Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Using the SEER-Medicare Data to Assess Incident Chronic Myeloid Leukemia and Bladder Cancer Cases Missed by Cancer Registries

Using the SEER-Medicare Data to Assess Incident Chronic Myeloid Leukemia and Bladder Cancer Cases... Abstract The growing use of oral systemic therapies and transition of some cancer treatments to the outpatient setting makes capturing all cancer case patients more difficult. We aim to develop algorithms to identify potentially missed incident case patients and estimate impact on incidence rates. We reviewed claims from SEER-Medicare 5% noncancer control patient sample to identify potentially missed chronic myeloid leukemia (CML) and bladder case patients based on diagnosis codes, cancer-related treatments, and oncology consultations. Observed rates of definite missed CML and definite and probable missed bladder case patients were calculated and the impact of missed case patients of these two cancers on SEER 65+ incidence rates were estimated. From 2008 to 2015, the algorithm estimated 781 definite CML case patients missed, increasing the number by 10.7%. From 2007 to 2015, the algorithm estimated 4629 definite and 5772 probable bladder case patients missed, increasing the number by 3.8% to 8.1%. Our algorithms offer potential methods for identifying missed case patients and validating the completeness of cancer registries. Cancer registries are mandated to capture all incident cancer case patients diagnosed in their defined catchment areas. Historically, the vast majority of cancer case patients were identified in the hospital setting. However, with the growing use of oral systemic therapies and transition of some cancer diagnoses and treatments to the outpatient setting (1–9), capturing all cancer case patients is becoming more difficult. Hematologic malignancies are increasingly being diagnosed and treated outside of hospitals, potentially leading to underascertainment by cancer registries. Studies from developed countries estimate the underreporting of hematologic malignancies to be as much as 18–37% (10–15). Chronic myeloid leukemia (CML) is a type of hematologic malignancy that may be underascertained because the treatment is primarily with oral agents that are prescribed in the outpatient setting (16). Similarly, bladder cancer patients with in situ disease are often diagnosed and treated exclusively in outpatient urology offices, which may fail to report the cancer case patients to the registries (5,17). Missed cancer cases would not be included in population-based incidence estimates, potentially leading to unknown underreporting for each of these cancer sites. This study is intended to describe an approach to identify missed incident cancer case patients and to report what impact the missed case patients may have on the National Cancer Institute Surveillance, Epidemiology and End Results (SEER) incidence rates for CML and bladder cancer. We made this assessment using the SEER-Medicare database, which contains Medicare claims data for cancer patients (eg, persons included in the SEER cancer registry data) and noncancer control patients (eg, persons not included in the SEER data) who reside in the SEER catchment areas. We reviewed the claims for the noncancer controls to identify persons with treatments, diagnoses, and oncology consultations that were consistent with what would be expected for CML or bladder cancer case patients. We then used claims-based algorithms to classify the identified persons as definite or probable missed incident cancer case patients. Methods Data Source The SEER-Medicare data comprise a linkage between two large population-based data sources: the National Cancer Institute’s SEER cancer registry data and Centers for Medicare and Medicaid Medicare enrollment and claims data (18). We included the following SEER registries (November 2017 submission): Connecticut, Hawaii, Iowa, New Mexico, Utah, rural and greater Georgia, California, Kentucky, Louisiana, New Jersey, and metropolitan areas including San Francisco–Oakland, Detroit, Seattle–Puget Sound, Atlanta, San Jose–Monterey, and Los Angeles. These areas represent approximately 30% of the total US population (19). Medicare provides federally funded health insurance for persons ages 65 years and older, representing approximately 50 million people in the United States (20). The SEER-Medicare data links 94% of SEER cancer case patients diagnosed at ages 65 years and older to Medicare data, including claims on all Medicare-covered services for beneficiaries with fee-for-service (FFS) coverage. The Centers for Medicare and Medicaid also makes claim data for a random 5% sample of beneficiaries available for research purposes. A SEER-Medicare noncancer control patient sample is created from the Medicare 5% sample by excluding persons who have not resided in the SEER catchment areas and persons who have been diagnosed with cancer (eg, persons included in the SEER data). This analysis is based on the noncancer control patients. The noncancer control patient data include information about where a beneficiary resided during each year of Medicare eligibility. The noncancer control patients with FFS coverage also have available Medicare claims that include hospital care (Part A), physician and outpatient services (Part B), and prescription drugs (Part D). Part B data include International Classification of Diseases (ICD-9 and ICD-10) diagnosis codes and Healthcare Common Procedure Coding System codes for treatments. Part D claims include National Drug Codes to identify prescription medications. Identification of Possible Missed Incident Case Patients We used the SEER-Medicare noncancer control patient sample to annually identify possible missed incident case patients of CML (2008–2015) and bladder cancer (2007–2015), using Medicare claims from the first year assessed through 2016. We reviewed claims for billing codes to identify persons who had a diagnosis or treatment consistent with CML or bladder cancer; these persons were considered possible missed cancer case patients. We focused on physician and outpatient (Part B) and, in the case patient of CML, prescription drug (Part D) claims. Other claim types (eg, from hospitals and hospices) do not provide the granularity of information needed for our assessment and, thus, were not considered. Site-specific inclusion and exclusion criteria for the current analyses were applied, as detailed below. A common inclusion criterion for both cancer sites was that, based on Medicare residency data, persons must have resided in a SEER registry area in the year prior to their index date, the date of the first claim that included a treatment deemed consistent with the cancer of interest. In addition, the persons must have been aged 66 years or older on the index date and had to have at least one year of claims prior to the index date. The rationale for these requirements was to provide the opportunity for the beneficiary to have resided in a SEER area long enough to be captured by the registry and have an adequate look-back period of Medicare claims. Persons who did not have sufficient look-back periods were not included in the assessments because it would not have been possible to ensure that any treatments or diagnoses found in the claims were for an incident cancer and not a prevalent cancer (eg, person diagnosed with cancer in a non-SEER area who then moved into a SEER area). Evidence of cancer-directed treatment was also given preference over a cancer-consistent diagnosis, because diagnosis codes could be listed on a claim for rule-out purposes. To prevent double-counting, persons identified as possible missed cancer case patients in one year were not counted in another year. In both CML and bladder cancer analyses, persons were required to have continuous FFS coverage for the year following the index date (follow-up period). Claims during the follow-up period were reviewed to look for additional evidence (eg, treatments, diagnoses, specialty consultations) that the person was indeed a missed cancer case patient. Potential missed bladder cancer case patients were required to have continuous Medicare Parts A and B enrollment during the follow-up period. Potential missed CML case patients had to have continuous Part D coverage as well as Parts A and B, because prescription drugs are a common treatment modality for CML. Cancer-Specific Diagnoses and Treatments Used to Identify Possible Missed Case Patients Chronic Myeloid Leukemia. All codes used to identify CML-specific diagnoses and treatments of interest are provided in Appendix Table 1 (available online). We used National Comprehensive Cancer Network Guidelines to identify treatments that are common for CML (16) and focused on three tyrosine kinase inhibitors (TKIs): imatinib, nilotinib, and dasatinib. First, we reviewed Part D claims among persons included in the noncancer control patient sample beginning in 2008. SEER-Medicare Part D data are available beginning in 2007, but all eligible persons assessed were required to have at least one year of Medicare claims prior to their index date (eg, first claim indicative of CML treatment); thus, our assessment began with 2008 claims. Table 1. Claims-based algorithm for identifying definite and probable missed bladder cancer case patients included in a population-based noncancer sample* . Had claim with bladder cancer† diagnosis . Had claim with oncology consultation . Missing case patient status . Had a cystectomy† . Yes . Yes . Definite missed incident case patient . Yes . No . Probable missed incident case patient . Had claim with bladder cancer† diagnosis Received BCG Received Mitomycin Had claim with oncology consultation Had claim with radiation therapy Missing case patient status Had a TURBT or bladder biopsy† Yes Yes Definite missed incident case patient Yes Yes Yes Definite missed incident case patient Yes Yes Probable missed incident case patient Yes Yes Probable missed incident case patient Yes Probable missed incident case patient . Had claim with bladder cancer† diagnosis . Had claim with oncology consultation . Missing case patient status . Had a cystectomy† . Yes . Yes . Definite missed incident case patient . Yes . No . Probable missed incident case patient . Had claim with bladder cancer† diagnosis Received BCG Received Mitomycin Had claim with oncology consultation Had claim with radiation therapy Missing case patient status Had a TURBT or bladder biopsy† Yes Yes Definite missed incident case patient Yes Yes Yes Definite missed incident case patient Yes Yes Probable missed incident case patient Yes Yes Probable missed incident case patient Yes Probable missed incident case patient * Total number of beneficiaries included in the SEER-Medicare 5% non-cancer control sample who were age 66 years of older and enrolled in Medicare, as of July 1 each calendar year. BCG = bacillus Calmette-Guerin; TURBT = transurethral resection of bladder tumor. † Determine based on Healthcare Common Procedure Coding System codes. ‡ Determined based on ICD-9 or ICD-10 diagnosis code. Open in new tab Table 1. Claims-based algorithm for identifying definite and probable missed bladder cancer case patients included in a population-based noncancer sample* . Had claim with bladder cancer† diagnosis . Had claim with oncology consultation . Missing case patient status . Had a cystectomy† . Yes . Yes . Definite missed incident case patient . Yes . No . Probable missed incident case patient . Had claim with bladder cancer† diagnosis Received BCG Received Mitomycin Had claim with oncology consultation Had claim with radiation therapy Missing case patient status Had a TURBT or bladder biopsy† Yes Yes Definite missed incident case patient Yes Yes Yes Definite missed incident case patient Yes Yes Probable missed incident case patient Yes Yes Probable missed incident case patient Yes Probable missed incident case patient . Had claim with bladder cancer† diagnosis . Had claim with oncology consultation . Missing case patient status . Had a cystectomy† . Yes . Yes . Definite missed incident case patient . Yes . No . Probable missed incident case patient . Had claim with bladder cancer† diagnosis Received BCG Received Mitomycin Had claim with oncology consultation Had claim with radiation therapy Missing case patient status Had a TURBT or bladder biopsy† Yes Yes Definite missed incident case patient Yes Yes Yes Definite missed incident case patient Yes Yes Probable missed incident case patient Yes Yes Probable missed incident case patient Yes Probable missed incident case patient * Total number of beneficiaries included in the SEER-Medicare 5% non-cancer control sample who were age 66 years of older and enrolled in Medicare, as of July 1 each calendar year. BCG = bacillus Calmette-Guerin; TURBT = transurethral resection of bladder tumor. † Determine based on Healthcare Common Procedure Coding System codes. ‡ Determined based on ICD-9 or ICD-10 diagnosis code. Open in new tab To be a possible missed incident CML case patient, the person must have had two claims for any of the above-mentioned TKIs at least 30 days apart. Definite missed CML case patients were defined as persons who had at least one claim that included a CML diagnosis code on or before the first TKI claim date (index date). Bladder Cancer. All codes used to identify bladder cancer–specific diagnoses and treatments of interest are in Appendix Table 2 (available online). We used National Comprehensive Cancer Network guidelines (21) and consulted with two board-certified urologists (AS, MN) to identify specific treatments that are typically indicated for bladder cancer. Specific surgical procedures included cystectomy (complete or partial), transurethral resection of bladder tumor (TURBT), and bladder biopsy, which are considered mainstay treatments for bladder cancer. Intravesical therapy included mitomycin and bacillus Calmette-Guerin (BCG). Radiotherapy was also taken into consideration. To be a possible missed incident bladder cancer case patient, persons must have had a claim for at least one of the specific surgical procedures. Table 2. Annual denominator calculations used to calculate the incidence rate of missing bladder cancer and CML case patients among persons ages 65 years and older included in the SEER-Medicare 5% noncancer control patient sample Year . Total beneficiary enrollment* . Exclusions† . Final denominators . Non-SEER area residents . Not enrolled in FFS . Did not have Medicare Parts A/B . Did not have Medicare Part D . Denominator for bladder cancer assessment‡ . Denominator for CML assessment§ . 2007 354 314 40 480 75 532 22 960 105 156 215 342 — 2008 364 983 40 765 83 739 24 621 103 089 215 858 112 769 2009 379 687 41 042 93 147 26 869 103 322 218 629 115 307 2010 393 600 41 147 100 581 28 765 104 451 223 107 118 656 2011 407 354 40 724 108 468 30 991 102 707 227 171 124 464 2012 422 167 40 276 116 908 33 144 95 181 231 839 136 658 2013 446 754 40 296 130 626 36 353 82 764 239 479 156 715 2014 469 449 39 780 143 717 38 554 83 398 247 398 164 000 2015 491 278 38 649 157 569 40 651 81 167 254 409 173 242 Year . Total beneficiary enrollment* . Exclusions† . Final denominators . Non-SEER area residents . Not enrolled in FFS . Did not have Medicare Parts A/B . Did not have Medicare Part D . Denominator for bladder cancer assessment‡ . Denominator for CML assessment§ . 2007 354 314 40 480 75 532 22 960 105 156 215 342 — 2008 364 983 40 765 83 739 24 621 103 089 215 858 112 769 2009 379 687 41 042 93 147 26 869 103 322 218 629 115 307 2010 393 600 41 147 100 581 28 765 104 451 223 107 118 656 2011 407 354 40 724 108 468 30 991 102 707 227 171 124 464 2012 422 167 40 276 116 908 33 144 95 181 231 839 136 658 2013 446 754 40 296 130 626 36 353 82 764 239 479 156 715 2014 469 449 39 780 143 717 38 554 83 398 247 398 164 000 2015 491 278 38 649 157 569 40 651 81 167 254 409 173 242 * Total number of beneficiaries included in the SEER-Medicare 5% noncancer control patient sample who were ages 66 years or older and enrolled in Medicare, as of July 1 each calendar year. CML = chronic myeloid leukemia; FFS = fee-for-service; SEER = Surveillance, Epidemiology and End Results. — designates the number of missed CML case patients was not assessed for 2007. † Based on individual beneficiary Medicare enrollment data, as of July 1 each calendar year. ‡ Total beneficiary enrollment minus non-SEER area residents and persons not enrolled in FFS Parts A and B, as of July 1 of each calendar year. § Total beneficiary enrollment minus non-SEER area residents and persons not enrolled in FFS Parts A, B, and D, as of July 1 of each calendar year. Open in new tab Table 2. Annual denominator calculations used to calculate the incidence rate of missing bladder cancer and CML case patients among persons ages 65 years and older included in the SEER-Medicare 5% noncancer control patient sample Year . Total beneficiary enrollment* . Exclusions† . Final denominators . Non-SEER area residents . Not enrolled in FFS . Did not have Medicare Parts A/B . Did not have Medicare Part D . Denominator for bladder cancer assessment‡ . Denominator for CML assessment§ . 2007 354 314 40 480 75 532 22 960 105 156 215 342 — 2008 364 983 40 765 83 739 24 621 103 089 215 858 112 769 2009 379 687 41 042 93 147 26 869 103 322 218 629 115 307 2010 393 600 41 147 100 581 28 765 104 451 223 107 118 656 2011 407 354 40 724 108 468 30 991 102 707 227 171 124 464 2012 422 167 40 276 116 908 33 144 95 181 231 839 136 658 2013 446 754 40 296 130 626 36 353 82 764 239 479 156 715 2014 469 449 39 780 143 717 38 554 83 398 247 398 164 000 2015 491 278 38 649 157 569 40 651 81 167 254 409 173 242 Year . Total beneficiary enrollment* . Exclusions† . Final denominators . Non-SEER area residents . Not enrolled in FFS . Did not have Medicare Parts A/B . Did not have Medicare Part D . Denominator for bladder cancer assessment‡ . Denominator for CML assessment§ . 2007 354 314 40 480 75 532 22 960 105 156 215 342 — 2008 364 983 40 765 83 739 24 621 103 089 215 858 112 769 2009 379 687 41 042 93 147 26 869 103 322 218 629 115 307 2010 393 600 41 147 100 581 28 765 104 451 223 107 118 656 2011 407 354 40 724 108 468 30 991 102 707 227 171 124 464 2012 422 167 40 276 116 908 33 144 95 181 231 839 136 658 2013 446 754 40 296 130 626 36 353 82 764 239 479 156 715 2014 469 449 39 780 143 717 38 554 83 398 247 398 164 000 2015 491 278 38 649 157 569 40 651 81 167 254 409 173 242 * Total number of beneficiaries included in the SEER-Medicare 5% noncancer control patient sample who were ages 66 years or older and enrolled in Medicare, as of July 1 each calendar year. CML = chronic myeloid leukemia; FFS = fee-for-service; SEER = Surveillance, Epidemiology and End Results. — designates the number of missed CML case patients was not assessed for 2007. † Based on individual beneficiary Medicare enrollment data, as of July 1 each calendar year. ‡ Total beneficiary enrollment minus non-SEER area residents and persons not enrolled in FFS Parts A and B, as of July 1 of each calendar year. § Total beneficiary enrollment minus non-SEER area residents and persons not enrolled in FFS Parts A, B, and D, as of July 1 of each calendar year. Open in new tab Possible missed bladder cancer case patients were categorized into definite and probable missed incident case patients (Table 1). The classification algorithm first separated persons by surgical procedure: cystectomy or TURBT or bladder biopsy. Because bladder cancer patients can live with the disease for extended periods of time, patients often undergo surveillance procedures; thus, if a true cancer case patient, we expected to see claims every 3–6 months with bladder cancer diagnosis codes. Therefore, among persons who had a claim for cystectomy, those who also had at least two claims with a bladder cancer diagnosis within any 120-day window during the one-year follow-up period (eg, sufficient evidence of a diagnosis) and a claim for an oncology consultation were categorized as a definite missed incident case patient. Persons who had a cystectomy and sufficient evidence of a bladder cancer diagnosis but no claim for an oncology consultation were categorized as probable missed incident bladder cancer case patients. Among persons who had a TURBT or bladder biopsy, to be considered a definite missed incident case patient, persons had to have sufficient evidence of a bladder cancer diagnosis and either receipt of BCG or an oncology consultation and radiation therapy. Patients were categorized as probable missed incident case patients if they had claims for TURBT or bladder biopsy and 1) had sufficient evidence of a bladder cancer diagnosis and received mitomycin, 2) had sufficient evidence of a bladder cancer diagnosis and a claim for an oncology consultation, or 3) received BCG. Statistical Analysis We identified the number of definite and probable missed incident cancer case patients for both CML and bladder based on the above algorithms. We then calculated the estimated incidence rates per 100 000 among the noncancer control patient sample, for both the definite and probable missed incident case patients by cancer site. Specifically for the bladder cancer analysis, the rate including only the definite missed incident cancers represents the most conservative estimate of missed case patients, whereas the rate including only the probable missed incident case patients represents the increase that could be expected when a less conservative definition of missed case patients was used. We then added the rates for the definite and probable missed case patients to calculate a combined rate of missed incident cancers. The correct denominators from the noncancer control patient sample were created using specific exclusion criteria, as follows (Table 2). Starting from 2007 through 2015, we identified the annual number of people included in the noncancer control patient sample who were aged 66 years or older and enrolled in Medicare, as of July 1, which was chosen as the calendar midway point. Three specific exclusion criteria were applied to these eligible beneficiaries to calculate the final denominators for the CML and bladder cancer assessments. Beneficiaries were excluded if on July 1 of the year they 1) resided in a non-SEER area, based on Medicare residency data, 2) were not enrolled in an FFS plan, and 3) did not have Medicare Parts A and B coverage. An additional exclusion factor was applied for the calculation of the final CML denominator; persons were excluded if they did not have Part D coverage on July 1 of the year. We then quantified the underascertainment of CML and bladder cancer in SEER-18 registry data, among persons 65 years of age and older. First, using SEER*Stat, we calculated annual CML and bladder crude rates per 100 000, based on the annual number of CML and bladder cancers reported to the SEER registries and the underlying annual population counts (denominators) (22). This allowed for a comparison of the SEER (registry-confirmed) and registry-missed incidence rates. Additionally, we applied the missed cancer incidence rates calculated from the noncancer control patient sample, which is based on a random 5% sample, to the SEER population counts to estimate the total number and percentage of CML and bladder cancer case patients that were missed by the registries. Results Chronic Myeloid Leukemia From 2008 to 2015, the number of CML case patients aged 65 years and older in the SEER data was 6518, ranging from 724 in 2008 to 930 in 2015; the crude rate increased moderately over this time period from 7.3 to 7.5 per 10 000 (Table 3). Using the CML missed case patients algorithm, among the 5% noncancer control patient sample, 10 definite missed incident CML case patients were identified between 2008 and 2015. The annual rate of definite missed incident CML case patients ranged from 0.0 to 1.9 per 100 000 during the study period. When the definite missed incidence rates were applied to the total SEER population, we estimated that approximately 781 definite CML incident case patients were missed. Inclusion of these missed case patients would increase CML case patients reported to SEER by 10.7%. Table 3. Number and incidence rate of missed CML case patients identified in beneficiaries ages 65 years and older within the SEER-Medicare 5% noncancer sample and the projected impact on broader SEER CML statistics Year . Calculating the rate of CML case patients found in the 5% noncancer sample* . Number and incidence rate of CML for persons age 65 years and older in SEER data† . Estimated number of missed incident case patients in SEER data . No. of definite missed incident cancers . Denominators for CML cancer case patients‡ . Rate per 100 000 of definite missed incident cancers . No. of CML cancers . SEER population . Crude rate per 100 000 . Estimated no. of definite missed incident cancers . Estimated percent of definite missed incident cancers . 2008 0 112 769 0.0 724 9 924 059 7.3 0.0 0.0 2009 1 115 307 0.9 725 10 163 003 7.1 88.1 10.8 2010 1 118 656 0.8 773 10 405 272 7.4 87.7 10.2 2011 1 124 464 0.8 792 10 662 369 7.4 85.7 9.8 2012 1 136 658 0.7 820 11 140 793 7.4 81.5 9.0 2013 3 156 715 1.9 862 11 548 160 7.5 221.1 20.4 2014 2 164 000 1.2 892 11 962 192 7.5 145.9 14.1 2015 1 173 242 0.6 930 12 377 108 7.5 71.4 7.1 Total 10 1 101 811 — 6518 88 182 956 — 781.4 10.7 Year . Calculating the rate of CML case patients found in the 5% noncancer sample* . Number and incidence rate of CML for persons age 65 years and older in SEER data† . Estimated number of missed incident case patients in SEER data . No. of definite missed incident cancers . Denominators for CML cancer case patients‡ . Rate per 100 000 of definite missed incident cancers . No. of CML cancers . SEER population . Crude rate per 100 000 . Estimated no. of definite missed incident cancers . Estimated percent of definite missed incident cancers . 2008 0 112 769 0.0 724 9 924 059 7.3 0.0 0.0 2009 1 115 307 0.9 725 10 163 003 7.1 88.1 10.8 2010 1 118 656 0.8 773 10 405 272 7.4 87.7 10.2 2011 1 124 464 0.8 792 10 662 369 7.4 85.7 9.8 2012 1 136 658 0.7 820 11 140 793 7.4 81.5 9.0 2013 3 156 715 1.9 862 11 548 160 7.5 221.1 20.4 2014 2 164 000 1.2 892 11 962 192 7.5 145.9 14.1 2015 1 173 242 0.6 930 12 377 108 7.5 71.4 7.1 Total 10 1 101 811 — 6518 88 182 956 — 781.4 10.7 * Medicare beneficiaries included in the SEER-Medicare 5% noncancer control patient sample who were ages 66 years or older each calendar year. CML = chronic myeloid leukemia; SEER = Surveillance, Epidemiology and End Results. † SEER-18 data. ‡ From Table 2. Open in new tab Table 3. Number and incidence rate of missed CML case patients identified in beneficiaries ages 65 years and older within the SEER-Medicare 5% noncancer sample and the projected impact on broader SEER CML statistics Year . Calculating the rate of CML case patients found in the 5% noncancer sample* . Number and incidence rate of CML for persons age 65 years and older in SEER data† . Estimated number of missed incident case patients in SEER data . No. of definite missed incident cancers . Denominators for CML cancer case patients‡ . Rate per 100 000 of definite missed incident cancers . No. of CML cancers . SEER population . Crude rate per 100 000 . Estimated no. of definite missed incident cancers . Estimated percent of definite missed incident cancers . 2008 0 112 769 0.0 724 9 924 059 7.3 0.0 0.0 2009 1 115 307 0.9 725 10 163 003 7.1 88.1 10.8 2010 1 118 656 0.8 773 10 405 272 7.4 87.7 10.2 2011 1 124 464 0.8 792 10 662 369 7.4 85.7 9.8 2012 1 136 658 0.7 820 11 140 793 7.4 81.5 9.0 2013 3 156 715 1.9 862 11 548 160 7.5 221.1 20.4 2014 2 164 000 1.2 892 11 962 192 7.5 145.9 14.1 2015 1 173 242 0.6 930 12 377 108 7.5 71.4 7.1 Total 10 1 101 811 — 6518 88 182 956 — 781.4 10.7 Year . Calculating the rate of CML case patients found in the 5% noncancer sample* . Number and incidence rate of CML for persons age 65 years and older in SEER data† . Estimated number of missed incident case patients in SEER data . No. of definite missed incident cancers . Denominators for CML cancer case patients‡ . Rate per 100 000 of definite missed incident cancers . No. of CML cancers . SEER population . Crude rate per 100 000 . Estimated no. of definite missed incident cancers . Estimated percent of definite missed incident cancers . 2008 0 112 769 0.0 724 9 924 059 7.3 0.0 0.0 2009 1 115 307 0.9 725 10 163 003 7.1 88.1 10.8 2010 1 118 656 0.8 773 10 405 272 7.4 87.7 10.2 2011 1 124 464 0.8 792 10 662 369 7.4 85.7 9.8 2012 1 136 658 0.7 820 11 140 793 7.4 81.5 9.0 2013 3 156 715 1.9 862 11 548 160 7.5 221.1 20.4 2014 2 164 000 1.2 892 11 962 192 7.5 145.9 14.1 2015 1 173 242 0.6 930 12 377 108 7.5 71.4 7.1 Total 10 1 101 811 — 6518 88 182 956 — 781.4 10.7 * Medicare beneficiaries included in the SEER-Medicare 5% noncancer control patient sample who were ages 66 years or older each calendar year. CML = chronic myeloid leukemia; SEER = Surveillance, Epidemiology and End Results. † SEER-18 data. ‡ From Table 2. Open in new tab Bladder From 2007 to 2015, 118 210 bladder cancer case patients were reported to the SEER registries among persons aged 65 years and older; crude rates per 100 000 declined from 128.2 in 2007 to 112.3 in 2015 (Table 4). Using the bladder cancer missed case patients algorithm, among the 5% noncancer control patient sample, 99 definite and 123 probable missed incident bladder cancer case patients were identified between 2007 and 2015. The annual rate of definite missed incident bladder cancer case patients decreased from 12.5 per 100 000 in 2007 to 2.4 per 100 000 in 2015. During the study period, the annual rate of probable missed incident bladder cancer case patients also decreased from 10.2 to 5.1 per 100 000, and the combined missed rate decreased from 22.8 to 7.5 per 100 000. When the definite and probable missed rates were applied to the total SEER population, we estimated that approximately 4629 definite and 5772 probable incident bladder cancer case patients were missed. Inclusion of these missed case patients would increase bladder cancer case patients reported to SEER by 3.8% (include only definite missed) to 8.1% (include definite and probable missed). Table 4. Number and incidence rate of missed bladder case patients identified in beneficiaries ages 65 years and older within the SEER-Medicare 5% noncancer sample and the projected impact on broader SEER bladder statistics Year . Calculating the rate of bladder case patients found in the 5% noncancer sample* . Number and incidence rate of bladder cancer for persons age 65 years and older in SEER data† . Estimated number of missed incident case patients in SEER data . No. of definite missed incident cancers . No. of probable missed incident cancers . Denominators for bladder cancer case patients‡ . Rate per 100 000 of definite missed incident cancers . Rate per 100 000 of probable missed incident cancers . Rate per 100 000 of combined missed incident cancers . No. of bladder cancers . SEER population . Crude rate per 100 000 . Estimated no. of definite missed incident cancers . Estimated percent of definite missed incident cancers . Estimated no. of probable missed incident cancers . Estimated percent of probable missed incident cancers . Estimated no. of combined missed incident cancers . Estimated percent of combined missed incident cancers . 2007 27 22 215 342 12.5 10.2 22.8 12 398 9 667 726 128.2 1212.2 8.9 987.7 7.4 2199.8 15.1 2008 8 16 215 858 3.7 7.4 11.1 12 518 9 924 059 126.1 367.8 2.9 735.6 5.6 1103.4 8.1 2009 13 12 218 629 5.9 5.5 11.4 12 635 10 163 003 124.3 604.3 4.6 557.8 4.2 1162.1 8.4 2010 7 11 223 107 3.1 4.9 8.1 13 016 10 405 272 125.1 326.5 2.4 513.0 3.8 839.5 6.1 2011 5 13 227 171 2.2 5.7 7.9 12 910 10 662 369 121.1 234.7 1.8 610.2 4.5 844.8 6.1 2012 8 14 231 839 3.5 6.0 9.5 13 531 11 140 793 121.5 384.4 2.8 672.8 4.7 1057.2 7.2 2013 11 11 239 479 4.6 4.6 9.2 13 580 11 548 160 117.6 530.4 3.8 530.4 3.8 1060.9 7.2 2014 14 11 247 398 5.7 4.4 10.1 13 723 11 962 192 114.7 676.9 4.7 531.9 3.7 1208.8 8.1 2015 6 13 254 409 2.4 5.1 7.5 13 899 12 377 108 112.3 291.9 2.1 632.5 4.4 924.4 6.2 Total 99 123 2 073 232 — — — 118 210 97 850 682 — 4629.1 3.8 5771.8 4.7 10 400.9 8.1 Year . Calculating the rate of bladder case patients found in the 5% noncancer sample* . Number and incidence rate of bladder cancer for persons age 65 years and older in SEER data† . Estimated number of missed incident case patients in SEER data . No. of definite missed incident cancers . No. of probable missed incident cancers . Denominators for bladder cancer case patients‡ . Rate per 100 000 of definite missed incident cancers . Rate per 100 000 of probable missed incident cancers . Rate per 100 000 of combined missed incident cancers . No. of bladder cancers . SEER population . Crude rate per 100 000 . Estimated no. of definite missed incident cancers . Estimated percent of definite missed incident cancers . Estimated no. of probable missed incident cancers . Estimated percent of probable missed incident cancers . Estimated no. of combined missed incident cancers . Estimated percent of combined missed incident cancers . 2007 27 22 215 342 12.5 10.2 22.8 12 398 9 667 726 128.2 1212.2 8.9 987.7 7.4 2199.8 15.1 2008 8 16 215 858 3.7 7.4 11.1 12 518 9 924 059 126.1 367.8 2.9 735.6 5.6 1103.4 8.1 2009 13 12 218 629 5.9 5.5 11.4 12 635 10 163 003 124.3 604.3 4.6 557.8 4.2 1162.1 8.4 2010 7 11 223 107 3.1 4.9 8.1 13 016 10 405 272 125.1 326.5 2.4 513.0 3.8 839.5 6.1 2011 5 13 227 171 2.2 5.7 7.9 12 910 10 662 369 121.1 234.7 1.8 610.2 4.5 844.8 6.1 2012 8 14 231 839 3.5 6.0 9.5 13 531 11 140 793 121.5 384.4 2.8 672.8 4.7 1057.2 7.2 2013 11 11 239 479 4.6 4.6 9.2 13 580 11 548 160 117.6 530.4 3.8 530.4 3.8 1060.9 7.2 2014 14 11 247 398 5.7 4.4 10.1 13 723 11 962 192 114.7 676.9 4.7 531.9 3.7 1208.8 8.1 2015 6 13 254 409 2.4 5.1 7.5 13 899 12 377 108 112.3 291.9 2.1 632.5 4.4 924.4 6.2 Total 99 123 2 073 232 — — — 118 210 97 850 682 — 4629.1 3.8 5771.8 4.7 10 400.9 8.1 * Medicare beneficiaries included in the SEER-Medicare 5% noncancer control patient sample who were ages 66 years or older each calendar year. SEER = Surveillance, Epidemiology and End Results. † SEER-18 data. ‡ From Table 2. Open in new tab Table 4. Number and incidence rate of missed bladder case patients identified in beneficiaries ages 65 years and older within the SEER-Medicare 5% noncancer sample and the projected impact on broader SEER bladder statistics Year . Calculating the rate of bladder case patients found in the 5% noncancer sample* . Number and incidence rate of bladder cancer for persons age 65 years and older in SEER data† . Estimated number of missed incident case patients in SEER data . No. of definite missed incident cancers . No. of probable missed incident cancers . Denominators for bladder cancer case patients‡ . Rate per 100 000 of definite missed incident cancers . Rate per 100 000 of probable missed incident cancers . Rate per 100 000 of combined missed incident cancers . No. of bladder cancers . SEER population . Crude rate per 100 000 . Estimated no. of definite missed incident cancers . Estimated percent of definite missed incident cancers . Estimated no. of probable missed incident cancers . Estimated percent of probable missed incident cancers . Estimated no. of combined missed incident cancers . Estimated percent of combined missed incident cancers . 2007 27 22 215 342 12.5 10.2 22.8 12 398 9 667 726 128.2 1212.2 8.9 987.7 7.4 2199.8 15.1 2008 8 16 215 858 3.7 7.4 11.1 12 518 9 924 059 126.1 367.8 2.9 735.6 5.6 1103.4 8.1 2009 13 12 218 629 5.9 5.5 11.4 12 635 10 163 003 124.3 604.3 4.6 557.8 4.2 1162.1 8.4 2010 7 11 223 107 3.1 4.9 8.1 13 016 10 405 272 125.1 326.5 2.4 513.0 3.8 839.5 6.1 2011 5 13 227 171 2.2 5.7 7.9 12 910 10 662 369 121.1 234.7 1.8 610.2 4.5 844.8 6.1 2012 8 14 231 839 3.5 6.0 9.5 13 531 11 140 793 121.5 384.4 2.8 672.8 4.7 1057.2 7.2 2013 11 11 239 479 4.6 4.6 9.2 13 580 11 548 160 117.6 530.4 3.8 530.4 3.8 1060.9 7.2 2014 14 11 247 398 5.7 4.4 10.1 13 723 11 962 192 114.7 676.9 4.7 531.9 3.7 1208.8 8.1 2015 6 13 254 409 2.4 5.1 7.5 13 899 12 377 108 112.3 291.9 2.1 632.5 4.4 924.4 6.2 Total 99 123 2 073 232 — — — 118 210 97 850 682 — 4629.1 3.8 5771.8 4.7 10 400.9 8.1 Year . Calculating the rate of bladder case patients found in the 5% noncancer sample* . Number and incidence rate of bladder cancer for persons age 65 years and older in SEER data† . Estimated number of missed incident case patients in SEER data . No. of definite missed incident cancers . No. of probable missed incident cancers . Denominators for bladder cancer case patients‡ . Rate per 100 000 of definite missed incident cancers . Rate per 100 000 of probable missed incident cancers . Rate per 100 000 of combined missed incident cancers . No. of bladder cancers . SEER population . Crude rate per 100 000 . Estimated no. of definite missed incident cancers . Estimated percent of definite missed incident cancers . Estimated no. of probable missed incident cancers . Estimated percent of probable missed incident cancers . Estimated no. of combined missed incident cancers . Estimated percent of combined missed incident cancers . 2007 27 22 215 342 12.5 10.2 22.8 12 398 9 667 726 128.2 1212.2 8.9 987.7 7.4 2199.8 15.1 2008 8 16 215 858 3.7 7.4 11.1 12 518 9 924 059 126.1 367.8 2.9 735.6 5.6 1103.4 8.1 2009 13 12 218 629 5.9 5.5 11.4 12 635 10 163 003 124.3 604.3 4.6 557.8 4.2 1162.1 8.4 2010 7 11 223 107 3.1 4.9 8.1 13 016 10 405 272 125.1 326.5 2.4 513.0 3.8 839.5 6.1 2011 5 13 227 171 2.2 5.7 7.9 12 910 10 662 369 121.1 234.7 1.8 610.2 4.5 844.8 6.1 2012 8 14 231 839 3.5 6.0 9.5 13 531 11 140 793 121.5 384.4 2.8 672.8 4.7 1057.2 7.2 2013 11 11 239 479 4.6 4.6 9.2 13 580 11 548 160 117.6 530.4 3.8 530.4 3.8 1060.9 7.2 2014 14 11 247 398 5.7 4.4 10.1 13 723 11 962 192 114.7 676.9 4.7 531.9 3.7 1208.8 8.1 2015 6 13 254 409 2.4 5.1 7.5 13 899 12 377 108 112.3 291.9 2.1 632.5 4.4 924.4 6.2 Total 99 123 2 073 232 — — — 118 210 97 850 682 — 4629.1 3.8 5771.8 4.7 10 400.9 8.1 * Medicare beneficiaries included in the SEER-Medicare 5% noncancer control patient sample who were ages 66 years or older each calendar year. SEER = Surveillance, Epidemiology and End Results. † SEER-18 data. ‡ From Table 2. Open in new tab Discussion In this study we developed algorithms to identify incident CML and bladder cancer case patients that were not ascertained by the SEER registries. We calculated the estimated rates of the missed case patients using the SEER-Medicare 5% noncancer control patient sample and assessed how inclusion of the missed case patients would impact the number of reported case patients and incidence rates for CML and bladder cancer. Although our results indicate that the number of missed case patients for both cancer sites was relatively low, missed case patients were still identified, warranting further investigation into the causes of the underascertainment. Prior studies have reported underascertainment of hematologic malignancies (10–15); however, our study is the first to assess the completeness of solely CML ascertainment in US population-based cancer registries. The percent increase in CML cancer was 10.7% in our study. The number of missed CML case patients is low because CML is a rare cancer site (23). It should be noted that our algorithm identified incident, not prevalent case patients, through our look-back period criteria. A prior study that assessed the utility of physician office claims to identify hematologic malignancies reported that 63.9% of leukemias found in claims were prevalent case patients (1). Therefore, we think the case patients that were identified as missed case patients using our algorithm were truly incident case patients. The most probable explanation for incident CML case patients being missed by the registries is that these case patients are being treated solely with oral therapies (16), specifically TKIs administered in an outpatient setting as opposed to a hospital setting where the cancer case patient would be captured. Given the increasing trend for CML to be treated with oral therapies, the registries are challenged to identify alternative ascertaining methods. This study demonstrated that prescription drug claims could be a potential ascertainment resource for registries. However, as TKIs are being used for other cancers (24), the registries may opt for requiring a confirmatory diagnosis of CML, as we did for the definite missed case patients. Because of the small number of CML case patients reported to an individual registry, each registry would need to weigh the resources required relative to the number of previously missed case patients that could be recouped. Our analysis of missed bladder cancer portrays a different scenario, compared with our CML analysis. We estimated that the inclusion of missed case patients would represent a smaller increase in total bladder cancer case patients ranging from 3.8% to 8.1%, depending on whether probable missed case patients were included. Because bladder cancer is a common cancer site, the number of missed case patients is larger; we estimated more than 7500 case patients might be missed by the registries. Many bladder cancer case patients are treated as outpatients in large urology oncology practices, and some of them may not be reporting incident case patients to the SEER registries. Our missed bladder cancer algorithm was procedure-based and required the patient to undergo cystectomy or TURBT and bladder biopsy. Our findings suggest that many bladder cancer case patients are not being reported to the registries at the time of surgery. Cystectomy is the more invasive of the two procedures considered and usually requires hospitalization, thereby increasing the likelihood that a person who has this procedure will be captured in the registry data. TURBTs and bladder biopsies are outpatient procedures. With the increase in freestanding ambulatory surgical centers, there has been a transition from TURBTs and bladder biopsies being performed in the hospital outpatient surgery setting to ambulatory surgical centers (25). In addition, intravesical BCG, which is exclusively used to treat bladder cancer, is administered in the physicians’ offices. Thus, treatment outside the hospital likely explains many of the missed bladder case patients. We estimated, in 2007, inclusion of missed case patients would increase the bladder cancer rate per 100 000 substantially, from 128.2 to 151. However, since 2007, bladder incidence rates have decreased, particularly the missed case patients rate. The North American Association of Central Cancer Registries is a recognized international organization establishing standards and certifications for data collection among cancer surveillance programs (26). The SEER registries have attained gold certification that requires at least 95% completeness of case ascertainment. Our results are reassuring and confirm this certification because we have estimated inclusion of missed case patients would represent an increase of 3.8–8.1% for bladder cancer and 10.7% for CML, although the number of CML case patients is very small, and thus these latter findings should be interpreted mindfully. Our study has limitations. We were only able to include potential cancer case patients ages 66 years and older because of eligibility requirements for Medicare; thus, our findings may not be completely generalizable to younger populations. However, the median age at diagnosis with CML is 66 years and 73 years for bladder cancer (22,27), so our findings were focused on the most appropriate age group. We were also only able to assess the missing case patients among FFS beneficiaries. All state-based cancer registries are mandated to capture incident case patients occurring in their state. In the event that a patient who is diagnosed is not a resident of the state, the registry is required to report the case patient to the registry where the patient permanently resides. However, it is possible that some cases were misclassified. If this occurred in our study, our data may include some patients diagnosed in a SEER area who were residents of a non-SEER area. To minimize this occurring, we required that patients in our analysis reside in a SEER area the year before their cancer diagnosis (residence identified through Medicare). Conversely, some cancer patients who resided in a SEER area may have been diagnosed in a non-SEER area and were included in the non-SEER registry data. These case patients would have been included by the cancer surveillance system nationally but would not have appeared in the SEER data. We excluded beneficiaries who were enrolled in health maintenance organizations (HMOs) instead of FFS. Beneficiaries who are treated by HMOs may be less likely to be missed case patients because HMOs usually have a consolidated medical record that includes care in the inpatient and outpatient settings. In addition, we excluded beneficiaries who did not have Medicare Parts A and B. These patients may have been more likely to be missed case patients if they were uninsured or underinsured. This group composed 7.5% of all Medicare beneficiaries in the noncancer file. Additionally, for our CML assessment, receipt of TKIs may have been missed if Medicare was not the primary payer. For our bladder cancer assessment, our findings may have been influenced by changes in coding rules. In 2007, the Facility Oncology Registry Data Standards Manual (28) coding rules changed for T1 bladder tumors. New T1 bladder tumors were diagnosed and reported using the same coding rules before and after 2007, but after 2007, a patient with a history of Ta (noninvasive) who had a subsequent T1 would essentially be counted twice and thus was considered to be a new incident case patient, which may have led to more missed case patients and may explain the larger estimated percentages of definite and probable missed incident bladder cancers in 2007. An additional limitation in our bladder analysis is that we did not capture diagnoses of papillary urothelial neoplasm of low malignant potential. The cancer registries were not required during this time period to capture papillary urothelial neoplasm of low malignant potential in abstracting cancer information, and therefore we do not have any estimates of this condition. This condition is uncommon and controversial in its classification as a cancer and would not impact our estimated rates. In conclusion, our algorithms provide a novel approach to identify missed incident case patients of CML and bladder cancer. This methodology can be used for other cancer sites where there are concerns that the SEER registries might be missing cancers. These methods can also be adjusted to be used with other sources of claims data. Our algorithms offer potential methods for case finding and validation for cancer registries to continue improving and expanding cancer surveillance and ultimately impacting our current understanding of bladder cancer and CML incidence trends. The diagnosis and treatment of cancer is likely to continue migrating into outpatient settings; therefore, complete ascertainment and surveillance of cancer is likely to become more complex. As a result, the SEER registries are continually assessing new data resources and more efficient methods to reduce the number of missed case patients. Funding No specific funding was provided for this research. Notes Affiliations of authors: Data Analytics Branch, Surveillance Research Program, Division of Cancer Control and Population Sciences, National Cancer Institute, Rockville, MD (AM, CL); Healthcare Assessment Research Branch, Healthcare Delivery Research Program, Division of Cancer Control and Population Sciences, National Cancer Institute, Rockville, MD (JW); UNC Department of Urology Oncology, UNC Lineberger Cancer Center, Chapel Hill, NC (AS, MN); Information Management Services, Inc., Calverton, MD (EB, MB). There are no conflict of interest disclosures from any authors. Author contributions: CL: data curation, investigation, formal analysis, writing original draft, writing review, editing; JW: data curation, investigation, formal analysis, writing original draft, writing review, editing; MN: investigation, formal analysis, writing original draft, writing review, clinical expertise; AS: investigation, formal analysis, writing original draft, writing review, clinical expertise; EB: data curation, investigation, formal analysis; MB: data curation, investigation, formal analysis; AM: data curation, investigation, formal analysis, writing original draft, writing review, editing. References 1 Penberthy L , McClish D, Peace S, et al. . Hematologic malignancies: an opportunity to fill a gap in cancer surveillance . Cancer Causes Control . 2012 ; 23 ( 8 ): 1253 – 1264 . Google Scholar Crossref Search ADS PubMed WorldCat 2 Decker SL , Schappert SM, Sisk JE. Use of medical care for chronic conditions . Health Aff (Millwood) . 2009 ; 28 ( 1 ): 26 – 35 . Google Scholar Crossref Search ADS PubMed WorldCat 3 Harlan LC , Clegg LX, Abrams J, Stevens JL, Ballard-Barbash R. Community-based use of chemotherapy and hormonal therapy for early-stage breast cancer: 1987-2000 . J Clin Oncol . 2006 ; 24 ( 6 ): 872 – 877 . Google Scholar Crossref Search ADS PubMed WorldCat 4 Cronin DP , Harlan LC, Clegg LX, Stevens JL, Yuan G, Davis TA. Patterns of care in a population-based random sample of patients diagnosed with non-Hodgkin’s lymphoma . Hematol Oncol . 2005 ; 23 ( 2 ): 73 – 81 . Google Scholar Crossref Search ADS PubMed WorldCat 5 Snyder C , Harlan L, Knopf K, Potosky A, Kaplan R. Patterns of care for the treatment of bladder cancer . J Urol . 2003 ; 169 ( 5 ): 1697 – 1701 . Google Scholar Crossref Search ADS PubMed WorldCat 6 Warren JL , Harlan LC, Fahey A, et al. . Utility of SEER-Medicare data to identify chemotherapy use . Med Care . 2002 ; 40(suppl ): IV – IV-61 . OpenURL Placeholder Text WorldCat 7 Harlan LC , Abrams J, Warren JL, Clegg L, Stevens J, Ballard-Barbash R. Adjuvant therapy for breast cancer: practice patterns of community physicians . J Clin Oncol . 2002 ; 20 ( 7 ): 1809 – 1817 . Google Scholar Crossref Search ADS PubMed WorldCat 8 Schenck A , Klabunde CN, Warren JL, et al. . Evaluation of claims, medical records, and self-report for measuring fecal occult blood testing among Medicare enrollees in fee for service . Cancer Epidemiol Biomarkers Prev . 2008 ; 17 ( 4 ): 799 – 804 . Google Scholar Crossref Search ADS PubMed WorldCat 9 Clarke CA , Glaser SL, Dorfman RF, Bracci PM, Eberle E, Holly EA. Expert review of non-Hodgkin’s lymphomas in a population-based cancer registry: reliability of diagnosis and subtype classification . Cancer Epidemiol Biomarkers Prev . 2004 ; 13 ( 1 ): 138 – 143 . Google Scholar Crossref Search ADS PubMed WorldCat 10 Seftel MD , Demers AA, Banerji V, et al. . High incidence of chronic lymphocytic leukemia (CLL) diagnosed by immunophenotyping: a population-based Canadian cohort . Leuk Res . 2009 ; 33 ( 11 ): 1463 – 1468 . Google Scholar Crossref Search ADS PubMed WorldCat 11 Zent CS , Kyasa MJ, Evans R, Schichman SA. Chronic lymphocytic leukemia incidence is substantially higher than estimated from tumor registry data . Cancer . 2001 ; 92 ( 5 ): 1325 – 1330 . Google Scholar Crossref Search ADS PubMed WorldCat 12 Cogle CR , Craig BM, Rollison DE, List AF. Incidence of the myelodysplastic syndromes using a novel claims-based algorithm: high number of uncaptured cases by cancer registries . Blood . 2011 ; 117 ( 26 ): 7121 – 7125 . Google Scholar Crossref Search ADS PubMed WorldCat 13 Turesson I , Linet MS, Bjorkholm M, et al. . Ascertainment and diagnostic accuracy for hematopoietic lymphoproliferative malignancies in Sweden 1964-2003 . Int J Cancer . 2007 ; 121 ( 10 ): 2260 – 2266 . Google Scholar Crossref Search ADS PubMed WorldCat 14 Skjelbakken T , Lochen ML, Dahl IM. Haematological malignancies in a general population, based on information collected from a population study, hospital records, and the Cancer Registry of Norway: the Tromso Study . Eur J Haematol . 2002 ; 69 ( 2 ): 67 – 75 . Google Scholar Crossref Search ADS PubMed WorldCat 15 Astrom M , Bodin L, Tidefelt U. Adjustment of incidence rates after an estimate of completeness and accuracy in registration of acute leukemias in a Sweden population . Leuk Lymphoma . 2001 ; 41 ( 5-6 ): 559 – 570 . Google Scholar Crossref Search ADS PubMed WorldCat 16 Radich JP , Deininger M, Abboud CN, et al. . Chronic myeloid leukemia, Version 1.2019, NCCN Clinical Practice Guidelines in Oncology . J Natl Compr Canc Netw . 2018 ; 16 ( 9 ): 1108 – 1135 . Google Scholar Crossref Search ADS PubMed WorldCat 17 Penberthy LT , McClish D, Agovino P. Impact of automated data collection from urology offices: improving incidence and treatment reporting in urologic cancers . J Registry Manag . 2010 ; 37 ( 4 ): 141 – 147 . Google Scholar PubMed OpenURL Placeholder Text WorldCat 18 Warren JL , Klabunde CN, Schrag D, Bach PB, Riley GF. Overview of the SEER-Medicare data: content, research applications, and generalizability to the United States elderly population . Med Care . 2002 ; 40(suppl ): IV3 – IV18 . OpenURL Placeholder Text WorldCat 19 National Cancer Institute Surveillance, Epidemiology and End Results. Registry Groupings in SEER Data and Statistics. http://seer.cancer.gov/registries/terms.html. Accessed July 2017. 20 Centers for Medicare and Medicaid Services. 2017 Medicare Enrollment Section. https://www.cms.gov/Research-Statistics-Data-and-Systems/Statistics-Trends-and-Reports/CMSProgramStatistics/2017/2017_Enrollment.html. Accessed July 2017. 21 Flaig TW , Spiess PE, Agarwal N, et al. . NCCN guidelines insights: bladder cancer, version 5.2018 . J Natl Compr Canc Netw . 2018 ; 16 ( 9 ): 1041 – 1053 . Google Scholar Crossref Search ADS PubMed WorldCat 22 National Cancer Institute Surveillance, Epidemiology, and End Results (SEER) Program SEER 9 Regs Research Data, Nov 2018 Sub (1975-2016) National Cancer Institute, DCCPS, Surveillance Research Program, released April 2019 , based on the November 2018 submission. www.seer.cancer.gov. Accessed July 2017. 23 DeSantis CE , Kramer JL, Jemal A. The burden of rare cancers in the United States . CA Cancer J Clin . 2017 ; 67 ( 4 ): 261 – 272 . Google Scholar Crossref Search ADS PubMed WorldCat 24 Arora A , Scholar EM. Role of tyrosine kinase inhibitors in cancer therapy . J Pharmacol Exp Ther . 2005 ; 315 ( 3 ): 971 – 979 . Google Scholar Crossref Search ADS PubMed WorldCat 25 Suskind AM , Dunn RL, Zhang Y, Hollingsworth JM, Hollenbeck BK. Ambulatory surgery centers and outpatient urologic surgery among Medicare beneficiaries . Urology . 2014 ; 84 ( 1 ): 57 – 61 . Google Scholar Crossref Search ADS PubMed WorldCat 26 North American Association of Central Cancer Registries. Certification Criteria. https://www.naaccr.org/certification-criteria/. Accessed July 2017. 27 Nielsen ME , Smith AB, Meyer AM, et al. . Trends in stage-specific incidence rates for urothelial carcinoma of the bladder in the United States: 1988 to 2006 . Cancer . 2014 ; 120 ( 1 ): 86 – 95 . Google Scholar Crossref Search ADS PubMed WorldCat 28 Commission on Cancer-American College of Surgeons. Facility Oncology Registry Data Standards: Revised for 2007. Chicago, IL: Commission on Cancer-American College of Surgeons; 2002 . Published by Oxford University Press 2020. This work is written by US Government employees and is in the public domain in the US. This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model) http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png JNCI Monographs Oxford University Press

Using the SEER-Medicare Data to Assess Incident Chronic Myeloid Leukemia and Bladder Cancer Cases Missed by Cancer Registries

Download PDF
8 pages

Loading next page...
 
/lp/oxford-university-press/using-the-seer-medicare-data-to-assess-incident-chronic-myeloid-0SV1tjzXHc

References (30)

Publisher
Oxford University Press
Copyright
Published by Oxford University Press 2020. This work is written by US Government employees and is in the public domain in the US.
ISSN
1052-6773
eISSN
1745-6614
DOI
10.1093/jncimonographs/lgz033
Publisher site
See Article on Publisher Site

Abstract

Abstract The growing use of oral systemic therapies and transition of some cancer treatments to the outpatient setting makes capturing all cancer case patients more difficult. We aim to develop algorithms to identify potentially missed incident case patients and estimate impact on incidence rates. We reviewed claims from SEER-Medicare 5% noncancer control patient sample to identify potentially missed chronic myeloid leukemia (CML) and bladder case patients based on diagnosis codes, cancer-related treatments, and oncology consultations. Observed rates of definite missed CML and definite and probable missed bladder case patients were calculated and the impact of missed case patients of these two cancers on SEER 65+ incidence rates were estimated. From 2008 to 2015, the algorithm estimated 781 definite CML case patients missed, increasing the number by 10.7%. From 2007 to 2015, the algorithm estimated 4629 definite and 5772 probable bladder case patients missed, increasing the number by 3.8% to 8.1%. Our algorithms offer potential methods for identifying missed case patients and validating the completeness of cancer registries. Cancer registries are mandated to capture all incident cancer case patients diagnosed in their defined catchment areas. Historically, the vast majority of cancer case patients were identified in the hospital setting. However, with the growing use of oral systemic therapies and transition of some cancer diagnoses and treatments to the outpatient setting (1–9), capturing all cancer case patients is becoming more difficult. Hematologic malignancies are increasingly being diagnosed and treated outside of hospitals, potentially leading to underascertainment by cancer registries. Studies from developed countries estimate the underreporting of hematologic malignancies to be as much as 18–37% (10–15). Chronic myeloid leukemia (CML) is a type of hematologic malignancy that may be underascertained because the treatment is primarily with oral agents that are prescribed in the outpatient setting (16). Similarly, bladder cancer patients with in situ disease are often diagnosed and treated exclusively in outpatient urology offices, which may fail to report the cancer case patients to the registries (5,17). Missed cancer cases would not be included in population-based incidence estimates, potentially leading to unknown underreporting for each of these cancer sites. This study is intended to describe an approach to identify missed incident cancer case patients and to report what impact the missed case patients may have on the National Cancer Institute Surveillance, Epidemiology and End Results (SEER) incidence rates for CML and bladder cancer. We made this assessment using the SEER-Medicare database, which contains Medicare claims data for cancer patients (eg, persons included in the SEER cancer registry data) and noncancer control patients (eg, persons not included in the SEER data) who reside in the SEER catchment areas. We reviewed the claims for the noncancer controls to identify persons with treatments, diagnoses, and oncology consultations that were consistent with what would be expected for CML or bladder cancer case patients. We then used claims-based algorithms to classify the identified persons as definite or probable missed incident cancer case patients. Methods Data Source The SEER-Medicare data comprise a linkage between two large population-based data sources: the National Cancer Institute’s SEER cancer registry data and Centers for Medicare and Medicaid Medicare enrollment and claims data (18). We included the following SEER registries (November 2017 submission): Connecticut, Hawaii, Iowa, New Mexico, Utah, rural and greater Georgia, California, Kentucky, Louisiana, New Jersey, and metropolitan areas including San Francisco–Oakland, Detroit, Seattle–Puget Sound, Atlanta, San Jose–Monterey, and Los Angeles. These areas represent approximately 30% of the total US population (19). Medicare provides federally funded health insurance for persons ages 65 years and older, representing approximately 50 million people in the United States (20). The SEER-Medicare data links 94% of SEER cancer case patients diagnosed at ages 65 years and older to Medicare data, including claims on all Medicare-covered services for beneficiaries with fee-for-service (FFS) coverage. The Centers for Medicare and Medicaid also makes claim data for a random 5% sample of beneficiaries available for research purposes. A SEER-Medicare noncancer control patient sample is created from the Medicare 5% sample by excluding persons who have not resided in the SEER catchment areas and persons who have been diagnosed with cancer (eg, persons included in the SEER data). This analysis is based on the noncancer control patients. The noncancer control patient data include information about where a beneficiary resided during each year of Medicare eligibility. The noncancer control patients with FFS coverage also have available Medicare claims that include hospital care (Part A), physician and outpatient services (Part B), and prescription drugs (Part D). Part B data include International Classification of Diseases (ICD-9 and ICD-10) diagnosis codes and Healthcare Common Procedure Coding System codes for treatments. Part D claims include National Drug Codes to identify prescription medications. Identification of Possible Missed Incident Case Patients We used the SEER-Medicare noncancer control patient sample to annually identify possible missed incident case patients of CML (2008–2015) and bladder cancer (2007–2015), using Medicare claims from the first year assessed through 2016. We reviewed claims for billing codes to identify persons who had a diagnosis or treatment consistent with CML or bladder cancer; these persons were considered possible missed cancer case patients. We focused on physician and outpatient (Part B) and, in the case patient of CML, prescription drug (Part D) claims. Other claim types (eg, from hospitals and hospices) do not provide the granularity of information needed for our assessment and, thus, were not considered. Site-specific inclusion and exclusion criteria for the current analyses were applied, as detailed below. A common inclusion criterion for both cancer sites was that, based on Medicare residency data, persons must have resided in a SEER registry area in the year prior to their index date, the date of the first claim that included a treatment deemed consistent with the cancer of interest. In addition, the persons must have been aged 66 years or older on the index date and had to have at least one year of claims prior to the index date. The rationale for these requirements was to provide the opportunity for the beneficiary to have resided in a SEER area long enough to be captured by the registry and have an adequate look-back period of Medicare claims. Persons who did not have sufficient look-back periods were not included in the assessments because it would not have been possible to ensure that any treatments or diagnoses found in the claims were for an incident cancer and not a prevalent cancer (eg, person diagnosed with cancer in a non-SEER area who then moved into a SEER area). Evidence of cancer-directed treatment was also given preference over a cancer-consistent diagnosis, because diagnosis codes could be listed on a claim for rule-out purposes. To prevent double-counting, persons identified as possible missed cancer case patients in one year were not counted in another year. In both CML and bladder cancer analyses, persons were required to have continuous FFS coverage for the year following the index date (follow-up period). Claims during the follow-up period were reviewed to look for additional evidence (eg, treatments, diagnoses, specialty consultations) that the person was indeed a missed cancer case patient. Potential missed bladder cancer case patients were required to have continuous Medicare Parts A and B enrollment during the follow-up period. Potential missed CML case patients had to have continuous Part D coverage as well as Parts A and B, because prescription drugs are a common treatment modality for CML. Cancer-Specific Diagnoses and Treatments Used to Identify Possible Missed Case Patients Chronic Myeloid Leukemia. All codes used to identify CML-specific diagnoses and treatments of interest are provided in Appendix Table 1 (available online). We used National Comprehensive Cancer Network Guidelines to identify treatments that are common for CML (16) and focused on three tyrosine kinase inhibitors (TKIs): imatinib, nilotinib, and dasatinib. First, we reviewed Part D claims among persons included in the noncancer control patient sample beginning in 2008. SEER-Medicare Part D data are available beginning in 2007, but all eligible persons assessed were required to have at least one year of Medicare claims prior to their index date (eg, first claim indicative of CML treatment); thus, our assessment began with 2008 claims. Table 1. Claims-based algorithm for identifying definite and probable missed bladder cancer case patients included in a population-based noncancer sample* . Had claim with bladder cancer† diagnosis . Had claim with oncology consultation . Missing case patient status . Had a cystectomy† . Yes . Yes . Definite missed incident case patient . Yes . No . Probable missed incident case patient . Had claim with bladder cancer† diagnosis Received BCG Received Mitomycin Had claim with oncology consultation Had claim with radiation therapy Missing case patient status Had a TURBT or bladder biopsy† Yes Yes Definite missed incident case patient Yes Yes Yes Definite missed incident case patient Yes Yes Probable missed incident case patient Yes Yes Probable missed incident case patient Yes Probable missed incident case patient . Had claim with bladder cancer† diagnosis . Had claim with oncology consultation . Missing case patient status . Had a cystectomy† . Yes . Yes . Definite missed incident case patient . Yes . No . Probable missed incident case patient . Had claim with bladder cancer† diagnosis Received BCG Received Mitomycin Had claim with oncology consultation Had claim with radiation therapy Missing case patient status Had a TURBT or bladder biopsy† Yes Yes Definite missed incident case patient Yes Yes Yes Definite missed incident case patient Yes Yes Probable missed incident case patient Yes Yes Probable missed incident case patient Yes Probable missed incident case patient * Total number of beneficiaries included in the SEER-Medicare 5% non-cancer control sample who were age 66 years of older and enrolled in Medicare, as of July 1 each calendar year. BCG = bacillus Calmette-Guerin; TURBT = transurethral resection of bladder tumor. † Determine based on Healthcare Common Procedure Coding System codes. ‡ Determined based on ICD-9 or ICD-10 diagnosis code. Open in new tab Table 1. Claims-based algorithm for identifying definite and probable missed bladder cancer case patients included in a population-based noncancer sample* . Had claim with bladder cancer† diagnosis . Had claim with oncology consultation . Missing case patient status . Had a cystectomy† . Yes . Yes . Definite missed incident case patient . Yes . No . Probable missed incident case patient . Had claim with bladder cancer† diagnosis Received BCG Received Mitomycin Had claim with oncology consultation Had claim with radiation therapy Missing case patient status Had a TURBT or bladder biopsy† Yes Yes Definite missed incident case patient Yes Yes Yes Definite missed incident case patient Yes Yes Probable missed incident case patient Yes Yes Probable missed incident case patient Yes Probable missed incident case patient . Had claim with bladder cancer† diagnosis . Had claim with oncology consultation . Missing case patient status . Had a cystectomy† . Yes . Yes . Definite missed incident case patient . Yes . No . Probable missed incident case patient . Had claim with bladder cancer† diagnosis Received BCG Received Mitomycin Had claim with oncology consultation Had claim with radiation therapy Missing case patient status Had a TURBT or bladder biopsy† Yes Yes Definite missed incident case patient Yes Yes Yes Definite missed incident case patient Yes Yes Probable missed incident case patient Yes Yes Probable missed incident case patient Yes Probable missed incident case patient * Total number of beneficiaries included in the SEER-Medicare 5% non-cancer control sample who were age 66 years of older and enrolled in Medicare, as of July 1 each calendar year. BCG = bacillus Calmette-Guerin; TURBT = transurethral resection of bladder tumor. † Determine based on Healthcare Common Procedure Coding System codes. ‡ Determined based on ICD-9 or ICD-10 diagnosis code. Open in new tab To be a possible missed incident CML case patient, the person must have had two claims for any of the above-mentioned TKIs at least 30 days apart. Definite missed CML case patients were defined as persons who had at least one claim that included a CML diagnosis code on or before the first TKI claim date (index date). Bladder Cancer. All codes used to identify bladder cancer–specific diagnoses and treatments of interest are in Appendix Table 2 (available online). We used National Comprehensive Cancer Network guidelines (21) and consulted with two board-certified urologists (AS, MN) to identify specific treatments that are typically indicated for bladder cancer. Specific surgical procedures included cystectomy (complete or partial), transurethral resection of bladder tumor (TURBT), and bladder biopsy, which are considered mainstay treatments for bladder cancer. Intravesical therapy included mitomycin and bacillus Calmette-Guerin (BCG). Radiotherapy was also taken into consideration. To be a possible missed incident bladder cancer case patient, persons must have had a claim for at least one of the specific surgical procedures. Table 2. Annual denominator calculations used to calculate the incidence rate of missing bladder cancer and CML case patients among persons ages 65 years and older included in the SEER-Medicare 5% noncancer control patient sample Year . Total beneficiary enrollment* . Exclusions† . Final denominators . Non-SEER area residents . Not enrolled in FFS . Did not have Medicare Parts A/B . Did not have Medicare Part D . Denominator for bladder cancer assessment‡ . Denominator for CML assessment§ . 2007 354 314 40 480 75 532 22 960 105 156 215 342 — 2008 364 983 40 765 83 739 24 621 103 089 215 858 112 769 2009 379 687 41 042 93 147 26 869 103 322 218 629 115 307 2010 393 600 41 147 100 581 28 765 104 451 223 107 118 656 2011 407 354 40 724 108 468 30 991 102 707 227 171 124 464 2012 422 167 40 276 116 908 33 144 95 181 231 839 136 658 2013 446 754 40 296 130 626 36 353 82 764 239 479 156 715 2014 469 449 39 780 143 717 38 554 83 398 247 398 164 000 2015 491 278 38 649 157 569 40 651 81 167 254 409 173 242 Year . Total beneficiary enrollment* . Exclusions† . Final denominators . Non-SEER area residents . Not enrolled in FFS . Did not have Medicare Parts A/B . Did not have Medicare Part D . Denominator for bladder cancer assessment‡ . Denominator for CML assessment§ . 2007 354 314 40 480 75 532 22 960 105 156 215 342 — 2008 364 983 40 765 83 739 24 621 103 089 215 858 112 769 2009 379 687 41 042 93 147 26 869 103 322 218 629 115 307 2010 393 600 41 147 100 581 28 765 104 451 223 107 118 656 2011 407 354 40 724 108 468 30 991 102 707 227 171 124 464 2012 422 167 40 276 116 908 33 144 95 181 231 839 136 658 2013 446 754 40 296 130 626 36 353 82 764 239 479 156 715 2014 469 449 39 780 143 717 38 554 83 398 247 398 164 000 2015 491 278 38 649 157 569 40 651 81 167 254 409 173 242 * Total number of beneficiaries included in the SEER-Medicare 5% noncancer control patient sample who were ages 66 years or older and enrolled in Medicare, as of July 1 each calendar year. CML = chronic myeloid leukemia; FFS = fee-for-service; SEER = Surveillance, Epidemiology and End Results. — designates the number of missed CML case patients was not assessed for 2007. † Based on individual beneficiary Medicare enrollment data, as of July 1 each calendar year. ‡ Total beneficiary enrollment minus non-SEER area residents and persons not enrolled in FFS Parts A and B, as of July 1 of each calendar year. § Total beneficiary enrollment minus non-SEER area residents and persons not enrolled in FFS Parts A, B, and D, as of July 1 of each calendar year. Open in new tab Table 2. Annual denominator calculations used to calculate the incidence rate of missing bladder cancer and CML case patients among persons ages 65 years and older included in the SEER-Medicare 5% noncancer control patient sample Year . Total beneficiary enrollment* . Exclusions† . Final denominators . Non-SEER area residents . Not enrolled in FFS . Did not have Medicare Parts A/B . Did not have Medicare Part D . Denominator for bladder cancer assessment‡ . Denominator for CML assessment§ . 2007 354 314 40 480 75 532 22 960 105 156 215 342 — 2008 364 983 40 765 83 739 24 621 103 089 215 858 112 769 2009 379 687 41 042 93 147 26 869 103 322 218 629 115 307 2010 393 600 41 147 100 581 28 765 104 451 223 107 118 656 2011 407 354 40 724 108 468 30 991 102 707 227 171 124 464 2012 422 167 40 276 116 908 33 144 95 181 231 839 136 658 2013 446 754 40 296 130 626 36 353 82 764 239 479 156 715 2014 469 449 39 780 143 717 38 554 83 398 247 398 164 000 2015 491 278 38 649 157 569 40 651 81 167 254 409 173 242 Year . Total beneficiary enrollment* . Exclusions† . Final denominators . Non-SEER area residents . Not enrolled in FFS . Did not have Medicare Parts A/B . Did not have Medicare Part D . Denominator for bladder cancer assessment‡ . Denominator for CML assessment§ . 2007 354 314 40 480 75 532 22 960 105 156 215 342 — 2008 364 983 40 765 83 739 24 621 103 089 215 858 112 769 2009 379 687 41 042 93 147 26 869 103 322 218 629 115 307 2010 393 600 41 147 100 581 28 765 104 451 223 107 118 656 2011 407 354 40 724 108 468 30 991 102 707 227 171 124 464 2012 422 167 40 276 116 908 33 144 95 181 231 839 136 658 2013 446 754 40 296 130 626 36 353 82 764 239 479 156 715 2014 469 449 39 780 143 717 38 554 83 398 247 398 164 000 2015 491 278 38 649 157 569 40 651 81 167 254 409 173 242 * Total number of beneficiaries included in the SEER-Medicare 5% noncancer control patient sample who were ages 66 years or older and enrolled in Medicare, as of July 1 each calendar year. CML = chronic myeloid leukemia; FFS = fee-for-service; SEER = Surveillance, Epidemiology and End Results. — designates the number of missed CML case patients was not assessed for 2007. † Based on individual beneficiary Medicare enrollment data, as of July 1 each calendar year. ‡ Total beneficiary enrollment minus non-SEER area residents and persons not enrolled in FFS Parts A and B, as of July 1 of each calendar year. § Total beneficiary enrollment minus non-SEER area residents and persons not enrolled in FFS Parts A, B, and D, as of July 1 of each calendar year. Open in new tab Possible missed bladder cancer case patients were categorized into definite and probable missed incident case patients (Table 1). The classification algorithm first separated persons by surgical procedure: cystectomy or TURBT or bladder biopsy. Because bladder cancer patients can live with the disease for extended periods of time, patients often undergo surveillance procedures; thus, if a true cancer case patient, we expected to see claims every 3–6 months with bladder cancer diagnosis codes. Therefore, among persons who had a claim for cystectomy, those who also had at least two claims with a bladder cancer diagnosis within any 120-day window during the one-year follow-up period (eg, sufficient evidence of a diagnosis) and a claim for an oncology consultation were categorized as a definite missed incident case patient. Persons who had a cystectomy and sufficient evidence of a bladder cancer diagnosis but no claim for an oncology consultation were categorized as probable missed incident bladder cancer case patients. Among persons who had a TURBT or bladder biopsy, to be considered a definite missed incident case patient, persons had to have sufficient evidence of a bladder cancer diagnosis and either receipt of BCG or an oncology consultation and radiation therapy. Patients were categorized as probable missed incident case patients if they had claims for TURBT or bladder biopsy and 1) had sufficient evidence of a bladder cancer diagnosis and received mitomycin, 2) had sufficient evidence of a bladder cancer diagnosis and a claim for an oncology consultation, or 3) received BCG. Statistical Analysis We identified the number of definite and probable missed incident cancer case patients for both CML and bladder based on the above algorithms. We then calculated the estimated incidence rates per 100 000 among the noncancer control patient sample, for both the definite and probable missed incident case patients by cancer site. Specifically for the bladder cancer analysis, the rate including only the definite missed incident cancers represents the most conservative estimate of missed case patients, whereas the rate including only the probable missed incident case patients represents the increase that could be expected when a less conservative definition of missed case patients was used. We then added the rates for the definite and probable missed case patients to calculate a combined rate of missed incident cancers. The correct denominators from the noncancer control patient sample were created using specific exclusion criteria, as follows (Table 2). Starting from 2007 through 2015, we identified the annual number of people included in the noncancer control patient sample who were aged 66 years or older and enrolled in Medicare, as of July 1, which was chosen as the calendar midway point. Three specific exclusion criteria were applied to these eligible beneficiaries to calculate the final denominators for the CML and bladder cancer assessments. Beneficiaries were excluded if on July 1 of the year they 1) resided in a non-SEER area, based on Medicare residency data, 2) were not enrolled in an FFS plan, and 3) did not have Medicare Parts A and B coverage. An additional exclusion factor was applied for the calculation of the final CML denominator; persons were excluded if they did not have Part D coverage on July 1 of the year. We then quantified the underascertainment of CML and bladder cancer in SEER-18 registry data, among persons 65 years of age and older. First, using SEER*Stat, we calculated annual CML and bladder crude rates per 100 000, based on the annual number of CML and bladder cancers reported to the SEER registries and the underlying annual population counts (denominators) (22). This allowed for a comparison of the SEER (registry-confirmed) and registry-missed incidence rates. Additionally, we applied the missed cancer incidence rates calculated from the noncancer control patient sample, which is based on a random 5% sample, to the SEER population counts to estimate the total number and percentage of CML and bladder cancer case patients that were missed by the registries. Results Chronic Myeloid Leukemia From 2008 to 2015, the number of CML case patients aged 65 years and older in the SEER data was 6518, ranging from 724 in 2008 to 930 in 2015; the crude rate increased moderately over this time period from 7.3 to 7.5 per 10 000 (Table 3). Using the CML missed case patients algorithm, among the 5% noncancer control patient sample, 10 definite missed incident CML case patients were identified between 2008 and 2015. The annual rate of definite missed incident CML case patients ranged from 0.0 to 1.9 per 100 000 during the study period. When the definite missed incidence rates were applied to the total SEER population, we estimated that approximately 781 definite CML incident case patients were missed. Inclusion of these missed case patients would increase CML case patients reported to SEER by 10.7%. Table 3. Number and incidence rate of missed CML case patients identified in beneficiaries ages 65 years and older within the SEER-Medicare 5% noncancer sample and the projected impact on broader SEER CML statistics Year . Calculating the rate of CML case patients found in the 5% noncancer sample* . Number and incidence rate of CML for persons age 65 years and older in SEER data† . Estimated number of missed incident case patients in SEER data . No. of definite missed incident cancers . Denominators for CML cancer case patients‡ . Rate per 100 000 of definite missed incident cancers . No. of CML cancers . SEER population . Crude rate per 100 000 . Estimated no. of definite missed incident cancers . Estimated percent of definite missed incident cancers . 2008 0 112 769 0.0 724 9 924 059 7.3 0.0 0.0 2009 1 115 307 0.9 725 10 163 003 7.1 88.1 10.8 2010 1 118 656 0.8 773 10 405 272 7.4 87.7 10.2 2011 1 124 464 0.8 792 10 662 369 7.4 85.7 9.8 2012 1 136 658 0.7 820 11 140 793 7.4 81.5 9.0 2013 3 156 715 1.9 862 11 548 160 7.5 221.1 20.4 2014 2 164 000 1.2 892 11 962 192 7.5 145.9 14.1 2015 1 173 242 0.6 930 12 377 108 7.5 71.4 7.1 Total 10 1 101 811 — 6518 88 182 956 — 781.4 10.7 Year . Calculating the rate of CML case patients found in the 5% noncancer sample* . Number and incidence rate of CML for persons age 65 years and older in SEER data† . Estimated number of missed incident case patients in SEER data . No. of definite missed incident cancers . Denominators for CML cancer case patients‡ . Rate per 100 000 of definite missed incident cancers . No. of CML cancers . SEER population . Crude rate per 100 000 . Estimated no. of definite missed incident cancers . Estimated percent of definite missed incident cancers . 2008 0 112 769 0.0 724 9 924 059 7.3 0.0 0.0 2009 1 115 307 0.9 725 10 163 003 7.1 88.1 10.8 2010 1 118 656 0.8 773 10 405 272 7.4 87.7 10.2 2011 1 124 464 0.8 792 10 662 369 7.4 85.7 9.8 2012 1 136 658 0.7 820 11 140 793 7.4 81.5 9.0 2013 3 156 715 1.9 862 11 548 160 7.5 221.1 20.4 2014 2 164 000 1.2 892 11 962 192 7.5 145.9 14.1 2015 1 173 242 0.6 930 12 377 108 7.5 71.4 7.1 Total 10 1 101 811 — 6518 88 182 956 — 781.4 10.7 * Medicare beneficiaries included in the SEER-Medicare 5% noncancer control patient sample who were ages 66 years or older each calendar year. CML = chronic myeloid leukemia; SEER = Surveillance, Epidemiology and End Results. † SEER-18 data. ‡ From Table 2. Open in new tab Table 3. Number and incidence rate of missed CML case patients identified in beneficiaries ages 65 years and older within the SEER-Medicare 5% noncancer sample and the projected impact on broader SEER CML statistics Year . Calculating the rate of CML case patients found in the 5% noncancer sample* . Number and incidence rate of CML for persons age 65 years and older in SEER data† . Estimated number of missed incident case patients in SEER data . No. of definite missed incident cancers . Denominators for CML cancer case patients‡ . Rate per 100 000 of definite missed incident cancers . No. of CML cancers . SEER population . Crude rate per 100 000 . Estimated no. of definite missed incident cancers . Estimated percent of definite missed incident cancers . 2008 0 112 769 0.0 724 9 924 059 7.3 0.0 0.0 2009 1 115 307 0.9 725 10 163 003 7.1 88.1 10.8 2010 1 118 656 0.8 773 10 405 272 7.4 87.7 10.2 2011 1 124 464 0.8 792 10 662 369 7.4 85.7 9.8 2012 1 136 658 0.7 820 11 140 793 7.4 81.5 9.0 2013 3 156 715 1.9 862 11 548 160 7.5 221.1 20.4 2014 2 164 000 1.2 892 11 962 192 7.5 145.9 14.1 2015 1 173 242 0.6 930 12 377 108 7.5 71.4 7.1 Total 10 1 101 811 — 6518 88 182 956 — 781.4 10.7 Year . Calculating the rate of CML case patients found in the 5% noncancer sample* . Number and incidence rate of CML for persons age 65 years and older in SEER data† . Estimated number of missed incident case patients in SEER data . No. of definite missed incident cancers . Denominators for CML cancer case patients‡ . Rate per 100 000 of definite missed incident cancers . No. of CML cancers . SEER population . Crude rate per 100 000 . Estimated no. of definite missed incident cancers . Estimated percent of definite missed incident cancers . 2008 0 112 769 0.0 724 9 924 059 7.3 0.0 0.0 2009 1 115 307 0.9 725 10 163 003 7.1 88.1 10.8 2010 1 118 656 0.8 773 10 405 272 7.4 87.7 10.2 2011 1 124 464 0.8 792 10 662 369 7.4 85.7 9.8 2012 1 136 658 0.7 820 11 140 793 7.4 81.5 9.0 2013 3 156 715 1.9 862 11 548 160 7.5 221.1 20.4 2014 2 164 000 1.2 892 11 962 192 7.5 145.9 14.1 2015 1 173 242 0.6 930 12 377 108 7.5 71.4 7.1 Total 10 1 101 811 — 6518 88 182 956 — 781.4 10.7 * Medicare beneficiaries included in the SEER-Medicare 5% noncancer control patient sample who were ages 66 years or older each calendar year. CML = chronic myeloid leukemia; SEER = Surveillance, Epidemiology and End Results. † SEER-18 data. ‡ From Table 2. Open in new tab Bladder From 2007 to 2015, 118 210 bladder cancer case patients were reported to the SEER registries among persons aged 65 years and older; crude rates per 100 000 declined from 128.2 in 2007 to 112.3 in 2015 (Table 4). Using the bladder cancer missed case patients algorithm, among the 5% noncancer control patient sample, 99 definite and 123 probable missed incident bladder cancer case patients were identified between 2007 and 2015. The annual rate of definite missed incident bladder cancer case patients decreased from 12.5 per 100 000 in 2007 to 2.4 per 100 000 in 2015. During the study period, the annual rate of probable missed incident bladder cancer case patients also decreased from 10.2 to 5.1 per 100 000, and the combined missed rate decreased from 22.8 to 7.5 per 100 000. When the definite and probable missed rates were applied to the total SEER population, we estimated that approximately 4629 definite and 5772 probable incident bladder cancer case patients were missed. Inclusion of these missed case patients would increase bladder cancer case patients reported to SEER by 3.8% (include only definite missed) to 8.1% (include definite and probable missed). Table 4. Number and incidence rate of missed bladder case patients identified in beneficiaries ages 65 years and older within the SEER-Medicare 5% noncancer sample and the projected impact on broader SEER bladder statistics Year . Calculating the rate of bladder case patients found in the 5% noncancer sample* . Number and incidence rate of bladder cancer for persons age 65 years and older in SEER data† . Estimated number of missed incident case patients in SEER data . No. of definite missed incident cancers . No. of probable missed incident cancers . Denominators for bladder cancer case patients‡ . Rate per 100 000 of definite missed incident cancers . Rate per 100 000 of probable missed incident cancers . Rate per 100 000 of combined missed incident cancers . No. of bladder cancers . SEER population . Crude rate per 100 000 . Estimated no. of definite missed incident cancers . Estimated percent of definite missed incident cancers . Estimated no. of probable missed incident cancers . Estimated percent of probable missed incident cancers . Estimated no. of combined missed incident cancers . Estimated percent of combined missed incident cancers . 2007 27 22 215 342 12.5 10.2 22.8 12 398 9 667 726 128.2 1212.2 8.9 987.7 7.4 2199.8 15.1 2008 8 16 215 858 3.7 7.4 11.1 12 518 9 924 059 126.1 367.8 2.9 735.6 5.6 1103.4 8.1 2009 13 12 218 629 5.9 5.5 11.4 12 635 10 163 003 124.3 604.3 4.6 557.8 4.2 1162.1 8.4 2010 7 11 223 107 3.1 4.9 8.1 13 016 10 405 272 125.1 326.5 2.4 513.0 3.8 839.5 6.1 2011 5 13 227 171 2.2 5.7 7.9 12 910 10 662 369 121.1 234.7 1.8 610.2 4.5 844.8 6.1 2012 8 14 231 839 3.5 6.0 9.5 13 531 11 140 793 121.5 384.4 2.8 672.8 4.7 1057.2 7.2 2013 11 11 239 479 4.6 4.6 9.2 13 580 11 548 160 117.6 530.4 3.8 530.4 3.8 1060.9 7.2 2014 14 11 247 398 5.7 4.4 10.1 13 723 11 962 192 114.7 676.9 4.7 531.9 3.7 1208.8 8.1 2015 6 13 254 409 2.4 5.1 7.5 13 899 12 377 108 112.3 291.9 2.1 632.5 4.4 924.4 6.2 Total 99 123 2 073 232 — — — 118 210 97 850 682 — 4629.1 3.8 5771.8 4.7 10 400.9 8.1 Year . Calculating the rate of bladder case patients found in the 5% noncancer sample* . Number and incidence rate of bladder cancer for persons age 65 years and older in SEER data† . Estimated number of missed incident case patients in SEER data . No. of definite missed incident cancers . No. of probable missed incident cancers . Denominators for bladder cancer case patients‡ . Rate per 100 000 of definite missed incident cancers . Rate per 100 000 of probable missed incident cancers . Rate per 100 000 of combined missed incident cancers . No. of bladder cancers . SEER population . Crude rate per 100 000 . Estimated no. of definite missed incident cancers . Estimated percent of definite missed incident cancers . Estimated no. of probable missed incident cancers . Estimated percent of probable missed incident cancers . Estimated no. of combined missed incident cancers . Estimated percent of combined missed incident cancers . 2007 27 22 215 342 12.5 10.2 22.8 12 398 9 667 726 128.2 1212.2 8.9 987.7 7.4 2199.8 15.1 2008 8 16 215 858 3.7 7.4 11.1 12 518 9 924 059 126.1 367.8 2.9 735.6 5.6 1103.4 8.1 2009 13 12 218 629 5.9 5.5 11.4 12 635 10 163 003 124.3 604.3 4.6 557.8 4.2 1162.1 8.4 2010 7 11 223 107 3.1 4.9 8.1 13 016 10 405 272 125.1 326.5 2.4 513.0 3.8 839.5 6.1 2011 5 13 227 171 2.2 5.7 7.9 12 910 10 662 369 121.1 234.7 1.8 610.2 4.5 844.8 6.1 2012 8 14 231 839 3.5 6.0 9.5 13 531 11 140 793 121.5 384.4 2.8 672.8 4.7 1057.2 7.2 2013 11 11 239 479 4.6 4.6 9.2 13 580 11 548 160 117.6 530.4 3.8 530.4 3.8 1060.9 7.2 2014 14 11 247 398 5.7 4.4 10.1 13 723 11 962 192 114.7 676.9 4.7 531.9 3.7 1208.8 8.1 2015 6 13 254 409 2.4 5.1 7.5 13 899 12 377 108 112.3 291.9 2.1 632.5 4.4 924.4 6.2 Total 99 123 2 073 232 — — — 118 210 97 850 682 — 4629.1 3.8 5771.8 4.7 10 400.9 8.1 * Medicare beneficiaries included in the SEER-Medicare 5% noncancer control patient sample who were ages 66 years or older each calendar year. SEER = Surveillance, Epidemiology and End Results. † SEER-18 data. ‡ From Table 2. Open in new tab Table 4. Number and incidence rate of missed bladder case patients identified in beneficiaries ages 65 years and older within the SEER-Medicare 5% noncancer sample and the projected impact on broader SEER bladder statistics Year . Calculating the rate of bladder case patients found in the 5% noncancer sample* . Number and incidence rate of bladder cancer for persons age 65 years and older in SEER data† . Estimated number of missed incident case patients in SEER data . No. of definite missed incident cancers . No. of probable missed incident cancers . Denominators for bladder cancer case patients‡ . Rate per 100 000 of definite missed incident cancers . Rate per 100 000 of probable missed incident cancers . Rate per 100 000 of combined missed incident cancers . No. of bladder cancers . SEER population . Crude rate per 100 000 . Estimated no. of definite missed incident cancers . Estimated percent of definite missed incident cancers . Estimated no. of probable missed incident cancers . Estimated percent of probable missed incident cancers . Estimated no. of combined missed incident cancers . Estimated percent of combined missed incident cancers . 2007 27 22 215 342 12.5 10.2 22.8 12 398 9 667 726 128.2 1212.2 8.9 987.7 7.4 2199.8 15.1 2008 8 16 215 858 3.7 7.4 11.1 12 518 9 924 059 126.1 367.8 2.9 735.6 5.6 1103.4 8.1 2009 13 12 218 629 5.9 5.5 11.4 12 635 10 163 003 124.3 604.3 4.6 557.8 4.2 1162.1 8.4 2010 7 11 223 107 3.1 4.9 8.1 13 016 10 405 272 125.1 326.5 2.4 513.0 3.8 839.5 6.1 2011 5 13 227 171 2.2 5.7 7.9 12 910 10 662 369 121.1 234.7 1.8 610.2 4.5 844.8 6.1 2012 8 14 231 839 3.5 6.0 9.5 13 531 11 140 793 121.5 384.4 2.8 672.8 4.7 1057.2 7.2 2013 11 11 239 479 4.6 4.6 9.2 13 580 11 548 160 117.6 530.4 3.8 530.4 3.8 1060.9 7.2 2014 14 11 247 398 5.7 4.4 10.1 13 723 11 962 192 114.7 676.9 4.7 531.9 3.7 1208.8 8.1 2015 6 13 254 409 2.4 5.1 7.5 13 899 12 377 108 112.3 291.9 2.1 632.5 4.4 924.4 6.2 Total 99 123 2 073 232 — — — 118 210 97 850 682 — 4629.1 3.8 5771.8 4.7 10 400.9 8.1 Year . Calculating the rate of bladder case patients found in the 5% noncancer sample* . Number and incidence rate of bladder cancer for persons age 65 years and older in SEER data† . Estimated number of missed incident case patients in SEER data . No. of definite missed incident cancers . No. of probable missed incident cancers . Denominators for bladder cancer case patients‡ . Rate per 100 000 of definite missed incident cancers . Rate per 100 000 of probable missed incident cancers . Rate per 100 000 of combined missed incident cancers . No. of bladder cancers . SEER population . Crude rate per 100 000 . Estimated no. of definite missed incident cancers . Estimated percent of definite missed incident cancers . Estimated no. of probable missed incident cancers . Estimated percent of probable missed incident cancers . Estimated no. of combined missed incident cancers . Estimated percent of combined missed incident cancers . 2007 27 22 215 342 12.5 10.2 22.8 12 398 9 667 726 128.2 1212.2 8.9 987.7 7.4 2199.8 15.1 2008 8 16 215 858 3.7 7.4 11.1 12 518 9 924 059 126.1 367.8 2.9 735.6 5.6 1103.4 8.1 2009 13 12 218 629 5.9 5.5 11.4 12 635 10 163 003 124.3 604.3 4.6 557.8 4.2 1162.1 8.4 2010 7 11 223 107 3.1 4.9 8.1 13 016 10 405 272 125.1 326.5 2.4 513.0 3.8 839.5 6.1 2011 5 13 227 171 2.2 5.7 7.9 12 910 10 662 369 121.1 234.7 1.8 610.2 4.5 844.8 6.1 2012 8 14 231 839 3.5 6.0 9.5 13 531 11 140 793 121.5 384.4 2.8 672.8 4.7 1057.2 7.2 2013 11 11 239 479 4.6 4.6 9.2 13 580 11 548 160 117.6 530.4 3.8 530.4 3.8 1060.9 7.2 2014 14 11 247 398 5.7 4.4 10.1 13 723 11 962 192 114.7 676.9 4.7 531.9 3.7 1208.8 8.1 2015 6 13 254 409 2.4 5.1 7.5 13 899 12 377 108 112.3 291.9 2.1 632.5 4.4 924.4 6.2 Total 99 123 2 073 232 — — — 118 210 97 850 682 — 4629.1 3.8 5771.8 4.7 10 400.9 8.1 * Medicare beneficiaries included in the SEER-Medicare 5% noncancer control patient sample who were ages 66 years or older each calendar year. SEER = Surveillance, Epidemiology and End Results. † SEER-18 data. ‡ From Table 2. Open in new tab Discussion In this study we developed algorithms to identify incident CML and bladder cancer case patients that were not ascertained by the SEER registries. We calculated the estimated rates of the missed case patients using the SEER-Medicare 5% noncancer control patient sample and assessed how inclusion of the missed case patients would impact the number of reported case patients and incidence rates for CML and bladder cancer. Although our results indicate that the number of missed case patients for both cancer sites was relatively low, missed case patients were still identified, warranting further investigation into the causes of the underascertainment. Prior studies have reported underascertainment of hematologic malignancies (10–15); however, our study is the first to assess the completeness of solely CML ascertainment in US population-based cancer registries. The percent increase in CML cancer was 10.7% in our study. The number of missed CML case patients is low because CML is a rare cancer site (23). It should be noted that our algorithm identified incident, not prevalent case patients, through our look-back period criteria. A prior study that assessed the utility of physician office claims to identify hematologic malignancies reported that 63.9% of leukemias found in claims were prevalent case patients (1). Therefore, we think the case patients that were identified as missed case patients using our algorithm were truly incident case patients. The most probable explanation for incident CML case patients being missed by the registries is that these case patients are being treated solely with oral therapies (16), specifically TKIs administered in an outpatient setting as opposed to a hospital setting where the cancer case patient would be captured. Given the increasing trend for CML to be treated with oral therapies, the registries are challenged to identify alternative ascertaining methods. This study demonstrated that prescription drug claims could be a potential ascertainment resource for registries. However, as TKIs are being used for other cancers (24), the registries may opt for requiring a confirmatory diagnosis of CML, as we did for the definite missed case patients. Because of the small number of CML case patients reported to an individual registry, each registry would need to weigh the resources required relative to the number of previously missed case patients that could be recouped. Our analysis of missed bladder cancer portrays a different scenario, compared with our CML analysis. We estimated that the inclusion of missed case patients would represent a smaller increase in total bladder cancer case patients ranging from 3.8% to 8.1%, depending on whether probable missed case patients were included. Because bladder cancer is a common cancer site, the number of missed case patients is larger; we estimated more than 7500 case patients might be missed by the registries. Many bladder cancer case patients are treated as outpatients in large urology oncology practices, and some of them may not be reporting incident case patients to the SEER registries. Our missed bladder cancer algorithm was procedure-based and required the patient to undergo cystectomy or TURBT and bladder biopsy. Our findings suggest that many bladder cancer case patients are not being reported to the registries at the time of surgery. Cystectomy is the more invasive of the two procedures considered and usually requires hospitalization, thereby increasing the likelihood that a person who has this procedure will be captured in the registry data. TURBTs and bladder biopsies are outpatient procedures. With the increase in freestanding ambulatory surgical centers, there has been a transition from TURBTs and bladder biopsies being performed in the hospital outpatient surgery setting to ambulatory surgical centers (25). In addition, intravesical BCG, which is exclusively used to treat bladder cancer, is administered in the physicians’ offices. Thus, treatment outside the hospital likely explains many of the missed bladder case patients. We estimated, in 2007, inclusion of missed case patients would increase the bladder cancer rate per 100 000 substantially, from 128.2 to 151. However, since 2007, bladder incidence rates have decreased, particularly the missed case patients rate. The North American Association of Central Cancer Registries is a recognized international organization establishing standards and certifications for data collection among cancer surveillance programs (26). The SEER registries have attained gold certification that requires at least 95% completeness of case ascertainment. Our results are reassuring and confirm this certification because we have estimated inclusion of missed case patients would represent an increase of 3.8–8.1% for bladder cancer and 10.7% for CML, although the number of CML case patients is very small, and thus these latter findings should be interpreted mindfully. Our study has limitations. We were only able to include potential cancer case patients ages 66 years and older because of eligibility requirements for Medicare; thus, our findings may not be completely generalizable to younger populations. However, the median age at diagnosis with CML is 66 years and 73 years for bladder cancer (22,27), so our findings were focused on the most appropriate age group. We were also only able to assess the missing case patients among FFS beneficiaries. All state-based cancer registries are mandated to capture incident case patients occurring in their state. In the event that a patient who is diagnosed is not a resident of the state, the registry is required to report the case patient to the registry where the patient permanently resides. However, it is possible that some cases were misclassified. If this occurred in our study, our data may include some patients diagnosed in a SEER area who were residents of a non-SEER area. To minimize this occurring, we required that patients in our analysis reside in a SEER area the year before their cancer diagnosis (residence identified through Medicare). Conversely, some cancer patients who resided in a SEER area may have been diagnosed in a non-SEER area and were included in the non-SEER registry data. These case patients would have been included by the cancer surveillance system nationally but would not have appeared in the SEER data. We excluded beneficiaries who were enrolled in health maintenance organizations (HMOs) instead of FFS. Beneficiaries who are treated by HMOs may be less likely to be missed case patients because HMOs usually have a consolidated medical record that includes care in the inpatient and outpatient settings. In addition, we excluded beneficiaries who did not have Medicare Parts A and B. These patients may have been more likely to be missed case patients if they were uninsured or underinsured. This group composed 7.5% of all Medicare beneficiaries in the noncancer file. Additionally, for our CML assessment, receipt of TKIs may have been missed if Medicare was not the primary payer. For our bladder cancer assessment, our findings may have been influenced by changes in coding rules. In 2007, the Facility Oncology Registry Data Standards Manual (28) coding rules changed for T1 bladder tumors. New T1 bladder tumors were diagnosed and reported using the same coding rules before and after 2007, but after 2007, a patient with a history of Ta (noninvasive) who had a subsequent T1 would essentially be counted twice and thus was considered to be a new incident case patient, which may have led to more missed case patients and may explain the larger estimated percentages of definite and probable missed incident bladder cancers in 2007. An additional limitation in our bladder analysis is that we did not capture diagnoses of papillary urothelial neoplasm of low malignant potential. The cancer registries were not required during this time period to capture papillary urothelial neoplasm of low malignant potential in abstracting cancer information, and therefore we do not have any estimates of this condition. This condition is uncommon and controversial in its classification as a cancer and would not impact our estimated rates. In conclusion, our algorithms provide a novel approach to identify missed incident case patients of CML and bladder cancer. This methodology can be used for other cancer sites where there are concerns that the SEER registries might be missing cancers. These methods can also be adjusted to be used with other sources of claims data. Our algorithms offer potential methods for case finding and validation for cancer registries to continue improving and expanding cancer surveillance and ultimately impacting our current understanding of bladder cancer and CML incidence trends. The diagnosis and treatment of cancer is likely to continue migrating into outpatient settings; therefore, complete ascertainment and surveillance of cancer is likely to become more complex. As a result, the SEER registries are continually assessing new data resources and more efficient methods to reduce the number of missed case patients. Funding No specific funding was provided for this research. Notes Affiliations of authors: Data Analytics Branch, Surveillance Research Program, Division of Cancer Control and Population Sciences, National Cancer Institute, Rockville, MD (AM, CL); Healthcare Assessment Research Branch, Healthcare Delivery Research Program, Division of Cancer Control and Population Sciences, National Cancer Institute, Rockville, MD (JW); UNC Department of Urology Oncology, UNC Lineberger Cancer Center, Chapel Hill, NC (AS, MN); Information Management Services, Inc., Calverton, MD (EB, MB). There are no conflict of interest disclosures from any authors. Author contributions: CL: data curation, investigation, formal analysis, writing original draft, writing review, editing; JW: data curation, investigation, formal analysis, writing original draft, writing review, editing; MN: investigation, formal analysis, writing original draft, writing review, clinical expertise; AS: investigation, formal analysis, writing original draft, writing review, clinical expertise; EB: data curation, investigation, formal analysis; MB: data curation, investigation, formal analysis; AM: data curation, investigation, formal analysis, writing original draft, writing review, editing. References 1 Penberthy L , McClish D, Peace S, et al. . Hematologic malignancies: an opportunity to fill a gap in cancer surveillance . Cancer Causes Control . 2012 ; 23 ( 8 ): 1253 – 1264 . Google Scholar Crossref Search ADS PubMed WorldCat 2 Decker SL , Schappert SM, Sisk JE. Use of medical care for chronic conditions . Health Aff (Millwood) . 2009 ; 28 ( 1 ): 26 – 35 . Google Scholar Crossref Search ADS PubMed WorldCat 3 Harlan LC , Clegg LX, Abrams J, Stevens JL, Ballard-Barbash R. Community-based use of chemotherapy and hormonal therapy for early-stage breast cancer: 1987-2000 . J Clin Oncol . 2006 ; 24 ( 6 ): 872 – 877 . Google Scholar Crossref Search ADS PubMed WorldCat 4 Cronin DP , Harlan LC, Clegg LX, Stevens JL, Yuan G, Davis TA. Patterns of care in a population-based random sample of patients diagnosed with non-Hodgkin’s lymphoma . Hematol Oncol . 2005 ; 23 ( 2 ): 73 – 81 . Google Scholar Crossref Search ADS PubMed WorldCat 5 Snyder C , Harlan L, Knopf K, Potosky A, Kaplan R. Patterns of care for the treatment of bladder cancer . J Urol . 2003 ; 169 ( 5 ): 1697 – 1701 . Google Scholar Crossref Search ADS PubMed WorldCat 6 Warren JL , Harlan LC, Fahey A, et al. . Utility of SEER-Medicare data to identify chemotherapy use . Med Care . 2002 ; 40(suppl ): IV – IV-61 . OpenURL Placeholder Text WorldCat 7 Harlan LC , Abrams J, Warren JL, Clegg L, Stevens J, Ballard-Barbash R. Adjuvant therapy for breast cancer: practice patterns of community physicians . J Clin Oncol . 2002 ; 20 ( 7 ): 1809 – 1817 . Google Scholar Crossref Search ADS PubMed WorldCat 8 Schenck A , Klabunde CN, Warren JL, et al. . Evaluation of claims, medical records, and self-report for measuring fecal occult blood testing among Medicare enrollees in fee for service . Cancer Epidemiol Biomarkers Prev . 2008 ; 17 ( 4 ): 799 – 804 . Google Scholar Crossref Search ADS PubMed WorldCat 9 Clarke CA , Glaser SL, Dorfman RF, Bracci PM, Eberle E, Holly EA. Expert review of non-Hodgkin’s lymphomas in a population-based cancer registry: reliability of diagnosis and subtype classification . Cancer Epidemiol Biomarkers Prev . 2004 ; 13 ( 1 ): 138 – 143 . Google Scholar Crossref Search ADS PubMed WorldCat 10 Seftel MD , Demers AA, Banerji V, et al. . High incidence of chronic lymphocytic leukemia (CLL) diagnosed by immunophenotyping: a population-based Canadian cohort . Leuk Res . 2009 ; 33 ( 11 ): 1463 – 1468 . Google Scholar Crossref Search ADS PubMed WorldCat 11 Zent CS , Kyasa MJ, Evans R, Schichman SA. Chronic lymphocytic leukemia incidence is substantially higher than estimated from tumor registry data . Cancer . 2001 ; 92 ( 5 ): 1325 – 1330 . Google Scholar Crossref Search ADS PubMed WorldCat 12 Cogle CR , Craig BM, Rollison DE, List AF. Incidence of the myelodysplastic syndromes using a novel claims-based algorithm: high number of uncaptured cases by cancer registries . Blood . 2011 ; 117 ( 26 ): 7121 – 7125 . Google Scholar Crossref Search ADS PubMed WorldCat 13 Turesson I , Linet MS, Bjorkholm M, et al. . Ascertainment and diagnostic accuracy for hematopoietic lymphoproliferative malignancies in Sweden 1964-2003 . Int J Cancer . 2007 ; 121 ( 10 ): 2260 – 2266 . Google Scholar Crossref Search ADS PubMed WorldCat 14 Skjelbakken T , Lochen ML, Dahl IM. Haematological malignancies in a general population, based on information collected from a population study, hospital records, and the Cancer Registry of Norway: the Tromso Study . Eur J Haematol . 2002 ; 69 ( 2 ): 67 – 75 . Google Scholar Crossref Search ADS PubMed WorldCat 15 Astrom M , Bodin L, Tidefelt U. Adjustment of incidence rates after an estimate of completeness and accuracy in registration of acute leukemias in a Sweden population . Leuk Lymphoma . 2001 ; 41 ( 5-6 ): 559 – 570 . Google Scholar Crossref Search ADS PubMed WorldCat 16 Radich JP , Deininger M, Abboud CN, et al. . Chronic myeloid leukemia, Version 1.2019, NCCN Clinical Practice Guidelines in Oncology . J Natl Compr Canc Netw . 2018 ; 16 ( 9 ): 1108 – 1135 . Google Scholar Crossref Search ADS PubMed WorldCat 17 Penberthy LT , McClish D, Agovino P. Impact of automated data collection from urology offices: improving incidence and treatment reporting in urologic cancers . J Registry Manag . 2010 ; 37 ( 4 ): 141 – 147 . Google Scholar PubMed OpenURL Placeholder Text WorldCat 18 Warren JL , Klabunde CN, Schrag D, Bach PB, Riley GF. Overview of the SEER-Medicare data: content, research applications, and generalizability to the United States elderly population . Med Care . 2002 ; 40(suppl ): IV3 – IV18 . OpenURL Placeholder Text WorldCat 19 National Cancer Institute Surveillance, Epidemiology and End Results. Registry Groupings in SEER Data and Statistics. http://seer.cancer.gov/registries/terms.html. Accessed July 2017. 20 Centers for Medicare and Medicaid Services. 2017 Medicare Enrollment Section. https://www.cms.gov/Research-Statistics-Data-and-Systems/Statistics-Trends-and-Reports/CMSProgramStatistics/2017/2017_Enrollment.html. Accessed July 2017. 21 Flaig TW , Spiess PE, Agarwal N, et al. . NCCN guidelines insights: bladder cancer, version 5.2018 . J Natl Compr Canc Netw . 2018 ; 16 ( 9 ): 1041 – 1053 . Google Scholar Crossref Search ADS PubMed WorldCat 22 National Cancer Institute Surveillance, Epidemiology, and End Results (SEER) Program SEER 9 Regs Research Data, Nov 2018 Sub (1975-2016) National Cancer Institute, DCCPS, Surveillance Research Program, released April 2019 , based on the November 2018 submission. www.seer.cancer.gov. Accessed July 2017. 23 DeSantis CE , Kramer JL, Jemal A. The burden of rare cancers in the United States . CA Cancer J Clin . 2017 ; 67 ( 4 ): 261 – 272 . Google Scholar Crossref Search ADS PubMed WorldCat 24 Arora A , Scholar EM. Role of tyrosine kinase inhibitors in cancer therapy . J Pharmacol Exp Ther . 2005 ; 315 ( 3 ): 971 – 979 . Google Scholar Crossref Search ADS PubMed WorldCat 25 Suskind AM , Dunn RL, Zhang Y, Hollingsworth JM, Hollenbeck BK. Ambulatory surgery centers and outpatient urologic surgery among Medicare beneficiaries . Urology . 2014 ; 84 ( 1 ): 57 – 61 . Google Scholar Crossref Search ADS PubMed WorldCat 26 North American Association of Central Cancer Registries. Certification Criteria. https://www.naaccr.org/certification-criteria/. Accessed July 2017. 27 Nielsen ME , Smith AB, Meyer AM, et al. . Trends in stage-specific incidence rates for urothelial carcinoma of the bladder in the United States: 1988 to 2006 . Cancer . 2014 ; 120 ( 1 ): 86 – 95 . Google Scholar Crossref Search ADS PubMed WorldCat 28 Commission on Cancer-American College of Surgeons. Facility Oncology Registry Data Standards: Revised for 2007. Chicago, IL: Commission on Cancer-American College of Surgeons; 2002 . Published by Oxford University Press 2020. This work is written by US Government employees and is in the public domain in the US. This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model)

Journal

JNCI Monographs – Oxford University Press

Published: May 1, 2020

There are no references for this article.