Search

PURPOSE: Oncology electronic health record (EHR) databases have increased in quality and availability over the past decade, yet it remains unclear whether these clinical practice data can be used to conduct reliable comparative effectiveness studies. We sought to emulate a clinical trial with EHR data in the advanced breast cancer population and compare our results against the trial. METHODS: This cohort study used EHR data from US oncology practices. All elements of the study were defined to mimic the PALOMA-2 trial as closely as possible. Patients with hormone-positive, HER-2 negative metastatic breast cancer with no prior treatment for metastatic disease were included. Patients initiating palbociclib and letrozole on the same day following the earliest record of metastasis were compared to those initiating letrozole only. The primary associational measure was the conditional hazard ratio for time-to-next treatment (TTNT). TTNT is well-measured in our data source and amenable for calibration against the randomized study results of the PALOMA-2 trial. We used multiple imputation for several patient characteristics with missing values. RESULTS: There were 3836 study-eligible women with advanced breast cancer. The hazard ratio for TTNT in the observational study (HR: 0.62; 95% CI: 0.56–0.68) was closely aligned with that of the randomized trial (HR: 0.64; 95% CI: 0.52–0.78). CONCLUSIONS: Under our assumptions on missing data and comparability of the two study populations, results from our non-randomized study closely matched that of the randomized trial. Further studies are needed to determine whether EHR data can yield reliable conclusions on treatment effects in oncology.

Background Racial and ethnic minority status, structural racism, low educational attainment, and poverty are consistently associated with cancer disparities and with higher rates of incarceration. The objective of this scoping review is to conduct a qualitative synthesis of the literature on cancer prevalence, incidence, mortality, and disparities in these outcomes for incarcerated and formerly incarcerated patients, as this literature is fragmented and heterogenous. Methods This scoping review included Bureau of Justice Statistics reports and searched PubMed in May 2021 for all English language studies published between 1990 and 30 April 2021, that reported on cancer prevalence, incidence, or mortality for incarcerated or formerly incarcerated individuals in the United States. Results Twenty studies were selected. Data on cancer prevalence and incidence were scarce but suggested that incarcerated and formerly incarcerated patients have a similar overall risk of cancer diagnosis as the general population, but elevated risk of certain cancers such as cervical, lung, colorectal, and hepatocellular carcinoma for which effective prevention and screening interventions exist. Cancer mortality data in state and local jails as well as prisons were robust and suggests that both incarcerated and formerly incarcerated patients have higher cancer mortality than the general population. Conclusions Incarcerated and formerly incarcerated patients likely have a higher risk of dying from cancer than the general population, but important gaps in our knowledge about the extent and drivers of disparities for this population remain. Additional research is needed to guide interventions to reduce cancer disparities for patients experiencing incarceration., This scoping review finds that individuals who are currently or formerly incarcerated in prison have higher cancer mortality than the general population. However, important knowledge gaps remain that must be addressed to inform targeted interventions to reduce cancer disparities in prisons and in the community.

BACKGROUND Smoking, the most common risk factor for bladder cancer (BC), is associated with increased complications after radical cystectomy (RC), poorer oncologic outcomes, and higher mortality. The authors hypothesized that the effect of smoking on the probability of major complications increases with increasing age among patients who undergo RC. METHODS The authors analyzed the American College of Surgeons National Surgical Quality Improvement Program database (2011-2017), identified all patients undergoing RC using Current Procedural Terminology codes, and formed two groups according to smoking status (active smoker and nonsmoker [included former and never-smokers]). Patient characteristics and 30-day postoperative complications using the Clavien-Dindo Classification (CDC) were assessed. A multivariable logistic regression model was constructed that included age, sex, race, body mass index, operative time, comorbidities, chemotherapy status, and type of diversion with major complications (CDC ≥III) as the outcome variable, and explored the interaction between age and smoking status. RESULTS A total of 10,528 patients underwent RC, including 22.8% who were active smokers. The authors identified an interaction between age and smoking status (P = .045). Older patients were found to experience a stronger smoking effect than younger patients with regard to the probability of major complications. The risk of a major complication was the same for 50-year-old nonsmokers and smokers, but it increased from 17.8% to 21.7% for 70-year-old nonsmokers and smokers, respectively (P < .001). CONCLUSIONS Up to 20% of patients who undergo RC are active smokers, and these individuals have an increased risk of major complications. The effect of smoking is stronger with increasing age; the difference with regard to complications for smokers versus nonsmokers was found to increase substantially, wherein older smokers are at an especially high risk of complications.

Background Historically, older patients with advanced lung cancer have often received no systemic treatment. Immunotherapy has improved outcomes in clinical trials, but its dissemination and implementation at the population level is not well‐understood. Methods A retrospective cohort study of patients with stage IV non‐small cell lung cancer (NSCLC) diagnosed age 66 or older from 2012 to 2015 was conducted using SEER‐Medicare. Treatment patterns within one year of diagnosis were ascertained. Outcomes included delivery of (a) any systemic therapy; (b) any second‐line infusional therapy, following first‐line infusional therapy; and (c) any second‐line immunotherapy, following first‐line infusional therapy. Trends in care patterns associated with second‐line immunotherapy approvals in 2015 were assessed using generalized additive models. Sociodemographic and clinical predictors of treatment were explored using logistic regression. Results Among 10 303 patients, 5173 (50.2%) received first‐line systemic therapy, with little change between the years 2012 (47.5%) and 2015 (50.3%). Among 3943 patients completing first‐line infusional therapy, the proportion starting second‐line infusional treatment remained stable from 2012 (30.5%) through 2014 (32.9%), before increasing in 2015 (42.4%) concurrent with second‐line immunotherapy approvals. Factors associated with decreased utilization of any therapy included age, black race, Medicaid eligibility, residence in a high‐poverty area, nonadenocarcinoma histology, and comorbidity; factors associated with increased utilization of any therapy included Asian race and Hispanic ethnicity. Among patients who received first‐line infusional therapy, factors associated with decreased utilization of second‐line infusional therapy included age, Medicaid eligibility, nonadenocarcinoma histology, and comorbidity; Asian race was associated with increased utilization of second‐line infusional therapy. Conclusion United States Food and Drug Administration (FDA) approvals of immunotherapy for the second‐line treatment of advanced NSCLC in 2015 were associated with increased rates of any second‐line treatment, but disparities based on social determinants of health persisted., United States Food and Drug Administration (FDA) approvals of immunotherapy for the second‐line treatment of advanced non‐small cell lung cancer in 2015 were associated with increased rates of any second‐line treatment in the Medicare population. However, disparities in immunotherapy uptake based on social determinants of health persisted, driven largely by differences in rates of any first‐line treatment.

OBJECTIVE To examine the incidence of perioperative complications after radical cystectomy (RC) and assess their impact on 90-day postoperative mortality during the index stay and upon readmission. PATIENTS AND METHODS A total of 57 553 patients with bladder cancer (unweighted cohort: 9137 patients) treated with RC, at 360 hospitals in the USA between 2005 and 2013 within the Premier Healthcare Database, were used for analysis. The 90-day perioperative mortality was the primary outcome. Multivariable regression was used to predict the probability of mortality; models were adjusted for patient, hospital, and surgical characteristics. RESULTS An increase in the number of complications resulted in an increasing predicted probability of mortality, with a precipitous increase if patients had four or more complications compared to one complication during hospitalisation following RC (index stay; 1.0-9.7%, P

Objective To compare alternative strategies for specifying cancer-free control cohorts for estimating cancer-attributable costs of care. Data Source, Study Design, Data Extraction Secondary data analysis of Surveillance, Epidemiology, and End Results data linked to Medicare claims among patients diagnosed with colorectal, lung, breast, and prostate cancers, 2007–2011. We estimated cancer-attributable costs using three alternative reference cohorts: (1) noncancer Medicare patients individually matched by demographic characteristics, (2) noncancer patients individually matched on demographic factors and comorbidity score, (3) cancer patients as their own control, using prediagnosis costs. Principal Findings Among 44,266 colorectal, 61,584 lung, 55,921 breast, and 67,733 prostate patients, mean total Medicare spending in the first year of diagnosis was $59,496, $54,261, $31,895, and $26,305, respectively. Estimates of cancer-attributable costs ranged from 79 percent to 82 percent of spending for colorectal, 76 percent–79 percent for lung, 65 percent–74 percent for breast, and 60 percent–75 percent for prostate cancers, depending on the reference cohort used. For all cancers, estimates were higher when patients were used as their own control, compared to demographic and comorbidity-matched controls. Conclusions Choice of reference group can have a substantial impact on proportion of total costs attributed to cancer and should be clearly defined in analyses of the costs of cancer care.

Objective To characterize spending patterns for Medicare patients with incident breast, prostate, lung, and colorectal cancer. Data Sources/Study Setting/Study Design 2007–2012 data from the Surveillance, Epidemiology, and End Results Program linked with Medicare fee-for-service claims. Data Collection/Extraction Methods We calculate per-patient monthly and yearly mean and median expenditures, by cancer type, stage at diagnosis, and spending category, over the years of diagnosis and death. Principal Findings Over the year of diagnosis, mean spending was $35,849, $26,295, $55,597, and $63,063 for breast, prostate, lung, and colorectal cancer, respectively. Over the year of death, spending was similar across different cancer types and stage at diagnosis. Conclusions Characterization of Medicare spending according to clinically meaningful categories may assist development of oncology alternative payment models and cost-effectiveness models.

Background Breast cancer 2 (BRCA2)-associated breast and ovarian cancers are sensitive to platinum-based chemotherapy. It is unknown whether BRCA2-associated prostate cancer responds favorably to such treatment. Methods A retrospective analysis of a single-institution cohort of men with castration-resistant, metastatic prostate cancer was performed to determine the association between carrier status of pathogenic BRCA2 germline variants and prostate-specific antigen response to carboplatin-based chemotherapy. From 2001 through 2015, 8081 adult men with prostate cancer who had a consultation and/or underwent treatment at Dana-Farber Cancer Institute provided blood samples and consented to analyses of biologic material and clinical records. A subgroup of 141 men received at least 2 doses of carboplatin and docetaxel for castration-resistant disease (94% were also taxane refractory). These patients were categorized according to the absence or presence of pathogenic germline mutations in BRCA2 based on DNA sequencing from whole blood. The primary outcome was the response rate to carboplatin/docetaxel chemotherapy, defined according to a decline in prostate-specific antigen that exceeded 50% within 12 weeks of initiating this regimen. Associations between BRCA2 mutation status and response to carboplatin-based chemotherapy were tested using the Fisher exact test, with a 2-sided P value Results Pathogenic germline BRCA2 variants were observed in 8 of 141 men (5.7%; 95% confidence interval, 2.5%-10.9%). Six of 8 BRCA2 carriers (75%) experienced prostate-specific antigen declines >50% within 12 weeks, compared with 23 of 133 noncarriers (17%; absolute difference, 58%; 95% confidence interval, 27%-88%; P Conclusions BRCA2-associated, castration-resistant prostate cancer is associated with a higher likelihood of response to carboplatin-based chemotherapy than non-BRCA2-associated prostate cancer. Cancer 2017;123:3532-9. © 2017 American Cancer Society.

Rationale, aims and objectives Frequent follow-up is recommended for the more than 3 million breast cancer survivors living in the USA. Given the multidisciplinary nature of breast cancer treatment, follow-up may be provided by medical oncologists, radiation oncologists, surgeons and primary care providers. This creates the potential for significant redundancy as well as gaps in care. The objective was to examine patterns of breast cancer follow-up provided by different types of oncologists and develop a statistical means of quantifying visit distribution over time. Methods We identified stage I–III breast cancer patients who underwent breast conservation from 2000 to 2006 (n = 12 139) within the SEER-Medicare linked database. Provider type was defined using Medicare specialty provider codes and AMA Masterfile. The coefficient of variation (CV) for time between oncologist follow-up visits was calculated. Ordinal logistic regression assessed factors associated with CV. Results Substantial variation in CV was observed. Sixty-seven per cent of patients with low CV (high visit regularity) received follow-up from a single oncologist type, versus 8% with high CV (low visit regularity). The number of oncologist types participating in follow-up had the greatest association with high CV (odds ratio 7.4 [6.7–8.3] and 15.4 [13.6–17.6] for two and three oncologist types). Conclusions Using a novel means of quantifying follow-up visit regularity, we determined that breast cancer patients with dispersed follow-up with more than one oncologist have more disordered care. The CV could be used in electronic medical records to identify cancer survivors with more disordered.

BACKGROUND Racial/ethnic and socioeconomic disparities persist in part because our current understanding of the care provided to minority and disadvantaged populations is limited. The authors evaluated the quality of breast cancer care in 2 large states to understand the disparities experienced by African Americans, Hispanics, Asian/Pacific Islanders (APIs), and Medicaid enrollees and to prioritize remediation strategies. METHODS Statewide cancer registry data for 80,436 women in New York and 121,233 women in California who were diagnosed during 2004 to 2009 with stage 0 through III breast cancer were used to assess underuse and overuse of surgery, radiation, chemotherapy, and hormone therapy based on 34 quality measures. Concordance values were compared across racial/ethnic and Medicaid-enrollment groups. Multivariable models were used to quantify disparities across groups for each treatment in each state. RESULTS Overall concordance was 76% for underuse measures and 87% for overuse measures. The proportions of patients who received care concordant with all relevant measures were 35% in New York and 33% in California. Compared with whites, African Americans were less likely to receive recommended surgery, radiation, and hormone therapy; Hispanics and APIs were usually more likely to receive recommended chemotherapy. Across states, the same racial/ethnic groups did not always experience the same disparities. Medicaid enrollment was associated with decreased likelihood of receiving all recommended treatments, except chemotherapy, in both states. Overuse was evident for hormone therapy and axillary surgery but was not associated with race/ethnicity or Medicaid enrollment. CONCLUSIONS Patient-level measures of quality identify substantial problems with care quality and meaningful disparities. Remediating these problems will require prioritizing low-performing measures and targeting high-risk populations, possibly in different ways for different regions. Cancer 2015. © 2015 American Cancer Society.

BACKGROUND Pancreatic cancer is a major cause of morbidity and mortality in the Medicare population. Whether the health care burden of pancreatic cancer has changed over the last decade is unknown. METHODS The authors used Medicare data from 2000 to 2010 to identify beneficiaries aged ≥ 65 years who were hospitalized for the management of pancreatic cancer. Annual trends were estimated for the age-sex-race–adjusted initial hospitalization rate, the age-sex-race-comorbidity–adjusted 1-year mortality rate after initial hospitalization, age-sex-race-comorbidity–adjusted procedure rates, 1-year all-cause rehospitalizations after initial pancreatic cancer hospitalization, and mean inflation-adjusted Medicare payment for initial hospitalization. RESULTS A total of 130,728 patients had ≥ 1 hospitalizations for pancreatic cancer and were identified from 56,642,071 beneficiaries during the study period. The age-sex-race–adjusted rate of initial hospitalization for pancreatic cancer was 50 per 100,000 person-years in 2010, representing a 0.5% annual increase since 2000 (95% confidence interval [95% CI], 0.3%-0.7%). In the same period, the age-sex-race-comorbidity–adjusted 1-year mortality rate decreased by 4.4% (95% CI, 3.9%-4.9%), and the age-sex-race-comorbidity–adjusted surgical resection rate increased by 6.9% (95% CI, 6.4%-7.5%). The mean inflation-adjusted Medicare payment for the initial hospitalization decreased, from $14,118 in 2000 to $13,318 in 2010, and the number of 1-year all-cause rehospitalizations after the initial hospitalization increased from 0.75 per patient in 2000 to 0.82 per patient in 2009 (all P

BACKGROUND Rates of screening mammography have plateaued, and the number of mammography facilities has declined in the past decade. The objective of this study was to assess changes over time and geographic disparities in the availability of mammography services. METHODS Using information from the US Food and Drug Administration and the US Census, county-level mammography capacity was defined as the number of mammography machines per 10,000 women aged ≥ 40 years. Cross-sectional variation and longitudinal changes in capacity were examined in relation to county characteristics. RESULTS Between 2000 and 2010, the number of mammography facilities declined 10% from 9434 to 8469, the number of mammography machines declined 10% from 13,100 to 11,762, and the median county mammography capacity decreased nearly 20% from 1.77 to 1.42 machines per 10,000 women aged ≥ 40 years. In cross-sectional analysis, counties with greater percentages of uninsured residents, less educated residents, greater population density, and higher managed care penetration had lower mammography capacity. Conversely, counties with more hospital beds per 100,000 population had higher capacity. High initial mammography capacity, growth in both the percentage of the population aged ≥ 65 years and the percentage living in poverty, and increased managed care penetration were all associated with a decrease in mammography capacity between 2000 and 2010. Only the percentage of rural residents was associated with an increase in capacity. CONCLUSIONS Geographic variation in mammography capacity and declines in capacity over time are associated with demographic, socioeconomic, and health care market characteristics. Maldistribution of mammography resources may explain geographic disparities in breast cancer screening rates. Cancer 2013;119:3847–3853. © 2013 American Cancer Society.

BACKGROUND Randomized trials report equivalent efficacy among various combinations of platinum-based regimens in advanced non–small cell lung cancer (NSCLC). Their relative effectiveness and comparability based on squamous versus nonsquamous histology is uncertain. METHODS The authors used the Surveillance, Epidemiology, and End Results (SEER)-Medicare linked data to identify first-line chemotherapy agents administered to Medicare beneficiaries with stage IIIB or IV NSCLC diagnosed from 2000 to 2007. Overall survival was compared between patients who received the 3 most common regimens: carboplatin-paclitaxel, carboplatin-gemcitabine, and carboplatin-docetaxel. Stratified analyses distinguished between the outcomes of patients with squamous versus nonsquamous cell histology. Multivariable Cox proportional hazards models and propensity score analyses facilitated adjustment for imbalance in measurable patient characteristics. RESULTS Of the 15,318 patients who received first-line chemotherapy, 43.1% received carboplatin-paclitaxel, 14.3% received carboplatin-gemcitabine, 8.5% received carboplatin-docetaxel, and 34.1% received other regimens. The median survival was 8.0 months (interquartile range [IQR], 3.5-17.4 months) for carboplatin-paclitaxel, 7.3 months (IQR, 3.4-15.2 months) for carboplatin-gemcitabine, and 7.5 months (IQR, 3.2-16.0 months) for carboplatin-docetaxel. Both multivariable and propensity score-adjusted Cox models demonstrated a slight inferiority associated with carboplatin-gemcitabine or carboplatin-docetaxel versus carboplatin-paclitaxel, with a hazard ratio of 1.10 (95% confidence interval, 1.04-1.15) and 1.09 (95% confidence interval, 1.02-1.16), respectively, in propensity score-stratified models. Among the subgroup of 2063 patients with squamous carcinoma, propensity score-stratified analyses had a higher risk of death (hazard ratio, 1.20; 95% confidence interval, 1.07-1.35) associated with carboplatin-gemcitabine versus carboplatin-paclitaxel. CONCLUSIONS Carboplatin-paclitaxel was associated with slightly better survival compared with carboplatin-gemcitabine or carboplatin-docetaxel within the Medicare population with advanced NSCLC, and this was most pronounced for patients who had squamous cell histology. Cancer 2013;119:2048–2060. © 2013 American Cancer Society.

BACKGROUND: In clinical trials, combined 5-fluorouracil (5FU) plus oxaliplatin improves the survival of patients who have resected, stage III colon cancer with manageable toxicity. However, the tolerability of this in the general population of patients with colon cancer is uncertain. METHODS: Adverse outcomes were compared in patients with stage III colon cancer who received either 5FU or 5FU/oxaliplatin within 120 days of undergoing resection versus a control group of patients with stage II colon cancer who did not receive chemotherapy in the Surveillance, Epidemiology, and End Results (SEER)-Medicare database and in data from the New York State Cancer Registry linked to Medicare and Medicaid. Hospitalizations, emergency room (ER) visits, and outpatient adverse events (AEs) were measured in claims from 30 days to 9 months after patients underwent resection. Multiple logistic regression was used to calculate adjusted odds ratios of events by treatment. Propensity score matching was used to minimize selection bias. RESULTS: Adverse outcomes were more frequent for chemotherapy recipients. AE rates were higher in patients who received 5FU/oxaliplatin (81%) compared with patients who received 5FU alone (72%), in the SEER-Medicare data. The effect of oxaliplatin on AEs was greater in older patients: The odds ratio was 2.10 (95% confidence interval, 1.53-2.87) for patients aged ≥75 years versus 1.75 (95% confidence interval, 1.39-2.21) for patients aged

This paper describes a linkage between the New York State Cancer Registry (NYSCR) and the New York State (NYS) Medicaid program, both housed within the New York State Department of Health (NYSDOH). Medicaid, the state-based health program for individuals and families with low incomes and resources in the United States, insures approximately one-sixth of New York adults aged 18–64 and one-fifth of elderly over 65, most of whom are dually enrolled in Medicare and Medicaid. Deficiencies in cancer care are known to disproportionately affect the poor as well as racial and ethnic minorities who are overrepresented in Medicaid (Agency for Healthcare Research and Quality 2005; Institute of Medicine 2005; Landon et al. 2007;). The Surveillance, Epidemiology and End Results (SEER)–Medicare linkage is the prototype for our effort (Warren et al. 2002; National Cancer Institute, Division of Cancer Control and Population Sciences, Health Services and Economics Branch 2009;). This linkage has enabled the development of a resource containing information on over 3 million elderly persons with cancer, which has yielded a large number of influential publications describing the patterns and outcomes of cancer care in the United States (Warren et al. 2002; Hershman et al. 2007; Wong et al. 2007; Gooden et al. 2008; Morris et al. 2008; White et al. 2008;). The SEER program collects demographic and diagnostic information for persons diagnosed with cancer, historically representing about 14 percent of the U.S. population and more recently expanded to cover 26 percent. SEER data provide detailed characterizations of cancer diagnosis, tumor stage, and initial treatment but do not track care longitudinally other than vital status. Linkage with Medicare claims for enrollees in fee-for-service plans enables characterization of covered health services longitudinally. A comparable match to Medicaid records has seldom been attempted owing to technical and procedural obstacles (Bradley, Given, and Roberts 2001; Bradley et al. 2007;). These obstacles include the fact that each state's Medicaid program is distinct, that Medicaid enrollment can be discontinuous, and that managed care penetration might be high and thus limit the utility of the claims for ascertaining care. The few attempts to construct registry—Medicaid linkages at the state level include those in Michigan (Bradley, Given, and Roberts 2001, 2002, 2003), California (Perkins et al. 2001; Chan et al. 2006;), Ohio (Koroukian et al. 2006a, 2006b), Washington (Ramsey et al. 2008), and Louisiana (Whitaker et al. 2009). The NYSCR–Medicaid linkage has two primary objectives. The first, discussed in this paper, is to create a deidentified analytic dataset for use in ongoing research projects with extramural partners. We will be using this dataset, for example, to compare community-dwelling cancer patients by Medicaid status, stage distribution, and receipt of quality care to identify whether racial and ethnic disparities exist within the Medicaid program. Findings will be related to timing and duration of enrollment, as those who enrolled in Medicaid in response to a cancer diagnosis might have quite different diagnostic or demographic characteristics than long-term enrollees. The second objective is to assess the degree to which Medicaid claims data add value to the cancer treatment information already collected by the NYSCR, with an emphasis on the treatment of breast and colorectal cancer. The collection of detailed treatment information is a relatively new undertaking for most U.S. cancer registries that are not part of the SEER program; treatment data for New York are considered complete only for cases diagnosed beginning in 2003. Even for this “complete” data, the amount of missing information is typically more than double that seen in SEER. For example, 3.2 percent of colorectal cancer cases diagnosed in NYS between 2004 and 2006 have unknown surgery information, compared with 1.3 percent in SEER. In addition, there is potentially useful information in claims records that cancer registries do not collect at all, such as screening utilization, prescription medications, and nursing home services. Claims data also provide a means for independently verifying that information being collected by central registries is correct. In this paper, we detail the data linkage process and the resulting analytic dataset. We report the percentage of cancer cases matching to the Medicaid enrollment files by cancer site, stage, age, marital status, race/ethnicity, and geography, and show how these differ in important ways from the cancer population generally. We then discuss ways in which the linkage enhanced the quality of registry data, some of which were unanticipated. Finally, we discuss file size issues as the single largest technical hurdle encountered.

BACKGROUND. Preoperative chemoradiation is the standard treatment for locally advanced rectal cancer. However, it is uncertain whether pretreatment clinical stage, degree of response to neoadjuvant treatment, or pathologic stage is the most reliable predictor of outcome. This study compared various staging elements and treatment-related variables to identify which factors or combination of factors reliably prognosticates disease-free survival in rectal cancer patients receiving neoadjuvant combined modality therapy. METHODS. From a prospectively maintained single institution database, 342 consecutive patients with locally advanced rectal cancer staged by endorectal ultrasound were identified. Patients underwent rectal resection 4 to 8 weeks after a 5.5-week course of pelvic radiotherapy/concurrent chemotherapy. The degree of tumor regression was histologically graded on each resected specimen using a previously reported response scale of 0% to 100%. Predictive models of disease-free survival were created utilizing available pretherapy and postoperative staging elements in addition to the degree of tumor regression noted histologically. Model accuracy was measured and compared by concordance index, with 95% confidence interval (CI). RESULTS. Stratifying patients by degree of tumor regression predicted outcome with a concordance index of 0.65 (95% CI, 0.59–0.71), which was significantly better than models using preoperative stage elements (concordance index of 0.54; 95% CI, 0.50–0.58). However, the model found to be most predictive of disease-free survival stratified patients by final pathologic T classification and N classification elements, with a concordance index of 0.75 (95% CI, 0.70–0.80). CONCLUSIONS. Tumor response to preoperative therapy is a strong predictor of disease-free survival. However, outcome is most accurately estimated by final pathologic stage, which is influenced by both preoperative stage and response to therapy. Cancer 2008. © 2008 American Cancer Society.

In recent years health services researchers have conducted 'volume-outcome' studies to evaluate whether providers (hospitals or surgeons) who treat many patients for a specialized condition have better outcomes than those that treat few patients. These studies and the inherent clustering of events by provider present an unusual statistical problem. The volume-outcome setting is unique in that 'volume' reflects both the primary factor under study and also the cluster size. Consequently, the assumptions inherent in the use of available methods that correct for clustering might be violated in this setting. To address this issue, we investigate via simulation the properties of three estimation procedures for the analysis of cluster correlated data, specifically in the context of volume-outcome studies. We examine and compare the validity and efficiency of widely-available statistical techniques that have been used in the context of volume-outcome studies: generalized estimating equations (GEE) using both the independence and exchangeable correlation structures; random effects models; and the weighted GEE approach proposed by Williamson et al. (Biometrics 2003; 59:36-42) to account for informative clustering. Using data generated either from an underlying true random effects model or a cluster correlated model we show that both the random effects and the GEE with an exchangeable correlation structure have generally good properties, with relatively low bias for estimating the volume parameter and its variance. By contrast, the cluster weighted GEE method is inefficient.

BACKGROUND. Although colorectal cancer (CRC) is the second leading cause of cancer death in the U.S., available interventions to reduce CRC mortality are disseminated only partially throughout the population. This study assessed the potential reduction in CRC mortality that may be achieved through further dissemination of current interventions for risk-factor modification, screening, and treatment. METHODS. The MISCAN-COLON microsimulation model was used to simulate the 2000 U.S. population with respect to CRC risk-factor prevalence, screening use, and treatment use. The model was used to project age-standardized CRC mortality from 2000 to 2020 for 3 intervention scenarios. RESULTS. Without changes in risk-factor prevalence, screening use, and treatment use after 2000, CRC mortality would decrease by 17% by the Year 2020. If the 1995 to 2000 trends continue, then the projected reduction in mortality would be 36%. However, if trends in the prevalence of risk-factors could be improved above continued trends, if screening use increased to 70% of the target population, and if the use of chemotherapy increased among all age groups, then a 49% reduction would be possible. Screening drove most (23%) of the projected mortality reduction with these optimistic trends; however, decreasing risk-factors (16%) and increasing use of chemotherapy (10%) also contributed substantially. The contribution of risk-factors may have been overestimated, because effect estimates could not be obtained from randomized controlled trials. CONCLUSIONS. Currently available interventions for risk-factor modification, screening, and treatment have the potential to reduce CRC mortality by almost 50% by the Year 2020. However, without action now to further increase the uptake of current effective interventions, the reduction in CRC mortality may be only 17%. Cancer 2006. © American Cancer Society.

The cost-effectiveness ratio is a popular statistic that is used by policy makers to decide which programs are cost-effective in the public health sector. Recently, the net monetary benefit has been proposed as an alternative statistical summary measure to overcome the limitations associated with the cost-effectiveness ratio. Research on using the net monetary benefit to assess the cost-effectiveness of therapies in non-randomized studies has yet to be done. Propensity scores are useful in estimating adjusted effectiveness of programs that have non-randomized or quasi-experimental designs. This article introduces the use of propensity score adjustment in cost-effectiveness analyses to estimate net monetary benefits for non-randomized studies. The uncertainty associated with the net monetary benefit estimate is evaluated using cost-effectiveness acceptability curves. Our method is illustrated by applying it to SEER-Medicare data for muscle invasive bladder cancer to determine the most cost-effective treatment protocol.

BACKGROUND The results of a randomized comparison study (N9741) showed that oxaliplatin and infusional fluorouracil (FU) (FOLFOX) was superior to the previous standard of care in the United States, irinotecan and bolus FU (IFL), as first-line therapy for patients with metastatic colon carcinoma. The trade-offs between costs and survival for these two regimens have not been explored. METHODS A post-hoc, incremental cost-effectiveness (ICE) projection using simulated cohorts of patients starting FOLFOX or IFL was tracked for major clinical events, toxicities, and survival. Recurrence and survival risks were based on clinical trial data. Resource use was projected using observed dose intensity, duration of therapy, delays in therapy, and toxicities Grade > 2 in N9741. The frequency, costs, and consequences of second-line therapy were examined. The time frame was 5 years, and the perspective was that of Medicare as a third-party payer. RESULTS Initial treatment with FOLFOX versus IFL had an average incremental cost of $29,523, a survival benefit of 4.4 months, and an ICE of $80,410 per life year (LY), $111,890 per quality-adjusted LY, and $89,080 per progression-free year. By using the 95% confidence interval for the time to progression observed in N9741, the ICE associated with FOLFOX ranged from $121,220 to $59,250 per LY. In the clinical trial, dose delays and skipped doses were frequent. If progression-free patients were treated without delay for the first year or lifetime, then the ICE for FOLFOX increased to $117,910 and $222,200 per LY, respectively. The ICE increased to $84,780 per LY when the model incorporated a revised IFL schedule with lower early toxicity and similar rates of treatment with second-line regimens. CONCLUSIONS FOLFOX provided substantial benefits that incurred substantial additional costs. The ICE for FOLFOX fell into the upper range of commonly accepted oncology interventions in the context of the United States healthcare system. Cancer 2005. © 2005 American Cancer Society.

BACKGROUND Temporal comparisons of case survival are commonly used to assess improvement in cancer treatment at the population level. However, such comparisons may be confounded by secular trends in disease prognosis, even within conventional stage categories. The objective of the current study was to characterize within-stage migration of tumor size in breast carcinoma, and to estimate the effect of this shift on reported breast carcinoma survival. METHODS Population-based Surveillance, Epidemiology, and End Results (SEER) cancer registry data were used to evaluate secular trends in tumor size at the time of diagnosis and relative survival among localized and regional invasive breast carcinoma patients diagnosed between 1975–1999. Outcomes were stage-specific tumor size distribution, 5-year relative survival, relative survival standardized to the tumor size distribution of the cohort diagnosed between 1975–1979, and the percentage of improvement in relative survival attributable to shifts in tumor size distribution. RESULTS Within each stage category, the proportion of smaller tumors increased significantly over time. Comparing patients diagnosed between 1995–1999 with those diagnosed between 1975–1979, within-stage migration of tumor size accounted for 61% and 28%, respectively, of the relative survival increases noted in localized and regional breast carcinoma. CONCLUSIONS The tumor size distribution of incident breast carcinomas in SEER has shifted toward smaller tumors. A substantial fraction of the improvement in breast carcinoma survival noted since 1975 may be attributable to within-stage migration of tumor size. Cancer 2005. © 2005 American Cancer Society.

BACKGROUND Logistic, cultural, educational, and other barriers can impede the delivery of high-quality cancer care to underserved patients. Patient navigation services represent one innovation for addressing perceived barriers to care encountered by disadvantaged patients. In this report, the authors have 1) defined patient navigation, distinguishing it from other cancer support services; 2) described how programs are organized; and 3) discussed the need for research on program effectiveness. METHODS Information was examined on navigation programs published in the scientific literature and on line. Qualitative research also was conducted, consisting of direct observation of patient care in cancer clinics with and without navigators in northern California, in-person interviews with personnel and patients in the clinics observed, and telephone interviews with navigators at four sites across the United States. RESULTS The authors found that navigation services have been implemented at all stages of cancer care: prevention, screening, treatment, and survival. Navigators differ from other cancer support personnel in their orientation toward flexible problem solving to overcome perceived barriers to care rather than the provision of a predefined set of services. There are no rigorous demonstrations of the effects and effectiveness of navigation, although such studies are underway. CONCLUSIONS Currently, patient navigation is understudied, and literature documenting its effects and effectiveness is scant. Rigorous studies are needed of the navigator role and program costs and benefits. Such studies will facilitate an assessment of program effectiveness, feasibility across a range of health care settings, and performance relative to alternative approaches for addressing barriers to care among the underserved. Cancer 2005. © 2005 American Cancer Society.

BACKGROUND Use of the Internet is common among patients with cancer and their companions. However, little is known about patterns of use of print or telephone-based resources amidst growing Internet utilization, nor is it known whether different types of information are sought from electronic compared with print media. It is not clear as to whether patients and their companions differ in their patterns of content seeking. METHODS A survey was developed to evaluate the use of electronic and nonelectronic informational resources by patients and their companions. During a 10-week period, this questionnaire was administered to 443 outpatients and 124 paired companions attending an urban academic cancer center. RESULTS In this cohort, 64% of patients and 76% of companions were computer owners, with home Internet access indicated by 58% and 68%, respectively. Use of the Internet to obtain cancer-related information was reported by 44% of patients and 60% of companions. Print resources were used by 79% of patients and 83% of companions, with telephone resources used by 22% and 23%, respectively. The majority of Internet users also read print content (85%), whereas one-half of print users did not access data electronically (52%). Topic areas sought via print and the Internet were similar, with the exception of nutrition-related information, which was more commonly sought in print texts. There was a high rate of concordance between patient and companion use of both electronic and nonelectronic resources. CONCLUSIONS Despite recent publicity and scrutiny focusing on the quality of Internet health care content, print products remain the most common source of information sought by patients with cancer. Future investigation should focus on the quality of print products used by patients. Cancer 2004. © 2004 American Cancer Society.