Dr. Schneeweiss has received consulting fees from Booz & Company (less than $10,000) and WHISCON, LLC (more than $10,000); he has received research grants from Pfizer, Novartis, and Boehringer Ingelheim.
Jessica A. Myers,
Brigham and Women's Hospital, Boston, Massachussetts
Dr. Solomon has received research support from Abbott Immunology, Amgen, Eli Lilly, and the Consortium of Rheumatology Researchers of North America (CORRONA) and an educational grant from Bristol-Myers Squibb; he serves as an unpaid member of an Executive Committee and a Data Safety Monitoring Board for two analgesic trials sponsored by Pfizer.
Previous study findings have suggested that patients with chronic diseases such as rheumatoid arthritis (RA) do not receive optimal preventive medical services, including cancer screening tests. This study was undertaken to evaluate cancer screening rates in RA patients compared to non-RA control populations.
Using data from a large US commercial insurance plan, we examined rates of screening tests for cervical, breast, and colon cancer in patients with RA compared to control subjects without RA (non-RA controls) or control subjects with hypertension. Individuals were included in the RA cohort if they had at least 2 visits coded for a diagnosis of RA and had received at least 1 prescription for a disease-modifying antirheumatic drug during the study period. Multivariable Cox proportional hazards models were used to compare the rates of different cancer screening tests between RA patients and non-RA controls.
RA patients (n = 13,314) and control subjects (non-RA and hypertension controls) (n = 212,324) were screened, on average, once every 3 years for cervical cancer and once every 2 years for breast cancer during the followup period (mean 2.3 years of followup). In the age-adjusted Cox regression model, women with RA were more likely to receive ≥1 Papanicolaou smear (hazard ratio [HR] 1.21, 95% confidence interval [95% CI] 1.17–1.24), ≥1 mammogram (HR 1.49, 95% CI 1.45–1.53), and ≥1 colonoscopy (HR 1.69, 95% CI 1.61–1.77) compared to female non-RA control subjects. Men with RA were also more likely to receive at least 1 colonoscopy (HR 1.52, 95% CI 1.40–1.64) than were male non-RA control subjects. These results were robust in multivariable analyses adjusted for age, number of physician visits, percentage of visits made to primary care physicians, and the Charlson Comorbidity Index.
Patients with RA did not appear to be at risk for receiving fewer cancer screening tests when compared to individuals without RA. The majority of both RA patients and non-RA control subjects were screened regularly for cervical, breast, and colon cancer, in accordance with current recommendations.
Patients with rheumatoid arthritis (RA) are known to have a decreased life expectancy, even with early and aggressive treatment, compared to the general population (1–4). One of the main causes of death in RA patients is cancer (2, 4). Screening and early detection of some common cancers can improve morbidity and mortality; thus, both the United States Preventive Services Task Force (USPSTF) and the American Cancer Society recommend that most adults be regularly screened for cervical, breast, and colon cancer (5, 6). The USPSTF recommends screening for cervical cancer every 3 years with the Papanicolaou (Pap) smear in women ages 21–65 years. For women ages 30–65 years who want to lengthen the screening interval, screening with a combination of the Pap smear and human papillomavirus (HPV) testing every 5 years is recommended. The USPSTF recommends biennial screening with mammography for breast cancer in women ages 50–74 years, and fecal occult blood (FOB) testing, sigmoidoscopy, or colonoscopy for colorectal cancer in adults ages 50–75 years. The guidelines from the American Cancer Society are similar to those from the USPSTF, except that the American Cancer Society recommends that mammography be performed annually in women ages 40 years and older.
Prior research has raised concerns that patients with chronic diseases such as RA do not receive optimal preventive medical services, including cancer screening tests (7–10). Data from a historical cohort of 1,335 adults with RA enrolled in a national fee-for-service insurance plan showed that RA patients received inadequate quality of care, including a lack of health maintenance tests such as mammography and Pap smears (8). Another study showed that 20–35% of older Americans with arthritis in the US Medicare beneficiary population never underwent mammography or colonoscopy over a period of 5 years (10). While it is important to recognize chronic diseases such as RA as a potential barrier to preventive medical services and to raise awareness of the importance of cancer screening, most of the published studies did not compare cancer screening rates in patients with RA with the screening rates in a non-RA population.
A previous analysis from the Nurses' Health Study suggested that cancer screening practices were similar among women with RA compared to women without RA (11). It is unclear whether that study yielded different results because of the characteristics of the study population, comprising all health care professionals. However, it is also possible that patients with RA are evaluated more often by physicians, and thus go through more medical examinations, including cancer screening tests, than would normally be expected.
In this study, we reviewed the claims data from a large cohort of subjects in a health care utilization database. The aims of our study were 1) to examine cancer screening rates in patients with RA compared to subjects without RA (non-RA controls), and 2) to assess the adherence to current adult cancer screening guidelines recommended by the USPSTF and the American Cancer Society (5, 6) in both groups.
PATIENTS AND METHODS
We conducted a cohort study using administrative claims data for the period January 1, 2001 through June 30, 2008 from a large US commercial health insurance plan (Blue Cross/Blue Shield), which provides medical and pharmacy coverage for more than 28 million fully-insured subscribers (primarily working adults and their family members, as well as a small Medicare population) in 14 plans across the US. This database contains longitudinal claims information, including data on medical diagnoses, procedures, hospitalizations, physician visits, and pharmacy-dispensed prescriptions. Personal identifiers were removed from the data set before analysis, to protect subject confidentiality, and therefore patient informed consent was not required. The study protocol was approved by the Partners Healthcare Institutional Review Board.
Adult subjects whose claims data showed at least 2 visits coded with an International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) code for RA (714.xx) were identified for the RA cohort. Subjects were entered into the RA cohort at the first receipt of a disease-modifying antirheumatic drug (DMARD) prescription. Thus, all subjects in the RA cohort were required to have received 2 coded diagnoses of RA and at least 1 filled prescription for a DMARD at the start of followup. A previous validation study showed that RA patients can be accurately identified using a combination of diagnosis codes and DMARD prescriptions in claims data (12).
As controls, we identified 2 cohorts of subjects without RA who were older than age 18 years. The first comparison cohort, described herein as the non-RA cohort, consisted of adults who had never received a diagnosis of RA during the study period. Followup for the non-RA cohort began at a randomly selected index date, and subjects were required to have at least 12 months of continuous health plan eligibility before the start of followup and at least 12 months of followup time after the index date. Subjects were then followed up until the first of any of the following censoring events: development of solid tumors, including breast, cervical, prostate, and colorectal cancer, as well as human immunodeficiency virus (HIV) infection, loss of eligibility, admission to a nursing home, end of study period, or death.
To compare the cancer screening rates between patients with RA and individuals with another chronic condition that required regular visits to a physician, subjects for whom at least 2 visits were coded for hypertension (ICD-9 code 401.X) were selected. Followup for this secondary control cohort began at the date of the second diagnosis of hypertension. Subjects in this cohort were required to meet eligibility criteria similar to those for the non-RA cohort, and were followed up until the same censoring events as described above.
Nursing home residents and subjects whose claims data included coding for solid tumors, hematologic malignancies, myelodysplastic syndrome, HIV infection, or chemotherapy were excluded from all cohorts.
Cancer screening tests.
By noting the diagnosis and procedure codes recorded in the claims database for each subject during the study period (details available from the corresponding author upon request), several screening tests for cervical cancer, breast cancer, and colon cancer, all of which were performed after the index date, were identified. Our primary outcome of interest was the Pap smear, mammogram, and both conventional and virtual computed tomography colonoscopy. As secondary outcomes, we identified the HPV DNA test, colposcopy, breast ultrasound, magnetic resonance imaging, breast biopsy, FOB test, barium enema, and flexible sigmoidoscopy as additional screening or diagnostic tests that were performed for cervical, breast, or colon cancer.
Variables potentially related to the requirement for cancer screening tests in each subject were assessed using data from the 12 months before the index date. These variables included demographic factors (age and sex), comorbidities, and health care utilization factors (number of visits to any physicians, to primary care physicians, and to rheumatologists, occurrence of acute-care hospitalizations, and number of different prescription drugs). To quantify the comorbidities in each subject, we calculated the Deyo-adapted Charlson Comorbidity Index, based on the relevant ICD-9-CM codes (13, 14). The Comorbidity Index is a summative score for 19 major medical conditions, including, among others, myocardial infarction, pulmonary disease, renal disease, hepatic disease, diabetes, cancer, and HIV infection. A score of 0 represents the absence of comorbidity, and a higher score indicates a greater number of comorbid conditions.
We compared the baseline characteristics between the RA patient cohort and non-RA control cohorts. We first estimated the crude rates of the 3 major cancer screening tests in both the RA and non-RA cohorts (along with 95% confidence intervals [95% CIs]) by calculating the number of subjects who had undergone at least 1 of these screening tests during the study period, divided by the total person-time of followup. Pap smear rates were calculated for all women ages 18 years or older, and then stratified by age. Mammogram rates were calculated for all women ages 40 years or older, and then stratified by age. For colonoscopy, either conventional or virtual, the rates were calculated for all subjects ages 50 years or older, and then stratified by age and sex. Rate ratios, with 95% CIs, were estimated by dividing the rate of the specific cancer screening test in RA patients by the rate in non-RA control subjects (15). For the secondary outcomes in which the rates of any screening or diagnostic test specific to each cancer were determined, similar analyses were carried out.
To adjust for potential confounders, separate Cox proportional hazards models were used to compare the screening test rates for each cancer type between RA patients and non-RA controls (16), with results expressed as hazard ratios (HRs). All of these data analyses were repeated in a secondary analysis for the comparison between RA patients and control subjects with hypertension. All data analyses were done using SAS statistical software (version 9.1; SAS Institute).
There were more than 1 million potentially eligible subjects in the study database. Initially, 92,827 subjects for whom there was at least 1 ICD-9-CM code indicating a diagnosis of RA after 365 days of enrollment and ∼920,697 subjects who had not received a diagnosis of RA during this time were identified. When a second diagnosis code for RA and at least 1 prescription for DMARDs were required for inclusion in the RA cohort, and when the exclusion criteria were applied, this left 23,131 RA patients and 237,933 non-RA control subjects remaining in the study cohorts. When we restricted the analysis to those subjects with at least 1 year of followup data, our final study cohorts included 13,314 RA patients and 131,989 non-RA control subjects (Figure 1). For the secondary analysis, 80,335 patients with hypertension were selected as controls.
Characteristics of the study subjects.
Table 1 presents the baseline characteristics of the subjects in our study cohorts, comprising 13,314 RA patients, 131,989 non-RA control subjects, and 80,335 control subjects with hypertension. The mean age was 52 years for the RA patients, 48 years for the non-RA control subjects, and 61 years for the control subjects with hypertension. The mean followup time was 2.54 years for the RA patients, 2.37 years for the non-RA control subjects, and 2.33 years for the control subjects with hypertension. Substantial differences in baseline characteristics were observed between the cohorts and, in particular, between the RA patients and non-RA control subjects. Overall, compared to subjects in the non-RA control cohort, there were more women in the RA cohort, and RA patients had more comorbid conditions and a greater number of physician visits. When comparing all 3 groups, the control subjects with hypertension were the oldest and had the most comorbid conditions and hospitalizations.
Table 1. Baseline characteristics of the study cohorts in the 12 months prior to the index date*
RA patients (n = 13,314)
Non-RA (n = 131,989)
Hypertension (n = 80,335)
RA = rheumatoid arthritis; PCP = primary care physician; NA = not applicable.
Followup, mean ± SD years
2.5 ± 1
2.4 ± 1
2.3 ± 1
Age, mean ± SD years
51.8 ± 12
47.8 ± 17
61.4 ± 15
Women, no. (%)
Charlson Comorbidity Index, mean ± SD
1.2 ± 1
0.2 ± 1
0.6 ± 1
Chronic kidney disease, no. (%)
Liver disease, no. (%)
Hypertension, no. (%)
Diabetes mellitus, no. (%)
Inflammatory bowel disease, no. (%)
Chronic obstructive pulmonary disease, no. (%)
Heart failure, no. (%)
Health care utilization
No. of all physician visits, mean ± SD
9.6 ± 8
3.6 ± 5
6.2 ± 6
No. of PCP visits, mean ± SD
3.9 ± 4
1.7 ± 3
3.2 ± 3
% visits to PCPs
No. of rheumatologist visits, mean ± SD
2.0 ± 3
% visits to rheumatologists
No. (%) of patients with at least 1 hospitalization
No. of prescription drugs, mean ± SD
10.1 ± 6
3.7 ± 5
5.7 ± 6
Screening test rates for cervical, breast, and colon cancer.
Among all women ages 18–69 years in the study cohorts, ∼41–55% of RA patients, ∼40–49% of non-RA control subjects, and ∼36–70% of control subjects with hypertension had undergone at least 1 Pap smear per year. Among all women ages 40 years and older, 55% of RA patients, 44% of non-RA control subjects, and 47% of control subjects with hypertension had undergone mammography at least once per year. Among all subjects ages 50 years and older, 12% of RA patients, 9.5% of non-RA control subjects, and 12% of control subjects with hypertension had undergone at least 1 colonoscopy per year. Interestingly, fewer patients were receiving FOB testing than were undergoing colonoscopy, across all 3 groups. Overall, the rates of Pap testing, mammography, FOB testing, and colonoscopy were lower in non-RA control subjects compared to all RA patients and control subjects with hypertension.
Table 2 presents the crude rates of the major cancer screening tests (i.e., the rates in subjects having undergone at least 1 test per 1,000 person-years) among the subjects in all 3 cohorts. Assuming the test rates remained constant over time, we estimated that ∼70–80% of all women ages 18–69 years would have undergone at least 1 Pap smear every 3 years, regardless of RA status. For breast cancer screening, performed in women ages ≥40 years, 60% of female control subjects (both non-RA subjects and those with hypertension) and 70% of female patients with RA would have undergone mammography at least once every other year. For colon cancer screening, performed among both male and female subjects ages 50 years and older, 70% of RA patients, 70% of non-RA control subjects, and 61% of the control subjects with hypertension would have undergone a colonoscopy every 10 years.
Table 2. Crude rates of cancer screening tests by sex and age among RA patients and subjects in the control cohorts*
Crude rate (95% CI)
Crude rate (95% CI)
Crude rate (95% CI)
Values are the crude rate (95% confidence interval [95% CI]) per 1,000 person-years for having undergone at least 1 of these cancer screening tests during followup in the rheumatoid arthritis (RA) patients, non-RA control subjects, and control subjects with hypertension.
All women ages ≥18 years
Women ages 18–29 years
Women ages 30–69 years
Women ages ≥70 years
All women ages ≥40 years
Fecal occult blood test
All subjects ages ≥50 years
All subjects ages ≥50 years
Among women ages 18–39 years who had undergone at least 1 Pap smear during the followup period, the average number of days between the first Pap smear and the second Pap smear was 437 days for RA patients, 423 days for non-RA control subjects, and 439 for control subjects with hypertension. Among women ages 40 years and older who had undergone mammography at least once during the followup period, the average number of days between the first mammogram and second mammogram was 414 days for RA patients, 409 days for non-RA control subjects, and 406 days for control subjects with hypertension.
In the age-adjusted Cox regression model, women with RA were more likely to have received ≥1 Pap smear (HR 1.21, 95% CI 1.17–1.24), ≥1 mammogram (HR 1.49, 95% CI 1.45–1.53), and ≥1 colonoscopy (HR 1.69, 95% CI 1.61–1.77) when compared to female non-RA control subjects. Men with RA were also more likely to have received at least 1 colonoscopy (HR 1.52, 95% CI 1.40–1.64) when compared to male non-RA control subjects.
The results were robust in multivariable analyses that were adjusted for age, number of physician visits, percentage of visits made to primary care physicians, and the Charlson Comorbidity Index. In contrast, there were no significant differences in the likelihood of having received at least 1 Pap smear, mammogram, or colonoscopy during the followup between patients with RA and control subjects with hypertension, in either age-adjusted or multivariable Cox regression models (Table 3).
Table 3. Likelihood of having at least 1 cancer screening test among patients with RA when compared with non-RA control subjects or control subjects with hypertension*
Values are the adjusted hazard ratio (HR) (95% confidence interval [95% CI]) for the likelihood of undergoing at least 1 of these cancer screening tests during followup in patients with rheumatoid arthritis (RA) relative to the control cohorts.
Adjusted for age, numbers of physician visits, percentage of visits made to primary care physicians, and Charlson Comorbidity Index.
Our study shows that the majority of patients with RA, albeit not all, were screened regularly for cervical, breast, and colon cancer, in accordance with current recommendations, and the rates of cancer screening tests were comparable with those in subjects without RA in the population and with the screening rates in subjects with a chronic condition such as hypertension. These results are somewhat different from the findings in previous studies, which suggested that preventive health service utilization was inadequate in patients with chronic diseases, including RA (8–10, 17–19). However, it is important to note that some of the previously published studies did not compare cancer screening rates in RA patients with those in a non-RA group.
In a previous study of 1,335 adults with RA enrolled in a US national fee-for-service insurance plan (8), the results indicated that RA patients received inadequate quality of care, including a lack of health maintenance tests such as mammography and Pap smears. In addition, patients who had contact with a primary care physician but had no contact with a relevant specialist had better health maintenance service than those patients who did not have contact with either primary or specialist doctors. In a small, single institution–based study (18), 68% of women younger than age 50 years had undergone mammography at least once by the end of 2 years, and 33% of women age 50 years or older had undergone mammography at least once by 1 year. In women without a history of hysterectomy, the probability of having a Pap smear within 3 years was 77% among women with RA, which is consistent with our results. The mean age of that study population was higher compared to the mean age of our study subjects. These 2 previous studies did not have a comparison group, and therefore their results could not elucidate whether the problem of inadequate (i.e., <100% utilization) health maintenance service was specific to patients with RA.
A study using claims data from a random sample of Medicare enrollees (5% of enrollees) showed that RA patients were significantly less likely to receive evaluation for hyperlipidemia or screening for malignancy compared to patients with osteoarthritis (10). Although that study included a large number of patients who were randomly selected from the US Medicare database, all of the patients were age 65 years or older. Thus, those results with regard to access to health care may not be generalizable to a broad range of patients.
Several characteristics of our study may further explain the differences in our results. First, unlike many previous studies, the present study compared RA patients with non-RA control subjects. Similar findings were noted in a previous study comparing RA patients with non-RA subjects in the Nurses' Health Study, which showed that care to prevent acute myocardial infarction and cancer screening practices were similar among women with RA compared with women without RA, except for the frequency of aspirin use (11). Furthermore, we conducted a secondary analysis comparing RA patients with control subjects with a different chronic condition, hypertension, so that both groups would have a similar health care utilization pattern.
In addition, since most preventive services and tests are provided by primary care physicians, we also collected data on the total number of visits to physicians and the percentage of visits made to rheumatologists as well as primary care physicians. As expected, RA patients had a higher number of visits to both primary care physicians and rheumatologists when compared to non-RA controls. The multivariable Cox models comparing RA patients with non-RA control subjects were adjusted for this difference, as well as for the number of comorbidities.
Second, our study was based on data from a large commercial health insurance plan that insures mainly working adults and their families and a small Medicare managed care population (∼10%) across the US. Reimbursement for preventive services, including screening tests for cervical, breast, and colon cancer, is provided both by the commercial health plan and by Medicare. The availability of health insurance as well as access to medical care in the study population may be different from that in other study populations. Therefore, our results may not be generalizable to persons without health insurance coverage for preventive services.
Third, there could be misclassification of the diagnoses of RA and hypertension, as we mainly relied on claims data to identify these diagnoses. However, the ICD-9 codes for both RA and hypertension have been previously validated (12, 20).
Fourth, although we excluded patients with a history of any malignancy and censored them if they had a diagnosis of any cancer during followup, it is still possible that some of the tests might have been performed for diagnostic rather than screening purposes.
The results of our study may have clinical implications for preventive care of RA patients. Although the RA patients in this study cohort received similar screening tests for cancer at a similar rate as that observed in the 2 non-RA cohorts, it is not clear what the appropriate rates of cancer screening are for patients with RA. Given the increased risk of some cancers in patients with RA and given existing concerns regarding an association between various types of RA treatment and malignancy (2, 21, 22), it may be worth investigating the effectiveness of current cancer screening guidelines in patients with RA and in subgroups of patients receiving specific treatments.
In conclusion, patients with RA did not appear to be at risk for receiving fewer cancer screening tests when compared to subjects without RA. The majority of both RA patients and non-RA control subjects appeared to undergo screening for specific cancers on a regular basis, as suggested by the current recommendations, although the results may not be generalizable to persons without medical insurance. Continuous efforts should be made to improve and maintain both patients' and physicians' awareness of the importance of preventive health services in patients with chronic diseases such as RA.
All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved the final version to be published. Dr. Kim had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Study conception and design. Kim, Schneeweiss, Liu, Solomon.
Acquisition of data. Kim, Liu.
Analysis and interpretation of data. Kim, Schneeweiss, Myers, Liu, Solomon.