Reporting of Human Immunodeficiency Virus

¶ … Reporting of Human Immunodeficiency Virus Diagnoses

Hall, Irene H., Ruiguang Song, John E. Gerstle III, & Lisa M. Lee (2006). Assessing the completeness of reporting of human immunodeficiency virus diagnoses in 2002 -- 2003: Capture-recapture methods. 164 (4): 391. Retrieved June 17, 2010 at http://aje.oxfordjournals.org/cgi/reprint/164/4/391

What is the research problem being investigated?

One of the most seismic epidemiologically-related disasters of recent memory is the spread of acquired immunodeficiency syndrome (AIDS). AIDS/HIV has had an impact on almost every aspect of modern life in America and around the globe. However, even today, the extent of the epidemic is hotly debated amongst epidemiologists

According to Hall (et al., 2006), the relative completeness or lack of completeness of the reporting of the epidemic in the U.S. remains in dispute. For example, some areas indicated a near 98% completeness rate of reporting, while others only 61% (Hall et al. 2006, p.391). The central purpose of Hall's study is the question of how to assess the completeness rate in a more accurate and consistent fashion.

What is the rationale for studying this problem?

The stated rationale is that a significant pattern of over- or conversely under-reporting of AIDS and HIV-positive status suggests that a different method of assessing the epidemic nation-wide is required. By implication as well, comprehensiveness or a lack thereof when reporting cases sheds light the need for more funding for AIDS treatment, tracking, and prevention.

Earlier studies were deemed to be of limited value because these studies assessed completeness of reporting of HIV or AIDS diagnoses for only a few geographic areas and for isolated time periods. This study strives to be ground-breaking in its approach.

How comprehensive is the literature review?

Given their relatively dim view of the effectiveness of previous studies to provide a portrait of the epidemic, the researchers instead elect to focus upon the impact of their chosen method, a capture-recapture method of assessing accuracy of reporting. "Capture-recapture studies," which originated in wildlife biology research, "are used to estimate the size of a population when a census may be infeasible or impossible to conduct. The basic idea of capture-recapture studies is to sample and identify individuals, or cases, from a population and then resample the population to see what fraction of individuals, or cases, in the second sample were identified in the first sample (i.e. The fraction of individuals, or cases, that were found in both samples)" (Notes, 2006, VALID).

Epidemics which may have been underreported due to demographic characteristics of the population, presumably due to factors such as poverty or drug use, or cultural factors such as shame may be ideal for this model. The state-by-state method of reporting cases, which creates a patchwork of diagnostic and recording methods, makes double-verification more reasonable in terms of its accuracy.

What is/are the research question(s) being addressed? What is/are the research hypothesis (es)?

Because assessing the AIDS epidemic remains such a complex challenge, the authors advocate a completely and radically different epidemiological paradigm of data-collection than has been used in the past. The traditional (or case-based) surveillance system, "when more than one report is received for an individual, for example, from a medical care provider and from a laboratory, the information from these reports is consolidated, and one record with all pertinent information is entered into the surveillance system for that person" is rejected by the authors in favor of using data from multiple sources to ensure greater accuracy (Hall et al. 2006, p.392).

Using a 'double-barreled' or capture and recapture approach is a more effective means of assessing the impact of the AIDS epidemic in America, given the challenges of incomplete rates of diagnosis and inconsistent reporting in many areas. Different data sets within the same area, but from different sources, may highlight systemic problems with tracking the epidemic.

What are the implications of the research? Are the above items clearly explained or implied?

As defined explicitly in the article, the implications of the research are not fully explored. The article derives from an epidemiology journal, so the focus is upon accuracy of reporting rather than upon what benefits might be derived in terms of setting public policy. As the article is from a peer-reviewed journal, there is also the presumption that the reader is familiar with what constitutes a 'case-based' versus 'capture and recapture' approach when these methods are explained. There is no significant regarding the over-reporting or underreporting of HIV / AIDS cases due to cultural factors or the psychology of underreporting.

The article thus presents itself more as a launching ground for future public policy initiatives, rather than attempts to prescribe an answer of how to better combat the epidemic. Through a more accurate portrait of HIV / AIDS cases in America, it is hoped that knowledge will foster more effective treatment efforts.

Design and methods: Who is the study sample population? Describe in detail.

The surveillance programs of seven states and several major urban areas, including New York, participated in collecting information on persons with a diagnosis of HIV. "Separate reports from each database were matched to records for HIV-infected persons reported to CDC's HIV / AIDS Reporting System" and completeness of reporting of HIV / AIDS diagnoses was calculated "by comparing the number of diagnoses reported to the surveillance program for a diagnosis year with the number of diagnoses expected to have been made during that year" (Hall et al. 2006, p.392).

Can this study sample be generalized? How is it generalized?

The study samples derive from a wide variety of sources, including states and specific demographic populations with a high rate of diversity. The findings are generalized to the AIDS epidemic as a whole in the United States, and are not said to be specific to urban or rural populations, or specific populations known to be affected disproportionately by AIDS, such as gay men or HIV drug users.

The study sample is assumed to be generalized to the United States, although not to other areas, given the conclusions drawn from the findings.

How was the research sample obtained? What biases, exist in the sample if any?

Reporting sources derived from a wide array of sources and "included laboratories, hospitals, physicians' offices/health-care providers, vital statistics departments (death and birth certificates), local health departments, and HIV-testing and -counseling sites. Reporting sources were coded as inpatient; outpatient; emergency room; screening, diagnosis, or referral agencies; laboratory; other databases; other facility records; or other records. Other databases might include death certificates, Medicaid or billing records, or other disease registries such as cancer or hepatitis. Other facilities' records were from correctional facilities or coroners" (Hall et al. 2006, p.392).

Overall, comprehensiveness of data and breadth of demographic and geographic characteristics was used to reduce bias. After obtaining these records, "separate reports from each database were matched to records for HIV-infected persons reported to CDC's HIV / AIDS Reporting System by use of the state-assigned unique case identification number," and the two sources were compared (Hall et al. 2006, p.392).

What is the research design of the study?

The research design was purely quantitative in nature: data was accessed without resort to any specific, anecdotal case studies or field research. The time in which the data collected was discrete, although various sources were compared over specific, pre-determined times, as part of the article's study design (Hall et al. 2006, p.392).

6 months of reporting was compared with 12 months of follow-up, to see if demographic data 'matched' over a longer duration of time, between the different sources used for information and cross-comparison.

What instruments were used? How are the reliability and validity of the instruments addressed?

Most states rely upon laboratory reporting for reporting cases of AIDS/HIV. Laboratory reports were compared with expected diagnoses, based upon reported cases obtained through the above-mentioned sources. Log-linear (correlation) models were used to assess dependencies between variables, and allow for possible underreporting -- i.e. cases that were diagnosed but not identified.

Overall completeness of cases diagnosed within a 1-year period versus 6 months found a total of 11,266 HIV diagnoses. Completeness of reporting of HIV diagnoses was 76% and "when allowing 6 months of reporting delay (range: 72 -- 95%) and improved to 81% (95% confidence interval: 72, 88) with 12 months' follow-up" (Hall et al. 2006, p.391).

What operational definitions are used for the study, if any? Would you recommend alternative study methods?

No specific diagnosis of how AIDS/HIV was diagnosed was stated, or if this might affect results.

At minimum, examining some specific cases in several of the areas where the study took place, such as New York vs. A clinic in a rural area, as well as segmenting and explicating different methods of diagnosis and tracking in the different areas under scrutiny might help pinpoint if any specific areas were more apt to over or under-diagnose or report, and for what reasons.

Results and Discussions: How are these data presented?