Updated guidelines for evaluating public health surveillance systems: recommendations from the Guidelines Working Group.

The purpose of evaluating public health surveillance systems is to ensure that problems of public health importance are being monitored efficiently and effectively. CDC's Guidelines for Evaluating Surveillance Systems are being updated to address the need for a) the integration of surveillance and health information systems, b) the establishment of data standards, c) the electronic exchange of health data, and d) changes in the objectives of public health surveillance to facilitate the response of public health to emerging health threats (e.g., new diseases). This report provides updated guidelines for evaluating surveillance systems based on CDC's Framework for Program Evaluation in Public Health, research and discussion of concerns related to public health surveillance systems, and comments received from the public health community. The guidelines in this report describe many tasks and related activities that can be applied to public health surveillance systems.

The epidemiology of diabetes and pregnancy in the U.S., 1988.

OBJECTIVE: To determine the prevalence of pregnancy complicated by diabetes in a representative sample of the U.S. population. RESEARCH DESIGN AND METHODS: We analyzed data from a multistaged cross-sectional probability sample of live births recorded in the U.S. in 1988 for women 15-49 years of age. The main outcome measure was pregnancy complicated by diabetes. RESULTS: Diabetes was present in congruent to 154,000 (4%) of all pregnancies in the U.S. Gestational diabetes mellitus (GDM) accounted for 135,000 of such pregnancies (88%), non-insulin-dependent diabetes mellitus (NIDDM) for 12,000 (8%), and insulin-dependent diabetes mellitus for 7,000 (4%). On average, the mothers with NIDDM (29.6 years) and GDM (29.3 years) were older than mothers whose pregnancies were not complicated by diabetes (26.2 years; P < 0.05). In multivariate analyses, the odds of having a pregnancy complicated by GDM increased significantly with maternal age and body mass index. CONCLUSIONS: Pregnancy is complicated by diabetes more often than was previously believed. More frequent testing may further increase the apparent prevalence of GDM.

Diabetes-associated mortality in Native Americans.

OBJECTIVE: To describe diabetes-associated mortality among Native Americans. RESEARCH DESIGN AND METHODS: In this population-based study, we analyzed diabetes-associated mortality data from the IHS and the NCHS. We also examined diabetes data from the 1986 NMFS. RESULTS: IHS area-specific diabetes mortality rates for 1984-1986 ranged from 10 to 93/100,000, compared with 15/100,000 for the total U.S. population. NCHS data for the same period listed diabetes as the underlying cause of 708 deaths among Native Americans and the contributory cause of 1252 deaths; 63% of the latter deaths were attributable to circulatory diseases. The 1986 NMFS demonstrated that Native American heritage is underreported by 65% on death certificates. Using deaths identified as Native American by NMFS, the age-adjusted mortality rate for diabetes as the underlying cause for Native Americans (96/100,000) was 4.3 times that for whites and two times that for blacks. Where diabetes was a contributory cause of death, the mortality rate for Native Americans (264/100,000) was 3.7 times that for whites and 2.4 times that for blacks. CONCLUSIONS: The excessive diabetes-associated mortality among Native Americans is consistent with other indicators of the magnitude of the diabetes problem in this population. Further epidemiological research and expanded diabetes control interventions are needed.

Trends in diabetes and diabetic complications, 1980-1987.

OBJECTIVE: Although diabetes is a major source of morbidity and mortality in the United States, only recently has a unified national surveillance system begun to monitor trends in diabetes and diabetic complications. RESEARCH DESIGN AND METHODS: We established a diabetes surveillance system using data for 1980-1987 from vital records, the National Health Interview Survey, the National Hospital Discharge Survey, and the Health Care Financing Administration's records to examine trends in the prevalence and incidence of diabetes, diabetes mortality, hospitalizations, and diabetic complications. RESULTS: From 1980 through 1987, the number of individuals known to have diabetes increased by 1 million--to 6.82 million. Age-standardized prevalence for diabetes increased 9% during this period, from 25.4 to 27.6/1000 U.S. residents (P = 0.03). The incidence of diabetes increased among women (P = 0.003), particularly among those greater than 65 yr old (P = 0.02). Age-standardized mortality rates (for diabetes as either an underlying or contributing cause) per 100,000 individuals with diabetes declined 12%, from 2350 to 2066. Annual mortality rates from stroke (as an underlying cause and diabetes as a contributing cause) and diabetic ketoacidosis declined 29% (P = 0.003) and 22% (P less than 0.001), respectively. During these 8 yr, hospitalization rates for major CVD and stroke (as the primary diagnoses and diabetes as a secondary diagnosis) increased 34% (P = 0.006) and 38% (P = 0.01), respectively. Also during this period, hospitalization rates increased 21% for diabetic ketoacidosis (P = 0.01) and 29% for lower-extremity amputations (P = 0.06). From 1982 through 1986, treatment for end-stage renal disease related to diabetes increased greater than 10% each year (P less than 0.001). The prevalence of diagnosed diabetes was nearly twice as high in blacks as in whites (P = 0.04). Blacks also had increased rates of lower-extremity amputation (P = 0.02), diabetic ketoacidosis (P less than 0.001), and end-stage renal disease (P = 0.01). CONCLUSIONS: Diabetes surveillance data will be useful in planning, targeting, and evaluating public health efforts designed to prevent and control diabetes and its complications.

Is race related to glycemic control? An assessment of glycosylated hemoglobin in two South Carolina communities.

To consider the relationship between race and long-term glycemic control, as measured by glycosylated hemoglobin (GHb), we analyzed data from a community-based sample of 3175 adults in the South Carolina Cardiovascular Disease Prevention Project. A clinically meaningful difference for mean GHb levels (10.5 vs 8.4%, P < 0.001) was present between black people and white people reporting diabetes. Similarly, a significant association between race and GHb was present among people reporting "borderline diabetes" or no diabetes. Logistic regression confirmed this finding in all three diabetic categories, however, controlling for insulin use in the diabetic group reduced (P < 0.001) the association between GHb and race. These findings confirm that further improvements in glycemic control are necessary, especially for black patients and that black people not reporting diabetes have higher GHb levels compared to white people, possibly due to undiagnosed diabetes.

The distribution of West Virginia women targeted for screening of breast and cervical cancers.

The primary strategy for the control of female breast and cervical cancers is early diagnosis and treatment from periodic screenings. To effectively provide screenings, it is important to target the populations that especially need these services. Based on state health objectives and an analysis of data from the 1990 Census, about 33% of the women in West Virginia are targeted for routine screening for breast cancer by clinical breast examination and mammograms, while about 77% are targeted for routine screening for cervical cancer by the Papanicolaou test. Since the racial and age distributions of women vary across Public Health Management Districts in the state, it is critical to determine the exact extent of breast and cervical cancers in West Virginia for geographic and other populations of women. In addition, the results demonstrate a need for age-adjustment when comparing measures of the cancers across geographic areas.

The West Virginia Cancer Registry: historical development and preliminary results.

This article is designed to provide the medical community with an overview of the West Virginia Cancer Registry (WVCR), a population-based registry, which was established in 1993 with both federal and state funding. The WVCR is managed through the Bureau for Public Health's Division of Surveillance and Disease Control in the Office of Epidemiology and Health Promotion, and all data is confidential. The staff of the WVCR consists of a director, a cancer data supervisor, two abstractors, a secretary, and a voluntary, 13-member Cancer Advisory Committee. The initial findings of the WVCR have shown that for each 100,000 West Virginia women, 90.1 were discovered with invasive breast cancer in 1992, an increase from 86.8 in 1991. The invasive cervical cancer rate increased from 10.8 to 12.9 during this same time period. In addition, the age-adjusted death rate for cervical cancer rose from 3.6 cases per 100,000 women in 1991 to 3.8 in 1992. However, the age-adjusted death rate for breast cancer fell to 24.7 per 100,000 women in 1992, compared to 25.5 in 1991.

Long-term trends in cancer mortality rates for West Virginia.

Using data on death certificates, we examined age-adjusted cancer mortality rates for West Virginia men and women from 1980-1994. The leading causes of cancer deaths among men in the state in 1994 were cancers of the lung, prostate, colon, and pancreas, and non-Hodgkin's lymphoma. Among West Virginia women in 1994, cancer deaths were most often due to cancers of the lung, breast, colon, pancreas, and ovary. Based on polynomial regression analyses, many of the leading cancer mortality rates significantly increased during 1980-1994. The most striking increase was a 76% rise in lung cancer mortality among West Virginia women. These mortality data underscore the continuing need for tobacco control and other cancer control practices. Even though barriers to medical care challenge the state's health care professionals, the burden of cancer in West Virginia can be reduced by prevention, early detection, and appropriate treatment.

Sensitivity and predictive value positive measurements for public health surveillance systems.

Two important measurements for the evaluation of a public health surveillance system are sensitivity and predictive value positive (PVP). The computation of sensitivity and PVP for a public health surveillance system, however, can be complicated by the absence of an appropriate gold standard. In addition, there are few references for the computation of sensitivity and PVP for a surveillance system. To determine how these attributes of evaluation have been reported in epidemiologic literature, I review papers that report sensitivity and PVP for public health surveillance systems. Of the 31 papers that met selection criteria, 21 (68%) included either a reference for the computation or a definition of the attributes, whereas 18 (58%) reported both attributes. All 31 papers reported sensitivity, and among the 31 papers, 24 (77%) reported more than one sensitivity measurement. Among the 18 papers that reported at least PVP, 13 (72%) reported more than one PVP measurement. This review provides guidance in computing sensitivity and PVP for a public health surveillance system.

Lessons learned from the first funding period of the CDC Assessment Initiative.

The Centers for Disease Control and Prevention Assessment Initiative enables state and local health departments to improve their capacity to use data for developing policy and ensuring that needed health services are provided. Using reports and comments submitted by staff from the participating states, lessons learned are summarized for the first funding period of the initiative. Although many activities were common among the participating states, the lessons learned varied for them. From a commitment to the concepts of assessment, policy development, and assurance, a vision for the national mission of public health is provided and the various roles in fulfilling the mission are defined.

Patient adherence to guidelines for diabetes eye care: results from the diabetic eye disease follow-up study.

Early detection and treatment of diabetic eye disease can prevent blindness, yet many persons with diabetes lack regular eye care. This study followed 569 people with diabetes participating in blindness prevention programs during 1985 through 1987; it was found that 35% of subjects received dilated eye examinations before entering the programs, in comparison with 60% afterward. About 85% of participants referred for proliferative retinopathy treatment began such treatment, and, of these, 85% completed treatment. A lack of knowledge about the disease and limited finances were primary reasons for nonadherence. To improve the effectiveness of prevention programs, eye care providers and program staff must strive to eliminate these educational and financial barriers.

Surveillance for diabetes mellitus--United States, 1980-1989.

PROBLEM/CONDITION: In the United States, diabetes mellitus is the most important cause of lower-extremity amputation and end-stage renal disease; the major cause of blindness among working-age adults; a major cause of disability, premature mortality, congenital malformations, perinatal mortality, and health-care costs; and an important risk factor for the development of many other acute and chronic conditions (e.g., diabetic ketoacidosis, ischemic heart disease, stroke). Surveillance data describing diabetes and its complications are critical to increasing recognition of the public health burden of diabetes, formulating health-care policy, identifying high-risk groups, developing strategies to reduce the burden of this disease, and evaluating progress in disease prevention and control. REPORTING PERIOD COVERED: In this report, data are summarized from CDC's diabetes surveillance system; trends in diabetes and its complications are evaluated by age, sex, and race for the years 1980-1989. DESCRIPTION OF SYSTEM: CDC has established an ongoing and evolving surveillance system to analyze and compile periodic, representative data on the disease burden of diabetes and its complications in the United States. Data sources currently include vital statistics, the National Health Interview Survey, the National Hospital Discharge Survey, and Medicare claims data for end-stage renal disease. RESULTS AND INTERPRETATION: In 1989, approximately 6.7 million persons in the United States reported that they had diabetes mellitus, and a similar number probably had this disabling chronic disease without being aware of it. The disease burden of diabetes and its complications is large and is likely to increase as the population grows older. Effective primary, secondary, and tertiary prevention strategies are needed, and these efforts need to be intensified among groups at highest risk, including blacks. Important gaps exist in periodic and representative data for describing the disease burden. ACTIONS TAKEN: CDC is assisting diabetes control programs in 26 states and one territory. These programs attempt to reduce the burden of diabetes by preventing blindness, lower-extremity amputations, cardiovascular disease, and adverse outcomes of pregnancy among persons with diabetes. Because of important limitations in measuring the burden of diabetes, CDC is exploring sources of surveillance data for blindness, adverse outcomes of pregnancy, and the public health burden of diabetes among minority groups.