Linked insurance-tumor registry database for health services research

Consistency Between State's Cancer Registry and All-Payer Claims Database in Documented Radiation Therapy Among Patients Who Received Breast Conservative Surgery

PURPOSE

Arkansas is one of only four known states that have linked All-Payer Claims Database (APCD) to state's cancer registry (Arkansas Cancer Registry [ACR]). We evaluated the reporting consistency of radiation therapy (RT) between the two sources.

METHODS

Women age ≥ 18 years diagnosed in 2013-2017 with early-stage hormone receptor-positive breast cancer who received breast-conserving surgery were identified. Patients must have continuous insurance coverage (any private plans, Medicaid, and Medicare) in the 13 months (month of diagnosis and 12 months after). Receipt of RT was identified independently from ACR and APCD. We calculated sensitivity, specificity, positive predictive value, and negative predictive value for receipt of RT coded by the registry compared with APCD billing claims as the gold standard. We assessed the degree of concordance between the data sources by Cohen's kappa statistics.

RESULTS

The final sample included 2,695 patients who were in both databases and satisfied our inclusion/exclusion criteria. Using APCD as the gold standard, there were high sensitivity (88.1%) and positive predictive value (87.7%) and moderate specificity (71.1%) and negative predictive value (71.8%). The overall agreement between the two sources was 83.0%, with a kappa statistic of 0.59 (95% CI, 0.56 to 0.63). Consistency measures varied by age, stage, and insurance type with Medicare fee-for-service coverage only having the best and private insurance only the worse consistency.

CONCLUSION

In patients with early-stage hormone receptor-positive breast cancer who received breast-conserving surgery, recording of RT receipt was moderately consistent between Arkansas APCD and ACR. Future studies are needed to identify factors affecting reporting consistency to better use this unique resource in addressing population health problems.

Feasibility of Capturing Cancer Treatment Data in the Utah All-Payer Claims Database

PURPOSE

Incompleteness of treatment data is a recognized limitation of cancer registry data. An all-payer claims database (APCD) is a tool that states use to capture health care information across systems and payer. We linked the Utah Cancer Registry (UCR) records to Utah's statewide APCD and evaluated how this linkage led to improvements in the capture of cancer treatment information.

METHODS

We linked cancers diagnosed and reported to the UCR with Utah APCD claims for the calendar years 2013 and 2014 using LinkPlus Software. For patients with breast or colorectal cancers, manual abstraction was completed to provide a gold-standard comparison for the treatment data obtained from the claims.

RESULTS

Among 10,759 reportable cancer occurrences linked to the APCD, the claims identified additional patients with cancer who received therapies that had been unknown to the registry, increasing the proportion treated with chemotherapy from 23.7% to 27.6%, hormone therapy from 14.1% to 18.8%, immunotherapy from 4.3% to 13.2%, and radiation therapy from 24.9% to 27.5%. The APCD increased the sensitivity of treatment variables compared with the abstraction gold standard. Notably, sensitivity of hormonal therapy for breast cancer increased from 78.6% to 95.2% when augmented with APCD claims data. However, the APCD alone did not achieve as high specificity for treatment data as did the data collected through traditional registry methods.

CONCLUSIONS

This is the first study, to our knowledge, showing that linking cancer registry data with a statewide claims database that covers multiple insurance companies improves cancer treatment data collection. Linking of cancer registry and APCD data can improve comprehensiveness of cancer registry treatment data.

Surveillance, Epidemiology, and End Results (SEER) and SEER-Medicare Databases: Use in Clinical Research for Improving Colorectal Cancer Outcomes

The Surveillance, Epidemiology, and End Results (SEER) program is a clinical database, funded by the National Cancer Institute (NCI), which was created to collect cancer incidence, prevalence, and survival data from U.S. cancer registries. By capturing approximately 30% of the U.S. population, it serves as a powerful resource for researchers focused on understanding the natural history of colorectal cancer and improvement in patient care. The linked SEER-Medicare database is a robust database allowing investigators to perform studies focusing on health disparities, quality of care, and cost of treatment in oncologic disease. Since its infancy in the early 1970s, the database has been utilized for thousands of studies resulting in novel publications that have shaped our management of colorectal cancer among other malignancies.

Factors That Contributed to Black-White Disparities in Survival Among Nonelderly Women With Breast Cancer Between 2004 and 2013

Purpose

To estimate the contribution of differences in demographics, comorbidity, insurance, tumor characteristics, and treatment to the overall mortality disparity between nonelderly black and white women diagnosed with early-stage breast cancer.

Patients and Methods

Excess relative risk of all-cause death in black versus white women diagnosed with stage I to III breast cancer, expressed as a percentage and stratified by hormone receptor status for each variable (demographics, comorbidity, insurance, tumor characteristics, and treatment) in sequentially, propensity-scored, optimally matched patients by using multivariable hazard ratios (HRs).

Results

We identified 563,497 white and black women 18 to 64 years of age diagnosed with stage I to III breast cancer from 2004 to 2013 in the National Cancer Data Base. Among women with hormone receptor–positive disease, who represented 78.5% of all patients, the HR for death in black versus white women in the demographics-matched model was 2.05 (95% CI, 1.94 to 2.17). The HR decreased to 1.93 (95% CI, 1.83 to 2.04), 1.54 (95% CI, 1.47 to 1.62), 1.30 (95% CI, 1.24 to 1.36), and 1.25 (95% CI, 1.19 to 1.31) when sequentially matched for comorbidity, insurance, tumor characteristics, and treatment, respectively. These factors combined accounted for 76.3% of the total excess risk of death in black patients; insurance accounted for 37.0% of the total excess, followed by tumor characteristics (23.2%), comorbidities (11.3%), and treatment (4.8%). Results generally were similar among women with hormone receptor–negative disease, although the HRs were substantially smaller.

Conclusion

Matching by insurance explained one third of the excess risk of death among nonelderly black versus white women diagnosed with early-stage breast cancer; matching by tumor characteristics explained approximately one fifth of the excess risk. Efforts to focus on equalization of access to care could substantially reduce ethnic/racial disparities in overall survival among nonelderly women diagnosed with breast cancer.

Patterns of cervical cancer screening, diagnosis, and follow-up treatment in a state Medicaid fee-for-service population

Despite being a screening-amenable cancer, cervical cancer is the third most common genital cancer among white women and the most common among African American women. The study objective was to use administrative claims data for CC disease surveillance among recipients enrolled in a state Medicaid fee-for-service (FFS) program. West Virginia (WV) Medicaid FFS administrative claims data for female recipients aged 21-64 years from 2003 to 2008 were used for this study. All medical and prescription claims were aggregated to reflect each recipient's medical care and prescription drug utilization. The yearly prevalence of Pap smear testing declined from 23.9% in 2003 to 15.8% in 2008 in the Medicaid FFS population. During the 6-year study period, persistence with Pap smear testing was low; 41.8% of recipients received no Pap smear testing. Only 73.1% of recipients received Pap smear testing during the year prior to their CC or precancerous cervical lesions (PCL) diagnosis. The likelihood of a CC diagnosis increased with a decrease in Pap smear testing persistence. Only 10.1% of recipients received appropriate follow-up care following a diagnosis of high-grade PCL; only 31.5% of the recipients received appropriate follow-up care for low-grade PCL diagnosis. Although CC preventive services such as screening and PCL follow-up care are covered under Medicaid programs, underutilization of these services by recipients in the Medicaid FFS population is a concern. Results of this study emphasize the need to address disparities in screening and appropriate PCL follow-up care among recipients in the Medicaid FFS population.

Analyzing partially missing confounder information in comparative effectiveness and safety research of therapeutics

PURPOSE

Electronic healthcare databases are commonly used in comparative effectiveness and safety research of therapeutics. Many databases now include additional confounder information in a subset of the study population through data linkage or data collection. We described and compared existing methods for analyzing such datasets.

METHODS

Using data from The Health Improvement Network and the relation between non-steroidal anti-inflammatory drugs and upper gastrointestinal bleeding as an example, we employed several methods to handle partially missing confounder information.

RESULTS

The crude odds ratio (OR) of upper gastrointestinal bleeding was 1.50 (95% confidence interval: 0.98, 2.28) among selective cyclo-oxygenase-2 inhibitor initiators (n = 43 569) compared with traditional non-steroidal anti-inflammatory drug initiators (n = 411 616). The OR dropped to 0.81 (0.52, 1.27) upon adjustment for confounders recorded for all patients. When further considering three additional variables missing in 22% of the study population (smoking, alcohol consumption, body mass index), the OR was between 0.80 and 0.83 for the missing-category approach, the missing-indicator approach, single imputation by the most common category, multiple imputation by chained equations, and propensity score calibration. The OR was 0.65 (0.39, 1.09) and 0.67 (0.38, 1.16) for the unweighted and the inverse probability weighted complete-case analysis, respectively.

CONCLUSIONS

Existing methods for handling partially missing confounder data require different assumptions and may produce different results. The unweighted complete-case analysis, the missing-category/indicator approach, and single imputation require often unrealistic assumptions and should be avoided. In this study, differences across methods were not substantial, likely due to relatively low proportion of missingness and weak confounding effect by the three additional variables upon adjustment for other variables.

Comparing spatio-temporal clusters of arthropod-borne infections using administrative medical claims and state reported surveillance data

Considered separately, notifiable disease registries and medical claims data have certain advantages (e.g., consistent case definitions and electronic records, respectively) and limitations (e.g., incomplete reporting and coding errors, respectively) within disease outbreak research. Combined however, these data could provide a more complete source of information. Using a retrospective space-time permutation scan statistic, zoonotic case information from a state registry system (TDH) was compared with administrative medical claims information from a managed care organization (MCO) to examine how data sources differ. Study observations included case information for four tick-borne (Lyme disease, ehrlichiosis, Rocky Mountain spotted fever, tularemia) and two mosquito-borne diseases (West Nile virus, La Crosse viral encephalitis) occurring in Tennessee. One hundred and three clusters were detected, of which nine were significant (P<0.05). Considering only significant clusters, no spatial or temporal overlapping between data sources occurred. In conclusion, data integration efforts and data limitations should be considered to provide more comprehensive case information.

Evaluation of record linkage between a large healthcare provider and the Utah Population Database

OBJECTIVE

Electronically linked datasets have become an important part of clinical research. Information from multiple sources can be used to identify comorbid conditions and patient outcomes, measure use of healthcare services, and enrich demographic and clinical variables of interest. Innovative approaches for creating research infrastructure beyond a traditional data system are necessary.

MATERIALS AND METHODS

Records from a large healthcare system's enterprise data warehouse (EDW) were linked to a statewide population database, and a master subject index was created. The authors evaluate the linkage, along with the impact of missing information in EDW records and the coverage of the population database. The makeup of the EDW and population database provides a subset of cancer records that exist in both resources, which allows a cancer-specific evaluation of the linkage.

RESULTS

About 3.4 million records (60.8%) in the EDW were linked to the population database with a minimum accuracy of 96.3%. It was estimated that approximately 24.8% of target records were absent from the population database, which enabled the effect of the amount and type of information missing from a record on the linkage to be estimated. However, 99% of the records from the oncology data mart linked; they had fewer missing fields and this correlated positively with the number of patient visits.

DISCUSSION AND CONCLUSION

A general-purpose research infrastructure was created which allows disease-specific cohorts to be identified. The usefulness of creating an index between institutions is that it allows each institution to maintain control and confidentiality of their own information.

Discontinuation of radiation treatment among medicaid-enrolled women with local and regional stage breast cancer

For women with nonmetastatic breast cancer, radiation therapy is recommended as a necessary component of the breast conserving surgery (BCS) treatment option. The degree to which Medicaid-enrolled women complete recommended radiation therapy protocols is not known. We evaluate radiation treatment completion rates for Medicaid enrollees aged 18-64 diagnosed with breast cancer. We determine clinical and socio-demographic factors associated with not starting treatment, and with interruptions or not completing radiation treatment. Using data from the Washington State Cancer Registry linked to Medicaid enrollment and claims records, we identified Medicaid enrollees diagnosed with breast cancer from 1997 to 2003 who received BCS. Among the 402 women who met inclusion criteria, 105 (26%) did not receive any radiation. Factors significantly associated with not receiving radiation included in situ disease and non-English as a primary language. Among those who received at least one radiation treatment, 65 (22%) failed to complete therapy and 71 (24%) patients had at least one 5 to 30 day gap in treatment. We found no significant predictors of interruptions in treatment or early discontinuation. A substantial proportion of Medicaid-insured women who are eligible for radiation therapy following BCS either fail to receive any treatment, experience significant interruptions during therapy, or do not complete a minimum course of treatment. More effort is needed to ensure this vulnerable population receives adequate radiation following BCS.

Cancer registry databases: an overview of techniques of statistical analysis and impact on cancer epidemiology

Cancer registries provide systematically collected information on cancer incidence, prevalence, mortality, and survival of different cancers. Aggregated and de-identified patient-level information on cancer is available for analysis from individual cancer registries, nationally from the Surveillance, Epidemiology, and End Results program, the Centers for Diseases Control and Prevention, the North American Association of Central Cancer Registries; and internationally from the International Association of Cancer Registries. Over the past few decades, the type and extent of cancer-related information captured by different cancer registries have been greatly expanded by linkage with other population-based information sources, such as the census data and the Centers for Medicare and Medicaid Services claims data. In addition, sophisticated statistical analytical techniques have been developed that have greatly expanded the traditional purview of cancer registries focused on descriptive epidemiology and disease quantification to a much broader analytical horizon ranging from study of cancer etiology; rare cancers in specific demographic groups; interaction of environmental and genetic factors in causation of cancer; impact of co-morbidities, race, geographic, socioeconomic, and provider-related factors on access, diagnosis, and treatment; outcomes and end results of cancer treatment; and cancer control initiatives to diverse areas of cancer care disparity, public health policy, public health education, and importantly, cost-effectiveness of cancer care. Thus, it is not surprising that cancer registries have increasingly become indispensable parts of local, national, and international cancer control programs, and it is certain that cancer registries will continue to be extraordinary resources of information for clinicians, researchers, scientists, policy makers, and the public in our fight against cancer.

Linking the US transplant registry to administrative claims data: expanding the potential of transplant research

OBJECTIVE

In the United States, data on transplanted and waitlisted patients collected by the Organ Procurement and Transplantation Network (OPTN) have been widely used in transplantation research. Administrative claims data, collected by health plans for reimbursement purposes, are also commonly used in health-services research. This study linked OPTN and private payer claims data to assess the relationship between data elements common to both sources.

METHODS

All transplanted or waitlisted patients in the registry were considered for inclusion. A multistep match algorithm was employed to link OPTN and payer data from years 1995 to 2004. Variables common to both datasets that contained relevant information for similar time periods were compared.

RESULTS

A total of 21,419 solid organ transplant recipients and 8808 waitlist patients were included in the final linked database. Organ type and demographic variable distributions in the linked dataset were similar to the overall OPTN database. Using claims as the reference group, sensitivity and specificity values were on average 0.72 and 0.69, respectively, and were highest for the indicators of immunosuppression use at discharge and follow-up.

CONCLUSION

This comparison of payer data with information reported by transplant centers to the OPTN provides important insight into the value of both data sources. Using administrative claims to augment the registry data with utilization and cost information will be useful for evaluation of both economic and clinical endpoints in solid organ transplantation.

Population-based assessment of hospitalizations for neutropenia from chemotherapy in older adults with non-Hodgkin's lymphoma (United States)

OBJECTIVE

To study neutropenia hospitalization (NH) incidence and risk factors in a population-based sample of older adults with non-Hodgkin's lymphoma (NHL) and evaluate the validity of inferences from Surveillance, Epidemiology and End Results (SEER)-Medicare linked databases.

METHODS

NHL cases receiving first-course chemotherapy were identified from Iowa SEER-Medicare. Survival methods evaluated NH risk factors. Medical record and Medicare claims data on chemotherapy and NH were compared.

RESULTS

Of 761 subjects, 165 (21.7%, 95% CI: 18.8, 24.6) were hospitalized for neutropenia. Of those hospitalized, 41% were hospitalized in cycle 1 and 22% in cycle 2. Significant multivariable risk factors for NH were diffuse large cell histology, renal disease, Charlson comorbidity index, and anthracycline chemotherapy but not patient age. Medicare and medical records agreed on month of chemotherapy initiation 95% of the time and chemotherapy type 95% of the time. ICD-9 code 288.0 sensitivity for NH was 80%.

CONCLUSIONS

Neutropenia hospitalizations were common in the first 2 chemotherapy cycles, especially among older adults with comorbidity. Findings conflict with a prior medical records study in which age was a risk factor for NH and dose intensity a negative confounder. Valid inferences about age effects on chemotherapy toxicity require more clinical detail than is available in administrative data.

Consistency and accuracy of diagnostic cancer codes generated by automated registration: comparison with manual registration

Background

Automated procedures are increasingly used in cancer registration, and it is important that the data produced are systematically checked for consistency and accuracy. We evaluated an automated procedure for cancer registration adopted by the Lombardy Cancer Registry in 1997, comparing automatically-generated diagnostic codes with those produced manually over one year (1997).

Methods

The automatically generated cancer cases were produced by Open Registry algorithms. For manual registration, trained staff consulted clinical records, pathology reports and death certificates. The social security code, present and checked in both databases in all cases, was used to match the files in the automatic and manual databases. The cancer cases generated by the two methods were compared by manual revision.

Results

The automated procedure generated 5027 cases: 2959 (59%) were accepted automatically and 2068 (41%) were flagged for manual checking. Among the cases accepted automatically, discrepancies in data items (surname, first name, sex and date of birth) constituted 8.5% of cases, and discrepancies in the first three digits of the ICD-9 code constituted 1.6%. Among flagged cases, cancers of female genital tract, hematopoietic system, metastatic and ill-defined sites, and oropharynx predominated. The usual reasons were use of specific vs. generic codes, presence of multiple primaries, and use of extranodal vs. nodal codes for lymphomas. The percentage of automatically accepted cases ranged from 83% for breast and thyroid cancers to 13% for metastatic and ill-defined cancer sites.

Conclusion

Since 59% of cases were accepted automatically and contained relatively few, mostly trivial discrepancies, the automatic procedure is efficient for routine case generation effectively cutting the workload required for routine case checking by this amount. Among cases not accepted automatically, discrepancies were mainly due to variations in coding practice.

The added value of claims for cancer surveillance: results of varying case definitions

OBJECTIVE

As cancer diagnosis and treatment has moved to the outpatient healthcare setting, traditional cancer surveillance tools are less effective for complete and unbiased capture of incident cases. This study evaluates the potential for Medicare data to supplement cancer surveillance in a unique manner by using a standard that is independent of a central cancer registry.

DESIGN

State cancer registry records were matched with Medicare data. Case validation included inpatient record abstraction combined with a mail/telephone survey of treating physicians. The positive predictive value (PPV), sensitivity (capture rate), and potential additional cases were calculated for 6 Medicare claims-based case definitions.

RESULTS

The PPV varied according to cancer site and definition, ranging from 70%-97% (prostate) to 87%-98% (breast). Sensitivity varied inversely with PPV, ranging from 51%-94% (breast) to 10%-88% (lung). The most important factors that predicted being missed by the registry were having no admission to an ACOS-certified hospital and no surgical treatment.

CONCLUSION

Medicare data represent a valid resource for supplementing state cancer registries in surveillance efforts. This potential is especially applicable to cancers predominantly diagnosed and treated outside the hospital setting.

Using Medicare Data to Estimate the Number of Cases Missed by a Cancer Registry

BACKGROUND

Cancer surveillance is essential for assessing patterns of cancer occurrence. State cancer registries do not capture all available cases potentially biasing results. Secondary data may be useful in identifying new cases and estimating the number of cases missed.

OBJECTIVE

We sought to create 2 distinct data sources from Medicare claims to use in combination with registry data as 3 sources for a capture-recapture analysis to estimate the capture rate and bias in capture of a statewide cancer registry.

METHODS

Data from the Virginia cancer registry (Registry) were merged with Medicare inpatient (Part A) as well as Medicare outpatient and physician claims (Part B) to provide 3 sources to estimate missing cases. A 3-source loglinear model was used to estimate the number of missing cancer cases for breast, lung, colorectal, and prostate cancer. Models included main effects and interactions. Additional analysis looked at the effect of demographic and comorbidity variables.

RESULTS

Loglinear models demonstrated mostly positive dependence between the 3 sources, implying that 2-source models would underestimate missing cases and overestimate capture rates. Using capture-recapture estimates of total number of cancer cases as the denominator, capture rates for Registry ranged from 59% (colorectal) to 74% (lung). When the aggregate of cases found by either Medicare or Registry were used the capture rates ranged from 74% (prostate) to 89% (breast). Further analysis indicated that capture rates differed by demographic characteristics.

CONCLUSION

We conclude that Medicare claims are useful to supplement a Registry, estimate the number of missing cases, and assess bias in capture.

Occurrence of cancer among people with mental health claims in an insured population

OBJECTIVE

The objective of this research is to determine whether people with mental disorders are at increased risk for the subsequent development of malignancies compared with people without mental disorders.

METHODS

This is a retrospective cohort study of administrative claims data. The study population included 722,139 adults who filed at least one medical claim from 1989 to 1993. The mental disorder cohort included people with a) one psychiatric hospitalization, b) one outpatient psychiatrist visit, or c) two outpatient mental health claims occurring at least 6 months before a cancer claim. The controls were subjects filing claims for medical services who had no mental health visits. We calculated age-stratified odds ratios (ORs) for development of malignancy.

RESULTS

People with mental disorders were no more or less likely to develop a malignancy than those without after adjusting for age (women: OR, 1.03; 95% confidence interval [CI], 0.95-1.12; men: OR, 1.10; 95% CI, 0.97-1.24). People with mental disorders, however, developed cancer at younger ages and had increased odds of primary central nervous system tumors (women: OR, 2.12; 95% CI, 1.40-3.21; men: OR, 2.09; 95% CI, 1.22-3.59) and respiratory system cancers (women: OR, 1.57; 95% CI, 1.13-2.19; men: OR, 1.52; 95% CI, 1.09-2.12).

CONCLUSIONS

Insured people with mental disorder claims had an increased risk of certain malignancies and developed malignancies at younger ages. The increased odds of respiratory tumors are likely secondary to increased rates of smoking among people with mental disorders and support use of smoking cessation interventions in this population. The increased odds for brain tumors may reflect only the early presence of mental symptoms, or a true association between the two conditions. Further study of these findings is mandated.

Relationship between depression and pancreatic cancer in the general population

OBJECTIVE

Prior research suggesting a relationship between pancreatic cancer and depression conducted on clinical populations has been subject to recall bias. We reexamined this association using longitudinal population-based data.

METHODS

This was a retrospective cohort study using longitudinal insurance claims data.

RESULTS

Men with mental disorders were more likely to develop pancreatic cancer than those without psychiatric claims (odds ratio 2.4, confidence interval 1.15-4.78). Depression more commonly preceded pancreatic cancer than it did other gastrointestinal malignancies (odds ratio 4.6, confidence interval 1.07-19.4) or all other cancers (odds ratio 4.1, confidence interval 1.05-16.0).

CONCLUSIONS

Depression and pancreatic cancer are associated in the general population.

A practical method of linking data from Medicare claims and a comprehensive electronic medical records system

BACKGROUND

Linking administrative and clinical databases provides opportunities for richer studies to improve healthcare, but linkage may require sophisticated algorithms. Linking US Medicare data with large databases used for everyday clinical practice is seldom described in detail in medical literature.

OBJECTIVES

Test a deterministic method of linking data from a local electronic medical records system to Medicare data, and report specific details of the algorithm used as well as lessons learned from the linkage process.

SUBJECTS

Medicare beneficiaries with medical encounters in selected Indiana counties in the 5-year period ending in 1999.

RESULTS

For 6,388 beneficiaries with Medicare data indicating inpatient encounters in the system, 98% had links to the clinical database. Of 7,231 patients hospitalized and registered in the local clinical system, 86% contained a link to Medicare data, and 69% contained a link even without using Social security number (SSN) as an identifier. Medicare data that conflicted with local hospital records by indicating no local hospitalization occurred in 1.8%. More than 2,000 claims contained hospital identifiers that did not exist in the hospital codebook.

CONCLUSIONS

Details of a practical, deterministic method of linking Medicare claims to a large electronic records system have been applied and described. Most records were linked without SSN. A variety of inconsistencies were found and these, along with missing or incomplete data, can influence linking. Integrity of specific variables must be assessed carefully.

Evaluation of an algorithm to identify incident breast cancer cases using DRGs data

Hospital databases have the potential to be inexpensive, timely and nationally representative sources of information about cancer. This study examines the utility of the French hospital database adapted from the Diagnosis Related Group (DRG) classification and named 'Programme de médicalisation des systèmes d'information (PMSI)', as an independent source to identify incident cancer cases. From the 19 679 women hospitalized and treated in 1998 in the public hospitals of the Somme area in France, we identified those diagnosed with breast cancer in the PMSI database. These women were matched with women in the cancer registry of the Somme area who had been diagnosed with breast cancer in 1998. An algorithm was used to identify cancer-related diagnoses and procedures reported to PMSI. The sensitivity, specificity and positive predictive value (PPV) of the PMSI database were calculated using the cancer registry as a gold standard. The PMSI database had 85% sensitivity, 99.9% specificity and 97% PPV for women hospitalized with breast cancer as a principal diagnosis. The sensitivity was higher by 9% for hospitalization with breast cancer as a secondary diagnosis but had a lower PPV (78%). In conclusion, the PMSI database seems to offer an interesting potential to assess breast cancer incidence, because of its high sensitivity, in particular when secondary diagnosis was considered, and its very high specificity and PPV. However, these preliminary results need to be confirmed by other studies in France before such databases are used, particularly in areas without cancer registries.

Using Medicare claims to identify second primary cancers and recurrences in order to supplement a cancer registry

BACKGROUND AND OBJECTIVE

The purpose of this study was to use Medicare claims to develop models to assist cancer registries in identifying cancer patients with second primaries or recurrences (an "event").

METHODS

Medicare inpatient and Part B data were merged with a cancer registry for patients first diagnosed in 1993-1994. Logistic regression was used to model the occurrence of an event at least 1 year after initial diagnosis, and to identify factors that could discriminate between recurrences and second primaries.

RESULTS

Predictors of an event included an inpatient cancer diagnosis, cancer diagnosis different from the initial diagnosis from any source, and radiation or surgery claims in Part B. The ROC curve area was 0.90 with all Medicare data; 0.84 when restricted to inpatient data. A cancer diagnosis different from the initial diagnosis or having surgery predicted a second primary; regional or distant stage disease, diagnosis of secondary malignancy, or an inpatient diagnosis of primary cancer in a position other than principal predicted recurrence.

CONCLUSIONS

Although Medicare claims have not been evaluated as a stand-alone system to identify second primaries and recurrences, Medicare claims may be useful tools for Cancer Registries in case ascertainment and follow-up.

Ability of Medicaid claims data to identify incident cases of breast cancer in the Ohio Medicaid population

BACKGROUND

The use of Medicaid data to study cancer-related outcomes would be highly desirable. However, the accuracy of Medicaid claims data in the identification of incident cases of breast cancer is unknown.

OBJECTIVES

(1) To estimate the sensitivity of Medicaid claims data for case ascertainment of breast cancer, and (2) to determine the positive predictive value (PPV) of diagnostic and procedure codes retrieved from Medicaid claims, using the Ohio Cancer Incidence Surveillance System (OCISS) as the gold standard.

METHODS

The study used the linked OCISS and Medicaid enrollment files, 1997-1998 (n = 1,648). The claims search yielded 2,635 incident cases, of which 1,132 were also identified through the OCISS-Medicaid files. Sensitivity and PPV of Medicaid data were calculated in subgroups of the population.

RESULTS

The overall sensitivity was 68.7 percent, but varied greatly across the subgroups of the population. It was lower among women enrolled in Medicaid only for part of the study year than those enrolled in Medicaid for 12 months of the study year (56.7 percent and 78.0 percent respectively, p < 0.0001), and lower among those who are dual Medicare-Medicaid eligible compared to those not participating in the Medicare program (63.1 percent and 78.6 percent respectively, p < 0.0001). The overall PPV was 43.0 percent, increasing up to 86.6 percent in the presence of procedure codes indicating the presence of mastectomy and lumpectomy, in addition to that of breast cancer diagnosis.

CONCLUSIONS

The sensitivity of Medicaid claims for case ascertainment of breast cancer is somewhat low, but improves considerably when accounting for women enrolled in Medicaid for the entire duration of the study year. The PPV is poor due to a high rate of false positives. The higher PPV obtained in the presence of procedure codes, in addition to diagnosis codes, will help researchers to correctly identify incident cases of breast cancer using Medicaid claims data.

Quality of breast cancer care: what do we know?

PURPOSE

To critically review studies that describe patterns of care for breast cancer patients and to examine the data sources used for case identification and determining patterns of care.

METHODS

We searched the MEDLINE database (National Library of Medicine, Bethesda, MD) in August 2001 for studies of breast cancer care published from January 1985 to June 2001. Thirty-eight articles, describing 32 studies, met the inclusion criteria for this review.

RESULTS

According to the patterns of care literature, approximately 10% of women do not have an axillary lymph node dissection, 11% to 26% do not have their hormone receptor status reported, 20% do not receive radiation after breast-conserving surgery, and 30% to 70% of women with lymph node-positive breast cancer are not prescribed tamoxifen. Twenty-five (78%) of the studies relied on cancer registries for case identification. Cancer registries (47%) and the medical record (38%) were the most frequent sources of data on process of care. Twenty percent of the articles reported using more than one data source to determine patterns of care.

CONCLUSION

Although more patterns of care research has taken place in breast cancer than in any other oncologic condition, we found the available data had many limitations. These limitations highlight the challenges of quality-of-care research. To track changes in the quality of cancer care that may result from our rapidly transforming health care system, we need reliable data on the quality of current practice.

Validity of cancer registry data for measuring the quality of breast cancer care

BACKGROUND

Various groups have called for a national system to monitor the quality of cancer care. The validity of cancer registry data for quality of cancer care has not been well studied. We investigated the validity of such information in the California Cancer Registry.

METHODS

We compared registry data associated with care with data abstracted from the medical records of patients diagnosed with breast cancer. We also calculated a quality score for each subject by determining the proportion of four evidence-based quality indicators that were met and then compared overall quality scores obtained from registry and medical record data. All statistical tests were two-sided.

RESULTS

Records of 304 patients were studied. Compared with the medical record data gold standard, the accuracy of registry data was higher for hospital-based services (sensitivity = 95.0% for mastectomy, 94.9% for lumpectomy, and 95.9% for lymph node dissection) than for ambulatory services (sensitivity = 9.8% for biopsy, 72.2% for radiation therapy, 55.6% for chemotherapy, and 36.2% for hormone therapy). On average, quality scores calculated from registry data were 11 percentage points (95% confidence interval [CI] = 9 to 13 percentage points, P<.001) lower than those calculated from medical record data. Quality scores calculated from registry data were 5 percentage points (95% CI = 3 to 7 percentage points) lower for patients with stage I breast cancer, 16 percentage points (95% CI = 12 to 20 percentage points) lower for patients with stage II breast cancer, and 20 percentage points (95% CI = 8 to 32 percentage points) lower for patients with stage III breast cancer than were corresponding scores calculated from medical record data (all P<.001). The greater difference in quality scores for stage II and III patients revealed that disease severity and setting of care affected the validity of registry data.

CONCLUSIONS

Cancer registry data for quality measurement may not be valid for all care settings, but registries could provide the infrastructure for collecting data on the quality of cancer care. We urge that funding be increased to augment data collection by cancer registries.

Observation interval for evaluating the costs of surgical interventions for older women with a new diagnosis of breast cancer

OBJECTIVE

To estimate the episodic costs of surgical treatments for breast cancer.

METHODS

The surgical treatment period as the 6 weeks following diagnosis is defined. Using a sample of 205 women aged 65 and older and their Medicare claim files, the cost of treatment is estimated and the progression from first to subsequent surgical procedures during the 6-week interval is demonstrated with a decision tree. Two equations are then estimated: the probability of mastectomy versus breast conserving surgery (BCS) as first surgery using Probit regression and the log of total charges using a generalized linear regression model.

RESULTS

It was found that only stage predicts the probability of mastectomy versus BCS and that 54% of women receiving BCS undergo a second surgery. Once all treatments in the initial surgical period are accounted, the difference between the adjusted cost of mastectomy alone and BCS followed by a second surgery was not statistically significant. Only a successful first BCS is statistically significantly (P <0.05) less costly than a mastectomy alone ($4,955 vs. $9,049).

CONCLUSIONS

By defining a 6-week surgical treatment episode it is shown that BCS followed by subsequent surgeries is the more costly option for initial treatment. Given the high prevalence of second surgeries, previous work may have underestimated the costs of surgical interventions for breast cancer.

Multivariate matching and bias reduction in the surgical outcomes study

BACKGROUND

Outcomes studies often need a level of detail that is not present in administrative data, therefore requiring abstraction of medical charts. Case-control methods may be used to improve statistical power and reduce abstraction costs, but limitations of exact matching often preclude the use of many covariates. Unlike exact matching, multivariate matching may allow cases to be matched simultaneously on hundreds of covariates.

OBJECTIVES

To develop multivariate matched case-control pairs in a study of death after surgery in the Medicare population.

RESEARCH DESIGN

Using 830 randomly selected index cases of patients who died within 60 days from admission, controls were found who did not die within that time period, matching on risk for death and other patient characteristics with up to 173 variables used simultaneously in the matching algorithms.

SUBJECTS

General and orthopedic Medicare surgical cases in Pennsylvania from 1995 to 1996. Controls were either selected from across the entire state (108,765 possible subjects), or from within the same hospital as the case.

MEASURES

Percent bias reduction and the average difference between cases and controls in units of standard deviations.

RESULTS

Matched controls were far more similar to cases (deaths) upon admission to the hospital than typical patients, both in statewide and within hospital matches. Bias reduction was usually greater than 50% and often approached 100%. The difference between cases and matched controls for most variables was usually below 0.2 SD.

CONCLUSIONS

Multivariate matching methods may aid in conducting studies with Medicare claims records by improving the quality of matches, thereby achieving a better understanding of the etiology of outcomes.

Administrative data for public health surveillance and planning

Electronically available administrative data are increasingly used by public health researchers and planners. The validity of the data source has been established, and its strengths and weaknesses relative to data abstracted from medical records and obtained via survey are documented. Administrative data are available from a variety of state, federal, and private sources and can, in many cases, be combined. As a tool for planning and surveillance, administrative data show great promise: They contain consistent elements, are available in a timely manner, and provide information about large numbers of individuals. Because they are available in an electronic format, they are relatively inexpensive to obtain and use. In the United States, however, there is no administrative data set covering the entire population. Although Medicare provides health care for an estimated 96% of the elderly, age 65 years and older, there is no comparable source for those under 65.