Abstract
Purpose:
Cancer treatment may be affected by comorbidities; however, studies are limited. The purpose of this study is to examine the frequency of comorbidities at visits by patients with breast, prostate, colorectal, and lung cancer and to estimate frequency of a prescription for antineoplastic drugs being included in the treatment received at visits by patients with cancer and concomitant comorbidities.
Methods:
We used nationally representative data on visits to office-based physicians from the 2010–2016 National Ambulatory Medical Care Survey and selected visits by adults with breast, prostate, colorectal, or lung cancer (n=4,672). Nineteen comorbid conditions were examined. Descriptive statistics were calculated for visits by cancer patients with 0, 1, and ≥2 comorbidities.
Results:
From 2010–2016, a total of 10.2 million physician office visits were made annually by adult patients with breast, prostate, colorectal, or lung cancer. Among US visits by adult patients with breast, prostate, colorectal, or lung cancer, 56.3% were by patients with ≥1 comorbidity. Hypertension was the most frequently-observed comorbidity (37.7%), followed by hyperlipidemia (19.0%) and diabetes (12.3%). Antineoplastic drugs were prescribed in 33.5% of the visits and prescribed at a lower percentage among visits by cancer patients with COPD (21.3% versus 34.3% of visits by cancer patients without COPD) and heart disease (22.7% versus 34.2% of visits by cancer patients without heart disease).
Conclusion:
Our study provides information about comorbidities in cancer patients being treated by office-based physicians in an ambulatory setting.
Keywords: Ambulatory health care, cancer visits, comorbidities, antineoplastic drugs prescription
Introduction
Cancer is the second leading cause of death in the United States(1). In 2018, 1,735,350 new cancer cases and 609,640 cancer deaths were projected to occur in the United States(1). Approximately 70% of patients with cancer are ≥65 years(2). With the number of adults ≥65 expected to increase from 35 million in 2000 to 72 million by 2030(2), an increasing number of older patients are expected to be diagnosed with cancer, require treatment, and have comorbidities (i.e., coexisting medical conditions distinct from the principal cancer diagnosis)(3). Data from Medicare beneficiaries in the United States show that 25% of patients with the most common cancer diagnoses, namely, breast, prostate, lung and colorectal cancer, have ≥1 chronic condition besides the cancer diagnosis, and 15% have ≥2 chronic conditions(4).
Cancer patients with comorbidities are often not included or underrepresented in clinical trials, and guidelines may not exist to inform chemotherapy treatment decisions in cancer patients with various comorbidities, which could lead to over- or under-treatment(2, 5–7). Yet the decision-making process for chemotherapy treatment in cancer patients with comorbidities must weigh the benefits of chemotherapy with the risks of toxicity, patient tolerability and future quality of life(8). Multiple studies have demonstrated that comorbidities are relevant to the prognosis of cancer patients, and considering comorbid conditions in the management of a patient’s cancer treatment may be important for prolonging overall survival and improving quality of life(9–12). The majority of studies investigating the impact of comorbidities on the administration of chemotherapy for the treatment of cancer report that patients with comorbidities were less likely to receive chemotherapy(10, 13–18). A few studies found no difference(9, 19–21), and in one study, men with diabetes were more likely to receive chemotherapy for prostate cancer than men without comorbidities(22). However, the majority of cancer-comorbidity studies did not assess the effect of specific comorbidities, but evaluated the collective effect of comorbidities by using a score(8). Moreover, the majority of the studies are based on population-based cancer registries linked to administrative data which might lack information on common comorbidities, such as hypertension or hyperlipidemia(4). Therefore, there are gaps in the literature regarding the description of cancer and specific comorbidities, and the receipt of chemotherapy, in particular no recent data are available on characteristics of outpatient visits made by patients with cancer and comorbidities. A recent study, using data from the Medical Expenditure Panel Survey has shown that among
cancer patients, ambulatory care visits accounted for the largest portion of health care expenditures (23). Due to the resources invested in cancer care in the outpatient office setting, it is important to examine the characteristics of visits and care provided at visits made by patients with cancer.
In this study, we use nationally-representative data on visits to physician offices from 2010–2016 to evaluate the frequency of comorbidities at visits by patients with breast, prostate, colorectal, and lung cancer. Finally, we estimate frequency of a prescription for antineoplastic drugs being included in the treatment received at visits by patients with cancer and concomitant comorbidities.
Methods
Data used in this study are from the 2010–2016 National Ambulatory Medical Care Survey (NAMCS), an annual survey conducted by the National Center for Health Statistics (NCHS). It is representative of ambulatory patient visits made to nonfederal, office-based physicians in the United States (i.e., 50 states and the District of Columbia). Weighted NAMCS physician participation rates, calculated by dividing the physicians who provided data on at least one patient visit by the total number of in-scope physicians, and multiplying by 100, ranged from 47.5% in 2016 to 59.3% in 2010. Detailed information regarding the survey instrument is available elsewhere (https://www.cdc.gov/nchs/ahcd/). To increase sample size and improve reliability, NAMCS data from the years 2010–2016 were merged.
To identify mutually-exclusive visits with breast, prostate, colorectal, and lung cancer diagnosis, we selected records in which the primary diagnosis field was coded as ICD-9-CM 174–175 and ICD-10-CM C50.0-C50.9 (breast cancer), ICD-9-CM 185 and ICD-10-CM C61 (prostate cancer), ICD-9-CM 153–154 and ICD-10-CM C18.0-C18.9, C19, C20 (colorectal cancer), or ICD-9-CM 162 and ICD-10-CM C34.00, C34.01, C34.02, C34.10, C34.11, C34.12, C34.2, C34.30, C34.31, C34.32, C34.80, C34.81, C34.82, C34.90, C34.91, C34.92 (lung cancer). Only visits by adult patients aged ≥18 years were included in our sample (n=4,672). We considered a cancer visit as any visit in which the primary diagnosis was breast, prostate, colorectal, or lung cancer, regardless of the patient’s chief complaint at the visit. The demographic variables used in the analyses were age, sex, and race/ethnicity (non-Hispanic white, non-Hispanic black, Hispanic). The physician specialty variable was divided into two categories: oncologists and non-oncologists as cancer patients may see non-oncologists as they are often treated for prolonged periods of time, have frequent follow-ups and may have comorbidities that require a multidisciplinary care approach(24). Moreover, some current chemotherapy treatments are administered in ambulatory settings by nurses(25). Finally, all expected sources of payment listed in the medical chart were combined into three mutually exclusive insurance groups: private insurance, Medicare or Medicaid, and other/unknown (includes worker’s compensation, self-pay, no charge/charity, and other sources of payment). For visits with more than one expected source of payment, a single expected source of payment is used based on the following hierarchy: Medicare, Medicaid/CHIP, private insurance, worker’s compensation, self-pay, no charge/charity, other, and unknown.
NAMCS medications were coded using Lexicon Plus®, a proprietary database of Cerner Multum, Inc. Antineoplastic drugs were considered prescribed/administered if visits were included in the Cerner Multum’s Lexicon first-level therapeutic category code “20 antineoplastics”. More than one prescription for the same chemotherapy cycle may be recorded during the same visit. A list of medications included in this category is provided in Supplementary Table 1.
Nineteen comorbid conditions, defined as coexisting medical conditions distinct from the principal cancer diagnosis, were included. We selected conditions that are part of the Charlson index (26, 27) as well as additional conditions (hypertension, hyperlipidemia, depression, osteoporosis, and asthma) included in NAMCS. In addition to the diagnoses codes, NAMCS asks about whether certain conditions are currently present using a checkbox (yes/no) format. Twelve of the conditions of interest for this analysis are included under this checkbox format: arthritis, asthma, chronic kidney disease, chronic obstructive pulmonary disease, congestive heart failure, ischemic heart disease (coronary artery disease, ischemic heart disease or history of myocardial infarction), depression, diabetes, hyperlipidemia, hypertension, osteoporosis, and stroke. A checked box (i.e., response of yes) for any of these conditions on NAMCS indicates that the medical record contains documentation that the patient currently has the condition, although it did not necessarily have to be diagnosed during the current visit. As described previously, NAMCS also separately collects up to five diagnoses for the current visit, indicated by ICD-9-CM codes (NAMCS 2010–2015) and ICD-10-CM codes (NAMCS 2016). We used these diagnosis codes for the conditions included in Charlson index that do not have a checkbox (Supplementary Table 2). We grouped similar conditions from the Charlson index together (e.g., liver disease [mild] and liver disease [moderate/severe]; diabetes and diabetes with complications). Finally, a trichotomous measure was created signifying: cancer and no comorbidity (i.e., absence of all the 19 chronic conditions listed above; but still could include some other diagnosis), cancer and one comorbidity, and cancer and ≥2 comorbidities.
Descriptive estimates were generated, including overall percentages of a prescription for antineoplastic drugs being included in the treatment received at visits by cancer patients, types of cancer, comorbid conditions, and potential confounders among visits by adult patients with cancer and zero and ≥1 comorbidity. Percentages for the most common comorbidities by adult patient cancer types were also examined. Differences among subgroups were evaluated with 2-tailed t test by using P<.05 as the level of significance. One-sample t-test was performed to compare visits by patients with cancer with no comorbidity versus visits by patients with cancer with ≥ 1 comorbidity in Table 2 and Table 3.
Table 2.
Percentages of physician office visits by adult patients with breast, prostate, colorectal and lung cancer with no comorbidities and ≥1 comorbidities, by selected patient characteristics: United States, 2010–2016
| Visits by adults with breast, prostate, colorectal, and lung cancer | |||||
|---|---|---|---|---|---|
| Total | with no comorbidity | with ≥1 comorbidities | P value No comorbidity vs. ≥1 comorbidities | ||
| No.(Weighted No. in 1000s) | % (SE) | % (SE) | % (SE) | ||
| Total | 4672 (10155) | 100 | 43.7 (1.7) | 56.3 (1.7) | <.001 |
| Sex | |||||
| Female | 2630 (5812) | 57.2 (1.9) | 46.2 (2.3) | 53.8 (2.3) | 0.03 |
| Male | 2042 (4343) | 42.8 (1.9) | 40.4 (2.1) | 59.6 (2.1) | <.001 |
| Age (Mean) (SE) | 65.8 (0.4) | 62.5 (0.8) | 68.4 (0.4) | ||
| MEDIAN (IQR) | 66.4 (57.2–74.4) | 62.9 (52–72) | 68.6 (60.5–75.8) | ||
| Race/ethnicity1 | |||||
| Non-Hispanic White | 3773 (7888) | 77.7 (2.0) | 42.0 (1.9) | 58.0 (1.9) | <.001 |
| Non-Hispanic Black | 450 (966) | 9.5 (0.9) | 47.9 (4.5) | 52.1 (4.5) | >0.05 |
| Hispanic | 310 (916) | 9.0 (1.3) | 50.0 (6.1) | 49.9 (6.1) | >0.05 |
| Provider | |||||
| Oncologist2 | 2518 (5216) | 51.4 (3.5) | 45.9 (3.0) | 54.1 (3.0) | <.001 |
| Other | 2154 (4939) | 48.6 (3.5) | 41.4 (1.9) | 58.6 (1.9) | <.001 |
| Expected source of payment3 | |||||
| Private insurance | 1825 (3657) | 36.0 (1.6) | 51.9 (2.4) | 48.1 (2.4) | >0.05 |
| Medicare/Medicaid | 2519 (5668) | 55.8 (2.1) | 37.9 (2.5) | 62.1 (2.5) | <.001 |
| Other/Unknown | 328(830)* | * | 47.1 (5.2) | 52.9 (5.2) | >0.05 |
Estimate does not meet NCHS standards of reliability.
Non-Hispanic other not shown (n=139)
Oncologist includes gynecological oncology, musculoskeletal oncology, hematology/oncology, medical oncology, and surgical oncology
Expected source of payment is based on a hierarchy because more than one insurance could be selected as expected source of payment (see methods section for more information)
Percentages of total visits are presented as distributions for patient characteristics. Percentages of visits with no comorbidity and with ≥ 1 comorbidity are presented as row percentages.
Abbreviations: No., unweighted sample size; SE, standard error; IQR, interquartile range.
Source: National Ambulatory Medical Care Survey 2010–2016 (n=4,672). Numbers may not add to totals due to rounding.
Table 3.
Percentage of physician office visits with antineoplastic drugs prescription by adult patients with breast, prostate, colorectal, or lung cancer with no comorbidities and ≥1 comorbidities: United States, 2010–2016
| Visits by adults with breast, prostate, colorectal, and lung cancer receiving antineoplastic drugs | |||||
|---|---|---|---|---|---|
| Total | with no comorbidity | with ≥1 comorbidities | P value No comorbidity vs ≥1 comorbidities | ||
| No.(Weighted No. in 1000s) | % (SE) | %(SE) | % (SE) | ||
| Total | 1578 (3406) | 100 | 42.9 (2.5) | 57.1 (2.5) | <.001 |
| Sex | |||||
| Female | 1023 (2207) | 64.8 (2.45) | 42.9 (3.3) | 57.1 (3.3) | <0.01 |
| Male | 555 (1199) | 35.2 (2.45) | 42.9 (3.3) | 57.1 (3.3) | <0.01 |
| Age (Mean) (SE) | 65.0(0.6) | 61.1 (1.6) | 68.0 (0.6) | ||
| MEDIAN (IQR) | 66.2 (56.1–73.4) | 62.5(50.8–71.2) | 68.3 (60.2–75.6) | ||
| Race1 | |||||
| Non-Hispanic White | 1250 (2597) | 76.2 (2.6) | 42.4 (3.1) | 57.6 (3.1) | <0.01 |
| Non-Hispanic Black | 157 (312) | 9.2 (1.1) | 44.4 (7.1) | 55.6 (7.1) | 0.030 |
| Hispanic | 123 (333) | 9.8 (1.5) | 36.7 (8.1) | 63.3 (8.1) | 0.020 |
| Provider | |||||
| Oncologist2 | 1243 (2588) | 76.0 (3.5) | 43.5 (3.1) | 56.5 (3.1) | <0.01 |
| Other | 335 (818) | 24.0 (3.5) | 40.7 (4.2) | 59.3 (4.2) | <0.01 |
| Expected source of payment3 | |||||
| Private insurance | 648 (1263) | 37.1 (2.3) | 53.8 (3.8) | 46.2 (3.8) | >0.05 |
| Medicare/Medicaid | 817 (1870) | 54.9 (2.6) | 35.0 (4.3) | 65.0 (4.3) | <0.001 |
| Other/Unknown | 113 (274)* | * | 46.2 (7.3) | 53.8 (7.3) | >0.05 |
Estimate does not meet NCHS standards of reliability.
Non-Hispanic other not shown (n=48)
Oncologist includes gynecological oncology, musculoskeletal oncology, hematology/oncology, medical oncology, and surgical oncology
Expected source of payment is based on a hierarchy because more than one insurance could be selected as expected source of payment (see methods section for more information)
Percentages of total visits with antineoplastic drugs prescription are presented as distributions for patient characteristics. Percentages of visits with no comorbidity and with ≥ 1 comorbidity are presented as row percentages.
Abbreviations: No., unweighted sample size; SE, standard error; IQR, interquartile range.
Source: National Ambulatory Medical Care Survey 2010–2016 (n=1,578). Numbers may not add to totals due to rounding.
Data are weighted to produce national estimates that account for the stratified complex sample design of NAMCS. All estimates presented are annual averages. Percentages were suppressed if they did not meet the NCHS standards for presentation of proportions(28) and weighted population counts were flagged if their relative standard errors were > 30%. Data analyses were performed using the statistical packages SAS version 9.4 (SAS Institute, Cary, N.C.) and SAS-callable SUDAAN version 11.0 (RTI International, Research Triangle Park, N.C.).
Results
For our analysis, we selected the most common cancers in our sample: breast, prostate, colorectal and lung cancer. A total of 10.2 million physician office visits were made annually by adult patients with breast, prostate, colorectal, or lung cancer (Table 1), hereafter referred to as “cancer visits.” Visits by adults with breast, prostate, colorectal, and lung cancer represented 41.2%, 25.4%, 16.8% and 16.6% of the total visits with cancer selected for our analysis, respectively. The mean age of patients making a cancer visit was 65.8 years. Overall, more visits by cancer patients were made by women than men (57.2% vs. 42.8%, P<.001). More visits were made by non-Hispanic white, than non-Hispanic black and Hispanic adults (77.7% vs 9.5%, P<.001 and 77.7% vs 9.0%, P<.001).
Table 1.
Characteristics of physician office visits by adult patients with breast, prostate, colorectal, or lung cancer, with and without comorbidities: United States, 2010–2016
| Visits by adults with breast, prostate, colorectal, and lung cancer | ||||||||
|---|---|---|---|---|---|---|---|---|
| Total | with no comorbidity | with 1 comorbidity | with ≥2 comorbidities | |||||
| No. (Weighted No. in 1000s) | %(SE) | No. (Weighted No. in 1000s) | % (SE) | No. (Weighted No. in 1000s) | % (SE) | No. (Weighted No. in 1000s) | % (SE) | |
| Total | 4672 (10155) | 2239 (4439) | 43.7 (1.7) | 1114 (2368) | 23.3 (0.9) | 1319 (3348) | 33.0 (1.6) | |
| Type of cancer | ||||||||
| Breasta | 1944 (4189) | 41.2 (1.7) | 1023 (2102) | 47.3 (2.1) | 446 (874) | 36.9 (2.7) | 475 (1213) | 36.2 (2.5) |
| Prostateb | 1297 (2576) | 25.4 (1.8) | 578 (1001) | 22.5 (2.1) | 325 (653) | 27.6 (2.6) | 394 (922) | 27.5 (2.7) |
| Colorectalc | 760 (1702) | 16.8 (1.0) | 382 (763) | 17.2 (1.4) | 170 (378) | 15.9 (1.9) | 208 (561) | 16.8 (1.9) |
| Lung | 671 (1688) | 16.6 (1.3) | 256 (573) | 12.9 (1.4) | 173 (464) | 19.6 (3.4) | 242 (652) | 19.5 (2.1) |
| Sex | ||||||||
| Femaled | 2630 (5812) | 57.2 (1.9) | 1326 (2683) | 60.4 (2.4) | 626 (1373) | 58.0 (3.0) | 678 (1756) | 52.4 (2.6) |
| Male | 2042 (4343) | 42.8 (1.9) | 913 (1756) | 39.6 (2.4) | 488 (995) | 42.0 (3.0) | 641 (1592) | 47.6 (2.6) |
| Age (Mean) (SE) | 65.8 (0.3) | 62.5 (0.7) | 66.4 (0.6) | 69.8 (0.5) | ||||
| MEDIAN (IQR) | 66.4 (57.2–74.4) | 62.9 (52–72) | 66.5 (59–74.3) | 69.8 (62.1–76.3) | ||||
| Race/Ethnicity1 | ||||||||
| Non-Hispanic Whitee | 3773 (7888) | 77.7 (2.0) | 1787 (3310) | 74.6 (2.9) | 901 (1884) | 79.5 (2.2) | 1085 (2695) | 80.5 (2.6) |
| Non-Hispanic Black | 450 (966) | 9.5 (0.9) | 220 (463) | 10.4 (1.3) | 108 (225) | 9.5 (1.4) | 122 (278) | 8.3 (1.3) |
| Hispanic | 310 (916) | 9.0 (1.3) | 170 (458) | 10.3 (2.0) | 75 (209) | 8.8 (1.8) | 65 (249) | 7.4 (1.6) |
| Provider | ||||||||
| Oncologist2 | 2518 (5216) | 51.4 (3.5) | 1264 (2393) | 53.9 (3.3) | 570 (1150) | 48.6 (3.9) | 684 (1673) | 50.0 (5.1) |
| Other | 2154 (4939) | 48.6 (3.5) | 975 (2045) | 46.1 (3.3) | 544 (1218) | 51.4 (3.9) | 635 (1675) | 50.0 (5.1) |
| Expected source of payment3 | ||||||||
| Private insurancef | 1825 (3657) | 36.0 (1.6) | 1009 (1898) | 42.8 (2.7) | 428 (899) | 38.0 (2.5) | 388 (859) | 25.6 (2.2) |
| Medicare/Medicaid | 2519 (5668) | 55.8 (2.1) | 1055 (2149) | 48.4 (2.6) | 617 (1266) | 53.5 (2.8) | 847 (2253) | 67.3 (3.2) |
| Other/Unknown | 328 (830)* | * | 175 (391) | 8.8 (2.3) | 69 (203)* | * | 84 (236)* | * |
Estimate does not meet NCHS standards of reliability.
Non-Hispanic other not shown (n=139)
Oncologist includes gynecological oncology, musculoskeletal oncology, hematology/oncology, medical oncology, and surgical oncology
Expected source of payment is based on a hierarchy because more than one insurance could be selected as expected source of payment (see Methods section for more information)
Significantly different from the other cancer types overall and across all comorbidity groups;
Significantly different from the other cancer types overall and among visits with no comorbidity and with ≥2 comorbidities; significantly different from colorectal cancer among visit with 1 comorbidity;
Significantly different from lung cancer among visits with no comorbidity;
Significantly different from male overall, among visits with no comorbidity and with 1 comorbidity;
Significantly different from the Non-Hispanic Black and Hispanic overall and across all comorbidity groups;
Significantly different from Medicare/Medicaid and other/unknown overall, among visits with 1 comorbidity, and among visits with ≥2 comorbidities.
Abbreviations: No., unweighted sample size; SE, standard error; IQR, interquartile range.
Source: National Ambulatory Medical Care Survey 2010–2016 (n=4,672). Numbers may not add to totals due to rounding.
Oncologists were seen at 51.4% of cancer visits, not different from the percentage of cancer visits made to non-oncologists (48.6%). The expected source of payment was Medicare/Medicaid for 55.8% of the visits, followed by private insurance (36.0%) (P <.001).
Patients making a breast cancer visit were younger (61.7 years) than patients making prostate, colorectal or lung cancer visits (Supplementary Table 3). More visits were made by non-Hispanic white, than non-Hispanic black and Hispanic adults across all cancer types. Oncologists were seen more frequently than other providers among visits made by patients with breast, colorectal and lung cancer whereas the opposite pattern was observed for visits by patients with prostate cancer. The expected source of payment was Medicare/Medicaid for most visits, followed by private insurance, across all cancer type.
Approximately 44% of cancer visits were by patients with cancer and no comorbidities, 23% of visits were by patients with cancer and one comorbidity, and 33.0% of visits by patients with cancer and ≥2 comorbidities.
The percentage of visits by patients with breast cancer was higher than the percentage of visits made by patients with prostate, colorectal or lung cancer across all three comorbidity groups (P<.001) (Table 1). The average patient age was 62.5 years for visits by patients with no comorbidities, 66.4 years for visits by patients with one comorbidity and 69.8 years for visits by patients with ≥2 comorbidities. There was no significant difference in the percentage of visits made to oncologists and non-oncologists across all three comorbidity groups. Oncologists were seen at 53.9% of visits by patients with cancer and no comorbidity, 48.6% of visits by patients with cancer and one comorbidity and 50.0% of visits by patients with cancer and ≥2 comorbidities. Medicare and Medicaid were the primary expected sources of payment for visits by patients with one comorbidity and by patients with ≥2 comorbidities, whereas no significant difference between Medicare/Medicaid and private insurance among visits by patients with cancer and no comorbidity.
Among visits by patients with breast cancer, 20.9% had one comorbidity while 29.0% had ≥2 comorbidities (p=0.001). Among visits by adults with prostate cancer, 25.3% had one comorbidity and 35.8% had ≥2 comorbidities (p=0.002). Among visits by colorectal cancer patients, 22.2% had one comorbidity and 32.9% were made by patients with ≥2 comorbidities (p=0.01). Among visits by adults with lung cancer, 27.5% had one comorbidity and 38.6% had ≥2 comorbidities (p=0.03) (Figure 1). Overall, hypertension was the most frequently-observed comorbidity, found in 37.7% of the visits by patients with cancer. This was followed by hyperlipidemia (19.0%), diabetes (12.3%), arthritis (11.4%), and depression (7.7%) (Figure 2). Regardless of cancer type, hypertension was the most frequently-observed comorbidity.
Figure 1. Percentage of physician office visits by adult patients by number of comorbidities and cancer type: United States, 2010–2016.
Percentage of physician office visits by adult patients by number of comorbidities and cancer type: United States, 2010–2016
Note: Error bars represent KornGraubard95% confidence interval. *Significantly higher than 1 comorbidity group in all cancer type. Source: National Ambulatory Medical Care Survey 2007–2016 (n=4,672)
Figure 2. Five most frequently-observed comorbidities among physician office visits by adult patients with breast, prostate, colorectal and lung cancer by cancer type: United States, 2010–2016.
Five most frequently-observed comorbidities among physician office visits by adult patients with breast, prostate, colorectal and lung cancer by cancer type: United States, 2010–2016
Abbreviation: COPD, chronic obstructive pulmonary disease.Note: Error bars represent KornGraubard95% confidence interval. Source: National Ambulatory Medical Care Survey, 2010–2016 (n=4,672).
When comparing visits by adult cancer patients with no comorbidities versus visits by adult cancer patients with ≥1 comorbidities, more visits were made by cancer patients with ≥1 comorbidities than by cancer patients with no comorbidities. Among visits by non-Hispanic white adults, 58.0% of the visits were made by cancer patients with ≥1 comorbidities. Among visits by cancer patients to non-oncologists, 58.6% were by patients with ≥1 comorbidities. Among visits in which the expected source of payment was Medicare/Medicaid, 62.1% were made by cancer patients with ≥1 comorbidities (Table 2).
Antineoplastic drugs were prescribed in approximately 3.4 million (or 33.5%) of visits by cancer patients. Precisely 5.4% of the visits in which antineoplastic drugs were prescribed also had radiation therapy prescribed or administered (data not shown). Cancer patients with ≥1 comorbidities made 57.1% of the visits with prescription for antineoplastic drugs (Table 3). Among visits with a prescription for antineoplastic drugs, visits made by cancer patients with ≥1 comorbidities were higher compared to visits made by cancer patients with no comorbidities across all subgroups with the exception of visits with private insurance and other/unknown as expected source of payment, for which no differences were found. Among visits with a prescription for antineoplastic drugs, 2.6 million of the visits were made to oncologists (76.0%).
In analysis by specific comorbidities, the percentage of visits with a prescription for antineoplastic drugs was significantly higher for visits by cancer patients with depression compared to visits by cancer patients without depression (43.7 versus 32.7% P=.02). We also found that antineoplastic drugs were prescribed less often among cancer patients with chronic obstructive pulmonary disease (COPD) than among cancer patients without COPD (21.3% vs 34.3% P=.005), and among cancer patients with ischemic heart disease than cancer patients without ischemic heart disease (34.2% versus 22.6% P=.02). Renal disease, cerebrovascular disease, congestive heart failure, peripheral vascular disease, peptic ulcer disease, hemiplegia or paraplegia, dementia, liver disease, and AIDS were not shown because the corresponding proportions do not meet standards of precision (Figure 3).
Fig. 3.
Percentage of physician office visits by adult patients with breast, prostate, colorectal, or lung cancer with antineoplasticdrugs prescription, by presence or absence of specific comorbidities: USA, 2010–2016. The percentage of visits with a prescription for antineoplastic drugs was significantly higher for visits by cancer patients with depression compared to visits by cancer patients without depression. Antineoplastic drugs were prescribed less often among cancer patients with chronic obstructive pulmonary disease (COPD) than among cancer patients without COPD, and among cancer patients with ischemic heart disease than cancer patients without ischemic heart disease
*P< .05 Abbreviation: COPD, chronic obstructive pulmonary disease.Note: Renal disease, cerebrovascular disease, congestive heart failure, peripheral vascular disease, peptic ulcer disease, hemiplegia or paraplegia, dementia, liver disease, and AIDS were not shown because the proportiondo not meet NCHS standards of precision. Source: National Ambulatory Medical Care Survey 2010–2016 (n=1,578)
Discussion
This study found that over half (56.3%) of the visits by cancer patients were by patients with ≥1 comorbidities. Edwards et al. reported US Medicare beneficiaries ≥66 years with breast and prostate cancers had similar prevalence of comorbidity as non-cancer patients (30%−32%), patients with colorectal cancer had higher prevalence (41%), and lung cancer patients had the highest prevalence of comorbidities (53%)(4). Although it is recognized that comorbidities are common among cancer patients, comparing comorbidity prevalence data across cancer studies is made difficult by the large variability of conditions used, as comorbidity varies based upon the measure used, the study population, and the cancer type. Comorbidities appear to be more common in studies restricted to older patients and less common in studies based on claims data, cancer registries or self-report(29)(8). Although these studies are not easily comparable to ours because the unit of analysis is different, a higher proportion of our study population had comorbidities compared to Edwards et al. This could be explained by the inclusion of additional comorbidities, common in the population, in our analysis, such as hypertension, hyperlipidemia, depression, asthma and osteoporosis.
In our analysis, the most common comorbid condition found in visits by adult cancer patients was hypertension, which is consistent with Piccirillo et al. who found hypertension was present in 37% of cancer patients(30). This finding is consistent with the high prevalence of hypertension in the general population(31).
We found that among cancer visits with a prescription for antineoplastic drugs, the percentage of visits made by patients with ≥1 comorbidities was higher compared to visits made by cancer patients with no comorbidities. A possible explanation is that people with multiple comorbidities may access the health care system more frequently, have a need to better manage their conditions (including cancer), and subsequently might receive more treatment. (32–34) (35, 36). It is also possible that patients with comorbidities receive modified cancer treatment that requires more frequent visits.(37, 38) Finally, since cancer stages influence treatment, and surgery and radiotherapy could be alternatives to chemotherapy (e.g., in patients with localized prostate cancer), (39, 40) it is also possible that patients with no comorbidities have cancers at stages that would not benefit from antineoplastic drugs administration.
We found that the percentages of visits with a prescription for antineoplastic drugs by cancer patients with COPD were significantly lower than the percentages of visits with prescription for antineoplastic drugs by cancer patients without COPD. In previous studies COPD has been shown to be associated with less adherence to chemotherapy in breast, prostate, colon and lung cancer(8, 15, 19, 22). We found that the percentages of visits with prescription for antineoplastic drugs by cancer patients with ischemic heart disease were significantly lower than the percentages of visits with a prescription for antineoplastic drugs by cancer patients without ischemic heart disease. It has been shown that acute coronary events can be precipitated by infusion of several chemotherapeutic agents(41–44). The percentages of visits with a prescription for antineoplastic drugs by cancer patients with depression were significantly higher than the percentages of visits with prescription for antineoplastic drugs by cancer patients without depression. Previous studies have found that, at least subclinical depression, was present in approximately 29% of cancer patients(45). In particular, some studies found high prevalence of depression in cancer patients prior to chemotherapy and increased depressive disorders triggered by chemotherapy(46, 47), which might explain our findings. We did not find any significant difference between cancer visits with and without hypertension and cancer visits with and without diabetes for prescription for antineoplastic drugs. Some studies have shown that breast cancer patients with diabetes still receive chemotherapy, although modified breast cancer treatment regimens were the preferred options(13, 48, 49). We also did not find any significant difference between the percentage of visits with a prescription for antineoplastic drugs in cancer visits with and without hyperlipidemia, arthritis, osteoporosis, or asthma.
Our study has limitations that can affect the interpretation of the results. First, stage of cancer and time since diagnosis is unknown and both factors may influence cancer treatment or chemotherapy suitableness. Also, because time since diagnosis is unknown and follow-up visits for cancer can continue long after the diagnosis, it is possible that visits made by cancer patients with cancer in remission could be included. However, we excluded codes listing a personal history of cancer, including only visits with ICD-9-CM and ICD-10-CM codes denoting a first listed, primary diagnosis of breast, colon, or colorectal cancer.
Second, our study is cross-sectional, and therefore causation cannot be inferred. It is possible that the comorbidities were a consequence of the chemotherapy; therefore, limiting further access to treatment. Third, subgroup analysis by specific comorbidities and cancer type could not be conducted due to small sample size. Fourth, although from 2010–2016 there was an increasing trend in the percentage of visits by adult cancer patients with ≥2 comorbidities, and a decreasing trend in the percentage of visits by adult cancer patients with no comorbidities (data not shown), we combined these data and presented a time-averaged association. Fifth, only ambulatory data (i.e., non-inpatient) are included in this analysis, therefore patients with more severe comorbidities or advanced cancer stage may not be captured by this analysis. Finally, the unit of analysis for NAMCS is an ambulatory care visit to a physician in the United States; thus, the number of visits rather than the number of people are measured, so it may be difficult to interpret the data at patient level as it is possible for the same person to be counted multiple times. However, the reporting period of the NAMCS is only one week, decreasing the likelihood of repeated office visits by the same patient during this period.
Our study also has several strengths. NAMCS is nationally-representative and provides data abstracted from patient medical records, which may result in more complete collection of data. Although the impact of comorbidities on cancer prognosis and treatment decision has been investigated in several studies (8, 11, 12, 37), many questions about specific comorbidities remain unanswered. Previous studies were based on cancer registries linked to administrative data, which may underestimate comorbidities because there is limited information on common chronic conditions such as hypertension, depression, or asthma. Our study also included comorbidities that are not listed in the Charlson index. The Charlson Index is an accepted measure of comorbidities and was shown to predict 1-year all-cause mortality(26); however, it does not account for the mentioned chronic conditions, which are expected to affect overall patient health, quality of life, and access to healthcare, especially in older patient populations(4).
As a nationally-representative study of office-based ambulatory medical care, our study provides information about comorbidities in visits by adult cancer patients and antineoplastic drugs prescription received during these visits. These findings complement the literature regarding cancer and comorbidities among adult patients.
Supplementary Material
Footnotes
Conflict of interest:
The authors state no conflict of interest.
Disclaimer:
The findings and conclusions in this report are those of the authors and do not necessarily represent the official position of the National Center for Health Statistics, Centers for Disease Control and Prevention.
References
- 1.Siegel RL, Miller KD, Jemal A. (2018) Cancer statistics, 2018. CA Cancer J Clin. 68: 7–30. [DOI] [PubMed] [Google Scholar]
- 2.Smith BD, Smith GL, Hurria A, Hortobagyi GN, Buchholz TA. (2009) Future of cancer incidence in the United States: burdens upon an aging, changing nation. J Clin Oncol. 27: 2758–65. [DOI] [PubMed] [Google Scholar]
- 3.Feinstein AR. (1970) The Pre-Therapeutic Classification of Co-Morbidity in Chronic Disease. J Chronic Dis. 23: 455–68. [DOI] [PubMed] [Google Scholar]
- 4.Edwards BK, Noone AM, Mariotto AB, et al. (2014) Annual Report to the Nation on the status of cancer, 1975–2010, featuring prevalence of comorbidity and impact on survival among persons with lung, colorectal, breast, or prostate cancer. Cancer. 120: 1290–314. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Townsley CA, Selby R, Siu LL. (2005) Systematic review of barriers to the recruitment of older patients with cancer onto clinical trials. J Clin Oncol. 23: 3112–24. [DOI] [PubMed] [Google Scholar]
- 6.Mohile SG, Dale W, Somerfield MR, Hurria A. (2018) Practical Assessment and Management of Vulnerabilities in Older Patients Receiving Chemotherapy: ASCO Guideline for Geriatric Oncology Summary. J Oncol Pract. 14: 442–6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Unger JM, Hershman DL, Fleury ME, Vaidya R. (2019) Association of Patient Comorbid Conditions With Cancer Clinical Trial Participation. JAMA Oncol. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Lee L, Cheung WY, Atkinson E, Krzyzanowska MK. (2011) Impact of comorbidity on chemotherapy use and outcomes in solid tumors: a systematic review. J Clin Oncol. 29: 106–17. [DOI] [PubMed] [Google Scholar]
- 9.Janssen-Heijnen ML, Smulders S, Lemmens VE, Smeenk FW, van Geffen HJ, Coebergh JW. (2004) Effect of comorbidity on the treatment and prognosis of elderly patients with non-small cell lung cancer. Thorax. 59: 602–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Janssen-Heijnen ML, Houterman S, Lemmens VE, Louwman MW, Maas HA, Coebergh JW. (2005) Prognostic impact of increasing age and co-morbidity in cancer patients: a population-based approach. Crit Rev Oncol Hematol. 55: 231–40. [DOI] [PubMed] [Google Scholar]
- 11.Meyerhardt JA, Catalano PJ, Haller DG, et al. (2003) Impact of diabetes mellitus on outcomes in patients with colon cancer. J Clin Oncol. 21: 433–40. [DOI] [PubMed] [Google Scholar]
- 12.Pal SK, Hurria A. (2010) Impact of age, sex, and comorbidity on cancer therapy and disease progression. J Clin Oncol. 28: 4086–93. [DOI] [PubMed] [Google Scholar]
- 13.van de Poll-Franse LV, Houterman S, Janssen-Heijnen ML, Dercksen MW, Coebergh JW, Haak HR. (2007) Less aggressive treatment and worse overall survival in cancer patients with diabetes: a large population based analysis. Int J Cancer. 120: 1986–92. [DOI] [PubMed] [Google Scholar]
- 14.Tammemagi CM, Neslund-Dudas C, Simoff M, Kvale P. (2004) In lung cancer patients, age, race-ethnicity, gender and smoking predict adverse comorbidity, which in turn predicts treatment and survival. J Clin Epidemiol. 57: 597–609. [DOI] [PubMed] [Google Scholar]
- 15.Gross CP, McAvay GJ, Guo Z, Tinetti ME. (2007) The impact of chronic illnesses on the use and effectiveness of adjuvant chemotherapy for colon cancer. Cancer. 109: 2410–9. [DOI] [PubMed] [Google Scholar]
- 16.Lemmens VE, Janssen-Heijnen ML, Verheij CD, Houterman S, Repelaer van Driel OJ, Coebergh JW. (2005) Co-morbidity leads to altered treatment and worse survival of elderly patients with colorectal cancer. Br J Surg. 92: 615–23. [DOI] [PubMed] [Google Scholar]
- 17.Sarfati D, Hill S, Blakely T, et al. (2009) The effect of comorbidity on the use of adjuvant chemotherapy and survival from colon cancer: a retrospective cohort study. BMC Cancer. 9: 116. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Hawfield A, Lovato J, Covington D, Kimmick G. (2006) Retrospective study of the effect of comorbidity on use of adjuvant chemotherapy in older women with breast cancer in a tertiary care setting. Crit Rev Oncol Hematol. 59: 250–5. [DOI] [PubMed] [Google Scholar]
- 19.Dy SM, Sharkey P, Herbert R, Haddad K, Wu AW. (2006) Comorbid illnesses and health care utilization among Medicare beneficiaries with lung cancer. Crit Rev Oncol Hematol. 59: 218–25. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Janssen-Heijnen ML, Lemmens VE, van den Borne BE, Biesma B, Oei SB, Coebergh JW. (2007) Negligible influence of comorbidity on prognosis of patients with small cell lung cancer: a population-based study in the Netherlands. Crit Rev Oncol Hematol. 62: 172–8. [DOI] [PubMed] [Google Scholar]
- 21.Houterman S, Janssen-Heijnen ML, Verheij CD, et al. (2004) Comorbidity has negligible impact on treatment and complications but influences survival in breast cancer patients. Br J Cancer. 90: 2332–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Bradley CJ, Dahman B, Anscher M. (2014) Prostate cancer treatment and survival: evidence for men with prevalent comorbid conditions. Med Care. 52: 482–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Park J, Look KA. (2019) Health Care Expenditure Burden of Cancer Care in the United States. Inquiry. 56: 46958019880696. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Nissen MJ, Beran MS, Lee MW, Mehta SR, Pine DA, Swenson KK. (2007) Views of primary care providers on follow-up care of cancer patients. Fam Med. 39: 477–82. [PubMed] [Google Scholar]
- 25.Jacobson JO, Polovich M, McNiff KK, et al. (2009) American Society of Clinical Oncology/Oncology Nursing Society chemotherapy administration safety standards. Oncol Nurs Forum. 36: 651–8. [DOI] [PubMed] [Google Scholar]
- 26.Charlson ME, Pompei P, Ales KL, MacKenzie CR. (1987) A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. J Chronic Dis. 40: 373–83. [DOI] [PubMed] [Google Scholar]
- 27.Klabunde CN, Potosky AL, Legler JM, Warren JL. (2000) Development of a comorbidity index using physician claims data. J Clin Epidemiol. 53: 1258–67. [DOI] [PubMed] [Google Scholar]
- 28.Parker JD TM, Malec DJ, et al. (2017) National Center for Health Statistics Data Presentation Standards for Proportions. National Center for Health Statistics. Vital Health Stat 2(175). [PubMed] [Google Scholar]
- 29.Sarfati D (2012) Review of methods used to measure comorbidity in cancer populations: no gold standard exists. J Clin Epidemiol. 65: 924–33. [DOI] [PubMed] [Google Scholar]
- 30.Piccirillo JF, Tierney RM, Costas I, Grove L, Spitznagel EL Jr. (2004) Prognostic importance of comorbidity in a hospital-based cancer registry. JAMA. 291: 2441–7. [DOI] [PubMed] [Google Scholar]
- 31.Yoon SS, Carroll MD, Fryar CD. (2015) Hypertension Prevalence and Control Among Adults: United States, 2011–2014. NCHS Data Brief. 1–8. [PubMed] [Google Scholar]
- 32.Gross CP, Andersen MS, Krumholz HM, McAvay GJ, Proctor D, Tinetti ME. (2006) Relation between Medicare screening reimbursement and stage at diagnosis for older patients with colon cancer. JAMA. 296: 2815–22. [DOI] [PubMed] [Google Scholar]
- 33.Zafar SY, Abernethy AP, Abbott DH, et al. (2008) Comorbidity, age, race and stage at diagnosis in colorectal cancer: a retrospective, parallel analysis of two health systems. BMC Cancer. 8: 345. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Ahn DH, Mehta N, Yorio JT, Xie Y, Yan J, Gerber DE. (2013) Influence of medical comorbidities on the presentation and outcomes of stage I-III non-small-cell lung cancer. Clin Lung Cancer. 14: 644–50. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Fleming ST, Pursley HG, Newman B, Pavlov D, Chen K. (2005) Comorbidity as a predictor of stage of illness for patients with breast cancer. Med Care. 43: 132–40. [DOI] [PubMed] [Google Scholar]
- 36.Yasmeen S, Xing G, Morris C, Chlebowski RT, Romano PS. (2011) Comorbidities and mammography use interact to explain racial/ethnic disparities in breast cancer stage at diagnosis. Cancer. 117: 3252–61. [DOI] [PubMed] [Google Scholar]
- 37.Sogaard M, Thomsen RW, Bossen KS, Sorensen HT, Norgaard M. (2013) The impact of comorbidity on cancer survival: a review. Clin Epidemiol. 5: 3–29. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Snyder CF, Frick KD, Herbert RJ, et al. (2015) Comorbid condition care quality in cancer survivors: role of primary care and specialty providers and care coordination. J Cancer Surviv. 9: 641–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Sanda MG, Cadeddu JA, Kirkby E, et al. (2018) Clinically Localized Prostate Cancer: AUA/ASTRO/SUO Guideline. Part II: Recommended Approaches and Details of Specific Care Options. J Urol. 199: 990–7. [DOI] [PubMed] [Google Scholar]
- 40.Sanda MG, Cadeddu JA, Kirkby E, et al. (2018) Clinically Localized Prostate Cancer: AUA/ASTRO/SUO Guideline. Part I: Risk Stratification, Shared Decision Making, and Care Options. J Urol. 199: 683–90. [DOI] [PubMed] [Google Scholar]
- 41.Lu JI, Carhart RL, Graziano SL, Gajra A. (2006) Acute coronary syndrome secondary to fluorouracil infusion. J Clin Oncol. 24: 2959–60. [DOI] [PubMed] [Google Scholar]
- 42.Wijesinghe N, Thompson PI, McAlister H. (2006) Acute coronary syndrome induced by capecitabine therapy. Heart Lung Circ. 15: 337–9. [DOI] [PubMed] [Google Scholar]
- 43.Gemici G, Cincin A, Degertekin M, Oktay A. (2009) Paclitaxel-induced ST-segment elevations. Clin Cardiol. 32: E94–6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 44.Mohanty BD, Mohanty S, Hussain Y, et al. (2017) Management of ischemic coronary disease in patients receiving chemotherapy: an uncharted clinical challenge. Future Cardiol. 13: 247–57. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 45.Linden W, Vodermaier A, Mackenzie R, Greig D. (2012) Anxiety and depression after cancer diagnosis: prevalence rates by cancer type, gender, and age. J Affect Disord. 141: 343–51. [DOI] [PubMed] [Google Scholar]
- 46.Haque R, Shi J, Schottinger JE, et al. (2016) Tamoxifen and Antidepressant Drug Interaction in a Cohort of 16,887 Breast Cancer Survivors. J Natl Cancer Inst. 108. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 47.Gregorowitsch ML, Ghedri A, Young-Afat DA, et al. (2019) The effect of chemotherapy on subjective cognitive function in younger early-stage breast cancer survivors treated with chemotherapy compared to older patients. Breast Cancer Res Treat. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 48.Yancik R, Wesley MN, Ries LA, Havlik RJ, Edwards BK, Yates JW. (2001) Effect of age and comorbidity in postmenopausal breast cancer patients aged 55 years and older. JAMA. 285: 885–92. [DOI] [PubMed] [Google Scholar]
- 49.Srokowski TP, Fang S, Hortobagyi GN, Giordano SH. (2009) Impact of diabetes mellitus on complications and outcomes of adjuvant chemotherapy in older patients with breast cancer. J Clin Oncol. 27: 2170–6. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.