Objective
To investigate cervical cancer treatment of patients enrolled under the Breast and Cervical Cancer Prevention and Treatment Act in Georgia.
Study Design
Georgia Comprehensive Cancer Registry and Medicaid enrollment/claims were used to identify enrollees with preinvasive disease (n = 1149) and invasive cervical cancer (n = 444). Logistic regressions were used to estimate factors associated with the odds of receiving: (1) cancer workup, (2) precancerous procedure, (3) surgery, (4) radiation, and (5) chemotherapy.
Results
Preinvasive disease cases with cervical intraepithelial neoplasia 3, in situ, a comorbidity or without a Commission on Cancer approved hospital nearby were more likely to receive surgery. Among invasive cases, later stage was associated with higher odds of receiving radiation or chemotherapy. Black patients were less likely to have surgery than white patients regardless of preinvasive ( P < .01) or invasive status ( P = .05).
Conclusion
Treatment patterns among Georgia Medicaid cases appear appropriate to stage but 18% with invasive cervical cancer received no cancer treatment, although Medicaid enrolled.
Cervical cancer is one of the most common female cancers in the United States. The American Cancer Society (ACS) estimates approximately 11,270 new cases will be diagnosed with invasive cervical cancer and that there will be 4070 related deaths in 2009. The incidence and mortality of invasive cervical cancer has declined dramatically since the mid 1940s because of the wide use of Papanicolaou (Pap) smear and early detection and treatment of cervical intraepithelial neoplasia (CIN). However, this disease is not evenly distributed across segments of the US population. Women who were low income, low education, or uninsured were more likely to have a higher incidence and mortality mainly because of failure to detect and treat the cancer early. Unlike the patterns for invasive cervical cancer, the number of cases of CIN has continuously increased, especially in young women.
Two important pieces of legislation created greater access to cancer screening, diagnosis, and treatment of this preventable and curable disease. The National Breast and Cervical Cancer Early Detection Program (NBCCEDP) funded by the Centers for Diseases Control and Prevention, has provided screening and diagnostic follow-up for low-income uninsured women since 1990. In program year 2008, NBCCEDP screened 301,209 women for breast cancer with mammography and found 3782 cases of this disease, whereas also screening 321,296 women for cervical cancer and detecting 5201 cancers and high-grade precancerous lesions. Research has indicated that the aging of NBCCEDP significantly improved screening rates of mammography, breast clinic examination, and Pap smear in the late 1990s and yet, because of low funding, this program can reach only an estimated 12-15% of those eligible.
Because of the NBCCEDP does not cover cancer treatment costs, Congress passed the Breast and Cervical Cancer Prevention and Treatment Act (BCCPTA) in 2000 to give states the option of offering women in the NBCCEDP access to treatment through a special Medicaid coverage. BCCPTA was adopted quickly by all 50 states, and Georgia began enrolling their patients into the Women’s Health Medicaid Program (WHMP) in July 2001. Although BCCPTA mandated that women be screened by the NBCCEDP, states had options to extend eligibility to women screened by non-NBCCEDP providers. Georgia is 1 of 12 states that selected this more expansive screening option. In 2005, only 25% of BCCPTA women were screened in health departments through Georgia’s NBCCEDP screening program (known as “Breast Test and More”); the majority (75%) of them were screened by private providers. Women will be continuously eligible for BCCPTA as long as they are considered “under cancer treatment” by their physicians, although states differ in their redetermination process. In 2003, BCCPTA women in Georgia could simply self-report to WHMP every 6 months indicating they were in active treatment.
Our earlier work found that BCCPTA implementation in Georgia led to an increase of 2-3 more women with these cancers enrolling in Medicaid in a given month and also shortened the enrollment process by 7-8 months. Hence, breast cancer patients can enroll and potentially, start treatment while still at an early stage of disease (data unavailable). This could, in turn, lead to better outcomes and longer survival. The simpler recertification process under BCCPTA also created both a stable insurance coverage and connection with the participants’ health care providers, resulting in a 50% decline in the rates of disenrollment from Medicaid for both cancers post-BCCPTA.
The ultimate goal of BCCPTA is to provide Medicaid coverage to assure quality treatment. In Georgia, BCCPTA breast cancer patients were more likely to receive any cancer treatment within 2 years of Medicaid enrollment than those under other Medicaid eligibility groups (data unavailable). Though there was no difference in receiving lumpectomy vs mastectomy; BCCPTA women were more likely than other groups to receive adjuvant therapy after surgery. Given the findings for breast cancer patients, it is important to understand the cancer treatment pattern of patients diagnosed with preinvasive disease and invasive cervical cancer under BCCPTA.
Materials and Methods
Institutional review board
This study was granted Institutional Review Board by the Georgia State Institutional Review Board, the American Cancer Society and Emory University.
Data sources
The major datasets used were the Georgia Cancer Comprehensive Registry (GCCR), Medicaid enrollment, and claims files. The GCCR is a state-wide population-based cancer registry that collects all cancer cases diagnosed in Georgia since 1999. Medicaid enrollment files offer monthly eligibility records and Medicaid claims files contain diagnostic and procedure information coded by the International Classification of Disease, 9th Revision, Clinical Modification (ICD-9-CM) and Current Procedural Terminology (CPT) schema for medical billing.
The county data were from Area Resource File (ARF), Commission on Cancer (CoC), and Consolidated Analysis Center, Incorporated (CACI). The ARF, assembled by the Health Resources and Services Administration, provides a variety of county-level data for the 159 counties in Georgia. The CoC, established by the American College of Surgeons, provides data on hospitals that are able to comply with established CoC standards to deliver cancer care. The CACI US Marketing Systems Group provides data on household’s income level at the county level.
Study sample
Incident cervical cancer cases in the GCCR between July 1, 2001, and Dec. 31, 2004, were linked to the Medicaid enrollment file using the patients’ encrypted social security number for those identified with a primary site of cancer of cervix (local stage and beyond) (n = 470). We excluded those who enrolled in Medicaid more than 6 months after being diagnosed based on the assumption that they did not require Medicaid for cancer treatment on diagnosis (n = 25). We note that the GCCR did not include preinvasive (CIN 2 and 3, in situ) cervical cases; therefore, we identified those cases as women who were ever enrolled under the BCCPTA eligibility category and were not invasive cervical or breast cancer cases in the GCCR (n = 1539).
Of the remaining sample (n = 1984), we excluded those who had more than 1 primary cancer site because their cancer treatment can be affected by other cancers. To ensure that all women can be followed for at least for 2 years, we excluded those over 63 years because their medical claims converted to Medicare at age 65. Those not continuously enrolled in Medicaid in the first 2 months were also excluded because they might not actually be eligible for Medicaid but, rather, were given presumptive eligibility. The final sample was 1937 patients covered by Medicaid, 1149 of whom were diagnosed with preinvasive disease, 444 of whom were diagnosed with invasive cervical cancer and 344 of whom we were unable to identify their stage either from GCCR or Medicaid claims. We omitted them in later analyses.
Cervical cancer treatment
In conjunction with the treatment guidelines of the National Cancer Institute, we examined cervical cancer treatment as the receipt of any: (1) cancer workup, (2) precancerous procedure, (3) surgery, (4) radiation, or (5) chemotherapy among patients with preinvasive and invasive cervical disease.
Cancer workup was defined as a series of tests and procedures that were needed to establish the diagnosis to prepare treatment plans. The standard treatments for patients with preinvasive disease included loop electrosurgical excision procedure (LEEP) and conization, and simple hysterectomy. Invasive surgery, radiation, and/or chemotherapy were treatment options for invasive cases. All treatments were identified from ICD-9-CM code or CPT codes contained in the Medicaid claims files.
Study variables
We included patient’s demographics, disease conditions, Medicaid enrollment status, and resident county characteristics to examine factors associated with the receipt of treatment. Patient’s covariates included: (1) age at enrollment, (2) race/ethnicity, (3) stage at diagnosis, (4) comorbidity index, (5) preenrolled in Medicaid, (6) Medicaid eligibility category, and (7) length of Medicaid enrollment.
The stage data were from either GCCR or based on Medicaid claims for services received in or after their month of enrollment. The stage system we used here is the Surveillance, Epidemiology, and End Results (SEER) Program summary stage, which can group cases into 5 main categories: (1) in situ, (2) local, (3) regional, (4) distant, and (5) unstaged. It can also be derived from the ICD-9-CM codes in medical claims, which helps us identify stage for those not from the GCCR. The full list of codes for identifying cervical cancer stage from the claims is available on request.
To adjust the severity of a noncancer medical illness, which might affect the treatment options, we adopted Romano’s modification of the comorbidity index originally developed by Charlson. All available medical claims up to 1 year after the first month of Medicaid enrollment were used to compute the Charlson comorbidity index, and patients were assigned the maximal score observed.
Because reasons for enrollment in Medicaid vary, we hypothesized that women who enrolled in Medicaid before cancer diagnosis would be a distinct group compared with women who enrolled in Medicaid after t heir cancer diagnosis. Thus, we created a dichotomous variable “preenrolled” to indicate whether subjects were previously enrolled in Medicaid 4 or more months before the month of their cancer diagnosis as recorded in the GCCR.
Medicaid eligibility category was defined based on the most frequent eligibility category during the initial 6 months of enrollment and classified into 3 categories: (1) BCCPTA, (2) disabled, and (3) other eligibility groups. We also created a hierarchy to break a tie (eg, BCCPTA before disabled and disabled before the “other Medicaid”). Finally, length of enrollment was defined as a dichotomous variable, which denoted whether she was continuously enrolled in Medicaid for 24 months or longer.
County covariates included: (1) urban/rural status of resident county, (2) percentage of household’s income <$15K, (3) whether county has at least 1 hospital with CoC certification, and (4) percentage of Obstetrician/Gynecologist (Ob/Gyn) per 1000 women.
Statistical analysis
Logistic regressions were used to estimate the odd ratios (ORs) of independent variables associated with the receipt of cervical cancer treatment. All analyses were conducted using STATA version 9.2 (STATA Corp, College Station, TX).
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
