Hemophilia refers to a group of rare hereditary disorders of blood coagulation. The most common are defects of clotting factors VIII (hemophilia A) or IX (hemophilia B) and affect approximately 1 in 10,000 and 1 in 40,000 persons in the United States, respectively (1). The combined prevalence for these two single-gene, single-protein, x-chromosome-linked defects is estimated to be 13 cases per 100,000 males or approximately 17,000 cases in the United States. Persons with hemophilia exhibit varying degrees of factor deficiency which correlate with their disease severity, and about 40 percent have the severe (< 1% of either factor VIII or factor IX) form of the disorder. Among the congenital bleeding disorders, von Willebrand disease (vWD) is the most common, affecting perhaps as much as 1% - 3% of the U.S. population (2-4).

Prior to the 1980s hemophilia-associated morbidity, mortality, and disability were due to excessive bleeding, particularly into weight-bearing joints and the central nervous system. The median age of death for persons with hemophilia was about 11.4 years from 1831 to 1920, but had reached the mid-20s by 1960. During the 1960s and 1970s, a better understanding of the medical basis of the disorders and improved therapy in the form of clotting factor concentrates lowered mortality and yielded nearly normal life expectancy for such persons. Since 1982, however, the AIDS epidemic, the result of HIV-transmission from infected plasma-derived clotting factor, has had a significant impact on mortality (5,6).

Treatment for the severe forms of hemophilia A and B and for their associated complications place large demands on health care resources. The annual cost of highly purified, viral-inactivated or recombinant clotting factor concentrates for the treatment or prevention of bleeding in persons with severe hemophilia can approach $150,000 per patient. Treatment for the preventable complications of the disease increase health care costs substantially beyond this fixed cost of replacement therapy(7), thus impacting significantly on the total U.S. health care budget. Conservative estimates suggest that the current cost of hemophilia care in the United States (including the cost of replacement product) may be as high as $1 billion annually.

Several important considerations influence the design of adequate health-care delivery systems for persons with hemophilia. First, hemophilia care is very specialized and appropriate training and experience is needed to avoid poor therapeutic decisions that can lead to severe disability and mortality. Second, because the disease is rare, maintaining the availability of trained and experienced physicians often can be achieved only by concentrating patient care into specialized centers. Third, the when complications occur, they are extremely expensive to treat which places a premium on a preventive approach to care. Finally, the organization of care must be structured to enable optimal delivery of care and allocation of resources.

Based on these considerations, in 1975 the Federal government established a network of specialized hemophilia treatment centers (HTCs) designed to provide a public health approach to the care of persons with hemophilia by integrating prevention and clinical practice. The goals of this approach were to 1) reduce the complications of the bleeding disorders in this population; 2) reduce the number of hospitalizations; 3) normalize life expectancy, 4) enable them to maintain a state of health adequate for employment; and 5) reduce the demands on the health care system by providing comprehensive primary and secondary rather than tertiary care. This centralized direction of care also provided a means of coordinating health care resources as well as evaluating outcomes among individuals with rare diseases. This information could then be used to improve services and ensure that effective programs were maintained. In addition, it was hoped that such a system could serve as a model for health care delivery for other chronic diseases.

Data are the essential to establish and maintain good health-care delivery systems. In the early 1970s, data from hemophilia patients were used to show that many of the disabling effects and much of the early mortality caused by hemophilia could be reduced or prevented by providing training for patients in self-therapy and home care. In the United States, these data served as the basis for the comprehensive care approach that was implemented in the HTC system. These data also provided the justification for applying similar principles of care to hemophilia populations in other developed countries, thus doubling the life span of those affected.

After the HTC system was established data were needed 1) plan local health care strategies and set priorities;2) target specific problems; 3) motivate government agencies to take action; 4) disseminate accurate information to the press; 5) gain financial support; 6) mobilize the patient community for action; and 7) measure the impact of the program to justify the continued support of governments and private organizations.

To collect these data, however, a surveillance system was needed. To develop the system, a working group comprising epidemiologists, clinicians, and other individuals with knowledge of the issues important to the hemophilia community. This group established the goals for the surveillance system, selected the type of system that would be used, and determined what data would be collected. Other steps included identification of necessary resources, establishment of functional systems, and the development of a data collection instrument.

Because hemophilia is a chronic disease of low prevalence, developing surveillance and other data collection systems presents special challenges . Most surveillance systems are designed to work with acute diseases or highly prevalent chronic diseases. Registry systems are one logical approach; however, because of privacy issues, including those raised during the HIV epidemic, a registry system for hemophilia has not been established. Other traditional sources of surveillance information, such as death certificates, voluntary reporting, and surveys, provide some types of information but are limited by issues such as timeliness and the degree of participation.

Two sources of data that may be appropriate for hemophilia are medical record reviews and active case reporting as part of a health-care system. Each of these methods has its limitations. Medical record review is time consuming and costly and can only provide retrospective data. The major issue complicating active surveillance is the need for an elaborate, organized reporting network within the health-care system. Such a network requires a high level of motivation and a common sense of purpose within the participants of the system.

To achieve a common sense of purpose, prevention program should seek a consensus on potential uses of data, types of surveillance models that might be used as a pattern for the system, and what data may be available. Generally, two types of approaches are most commonly used: the public health database system and the clinical study database system. The former comprises the minimal amount of data necessary to measure trends and identify specific problems, whereas the latter consists of extensive data sets. For most applications, the public health database system is the most practical.

The process of identifying data needs and uses is done out by a group that includes consultants or participants representing the patient community, and representatives from public health and other organizations that will use the data. This group will determine the direction and type of system to be used, and the content of the data collection instrument. Because these representatives often have competing needs, the availability of resources will determine which needs will be addressed. Although the practicality of collecting data for each need should be reviewed, such a review should occur only in the latter stages of the group's discussion so as not to inhibit ideas.

To select the most appropriate system for data collection, the program should review existing surveillance systems models with regard to their suitability for collecting data on the hemophilia population, set priorities for surveillance content, and devise potential surveillance questions for each priority content area to be addressed by the system. Once these issues have been addressed, the type of system needed often become apparent.

The discussion of possible data collection systems and content should also consider the following aspects: location of data, barriers to data collection, cost of data collection, extent of data, and the sensitivity, and validity of the data. The program will have the most difficulty achieving agreement on the data collection instrument. Since it should address the needs and potential uses of the proposed system. Before being included in the data collection instrument, each data topic should be discussed thoroughly and be prioritized according to a criteria scale established beforehand. The volume of data desired will likely exceed that which may be realistically be collected, and the priority scale will allow only the most valuable data to become part of the collection system. While the group will want to proceed directly to the composition of questions to be included on the instrument, such an approach may be disastrous. We have found the following approach to this process to be the most efficient and effective:

1. State the goals of the data collection. These goals should be the driving force for determining what data are to be collected. Collect data only for specific needs and uses as defined by these goals. Possible goals might include:

a. Determine the prevalence and incidence of hemophilia (e.g., disease manifestation, demographics).

b. Study the occurrence of complications and the use of health-care resources over time (e.g., nature/type of complications, source of medical care). .

c. Develop population-based data on the social and economic impact of hemophilia and its complications (e.g., medical care, hospitalizations, lost days of work).

d. Detect emerging situations that require intervention.

e. Understand the demographics of HIV/AIDS.

2. Identify the data topics needed to meet the stated goals. These topics will be chosen based upon the priority criteria previously determined by the group. The topics should be specific and data should be obtainable using the resources available. Trying to collect data that require more resources than available will be very discouraging and may cause the entire system to fail. Limit the data elements to the minimum required to answer the questions. The data system is not a medical chart; its success will depend upon its simplicity.

3. Establish definitions for each data topic. These definitions will determine what data will be collected and the minimum amount of data necessary and will ensure that the data are uniform. For instance, if hemophilia is defined differently among participants, then the data are not compatible. This step will also allow data to be abstracted by persons of varied training backgrounds. The definitions will determine the guidelines for inclusions of data. No questions should be developed until these definitions are determined.

4. Develop the data items. This is a key step in the development of a useful and effective data collection tool. There are many aspects to constructing a data collection instrument that contribute to its effectiveness in obtaining the best possible information in the most efficient and effective way. It is crucial that epidemiologists, statisticians, or others with experience in designing data collection instruments have a role in this process. An additional consideration at this point is to make sure that all data requested by the instrument will be available in the setting in which the instrument will be used. For example, it is unreasonable to expect that persons working in a clinic will be able to provide information that is only available in a hospital.

5. Field test the data instrument. This step should be done by the participating centers and is crucial to determine the feasibility of collecting the needed data. Field testing will often identify problems with missing data, definitions, and interpretation of results. After the field test, the definitions and data collection instrument should again be modified to correct any problems encountered.

The first phase in establishing a functional system is to determine the logistics. The who, what, where, and how must be evaluated according to available resources. In many respects, the broad scheme will be primarily determined by the criteria previously discussed and the types of resources available for the system.

Planning the details is the most time-consuming task. Many details have to be considered. Will the personnel be volunteers or salaried? Who will abstract the data? How will errors be detected? What are the communication channels? Who will do the analysis? How will conflicts in data be resolved? How will the data be tabulated? How will the data collection forms be printed? Who will pay for postage? How and to whom will the results be communicated? How and by whom will personnel involved in data collection and management be trained? Who will provide technical assistance?

The goals of data collection for the hemophilia public health program in the United States are to determine the burden of disease, the source and extent of health resources utilized, and the level of disability and death produced by the disease burden. In addition, the program needs a monitoring system to evaluate the impact of programs which were directed at reducing these burdens. Especially important issues include blood-transmitted infectious agents and the level of joint and liver disease resulting from chronic bleeding episodes or as a complication of blood-borne infectious agents.

To accomplish these goals, two different systems are utilized, the Hemophilia Surveillance System (HSS) and the Universal Data Collection System (UDC). Each was designed for specific purposes and is providing valuable information to federal agencies and health care providers to improve the health care of persons with hemophilia by prevention and early intervention.

HSS is an active public health surveillance system which identifies and collects data on all patients with hemophilia A and hemophilia B in six states. The states include Colorado, Georgia, Louisiana, Massachusetts, New York, and Oklahoma which comprise approximately 20 percent of the U.S. hemophilia population (1). Operation of the system is through state health departments collaborating with hemophilia treatment centers and hemophilia associations in these states. Beginning in 1995, surveillance staff in each state, under the authority of the state health department, implemented methods of case finding best suited for a given locality. Standardized data are abstracted from medical and clinical records and include demographic and basic clinic information, detailed information on source of care and reimbursement, number of bleeding episodes and the amount and sources of clotting factor used, the results of testing for exposure to infectious diseases, information from comprehensive assessment of joints, and complete information on all hospitalizations. Data are also collected concerning the immediate and underlying causes of death for individuals who die during the surveillance period. The data are stored in local databases and periodically transmitted without patient identifiers to CDC for analysis. All transmitted data are screened for omissions, inconsistencies, and possible abstraction or data entry errors.

Because this system is extremely labor intensive and expensive, it was envisioned to collect data for only five years. However, it has already provided invaluable, previously non-existent data on the hemophilia population and their health care needs as well as information on target areas for prevention. It has also provided information demonstrating the effectiveness of the comprehensive care approach to hemophilia care in reducing mortality.

Universal Data Collection (UDC) is a nationwide surveillance system developed in response to a congressional mandate to CDC to reduce or prevent the complications of bleeding disorders. Designed to prospectively collect a uniform set of clinical and outcome data, it focuses primarily upon blood-borne diseases and joint disease. Begun in May 1998 and conducted by the CDC's Hematologic Diseases Branch, UDC will eventually supplant other surveillance activities in this community such as the HSS.

Data for UDC are gathered at federally funded HTC across the United States. HTC staff obtain information on participant demographics, bleeding disorder diagnosis, treatment regimens and factor replacement products used, history of illness due to and treatment for or vaccination against blood-borne infections, and history of joint disease and its affect upon daily activities. Staff perform joint range of motion measurements and draw blood for testing for hepatitis A, B, and C viruses and HIV-1 and HIV-2. Participation is voluntary and open to persons who meet at least one of the following criteria:

1. Age 2 years or older and diagnosed with a bleeding disorder due to congenital deficiency or acquired inhibitors in which any of the coagulation proteins are missing, reduced, or defective and have a functional level of less than 50 percent.

2. Age 2 years or older with von Willebrand disease diagnosed by a physician.

Persons with thrombophilia, coagulation protein deficiencies due to liver failure, or an exclusive diagnosis of a platelet disorder are specifically excluded. Persons must consent to both data collection and blood testing in order to participate. The goal is to collect a complete set of information and a blood specimen from each participant annually.

Staff complete an "Annual Visit" form and a "Laboratory" form for each participant each time he or she participates. A "Registration" form acknowledging consent is also completed once for each participant. For eligible HTC attendees who decline to participate, staff complete a "Patient Refusal" form. The Registration, Annual Visit, and Patient Refusal forms are sent to CDC where their information is entered into the UDC database. The Laboratory form is not used for data collection by CDC but is used by the hepatitis testing laboratory when interpreting test results; the form is routed to this lab by way of the serum bank with aliquots of the blood specimens. Personal identifiers of participants are not reported to CDC or the hepatitis testing lab. Instead, the HTC generates a random, yet unique and unchanging, identification number for each participant using computer software supplied by CDC. All communications between CDC, the hepatitis lab, and the HTC concerning the participant are conducted using this anonymous number. Only the HTCs maintain a list linking the identification number with the patient.

Blood specimens (plasma) are collected and shipped to the CDC serum bank in Lawrenceville, Georgia. The first aliquot of each specimen (as well as any remaining after allocation for routine screening) is banked at this facility to create a specimen pool which may be accessed if special viral investigations become necessary. Additional aliquots are distributed to the HIV testing laboratory at CDC in Atlanta and to the Eugene B. Casey Hepatitis Laboratory at Baylor College of Medicine in Houston, Texas. Appropriate EIA and Western blot tests are performed for HIV-1 and HIV-2. In rare cases, additional tests, such as the P1 and P2 protein dot blot tests, may be employed to distinguish the HIV subtypes. Testing for hepatitis A, B, and C is performed as outlined in the attached testing algorithms. Results of serologic tests are entered into the UDC database and reported to the HTC for use in patient care.

Analyses of these data will be directed towards identifying and describing the occurrence of complications among various subgroups of the hemophilia community so that prevention efforts can be appropriately initiated and implemented. Reports of these analyses will be periodically distributed for dissemination to all participating HTCs. Results of viral testing will be used to determine prevalence and seroconversion rates of viral infection among persons with bleeding disorders, as well as to monitor for significant increases which may prompt special investigations. At regular intervals, the data collected from each of the 12 federal hemophilia treatment center regions will be distributed in electronic form to the coordinators of each region for distribution to the HTCs for their use.

Prevention of transfusion-transmissable infectious diseases is a high priority for public health programs in patients with hereditary bleeding disorders. The UDC system will provide an "early warning system" for the nation's blood supply by establishing a serum bank on all HTC patients participating in the program, thereby allowing CDC to respond quickly to concerns regarding potential contamination of the blood supply. The UDC and HTC network will provide an important model for testing prevention effectiveness in patients with hereditary bleeding disorders and following their health outcomes. These mechanisms will allow CDC, in collaboration with health care providers in the HTCs, State Department of Health Agencies, the National Hemophilia Foundation, the Federal Drug Administration (FDA), the Maternal and Child Health Bureau of the Health Resources and Services Administration (HRSA) and the National Heart Lung, and Blood Institute (NHLBI), to quickly test and implement aggressive prevention strategies developed for virtually any of the complications of hereditary bleeding disorders.

In summary, surveillance programs for patients with hereditary bleeding disorders have been deployed widely by CDC across the United States through a network of interactive programs in State Departments of Public Health and Hemophilia Treatment Centers. These surveillance programs are providing valuable information for the design and implementation of prevention strategies, which have been developed in collaboration with other involved health care agencies, health care providers and patient advocacy groups. This powerful prevention coalition can use this unique system to rapidly respond to public health problems among persons with hereditary bleeding disorders.

REFERENCES

1. Soucie JM, Evatt B, Jackson D and the Hemophilia Surveillance System Project Investigators. The occurrence of hemophilia in the United States. Am J Hematol. 1998; In Press.

2. Montgomery RR, Coller BS. Von Willebrand disease. In: Colman RW, Hirsh J, Marder VJ, Salzman EW, eds. Hemostasis and Thrombosis: Basic Principles and Clinical Practice, 2nd ed. Philadelphia, Pa: JB Lippincott Co, 1994:134-168.

3. Rick ME. Diagnosis and management of von Willebrand's syndrome. Med Clin North Am. 1994;78:609-623.

4. Miller CH, Lenzi R, Breen C. Prevalence of von Willebrand's disease among US adults. Blood. 1987;70 (suppl):377a (Abstract).

5. Chorba TL, Holman RC, Strine TW, Clarke MJ, Evatt BL. Changes in longevity and causes of death among persons with hemophilia A. Am J Hematol. 1994;45:112-121.

6. Eyster ME, Schaefer JH, Ragni MV, Gorenc TJ, Shapiro S, Cuttler S, Kajani MK, Abrams J, Barron LE, Odenwelder A, et.al. Changing causes of death in Pennsylvania's hemophiliacs 1976-1991: Impact of liver disease and acquired immunodeficiency syndrome. Blood. 1992;79:2494-2495 (letter).

7. Smith PS, Teutsch SM, Shaffer PA, Rolka H, Evatt BL. Episodic versus prophylactic infusions for hemophilia A: A cost-effectiveness analysis. J Pediatr. 1996;129:424-431.