|
Background
In 1981, after early reports of Pneumocystis carinii pneumonia, Kaposi's
sarcoma, and other opportunistic infections in young homosexual men in Los
Angeles, New York, and San Francisco, the Centers for Disease Control and
Prevention (CDC) began surveillance for a newly recognized constellation of
diseases, now termed the acquired immunodeficiency syndrome (AIDS). CDC
developed a surveillance case definition for this syndrome and initially
received case reports directly from health care providers and state and local
health departments. As the epidemic spread, state and local health departments
assumed responsibility for AIDS surveillance, and by 1985 all states had
regulations requiring physicians and other health care providers to report AIDS
cases directly to the state or local health department. These health departments
then share the reports with CDC, which produces the national AIDS surveillance
data set.
The goals of AIDS surveillance have been to monitor both trends in AIDS cases
and the scope of severe morbidity due to infection with the human
immunodeficiency virus (HIV). AIDS surveillance data are used to allocate
resources for patient care, target HIV prevention programs, and evaluate the
impact of public health recommendations. Advances in the understanding of the
epidemiology and manifestations of HIV infection and changing diagnostic
practices, however, present multiple challenges to those analyzing and
interpreting the AIDS surveillance data. The following are a few examples:
-
A wide variety of persons are at risk
for HIV, including men who have sex with men, injecting drug users, person who
received a transfusion or who were tissue transplant recipients before March
1985, heterosexual partners of infected persons, children born to infected
mothers, and persons with mucous membrane or percutaneous exposure to blood or
body fluids of infected persons (e.g., health care workers). Because men who
have sex with men comprise such a large proportion of the total number of AIDS
cases, trends in this subgroup will overshadow those in other groups unless the
data are examined separately. Analysis of data, without regard to specific
subgroups, may conceal information or lead to misinterpretation of the data.
-
The etiologic agent of AIDS, HIV, has
been identified, and diagnostic tests for infection with this virus have been
developed. As a result, the surveillance of AIDS, initially dependent on the
presence of certain indicator diseases specific for the infection, was expanded
in 1985, 1987, and 1993 to include additional conditions (some conditions may be
less specific for HIV infection) in the presence of laboratory evidence for
infection, and in 1993 to include HIV-infected persons with laboratory evidence
of severe immunosuppression. The addition of these conditions to the AIDS case
definition has affected trends in reported AIDS cases, as well as trends in
reporting of AIDS-defining opportunistic conditions.
-
Diagnostic practices have changed over
time and vary geographically. AIDS is now a common diagnosis in many hospitals
and clinics, and definitive diagnostic tests for manifestations of HIV infection
(e.g., Pneumocystis carinii pneumonia or esophageal candidiasis) may not be
done. HIV testing is not available for all patients and some patients choose not
to be tested. Geographic variations in diagnostic practices and surveillance
procedures, and changes over time could markedly affect trends in AIDS
surveillance.
Return to Manual
Source of AIDS Surveillance Data
CDC maintains national AIDS surveillance through receipt of AIDS case
reports submitted by individual state and local health departments. Health
departments report cases electronically through a CDC-developed
microcomputer system. All 50 states, the District of Columbia, U.S.
dependencies and possessions, and independent nations in free association
with the United States (Puerto Rico, the U.S. Virgin Islands, Guam, American
Samoa, the Republic of Palau, the Republic of the Marshall Islands, the
Commonwealth of the Northern Mariana Islands, and the Federated States of
Micronesia) report AIDS cases to CDC.
Although state and local health departments share AIDS surveillance data
with CDC, the responsibility and authority for AIDS surveillance rests with
the individual health departments. Like any reportable disease, the
completeness of AIDS reporting reflects how actively health departments
solicit case reports. Historically, disease surveillance systems have been
categorized as passive or active, i.e., health departments may passively
receive case reports from health care providers, depending on health care
providers to know and comply with reporting requirements; or they may
actively contact and interact with health care facilities or individual
providers to stimulate disease reporting, sometimes directly assuming the
primary responsibility of reporting cases from large or high-volume
institutions.
CDC provides funding and technical assistance to health departments to
actively stimulate AIDS case reporting and has encouraged them to take an
active rather than passive approach to AIDS surveillance. Through
surveillance cooperative agreements supported by CDC, health departments are
encouraged to identify health care facilities that serve AIDS patients and
work closely with these facilities to encourage reporting. They are also
encouraged to send newsletters to health care providers and attend
professional organization meetings, and to use other data sources to
identify AIDS cases, including death certificates, laboratory reports, and
tuberculosis and tumor registries. States vary in the structure and
organization of their surveillance systems and, therefore, in the
completeness of their case reporting (see below).
Return to Manual
Case Definition
Before HIV was identified as the etiologic agent for AIDS, CDC defined a
case of AIDS (for surveillance purposes) as a disease, at least moderately
indicative of a defect in cell-mediated immunity, occurring in a person with
no known cause for diminished resistance to the disease. Such diseases
included Pneumocystis carinii pneumonia, Kaposi's sarcoma, and many other
serious opportunistic infections (see American Journal of Medicine, March
1984, pages 493-500). With identification of HIV as the causative agent for
AIDS and the availability of laboratory tests to detect HIV antibody, the
case definition was expanded to reflect an increased understanding of HIV
infection in 1985 (see CDC's
Morbidity and Mortality Weekly Report, June 28,
1985, pages 373-375) and in 1987 (see
Morbidity and Mortality Weekly Report,
August 14, 1987, supplement, pages 3S-15S). These revisions applied to
persons with laboratory evidence for HIV infection. Among diseases added
in
1985 were disseminated histoplasmosis, chronic isosporiasis, and certain
non-Hodgkin's lymphomas. Among those added in 1987 were extrapulmonary
tuberculosis, HIV encephalopathy, and HIV wasting syndrome. In children,
recurrent, serious bacterial infections were also added. In addition, the
1987 revision allowed certain indicator diseases to be diagnosed
presumptively based on clinical presentation rather than "confirmed" by
laboratory or diagnostic methods.
To be consistent with standards of medical care for HIV-infected persons and
to more accurately reflect the number of persons with severe HIV-related
immunosuppression who are at highest risk for HIV-related morbidity and most
in need of close medical follow-up, the surveillance definition was expanded
on January 1, 1993 (see CDC's
Morbidity and Mortality Weekly Report,
Recommendations and Reports, December 18, 1992). This expansion includes all
HIV-infected adults and adolescents who have less that 200 CD4+ T-lymphocytes/µL
or a CD4+ T-lymphocyte percent of total lymphocytes less
than 14, or who have been diagnosed with pulmonary tuberculosis, invasive
cervical cancer, or recurrent pneumonia. The addition of pulmonary
tuberculosis, recurrent pneumonia, and invasive cervical cancer in
HIV-infected adults and adolescents to the 23 clinical conditions listed in
the 1987 surveillance definition reflects their documented or potential
importance in the HIV epidemic.
While the reported incidence of AIDS increased only 3 to 4 percent as a
result of the 1985 revision, the 1987 revision greatly increased the numbers
of reported cases. Roughly one fourth of all adults/adolescents who were
both diagnosed and reported in the year following the 1987 revision were
reported based only on the additional criteria included in the 1987
revision. Furthermore, the proportion of cases meeting only the revised
criteria was higher in Hispanics and non-Hispanic blacks than in
non-Hispanic whites, higher in heterosexual injecting drug users, and lower
in men who have sex with men. The 1993 revision has had substantial impact
on the number of reported cases. The immediate increase in case reporting
was largely attributed to the addition of severe immunosuppression to the
definition; a smaller impact was due to the addition of pulmonary
tuberculosis, recurrent pneumonia, and invasive cervical cancer, since many
persons with these diseases also have a CD4+ T-lymphocyte count
of less than 200 cells/µL. The early effects of expanded surveillance were
greater than long-term effects because prevalent as well as incident cases
of immunosuppression
were reported after implementation of the expanded
surveillance case definition. In recent years, the effect on the number of
reported cases has been smaller. Due to the large number of cases reported
based on criteria in only the revised case definitions and to the
inconsistent use of the revised case definitions in different populations,
analyses of trends in AIDS cases must take these revisions into account.
Return to Manual
Case report form
Separate case report forms are used for pediatric patients
(patients less than 13 years of age at the time of diagnosis) and
adult/adolescent patients (patients 13 years of age or older at the time of
diagnosis). Although the forms are similar, the pediatric form includes
behavioral risk information on the child's mother. These forms are completed
by the health care provider or by the AIDS surveillance staff in the local
or state health department. In addition, a laboratory report of an
AIDS-defining condition sent to health departments may initiate a case
report. In these cases, follow-up with the health care provider is required
to obtain complete information.
Names
are retained by the state or local health department and are converted
to an alpha-numeric
code called “soundex” for use by CDC.
CDC does not receive names of persons with AIDS. Because more than one state
may report an individual case, CDC screens reported cases by soundex
code,
date of birth, sex, and state of residence to cull presumed duplicate
reports. States also cooperate in this process by reporting
out-of-jurisdiction cases to the patient's state of residence.
The variables available on the AIDS data set are listed in
the next section. However, a few deserve special comment.
-
Vital status. Patients survive for a variable amount
of time following the diagnosis of AIDS. Because death usually occurs after
the initial report to CDC, case reports may not be updated to reflect the
change in vital status. As a result, reporting of deaths among AIDS patients
may be delayed or incomplete. However, states are required to perform
periodic reviews of death certificates and state death registries to
identify unreported cases, and to update vital status of known cases. In
addition, 16 states participated in a special project to match their case
registries to the National Death Index to assess the completeness of
reporting and to identify deaths among cases that died out-of-jurisdiction.
-
Exposure category. Some patients may have more than
one mode of exposure to HIV. For surveillance purposes, AIDS cases are
counted only once in a hierarchy of exposure categories. Persons with more
than one reported mode of exposure are listed in the category that appears
first in the exposure hierarchy, except for men with both a history of
sexual contact with other men and injecting drug use. They make up a
separate exposure category.
-
AIDS definition category. Patients may develop
additional conditions indicative of AIDS after their initial AIDS diagnosis.
The case report form may not be updated to reflect additional conditions.
Some persons reported as meeting only the immunologic criteria may have
concurrent or prior opportunistic infections or conditions that are not
included in the case report. Therefore, cases reported as meeting only the
criteria added to the case definition in 1993 may include persons who meet
the criteria in 1987 definition.
-
Date of diagnosis. CDC collects dates of diagnosis
for each AIDS-indicator disease, and, for patients with severe
immunosuppression, the date of the CD4+ T-lymphocyte test. From
this information, a single date of diagnosis is calculated for each patient;
it is the earliest of these dates.
Return to Manual
Delay in Reporting
The timeliness of AIDS case reporting to CDC is dependent on
a number of factors, including the volume of cases reported from a state or
locality, the cooperation of health care providers and medical institutions,
the availability of staff to complete case report forms, and changes in the
case definition. In many instances initial case reports are incomplete and
require additional follow-up by state and local health department staff,
including reviews of other record systems and contact with health care
providers.
Based on estimates calculated using AIDS surveillance data
reported between 1995 and 2000, about 50 percent of all cases were reported
to CDC within 4 months of the date of diagnosis, but about 20 percent were
reported more than 1 year after diagnosis. Delays vary widely among
geographic, age, exposure, sex, and racial/ethnic categories. They are
substantially longer for pediatric cases and shorter for AIDS cases
previously reported with HIV infection, for example. Due to the reporting
delay, the number of cases diagnosed during any period often exceeds the
number reported during that period. This is particularly important in
examining trends over time, since many cases in recent periods of time will
not yet be reported.
To account for delays in the reporting of cases, the
variable adjwgt is included in the data set. This variable may be
used to weight each case on the data set and obtain adjusted case counts.
For example, summing adjwgt for cases would estimate the number of
cases diagnosed through the time period covered by the data set that will
eventually be reported to CDC. To use this variable, select the adjustment
weight option from the Tools menu. Once you turn the option on, all
subsequent tabulations will be adjusted for reporting delay. The adjustment
weight and resulting tabulations are not reliable for cases diagnosed during
the most recent 6 to 9 months.
Return to Manual
Effect of CD4 Reporting on AIDS Case Trends
As a result of the case definition change in 1993, trends in
AIDS case counts showed an artifactual peak early in 1993, even after
adjustment for reporting delay. To examine trends over time using
a constant case definition, i.e., diagnoses of opportunistic illnesses that
were included in the 1987 or the 1993 case definition, CDC developed methods
that estimated incidence of 1987 or 1993 definition opportunistic infections
for cases that met only the 1993 immunologic (CD4+) criteria.
These estimates showed that the number of diagnoses of AIDS-defining
opportunistic infections increased during 1992 and 1993 by approximately 2
percent and 3 percent, respectively (see Morbidity and Mortality Weekly
Report, November 18, 1994). The temporary distortion of the AIDS
incidence curve caused by the 1993 expansion of the AIDS case definition had
almost entirely waned by 1996.
Return to Manual
Effect of Therapy on AIDS Incidence
Continuing the pattern first observed from 1995 to 1996,
AIDS incidence decreased again from 1996 to 1997 and from 1997 to 1998.
These decreases are mostly due to the effect of therapies for HIV infection
and AIDS, which have altered the natural history of HIV infection and slowed
progression to AIDS. AIDS incidence increasingly represents persons who were
not diagnosed with HIV infection until they developed AIDS, persons who did
not access treatment, or persons for whom treatment failed. Caution should
be used when interpreting trends in AIDS incidence; the contribution of
these effects to the AIDS incidence curve is currently being evaluated. See
Morbidity and Mortality Weekly Report,
September 19, 1997 and
April
24, 1998.
Return to Manual
Early Reporting Dates
Before 1990, CDC occasionally received reports on patients
before they met the CDC AIDS case definition. If such patients were later
diagnosed with AIDS, the diagnosis date on their record (when they first met
the CDC definition) would be after the report date (when CDC first received
information about the patient). Such records should be excluded from certain
analyses, such as survival analysis and analysis of reporting delay. CDC's
AIDS surveillance data base no longer receives reports on patients who do
not meet the AIDS case definition.
Return to Manual
Follow-up of Reported AIDS Cases
AIDS case records maintained at CDC contain all information
reported to date from state and local health departments. As patients
progress through their illness, additional conditions may be reported, or
the patient's vital status may change. However, not all health departments
have the resources to routinely follow-up patients for additional
information. For this reason and because many patients move out of the
reporting health department's jurisdiction, CDC records do not always
contain all current information for each patient.
AIDS cases reports that do not include mode of HIV exposure
information are routinely followed up by state and local health departments.
As of December 1999, excluding cases which were not yet investigated, mode
of exposure information has been identified for 78 percent of cases.
Twenty-one percent of cases were closed with incomplete information because
the patient died, declined interview, or was lost to follow-up; 1 percent of
cases remained without a reported risk for HIV infection after complete
investigation (see Centers for Disease Control and Prevention.
HIV/AIDS
Surveillance Report, 1999;11 (no.2): 27). The demographic profile of
persons who remain without risk information is more similar to that of other
persons reported with AIDS than with the general U.S. population.
Return to Manual
Evaluation of AIDS Surveillance
Cases of AIDS may not be reported to CDC for a variety of
reasons. The diagnostic tests needed to confirm the diagnosis of certain
AIDS-indicator conditions may not be performed, or physicians and hospital
personnel may fail to report cases to the health department. Further, some
patients with HIV disease may be ill or die from diseases or conditions not
included in the current AIDS surveillance definition or from causes
unrelated to their HIV infection.
Both CDC and state and local health departments have
commissioned a variety of studies to evaluate the completeness of AIDS
surveillance. Most evaluation projects have used alternate data resources if
they are independent of routine case finding, such as death certificates,
hospital discharge records, and laboratory records. Individual records from
these alternate sources have then been matched against records in AIDS
surveillance data bases. If an alternative source is found to be a
productive source of case reports, it may be added to routine case finding
methods. Evaluation projects have varied in size and scope (e.g., varying
numbers of ICD-9 codes from death certificates or computerized discharge
records), geographic area covered, detection of both inpatient and
outpatient cases, and time frames. In general, evaluation studies suggest
that reporting of AIDS cases is fairly complete; but, depending on the
setting and evaluation method used, the level of reporting completeness may
vary. High prevalence areas for AIDS appear to have more complete reporting
than low prevalence areas. Following implementation of active case finding
under the 1987 case definition, with funding support from CDC, completeness
of case reporting increased in most areas and was estimated to be more than
85 percent complete (see Journal of Acquired Immunodeficiency Syndrome,
1992;5:257-64 and American Journal of Public Health 1992;82:1495-99).
Return to Manual
Summary
Public health surveillance represents an ongoing and regular
collection, analysis, interpretation, and application of health data for
disease prevention and control. AIDS surveillance, like other national
surveillance efforts, depends on health care providers and the state and
local health departments and, thus, requires a balance between information
needs versus practical limitations. AIDS surveillance in the United States
represents an unprecedented public health enterprise and has achieved an
unusually high degree of completeness. In addition, surveillance has changed
as understanding of AIDS and HIV infection have grown. Users of the public
information data set should be familiar with the characteristics of public
health surveillance in general as well as with the evolution of AIDS
surveillance.
Return to Manual
|
