Receipt of Next-generation Genomic Sequencing among Patients with Metastatic Colorectal Cancer (mCRC) in a Real-World Cohort

Purpose of the Study: Disparities in genomic precision medicine approaches, through molecular profiling or next- generation sequencing (NGS), by race/ethnicity, insurance, and poverty have been identified in lung cancer, but not mCRC. Our goal was to examine disparities in receipt of NGS in patients with mCRC. Methods: We used all-payer electronic health record (EHR)-derived de-identified data from the Flatiron Health database generated from routine clinical care across the United States. Our study population included 26,524 patients with mCRC during the years 2013–2020. In addition to date of NGS testing, the FH-EHR data include demographics (age, sex, and race/ethnicity), payer type, and Eastern Cooperative Oncology Group (ECOG) performance status. We conducted descriptive analyses and multivariable logistic regression analysis to identify correlates of receipt of NGS within 6 months of metastatic diagnosis. Results: Among the 26,524 people with mCRC, 45% (n = 11,946) were women, 48% (n = 12,732) had a Commercial Health Plan, and the majority were seen in a community practice (92%) vs academic hospitals. Over 70% of the patients were White, 12% Black or African-American (AA), and 14% Other. Thirty-three percent (n = 8,821) of patients had documentation in the EHR of having received NGS. After simultaneously adjusting for other factors in the model, older age (ORper year increase: 0.97, 95% CI: 0.96–0.98) and Black/AA race (OR: 0.74, 95% CI: 0.68–0.81), compared to White, was associated with lower odds of receiving NGS testing. Conversely, female sex, better ECOG performance status, later calendar year, being seen in an academic practice, and having a Commercial Health Plan were associated with greater odds of receiving NGS. Conclusions: Our findings indicate that NGS is not received uniformly by all patients with mCRC. Future analyses will incorporate receipt of individual molecular biomarker tests, as recommended by professional societies, as well as their results (e.g., KRAS, NRAS, BRAF, MMR/MSI), treatment information, and survival.

Abstract P6-02-01: Perceived life expectancy, multimorbidity, and breast cancer screening in older women

Objective: Screening mammography (SM) benefits are maximal in women who have a several years' life expectancy. Perceived life expectancy (PLE), however, can be compromised by older age, and/or poor health. Prior studies have shown low rates of SM in younger and/or healthier women, while women who are older and/or those presenting with multimorbidity (MM) continue to undergo screening. The resulting under/overuse of SM causes an imbalance in the use of finite resources that should be rectified as we prepare for population aging. In this study, we investigate the use of SM in women 50 years of age or older in the context of age, PLE, MM, and other social determinants of health.

Methods: This is a cross-sectional study using the nationally-representative 2012 U.S. Health and Retirement Study (n= 8934 women). In addition to demographics, we examined a broad range of variables on social determinants of health; as well as conditions constituting MM, including self-reported chronic conditions (e.g., heart disease), functional limitations (e.g., strength, and mobility limitations), and geriatric syndromes (e.g., low cognitive performance). We defined MM0-MM3 as gradients of MM, based on the occurrence/co-occurrence of chronic conditions, functional limitations, and/or geriatric syndromes. PLE was calculated based on respondent's age-specific assessment of the chance they would live another 10-15 years. The outcome variable was self-reported mammography in the past 2 years. We conducted logistic regression analysis to evaluate the independent and interactive association between each of PLE and MM relative to SM, adjusting for potential confounders.

Results: The median age was 65.7 years; 10.5% were Non-Hispanic Black (NHB), 7.8% were Hispanic, and 3.2% were Other Race. The percent of women with SM was 71.5% in the total population. The median assessment of PLE was a 50% chance of living another 10-15 years. 71.9% of women presented with chronic conditions, 36.8% with functional limitations, and 58.2 with geriatric syndromes. Adjusting for confounders, receipt of mammogram was positively associated with greater certainty of PLE (AOR for an increase of 10% –1.03 (1.01, 1.05)). Compared to those with no conditions (MM0), the co-occurrence of chronic conditions, functional limitations, and/or geriatric syndromes was also positively associated with SM (AOR MM1– 1.48 (1.19, 1.84); AOR MM2– 1.35 (1.09, 1.69); and AOR MM3– 1.29 (1.02, 1.62)). In addition, PLE further strengthened the association between MM and SM. Although age was negatively associated with SM, this interacted with PLE such that the likelihood of having a mammogram was unrelated to age among women 100% certain they would live another 10-15 years.

Conclusion: Both multimorbidity (MM) and perceived life expectancy (PLE) are independently and interactively associated with increased screening mammography (SM), suggesting overuse. Indeed, even among women 75 years of age or older, when SM may be least beneficial, receipt of SM in the past two years was 59.6% in the presence of highest MM gradient, and 68.1% when they were at least 75% certain they will live 10-15 years. A more detailed examination of the basis for PLE is warranted to understand the context in which screening recommendation is made.

Citation Format: Koroukian SM, Warner DF, Schiltz NK, Cooper GS, Owusu C, Stange KC, Berger NA. Perceived life expectancy, multimorbidity, and breast cancer screening in older women [abstract]. In: Proceedings of the 2017 San Antonio Breast Cancer Symposium; 2017 Dec 5-9; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2018;78(4 Suppl):Abstract nr P6-02-01.

Polypharmacy in elders with cancer: An analysis of the Ohio medicaid population

19550

Background: Polypharmacy in the elderly complicates therapy, increases cost of treatment and is a challenge to healthcare agencies. However, the impact of polypharmacy on the care of the older cancer patient has been poorly described. The objective of this study was to characterize the use of pharmacy drugs in the month preceding cancer diagnosis among Ohio Medicaid beneficiaries 65 years of age or older, and diagnosed with incident breast, prostate, or colorectal cancer during the study period 1997–2001. Methods: Medicaid beneficiaries were identified by linking data from the Ohio Cancer Incidence Surveillance System (OCISS) with Medicaid enrollment and claims files on a year-by-year basis. Because of potentially incomplete claims history, the study was limited to patients who were enrolled in Medicaid continuously in the year preceding initial cancer diagnosis. Additionally, those with spenddown or nursing home stay in that time frame were excluded from the analysis. Demographics, anatomic cancer site and stage were retrieved from the OCISS, and all non-pharmacy claims for services received in the year prior to cancer diagnosis were accounted for in deriving the Charlson comorbidity score. Results: The study population included 652 patients: 282 with incident breast cancer, 111 with prostate cancer, and 259 with colorectal cancer. The median age was 73, 74, and 75 years among breast, prostate, and colorectal cancer patients, respectively. Nearly 45% were prescribed at least one type of medication in the month preceding cancer diagnosis. Of those with any prescription in that time period, nearly 49% were prescribed 5 or more types of medications. The most commonly prescribed types of medication were narcotic analgesics, diuretics and potassium replacement, anti- hypertensive agents, laxatives, antianxiety drugs, and antidpressants. No significant variations in the use of medications were observed across the different subgroups of the study population. Conclusions: Polpypharmacy (5 or more types of medication) was present in more than 22% of patients in this study population. The effects of polpypharmacy above and beyond that of comorbidities, disability, and other geriatric syndromes, relative to cancer-related outcomes should be determined.

No significant financial relationships to disclose.

Comparison of cesarean section rates in fee-for-service versus managed care patients in the Ohio Medicaid population, 1992-1997

OBJECTIVE

To examine changes over time in the cesarean section rates for fee-for-service (FFS) beneficiaries versus enrollees of managed care programs (MCPs) in the Ohio Medicaid population.

STUDY DESIGN

Cross-sectional study using linked Ohio birth certificates and Medicaid files.

PATIENTS AND METHODS

Study patients were Medicaid-enrolled residents of urban counties who had singleton, live births from 1992 through 1997 (n = 86,459). Changes in primary and repeat cesarean section rates were analyzed in the FFS and MCP groups. The test of homogeneity of odds ratios was used to measure the statistical difference between unadjusted odds ratios. Logistic regression analysis was conducted to adjust for risk factors.

RESULTS

From 1992 to 1997, the difference in the rates of primary and repeat cesarean sections between FFS and MCP patients decreased. The unadjusted odds ratio (OR) increased from 0.66 to 0.81 (P = .06) for primary cesarean sections and from 0.67 to 1.04 (P = .03) for repeat cesarean sections; this indicated that the likelihood of undergoing a cesarean section increased over time for MCP enrollees compared with FFS beneficiaries. The results of the multivariate analysis indicated that the interaction term of payment source by year was not significant for primary cesarean sections (adjusted OR = 0.93; 95% confidence interval = 0.83, 1.04), but was highly significant for repeat cesarean sections (adjusted OR = 0.53; 95% confidence interval = 0.44, 0.64).

CONCLUSION

We observed a reduction in the difference between the rates of both primary and repeat cesarean sections in FFS and MCP patients over time. The reduction was not statistically significant for primary cesarean sections. For repeat cesarean sections, however, we observed a convergence of the rates for FFS and MCP patients.

Declining trends in cesarean deliveries, Ohio 1989-1996: an analysis by indications

BACKGROUND

Similar to trends observed nationwide, the rates of cesarean deliveries declined in Ohio during the late 1980s and the early 1990s. This study examined the trends in cesarean deliveries in Ohio from 1989 through 1996, in the presence or absence of indications, and in relation to the use of obstetric procedures.

METHODS

Birth certificate data for all singleton, liveborn infants in Ohio (n = 1,204,859) were used to analyze temporal trends in cesarean sections.

RESULTS

The rates of primary and repeat cesarean deliveries declined, respectively, from 15.7 to 12.4 percent and from 83 to 63.3 percent during the 8-year study period. Significant declines in repeat cesarean deliveries were observed both in the presence and absence of documented medical conditions that could present a potential indication for the procedure. The rates of repeat cesareans remained comparable among women with and without documented indications for cesarean section (64% and 61%, respectively). In addition, 45 and 30 percent of repeat cesareans in 1989 and 1996, respectively, were performed in the absence of any documented indications, or on an elective basis. The declines in cesarean delivery rates during the 8-year study period occurred simultaneously with an increase in the use of electronic fetal monitoring, induction, and stimulation of labor.

CONCLUSIONS

The findings suggest that a sizable proportion of repeat cesarean deliveries in 1996 may be unnecessary, even though a marked decline in the procedure has occurred between 1989 and 1996.

Estimating the proportion of unnecessary Cesarean sections in Ohio using birth certificate data

The main objective of the study is to present a method that estimates the proportion of unnecessary Cesarean sections (C-sections) using birth certificate data. This population-based cross-sectional study uses two major databases--Ohio birth certificates and Medicaid eligibility files--and includes singleton infants born during the period July 1991 through June 1993 (n = 262,013). A total of 57 variables indicative of adverse events, including maternal medical risk factors, complications of labor and delivery, and congenital anomalies that are available on the birth certificate, are examined to estimate the rate of unnecessary C-sections. The results obtained through this method indicate that nearly 40% of the repeat C-sections had no documented abnormalities on the birth certificate to justify a C-section. Because studies using medical records have yielded similar results, we believe that using birth certificate data may be a reliable method to measure and monitor the rate of unnecessary C-sections.