Tobacco control indicators and lung cancer rates in young adults by state in the United States.

OBJECTIVE: To determine the association between a tobacco control index (TCI) for 1992-93 for each US state and lung and bronchus cancer mortality and incidence rates by state for younger adults, as the best indicator of the effects of recent progress in tobacco control on lung cancer control. DESIGN: For all 51 US areas (50 states and the District of Columbia), correlation coefficients between the state's TCI and lung cancer rate were analysed. Multiple linear regression models (MLR) predicting cancer rates included sociodemographic variables by state (from the 2000 census). In addition, the 51 areas also were divided into tertiles, from highest to lowest TCI, and means for lung cancer rates were compared. SUBJECTS AND SETTINGS: All areas with available data on lung cancer mortality and incidence rates. MAIN OUTCOME MEASURES: Age-standardised mortality rates were available for all 51 areas (50 states and the District of Columbia) for 1989-93, 1994-98 and 1999-2003 for lung cancer at age 15-44 years. Lung cancer incidence rates for 1999-2002 were available for age 20-44 years for 44 states. RESULTS: The correlation between the TCI and lung cancer mortality rate increased in magnitude from 1989-93 to 1999-2003, and the association was statistically significant in MLR models for 1994-98 and 1999-2003 (but not 1989-93). The TCI was statistically significantly correlated with the lung cancer incidence rate in 1999-2002 by state, and the association persisted in an MLR model. CONCLUSIONS: Within the limitations of ecologic analyses, findings are consistent with effects of state tobacco control efforts on reducing state-wide lung cancer rates in younger adults.

Type of breast reconstructive surgery among breast cancer patients: a population-based study.

The type of breast reconstructive surgery (implant versus flap) was examined among all Connecticut-resident breast cancer patients diagnosed between 1994 and 1997 and identified from a population-based cancer registry. Type of reconstruction was obtained primarily from questionnaires sent to hospitals, but physicians were contacted about selected patients. Among 526 patients who underwent reconstruction, reconstruction with a flap (with or without an implant; 367 patients, or 69.8 percent) was more frequent than reconstruction with an implant alone (111 patients, or 21.1 percent); the type of reconstruction was unknown for 48 patients (9.1 percent). Some disagreement was found between reports from physicians and hospitals in a subsample of patients diagnosed in 1997. This study describes the baseline data and methods for examining trends in type of reconstruction among breast cancer patients in a defined population.

Poverty, comorbidity, and survival of colorectal cancer patients diagnosed in Connecticut.

Studies have reported reduced survival rates for colorectal cancer patients in lower socioeconomic status categories, but this finding could be due (at least in part) to higher comorbidity. This study involved 1,219 patients diagnosed with invasive colorectal cancer in 1992 who were reported to the population-based Connecticut Tumor Registry and followed to their death or through the end of 1997. Risk of death was elevated for patients living in census tracts in the highest quintile for poverty rate, independent of comorbidity (as recorded in a hospital discharge database), age, and stage at diagnosis. Patients living in census tracts with a poverty rate of 20 percent or higher had the highest risk of death. The explanation for these findings requires further study, in order to reduce socioeconomic status disparities in survival rates.

How frequent is postmastectomy breast reconstructive surgery? A study linking two statewide databases.

Studies of the frequency of reconstructive surgery among breast cancer patients have involved non-population-based sources, incomplete ascertainment, or both. Postmastectomy breast reconstructive surgery among 4688 breast cancer patients (diagnosed between 1992 and 1996) was estimated by linking a population-based cancer registry with a statewide hospital discharge database in Connecticut. Of these 4688 patients, 585 (12.5 percent) had reconstruction coded in one database or both databases. The reconstruction rates were higher than in a previous study in Connecticut and increased from 9 percent in 1992 to 16 percent in 1996. Reconstruction was not related to the patient's tumor size, marital status, or race (black versus white), but declined with increasing age at diagnosis and with poverty rate (of the census tract of residence). These associations were similar to those reported from previous studies.

Agreement in race-ethnicity coding between a hospital discharge database and another database.

OBJECTIVE: This study assessed the agreement in coding of race and Hispanic ethnicity for the same patients in a hospital discharge database and another database. METHODS: Race-ethnicity coding was examined for 72,276 cancer patients discharged from Connecticut hospitals (1992-1997) who were linked with the statewide cancer registry that included information on birthplace (for 76% of patients) and surname (for all patients). Surnames in the cancer registry were also linked with a list of Spanish surnames, to improve ascertainment of ethnicity among patients of probable Hispanic origin. RESULTS: Kappa coefficients (kappa) indicated substantial agreement (beyond that expected by chance) for White (kappa = .74) and Black (kappa = .93) race, Hispanic ethnicity (kappa = .73) and non-Hispanic White race-ethnicity (kappa = .83) categories. Kappa was moderate (ie, .52) for the Asian-Pacific Islander race category. Only 42% of Asian-Pacific Islanders and 62% of Hispanics in the cancer registry were similarly coded in the discharge database. CONCLUSIONS: Although both databases are imperfect, the findings suggest that ascertainment of certain racial-ethnic groups in hospital discharge databases could be improved if birthplace and surname were available.

Disparities in surgical treatment of early-stage non-small-cell lung cancer.

This study involved 1,564 black or white patients diagnosed in 1992 to 1997 with non-small-cell lung cancer, reported to the population-based Connecticut Tumor Registry, who were linked with a statewide hospital discharge database that provided information on comorbid conditions. While only 11.4 percent of patients did not receive surgical treatment (lung resection), this proportion increased with rising age and was higher among patients who resided in a census tract in the highest poverty-rate quintile, were black, not married and had one or more selected comorbid conditions. These associations persisted in logistic regression models that included all of the variables as predictors of surgery. Studies are needed to explain these disparities.

Hospital-based smoking cessation programs: a statewide survey.

Hospitalization is regarded as a "window of opportunity" for initiating smoking-cessation programs, but the proportion of hospitalized smokers reached by such programs is uncertain. While 25 of 31 Connecticut acute-care hospitals surveyed (March-April 2000) had a smoking cessation program, only three had a program specifically for hospital patients. Each of these three programs counseled several hundred patients per year. Programs at the other hospitals included few hospital patients. Only a small proportion of the estimated hospitalized (inpatient) smokers were reached by a hospital-based cessation program. Hospitalized smokers remain largely a missed opportunity with regard to enhancing long-term smoking cessation rates.

Inpatient hospital admission through an emergency department in relation to stage at diagnosis of colorectal cancer.

The purpose of this study was to use population-based sources to estimate the frequency and characteristics of first inpatient hospital admission through an emergency department (ED) among 11,023 patients with diagnosed colorectal cancer between 1992 and 1996. Patients were identified from the population-based Connecticut Tumor Registry. Linkage with a statewide hospital discharge database (inpatient only) disclosed that 20% had a first hospital inpatient admission through an ED. Inpatient admission through an ED was statistically significantly associated with older age and race and was a statistically significant risk factor for distant stage at diagnosis. Studies are needed of the roles of patient delay and lack of screening in influencing ED presentation, especially in the elderly.

Black-white differences in sentinel causes of death: counties in large metropolitan areas.

The black/white ratio of death rates (before 65 years of age) in 1994-1996 for a group of "sentinel" causes, regarded as preventable by medical treatment and as useful in assessing overall quality of health care, was examined for 60 US counties located in large metropolitan areas. Counties with the highest black/white death rate ratios (>3.5) and the highest death rates for blacks included the District of Columbia; Essex (Newark), New Jersey; Cook (Chicago), Illinois; Wayne (Detroit), Michigan; and Dade (Miami), Florida. In these five counties, in contrast to the US, the death rate from the sentinel causes for blacks had not declined from 1979-1981 to 1994-1996. The findings suggest that racial inequities in health care may be unusually great in certain counties in large metropolitan areas, and that further studies are needed to explain the variation among counties in the black-white ratio of mortality from the sentinel causes.

Hospital charges for colorectal cancer patients first admitted through an emergency department.

Some 11,023 colorectal cancer patients diagnosed in 1992-96 in Connecticut first were admitted to a hospital through a hospital emergency department. The average hospital inpatient charges and average length of stay were about 60 percent higher for emergency department versus nonemergency department first admissions. Emergency department status was an important predictor of charges independent of age at diagnosis and length of stay. Applying these data to the approximately 131,000 colorectal cancers diagnosed in 1998 in the U.S., estimated excess hospital costs due to emergency department versus nonemergency department first admissions for colorectal cancer were about $328 million.

Later-stage cancer in relation to medically underserved areas in Connecticut.

Health status indicators, including advanced stage at diagnosis of cancer, have been proposed as indicators of health care access and quality to be considered in revising the definition of medically underserved areas (MUAs). Using the population-based Connecticut Tumor Registry, "outlier" census tracts were defined as those that had a high proportion of breast, colorectal, or cervical cancer diagnosed at later stage, relative to all tracts in the state. In the six Connecticut cities that had MUAs, MUAs comprised the majority of outlier tracts, but non-MUA outliers were often located on the fringes of MUAs. The findings are discussed in relation to revising the criteria for defining MUAs and to targeting interventions for early detection of cancer in urban areas.

Geographic variation in postmastectomy breast reconstruction rates.

Data on postmastectomy breast reconstructive surgery were examined for 52,357 female breast cancers that were treated with mastectomy and diagnosed in geographic areas covered by the National Cancer Institute's Surveillance, Epidemiology, and End Results Program. The proportion of cancers that involved reconstruction varied in these geographic areas in each age group (under age 70 years) by a factor of about four or five, even after adjustment for stage at diagnosis, marital status, and poverty rate of county of residence at diagnosis. Studies are needed to explain the large differences in reconstruction rates by geographic area.

Trends in late-stage breast, cervical and colorectal cancers in blacks and whites.

OBJECTIVE: This study examined the black-white disparities in diagnosis of late-stage breast, cervical, and colorectal cancers, which could change over time, especially in view of changes in the health-care system. METHODS: Recent trends (1988-95) in stage distribution were examined by age group (<65 and 65+ years) for the geographic areas covered by the National Cancer Institute's Surveillance, Epidemiology and End Results (SEER) Program. RESULTS: Proportions of breast and cervical (but not colorectal) cancers that were late-stage declined in both races and black-white disparities had narrowed by 1995, especially for patients diagnosed at age 65 years or older. The proportion of breast and cervical cancers that were late stage declined for blacks in counties with a black poverty rate of 20% or higher. CONCLUSIONS: Some progress toward black-white equity in stage at diagnosis was evident, consistent with reported trends in use of screening tests, but further surveillance and research are needed.

Postmastectomy breast reconstruction in Connecticut: trends and predictors.

Using the population-based Connecticut Tumor Registry, postmastectomy breast reconstruction was examined in the 10,756 breast cancers in Connecticut women (including 10,133 in white women and 554 in black women) diagnosed from 1988 to 1995. Reconstruction increased from 6.4 percent of cancers in 1988 to 9.1 percent in 1991, but it declined to 4.7 percent of cancers in 1992 (when the Food and Drug Administration instituted a restriction on the use of silicone gel implants); by 1995, the rate had recovered to 8.5 percent. Reconstruction was negatively associated with age, poverty rate of the census tract of residence, and black (versus white) race; these associations require further study.

Regional variation in smoking among African Americans.

BACKGROUND: The impact of geographic region and metropolitan residence on smoking prevalence among African Americans has not been adequately examined. METHODS: This study analyzed 5 years of data from the National Health Interview Survey (1990-1994) on current smoking and regional variation among 16,738 African Americans. Results. Respondents in the West had the lowest unadjusted smoking prevalence rates and Midwest residents had the highest. Current smoking was lower among African Americans living in non-central cities than in central cities even after adjusting for several sociodemographic covariates. Multivariate logistic regression analysis revealed that black women in the South were significantly less likely to be smokers compared with any other gender/region group. CONCLUSIONS: These findings suggest the significance of gender and regional factors such as the social history of migration, social stress and racism, exposure to tobacco advertisement, variations in cultural influences, community structures, and coping strategies in under standing African American smoking behavior.