Validation studies for models projecting the risk of invasive and total breast cancer incidence.

BACKGROUND: In 1989, Gail and colleagues developed a model for estimating the risk of breast cancer in women participating in a program of annual mammographic screening (designated herein as model 1). A modification of this model to project the absolute risk of developing only invasive breast cancer is referred to herein as model 2. We assessed the validity of both models by employing data from women enrolled in the Breast Cancer Prevention Trial. METHODS: We used data from 5969 white women who were at least 35 years of age and without a history of breast cancer. These women were in the placebo arm of the trial and were screened annually. The average follow-up period was 48.4 months. We compared the observed number of breast cancers with the predicted numbers from the models. RESULTS: In terms of absolute risk, the ratios of total expected to observed numbers of cancers (95% confidence intervals [CIs]) were 0.84 (0. 73-0.97) for model 1 and 1.03 (0.88-1.21) for model 2, respectively. Within the age groups of 49 years or less, 50-59 years, and 60 years or more, the ratios of expected to observed numbers of breast cancers (95% CIs) for model 1 were 0.91 (0.73-1.14), 0.96 (0.73-1. 28), and 0.66 (0.52-0.86), respectively. Thus, model 1 underestimated breast cancer risk in women more than 59 years of age. For model 2, the risk ratios (95% CIs) were 0.93 (0.72-1.22), 1.13 (0.83-1.55), and 1.05 (0.80-1.41), respectively. Both models exhibited a tendency to overestimate risk for women classified in the higher quintiles of predicted 5-year risk and to underestimate risk for those in the lower quintiles of the same. CONCLUSION: Despite some limitations, these methods provide useful information on breast cancer risk for women who plan to participate in an annual mammographic screening program.

Issues and findings in the evaluation of occupational risk among women high nickel alloys workers.

BACKGROUND: We present the mortality experience for a cohort of women (n = 2,877) from a large epidemiologic study of production and fabrication high nickel alloys workers (n = 31,165). All the plants were located within the United States and cohort eligibility required some work experience within the period of the late 1940s through the mid 1960s. METHODS: Vital status follow-up was through the end of 1988 and incorporated information from multiple sources. Cause-specific mortality was evaluated by comparing cohort mortality to the general United States female population and to local populations in geographic proximity to the plants. Relative risk estimates were determined for 62 cause of death categories using the Standardized Mortality Ratio (SMR) and were adjusted for age, race, gender, and calendar time by the indirect method using a modified life table technique. RESULTS: Relative risks for all causes (0.98), all cancers (0.90), lung cancer (1.34), and breast cancer (0.96) were nonsignificant when mortality was compared to the US female population. No relationship between mortality and length of time employed in the industry or work area was identified. DISCUSSION: Although there were some difficulties in tracing women due to name changes, comprehensive follow-up was obtained when using multiple sources of information. Our strategy resulted in resolving vital status for over 95% of the women, which is comparable to that of the male cohort. The type of jobs and work activities differed between genders. Females were employed predominantly in two work areas (allocated services, 87%, and grinding, 46%), whereas males were employed in several work areas (allocated services, 76%, grinding, 27%, hot working, 20%, and cold working, 17%). Considerable variation was noted among the study plants with respect to the percent of female production workers in the workforce. Generally, the patterns of relative risks derived for the total cohort and various subgroups are similar across the different comparison populations. Estimated elevated risks are usually lower when cohort mortality is compared to that of local populations. No increased risks were identified for any site-specific cancers or nonmalignant causes of death.

Using alternative comparison populations to assess occupation-related mortality risk. Results for the high nickel alloys workers cohort.

The focus of this article is to examine how the choice of comparison group affects the identification and interpretation of cause-specific health risks in occupational cohorts when different external control populations are used. The mortality experience of approximately 31,000 high nickel alloys workers is compared with the total US population and to local populations in geographic proximity to the plants. Generally, the patterns of relative risks derived for the total cohort and various subgroups are similar across the different comparison populations. Estimated elevated risks are usually lower when cohort mortality is compared with that of local populations. An overall significant 13% risk for lung cancer is noted when compared with that of the total US population. However, no significant excess is identified when local populations are used. Subset analysis identified significant excesses of colon cancer among nonwhite males (50%-150%) and kidney cancer among white male workers employed in melting (approximately 100%), irrespective of the comparison population.

Tamoxifen for prevention of breast cancer: report of the National Surgical Adjuvant Breast and Bowel Project P-1 Study.

BACKGROUND: The finding of a decrease in contralateral breast cancer incidence following tamoxifen administration for adjuvant therapy led to the concept that the drug might play a role in breast cancer prevention. To test this hypothesis, the National Surgical Adjuvant Breast and Bowel Project initiated the Breast Cancer Prevention Trial (P-1) in 1992. METHODS: Women (N=13388) at increased risk for breast cancer because they 1) were 60 years of age or older, 2) were 35-59 years of age with a 5-year predicted risk for breast cancer of at least 1.66%, or 3) had a history of lobular carcinoma in situ were randomly assigned to receive placebo (n=6707) or 20 mg/day tamoxifen (n=6681) for 5 years. Gail's algorithm, based on a multivariate logistic regression model using combinations of risk factors, was used to estimate the probability (risk) of occurrence of breast cancer over time. RESULTS: Tamoxifen reduced the risk of invasive breast cancer by 49% (two-sided P<.00001), with cumulative incidence through 69 months of follow-up of 43.4 versus 22.0 per 1000 women in the placebo and tamoxifen groups, respectively. The decreased risk occurred in women aged 49 years or younger (44%), 50-59 years (51%), and 60 years or older (55%); risk was also reduced in women with a history of lobular carcinoma in situ (56%) or atypical hyperplasia (86%) and in those with any category of predicted 5-year risk. Tamoxifen reduced the risk of noninvasive breast cancer by 50% (two-sided P<.002). Tamoxifen reduced the occurrence of estrogen receptor-positive tumors by 69%, but no difference in the occurrence of estrogen receptor-negative tumors was seen. Tamoxifen administration did not alter the average annual rate of ischemic heart disease; however, a reduction in hip, radius (Colles'), and spine fractures was observed. The rate of endometrial cancer was increased in the tamoxifen group (risk ratio = 2.53; 95% confidence interval = 1.35-4.97); this increased risk occurred predominantly in women aged 50 years or older. All endometrial cancers in the tamoxifen group were stage I (localized disease); no endometrial cancer deaths have occurred in this group. No liver cancers or increase in colon, rectal, ovarian, or other tumors was observed in the tamoxifen group. The rates of stroke, pulmonary embolism, and deep-vein thrombosis were elevated in the tamoxifen group; these events occurred more frequently in women aged 50 years or older. CONCLUSIONS: Tamoxifen decreases the incidence of invasive and noninvasive breast cancer. Despite side effects resulting from administration of tamoxifen, its use as a breast cancer preventive agent is appropriate in many women at increased risk for the disease.

Long-term tamoxifen citrate use and potential ocular toxicity.

PURPOSE: To estimate the prevalence of abnormalities in visual function and ocular structures associated with the long-term use of tamoxifen citrate. METHODS: A single-masked, cross-sectional study involving multiple community and institutional ophthalmologic departments was conducted with a volunteer sample of 303 women with breast cancer currently taking part in a randomized clinical trial to determine the efficacy of tamoxifen (20 mg/day) in preventing recurrences. Participants included women who had never been on drug (n=85); women who had taken tamoxifen for an average of 4.8 years, then been off the drug for an average of 2.7 years (n=140); and women who had been on tamoxifen continuously for an average of 7.8 years (n=78). Women were evaluated by questionnaire, psychophysical testing, and clinical examination to determine any abnormalities in visual function and the comparative prevalences of corneal, lens, retinal, and optic nerve pathology. RESULTS: There were no cases of vision-threatening ocular toxicity among the tamoxifen-treated participants. Compared with nontreated participants, the tamoxifen-treated women had no differences in the activities of daily vision, visual acuity measurements, or other tests of visual function except for color screening. Intraretinal crystals (odds ratio [OR]=3.58, P=.178) and posterior subcapsular opacities (OR=4.03, P=.034) were more frequent in the tamoxifen-treated group. CONCLUSIONS: Women should have a thorough baseline ophthalmic evaluation within the first year of initiating tamoxifen therapy and receive appropriate follow-up evaluations.

Prognosis among African-American women and white women with lymph node negative breast carcinoma: findings from two randomized clinical trials of the National Surgical Adjuvant Breast and Bowel Project (NSABP).

BACKGROUND: A disparity in breast carcinoma survival between African-American and white women has been noted over the past several decades. A major factor implicated in this disparity is stage of disease at diagnosis. In this study, survival and related endpoints were examined among African-American women and white women with lymph node negative breast carcinoma who participated in two randomized clinical trials of the National Surgical Adjuvant Breast and Bowel Project (NSABP). METHODS: Patients from two studies, one conducted among patients with estrogen receptor (ER) negative tumors and the other among patients with ER positive tumors, were included. Study goals were to determine whether African-Americans and whites had comparable outcomes, accounting for ER status and differences in patient characteristics at diagnosis, and to determine whether treatment response was similar for African-Americans and whites. RESULTS: Five-year survival rates were 83% for African-Americans and 85% for whites among ER negative patients, and 93% for African-Americans and 92% for whites among ER positive patients. Rates of disease free survival (DFS) (i.e., time to disease recurrence, second primary cancer, or death) were 71% for African-Americans and 74% for whites at 5 years among ER negative patients, and 81% for African-Americans and 80% for whites among ER positive patients. African-Americans tended to have less favorable baseline prognostic characteristics. Adjusted relative risk (RR) estimates indicated similar prognosis for African-Americans compared with whites for mortality (African-American/white RR = 1.02 with 95% confidence interval [CI], 0.66-1.56 among ER negative patients; RR = 1.14 with 95% CI, 0.84-1.54 among ER positive patients) and DFS (RR = 0.98 with 95% CI, 0.70-1.37 for ER negative patients; RR = 0.96 with 95% CI, 0.75-1.22 for ER positive patients). Estimated percent reductions in DFS events for patients receiving adjuvant therapy were 32% for ER negative African-Americans, 36% for ER negative whites, 20% for ER positive African-Americans, and 39% for ER positive whites. CONCLUSIONS: African-American and white patients with localized breast carcinoma had similar outcomes and benefited equally from systemic therapy. These results suggest that early detection and appropriate therapy among African-American patients could result in a reduction in the current disparity in breast carcinoma mortality between African-Americans and whites.

Design and current status of the NSABP breast cancer prevention trial.

From its inception the BCPT has been the focus of bitter sustained controversy. In spite of reassessments by numerous advisory groups that have deemed the trial to be scientifically, clinically and ethically sound, its opponents have continued their assaults. That the trial has survived is a tribute to Dr. Fisher's leadership and the dedication of many clinical investigators, who, in spite of the injustice that they have endured, continue to believe in the goals of the NSABP clinical trials. Even greater credit and recognition should be given to the women randomized who have continued their commitment to the study. Their altruism and courage are to be acknowledged. These women undoubtedly understand that the trial affords potential risks, as well as benefits, and that even if the drug is eventually found effective in preventing breast cancer, half of them will have received a placebo in the trial. The ultimate tragedy will be if the study fails to answer definitively the study hypotheses because of what has occurred this past year. We are hopeful that, despite this disruption, the trial will continue to a successful completion. While larger multicenter randomized clinical trials clearly have vulnerabilities and imperfections, we should not lose sight that they are clearly the best mechanism for rigorously testing new therapies in a manner that will provide reliable, generalizable findings.

Statistical concepts and biomedical research.

An understanding of statistical concepts is important for biomedical researchers. Statistics and probability are the scientific basis for dealing with sources of variation inherent in most biologic processes. This paper provides an introduction to the statistical framework that is fundamental to scientific methods of inquiry. Topics covered include the role of statistics, hierarchy of evidence, reliability and bias, sample size and power, random samples and random allocation, hypothesis testing, estimation, statistical inference, and publication bias. Future papers will focus on specific statistical techniques commonly used in biomedical research.

Determinants of black/white differences in colon cancer survival.

BACKGROUND: Blacks have lower survival rates for colon cancer than whites, possibly related to more advanced stages of disease at diagnosis and to socioeconomic differences between blacks and whites. While the black/white difference in colon cancer survival is well documented, the few studies that have investigated this difference have been limited by the modest number and type of explanatory factors that were considered. PURPOSE: We analyzed data from the National Cancer Institute Black/White Cancer Survival Study to determine 1) what characteristics might contribute to the racial difference in colon cancer survival and 2) if a survival disparity remained between black and white patients after adjustment was made for these characteristics. METHODS: This prospective study included 454 blacks and a stratified random sample of 521 whites, aged 20-79 years, with cancer of the colon diagnosed from January 1, 1985, through December 31, 1986, and who were residents of the metropolitan areas of Atlanta, New Orleans, and San Francisco/Oakland. Follow-up was truncated on December 31, 1990. Cox proportional hazards regression was used to estimate the death rate among blacks relative to that among whites after adjustment for potential explanatory factors, including sociodemographic factors, concurrent (comorbid) medical conditions, stage at diagnosis, tumor characteristics, and treatment. All P values were calculated from two-tailed tests of statistical significance. RESULTS: After adjustment for age, sex, and geographic area, the black-to-white mortality hazard ratio (HR) was 1.5 (95% confidence interval [CI] = 1.2-1.9), indicating that the risk of death among black patients was 50% higher than that among white patients. Further adjustment for stage reduced the excess cancer mortality to 20% (HR = 1.2; 95% CI = 1.0-1.5), decreasing the overall racial difference in excess mortality from 50% to 20% or to a 60% reduction in excess mortality. Although adjustment for poverty reduced the excess mortality by 20%, adjusting for both stage and poverty did not further reduce the racial difference. Among patients with stages II and III disease, blacks had lower survival rates than whites (HR = 1.8; 95% CI = 1.0-3.1 and HR - 1.5; 95% CI = 1.0-2.3, respectively). Among those patients with metastatic disease (stage IV), survival was similar for whites and blacks. CONCLUSIONS: Stage at diagnosis accounted for more than half of the excess colon cancer mortality observed among blacks. Poverty and other socioeconomic conditions, general health status, tumor characteristics, and general patterns of treatment did not further explain the remaining survival disadvantage among blacks. IMPLICATIONS: Because the racial disparity was confined to earlier stages, future studies should investigate whether blacks have more advanced disease at diagnosis and whether less aggressive treatment is provided because of understanding.

Reanalysis and results after 12 years of follow-up in a randomized clinical trial comparing total mastectomy with lumpectomy with or without irradiation in the treatment of breast cancer.

BACKGROUND: Previous findings from a clinical trial (Protocol B-06) conducted by the National Surgical Adjuvant Breast and Bowel Project (NSABP) indicated the worth of lumpectomy and breast irradiation for treating breast cancer. After the discovery by NSABP staff members of falsified information on patients enrolled in the study by St. Luc Hospital in Montreal, separate audits were conducted at St. Luc Hospital and other participating institutions. We report the results of both audits and update the study findings through an average of 12 years of follow-up. METHODS: Patients with either negative or positive axillary nodes and tumors 4 cm or less in diameter were randomly assigned to one of three treatments: total mastectomy, lumpectomy followed by breast irradiation, or lumpectomy without irradiation. Three cohorts of patients were analyzed. The first cohort included all 2105 randomized patients, who were analyzed according to the intention-to-treat principle. The second cohort consisted of 1851 eligible patients in the first cohort with known nodal status who agreed to be followed and who accepted their assigned therapy (among those excluded were 6 patients from St. Luc Hospital who were declared ineligible because of falsified biopsy dates). The third cohort consisted of the patients in the second cohort minus the 322 eligible patients from St. Luc Hospital (total, 1529 patients). RESULTS: Regardless of the cohort, no significant differences were found in overall survival, disease-free survival, or survival free of disease at distant sites between the patients who underwent total mastectomy and those treated by lumpectomy alone or by lumpectomy plus breast irradiation. After 12 years of follow-up, the cumulative incidence of a recurrence of tumor in the ipsilateral breast was 35 percent in the group treated with lumpectomy alone and 10 percent in the group treated with lumpectomy and breast irradiation (P < 0.001). CONCLUSIONS: Our findings continue to indicate that lumpectomy followed by breast irradiation is appropriate therapy for women with either negative or positive axillary nodes and breast tumors 4 cm or less in diameter.

Occupationally related cancer risk among coke oven workers: 30 years of follow-up.

This study concerns the update of cause-specific mortality among coke oven workers. Updated information provides 3 decades of work history and vital status follow-up on 15,818 workers. Mortality patterns are summarized by race, cumulative exposure, and period of follow-up. The findings are consistent with those from earlier assessments, indicating that occupational exposure to coke oven emissions is associated with significant excess mortality from cancer of the respiratory system and of the prostate. Depending on the segment of the population considered, the respiratory cancer risk for coke oven workers ranged as high as 4.45 times that for non-oven workers. Relative risk values for cancer of the prostate ranged as high as 1.93. Rates of respiratory cancer across period of follow-up are declining, suggesting that the implementation of emissions control and occupational exposure limits has been beneficial.

Racial differences in survival from breast cancer. Results of the National Cancer Institute Black/White Cancer Survival Study.

OBJECTIVE: To examine the ability of recognized prognostic factors for breast cancer to account for the observed poorer survival in blacks compared with their white counterparts. DESIGN AND PARTICIPANTS: Subjects included 1130 women (612 blacks and 518 whites) aged 20 to 79 years residing in metropolitan Atlanta, Ga, New Orleans, La, or San Francisco/Oakland, Calif, who were diagnosed with primary invasive breast cancer. Information on stage, tumor characteristics, treatment, comorbid conditions, and sociodemographic factors was obtained from personal interview, physician and hospital records, and a pathology review of biopsy and surgical specimens. MAIN OUTCOME MEASURE: Multivariable survival models were used to estimate the hazard ratio (relative risk of mortality) for blacks compared with whites, adjusting for various combinations of potential explanatory factors. RESULTS: After controlling for geographic site and age, the risk of dying was 2.2 times (95% confidence interval [CI], 1.8 to 2.8) greater for blacks than whites. Adjustment for stage reduced the risk from 2.2 to 1.7; further adjustment for sociodemographic variables had no effect. Treatment was not a contributing factor once stage and tumor pathology were in the model. After adjusting for stage, treatment, comorbid illness, and pathologic and sociodemographic variables, blacks continued to demonstrate a slightly increased, but not statistically significant, risk of death (hazard ratio = 1.3; 95% CI, 1.0 to 1.8). Results were similar for all-cause mortality and breast cancer-specific mortality. CONCLUSIONS: Approximately 75% of the racial difference in survival was explained by the prognostic factors studied. Sociodemographic variables appeared to act largely through racial differences in stage at diagnosis, which may be amenable to change through improved access to and use of screening for black women.

Endometrial cancer in tamoxifen-treated breast cancer patients: findings from the National Surgical Adjuvant Breast and Bowel Project (NSABP) B-14.

BACKGROUND: Tamoxifen is advantageous in treating all stages of breast cancer. However, studies have suggested that incidence and severity of endometrial cancer increase in women treated with tamoxifen. PURPOSE: We compared rates of endometrial and other cancers in tamoxifen- and non-tamoxifen-treated patients and described the pathologic characteristics of the endometrial cancers. METHODS: Data were analyzed on 2843 patients with node-negative, estrogen receptor-positive, invasive breast cancer randomly assigned to placebo or tamoxifen (20 mg/d) and on 1220 tamoxifen-treated patients registered in NSABP B-14 subsequent to randomization. Average time on study is 8 years for randomly assigned patients and 5 years for registered patients. RESULTS: The incidence rates of liver, gastrointestinal, urinary tract, and nonuterine genital tumors were not increased by tamoxifen treatment. Twenty-five endometrial cancers were originally reported, one of which was reclassified after subsequent review. Two cases occurred in the placebo group in patients whose medical status subsequent to random assignment had required tamoxifen treatment. Twenty-three occurred in the tamoxifen groups. Twenty-one of the 24 originally reported endometrial cancers were FIGO stage 1; 18 of 23 gradable cases were of good to moderate histologic grade. Four tamoxifen-treated women died of uterine cancer. The average annual hazard rate of endometrial cancer as a first event within the first 5 years of follow-up in the randomized, tamoxifen-treated group was 1.2/1000 patient-years; the cumulative hazard rate was 6.3/1000. Findings for the registered, tamoxifen-treated group were similar. Including all originally reported endometrial cancers, the annual hazard rate through all follow-up was 0.2/1000 in the placebo group and 1.6/1000 in the randomized, tamoxifen-treated group; the relative risk of endometrial cancer for the latter versus the former group was 7.5. Again for the latter group, using population-based rates of endometrial cancer from SEER data and information from another NSABP (B-06) trial, relative risks were 2.2 and 2.3, respectively. The 5-year cumulative hazard rate for disease-free survival in the randomized tamoxifen group was 38% less than that in the placebo group. Some data in this paper were provided by an investigator who submitted fraudulent data to the NSABP [see the "News" section]; therefore, the reader must read the entire text including Table 10 and the Editor's notes. In brief, data on 182 of the 2843 randomly assigned patients and 37 of the 1220 registered patients were provided by the investigator in question. After review, 24 of the 182 records showed falsification, all involving characteristics of patients prior to random assignment. Of the 37 registered-patient records, 8 showed falsification. CONCLUSIONS: Risk of endometrial cancer increases following tamoxifen therapy for invasive breast cancer; however, net benefit greatly outweighs risk. Endometrial cancers occurring after tamoxifen therapy do not appear to be of a different type with a worse prognosis than are such tumors in non-tamoxifen-treated patients. IMPLICATIONS: Tamoxifen treatment for breast cancer should continue. In addition, the relative risk of endometrial cancer observed in B-14 tamoxifen-treated patients is consistent with the twofold relative risk used in the initial risk-benefit computation for the NSABP breast cancer prevention trial.

Breast cancer: factors associated with stage at diagnosis in black and white women. Black/White Cancer Survival Study Group.

BACKGROUND: Numerous studies have reported differences in cancer staging at diagnosis and in survival between Black and White patients with breast cancer. Utilizing data obtained from the National Cancer Institute's (NCI's) Black/White Cancer Survival Study for the period 1985-1986, a new study is presented here that systematically examines multiple explanatory factors (e.g., lack of mammograms) associated with these cancer-staging differences. PURPOSE: We evaluated within a single study the relationship of selected demographic, lifestyle, antecedent medical experiences, and health care access factors to cancer staging at diagnosis in Black and White breast cancer patients. METHODS: Data utilized in this population-based cohort study of 1222 eligible women (649 Black and 573 White) newly diagnosed for the period 1985-1986 with histologically confirmed primary breast cancer were obtained from the NCI's Black/White Cancer Survival Study. Sources of data included abstracts of hospital medical records, central review of histology slides by a study consultant pathologist, and patient interviews obtained from three metropolitan areas: Atlanta, New Orleans, and San Francisco-Oakland. Within each area, 70% of all Black incident cases were randomly selected, and a sample of White cases, frequency matched by age groups (20-49 years, 50-64 years, and 65-79 years), was selected for comparison. Stage of breast cancer at diagnosis was classified according to the international tumor-lymph node-metastases (TNM) system. Statistical models utilized in this study included the log-linear and polychotomous logistic regression with multiple predictor variables. RESULTS: Factors associated with cancer staging were differentially expressed in Blacks and Whites. Indicators of access to health care, a lack of mammograms, and an increased body mass index significantly (P < .02) contributed to stage differences in Blacks, whereas income was marginally associated (P = .06) with stage for Whites only. Nuclear grade, having a breast examination by a physician, and a history of patient delay explained approximately 50% of the excess risk for stage III-IV cancer versus stage I-IIN0 cancer among Blacks compared with Whites (odds ratio reduction from 2.19 to 1.68). CONCLUSION: These findings suggest that no single factor or group of factors can explain more than half of the race-stage differences noted in this study with respect to Black and White breast cancer patients.

Computations of adjusted rates and lifetime risks from occupational cohort data: a program package using FORTRAN and GLIM.

A program package using FORTRAN and GLIM is presented to compute lifetime risks of dying from a particular cause of death for a worker subjected to specific risk exposures using death rates adjusted for selected covariates (risk factors). Calculations of the exposure index and adjusted rates depend on several commonly used procedures. Tests of homogeneity and trend for adjusted rates are provided. Lifetime risks are calculated in two different ways: adjusting or ignoring competing causes of death.

Recent developments in the multistage modeling of cohort data for carcinogenic risk assessment.

The modeling of cohort data based on the Armitage-Doll multistage model of the carcinogenic process has gained popular acceptance as a methodology for quantitative risk assessment for estimating the dose-related relationships between different occupational and environmental carcinogenic exposures and cancer mortality. The multistage model can be used for extrapolation to low doses relevant for setting environmental standards and also provides information regarding whether more than one stage is dose-related, which assists in determining whether different carcinogens affect different stages of the cancer process. This paper summarizes recent developments in the multistage modeling of cohort data and emphasizes practical issues such as handling data arising from large epidemiologic follow-up studies, the time-dependent nature of exposures and statistical issues such as multicollinearity in time-related variables, robustness of parameter estimates, validating of the fitted models, and routine inferential procedures. Problems related to uncertainties of risk estimates are discussed also. Computer programs for fitting multistage models with one or two dose-related stages to cohort data incorporating time-dependent exposure patterns; constructing confidence regions for model parameters; and predicting lifetime risks of dying from cancer adjusting for competing causes of death are detailed. Illustrations are provided using lung cancer mortality in a cohort of nonwhite male coke oven workers exposed to coal tar pitch volatiles.

Analysis of clinical, epidemiologic, and laboratory data on chronic fatigue syndrome.

Much of the research conducted on chronic fatigue syndrome (CFS) is exploratory. The researchers' overall goal is to use clinical, epidemiologic, and laboratory data to provide clues about the etiology of this syndrome. In preparation for this symposium, a review of numerous publications on CFS has indicated that the literature generally does not reflect the application of optimal statistical methods for exploration of data. Whenever the researchers' aim is to generate hypotheses, modern methods designed specifically for exploratory data analysis are likely to provide greater insights into any patterns of data than are the traditional approaches to hypothesis testing. In addition, the use of formal methods of data synthesis for ongoing and future research on CFS is a means of strengthening collaborative efforts and of improving the ability of researchers to interpret the evidence available that relates to specific etiologic factors. The inclusion on the research team of experienced biostatisticians, who would oversee the statistical methods and the development of innovative analyses, is recommended.

Multistage modeling of lung cancer mortality among arsenic-exposed copper-smelter workers.

Multistage modeling incorporating a time-dependent exposure pattern is applied to lung cancer mortality data obtained from a cohort of 2802 arsenic-exposed copper-smelter workers who worked 1 or more years during the period 1940-1964 at a copper smelter at Tacoma, Washington. The workers were followed for death through 1976. There were 100 deaths due to lung cancer during the follow-up period. Exposures to air arsenic levels measured in micrograms/m3 were estimated from departmental air arsenic and workers urinary arsenic measurements. Relationships of different temporal variables with excess death rates are examined to judge qualitatively the implications of the multistage cancer process. Analysis to date indicates a late stage effect of arsenic although an additional early stage effect cannot be ruled out.