A third evaluation developed long-term health care projections for alcohol-related hospitalizations. The projections--which were based on a new method called "long-term projection methodology"--sought to determine the impact of changes in hospitalizations as a result of advances and innovations in treatment, tech-nology, and behavior. The technique forecast a 20 percent decline in alcohol-related hospitalizations. While these projections were encouraging, they were insufficient to meet Healthy People 2000 objectives. This finding is one of many that have led to a renewed IHS emphasis on prevention of substance abuse. This study will be used specifically to guide evaluations such as that planned on regional adolescent alcoholism treatment centers (described below under "New Directions for Evaluation") and will provide tribes with information relevant to alcoholism program planning, policy development, and program evaluation. For example, the study informs IHS providers that the only age-gender group showing increases in alcohol-related hospitalizations is American Indian women over the age of 45.

Evaluations in Progress The IHS currently is sponsoring 21 evaluations covering service delivery, management, and prevention. Two of these evaluations underscore the IHS commitment to the development of new programs to prevent child abuse and family violence. Violence and abuse are serious problems that erode the quality of family life and can result in death, serious injury, or long-term health and behavioral problems. Family violence and child abuse are difficult to study because of underreporting and shifting definitions of what is acceptable behavior. The diversity of cultures, languages, and customs of American Indian tribes also heightens the difficulty of determining the extent of these problems and evaluating what works to prevent them.

One report, which is expected to be completed in FY 1995, is an assessment of the extent of child abuse and neglect among American Indian tribes and the ways in which the IHS responds to these problems. The evaluation also develops a model program to combat child abuse and neglect. Preliminary results from survey data indicate about 34 percent of American Indian children are at risk of becoming victims, and substance abuse commonly plays a role. The intervention program under development is being modeled after a successful program developed in 1985 by the State of Hawaii. The program provides home visitor services to new mothers, ensures continuity of medical care for children, and links families with essential services.

The other evaluation report is a case study of family violence in four distinct American Indian communities. Researchers conducted more than 100 unstructured interviews at the four sites to probe the nature and extent of family violence, which was defined as spousal abuse, child abuse and neglect, child sexual abuse, or elder abuse. They found little consensus across the four study sites about which type of family violence represents the gravest problem. These and other study findings are being incorporated into a model to guide the development of prevention programs. Eventually, this model will be available for use by an individual, family, group, or tribe to develop programs to prevent or reduce family violence in American Indian communities.

New Directions for Evaluation The IHS foresees the need for evaluations in the following areas: mental health services for urban Indians, regional treatment centers (RTCs) for substance abuse disorders, and health services for elderly American Indians. Urban American Indians, who account for approximately 56 percent of the American Indian population according to the 1990 Census, have important mental health needs. The urban American Indian usually lives in poverty and has little or no support system, in contrast to the American Indian living on a reservation among others who live at the same economic level and rely on the security of extended family and housing assistance. Indians who come to the cities encounter a physical environment, social organization, interpersonal behavior, attitudes, values, and sometimes even language that are foreign to their experience. They may suffer from feelings of isolation, depression, desperation, and anxiety, and may have problems with self-esteem. The traditional support of the extended family may be totally lacking. In 1976, Congress passed the Indian Health Care Improvement Act (P.L. 94-437), which was landmark legislation for all Indian health concerns but particularly for urban populations. Title V of the legislation specifically authorized health outreach and referral and the delivery of services to American Indian people in urban areas. Before this, health care delivery was not permitted off the reservation. The urban health program was expanded in 1988 in the Indian Health Care Amendments (P.L. 100-713). These amendments clarified the types of primary care and outreach services that could be provided, paving the way for the provision of mental health services. Mental health services are considered ripe for evaluation because the IHS supports extensive services in some urban areas and few, if any, services in other urban areas. This disparity needs to be evaluated.

Nine RTCs offer residential substance abuse treatment to Indian youth, most of them aged 12 to 19, in recognition of the fact that young people are especially vulnerable to long-term patterns of alcohol and drug abuse. Although alcoholism is decreasing among older American Indians, it remains stable in adolescents. Alcohol abuse among American Indian youth usually begins during adolescence or earlier. The RTC programs were legislated into existence by Congress and are an important part of the continuum of care for American Indian youth. Programs that began in the late 1980s and were accredited by the Joint Commission on Accreditation of Health Care Organizations are now deemed ready for evaluation because they have been in operation for several years and they appear to play an important role in reducing substance abuse. Evaluations are needed to document the impact of these programs, including outcome measures; to determine the level of agreement between referring and discharge diagnoses of individuals admitted to RTCs; to examine the characteristics and histories of youth and staff at the RTCs; and to determine the impact these centers have had in the local areas. The results can be used to redefine the operation of existing RTCs and guide the design of future centers.

Longer life expectancy among American Indians is posing new challenges for IHS service delivery. Demographics showing that life expectancy has increased from 61 years in 1972-74 to 71.6 years in 1986-88 have led to the creation and expansion of special services geared to the elderly. Evaluation of health services for elderly persons who are ambulatory, homebound, or institutionalized is considered critical if the IHS is to effectively meet the needs of this growing population. Evaluations should address the increasing need for long-term care. The emerging nature of the problem gives the IHS an opportunity to define and develop the most appropriate services for American Indian elders before the need overwhelms American Indian communities and leaves many without services. Improving the quality of life for elderly American Indians is thought to be a critical need in the next few decades.

National Institutes of Health MISSION: To discover and disseminate new knowledge leading to improved health for all Americans.

NIH Evaluation Program Evaluation is an integral part of the role of the National Institutes of Health (NIH) in the support of biomedical research, training, and public education. Evaluation studies are undertaken to ensure that NIH meets its specific goals to--

A distinguishing feature of the NIH Evaluation Program is the variety of evaluation instruments it employs. The most familiar instrument is a formal evaluation study that examines whether a program has successfully met its objectives. But NIH supports a host of evaluation strategies that go beyond traditional program evaluations. The NIH peer review system is one type of evaluation strategy: research proposals from scientists around the Nation are subjected to a rigorous assessment by fellow scientists, and only the most meritorious proposals receive funding. Other evaluation strategies include national advisory councils, boards of scientific counselors, consensus development conferences, and committees. These groups are charged with assessing a body of research to establish priorities, developing long-range goals and strategies, addressing emerging issues, identifying significant opportunities, assessing needs for new programs and activities, and recommending expansion, realignment, or continuation of ongoing programs.

The reason for the diversity of evaluation instruments lies in the nature of research. Research--especially basic research--depends on pursuing the unknown. The results of a research program and the generation of new knowledge usually cannot be anticipated. Consequently, research does not readily lend itself to the most common type of evaluation, an outcomes evaluation. Programs that provide services or promulgate regulations are most suited to outcomes evaluation because they are intended to achieve explicit, preconceived objectives. NIH attempts to evaluate its research programs with methods suited to the serendipitous nature of the research enterprise.

Another distinguishing feature of the NIH Evaluation Program is its use of the 1 percent set-aside evaluation fund strictly for programs that transcend individual NIH Institutes. The focus on NIH-wide evaluations is a self-imposed policy. NIH relies on its component Institutes, Centers, and Divisions (ICDs) to generate requests for funding of NIH-wide projects from the 1 percent set-aside, in addition to those that are centrally directed. The ICDs also conduct individual evaluations supported by their own program funds.

In June 1991, the Office of the Assistant Secretary for Health (OASH) authorized NIH to approve all set-aside funded evaluations, whatever the budget, while it maintained an ex officio presence in the review process. A two-tiered system is used to review project requests for 1 percent set-aside funding. One tier is the Evaluation Policy Oversight Committee (EPOC), and the other is the Technical Merit Review Committee (TMRC). The EPOC includes representatives from the Office of the Director, NIH, and ICD representatives at the level of Director, Deputy Director of an ICD, or Associate Director of the Office of the Director. EPOC conducts policy- level concept reviews of proposals for NIH-wide evaluation studies that use set-aside funds, establishes the overall NIH set-aside budget, and oversees the process. EPOC recommendations are approved by the Director, NIH, or designee before the initiation of any study. The TMRC is responsible for the technical review of the submissions and for recommending to the EPOC whether a project fits within departmental guidelines for the set-aside fund.

Evaluations and evaluation priorities pertaining to individual ICDs are shaped by ICD Directors and Deputy Directors. The results help ICDs and the Director, NIH, establish priorities, develop long-range goals and strategies, and review programs in terms of scientific excellence, relevance, cost, and uniqueness.

Summary of FY 1994 Evaluations The eight evaluations completed in FY 1994 addressed almost all elements of NIH's mission, from research to public education. Four of the eight evaluations are highlighted in chapter II. The evaluations summarized below illustrate the diversity of the NIH Evaluation Program through a commitment to evaluating a body of research that informs public policy, evaluating the Nation's need for scientific manpower, and evaluating public health information.

One of these four evaluations addressed the safety of selected childhood vaccines. This study was mandated by Congress under Section 313 of Public Law 99-660 to yield essential information that would help the Public Health Service draft recommendations about the use of, and compensation for adverse reactions to, vaccines against tetanus, diphtheria, measles, mumps, polio, Haemophilus influenzae type b, and hepatitis B. The study entailed an expert review by a committee of the Institute of Medicine. The committee examined all relevant medical and scientific literature on the potentially serious risks associated with currently licensed childhood vaccines. Its findings about each vaccine have been incorporated into brochures given to parents before children are vaccinated and into proposals to revise the list of compensable injuries presumed to have been caused by certain vaccines.

Another evaluation was on national needs for biomedical and behavioral research personnel. This evaluation addressed the Nation's future need for biomedical and behavioral research scientists and the contribution of NIH training grants called National Research Service Awards (NRSAs). The study was the 10th in a continuing series of reports to NIH and the U.S. Congress on this topic.

The National Research Service Award Act of 1974 consolidated all previous training authorities into the NRSA program. The Act authorized both predoctoral and postdoctoral support to individuals and to institutions. To implement the Act, NIH set up individual fellowships and grants to institutions for training predoctoral and postdoctoral students. Close to $400 million is spent annually on these training grants. A National Research Council expert committee, under contract to NIH, found that although the NRSA program is relatively small in terms of the total number of trainees, it is enormously powerful in its ability to change research emphasis and to attract the highest quality individuals to research careers. It is viewed as a prestigious, highly competitive program, and it is clear that initiatives introduced through the NRSA program can have a powerful impact on intended new research directions or constituencies. A final evaluation was performed to determine the feasibility of assessing NIH-supported research to increase condom use. This project was a feasibility study, or evaluability assessment, to determine if an outcomes evaluation could be performed in a second phase and if it would be useful to do so. The evaluation was motivated by public health efforts to prevent the spread of AIDS through the use of condoms. Its objectives were to assess the findings of condom use research efforts, guide the development of future program areas, and suggest methodological guidelines to facilitate the evaluation of future condom research programs.

The evaluation identified and inventoried the universe of condom use research studies supported by NIH grants. A reproducible methodology that combined automated database searching of NIH grants with human judgment was established. The methodology identified more than 500 relevant studies. A sample of 76 studies was examined in detail to identify how well demographic characteristics were defined, which sampling methodologies were employed, whether and what type of comparison group was used, and so on. The final report was widely disseminated. The evaluation was valuable because it generated useful methodological tools for meta-analysis and because it led to a decision not to pursue a larger scale evaluation. What emerged was a research agenda that the ICDs (seven had participated in the technical advisory group for the evaluation) could use to address the vital questions related to condom use most closely related to their institutional missions.

Evaluations in Progress NIH has 24 evaluations in progress. They range from small- to large-scale assessments, from evaluability studies to full-blown evaluations. One study builds on a longstanding role in ensuring the training and continued availability of superior biomedical and behavioral scientists through the NRSAs. The first study objective is to conduct an evaluation design study, that is, develop a detailed plan for a comprehensive evaluation of career outcomes of the predoctoral and postdoctoral trainees and fellows and the NRSA programs in which they have participated. The second objective is to develop an approach to characterize the nature and quality of the training actually experienced by present and former trainees and fellows and to differentiate between a good training program and simply good trainees. No baseline data are available on trainees, and program versus selection effects have not been studied. The third objective is to develop an approach to tap the perceptions of NIH staff, present and former NRSA trainees and fellows, and university administrators about the nature and impact of the training program. This study is expected to be completed in 1995.

A second example is a study of the Physician Data Query (PDQ), a comprehensive cancer database intended primarily for cancer health professionals. The database contains state-of-the-art treatment summaries and information on supportive care, screening, prevention, and experimental drug therapies. The objectives of the study are to survey PDQ database users to determine who is using the database and how the information is used, and to assess user satisfaction with the information and the method of retrieval, for example, CD-ROM, on-line, or hard copy.

The National Cancer Institute is directing a study to identify ways to increase target audiences. The study will produce a written summary of the activities and analyses of the project, data tapes and documentation of all questionnaire responses, and a computerized system of data collection, tabulation, and analysis. It will also include suggestions for improvements to the PDQ database to ensure that it meets user needs.

The third evaluation examines research resources available to primate researchers throughout the Nation. Through special grants, NIH supports seven Regional Primate Research Centers, a unique national network of nonhuman primate research and resource laboratories established in the early 1960s. The centers are geographically located throughout the United States, and each is closely affiliated with an academic institution. Center activities include the conduct of biomedical and behavioral research; research resource support to investigators funded by other sources; the maintenance of more than 18,000 nonhuman primates; the establishment of breeding programs to meet the centers' research requirements; conservation and preservation programs; expert professional and technical support to investigators; and training of pre- and postdoctoral professionals in primatology research. When the nonhuman primate is the most appropriate species to study, the centers provide a cost-effective response to the need for national repositories of nonhuman primates, scientific expertise, and specialized facilities and equipment.

The evaluation is assessing all aspects of the Regional Primate Research Centers program, such as its effectiveness in meeting current program objectives and guidelines, scientific distribution and emphasis of research programs, future planning decisions, and collaborations with and access by non-center investigators. Other issues to be addressed are compliance with policies and guidelines, financial management, host relationship, grant award process, peer review, status in the scientific community, and reporting and dissemination of program results. The centers are being compared with several selected non-center institutions with significant NIH-supported nonhuman primate research. The results of the study will be used to improve and refine the centers program.

New Directions for Evaluation In peer review, for example, NIH has been examining streamlining review of grant applications to make more effective use of reviewers' time. Streamlining has been tried on an experimental basis and is now being implemented fully. The results will be closely watched. The Government Performance Results Act (GPRA) of 1993 requires NIH to develop a strategic plan, an annual performance plan, and performance indicators by September 1997. This effort is currently a central focus of NIH evaluation activities. NIH is exploring a variety of indicators for potential areas in the strategic plan such as information dissemination, technology transfer, and its reinvention program. In addition, priority will be given to funding 1 percent set-aside project proposals submitted by the ICDs and the NIH Office of the Director that relate to the GPRA. Also, the EPOC will reexamine and redefine NIH evaluation priorities for the future.

Substance Abuse and Mental Health Services Administration MISSION: To improve the quality and availability of prevention, treatment, and rehabilitation services for substance abuse and mental illness.

SAMHSA Evaluation Program The Substance Abuse and Mental Health Services Administration (SAMHSA) is committed to evaluating its overall programs and individual grant projects to assess--

To the greatest extent appropriate and feasible, SAMHSA will encourage the use of comparable data elements and instruments across its evaluations in order to work toward a comprehensive evaluation system and to minimize respondent burden.

SAMHSA conducts grant programs under a variety of legislative authorities. These authorities can generally be grouped into two types: services and demonstrations. The evaluation required for a particular grant program depends on the type and purpose of the program. SAMHSA evaluates its service programs to provide information to program managers about the accountability of Federal funds. Its evaluations of demonstration programs generate new knowledge to lead the field in developing policies that improve services. Evaluation results from the two types of programs are used to improve the performance of SAMHSA and the field of substance abuse and mental health services. Program and evaluation staff work together to identify the questions or goals each grant program should address and to propose appropriate evaluation strategies.

The two types of grant programs represent two facets of SAMHSA's mission: service delivery and knowledge development. SAMHSA's leadership in the field depends on the successful interaction of these two facets of its mission. SAMHSA uses evaluation to identify effective approaches to prevention and treatment and uses service delivery funds to provide incentives to the field to implement effective approaches. Major emphases of SAMHSA's mission are the development, identification, and dissemination of effective strategies and systems for treatment and prevention.

SAMHSA is now implementing a new, integrated model of evaluation and planning. Strategic planning will identify priorities, such as managed care, that drive the development of grant programs and evaluations. In compliance with the Government Performance and Results Act (GPRA), SAMHSA is attempting to improve performance by identifying performance goals and indicators as part of its strategic planning process. The formulation of programmatic and evaluation priorities will include consultation with the SAMHSA and center advisory councils and with other experts in the fields of evaluation and service delivery. Early and continuous coordination of program planning and evaluation design will result in the articulation of evaluable programmatic objectives. Evaluations will show how well the overall grant programs have achieved their objectives, and SAMHSA will translate these results into information that can be used for program and policy development. The strategic planning and policy development processes will then use these results to refine SAMHSA's priorities and objectives.

This evaluation policy will help SAMHSA achieve its goal of continually informing policy and program development with knowledge culled from past performance. In this way SAMHSA can best serve its customers by enhancing the quality of public substance abuse and mental health services.

In compliance with the Public Health Service (PHS) guidelines for the technical review of evaluations, SAMHSA has a standing review committee of PHS staff who are evaluation specialists from Agency for Health Care Policy and Research, Centers for Disease Control and Prevention, Health Resources and Services Administration, National Institutes of Health, and SAMHSA. Representatives of the Office of the Assistant Secretary for Health and the Office of the Assistant Secretary for Planning and Evaluation serve as ex officio members, and the SAMHSA evaluation coordinator serves as committee chair.

Evaluation project proposals are generally prepared by SAMHSA program staff in the various centers. The standing committee reviews each proposal on the following criteria: clarity of evaluation objectives; appropriateness and feasibility of the specifications for evaluation design and methods; appropriateness of the plans for dissemination of results; and use of previous relevant evaluations and existing program data systems.

Each proposal must clearly state the relationship of the evaluation to SAMHSA's overall policy priorities and evaluation program. The committee generally reviews proposals for broad, comprehensive evaluations of SAMHSA program activities, such as the cross-site evaluations of demonstration grant programs. It does not review the evaluation proposals of individual grantees.

Summary of FY 1994 Evaluations During FY 1994, SAMHSA completed six evaluations. The six studies reflect some of the population groups SAMHSA has identified as being in greatest need of substance abuse and mental health services. These include pregnant and postpartum women and their infants, children with serious emotional disturbance, high-risk youth, and the homeless mentally ill.

The evaluation of a demonstration program serving pregnant and postpartum women and their infants examined the results of improved coordination, availability, and accessibility of health and substance abuse-related services. A substantial number of the women served by these programs reduced their substance use, and their babies were generally healthy. SAMHSA will use these findings to encourage improvements in the performance of service programs for substance-abusing pregnant and postpartum women and their infants.

SAMHSA completed the design for an evaluation of a comprehensive community services program for children with serious emotional disturbance. The purpose of the evaluation is to collect information on client and systems outcomes. The findings should help both policymakers and service providers improve the performance and cost-effectiveness of mental health services available to children.

Another evaluation conducted this year examined the cost-effectiveness of a drug treatment enrichment program in curbing drug use among students receiving training at Job Corps Centers. Job Corps Centers are residential employment and training programs for high-risk youth. Students participating in the enrichment program showed a significant decrease in drug use compared with a control group. The findings of this study are useful in verifying the effectiveness of substance abuse treatment and in informing the field about appropriate substance abuse services for high-risk youth.

An evaluation of a demonstration program for homeless mentally ill adults is highlighted in chapter II of this report. This study examined the results of improved linkage, integration, and availability of housing and social and medical services. The program was successful in reducing homelessness and improving the mental health of study participants. These results are encouraging communities to better integrate services for the homeless mentally ill.

An evaluation of demonstration projects for community partnerships to prevent substance abuse has produced information on the successful implementation of substance abuse prevention approaches that employ a comprehensive, coordinated, communitywide strategy. Evaluation results of the impact of the partnership program are forthcoming. The Second Report to Congress on Alcohol and Drug Abuse Prevention presents preliminary results of the National Structured Evaluation (NSE). The NSE is a systematic assessment of substance abuse prevention approaches implemented across the country. Final results of this study, to be presented in a third and final report to Congress, will have important implications for improving the performance of substance abuse prevention programs.

Evaluations in Progress SAMHSA currently has 10 major evaluations under way in the following general areas: program accountability, evaluation of demonstrations, reinforcing behavioral health, and commitment to customer service. Each type is described below, with some examples.

Program accountability evaluations are undertaken to inform program management and help managers refine program operations. This is the primary type of evaluation conducted on SAMHSA's service grant programs. For example, the children's mental health service program has an ongoing evaluation. The evaluation will yield continuous information on program implementation and on outcomes for children and families served. This information will be used for reports to Congress, feedback to grantees, program development, and performance improvement.

Evaluations of demonstrations are designed to generate new knowledge for policy development. The primary purpose of SAMHSA's demonstration programs is to generate new knowledge to lead the field in the development of policies that improve services. For example, evaluation results on substance abuse prevention for high-risk youth will allow policymakers to draw inferences about the effectiveness of certain interventions for this population. The Job Corps evaluation will continue to assess an enriched substance abuse treatment program for adolescents. An evaluation of the program for access to community care and effective services and supports (ACCESS) will provide information for the design of ongoing service programs at the Federal, State, and local levels.

Other SAMHSA evaluations strive to reinforce behavioral health as a cornerstone of public health. In the context of managed care and health care reform, SAMHSA's interagency collaborations promote the integration of services and reinforce the role of alcohol, drug, and mental health services as a critical component of general health care. Illustrations of such activities include collaboration with the Administration on Children and Families to evaluate a substance abuse prevention program for pregnant and postpartum women and infants.

The National Treatment Improvement Evaluation Study (NTIES) was designed as one of a family of studies of substance abuse treatment, including two studies conducted by the National Institutes of Health (NIH). This collaboration will enhance SAMHSA's ability to draw inferences about treatment effectiveness. SAMHSA will continue working with the Department of Housing and Urban Development to design a method by which therapeutic communities can be incorporated into public housing areas. An HIV/AIDS mental health services demonstration program is jointly funded by NIH and HRSA.

Because of its commitment to the recipients of alcohol, drug abuse, and mental health services, SAMHSA uses evaluation to guide the field in improving the integration of services and their responsiveness to the needs of clients. For example, the evaluation of the community partnership program will determine the part-nership's ability to achieve integration and responsiveness at the user level. Primary purposes of the community coalitions demonstration program are to increase the efficiency of services at the community level and to enhance the service delivery system's responsiveness to its users. The evaluation of the program will assess the effectiveness of this approach.

SAMHSA has promulgated a comprehensive treatment model that includes a broad array of treatment and ancillary services targeted for individuals whose substance abuse has severely affected many domains of their lives. The services include housing, child care, transportation, social services assistance, legal aid, and mental and physical health screenings and treatment, as appropriate. Every SAMHSA grant announcement for substance abuse treatment will require that grantees implement the comprehensive treatment model. Evaluations will assess the effectiveness of the model.

New Directions for Evaluation SAMHSA is designing its evaluation activities to guide programmatic and policy decisions. The evaluation activities will complement the SAMHSA and PHS strategic plans and will respond to emerging trends such as managed care and health care reform. For example, in FY 1995 SAMHSA will initiate a program of State contracts to evaluate managed care and treatment outcome. The program will provide support to States to conduct short- and long-term studies at the State and provider levels of the effects of managed care on substance abuse treatment access, cost, and outcomes. SAMHSA will support States' use of evaluation methodologies including State financial, provider, and client databases.

These evaluation activities will reinforce the critical role of behavioral health in general health care. Results of the evaluations will ensure accountability and generate new knowledge. SAMHSA will use these results to improve integration of services and responsiveness to the needs of individuals suffering from or at risk for alcohol, drug abuse, and mental health problems.

Office of the Assistant Secretary for Health MISSION: To provide public health advice to the Secretary of Health and Human Services and executive direction to the PHS Agencies through coordination and oversight of their programs.

OASH Evaluation Program The primary role of the Office of the Assistant Secretary for Health (OASH) is to coordinate and develop evaluations across the entire Public Health Service. The Assistant Secretary for Health (ASH) issues guidance each fiscal year to the PHS Agencies for preparation of their plans outlining evaluation priorities and projects proposed for implementation in that fiscal year (as mentioned in chapter I). OASH reviews the Agency plans to identify gaps in evaluation data needed for program or policy development. OASH also identifies potential program areas that could benefit from a collaborative evaluation.

In addition to its PHS-wide coordination function, OASH uses the 1 percent set-aside funds to initiate its own evaluations or policy studies, usually in consultation with relevant Agency program offices. The projects conducted by OASH are generally short, quick-turnaround policy studies to support initiatives of the ASH. Occasionally, OASH will initiate or participate in large-scale evaluations, particularly those that cut across PHS program areas.

The OASH review and approval of evaluation projects is performed by the Evaluation Policy Review Committee (EPRC), which makes recommendations to the ASH on OASH evaluation priorities for each fiscal year; reviews project proposals for relevance to PHS strategic planning objectives; and provides advice to the ASH in determining the OASH use of 1-percent evaluation funds and allocation of those funds to approved projects. Among the criteria used by the EPRC to review proposed projects are relevance to ASH priorities for policy, legislative, budget, program development, or crosscutting PHS issues; relevance to departmental and PHS strategic planning objectives and related evaluation strategies; coordination with other PHS and departmental offices; and relevance to the Healthy People 2000 objectives. OASH evaluation staff review all EPRC-approved projects for technical quality and feasibility, providing feedback to OASH program offices on project proposals and statement of work for contract procurements. In addition, the staff serves as a resource group for individual technical assistance and consultation to project officers in developing evaluation proposals.

Summary of FY 1994 Evaluations In FY 1994, OASH completed 16 evaluations in diverse areas of public health: health care reform, immunizations, adolescent and school health, nutrition, primary care, emergency preparedness, international health, and other smaller regional studies of health services delivery. Most of these evaluations focused on policy analysis to support initiatives in health care reform. Despite the failure of health care reform legislation, these studies are relevant because health care reform proposals are still being considered by Congress and because market-based reforms are occurring in the absence of legislation.

One study on the returns on investment in public health examined the contributions of public health programs in reducing the incidence of preventable illnesses and injuries. This work summarized the effectiveness of public health strategies from a cost-benefit analysis perspective to underscore the importance of the public health infrastructure in realizing the two primary goals of health care reform: to expand access to care and to control costs.

Two related studies examined the implications of health care reform on the health care workforce. The study on a proposed Graduate Nursing Education Account, which is highlighted in chapter II, documented the projected scarcity of advance practice nurses such as nurse practitioners, certified nurse-midwives, certified registered anesthetists, and clinical nurse specialists. It analyzed several options for channeling funds to support advance practice nursing education to meet future needs. Another study examined the supply, training, and distribution of primary care providers as part of health care reform efforts. It examined the rationale for new proposals to control the supply of new physicians, proposals for an all-payer pool for graduate medical education, and estimates of the number of advanced practice nurses that would be needed as substitutes in the case of a reduction in the number of physicians.

In recent years, OASH has given priority to evaluations of PHS immunization policies and programs. One study, also highlighted in chapter II, examined the economic underpinnings of the vaccine supply. A major finding was that States that supply vaccines at low prices to physician offices and encourage parents to have their children vaccinated have higher rates of immunization. In another study, OASH developed a design to assess the functional capabilities of the Vaccine Adverse Event Reporting System (VAERS). This design project will help PHS officials evaluate VAERS to determine whether it is providing reliable and accurate information, and how health care providers, manufacturers, and consumers perceive the effectiveness of the system.

Support for nutrition policy studies has been another OASH evaluation priority. In FY 1994, OASH completed a project with the American Institute of Nutrition to examine state-of-the-art techniques for measuring nutritional status and to make recommendations for a core set of nutritional status indicators for low-income populations. Development of these indicators will be important for the development of national and State nutrition monitoring systems, particularly in expanding the survey coverage of low-income populations.

Adolescent health, including school heath education programs, is another OASH program evaluation priority area. Two projects in this area, both described in chapter II, were completed in FY 1994. The first study sought ways to evaluate the academic benefits of school health programs. It developed a general framework, with alternative research designs, for assessing the effects of school health interventions on students' school performance. The second study was a focus group study of the knowledge, attitudes, behaviors, beliefs, and environments of youth engaged in high-risk behaviors. The results of this study will be used to develop appropriate health messages to youth about high-risk behavior.

Evaluations in Progress OASH has 14 program and policy evaluations under way in two major areas: health care reform and population-based services. In addition, OASH is supporting several smaller projects investigating crosscutting PHS issues and several evaluations specific to OASH program offices. Some of the evaluations that are likely to be completed in FY 1995 are described below.

With respect to health care reform, OASH is continuing to examine two major issues: the role of Federal, State, and local health information systems in achieving public health objectives, and the impact of health care reform on the health care workforce and the role of academic health centers (AHCs).

Health information systems at all levels of government are the focus of three projects. One is developing sets of indicators that communities can use to assess performance and to foster closer collaboration between public health and personal health care systems and employers in achieving specific public health goals. Another is reviewing current State laws, policies, practices, and environments governing public health reporting systems, health care information systems, health data confidentiality, and access and linkage policies. This project will identify aspects of existing laws, policies, or practices that create barriers to standardization, integration, data linkage, and access relating to health information systems within and across States. Third, OASH is supporting the National Committee on Vital and Health Statistics in a project to conduct analyses for the development of uniform health data sets for enrollment and encounter information from health care providers in reformed health care systems. Data derived from the uniform data sets will provide a foundation for monitoring and evaluating public health programs as well as systems of care in the framework of national health information networks.

Health care reform's impact on the health care workforce and on the AHCs, where many health professionals are trained, is the subject of two projects. First, OASH initiated a project to study how the growth of competitive health care financing and service delivery systems based on managed care affects the financial support available to AHCs and graduate medical education. The project focuses on AHCs in three cities: San Diego, Minneapolis-St. Paul, and Washington, D.C. Another project is examining the extent to which teaching hospitals associated with the AHCs are reorganizing and/or reducing services and other activities in areas in which competitive health plans hold a substantial share of the market. The project will assess the impact of such changes on hospital training programs for physicians.

In addition to these analytic studies, OASH will support a series of roundtable discussions among leading national experts in government, academic, clinical, and research settings regarding the role of AHCs in an era of health care reform. The purpose of these discussions is to exchange views and reach a national consensus on questions of the potential impact of health care reform on AHCs. Many questions are being raised about their mission, financial capability, and contribution to patient care and public health; how they can best participate in a national shift from training specialist physicians to training primary care practitioners; how health care reform affects clinical research; and what strategies may best protect the quality of clinical research and its contributions to public health.

The OASH evaluation program has also given priority to evaluations in the area of clinical preventive services. The Office of Disease Prevention and Health Promotion (ODPHP) will soon complete the design for the "Put Prevention Into Practice" initiative, a package of resource materials for health personnel in clinical settings to promote more effective use of clinical preventive services for their patients. The design will be implemented in late FY 1995. The results of this evaluation will help national primary care provider organizations determine the optimal use of these materials in their respective clinical settings.

ODPHP is also supporting a major effort to improve methodologies for evaluating the cost-effectiveness of clinical preventive services. A group of specialists in cost-effectiveness analysis, both inside and outside government, is working to develop guidelines to apply cost-effectiveness methods to clinical preventive services. The group will address current controversies about these methods and propose alternatives for improving the comparability of cost-effectiveness analyses.

As mentioned above, OASH often initiates evaluations of programs or policy issues that cut across the PHS Agencies. Several OASH projects in progress illustrate collaborative evaluations. Since 1993, OASH has worked with the Food and Drug Administration (FDA), the National Institutes of Health (NIH), and the Substance Abuse and Mental Health Services Administration to evaluate the departmental methadone regulations, a project expected to be completed in FY 1995. The study is examining the impact of Federal regulations on the provision of methadone treatment services and on the development of new anti-addictive medications. The final report will contain recommendations for improving the current methadone regulations.

In another project, OASH is working with the Centers for Disease Control and Prevention, the FDA, and NIH to study HIV transmission to hemophiliacs through blood products. In the early period of the AIDS epidemic, before the invention of rapid laboratory tests for identifying HIV infection, blood transfusion services for hemophiliacs operated without effective HIV screening or production methods to eliminate the virus. The Secretary of the Department of Health and Human Services, Donna Shalala, launched an investigation in 1993 to get a more complete understanding of the events that occurred in those early years. The results, expected in FY 1995, will be helpful for ensuring the safety of the Nation's blood supply against future challenges.

OASH also uses evaluation funds to support projects that measure the effectiveness of program activities managed by the various OASH Offices. For example, the Office of Minority Health presently is evaluating its HIV/AIDS Education/Prevention Grant Program. The grants program is providing assistance to community-based organizations in developing innovative approaches for reaching persons living in minority communities who are at risk for HIV infection. The evaluation is now analyzing data on disease prevention intervention models that have a likelihood of successful replication in other minority communities.

Another example of OASH evaluation activity is the efforts of the OASH Office of Research Integrity (ORI) to examine the consequences of being accused of research misconduct, as well as the consequences for the whistleblower. The results of these two projects will be useful to ORI for monitoring institutional compliance with PHS regulations on research misconduct, planning educational activities, and making improvements in PHS policies and procedures for responding to allegations of research misconduct.

New Directions for Evaluation The PHS FY 1995 Strategic Plan has a goal to strengthen the public health infrastructure. In recognition of the danger faced by health departments as a result of years of shrinking budgets, the PHS Agencies and OASH are striving to augment the capability of the Federal Government, tribal governments, States, and communities to identify and address high-priority health problems for their populations. One approach is to expand the public health knowledge base by supporting an appropriate balance of focused and multidisciplinary research in the areas of basic biological science, clinical medicine, public health practice, behavioral and social sciences, epidemiology, health systems and services, nutrition, and occupational and environmental health. In coming years, OASH will support the development and implementation of measures that will assess the capacity of State and local public health agencies to perform population-based functions and the extent to which States meet PHS program goals.

Managed care is changing the way health care is paid for and delivered and shifting responsibilities for care. This has serious implications not only for vulnerable populations but also for public health "safety net" providers, academic health centers, and public health agencies. This year, OASH will participate with the PHS Agencies and the Health Care Financing Administration to examine the impact of managed care on public health systems, primarily through demonstrations of the Medicaid Waiver (Section 1115) in five States. The Medicaid Waiver allows States to experiment with alternative approaches to improving the access and quality of health care services for the uninsured and reducing the costs of care. The analysis is expected to focus on the number of eligible participants, the costs to providers, the quality and types of care, and the access for special populations--such as American Indians and Native Alaskans, children with special health care needs, people in underserved areas, substance abusers, and the chronically mentally ill.