Targeting the underserved for breast and cervical cancer screening: the utility of ecological analysis using the National Health Interview Survey.

OBJECTIVES: This study tested the utility of ecological variables created from the National Health Interview Survey (NHIS) for strategic targeting of health services for the underserved. METHODS: Ecological variables were created using the 1989-1991 survey years of the NHIS public use data files. Segments, the NHIS secondary sampling units, permit computation of secondary sampling characteristics by percentage Black, percentage Hispanic, percentage below poverty, percentage unemployed, median education, median income, median age, and percentage residing in the United States for 5 years or less. These variables were analyzed with the NHIS Health Promotion and Disease Prevention 1990 supplement reporting mammogram, clinical breast examination, and Pap test use. RESULTS: Median education of areas was inversely related to never having mammograms. Areas with a high proportion (70%-100%) of Hispanic respondents also were more likely not to have mammograms. Women residing in areas with moderate or high proportions of Hispanic respondents were more likely never to have clinical breast examinations and Pap tests, as were those in areas with low income, poverty, and respondents who had resided in the United States 5 years or less. CONCLUSIONS: The new methodology of constructing ecological variables using the NHIS demonstrates an application that may help identify underserved areas or areas with underutilized services. More studies using this methodology are warranted.

Improving question wording in surveys of culturally diverse populations.

PURPOSE: The purpose of this paper is to briefly describe a theoretical model articulating cognitive theory and sources of potential response bias resulting from racial or ethnic cultural experience to survey questions that deal with health behavior. The theory components are then evaluated using questions obtained from national health surveys conducted by the National Center for Health Statistics and Centers for Disease Control and Prevention. The analysis explores the effects of four cognitive tasks involved in responding to questions as specified by the model: question interpretation, information retrieval from memory, judgment formation, and response editing. Implications for epidemiological research are considered. METHODS: Data were collected from a purposive sample of 423 adults aged 18 through 50 who were recruited to ensure equal numbers of African American, Puerto Rican, Mexican American, and non-Hispanic white respondents, stratified by age, gender, and education. Individual questions were selected for evaluation to ensure variation by topic and question format. Probes related to each of the cognitive tasks were designed to obtain insight into the underlying cognitive processes used by respondents to answer survey questions. All statistical analyses used logistic regression or ordinary least squares multiple regression as appropriate. RESULTS: Variation by race/ethnicity was found in the way respondents defined physical activity in a series of questions used in the Centers for Disease Control and Prevention Behavioral Risk Factor Surveillance System (BRFSS). Gender and race/ethnicity appeared to influence interpretation in the absence of specific cues in the question format about how to respond. Strategies used to retrieve information from memory did not appear to be influenced by respondent culture; however, frequency of the event was associated with the recall strategy in that more frequent or regular events were more likely to result in estimates about frequency, whereas unusual or seldom occurring events were counted. Effects of race/ethnicity on judgment formation seem to be reflected in the propensity of respondents' willingness to use extreme response categories. Most effects due to race/ethnicity were found in respondent editing of answers. Race/ethnicity was found to be associated with a social desirability trait; with willingness to disclose socially undesirable behavior, particularly to interviews from racial or ethnic groups that differed from the respondent; and with the tendency to overreport socially desirable behavior. CONCLUSIONS: Overall, the results of this research suggest several ways in which the validity of questions about risk behavior can be improved. In designing such questions, the investigator should envision the interview as a structured conversation in which ordinary conversational norms apply. Thus, questions that might request redundant information or that are threatening to the respondent need to be asked in ways that minimize these effects. Using interviewers of the same racial or ethnic group is important. Attending to the order of questions to ensure that redundant information is not requested is important. Writing questions to ensure that where response cues occur they lead the respondent to answer in unbiased ways is also important. Testing questions for potential racial or ethnic bias before using them is also important, even if the questions have been used successfully with population groups other than that or those included in a study.

Who participates in cardiovascular disease risk factor screenings? Experience with a religious organization-based program.

Adult members who declined participation in cardiovascular disease risk factor screenings offered at religious organizations were randomly selected and asked to participate in screenings at their homes. Relationships between screening participation and sociodemographic, behavioral, and physiological measures were examined. Age, knowledge of cardiovascular disease risk factors, body mass index, current smoking status, previous report of elevated blood pressure, current diastolic blood pressure measurement, frequency of worship service attendance, and residential distance from the religious organization screening site were important predictors of screening response. Those with conspicuous risk factors appeared less likely to initially respond to religious organization site screening invitations.

Who in America is trying to lose weight?

OBJECTIVES: To describe the characteristics and weight-loss methods of persons who are trying to lose weight; to compare the knowledges and practices regarding weight loss between those persons trying to lose weight and those not trying to lose weight; and to evaluate trends in these knowledges and practices between 1985 and 1990. DESIGN: Large (approximately 120,000 persons per year), nationally representative random sample of the U.S. population. SETTING: The 1985 and 1990 Health Promotion Disease Prevention Current Health Topics Supplements to the National Health Interview Survey. PARTICIPANTS: Random sample of the U.S. population. MEASUREMENTS: Proportions of the NHIS sample with characteristics of interest weighted to be representative of the U.S. population. RESULTS: Approximately 44 million persons 25 years or older were trying to lose weight in 1990; 61.8% of men and 59.6% of women were doing so by increasing their physical activity. Both proportions are significantly increased compared to the 1985 proportions of 56.9% and 56.2%, respectively. Twenty-seven percent of those persons who saw themselves as overweight were not trying to lose weight. CONCLUSIONS: More than one third of Americans see themselves as overweight, but fewer than two thirds of these persons are trying to lose weight. About 4% of self-perceived underweight persons and 11.4% of persons who think their weight was about right are also trying to lose weight. Most persons who are trying to lose weight are doing so by eating less, by increasing their physical activity, or by a combination of these methods.

Does cancer survival differ for older patients?

The relation of age to 5-year relative survival rates was examined for leading sites of cancer resulting in death among 127,554 patients; data from 1978 to 1982 were studied for four areas of the Surveillance, Epidemiology and End Results program of the National Cancer Institute. Overall and stage-stratified relative survival rates declined with advancing patient age for cancer of the lung, prostate, pancreas, bladder, oral cavity, uterus, cervix, ovary, and large bowel (women only). In men, this trend was not explained by age differences in stage of diagnosis, whereas, among women, age was associated with more advanced disease for most sites examined. Although overall survival rates were lower in black patients compared with white patients, the age-survival and age-stage trends were similar in the two racial groups.

Stage at diagnosis in breast cancer: race and socioeconomic factors.

Cancer incidence data from three US metropolitan areas were coupled with census tract indicators of education and income. The data suggest that both Black and White cancer patients living in census tracts with lower median education/income values are diagnosed in later disease stages than are patients in tracts with higher median education/income values. Within education and income strata, Black women had a less favorable stage of disease at diagnosis than Whites. The exception was in upper education/income levels, where the disadvantage for Blacks disappeared. These data provide additional evidence that women of low socioeconomic status could benefit from targeted screening.

Representativeness of the surveillance, epidemiology, and end results program data: recent trends in cancer mortality rates.

BACKGROUND: Mortality, incidence, and survival rates are the primary measures used by the National Cancer Institute (NCI) to monitor cancer in the United States. The Surveillance, Epidemiology, and End Results (SEER) data system collects data on all cancers diagnosed among residents in geographically defined populations, which comprise about 10% of the U.S. population. This data system is the major component of the NCI system for tracking these rates. Thus, it is important to assess the degree to which SEER data are representative of the entire U.S. population. PURPOSE: National data on mortality, but not on incidence or survival, are available from the National Center for Health Statistics. These data provide a census against which mortality data from the subset of the SEER regions may be compared. METHODS: Multivariate regression analyses of age-adjusted mortality rates from 1975 to 1988, computed for the SEER areas and for the entire United States, were performed for race- and sex-specific data from 15 cancer sites. Representativeness was evaluated by testing for differences in trends and levels between the data from the U.S. population and those from the SEER Program. RESULTS: Data from the SEER regions reflected the correct direction of trend for all sites, although some race-, sex-, and site-specific differences existed for the magnitude of the trends and levels of mortality when compared with data from the U.S. population. CONCLUSIONS: The demonstration that data from the SEER population do occasionally yield mortality rates that differ from those for the entire U.S. population suggests that data from the SEER coverage population are, in some cases, not representative of the greater U.S. population. IMPLICATIONS: This issue is of particular relevance to the interpretation of incidence measures, computed from the SEER data, for which there is no national database. Future efforts should be directed at a better understanding of how the SEER population differs from the U.S. population so that SEER rates can be adjusted to be more nationally representative.

Projections of the incidence of non-Hodgkin's lymphoma related to acquired immunodeficiency syndrome.

Advances in antiretroviral therapy and treatment or prophylaxis against opportunistic infection have resulted in prolongation of the survival of patients with acquired immunodeficiency syndrome (AIDS). Previous research has demonstrated an association between AIDS and risk of non-Hodgkin's lymphoma (NHL). In addition to the approximately 3% of individuals found to have NHL at the time of AIDS onset, others continue to develop NHL following AIDS diagnosis. Data from the Surveillance, Epidemiology, and End Results (SEER) Program of the National Cancer Institute demonstrated a sharply increasing incidence of NHL among men in the age range 20-49 years since 1983 in the United States. Based on new data on the risk of NHL following AIDS diagnosis, on estimates of improved survival following AIDS diagnosis, and on projections of future AIDS incidence, we considered four sets of assumptions and estimated the number of AIDS-related NHL cases in 1992 to be between 2900 and 9800. Three of these projections were higher than the estimate of 4700 cases obtained by linear extrapolation of SEER incidence trends. These projections of AIDS-related NHL incidence suggest that between 8% and 27% of all NHL cases that occur in the United States in 1992 will arise as a consequence of infection with the human immunodeficiency virus (HIV), imposing a substantial health care burden. More research into the pathogenesis of lymphoma and new approaches to antiretroviral and antilymphoma therapy will be necessary to prevent and treat this formidable complication of infection with HIV.

Socioeconomic factors and cancer incidence among blacks and whites.

Findings from previous studies suggest that differences in socioeconomic status may be responsible for some, if not all, of the elevated incidence of cancer among blacks as compared with whites. Using incidence data from the National Cancer Institute's Surveillance, Epidemiology, and End Results (SEER) Program, we tested this hypothesis by correlating black and white cancer incidence rates in three US metropolitan areas between 1978 and 1982 with data from the 1980 census on socioeconomic status within individual census tracts. The study analyzed data on the incidence of cancer at all sites combined (greater than 100 cancer sites) and at seven major sites separately. As in other studies, income and educational levels served as surrogates for socioeconomic status. The present study also used census-tract data on population density as a surrogate factor. Each of these measures of socioeconomic status was analyzed independently. Before correlation with census-tract data, age-adjusted data on cancer incidence showed statistically significant elevated risks among blacks for cancer at all sites combined and at four of the seven separate sites; whites showed an elevated risk for cancer at two sites. Cancer at only one site, the colon, showed no significant association with race. When age-adjusted incidence data were correlated with socioeconomic status, the comparative black-white risks changed: Whites showed an elevated risk of cancer at all sites combined and at three of the seven separate sites; blacks maintained their elevated risk at three sites. These findings suggest that the disproportionate distribution of blacks at lower socioeconomic levels accounts for much of the excess cancer burden among blacks. They also suggest that for both blacks and whites unidentified racial factors, which may be either cultural or genetic and which are not closely linked to socioeconomic status, may play a role in the incidence of some cancers.

Increasing incidence of cancers associated with the human immunodeficiency virus epidemic.

We examined data from San Francisco and other areas participating in the Surveillance, Epidemiology, and End Results (SEER) Program to determine the effect of the human immunodeficiency virus (HIV) epidemic on cancer incidence between 1973 and 1987. In this period, non-Hodgkin's lymphoma incidence has increased over 10-fold and Kaposi's sarcoma incidence has increased over 5000-fold in single San Francisco men 20 to 49 years of age. Increases in non-Hodgkin's lymphoma have been restricted to high-grade and diffuse large-cell (intermediate-grade) histological types. With the exceptions of non-Hodgkin's lymphoma and Kaposi's sarcoma, no other tumor has significantly increased in incidence. During 1987, we estimate that HIV-seropositive men in San Francisco had a 0.47% risk of developing non-Hodgkin's lymphoma and a 1.6% risk of developing Kaposi's sarcoma. The relative risks for non-Hodgkin's lymphoma and Kaposi's sarcoma associated with HIV infection were 104 and 40,000, respectively. For 1987, HIV was associated with 14% of all reported cancers (except non-melanoma skin cancer) in men aged 20 to 49. We expect that 1,890 to 2,730 excess cases of non-Hodgkin's lymphoma and 6,490 to 8,320 excess cases of Kaposi's sarcoma will occur in the United States in 1990.

Current status of prostate cancer in North American black males.

The National Cancer Institute's Surveillance, Epidemiology, and End Results (SEER) program is used to examine the most recent data available to draw inferences about black and white males in the United States with prostate cancer. Findings include a continuing rise in the incidence of prostate cancer which, as of 1985 SEER data, is 50% higher in the black male population than in white males. With the exception of minor fluctuations over the last 17 years, the mortality rate for black males demonstrates an upward trend. Combining all stages and ages, the survival rate for black males is 10% poorer than for white males. These data provide a glimpse into the problem of prostatic carcinoma in the United States today. To develop preventive strategies and cancer control interventions, a fuller understanding of the nature of the disease and its biologic course is necessary. Epidemiologic questions concerning socioeconomic status among and within racial groups, lifestyles, and behaviors that affect health seeking and diagnosis and treatment of prostatic cancer must be answered. By examining SEER data for prostatic cancer, we update the current status of this disease in North American blacks and infer possible directions for future epidemiologic surveys and cancer control intervention research.

Rise in prostatic cancer incidence associated with increased use of transurethral resection.

We examined the association between prostatic cancer incidence rates and the rates of transurethral prostatectomy to explore reasons for the nationally reported dramatic increases in incidence rates of prostatic cancer from 1973 through 1986. There was a strong correlation between both incidence of all stages of prostatic cancer combined and of localized disease and the increasing use of transurethral resection, a common surgical procedure usually performed to relieve urinary obstruction due to benign enlargement of the prostate. Our analyses suggest that increased detection of existing tumors via transurethral resection was the primary reason for the observed increase in incidence rates of prostatic cancer. However, analyses of mortality trends, particularly among nonwhites, and laboratory studies of the histologic nature of clinically asymptomatic tumors suggest that part of the increase may reflect changes in the real risk of prostatic cancer.

The incidence of retinoblastoma in the United States: 1974 through 1985.

We have estimated the incidence of retinoblastoma in the United States from data available form the Surveillance, Epidemiology, and End Results program of the National Cancer Institute, Bethesda, Md (1974 through 1985). The larger number of cases available (220) compared with those of previous US population-based studies enabled us to evaluate change in incidence over time and to more precisely estimate incidence according to various demographic characteristics. The incidence of retinoblastoma has been almost uniform form 1974 to 1985. The average annual incidence of retinoblastoma was 5.8 per million for children younger than 10 years and 10.9 per million for children younger than 5 years. There was no difference in the incidence of retinoblastoma by either sex or race. The overall 5-year cumulative survival rate was 91% (95% confidence interval, 87% to 95%). The data indicate a worsening survival with increasing age at diagnosis, through age 2 years, but a less clear relationship of survival with diagnosis beyond age 2 years.

Estimating cancer mortality rates from SEER incidence and survival data.

A method to estimate site-specific cancer mortality rates using Surveillance, Epidemiology, and End Results (SEER) Program incidence and survival data is proposed, calculated, and validated. This measure, the life table-derived mortality rate (LTM), is the sum of the product of the probability of being alive at the beginning of an interval times the probability of dying of the cancer of interest during the interval times the annual age-adjusted incidence rate for each year that data have been collected. When the LTM is compared to death certificate mortality rates (DCM) for organ sites with no known misclassification problems, the LTM was within 10 percent of the death certificate rates for 13 of 14 organ sites. In the sites that have problems with the death certificate rates, there were major disagreements between the LTM and DCM. The LTM was systematically lower than the DCM for sites if there was overreporting on the death certificates, and the LTM was higher than the DCM for sites if there was underreporting. The limitations and applications of the LTM are detailed.

Person-years of life lost due to cancer in the United States, 1970 and 1984.

The number of deaths due to cancer in the United States reached an all-time high of 453,450 deaths in 1984 and, due to the dynamics of population growth, will continue to increase if the risk of dying from cancer does not change. Between 1970 and 1984, the total Person-Years of Life Lost (PYLL), the sum of the difference between the actual age at death and the expected remaining lifetime for each person who died of cancer, increased for most cancer sites as well as for all sites combined. In 1984, 6,881,281 person-years of life were lost due to cancer deaths, up from 5,303,668 in 1970. The exceptions are those cancers for which there has been major progress in either prevention or treatment; e.g., stomach and cervix uteri (prevention) and testicular, Hodgkin's disease, leukemia, and childhood cancers (treatment). The Average Years of Life Lost (AYLL) per person dying from cancer in 1984 was generally less than in 1970. Overall, each person who died from cancer in 1984 died 15.2 years earlier than his/her life expectancy. The greatest loss was for those who died of childhood cancers (66.9 years earlier), followed by testicular cancer (35.8 years earlier). The least loss relative to the expectation of life was for those who died of prostate cancer. The 25,400 men who died from prostate cancer in 1984 died an average of nine years earlier than otherwise expected.

Contribution of socioeconomic status to black/white differences in cancer incidence.

Blacks and Whites have very different cancer incidence rates for many sites, but this may largely be due to the racial differences in socioeconomic status (SES). The authors tested this hypothesis by determining the effect of adjustment for SES on the black/white incidence ratios for 12 cancer sites. Race-specific census tract-level SES variables (median family income, percent below poverty level, and years of education) were obtained from the 1980 US census and applied to approximately 20,000 black and 88,000 white cancer cases from the National Cancer Institute's Surveillance, Epidemiology, and End Results (SEER) program for the years 1978 to 1982. For each cancer site (with each sex considered separately), Poisson regression was used to produce age-adjusted black/white incidence ratios, with and without adjustment for SES. The SES variable with the strongest adjusting power was percent below poverty level. For many sites (breast, in situ and invasive cervix, esophagus, male lung, pancreas, stomach) poverty accounted for much or all of the racial differences. For several sites (bladder, multiple myeloma, prostate, uterine corpus), large racial differences persisted after adjustment for poverty, and these findings suggest directions for investigating the etiology of these cancers.

Population-based estimate of survival and determinants of prognosis in patients with mycosis fungoides.

Mycosis fungoides is the most common cutaneous lymphoma in the US, and it is increasing rapidly in both incidence and mortality. Our knowledge of its prognosis derives primarily from case series, which are subject to possible selection bias and other limitations. The current analysis examines trends in survival and prognostic factors in data from nine population-based cancer registries in the US. Among the 650 cases followed with known dates of diagnosis and no history of prior malignancy, the median survival time was 7.8 years. Advanced age, black race, prior malignancy, and Sezary syndrome presence at the time of diagnosis were each independently associated with poor prognosis. Fatality was not influenced by sex or geographic area. There was no consistent evidence of improved survival with time over the 12-year study period.

Mycosis fungoides in the United States. Increasing incidence and descriptive epidemiology.

The etiology of mycosis fungoides is obscure, and the risk factors for its occurrence are poorly documented. This investigation uses data from nine US population-based cancer registries to investigate the descriptive epidemiology of this disorder. From 1973 through 1984, 721 newly diagnosed cases of mycosis fungoides were reported to these registries (0.29 cases per 100,000 population per year). A dramatic increase in the incidence of mycosis fungoides was noted over the period of this study. The incidence was highest among the elderly. Blacks were twice as likely to be afflicted as whites, and the incidence among men was more than twice the incidence among women. The geographic variation in incidence was associated with several demographic variables, including population density, family income, and concentration of physicians. Analysis of mortality among these patients revealed no evidence of detection bias.