Search

IMPORTANCE: In the setting of a new cancer diagnosis, the focus is usually on the cancer as the main threat to survival, but people may have other conditions that pose an equal or greater threat to their life than their cancer: a competing risk of death. This is especially true for patients who have cancer of the oral cavity, because prolonged exposure to alcohol and tobacco are risk factors for cancer in this location but also can result in medical conditions with the potential to shorten life expectancy, competing as a cause of death that may intervene in conjunction with or before the cancer. OBSERVATIONS: A calculator designed for public use has been released that allows patients age 20 to 86 years who have a newly diagnosed oral cancer to obtain estimates of their health status–adjusted age, life expectancy in the absence of the cancer, and probability of surviving, dying of the cancer, or dying of other causes within 1 to 10 years after diagnosis. The models in the calculator showed that patients with oral cavity cancer had a higher than average risk of death from other causes than the matched US population, and this risk increases by stage. CONCLUSIONS AND RELEVANCE: The Surveillance, Epidemiology and End Results Program Oral Cancer Survival Calculator supports a holistic approach to the life of the patient, and the risk of death of other causes is treated equally to consideration of the probability of death of the cancer. This tool may be usefully paired with the other available prognostic calculators for oral cancer and is an example of the possibilities now available with registry linkages to partially overlapping or independent data sets and statistical techniques that allow the use of 2 time scales in 1 analysis.

IMPORTANCE: Standard cancer prognosis models typically do not include much specificity in characterizing competing illnesses or general health status when providing prognosis estimates, limiting their utility for individuals, who must consider their cancer in the context of their overall health. This is especially true for patients with oral cancer, who frequently have competing illnesses. OBJECTIVE: To describe a statistical framework and accompanying new publicly available calculator that provides personalized estimates of the probability of a patient surviving or dying from cancer or other causes, using oral cancer as the first data set. DESIGN, SETTING, AND PARTICIPANTS: The models used data from the Surveillance, Epidemiology, and End Results (SEER) 18 registry (2000 to 2011), SEER-Medicare linked files, and the National Health Interview Survey (NHIS) (1986 to 2009). Statistical methods developed to calculate natural life expectancy in the absence of the cancer, cancer-specific survival, and other-cause survival were applied to oral cancer data and internally validated with 10-fold cross-validation. Eligible participants were aged between 20 and 94 years with oral squamous cell carcinoma. EXPOSURES: Histologically confirmed oral cancer, general health status, smoking, and selected serious comorbid conditions. MAIN OUTCOMES AND MEASURES: Probabilities of surviving or dying from the cancer or from other causes, and life expectancy in the absence of the cancer. RESULTS: A total of 22 392 patients with oral squamous cell carcinoma (13 544 male [60.5%]; 1476 Asian and Pacific Islander [6.7%]; 1792 Black [8.0%], 1589 Hispanic [7.2%], 17 300 White [78.1%]) and 402 626 NHIS interviewees were included in this calculator designed for public use for patients ages 20 to 86 years with newly diagnosed oral cancer to obtain estimates of health status–adjusted age, life expectancy in the absence of the cancer, and the probability of surviving, dying from the cancer, or dying from other causes within 1 to 10 years after diagnosis. The models in the calculator estimated that patients with oral cancer have a higher risk of death from other causes than their matched US population, and that this risk increases by stage. CONCLUSIONS AND RELEVANCE: The models developed for the calculator demonstrate that survival estimates that exclude the effects of coexisting conditions can lead to underestimates or overestimates of survival. This new calculator approach will be broadly applicable for developing future prognostic models of cancer and noncancer aspects of a person’s health in other cancers; as registries develop more linkages, available covariates will become broader, strengthening future tools.

Joinpoint regression can model trends in time-specific estimates from aggregated data. These methods have been developed mainly for nonsurvey data such as cancer registry data assuming that the time-specific estimates are uncorrelated from time point to time point. This independence assumption can be violated for trends in time-specific estimates from complex survey samples due to using the same primary sampling units across time and, therefore, the full variance–covariance matrix of the time-specific estimates should be incorporated into the regression model fitting. This article extends these joinpoint methods for analyzing complex survey data within the National Cancer Institute’s Joinpoint software and empirically compares the extended method to existing methods for analyses of time trends in three surveys.

With increased attention to the financing and structure of healthcare, dramatic increases in the cost of diagnosing and treating cancer, and corresponding disparities in access, the study of healthcare economics and delivery has become increasingly important. The Healthcare Delivery Research Program (HDRP) in the Division of Cancer Control and Population Sciences at the National Cancer Institute (NCI) was formed in 2015 to provide a hub for cancer-related healthcare delivery and economics research. However, the roots of this program trace back much farther, at least to the formation of the NCI Division of Cancer Prevention and Control in 1983. The creation of a division focused on understanding and explaining trends in cancer morbidity and mortality was instrumental in setting the direction of cancer-related healthcare delivery and health economics research over the subsequent decades. In this commentary, we provide a brief history of health economics and healthcare delivery research at NCI, describing the organizational structure and highlighting key initiatives developed by the division, and also briefly discuss future directions. HDRP and its predecessors have supported the growth and evolution of these fields through the funding of grants and contracts; the development of data, tools, and other research resources; and thought leadership including stimulation of research on previously understudied topics. As the availability of new data, methods, and computing capacity to evaluate cancer-related healthcare delivery and economics expand, HDRP aims to continue to support this growth and evolution.

IMPORTANCE: The US Preventive Services Task Force (USPSTF) issued its 2021 recommendation on lung cancer screening, which lowered the starting age for screening from 55 to 50 years and the minimum cumulative smoking exposure from 30 to 20 pack-years relative to its 2013 recommendation. Although costs are expected to increase because of the expanded screening eligibility criteria, it is unknown whether the new guidelines for lung cancer screening are cost-effective. OBJECTIVE: To evaluate the cost-effectiveness of the 2021 USPSTF recommendation for lung cancer screening compared with the 2013 recommendation and to explore the cost-effectiveness of 6 alternative screening strategies that maintained a minimum cumulative smoking exposure of 20 pack-years and an ending age for screening of 80 years but varied the starting ages for screening (50 or 55 years) and the number of years since smoking cessation (≤15, ≤20, or ≤25). DESIGN, SETTING, AND PARTICIPANTS: A comparative cost-effectiveness analysis using 4 independently developed microsimulation models that shared common inputs to assess the population-level health benefits and costs of the 2021 recommended screening strategy and 6 alternative screening strategies compared with the 2013 recommended screening strategy. The models simulated a 1960 US birth cohort. Simulated individuals entered the study at age 45 years and were followed up until death or age 90 years, corresponding to a study period from January 1, 2005, to December 31, 2050. EXPOSURES: Low-dose computed tomography in lung cancer screening programs with a minimum cumulative smoking exposure of 20 pack-years. MAIN OUTCOMES AND MEASURES: Incremental cost-effectiveness ratio (ICER) per quality-adjusted life-year (QALY) of the 2021 vs 2013 USPSTF lung cancer screening recommendations as well as 6 alternative screening strategies vs the 2013 USPSTF screening strategy. Strategies with a mean ICER lower than $100 000 per QALY were deemed cost-effective. RESULTS: The 2021 USPSTF recommendation was estimated to be cost-effective compared with the 2013 recommendation, with a mean ICER of $72 564 (range across 4 models, $59 493-$85 837) per QALY gained. The 2021 recommendation was not cost-effective compared with 6 alternative strategies that used the 20 pack-year criterion. Strategies associated with the most cost-effectiveness included those that expanded screening eligibility to include a greater number of former smokers who had not smoked for a longer duration (ie, ≤20 years and ≤25 years since smoking cessation vs ≤15 years since smoking cessation). In particular, the strategy that screened former smokers who quit within the past 25 years and began screening at age 55 years was associated with screening coverage closest to that of the 2021 USPSTF recommendation yet yielded greater cost-effectiveness, with a mean ICER of $66 533 (range across 4 models, $55 693-$80 539). CONCLUSIONS AND RELEVANCE: This economic evaluation found that the 2021 USPSTF recommendation for lung cancer screening was cost-effective; however, alternative screening strategies that maintained a minimum cumulative smoking exposure of 20 pack-years but included individuals who quit smoking within the past 25 years may be more cost-effective and warrant further evaluation.

Analysis of trends in health data collected over time can be affected by instantaneous changes in coding that cause sudden increases/decreases, or “jumps,” in data. Despite these sudden changes, the underlying continuous trends can present valuable information related to the changing risk profile of the population, the introduction of screening, new diagnostic technologies, or other causes. The joinpoint model is a well-established methodology for modeling trends over time using connected linear segments, usually on a logarithmic scale. Joinpoint models that ignore data jumps due to coding changes may produce biased estimates of trends. In this article, we introduce methods to incorporate a sudden discontinuous jump in an otherwise continuous joinpoint model. The size of the jump is either estimated directly (the Joinpoint-Jump model) or estimated using supplementary data (the Joinpoint-Comparability Ratio model). Examples using ICD-9/ICD-10 cause of death coding changes, and coding changes in the staging of cancer illustrate the use of these models.

Analysis of trends in health data collected over time can be affected by instantaneous changes in coding that cause sudden increases/decreases, or “jumps,” in data. Despite these sudden changes, the underlying continuous trends can present valuable information related to the changing risk profile of the population, the introduction of screening, new diagnostic technologies, or other causes. The joinpoint model is a well-established methodology for modeling trends over time using connected linear segments, usually on a logarithmic scale. Joinpoint models that ignore data jumps due to coding changes may produce biased estimates of trends. In this article, we introduce methods to incorporate a sudden discontinuous jump in an otherwise continuous joinpoint model. The size of the jump is either estimated directly (the Joinpoint-Jump model) or estimated using supplementary data (the Joinpoint-Comparability Ratio model). Examples using ICD-9/ICD-10 cause of death coding changes, and coding changes in the staging of cancer illustrate the use of these models.

Objectives. We compared national and state-based estimates for the prevalence of mammography screening from the National Health Interview Survey (NHIS), the Behavioral Risk Factor Surveillance System (BRFSS), and a model-based approach that combines information from the two surveys.Methods. At the state and national levels, we compared the three estimates of prevalence for two time periods (1997–1999 and 2000–2003) and the estimated difference between the periods. We included state-level covariates in the model-based approach through principal components.Results. The national mammography screening prevalence estimate based on the BRFSS was substantially larger than the NHIS estimate for both time periods. This difference may have been due to nonresponse and noncoverage biases, response mode (telephone vs. in-person) differences, or other factors. However, the estimated change between the two periods was similar for the two surveys. Consistent with the model assumptions, the model-based estimates were more similar to the NHIS estimates than to the BRFSS prevalence estimates. The state-level covariates (through the principal components) were shown to be related to the mammography prevalence with the expected positive relationship for socioeconomic status and urbanicity. In addition, several principal components were significantly related to the difference between NHIS and BRFSS telephone prevalence estimates.Conclusions. Model-based estimates, based on information from the two surveys, are useful tools in representing combined information about mammography prevalence estimates from the two surveys. The model-based approach adjusts for the possible nonresponse and noncoverage biases of the telephone survey while using the large BRFSS state sample size to increase precision.

Screening for fecal occult blood can be effective in reducing colorectal cancer mortality only if positive tests are appropriately followed up with complete diagnostic evaluation (ie, colonoscopy or flexible sigmoidoscopy with double contrast barium enema) and treatment.

In this article, we develop the first detailed illustration of the use of a cluster detection method using a spatial scan statistic based on an exponential survival model. We use this approach to study the spatial patterns of survival of patients with stage III or stage IV colorectal cancer or with stage I/II, stage III, or stage IV lung cancer in the State of California and the County of Los Angeles (LA) diagnosed during 1988 through 2002. We present the location of the detected clusters of short survival or long survival and compute nonparametric estimates of survival inside and outside of those detected clusters confirming the survival pattern detected by the spatial scan statistic in both areas. In LA County, we investigate the possible relationship between the cluster locations and race, sex, and histology using nonparametric methods, and we compare socioeconomic factors such as education, employment, income, and health insurance inside and outside of the detected clusters. Finally, we evaluate the effect of related covariates on statistically significant long and short survival clusters detected in LA County using logistic regression models. This article illustrates a new way to understand survival patterns that may point to health disparities in terms of diagnosis and treatment patterns.

Patients with newly diagnosed cancer may request an estimate of their prospects for long‐term survival. Unfortunately, standard estimates of survival may be outdated, because they do not reflect recent advances. The authors present a projection method that incorporates trends in survival and provides more up‐to‐date estimates of long‐term survival for newly diagnosed patients.The projection method fits a regression model to interval relative survival and includes a parameter associated with a trend on diagnosis year. The cumulative relative survival rate (CRS) in a target year is calculated by multiplying the projected interval survival rates for that year. To investigate the predictive ability of the projection approach and to develop model‐selection rules, data from the Surveillance, Epidemiology, and End Results Program and the Connecticut tumor registry were used to recreate data that were available at a particular time in the past, and those data were used to project survival for specified target years.The projection method was better at predicting the survival of recently diagnosed patients than current methods, especially long‐term survival for patients who had disease sites with an increasing and stable trend in survival. The authors predicted that the 15‐year CRS for patients who were diagnosed in 2003 will be 61% for all cancer sites combined, 57% for colorectal cancer, 82% for female breast cancer, 53% for ovarian cancer, and 97% for prostate cancer.Although the projection method was more speculative than other methods that are aligned more closely with current observed data, it offered the possibility of providing improved estimates of long‐term survival for recently diagnosed patients. Caution should be used when applying these methods for cancer sites where there has been a dramatic uptake of screening, e.g., prostate cancer, for which the projected results may be overly optimistic. Cancer 2006. Published 2006 by the American Cancer Society.

Estimates of the probability of developing or dying from cancer, either over a lifetime or over a specified number of years, are useful summary measures of the burden of cancer in a population.The authors used publicly available DevCan software and new, detailed, racial/ethnic data bases that were developed in the Surveillance Research Program of the National Cancer Institute to produce risk estimates for selected major cancers among American Indian/Aleut/Eskimo, black, Chinese, Filipino, native Hawaiian, Japanese, white (total, non‐Hispanic), and Hispanic populations.Japanese and non‐Hispanic white men had the highest lifetime risk for developing cancer (47.94% and 47.41%, respectively), and the American Indian/Eskimo/Aleut population (excluding Alaska) had the lowest lifetime risk among men (24.30%). Among women, white and American Indian/Eskimo/Aleut (in Alaska) populations had higher lifetime risks than Japanese women, whereas American Indian/Eskimo/Aleut (excluding Alaska) women had the lowest risk. The age‐conditional probabilities of developing cancer within the next 10 years among men and women age 60 years and the lifetime probabilities of dying from cancer also were reported by racial/ethnic group.Racial/ethnic disparities in the lifetime risk of cancer may be because of differences in cancer incidence rates, but they also may reflect differential mortality rates from causes other than the cancer of interest. Furthermore, because cross‐sectional incidence and mortality rates are used in calculating the DevCan lifetime risk estimates, results must be interpreted with caution when events, such as the widespread and rapid implementation of a new screening test, are known to have influenced disease rates. Cancer 2006. © 2005 American Cancer Society.

Abstract Background: Breast cancer mortality rates in women have been declining at the same time as breast cancer incidence rates, mammography rates and use of effective adjuvant therapy have been increasing. The objective of this study was to examine population data on breast cancer screening and breast cancer mortality to see if there is any geographic association between mammographic screening and breast cancer mortality reduction in the US, after adjusting for therapy use.

Abstract Objective: This paper presents a methodology for piecing together disparate data sources to obtain a comprehensive model for the use of mammography screening in the US population for the years 1975 – 2000. Methods: Two aspects of mammography usage, the age that a woman receives her first mammography and the interval between subsequent mammograms, are modeled separately. The initial dissemination of mammography is based on cross-sectional self report data from national surveys and the interval length between screening exams is fit using longitudinal mammography registry data. Results: The two aspects of mammography usage are combined to simulate screening histories for individual women that are representative of the US population. Simulated mammography patterns for the years 1994 – 2000 were found to be similar to observed screening patterns from the state level mammography registry for Vermont. Conclusions: The model presented gives insight into screening practices over time and provides an alternative public health measure for screening usage in the US population. The comprehensive description of mammography use from its introduction represents an important first step to understanding the impact of mammography on breast cancer incidence and mortality.

Prevalence usually is defined as the proportion of individuals alive who previously had a diagnosis of the disease, regardless of whether the individuals still are receiving treatment or are cured. The objective of this study was to estimate the proportion of elderly patients with colorectal carcinoma (CRC) in the U.S. that actually were receiving care for their disease as a better quantification of the burden of CRC. The authors used data from the National Cancer Institute's Surveillance, Epidemiology, and End Results (SEER) Program linked to Medicare claims. Four phases of CRC care were defined: initial diagnosis and treatment, postdiagnostic monitoring, treatment for recurrent/metastatic disease, and terminal care. CRC care prevalence measures by phase were extrapolated to the U.S. population age 65 years and older. For all patients with CRC who were diagnosed between 1975 and 1996, 62% received at least 1 service related to their CRC in 1996, and patients received an average of 2.1 months per person of CRC care. Among the U.S. population age 65 years and older, 1.81% had 1 diagnosis of CRC, and (1.81% × 0.62%) = 1.12% received at least 1 service related to their CRC. This translated to 380,783 individuals who received care and 1,210,121 person months of care during 1996. To the authors' knowledge, this is the first report in which care prevalence has been estimated directly. The classification of CRC care by phases of care provides a very detailed picture of the amount of care delivered in the U.S. population. Person-month estimates can be used to estimate the cost of CRC. Cancer 2003;98:1253–61. Published 2003 by the American Cancer Society. DOI 10.1002/cncr.11631

Summary. We propose a new Poisson method to estimate the variance for prevalence estimates obtained by the counting method described by Gail et al. (1999, Biometrics55, 1137–1144) and to construct a confidence interval for the prevalence. We evaluate both the Poisson procedure and the procedure based on the bootstrap proposed by Gail et al. in simulated samples generated by resampling real data. These studies show that both variance estimators usually perform well and yield coverages of confidence intervals at nominal levels. When the number of disease survivors is very small, however, confidence intervals based on the Poisson method have supranominal coverage, whereas those based on the procedure of Gail et al. tend to have below‐nominal coverage. For these reasons, we recommend the Poisson method, which also reduces the computational burden considerably.

The incidence of distant stage prostate carcinoma was relatively flat until 1991 and then started declining rapidly. This decline probably was caused by the shift to earlier stage disease associated with the rapid dissemination of prostate specific antigen (PSA) screening. Prostate carcinoma mortality rates started falling at approximately the same time. In this article, the authors model the potential impact of this stage shift on prostate carcinoma mortality rates given various assumptions concerning the survival of patients with screen-detected local-regional disease. The authors used the CAN*TROL 2 computer model to shift each deficit in the number of patients with distant stage disease to local-regional stage disease and modeled the implications on mortality using a set of base, optimistic, and pessimistic survival assumptions. A base survival assumes that a patient with screen-detected local-regional disease of a certain histologic grade has the same prognosis as a patient with clinically detected local-regional disease of same grade (i.e., an assumption of no length bias for patients with screen-detected disease), whereas the optimistic and pessimistic scenarios assume that survival is better or worse, respectively, than the base survival (i.e., complete cure for patients with favorable grade for the optimistic scenario and no improvements in survival for patients with unfavorable grade for the pessimistic scenario). Model results were compared with observed mortality trends. Rising age-adjusted mortality rates peaked in 1991 for white males and in 1993 for black males and then fell 21% and 13% for white males and black males, respectively, from 1990 through 1999. Under the modeled stage-shift intervention, mortality rates would fall 18%, 8%, and 19% for both white males and black males under the base, pessimistic, and optimistic assumptions, respectively. It is impossible to know what the mortality trends would have been in the absence of the introduction of PSA screening. However, under the base assumption, it appears that the decline in distant stage disease can have a fairly sizable and rapid impact on population mortality. The optimistic scenario is not much improved over the base scenario, which is indicative of the facts that the survival of patients diagnosed with clinical local-regional prostate carcinoma is quite good and that further survival improvements can have only a marginal impact. Under the pessimistic scenario, it appears that something else must be responsible for much of the decline in mortality. Screening trial results from the United States and Europe may verify and isolate the size of any mortality benefit associated with PSA screening. Trial results eventually can be put back into these population models to help quantify the impact of screening, treatment, and other factors on population trends. Cancer 2002;95:870–80. Published 2002 by the American Cancer Society. DOI 10.1002/cncr.10726

Background:Postoperative colon surveillance has been recommended for patients with a diagnosis of local/regional colorectal cancer. The extent to which these recommendations are followed in practice is poorly characterized. Patterns of surveillance after surgery for colorectal cancer were determined by using a large population-based database. Methods:This is a retrospective cohort study with cancer registry data linked to Medicare claims. Identified were 52,283 patients treated for local/regional colorectal cancer between 1986 and 1996, and surveillance patterns through 1998 were determined. Surveillance patterns were analyzed by using survival analysis and by computing the proportion of surviving patients who underwent procedures during 4 time periods after treatment: 2 to 14 months, 15 to 50 months, 51 to 86 months and more than 87 months. Results:Median times to first through fifth surveillance events were 20, 14, 15, 15, and 15 months, respectively. For 17% of the cohort there was no surveillance event. Younger patients were more likely to undergo surveillance. Surveillance patterns were not affected by stage. The proportions of the cohort that underwent no surveillance during the 4 respective time periods were 54%, 52%, 60%, and 69%. The percentages of patients who underwent surveillance annually or more frequently in the latter 3 time periods, respectively, were 19%, 10%, and 5%, or 11% overall, treating the data for the 3 events as a whole. Over the period from 1986 to 1998, the proportion of patients who had no surveillance procedures gradually decreased, whereas the proportion of those who underwent procedures annually or more frequently remained relatively constant. Conclusions:During the period from 1986 to 1998 there was low utilization of postdiagnosis colon surveillance in a substantial proportion of elderly patients with a diagnosis of local/regional colorectal cancer. Over time there was a trend toward increasing receipt of any surveillance procedures. The percentages of patients undergoing surveillance annually or more frequently did not change between earlier and later periods. (Gastrointest Endosc 2001;54:563-71.)

A common population-based cancer progress measure for net survival (survival in the absence of other causes) of cancer patients is relative survival. Relative survival is defined as the ratio of a population of observed survivors in a cohort of cancer patients to the proportion of expected survivors in a comparable set of cancer-free individuals in the general public, thus giving a measure of excess mortality due to cancer. Relative survival was originally designed to address the question of whether or not there is evidence that patients have been cured. It has proven to be a useful survival measure in several areas, including the evaluation of cancer control efforts and the application of cure models. However, it is not representative of the actual survival patterns observed in a cohort of cancer patients. This paper suggests a measure for cumulative crude (in the presence of other causes) cause-specific probability of death for a population diagnosed with cancer. The measure does not use cause of death information which can be unreliable for population cancer registries. Point estimates and variances are derived for crude cause-specific probability of death using relative survival instead of cause of death information. Examples are given for men diagnosed with localized prostate cancer over the age of 70 and women diagnosed with regional breast cancer using Surveillance, Epidemiology and End Results (SEER) Program data. The examples emphasize the differences in crude and net mortality measures and suggest areas where a crude measure is more informative. Estimates of this type are especially important for older patients as new screening modalities detect cancers earlier and choice of treatment or even ‘watchful waiting’ become viable options. Published in 2000 by John Wiley & Sons, Ltd.

A common population‐based cancer progress measure for net survival (survival in the absence of other causes) of cancer patients is relative survival. Relative survival is defined as the ratio of a population of observed survivors in a cohort of cancer patients to the proportion of expected survivors in a comparable set of cancer‐free individuals in the general public, thus giving a measure of excess mortality due to cancer. Relative survival was originally designed to address the question of whether or not there is evidence that patients have been cured. It has proven to be a useful survival measure in several areas, including the evaluation of cancer control efforts and the application of cure models. However, it is not representative of the actual survival patterns observed in a cohort of cancer patients. This paper suggests a measure for cumulative crude (in the presence of other causes) cause‐specific probability of death for a population diagnosed with cancer. The measure does not use cause of death information which can be unreliable for population cancer registries. Point estimates and variances are derived for crude cause‐specific probability of death using relative survival instead of cause of death information. Examples are given for men diagnosed with localized prostate cancer over the age of 70 and women diagnosed with regional breast cancer using Surveillance, Epidemiology and End Results (SEER) Program data. The examples emphasize the differences in crude and net mortality measures and suggest areas where a crude measure is more informative. Estimates of this type are especially important for older patients as new screening modalities detect cancers earlier and choice of treatment or even ‘watchful waiting’ become viable options. Published in 2000 by John Wiley & Sons, Ltd.

A Correction has been published for this article in

A Correction has been published for this article in Statistics in Medicine 2001; 20:655. The identification of changes in the recent trend is an important issue in the analysis of cancer mortality and incidence data. We apply a joinpoint regression model to describe such continuous changes and use the grid-search method to fit the regression function with unknown joinpoints assuming constant variance and uncorrelated errors. We find the number of significant joinpoints by performing several permutation tests, each of which has a correct significance level asymptotically. Each p-value is found using Monte Carlo methods, and the overall asymptotic significance level is maintained through a Bonferroni correction. These tests are extended to the situation with non-constant variance to handle rates with Poisson variation and possibly autocorrelated errors. The performance of these tests are studied via simulations and the tests are applied to U.S. prostate cancer incidence and mortality rates. Copyright © 2000 John Wiley & Sons, Ltd.

Background.Lifetime risk estimates of disease are limited by long-term data extrapolations and are less relevant to individuals who have already lived a period of time without the disease, but are approaching the age at which the disease risk becomes common. In contrast, short-term age-conditional risk estimates, such as the risk of developing a disease in the next 10 years among those alive and free of the disease at a given age, are less restricted by long-term extrapolation of current rates and can present patients with risk information tailored to their age. This study focuses on short-term age-conditional risk estimates for a broad set of important chronic diseases and nondisease causes of death among white and black men and women.

We looked for factors predicting the therapeutic outcome in 66 patients with severe ulcerative colitis treated with intravenous hydrocortisone or corticotropin (ACTH) for 10 days. Patients were randomized before therapy within strata defined by whether they had received oral corticosteroids continuously before the study (group A, 35 patients) or not (group B, 31 patients). Comparisons were made between groups receiving what we considered optimal corticoid therapy, hydrocortisone for group A and ACTH for group B. Overall, therapeutic success was achieved in 28 (42), with a median time of 7.5 days. Favorable factors measured on admission to the study were those suggesting less severe colitis activity: absence of fulminant disease, limited disease extent, a shorter duration of the present attack, fewer stools, a lower eryth-rocyte sedimentation rate (ESR), and a higher hemoglobin. Factors compatible with more severe colitis including fulminant activity, more extensive disease, a shorter total disease duration, bloody stools, and fewer bowel movements, favored an early response among those patients who were to achieve a remission. Prolonging therapy beyond 10 days by switching to the alternate corticoid drug did not improve the remission rate. Achieving remission during the initial therapy period, especially when it occurred early, was the most important predictive factor for a favorable clinical course during the following year. Prolonging therapy did not improve the 1-year remission rate. In fact, a higher proportion of patients who continued to require therapy underwent colectomy than those who received one treatment course.

Random fecal α1-antitrypsin levels, determined in 35 patients with Crohn's disease, showed a strong correlation with clinical activity (Spearman r0.44, p0.01). In 85 of the patients, both clinical and fecal measurements of Crohn's disease activity agreed (κ0.368, p0.011). Similar agreement occurred in those with colitis (κ0.385, p0.035) or ileitis (binomial, p< 0.001) and those with (binomial, p0.006) or without (κ0.492, p0.01) prior surgery. There was a reduction in the mean clinical score and fecal levels among all patient groups after treatment. There was good agreement between both disease activity measurements after medical therapy. However, even after apparent surgical “cure”, fecal protein levels generally remained at values consistent with diffuse occult intestinal disease. When surgical removal of all diseased bowel was not possible, fecal measurements again appeared superior to clinical assessment in reflecting the residual disease. A good correlation existed between the anatomical extent of disease and fecal levels (r0.606, p0.028), in contrast to the relationship between extent and clinical score (r0.14, p0.64). Random fecal α1arantitrypsin determinations provide a measure of the intestinal activity and extent of Crohn's disease. They may be useful in monitoring the response to therapy and the presence of residual disease after surgery.

Summary Mononuclear cells from 12 normal controls (co), 10 advanced untreated (c1), and 6 advanced treated cancer patients (c2) have been isolated. The numbers of mononuclear cells bearing Leu1, Leu2, Leu3, Leu2/HLA-DR and LeuM3 were measured with a fluorescence-activated cell sorter. Only the quantity of helper T cells (Leu3) was decreased in cancer patients (co: 0.89, c1: 0.32, c2: 0.44 × 109/1). Expression of all other markers, including activated suppressor T cells (Leu2/HLA-DR), did not differ significantly from the control. The proliferation of the lymphocytes was determined in a phytohemagglutininculture assay. The cancer groups showed a significantly decreased response (co: 95.8 × 109, cl: 28.7 × 109, c2: 25.7 × 109 cpm). These values correlated with the number of helper T cells but not with the suppressor T cells. Monocytes of cancer patients adsorbed significantly more immunoglobulins than the monocytes of controls. The addition of indomethacin or isoprinosine to phytohemagglutinin-culture assay increased the proliferation of lymphocytes from both the cancer patients and normal controls.

We offer an approximation to central confidence intervals for directly standardized rates, where we assume that the rates are distributed as a weighted sum of independent Poisson random variables. Like a recent method proposed by Dobson, Kuulasmaa, Eberle and Scherer, our method gives exact intervals whenever the standard population is proportional to the study population. In cases where the two populations differ non-proportionally, we show through simulation that our method is conservative while other methods (the Dobson et al. method and the approximate bootstrap confidence method) can be liberal. © 1997 by John Wiley & Sons, Ltd.

We introduce a method for estimating overdispersion in Poisson models for vital rates. We assume smoothness conditions on the counts to obtain pointwise variance estimates that we combine to obtain an estimate of the overdispersion parameter. We create confidence intervals about the observed rates using this estimate and an approximation based on the gamma distribution. The advantage of this method is that the estimates of the superpopulation rates do not depend on the smoothness assumption, yet when this assumption is met we obtain approximately unbiased estimates of the overdispersion parameter. Thus, we may calculate confidence intervals for vital rates under an overdispersed Poisson model without making parametric assumptions on the mean rates. © 1997 John Wiley & Sons, Ltd.

Triage of grant application at the National Institutes of Health (NIH) is a process whereby an initial screening of applications by a scientific peer review group eliminates applications that are not competitive for awards. The process of application triage has been limited to those applications submitted to the NIH in response to an RFA (Request for Applications). A hypergeometric model was developed to determine the extent to which five, six, seven, or eight member triage teams or subsets of 12‐to‐20 member full committees could provide a statistically defensible triage decision. Although the intent of triage is to remove from review those applications that are noncompetitive, the model was weighted in favor of the applicant to minimize the likelihood that highly competitive applications would be eliminated. Within the assumptions and rules developed, it was determined that there was little likelihood that the latter would occur. For example, in the worst case scenario, the greatest probability that a highly competitive application would be knocked out of competition is P≤ 0.014 in the case of a five‐member triage subset of a 20‐member committee. Using the latter case, the model was tested on a set of 73 applications that were submitted to the National Cancer Institute for action at the February 1993 National Cancer Advisory Board. The model selected for triage required that each application be assigned to five reviewers, that each reviewer be blinded to the review assignments of the other reviewers, and that four noncompetitive votes be registered to triage out an application. Each of 19 applications received four to five noncompetitive votes, and were triaged out of the review process. The remaining 54 applications were then reviewed according to the usual NIH review process. Four of the applications received three noncompetitive triage votes each and were either rated as not recommended for further consideration (NRF, n= 2)) or received priority scores e 250 (n= 2) (The smaller the priority score the better the technical merit). Thirteen of the 53applications received two noncompetitive votes. Of the latter, two were not recommended for further consideration and the remaining 11 received priority scores between in excess of 200. The distribution of competitive applications was such that funding was limited to those applications with priority scores of less than 190. Thus, the data suggest that the conservative model is valid such that the likelihood of eliminating a highly competitive application from consideration for funding is remotely small. With this model, the process of triage is fair to applicants on the one hand and is also effective in reducing consultant workloads on the other. The model could be applied to many different types of review situations (private sponsors as well as federal), especially when few awards are to be issued relative to the number of competing applications.—Vener, K. J., Feuer, E. J., Gorelic, L. A statistical model validating triage for the peer review process: keeping the competitive applications in the review pipeline. FASEB J.7: 1312‐1319; 1993.

There is concern that financial incentives in health maintenance organizations (HMOs) might result in pressures to induce sicker members to disenroll. The authors compared disenrollment rates of Medicare HMO enrollees with cancer with disenrollment rates for cancer-free enrollees, using Medicare enrollment files linked to population-based tumor registry data from the Surveillance, Epidemiology, and End Results (SEER) Program.

Breast cancer incidence in the United States has been rising dramatically since 1982, as shown in data collected by the Surveillance, Epidemiology, and End Results (SEER) Program. In women aged 50 and older, incidence rates for in situ and localized invasive tumors have increased over the period 1982–86, while rates for regional and distant tumors have remained stable. The incidence of localized tumors <1.0 cm, 1.0–1.9 cm, and 2.0–2.9 cm in size has increased more rapidly than that of tumors 3.0 cm or more in size among women over age 50. Survival rates also have improved among cases diagnosed over this time period. These data suggest that early detection may be playing a role in the recent increase in female breast cancer incidence, though other factors cannot be ruled out. Conclusions regarding improved cancer control await confirmation by reduced breast cancer mortality.

Previous estimates of the lifetime risk of developing breast cancer have used cross-sectional estimates of incidence. Cross-sectional rates, however, yield a biased picture of cohort risks when rates are unstable, as breast cancer trends have been. We developed cohort life tables for Connecticut women born from 1888–1892 to 1948–1952 to generate more specific estimates of breast cancer risk to age 85. Multiple decrement life tables were produced for each birth cohort. We included as cases only the first reports of breast cancer in women with no earlier malignancy. Our results indicate that widely circulated lifetime risks of 1 in 9 may be inflated slightly owing to changing incidence. We estimate that of those women 40–44 years old in 1992, 1 woman in 10 will develop breast cancer by age 85. For women born between 1928 and 1932, 1 in 13 will be diagnosed with breast cancer by age 85. The results are insensitive to mortality trends in the past. Errors in the estimates are more likely to arise from changes in incidence and mortality in the future.

AbstractObjectives:To characterize and enumerate central venous catheter (CVC)-related complications among children with chronic illnesses, and to reduce the complication rate through changes in CVC management and education.Design:A prospective obser vational study followed by an educational program and a nonrandomized inter ventional trial.Setting:The Children's Hospital of Philadelphia, a tertiary, pediatric facility.Patients:268 children with Broviac, Hickman, or Infusaport catheters in place during 58,290 catheter days.Interventions:Development and implementation of protocols for cleaning insertion site and hub, use of nonocclusive dressings, and manipulation of access; formal staff and parental education about protocols.Results:CVC-related infections fell from 4.58/1,000 catheter-days preinter vention to 3.83 postintervention (risk ratio [RR], 0.20; 95% confidence interval [CI95], 0.89-1.622; P=.25); exit-site infections fell from 0.58 to 0.11 (CI95, 1.22-45.64; P=.02); rates among infants on the surgical service fell from 15.46 to 6.67 (RR, 2.31; CI95, 1.10-4.30; P=.02).Conclusions:Education and changes in management protocols reduced the incidence of exit-site infections among all patients and reduced the overall infectious complication rate among the infants receiving parenteral nutrition on the surgical ser vice. Other interventions are needed to decrease further the infectious complications in these children

Random fecal α1-antitrypsin levels were determined in 34 patients, 24 with Crohn's disease, and 10 “controls” having diarrheal illnesses not associated with intestinal protein loss, in an effort to evaluate its usefulness as a measure of Crohn's disease intestinal activity. In the control group, all α1-antitrypsin levels were <2 mg/g dry wt of stool. The mean fecal level among those with Crohn's disease was 52.9 mg/ g (range <2 to ≥200). There was a strong correlation between disease activity, as measured by a clinical score, and the α1-antitrypsin levels (Spearman r = 0.65, p = 0.001). This correlation was similarly strong among those with colitis or ileitis. A fecal value >20 mg/g may provide a rough guideline to separate patients with clinically active disease from those with inactive Crohn's disease, despite a considerable range of fecal levels among patients with a particular clinical score. Fecal α1-antitrypsin levels correlated with several other laboratory measures that have been proposed as indicators of Crohn's disease activity. The serum orosomucoid, C-reactive protein, and albumin correlated with the clinical activity score among some of our patient groups. Both clinical scores and laboratory parameters, however, may have limited usefulness in a variety of circumstances. Random fecal α1-antitrypsin determinations seem to provide a reliable, although not directly quantitative, measure of the intestinal activity among patients with Crohn's disease, especially when other methods may be inconclusive.

AbstractObjectives:To characterize and enumerate central venous catheter (CVC)-related complications among children with chronic illnesses, and to reduce the complication rate through changes in CVC management and education.Design:A prospective obser vational study followed by an educational program and a nonrandomized inter ventional trial.Setting:The Children's Hospital of Philadelphia, a tertiary, pediatric facility.Patients:268 children with Broviac, Hickman, or Infusaport catheters in place during 58,290 catheter days.Interventions:Development and implementation of protocols for cleaning insertion site and hub, use of nonocclusive dressings, and manipulation of access; formal staff and parental education about protocols.Results:CVC-related infections fell from 4.58/1,000 catheter-days preinter vention to 3.83 postintervention (risk ratio [RR], 0.20; 95% confidence interval [CI95], 0.89-1.622; P=.25); exit-site infections fell from 0.58 to 0.11 (CI95, 1.22-45.64; P=.02); rates among infants on the surgical service fell from 15.46 to 6.67 (RR, 2.31; CI95, 1.10-4.30; P=.02).Conclusions:Education and changes in management protocols reduced the incidence of exit-site infections among all patients and reduced the overall infectious complication rate among the infants receiving parenteral nutrition on the surgical ser vice. Other interventions are needed to decrease further the infectious complications in these children

A risk-adjusted method is proposed for estimating cancer incidence rates from data collected by the Surveillance, Epidemiology, and End Results (SEER) program of the US National Cancer Institute. Unlike the conventional incidence-rate estimates reported by SEER, this method considers only the first primary cancer, and adjusts for population-based cancer prevalence, as well as for surgical procedures which remove an individual from risk of developing a given cancer. Thus, risk-adjusted incidence-rates more accurately reflect the average cancer risk for individuals in the cancer-free, at-risk population. The results of the analysis indicate that, in general, incidence-rate estimates are fairly similar between the conventional and risk-adjusted methods. However, this is not the case for certain cancer sites which may have: (i) a high number of subsequent primary cancers (e.g., melanomas—skin); (ii) a high cancer prevalence proportion (e.g., prostate cancer); (iii) a high number of subsequent primary cancers and prevalence proportion (e.g., female breast); or (iv) a high prevalence of removal of the organ in question (e.g., cervical and uterine cancers). For example, by applying the risk-adjusted incidence method for in situ and invasive cervical cancer, we found that in the period 1990–92 the age-adjusted incidence-rate estimate increased from 57.8 to 66.3 (15 percent) per 100,000 person-years; the greatest increase in the incidence-rate estimate occurred for women aged 65 to 69 years, from 40.1 to 63.2 (58 percent) per 100,000 person-years; and the lifetime risk of developing cervical cancer increased from 4.6 (1 in 22) to 5.5 (1 in 18) percent.

A number of studies have attributed much of the sharp increase in breast cancer incidence in the United States during the 1980s to the increased detection through mammography. The most recent breast cancer data from the US National Cancer Institute's Surveillance, Epidemiology, and End Results (SEER) program show that the incidence trend has slowed, while results from the National Health Interview Survey (NHIS) of 1987 and 1990 indicate that the percentage of women receiving mammograms continues to increase. This phenomenon suggested the need to reassess the relationship between increasingly early detection of breast cancer and overall incidence trends. A polynomial age-cohort model was used to establish the secular trend in incidence rates excluding the impact of recent increases in detection due to the rising use of mammography. Based on the model, the incidence trend in the youngest age group (40 to 49 years) would peak and then begin to decline in the early 1980s. This pattern would manifest itself later in successively older age groups as these younger cohorts age. Breast cancer trends are seen to be generally consistent with the impact of the increased use of mammography when its effect is superimposed upon the background of declining or slowing secular trends. These results support previous reports linking incidence rates with the increase in screening-mammography.

We introduce a method for estimating overdispersion in Poisson models for vital rates. We assume smoothness conditions on the counts to obtain pointwise variance estimates that we combine to obtain an estimate of the overdispersion parameter. We create confidence intervals about the observed rates using this estimate and an approximation based on the gamma distribution. The advantage of this method is that the estimates of the superpopulation rates do not depend on the smoothness assumption, yet when this assumption is met we obtain approximately unbiased estimates of the overdispersion parameter. Thus, we may calculate confidence intervals for vital rates under an overdispersed Poisson model without making parametric assumptions on the mean rates. © 1997 John Wiley & Sons, Ltd.