Annual report to the nation on the status of cancer (1973 through 1998), featuring cancers with recent increasing trends

BACKGROUND

The American Cancer Society, the National Cancer Institute (NCI), the North American Association of Central Cancer Registries, and the Centers for Disease Control and Prevention, including the National Center for Health Statistics (NCHS), collaborate to provide an annual update on cancer occurrence and trends in the United States. This year's report contains a special feature that focuses on cancers with recent increasing trends.

METHODS

From 1992 through 1998, age-adjusted rates and annual percent changes are calculated for cancer incidence and underlying cause of death with the use of NCI incidence and NCHS mortality data. Joinpoint analysis, a model of joined line segments, is used to examine long-term trends for the four most common cancers and for those cancers with recent increasing trends in incidence or mortality. Statistically significant findings are based on a P value of.05 by use of a two-sided test. State-specific incidence and death rates for 1994 through 1998 are reported for major cancers.

RESULTS

From 1992 through 1998, total cancer death rates declined in males and females, while cancer incidence rates declined only in males. Incidence rates in females increased slightly, largely because of breast cancer increases that occurred in some older age groups, possibly as a result of increased early detection. Female lung cancer mortality, a major cause of death in women, continued to increase but more slowly than in earlier years. In addition, the incidence or mortality rate increased in 10 other sites, accounting for about 13% of total cancer incidence and mortality in the United States.

CONCLUSIONS

Overall cancer incidence and death rates continued to decline in the United States. Future progress will require sustained improvements in cancer prevention, screening, and treatment.

The annual report to the nation on the status of cancer, 1973-1997, with a special section on colorectal cancer

BACKGROUND

This annual report to the nation addresses progress in cancer prevention and control in the U.S. with a special section on colorectal cancer. This report is the joint effort of the American Cancer Society, the National Cancer Institute (NCI), the North American Association of Central Cancer Registries (NAACCR), and the Centers for Disease Control and Prevention (CDC), including the National Center for Health Statistics (NCHS).

METHODS

Age-adjusted rates were based on cancer incidence data from the NCI and NAACCR and underlying cause of death as compiled by NCHS. Joinpoint analysis was based on NCI Surveillance, Epidemiology, and End Results (SEER) program incidence rates and NCHS death rates for 1973-1997. The prevalence of screening examinations for colorectal cancer was obtained from the CDC's Behavioral Risk Factor Surveillance System and the NCHS's National Health Interview Survey.

RESULTS

Between 1990-1997, overall cancer incidence and death rates declined. Joinpoint analyses of cancer incidence and death rates confirmed the declines described in earlier reports. The incidence trends for colorectal cancer have shown recent steep declines for whites in contrast to a leveling off of the rates for blacks. State-to-state variations occurred in colorectal cancer screening prevalence as well as incidence and death rates.

CONCLUSIONS

The continuing declines in overall cancer incidence and death rates are encouraging. However, a few of the top ten incidence or mortality cancer sites continued to increase or remained level. For many cancer sites, whites had lower incidence and mortality rates than blacks but higher rates than Hispanics, Asian and Pacific Islanders, and American Indians/Alaska Natives. The variations in colorectal cancer incidence and death rates by race/ethnicity, gender, age, and geographic area may be related to differences in risk factors, demographic characteristics, screening, and medical practice. New efforts currently are underway to increase awareness of screening benefits and treatment for colorectal cancer.

The annual report to the nation on the status of cancer, 1973-1997, with a special section on colorectal cancer

BACKGROUND

This annual report to the nation addresses progress in cancer prevention and control in the U.S. with a special section on colorectal cancer. This report is the joint effort of the American Cancer Society, the National Cancer Institute (NCI), the North American Association of Central Cancer Registries (NAACCR), and the Centers for Disease Control and Prevention (CDC), including the National Center for Health Statistics (NCHS).

METHODS

Age-adjusted rates were based on cancer incidence data from the NCI and NAACCR and underlying cause of death as compiled by NCHS. Joinpoint analysis was based on NCI Surveillance, Epidemiology, and End Results (SEER) program incidence rates and NCHS death rates for 1973-1997. The prevalence of screening examinations for colorectal cancer was obtained from the CDC's Behavioral Risk Factor Surveillance System and the NCHS's National Health Interview Survey.

RESULTS

Between 1990-1997, overall cancer incidence and death rates declined. Joinpoint analyses of cancer incidence and death rates confirmed the declines described in earlier reports. The incidence trends for colorectal cancer have shown recent steep declines for whites in contrast to a leveling off of the rates for blacks. State-to-state variations occurred in colorectal cancer screening prevalence as well as incidence and death rates.

CONCLUSIONS

The continuing declines in overall cancer incidence and death rates are encouraging. However, a few of the top ten incidence or mortality cancer sites continued to increase or remained level. For many cancer sites, whites had lower incidence and mortality rates than blacks but higher rates than Hispanics, Asian and Pacific Islanders, and American Indians/Alaska Natives. The variations in colorectal cancer incidence and death rates by race/ethnicity, gender, age, and geographic area may be related to differences in risk factors, demographic characteristics, screening, and medical practice. New efforts currently are underway to increase awareness of screening benefits and treatment for colorectal cancer.

The annual report to the nation on the status of cancer, 1973-1997, with a special section on colorectal cancer

BACKGROUND

This annual report to the nation addresses progress in cancer prevention and control in the U.S. with a special section on colorectal cancer. This report is the joint effort of the American Cancer Society, the National Cancer Institute (NCI), the North American Association of Central Cancer Registries (NAACCR), and the Centers for Disease Control and Prevention (CDC), including the National Center for Health Statistics (NCHS).

METHODS

Age-adjusted rates were based on cancer incidence data from the NCI and NAACCR and underlying cause of death as compiled by NCHS. Joinpoint analysis was based on NCI Surveillance, Epidemiology, and End Results (SEER) program incidence rates and NCHS death rates for 1973-1997. The prevalence of screening examinations for colorectal cancer was obtained from the CDC's Behavioral Risk Factor Surveillance System and the NCHS's National Health Interview Survey.

RESULTS

Between 1990-1997, overall cancer incidence and death rates declined. Joinpoint analyses of cancer incidence and death rates confirmed the declines described in earlier reports. The incidence trends for colorectal cancer have shown recent steep declines for whites in contrast to a leveling off of the rates for blacks. State-to-state variations occurred in colorectal cancer screening prevalence as well as incidence and death rates.

CONCLUSIONS

The continuing declines in overall cancer incidence and death rates are encouraging. However, a few of the top ten incidence or mortality cancer sites continued to increase or remained level. For many cancer sites, whites had lower incidence and mortality rates than blacks but higher rates than Hispanics, Asian and Pacific Islanders, and American Indians/Alaska Natives. The variations in colorectal cancer incidence and death rates by race/ethnicity, gender, age, and geographic area may be related to differences in risk factors, demographic characteristics, screening, and medical practice. New efforts currently are underway to increase awareness of screening benefits and treatment for colorectal cancer.

Effects of physician outreach programs on rural-urban differences in breast cancer management

In 1986 to 1987, urban-rural differences in several breast cancer management practices were noted in Illinois data. Several intervention programs for physicians were initiated to improve rural patients' access to state-of-the-art breast cancer management to reduce these differences. This report compares an intensive rural oncology outreach intervention program with a lower intensity physician education program. Medical records from 1986 to 1991 were reviewed on 2,277 breast cancer patients in a 12-county study area. The care received by urban patients was compared with three groups of rural patients: those managed in rural hospitals with intensive oncology outreach programs beginning in 1988 (Rural group 1), and in those rural hospitals with less intensive interventions using an audit with feedback strategy beginning in 1989 (Rural group 2). Rural patients who traveled to one of the urban facilities also were included in the analysis because the less intensive interventions also took place in these facilities, and these patients showed unique patterns of care in the baseline analysis (Rural Group 3). The years 1986 to 1987 constituted the baseline, and 1990 to 1991 constituted the final evaluation period. Chi square and multivariate analyses were conducted to compare the effect of the two types of interventions on changing breast cancer management practices and reducing the urban-rural differences. By the final evaluation period, the high intensity intervention was not more successful in reducing or eliminating the urban-rural differences than the low intensity intervention for many practices. However, often the frequency estimates were higher in Rural Group 1, which received the high intensity intervention. The changes noted in Rural Group 3 were not always the same as in Rural Group 2, even though both received the same low intensity interventions, lending evidence to the observation that travel distance and other nonmedical factors affect the choices of management modalities for these patients. Finally, given the nonrandomized study design, other explanations for the changes could not be ruled out.

Increased risk of malignancy in patients with systemic lupus erythematosus

BACKGROUND

Systemic lupus erythematosus is a chronic, multisystem, autoimmune disorder that primarily affects women. Morbidity and mortality have improved for lupus patients during the last 15 years. An increased risk of malignancy in patients with lupus has been shown in some, but not all studies. The purpose of this study was to ascertain cancer risk in lupus patients by linking two disease registries.

METHODS

Participants in the Chicago Lupus Cohort included 616 women with lupus who were residents of Cook County, Illinois. They were seen during 1985-1995 at 4 University, inner city, and suburban inpatient and outpatient clinics in Chicago. Malignancies occurring in these subjects during the study interval, 1985-1995, were identified from the Illinois State Cancer Registry by matching name, birthdate, and social security number. Standardized incidence ratios (SIRs) were estimated for all malignancies in this cohort of lupus patients using age, gender, and all race or race-stratified specific cancer incidence data from Cook County, Illinois.

RESULTS

The registry linkage study with the Illinois State Cancer Registry documented that 30 women with lupus had a malignancy. The expected number of malignancies for women in the lupus cohort was 15.0. There were 8 cases of breast cancer and 4 each of lung and cervical cancer. In the remaining 14 women, 12 different types of cancers were noted. The SIR and 95% confidence interval (CI) for malignancy for all women with lupus in the study were 2.0 (1.4, 2.9) and lung cancer was the only individual cancer increased in all women--SIR and 95% CI were 3.1 (1.3, 7.9). In the analysis stratified by race, the risk of malignancy (SIR and 95% CI) was increased in Caucasian women, 2.3 (1.4, 3.9). Breast cancer was the only individual cancer increased in Caucasian women with lupus with an SIR and 95% CI of 2.9 (1.4, 6.4).

CONCLUSIONS

Lupus patients have an increased risk of malignancy. Breast, lung, and gynecological malignancies were the most common malignancies observed in the cohort and breast cancer was significantly increased in Caucasian women.

Toward a broader definition of heat-related death: comparison of mortality estimates from medical examiners' classification with those from total death differentials during the July 1995 heat wave in Chicago, Illinois

The classification of heat-related deaths solely according to body temperatures underestimates the magnitude of heat-related mortality during heat waves. Broader classifications are often used, but their utility in defining the mortality and identifying at-risk populations has not been evaluated. Using death data from the July 1995 heat wave in Chicago, the authors compared heat-related mortality rates based on the classification of heat-related deaths by the Cook County Medical Examiner's Office (CCMEO), with excess mortality rates based on total mortality differentials during and before the heat wave. In July 1995, the overall mortality in Chicago was 19 deaths per 100,000 population for heat-related mortality and 24 to 26 deaths per 100,000 population for excess mortality. Across Chicago community areas, the two mortality rates were closely related (r = 0.73-0.79; p < .01), but heat-related mortality rates were lower than excess mortality rates in community areas where excess mortality rates were higher (slope < 1; p < .01), a finding indicating an underestimation of heat-related deaths in such areas. The underestimation could not be explained by uncertainties in estimating excess mortality rates or by differences in socioeconomic and demographic characteristics among communities. These results support using the broader CCMEO classification of heat-related deaths as a relative indicator to target communities for prevention and relief efforts, but not as an adequate measure of actual heat-related mortality in a high-risk neighborhood.

Cancer incidence and age at northern migration of African Americans in Illinois, 1986-1991

We compared the proportional cancer incidence of Illinois-born African Americans with those who migrated to Illinois from southern US states as children and adults, and with African American residents of the south. Adult Illinois residents, born between 1913 and 1966, who were diagnosed with cancer from 1986 through 1991 were classified by both birthplace and the state and year their social security number was assigned to determine their migration status: native, early (as child) migrant or late (as adult) migrant. African Americans of Atlanta were used to represent southern homeland ratios. Only lung cancer in African American females showed a statistically significant trend among the four groups, with Illinois native having the highest ratio. Although no trend was identified, Illinois natives had statistically significantly different ratios than both migrant groups and the southern homeland for cancers of the oral cavity (males), colon (females) and leukemias (females). The data also suggested that US regional differences in cancer ratios among African Americans exist (cancers of the prostate and testis, and in females, cancers of the oral cavity, esophagus and kidney), and among those African Americans that migrate to the north from the south, some cancer ratios also change (in males, cancers of the stomach colon, bladder and myeloma and in females, rectal cancer). Further, evidence was found in some cancer sites for the effect of the timing of northern migration on cancer risk (cancer of the rectum (males), liver (both sexes), and in females, cancer of the breast, stomach and nervous system).

Comparability and compatibility: issues in combining data from central cancer registries

Before combining or comparing data from different registries, one should consider similarities and differences in data collection methods, data quality, and underlying populations. What are important population demographic differences? What about differences in data quality: How can these be measured and evaluated? What factors can affect data compatibility? How can one assess data comparability? If registries are compatible, are they always comparable? Are comparable data from registries compatible data? When data are combined, what issues should be considered to determine whether the combined result is meaningful? These are some of the common questions that need to be addressed to determine whether and when data from different registries should be combined or compared.

Heat-related deaths during the July 1995 heat wave in Chicago

BACKGROUND

During a record-setting heat wave in Chicago in July 1995, there were at least 700 excess deaths, most of which were classified as heat-related. We sought to determine who was at greatest risk for heat-related death.

METHODS

We conducted a case-control study in Chicago to identify risk factors associated with heat-related death and death from cardiovascular causes from July 14 through July 17, 1995. Beginning on July 21, we interviewed 339 relatives, neighbors, or friends of those who died and 339 controls matched to the case subjects according to neighborhood and age.

RESULTS

The risk of heat-related death was increased for people with known medical problems who were confined to bed (odds ratio as compared with those who were not confined to bed, 5.5) or who were unable to care for themselves (odds ratio, 4.1). Also at increased risk were those who did not leave home each day (odds ratio, 6.7), who lived alone (odds ratio, 2.3), or who lived on the top floor of a building (odds ratio, 4.7). Having social contacts such as group activities or friends in the area was protective. In a multivariate analysis, the strongest risk factors for heat-related death were being confined to bed (odds ratio, 8.2) and living alone (odds ratio, 2.3); the risk of death was reduced for people with working air conditioners (odds ratio, 0.3) and those with access to transportation (odds ratio, 0.3). Deaths classified as due to cardiovascular causes had risk factors similar to those for heat-related death.

CONCLUSIONS

In this study of the 1995 Chicago heat wave, those at greatest risk of dying from the heat were people with medical illnesses who were socially isolated and did not have access to air conditioning. In future heat emergencies, interventions directed to such persons should reduce deaths related to the heat.

Chorionic villus sampling and transverse digital deficiencies: evidence for anatomic and gestational-age specificity of the digital deficiencies in two studies

Several but not all studies indicate that chorionic villus sampling (CVS) is associated with an increased risk for transverse limb deficiencies, including digital deficiencies. It has been suggested that variations in results regarding the transverse digital deficiencies (TDDs) may be due to the use of different classification criteria. We present the combined analysis of two case-control studies, the U.S. Multistate CVS (US) study and the Italian Multicentric Birth Defects (IP-IMC) study, using two different definitions of TDDs. We compared the frequency of CVS exposure in control infants with that among those infants with any number of affected digits (any TDD), and those with all five digits of at least one limb affected (extensive TDDs). The estimated relative risk (RR) for any TDD following CVS was 10.6 (IPIMC) and 6.6 (US). For the extensive TDDs, the RR was 30.5 (IPIMC) and 10.7 (US). In both studies, extensive TDDs were less than 25% of all TDDs. Compared to all TDDs, extensive TDDs were more likely to occur after CVS performed earlier in the first trimester (before 10-11 weeks' gestation). These findings suggest a relationship between the timing of CVS and the severity of TDDs; indicate that using a restrictive definition of TDDs (all five digits affected) may limit the ability to evaluate the association between CVS and TDDs in populations in whom CVS is usually performed at or after 10 weeks' gestation; and highlight the necessity to consider gestational age in any evaluation of the relative risk for limb deficiencies associated with CVS.

Cancer incidence of Hispanics and non-Hispanic whites in Cook County, Illinois

BACKGROUND

Variations in cancer incidence and mortality rates between Hispanics and non-Hispanic whites have been reported in several regions in the United States. This report describes the ethnic variation in selected cancer sites in Cook County, Illinois.

METHODS

Cancer incidence, age-specific, and age-standardized relative rates, and 95% confidence intervals were estimated among Hispanics and non-Hispanic whites in Cook County, Illinois, for 1986-1987. Hispanics were identified using surnames and maiden names with the Generally Useful Ethnic Search System (GUESS).

RESULTS

Rates of lung cancer in Hispanics were approximately half of those observed for non-Hispanic whites. Hispanics also had lower rates of colon, breast, and bladder cancer. Hispanic females had rates of invasive cervical cancer that were approximately two times higher than those of non-Hispanic whites.

CONCLUSIONS

These results are consistent with previous studies and suggest that Hispanics residing in the United States may retain some of the risk profile of those living in their home country.

Using central cancer-registry data to monitor progress in early detection of breast and cervical cancer (Illinois, United States)

Cases of breast and cervical cancer account for almost 40 percent of all cancers diagnosed in Illinois (United States) women. Information on screening rates, however, is not collected routinely for the populations at risk. This paper reports on surveillance indicators designed to identify target populations and evaluate programs. All cases of cancers of the breast (n = 38,824, including in situ) and invasive cervix (n = 2,763) with a known stage, among women aged 40 to 74, were identified through the population-based Illinois State Cancer Registry for 1986 to 1992. The proportion of breast cancer cases with in situ disease-stage and cervical cancer cases with a late invasive stage were selected as surveillance indicators. Differences by age and race were evaluated, as were age- and race-specific trends. The data suggest that Black women, aged 40 to 74 years, and White women, aged 65 to 74 years, should be targeted for breast-cancer-screening interventions. All women, aged 40 to 74, should be targeted for enhanced cervical-cancer-screening interventions. Significant trends in in situ breast cancer diagnoses were apparent in all age-race groups, however no significant decline in invasive cervical cancer was found for any age-race group. The indicators identified the age- and race-specific disparities among potential target populations for breast and cervical cancer screening.

Increased risk for transverse digital deficiency after chorionic villus sampling: results of the United States Multistate Case-Control Study, 1988-1992

Although numerous infants have been reported with transverse limb deficiencies after their mothers had undergone chorionic villus sampling (CVS), it has been unclear whether the procedure caused these defects. We report the results of the first multistate case-control study to assess and quantify the risk for specific limb deficiencies associated with CVS. Case subjects were 131 infants with nonsyndromic limb deficiency ascertained from 7 population-based birth defect surveillance programs, and born from 1988-1992 to mothers 34 years of age or older. Control subjects were 131 infants with other birth defects. We ascertained exposure to CVS from medical records and maternal and physician questionnaires. We assessed rates and timing of exposure to CVS, and estimated relative and absolute risks for anatomic subtypes of limb deficiency. The odds ratio for all types of limb deficiency after CVS from 8-12 weeks' gestation was 1.7 (95% confidence interval, 0.4-6.3). For specific anatomic subtypes, the strongest association was for transverse digital deficiency (odds ratio = 6.4; 95% confidence interval, 1.1-38.6). The risk for transverse digital deficiency increased with earlier gestational exposure (P < 0.01 for trend). We estimated that the absolute risk for transverse digital deficiency in infants after CVS was 1 per 2,900 births (0.03%). Exposure to CVS was associated with a sixfold increase in risk for transverse digital deficiency. The causality of this association is supported by its strength, specificity, biologic plausibility, and consistency with the results of previous studies. Although some centers already inform patients about risk for limb deficiency, this study quantifies the magnitude of risk associated with CVS from 8-12 weeks' gestation.

Education of rural physicians about breast cancer through an oncology outreach program

In 1990, the Rural Illinois Cancer Consortium initiated an intervention in the management of breast cancer for all rural hospitals. Regional data from a 12-county area were used to identify issues and develop an intervention emphasis. The data suggested two management issues: eliminate unilateral diagnostic mammography and increase the number of patients that have their tumor staged. The intervention involved seminars to provide feedback to physicians on management of breast cancer patients in the region. A series of personalized mailings emphasizing the intervention message were also deployed. Although data are not yet available to measure the intervention impact, immediate feedback on the interventions was sought through a physician survey and several process measures. The immediate feedback measures were assessed. These measures were the penetration of the rural hospital seminars, physician behavior self-reported by mail survey, and number of inquiries to the Physician Data Query. Each of the nine hospitals held a seminar, and 39 percent of the rural physicians treating breast cancer patients attended. Survey data showed physician behavior change in the desired direction, compared with data from the baseline medical record audit conducted in 1986-87. Intervention feedback was useful in defining the implementation success of the interventions. The outcome evaluation, based on medical record audits, is in progress.

Case-control studies of cancer in Illinois farmers using data from the Illinois State Cancer Registry and the U.S. Census of Agriculture

A series of case-control studies have been carried out to compare farmers reported to the Illinois State Cancer Registry (ISCR) with other males reported to the ISCR between 1986 and 1988. Data on the number of farms in each Illinois county producing given agricultural commodities were obtained from the United States Census of Agriculture and used as surrogate exposure indicators. Employment as a farmer was found to be associated with cancer of the eye [odds ratio (OR) = 6.49, 95% confidence interval (CI) = 1.78, 23.71], lip (OR = 4.42, 95% CI = 2.46, 7.94), prostate (OR = 1.15, 95% CI = 0.99, 1.35) and leukaemia (OR = 1.51, 95% CI = 1.01, 2.25). Wheat and soybean production were found to be positively associated with leukaemia. Hay and beef production were found to be positively associated with cancer of the prostate.

Birth-weight-specific infant mortality risks and leading causes of death. Illinois, 1980-1989

OBJECTIVE

To describe birth-weight-specific differences in mortality risks between white and black Illinois infants by age at death and leading cause of death.

DESIGN

Population-based birth cohort study.

SETTING

State of Illinois.

PATIENTS

All Illinois infants who were born from 1980 through 1989 and reported to the Illinois Department of Public Health. The death certificates of these infants were matched to corresponding birth certificates using a computerized linkage algorithm.

INTERVENTIONS

None.

RESULTS

The high black infant mortality rate is attributable to higher mortality risks in the neonatal period for black, normal birth-weight infants and in the postneonatal period for all black infants, regardless of birth weight.

CONCLUSION

Efforts to narrow the black-white gap in infant mortality and to reduce black mortality should not be limited to reduction of low birth weight and premature birth in black infants but should also include efforts to reduce risk factors associated with mortality among normal birthweight black infants.

Age differences in the breast cancer management of rural women

This article describes the effect of age on the pattern of access to breast cancer care among rural women diagnosed in 1986-1989. Cases were identified by the Illinois State Cancer Registry and information on breast cancer management was obtained through review of hospital records and physician survey. Case follow-back was more than 99% complete. State-of-the-art breast cancer management was defined by the Physician Data Query (PDQ) and included diagnostic evaluation, prognostic evaluation, and stage-specific definitive treatment. The data did not indicate age differences in bilateral diagnostic mammography, performance of hormone receptor assays, radiation therapy, and stage-appropriate treatment. Tumor staging and axillary node dissection did differ by age with significant negative trends in the multivariate model. Also, women younger than 55 years were the most likely to receive separate diagnostic biopsy, limited surgery, and chemotherapy, whereas women 55 to 74 years were most likely to receive hormone therapy. Although rural populations may experience barriers to some aspects of state-of-the-art breast cancer management, age does not always exacerbate them.

Cancer in Illinois construction workers: a study

A series of case-control studies using subjects from the Illinois State Cancer Registry have been conducted. Logistic regression was used to control for age and history of tobacco and alcohol use. Construction workers were consistently found to be younger than other subjects and to have used alcohol and tobacco more often. Significant positive associations between cancer of the stomach and welding (odds ratio [OR] = 2.11, 95% confidence interval [CI] = 1.09, 4.09), lung cancer and employment in the construction industry (OR = 1.18, 95% CI = 1.02, 1.26), and lung cancer and welding (OR = 1.68, 95% CI = 1.03, 2.76) were found. Significant negative associations between cancer of the colon and welding (OR = .54, 95% CI = .29, 1.00), cancer of the prostate and employment in the construction industry (OR = .76, 95% CI = .65, .89), cancer of the prostate and plumbing (OR = .44, 95% CI = .38, .50), cancer of the prostate and metal working (OR = .43, 95% CI = .19, .93), and bladder cancer and employment as an electrician (OR = .60, 95% CI = .36, 1.00) suggests that construction workers did not consistently experience excesses of cancers known to be associated with tobacco use, and an overall excess of sites not known to be related to tobacco use may have occurred.

Relation between population density and cancer incidence, Illinois, 1986-1990

Many investigators have examined urbanization gradients in cancer rates. The authors used incidence data for 1986 through 1990 from the Illinois State Cancer Registry, a large, population-based incidence registry, to identify race-specific, urban-rural trends in cancer rates. Using population density, they categorized an urbanization gradient into four groups. Five-year, average annual age-adjusted, site-specific incidence rates were calculated for all sex-race strata within each population density group. Monotonic and statistically significant cancer incidence trends across all race-sex groups were found for cancers of the esophagus, liver, lung, female breast and cervix, male prostate, nervous system, non-Hodgkin's lymphomas, and all cancers combined. No trend was observed for blacks that was not also seen for whites; however, significant trends for cancer of the pancreas and Hodgkin's disease were seen for whites but not for blacks. Colon cancer in males was the only sex-specific trend in cancer that can occur in both sexes. Analytic studies for sites with consistent urban-rural trends across all race-sex groups may be fruitful in identifying the aspect of population density, or other unmeasured factor, that contribute to these trends.

Risk factors for lung cancer among nonsmoking Illinois residents

The purpose of this study was to examine possible risk factors for lung cancer among nonsmokers. The Illinois State Cancer Registry was used to identify all nonsmoking lung cancer cases diagnosed between 1985 and 1987. Subjects were classified as nonsmokers only if their medical record specifically stated that they had never smoked during their lifetime. These cases were compared with nonsmoking colon cancer cases. White male nonsmoking lung cancer cases were more likely to have worked in the construction industry than controls [odds ratio (OR) = 1.6, 95% confidence interval (CI) = 1.2-2.3] and to have worked in the bus service and urban transit industry (OR = 2.6, 95% CI = 1.0-6.9), in the trucking service industry (OR = 2.1, 95% CI = 1.3-3.6), and in blast furnaces, steelworks, and rolling and finishing mills (OR = 1.9, 95% CI = 1.0-3.6). White female cases were more likely to have worked as registered nurses than were the controls (OR = 1.9, 95% CI = 1.0-3.5). Negative associations between lung cancer and farming were found in both white males (OR = 0.6, 95% CI = 0.5-0.7) and white females (OR = 0.1, 95% CI = 0.01-0.6). Several other less plausible associations between employment and lung cancer were also found. To determine whether urban residence and associated air pollution increased the risk of lung cancer for nonsmokers, rates among nonsmokers in Cook County were compared with those in the remainder of Illinois. Cook County rates of nonsmoking lung cancer were elevated among white females and nonwhite females, but not among males. Residences of the white female and nonwhite female lung cancer cases were mapped to determine whether clustering within Chicago had occurred. The absence of observable clustering suggests that the excess of female lung cancer cases in Cook County is not attributable to pollution.

Urban-rural differences in the management of breast cancer

We conducted a population-based study with medical-record review of breast cancer patients as part of a larger study of an oncology outreach program to improve cancer management for rural patients within their community hospital. This analysis compares the breast cancer care received by the rural population before the oncology outreach program with that of the contiguous urban population. All 1986-87 cases in selected Illinois (United States) counties were identified using the Illinois State Cancer Registry. Data were obtained by hospital record audit and physician survey. Case follow-back was 98 percent complete. Rural cases were evaluated separately when they were diagnosed in an urban facility. State-of-the-art management was defined by the 1986-87 Physician Data Query and included diagnostic evaluation, prognostic evaluation, and stage-specific treatment. A summary variable incorporated whether or not the most appropriate management was offered. Compared with urban cases, rural cases diagnosed in rural hospitals were less likely to have staged tumors and more likely to have node dissections. Rural cases traveling to urban centers were less likely to have limited surgery, hormone therapy, and a biopsy as a first-step surgical procedure, and more likely to have node dissection.

Repeat mammography among women over 50 years of age

Mammography decreases mortality among women 50 years of age and older. Although recent surveys show that mammography use has increased since 1983, it continues to be underused by women at risk for breast cancer. The frequency of repeat screening at recommended intervals remains an important unanswered question. This record audit study included all visits from 1986 to 1988 for active female patients, 51-64 years of age, in a family medicine practice. The practice has a disproportionately black patient population, many of whom are on public assistance, characteristics associated with lower compliance with cancer screening guidelines. I reviewed medical records for a physician's recommendation for mammography and also for a radiology report documenting receipt of the mammogram. I also abstracted from the medical record the reason for mammography, a history of breast cancer risk factors, and sociodemographic information. In addition, I noted documentation of a clinical breast examination (CBE) and CBE results. Records for 150 patients were included in the analysis. The results indicate that repeat screening mammography is not common: 3% had three mammograms during the study period; 19% had two; 33% had one; and 44% had none. Physician recommendation for first-time mammography and clinical examination occurred with low frequency. As others report, mammography use is strongly associated with physician recommendation for a mammogram.

A systematic consideration of the neoplastic spectrum of AIDS: registry linkage in Illinois

To examine unexplored aspects of the association between AIDS and neoplasia, the Illinois AIDS and Cancer Registries were linked. The method integrated use of a personal computer to find exact matches on names and dates of birth with manual review to assure satisfaction of a match definition. Of the factors examined, white race and homosexuality predicted Kaposi's sarcoma (KS) among people with AIDS (PWAs), and white race predicted non-Hodgkin's lymphoma (NHL). Earlier reports of a declining proportion of PWAs with KS were confirmed. Lymphoma (mixed lymphocytic/histiocytic type), while not currently diagnostic of AIDS, occurred more frequently among PWAs than in the Illinois population. For the first time, rates of cancers other than KS and NHL were demonstrated to be significantly increased among PWAs compared with general populations. In the light of these findings, reconsideration of current neoplastic definitions may be useful.

An algorithm for matching anonymous hospital discharge records used in occupational disease surveillance: anonymous record matching algorithm

The expense of collecting primary data, coupled with limited authority to mandate reporting, requires alternative methods of implementing an occupational disease registry in Illinois. One alternative data source for surveillance of some occupational diseases is hospital discharge records. Because these records lack personal identifiers, it has been impossible historically to match records belonging to the same individual and obtain reliable case estimates. To circumvent this difficulty, an algorithm has been developed to match anonymous hospital discharge records collected from all Illinois hospitals. The algorithm was based on the assumption that specific combinations of occupational disease code, sex, zip code, and date of birth would identify an individual to whom multiple hospitalizations belong. Matching with the algorithm reduced the 1986 case estimates from 597 to 499 for all cases of coal workers' pneumoconiosis, asbestosis, and silicosis.

Monitoring for clusters of disease: application to leukemia incidence in upstate New York

The authors propose a procedure for the detection of significant clusters of chronic diseases, with particular reference to cancer. The procedure allows for variations in population density and avoids the problem of "post hoc" formation of hypotheses or self-defined populations. This accounts for several of the principal problems of cluster evaluations. The techniques are practical but "computer-intensive." The procedure, termed the "cluster evaluation permutation procedure," is applied to leukemia incidence data for an Upstate New York region obtained from the New York State Cancer Registry and census files. Comparisons are made with two other recently proposed clustering methods, namely the U-statistic method of Whittemore et al. (Biometrika 1987;74:631-7) and the "geographical analysis machine" of Openshaw et al. (Lancet 1988;1:272-3). Routine examination of disease occurrence with the cluster evaluation permutation procedure would allow state health officials to prioritize case investigations and to respond in a timely and efficient manner to inquiries of reported clusters.

Predicting public concern regarding toxic substances in the environment

The purpose of this research was to identify the variables that increase concern about the health, environmental contamination, and economic consequences of toxic substances in the environment. A mail survey was sent to a New York State sample, and a 66% response was obtained. Seven indices were developed from specific concerns about toxic substances in the environment including, among others, exposure, health effects, pollution, and economic consequences. Stepwise regression analysis was conducted for each concern index. The results suggest that the number of information sources regarding environmental issues was a strong predictor of concern in nearly all models. Other variables that were repeatedly found to be important contributors to the models were years of education, attitudes about government involvement in private industry, knowledge of epidemiology and the scientific method, and the perceived proximity to sources of potential contamination. Unlike other research, women and mothers of sick children did not make large contributions to the model.

Public concern about chemicals in the environment: regional differences based on threat potential

While the hazards of chronic environmental pollution remain unclear, people are making decisions about their exposure to pollution and its possible effects on their health. To compare people's concerns about environmental problems, a systematic, stratified sample was surveyed. The sample was made up of residents, ages 25 through 74 years, of three areas of New York State. The three areas were western New York, with a high density of toxic dump sites; Long Island, with a major shallow ground water aquifer; and the remainder of the State, excluding New York City, as a comparison area. The sampling list was obtained from records of licensed drivers of the New York State Department of Motor Vehicles. A 66 percent response rate was obtained to the mailed survey. As expected, most concerns were greater for western New York and Long Island, the two areas with highest threat potential for exposure or contamination, than for the comparison area. The single exception was that no regional differences were noted for concerns about environmental pollution and contamination. All concerns were associated with perceived distance between one's residence and a source of potential exposure. Regardless of region, women were more concerned than men about exposures, pollution, and related health effects. No sex differences, however, were noted for economic concerns.

Sentinel health event surveillance: skin cancer of the scrotum in New York State

Skin cancer of the scrotum is a disease that has been identified as a sentinel health event (occupational) (SHE(O] that is necessarily occupationally related. The present paper examines the feasibility of using this disease in active cancer surveillance in New York State. After consulting cancer case reports, hospital records, death certificates, and city directories, we obtained occupational data for 17 of 22 cases of nonmelanoma skin cancer of the scrotum diagnosed between 1979 and 1984. Only three cases had occupations previously linked to scrotal cancer, while 11 of 19 cases had one or more possibly contributory medical conditions. The probable under-reporting of scrotal cancer cases and incompleteness of occupational data may limit the usefulness of scrotal cancer as a sentinel health event (occupational) unless additional steps are taken.

Early abortion and breast cancer risk among women under age 40

In New York State, incidence of cancer and fetal death are reportable health events mandated by state law. These data enabled a population-based record linkage study of the effect of early pregnancy termination on breast cancer risk to be conducted. In upstate New York 1451 cases under age 40 were reported to the Cancer Registry during 1976-1980. Cases were matched with 1451 population controls by year of birth and by residence using zip codes. All names including those changed by marriage were matched with the reports of fetal deaths occurring between 1971 and 1980. Matched pairs analyses revealed an excess of early pregnancy terminations among cases in all categories. Odds ratios (OR) were significantly elevated among those with an induced abortion (OR = 1.9) and a spontaneous abortion (OR = 1.5). Elevated risks were also noted for consecutive abortion events without intervening livebirths.

Cancer incidence following exposure to drinking water with asbestos leachate

In November 1985, the New York State Department of Health was altered to extraordinary concentrations of asbestos leachate in the drinking water in the Town of Woodstock. Concentrations of 3.2 million fibers per liter (MFL) to 304.5 MFL were found, depending on location. An investigation of cancer incidence in the area was conducted for the period 1973-83 using the State Cancer Registry to compute standardized incidence ratios. No evidence was found of elevated cancer incidence at sites associated with asbestos exposure. A statistically non-significant excess of kidney cancer was seen among men, but not women. Colon cancer among men was significantly low, but incidence among women was similar to that expected. Lung cancer incidence was lower than expected for both sexes. Ovarian cancer rates were not different from expected rates. At sites not previously related to asbestos exposure, cancer of the oral cavity was significantly high, with most affected persons having a history of cigarette smoking. Surveillance of the community is continuing because of an insufficient latent period for some exposed groups.

A comparison of actual and perceived residential proximity to toxic waste sites

Studies of Memphis and Three Mile Island have noted a positive association between actual residential distance and public concern about exposure to the potential of contamination, whereas none was found at Love Canal. In this study, concern about environmental contamination and exposure was examined in relation to both perceived and actual proximity to a toxic waste disposal site (TWDS). It was hypothesized that perceived residential proximity would better predict concern levels that would actual residential distance. The data were abstracted from a New York State, excluding New York City, survey using all respondents (N = 317) from one county known to have a large number of TWDSs. Using linear regression, the variance explained in concern scores was 22 times higher with perceived distance than for actual distance. Perceived residential distance was a significant predictor of concern scores, while actual distance was not. However, perceived distance explained less than 5% of the variance in concern scores.

Recency of Pap smear screening: a multivariate model

Most descriptive reports of women who have not received recent Pap smear screening have been limited to bivariate descriptions. The purpose of this study was to develop a multivariate model to predict the recency of Pap smear screening. A systematic sample of women residents, aged 25 to 74 years, in upstate New York was selected. The women were asked to report use of Pap smear screening during several time periods, their congruence with recommended medical practice, general use of medical services, and a variety of sociodemographic indicators. A log linear weighted least squares regression model was developed, and it explained 30 percent of the variance in recency of Pap smear screening behavior. While the sociodemographic variables were important predictors in the model, the medical care variables were the strongest predictors of recent Pap smear use. A significant relationship between race and recency of Pap smear testing was not supported by these data.

Breast self-examination palpation skill: a methodological note

A current issue in research of breast self-examination (BSE) efficacy is the competency of BSE performance. A nonverbal method for testing two palpation skills were developed. They were designed for mail surveys, or other cost-effective methods of following up subjects in a longitudinal study design. They were also designed to avoid semantics problems associated with verbal descriptions of BSE performance, as noted in earlier research of the author. These pictorial questions were included in three mail surveys of the general female population. The data were consistent across all three surveys for both the verbal and nonverbal questions of palpation skill. The diagram index of the hand part used in BSE suggested that a higher proportion of BSE practicers are using the recommended hand part than their verbal reports would indicate. However, the data from the diagram assessing the amount of pressure used for BSE were not very different from the results of the verbal questions.

Physicians' attitudes toward breast self-examination: a pilot study

Recent case control studies suggest that breast self-examination (BSE) is efficacious in the early clinical diagnosis of breast cancer. Population studies also indicate that adoption of this behavior depends largely on physicians. This pilot study was designed to determine whether physicians' attitudes on this subject could be obtained through a mail questionnaire. The results show that a self-administered mail questionnaire can be used successfully in physician studies. Although questionnaire length does affect response rates (78% one page, 69% four page, and 57% eight page), it does not affect either respondent characteristics or question response.

Age-specific hysterectomy and oophorectomy prevalence rates and the risks for cancer of the reproductive system

A sample survey was conducted in 1982 to determine the prevalence of hysterectomy and oophorectomy among upstate New York women, ages 25-74. The effects of this surgery on age-specific estimates of the risk for cancer of the uterus, cervix, and ovary were calculated. Overall, 16.9 per cent of the women reported having had a hysterectomy and 9.9 per cent reported a bilateral oophorectomy. The adjustment for age-specific hysterectomy increased the 1977-1979 average annual incidence rate of cervical and uterine cancer by 21 per cent. In several five-year age categories, the increase reached 54 per cent. The 1977-1979 average annual incidence rate of ovarian cancer increased by 12 per cent after adjusting for age-specific bilateral oophorectomy. The increase reached 29 per cent in one five-year age group. The sample results show a lower prevalence of hysterectomy among women 25 to 40 years old and among women 70 to 74 than estimates based on the application of mathematical models to data on surgical incidence.

Increasing efficiency in evaluation research: the use of sequential analysis

This paper describes Wald's sequential analysis and briefly reviews the history of its applied use. Two public health applications are presented as examples of how the method helped overcome two common problems associated with evaluation research. In one case, the sequential technique reduced dramatically the workload in an evaluation project where hospital records were being reviewed. In the second case, prompt feedback of data to a breast screening program was facilitated by the sequential method allowing program administrators to refocus their efforts on problem areas before the end of the funding period. On the average, Wald's sequential method results in a savings of 50 per cent in observations as compared to classical sampling procedures. The sample sizes will always terminate with a finite number and will rarely exceed three times the average sample size for any single sample.

Social factors associated with breast self-examination among high risk women

A sample of 708 women, who by sociodemographic characteristics are high risk to breast cancer, were interviewed by telephone about their knowledge, attitude, and practice of breast self-examination (BSE). Reported frequency of BSE, knowledge of BSE and breast cancer, and BSE attitude in this sample are comparable to data reported by others. This report analyzes the associations between the frequency of a breast self-examination practice and the variables, age, education, detection confidence, social influence, modesty, preventive health behaviors, and memory. These relationships are discussed and several new hypotheses are proposed. Since the data were collected retrospectively, they are not able to describe causal relationships.

Enhancing the effectiveness of media messages promoting regular breast self-examination: messages based on innovation adoption principles

As public health departments have come to rely more and more on the mass media for the promotion of healthful behavior, it has become increasingly desirable to determine the motivating capability of the messages used in this promotion. A fact that is well understood is that many times any message at all is better than none. However, the effectiveness of various media health messages has not been well researched.A study was therefore undertaken of the comparative effectiveness of two different messages describing how to do a breast self-examination. An "experimental message" for the study was designed by applying the principles that facilitate innovation adoption to the message's format and presentation. This message's impact was then compared with that of the American Cancer Society's pamphlet "How To Examine Your Breasts." This pamphlet had been mailed to a sample of women similar to those receiving the experimental message, but who lived in a different geographic area. The use in each area of control groups who had received no messages afforded an opportunity to study maturation effects (other factors than the mailings that might have influenced study results).The experimental message proved more successful in persuading women to adopt breast self-examination than the comparison message. The women who reported a change in breast self-examination practice following the mailing could be characterized as having a more extensive social support system to promote breast examinations and as having a pre-experiment perception that breast self-examination was a complex practice to perform repeatedly at regular intervals.