Skip to main content
PMC home page
Journal of Oncology Practice logoLink to Journal of Oncology Practice
. 2016 Feb;12(2):151–152. doi: 10.1200/JOP.2015.004812

ReCAP: Hospitalizations in Older Adults With Advanced Cancer: The Role of Chemotherapy

Caitriona B O’Neill 1, Coral L Atoria 1, Eileen M O’Reilly 1, Martin C Henman 1, Peter B Bach 1, Elena B Elkin 1,✉, Caitriona B O’Neill 2, Coral L Atoria 2, Eileen M O’Reilly 2, Martin C Henman 2, Peter B Bach 2, Elena B Elkin 2,✉
PMCID: PMC5702789  PMID: 26869655

Abstract

CONTEXT AND QUESTION ASKED:

In patients with metastatic cancer, chemotherapy may provide symptom control, prevent complications, prolong life, or improve quality of life. Except in rare cases, however, patients with metastatic disease will not be cured. In older patients with metastatic cancer, hospitalization for treatment toxicity may reduce the quality of an already limited life expectancy. We evaluated the association between chemotherapy for metastatic cancer and risk of hospitalization.

MAIN CONCLUSION:

Hospitalizations are common in patients with incurable advanced malignancies and are more likely among those who receive chemotherapy.

APPROACH:

In the linked SEER-Medicare dataset, we identified Medicare beneficiaries aged 66 years or older with a primary diagnosis of metastatic breast, colorectal, ovarian, bladder, lung, pancreas, esophageal, stomach, or prostate cancer between 2001 and 2009 who died by the end of 2010. Chemotherapy recipients and nonrecipients were pair-matched by age, sex, race, comorbidity, geographic region and survival duration. The primary endpoint was hospital admission, identified in inpatient claims between cancer diagnosis and the first of hospice admission or death. We also identified the subset of admissions associated with a primary or secondary diagnosis code suggestive of an adverse effect of chemotherapy. The association between chemotherapy and hospitalization was estimated in separate multivariable Cox proportional-hazards regression models for each cancer site, accounting for the matched-pairs design and controlling for unmatched demographic and disease characteristics.

RESULTS:

Of 18,486 patients who received chemotherapy for metastatic cancer, 92% were hospitalized at least once for any reason, including 51% hospitalized for a likely toxicity. The corresponding rates among matched non-recipients were 83% and 34% (Figure). In nearly all cancers, chemotherapy recipients had a greater risk of hospitalization for a likely toxicity or for any cause. Chemotherapy recipients had substantially higher hospitalization for infection or fever (21% v 15%), hematologic complications (11% v 3%), dehydration (13% v 6%), and PE or DVT (9% v 4%) compared with nonrecepients. Chemotherapy was associated with a significantly increased risk of likely toxicity-related hospitalization in nearly all cancers, controlling for sociodemographic characteristics and other treatment. The association was greatest in patients with metastatic esophageal cancer (adjusted hazard ratio, 2.00; 95% CI, 1.11 to 3.60) and smallest in patients with metastatic prostate cancer (adjusted hazard ratio, 1.22; 95% CI, 1.01 to 1.47).

INTERPRETATION:

Older patients receiving chemotherapy for incurable advanced cancers are at high risk of hospitalization, of which a non-negligible proportion is likely attributable to adverse effects of treatment. Infection, fever, dehydration, and hematologic complications constitute a large proportion of these events, some of which may be preventable through evidence-based patient management, prophylactic interventions, and effective outpatient care. Our findings might be limited to older patients with advanced cancer who have a generally poor prognosis or limited expected survival.

SIGNIFICANCE OF FINDINGS:

Understanding the common reasons for hospital admissions and developing toxicity management programs and educational resources may help patients and their families make informed treatment decisions, minimize adverse effects and reduce hospitalizations in this population with limited life expectancy.

FIG 1.

FIG 1.

Open in a new tab

Hospitalization by cause and receipt of chemotherapy.

J Oncol Pract. 2016 Feb;12(2):e138–e148. doi: 10.1200/JOP.2015.004812

Original Contribution: Hospitalizations in Older Adults With Advanced Cancer: The Role of Chemotherapy

Abstract

Purpose:

In older patients with metastatic cancer, hospitalization for treatment toxicity may reduce the quality of an already limited life expectancy. We evaluated the association between chemotherapy for metastatic cancer and risk of hospitalization.

Methods:

In the population-based SEER-Medicare dataset, we identified patients 66 years or older diagnosed with metastatic cancer of the bladder, breast, prostate, colon or rectum, esophagus, pancreas, stomach, ovaries, or lung in 2001 to 2009 who died by December 31, 2010. Chemotherapy recipients were matched to nonrecipients by age, sex, race, geographic region, comorbidity, and survival duration. We identified hospitalizations for any cause and for likely chemotherapy-related toxicity, comparing chemotherapy recipients with their matched peers.

Results:

Of 18,486 patients who received chemotherapy for metastatic cancer, 92% were hospitalized at least once for any reason, including 51% hospitalized for a likely toxicity. The corresponding rates among matched nonrecipients were 83% and 34%. Chemotherapy was associated with a significantly increased risk of likely toxicity-related hospitalization in nearly all cancers, controlling for sociodemographic characteristics and other treatment. The association was greatest in patients with metastatic esophageal cancer (adjusted hazard ratio, 2.00; 95% CI, 1.11 to 3.60) and smallest in patients with metastatic prostate cancer (adjusted hazard ratio, 1.22; 95% CI, 1.01 to 1.47).

Conclusion:

Hospitalizations are common in patients with incurable advanced malignancies and more likely among those who receive chemotherapy. Understanding common reasons for these events may help reduce adverse effects of chemotherapy for metastatic cancer and help patients and their families make informed treatment decisions.

INTRODUCTION

In the United States, half of all men and one third of women will develop cancer, and many are diagnosed with or eventually develop metastatic disease.1,2 More than half a million Americans die of cancer each year,3 most after the development of metastasis.

Chemotherapy is the primary treatment for most patients with metastatic cancer, but prognosis varies depending on cancer site, the effectiveness of therapy, comorbidity, and functional status.4,5 Except in rare cases, however, patients with metastatic disease will not be cured. Chemotherapy may provide symptom control, prevent complications, prolong life, or improve quality of life.

Despite scientific advances, improvements in the treatment of metastatic cancer are typically evaluated in weeks or months rather than years, and drug efficacy, measured by tumor response or disease progression, does not always correspond with improvement in survival or quality of life.4 In many settings, especially late lines of therapy for most cancers, there may be little or no evidence of benefit from chemotherapy. Yet nearly all drugs and regimens bear some level of toxicity and detriment to quality of life.4,6 Furthermore, although palliative chemotherapy is a reasonable treatment choice for many patients, most have inaccurate expectations of its curative potential, possibly compromising their ability to make informed treatment decisions consistent with their preferences.7

Recent studies have reported substantial and increasing chemotherapy use near the end of life, but less attention has been paid to the burden of treatment in terms of its adverse effects.8-12 Studies that have taken a retrospective approach, evaluating treatment received by decedents, may yield a biased portrait of terminal care and its outcomes.13 Our objective was to estimate the impact of systemic chemotherapy on the risk of hospitalization in patients with metastatic cancer.

METHODS

Data

We used SEER cancer registry data linked with Medicare claims.14 SEER is a National Cancer Institute–sponsored consortium of population-based cancer registries that now cover almost 30% of Americans. For all incident cancers, the SEER registries collect information regarding site and extent of disease, sociodemographic characteristics, and first course of cancer-directed therapy, with active follow-up for date and cause of death. Medicare is the primary health insurer for 97% of the population of the United States age 65 years and older, covering inpatient hospital care, outpatient care, physician services, and, since 2006, outpatient prescription medications. This study was deemed exempt research by the Institutional Review Board at Memorial Sloan Kettering Cancer Center, and the SEER-Medicare files were used in accordance with a data use agreement from the National Cancer Institute.

Study Cohort

We identified Medicare beneficiaries age 66 years or older with a primary diagnosis of metastatic breast, colorectal, ovarian, bladder, lung, pancreas, esophageal, stomach, or prostate cancer between January 1, 2001 and December 31, 2009 who died by December 31, 2010. Identification of metastatic disease was based on disease information collected in SEER (Appendix Table A1, online only). We excluded patients enrolled in a Medicare managed care plan and those who did not have continuous Medicare coverage from 1 year before diagnosis through death, patients diagnosed only at death, and those with a history of another malignancy.

The study population was divided into two cohorts: those who received systemic chemotherapy and those who did not. Receipt of chemotherapy was identified in Medicare claims from the date of cancer diagnosis through death or the end of follow-up and included both intravenous and oral chemotherapy drugs covered under Medicare Part B (Appendix Table A2).

Chemotherapy recipients and nonrecipients were pair-matched by age at diagnosis (± 5 years), sex, race (white, black, or other), comorbidity index score, geographic region (West, Midwest, South, or Northeast), and duration of survival. Comorbidity was estimated using the Klabunde modification of the Charlson comorbidity index on the basis of inpatient, outpatient, and physician claims in the year before cancer diagnosis, and scores were categorized by standard practice as 0, 1, or 2+.15 Patients with better-prognosis advanced cancers—breast, bladder, colorectal, ovarian, and small-cell lung cancer (SCLC)—were matched within ± 3 months survival duration. Patients with poorer-prognosis advanced cancers—pancreatic, esophageal, stomach, non–small-cell lung cancer (NSCLC), and prostate cancer—were matched within ± 1 month survival duration.5

Outcomes

The primary end point was hospital admission, identified in inpatient claims between cancer diagnosis and the first of hospice admission or death. We also identified the subset of admissions associated with a primary or secondary diagnosis code suggestive of an adverse effect of chemotherapy. These included hematologic complications, gastrointestinal complications, dehydration, deep venous thrombosis or pulmonary embolus, infection or fever, malnutrition or failure to thrive, cardiac complications, and constitutional symptoms (delirium, drug psychoses, and malaise or fatigue).16-18

Covariates

In addition to the characteristics used in matching, we examined marital status, urban or rural residence, and year of diagnosis. Median income in the census tract of residence was used as a marker of socioeconomic status.

Analysis

Unadjusted associations between receipt of chemotherapy and unmatched characteristics were assessed using χ2 statistics. Differences between matched cohorts in the frequency of hospital admission were evaluated for all hospitalizations and for the subset of likely toxicity-related events. We estimated unadjusted relative risks by cancer site and compared hospitalization rates using McNemar's test for matched pairs.

The association between chemotherapy and hospitalization was estimated in separate multivariable Cox proportional hazards regression models for each cancer site. The time origin was month of diagnosis, and chemotherapy was treated as a time-dependent exposure from the time of first claim for chemotherapy. Analyses accounted for the matched-pairs design and included unmatched demographic and disease characteristics as covariates. Observations were censored at the time of hospice admission or death. All analyses were performed in SAS version 9.2 (SAS Institute, Cary, NC).

RESULTS

Cohort Characteristics

We identified 36,676 patients who received chemotherapy for a metastatic cancer diagnosis, of whom 50% were matched to a patient with the same diagnosis who did not receive chemotherapy. The rate of successful matching varied from 31% of patients with esophageal cancer to 60% of patients with NSCLC. Distributions of sex, marital status, and comorbidity and region were similar between chemotherapy recipients who were successfully matched and those who were not. More of the matched chemotherapy recipients were from the South (26% v 17%), more were white (87% v 83%), and the matched patients were slightly older than the unmatched patients (mean age, 76 v 74 years). Median overall survival was substantially shorter in chemotherapy recipients who were successfully matched compared with those who were unmatched. The difference in median survival between matched and unmatched chemotherapy recipients varied from 6 months in patients with pancreatic cancer (3 months v 9 months) to 20 months in patients with breast cancer (9 months v 30 months).

Lung cancer accounted for the greatest proportion of cases included in the analysis (41%, NSCLC and SCLC combined), followed by colorectal cancer (20%) and pancreatic cancer (15%; Table 1). Just over half of the patients were men, 87% were white, and 58% had a Charlson comorbidity score of zero. Thirty-five percent of chemotherapy recipients also received radiation therapy, compared with 24% of those who did not receive chemotherapy. The chemotherapy cohort included a greater proportion of patients who were married, resided in urban areas, and lived in higher-income census tracts. Although chemotherapy recipients and nonrecipients were matched on survival duration (± 3 months or ± 1 month), in the better-prognosis cancers, survival was approximately 1 month greater on average in patients who received chemotherapy compared with those who did not. In the poorer-prognosis cancers, the difference in mean survival was less than 0.5 months. Fifty-six percent of chemotherapy recipients and 62% of nonrecipients entered hospice at any time before death.

Table 1.

Characteristics of Cohort by Receipt of Chemotherapy

Characteristic Chemotherapy No Chemotherapy
No. Col % No. Col % P
All cancers 18,486 — 18,486 —
Site
 Bladder 206 1 206 1
 Breast 1,127 6 1,127 6
 Colorectal 3,625 20 3,625 20
 Esophageal 278 2 278 2
 Non–small-cell lung 4,705 25 4,705 25
 Ovary 630 3 630 3
 Pancreas 2,813 15 2,813 15
 Prostate 1,321 7 1,321 7
 Small-cell lung 2,996 16 2,996 16
 Stomach 785 4 785 4
Age at diagnosis, years
 66-69 3,590 19 3,071 17 < .001
 70-74 4,934 27 4,418 24
 75-79 5,076 27 4,777 26
 80-84 3,384 18 4,029 22
 ≥ 85 1,502 8 2,191 12
Sex
 Male 9,641 52 9,641 52
 Female 8,845 48 8,845 48
Race
 White 16,105 87 16,105 87
 Black 1,427 8 1,427 8
 Other 954 5 954 5
Census tract median income, quartile
 1st 4,094 22 4,899 27 < .001
 2nd 4,558 25 4,799 26
 3rd 4,742 26 4,537 25
 4th 5,059 27 4,198 23
 Unknown 33 < 1 53 < 1
Urban-rural residence
 Metropolitan 15,464 84 15,057 81 < .001
 Nonmetropolitan 3,022 16 3,429 19
Region
 Northeast 4,095 22 4,095 22
 South 4,880 26 4,880 26
 Midwest 2,210 12 2,210 12
 West 7,301 39 7,301 39
Married
 Yes 10,310 56 8,341 45 < .001
 No 7,659 41 9,518 51
 Unknown 517 3 627 3
Charlson comorbidity score
 0 10,745 58 10,745 58
 1 4,537 25 4,537 25
 2+ 3,204 17 3,204 17
Radiation therapy
 Yes 6,562 35 4,369 24 < .001
 No 11,924 65 14,117 76
Open in a new tab

NOTE. Column percentages show distribution of characteristics within each group (chemotherapy recipients and nonrecipients). Chemotherapy recipients and nonrecipients were pair-matched by age, sex, race, SEER region, comorbidity score, and survival duration.

Hospitalizations

Ninety-two percent of chemotherapy recipients had at least one hospital admission for any cause, compared with 83% of matched nonrecipients (Fig 1). Just over half of the chemotherapy cohort had a likely toxicity-related hospitalization, compared with 34% of their matched peers. The mean number of hospitalizations was also greater among those who received chemotherapy than those who did not (2.62 v 1.96). Of the substantial proportion of patients in both groups who ever entered hospice, less than 2% had a first hospitalization after hospice admission—events that were excluded from the primary end point.

FIG 1.

FIG 1.

Open in a new tab

Hospitalization by cause and receipt of chemotherapy.

The proportion of patients with toxicity-related hospitalization and the unadjusted relative risk associated with chemotherapy varied by cancer site (Table 2). The largest difference in risk of likely toxicity-related hospitalization, comparing patients who did and did not receive chemotherapy, was seen in SCLC (56% v 30%; relative risk, 1.87) and the smallest difference was in prostate cancer (56% v 48%; relative risk, 1.17). The unadjusted probability of a toxicity-related hospitalization was significantly greater in the group that received chemotherapy (P < .05).

Table 2.

Impact of Chemotherapy on Risk of Hospitalization by Cancer Site

Cancer Site Chemotherapy (%) No Chemotherapy (%) Unadjusted RR Adjusted HR (95% CI) P
Toxicity-related
 Bladder 57 43 1.33 0.99 (0.56 to 1.76) NS
 Breast 54 45 1.20 1.43 (1.15 to 1.79) .0016
 Colorectal 51 34 1.50 1.49 (1.29 to 1.72) < .001
 Esophageal 44 30 1.47 2.00 (1.11 to 3.60) .0208
 Non–small-cell lung 52 36 1.44 1.54 (1.37 to 1.72) < .001
 Ovary 55 35 1.57 1.38 (0.96 to 1.98) NS
 Pancreas 41 27 1.52 1.59 (1.34 to 1.87) < .001
 Prostate 56 48 1.17 1.22 (1.01 to 1.47) .0443
 Small-cell lung 56 30 1.87 1.87 (1.58 to 2.20) < .001
 Stomach 48 30 1.60 1.78 (1.34 to 2.35) < .001
Any cause
 Bladder 95 83 1.14 1.35 (0.64 to 2.86) NS
 Breast 91 88 1.03 1.47 (1.13 to 1.90) .0036
 Colorectal 95 89 1.07 1.44 (1.16 to 1.78) .001
 Esophageal 90 74 1.22 2.55 (1.55 to 4.20) < .001
 Non–small-cell lung 90 81 1.11 1.59 (1.40 to 1.81) < .001
 Ovary 97 88 1.10 2.31 (1.64 to 3.26) < .001
 Pancreas 88 81 1.09 1.28 (1.09 to 1.51) .0029
 Prostate 89 80 1.11 1.30 (1.04 to 1.62) .0230
 Small-cell lung 94 81 1.16 2.31 (2.02 to 2.63) < .001
 Stomach 92 84 1.10 1.72 (1.28 to 2.32) < .001
Open in a new tab

NOTE. Chemotherapy recipients and nonrecipients were pair-matched by age, sex, race, SEER region, comorbidity score, and survival duration. Unadjusted relative risk compares the proportion of patients with a hospitalization in the chemotherapy group to the proportion in the no chemotherapy group. The unadjusted probability of toxicity-related rate was significantly greater in the group that received chemotherapy (P < .05) by McNemar's test for matched pairs. Adjusted hazard ratios from multivariable regression model, where chemotherapy was modeled as a time-dependent exposure. Hazard ratios for the impact of chemotherapy on risk of each outcome were adjusted for additional characteristics, including urban-rural residence, marital status, income quartile, and receipt of radiation therapy.

Abbreviations: HR, hazard ratio; NS, not statistically significant; RR, relative risk.

Among specific causes of toxicity-related hospitalizations, some diagnoses were more common among chemotherapy recipients than their matched peers (Appendix Fig A1). Compared with those who do not receive chemotherapy, chemotherapy recipients had substantially higher rates of hospitalization for infection or fever (21% v 15%), hematologic complications (11% v 3%), dehydration (13% v 6%), and pulmonary embolus or deep venous thrombosis (9% v 4%). The proportions of patients hospitalized for gastrointestinal complications, malnutrition, cardiac complications, or constitutional symptoms were somewhat greater in chemotherapy recipients, although the differences were modest.

In nearly all cancers, chemotherapy recipients had a greater risk of hospitalization—for likely treatment toxicities and for any cause—than those who did not, accounting for the matched-pairs design and controlling for additional demographic characteristics (Table 2). The adjusted impact of chemotherapy was greatest among patients with esophageal cancer, on both the risk of likely toxicity-related hospitalization (adjusted hazard ratio, 2.00; 95% CI, 1.11 to 3.60; P < .02) and the risk of hospitalization for any cause (adjusted hazard ratio, 2.55; 95% CI, 1.55 to 4.20; P < .001). In patients with bladder cancer, chemotherapy did not have a significant impact on the risk of toxicity-related or any-cause hospitalization.

DISCUSSION

Cancer remains the second leading cause of mortality in the United States.3 Despite advances in tumor profiling and drug development, improvements in health outcomes have been limited, and many new drugs are associated with both adverse effects and high costs.19-21 Thus, for patients with metastatic cancer, the expected benefits of chemotherapy must be weighed against the possibility of treatment-related toxicities.6,22 In our analysis of a population-based cohort of older adults with advanced cancer, chemotherapy was associated with increased risk of hospitalization, and a considerable proportion of admissions were attributable to likely chemotherapy-related toxicities.

Although some amount of chemotherapy-related inpatient care may be unavoidable, hospitalizations may be occurring more frequently than they should.23,24 One study found that approximately one third of hospital admissions in patients receiving chemotherapy were treatment related, and almost half of those patients were receiving treatment with palliative, as opposed to curative, intent.25 Some hospitalizations may reflect a failure to adequately manage common problems with prophylactic interventions and efficient outpatient care.4,26-29 In our analysis, infection and fever, dehydration, and hematologic complications were the most common treatment-related diagnoses associated with hospitalization. Neutropenic fever is potentially preventable with evidence-based use of growth factors for patients receiving myelosuppressive chemotherapy regimens.30 Some patients with infection and neutropenia can be, and prefer to be, managed with close observation and oral antibiotics in the outpatient setting, incurring substantially lower costs and reducing their risk of hospital-acquired infection.26,27,30-32

In many advanced cancers the efficacy of chemotherapy, especially in early lines of treatment, has been well established by numerous clinical trials. In these instances, the expected benefits of chemotherapy—in terms of symptom control and extension of progression-free survival—are nontrivial and may clearly exceed the burden of adverse effects. More rarely, chemotherapy in the metastatic setting has been shown to extend overall survival.33,34 We would not dissuade patients from treatment when evidence supports a likely and meaningful improvement in quality or length of life. However, the less sensitive a cancer is to chemotherapy, the greater the consideration needed when making treatment decisions.8 Compared with the other tumor types we evaluated, advanced NSCLC and metastatic cancers of the stomach, esophagus, pancreas, and prostate are generally less sensitive to chemotherapy.5,8 However, we found that chemotherapy almost doubled the risk of likely toxicity-related hospitalization for patients with metastatic cancer of the stomach or esophagus and increased the risk by more than 50% in metastatic pancreatic cancer and NSCLC. Given the limited expected benefit of chemotherapy in these cancers, it is not clear how many patients would choose chemotherapy if they understood the likely toxicities and resulting detriment to the quality of their already reduced life expectancy.

The point at which the risks of chemotherapy exceed the benefits is highly subjective and often unknown. As disease progresses despite palliative chemotherapy, the risk increases that survival improvements and symptom relief may be marginal and that the harmful effects of treatment may dominate.35 For physicians, expert recommendations provide little clear or evidence-based guidance about when to stop offering or administering chemotherapy. For example, the National Comprehensive Cancer Network guidelines recommend multiple lines of cancer-directed therapy for patients with tumors of limited survival times, despite a paucity of high-quality evidence supporting a meaningful benefit in terms of survival or quality of life.36 In contrast, 40% of patients with breast cancer will have some disease control from fourth-line chemotherapy for up to 4 months, even if survival is not extended.37 Regardless of cancer type, changing the focus of decision making from which drug to try next to how best to preserve function and quality of life can be especially challenging for clinicians.38

Given the shifting and subjective nature of the tradeoff between benefits and harms of chemotherapy in advanced cancer, individualization of treatment and early palliative care may be the optimal approach.39,40 The survival benefit found in a recent randomized trial of this strategy among patients with lung cancer was unexpected, but plausible.39 Early palliative care may enhance the management of adverse effects, allowing patients to receive more regimens of chemotherapy. The integrated model of care may also facilitate cessation of anticancer therapy at the end of life when the harmful adverse effects of aggressive treatment exceed the possible benefits.40

There are several limitations to our analysis. We compared cohorts of patients with metastatic cancer who did and did not receive chemotherapy, matched by key demographic and health characteristics and controlling for important unmatched characteristics, to estimate the difference in risk of hospitalization likely attributable to chemotherapy. This approach, consistent with previous studies of treatment complications41-43 and with the epidemiologic concept of attributable or excess risk,44 does not require direct attribution of each hospitalization to a specific cause, a task that is challenging even with detailed clinical information.45 However, there may have been residual confounding by unmeasured factors, such as pretreatment symptom burden, performance status, and functional status, other risk factors for treatment-related adverse effects, patient preferences, or physician recommendations. Omission of these unobserved factors that may be associated with receipt of chemotherapy and with the risk of hospitalization could have biased our estimates.46 Similarly, our definition of likely toxicity-related hospitalizations does not definitively distinguish hospitalizations directly attributable to chemotherapy toxicity from those attributable to cancer symptoms. Indeed, many patients who did not receive chemotherapy had a hospitalization associated with a likely toxicity-related diagnosis. However, by identifying hospitalizations for diagnoses that reflect common chemotherapy toxicities, we could analyze the subset of hospitalizations most likely to be associated with adverse effects of chemotherapy.

Although Medicare claims can be used to identify receipt of chemotherapy, we could not reliably distinguish agents, regimens, or dosages that were appropriate and potentially beneficial from those that were not. We could not reliably identify distinct, successive lines of therapy, nor could we identify patients who received integrated early palliative care. We identified only treatment-related complications severe enough to warrant a hospital admission, and thus we may have underestimated the incidence of some toxicities.47 And although the complications we identified are likely associated with changes in physical function, social function, and quality of life, we were not able to assess these end points directly. Finally, our results may not be generalizable to all older adults with metastatic solid tumors. Although the SEER-Medicare population is fairly representative of the nationwide Medicare population,14 our analysis was limited to the subset of patients who received chemotherapy and could be matched to nonrecipients. Matched and unmatched chemotherapy recipients were similar with respect to important demographic and health characteristics, including age and comorbidity. However, even though all patients had a diagnosis of metastatic cancer, unmatched chemotherapy recipients had substantially longer survival after diagnosis, suggesting that they were healthier or had more favorable disease characteristics than the matched patients included in the study. Thus, our findings might be limited to older patients with advanced cancer who have a generally poor prognosis or limited expected survival.

Older patients receiving chemotherapy for incurable advanced malignancies are at high risk of hospitalization, of which a nonnegligible proportion is likely attributable to adverse effects of treatment. Infection, fever, dehydration, and hematologic complications constitute a large proportion of these events, some of which may be preventable through evidence-based patient management, prophylactic interventions, and effective outpatient care. Understanding the common reasons for hospital admissions and developing toxicity management programs and educational resources for patients and their families may help minimize adverse effects and reduce hospitalizations in this population with limited life expectancy. Furthermore, the integration of early palliative care into the treatment pathway of patients with advanced cancer may improve health outcomes and preserve quality of life and of death.

Our study provides information about the risks of hospitalization in older adults with advanced cancer and the substantial increase in these risks associated with chemotherapy. This information is applicable to the many patients diagnosed with metastatic cancer annually who face difficult decisions about the risks they are willing to accept for therapies with varying benefits in terms of cancer control, quality of life, and survival. Our findings are especially important now, given the high cost of new chemotherapy drugs, many of which offer only modest benefits. Our results also inform the movement toward patient-centered care at the end of life.

Acknowledgment

Supported by the Health Research Board (Ireland) through the Health Research Board PhD Scholars Program in Health Service Research Grant No. PHD/2007/16, and by a Cancer Center Support Grant from the National Cancer Institute (P30 CA008748).

Appendix

FIG A1.

FIG A1.

Open in a new tab

Specific causes of toxicity-related hospitalizations. DVT, deep venous thrombosis; PE, pulmonary embolus.

Table A1.

Site and Extent-of-Disease Codes Used to Identify Cancers Metastatic at Diagnosis

Tumor Type SEER Site Recode (CODKM) ICD-O-3 Topography Code EOD-10 Extent Code Collaborative Staging Metastasis Code
Breast 46 C50.0-C50.6, C50.8-C50.9 85 10, 40, 42, 44, 50
Bladder 58 C67.0-C67.9 85 10, 11, 40, 50
Colorectal 15-23, 25-26 C18.0-C18.9, C19.9, C20.9 85 08, 10, 40, 50 (colon)
10, 11, 12, 40, 50 (rectum)
Ovarian 50 C56.9 85 10, 40, 50
Stomach 12 85 10, 40, 50
Pancreas 33 C25.0-C25.9 85 10, 40, 50
Esophagus 11 C15.0-C15.5, C15.8, C15.9 85 10, 11, 12, 40, 50
Lung 39 C34.0-C34.3, C34.8, C34.9 85 10, 35, 37, 39, 40, 50
Prostate 54 C61.9 85 11, 12, 30, 35, 40, 45, 50, 55
Open in a new tab

Abbreviations: EOD, extent of disease; ICD-O-3, International Classification of Diseases for Oncology, 3rd edition.

Table A2.

Billing Codes Used to Identify Chemotherapy in Medicare Claims

Modality ICD-9 Procedure Codes Health Care Common Procedure Coding System Codes*
Systemic chemotherapy† 99.25, V58.1, V66.2, V67.2, E93.07, E93.31 J9000-J9999, J8520-J8999, G0355-G0362, G9021-G9032, C9414-C9419, C9420-9438, S9325-S9379, S9494-S9497, 96400-96599, 36260, 36640, 95990, 95991, A4301, C1166, C1167, C1168, C9110, C9205, C9207, C9213, C9214, C9215, C9216, C9411, E0782, E0783, E0785, E0786, J0640, J2405, K0415, KO416, Q0083, Q0084, Q0085, Q0179, S0177, S0181
Open in a new tab

Abbreviation: ICD-9, International Classification of Diseases, 9th edition.

*

Codes for hormonal therapy (J9003, J9165, J9175, J9202, J9209, J9212, J9213, J9214, J9215, J9216, J9217, J9218, J9219, J9225, J9226, J9240, J9395) were excluded.

†

Includes intravenous, intramuscular, and orally administered agents, as indicated by specific codes.

AUTHOR CONTRIBUTIONS

Conception and design: Caitriona B. O’Neill, Eileen M. O’Reilly, Martin C. Henman, Peter B. Bach, Elena B. Elkin

Data analysis and interpretation: All authors

Manuscript writing: All authors

Final approval of manuscript: All authors

AUTHORS’ DISCLOSURES OF POTENTIAL CONFLICTS OF INTEREST

Hospitalizations in Older Adults With Advanced Cancer: The Role of Chemotherapy

The following represents disclosure information provided by authors of this manuscript. All relationships are considered compensated. Relationships are self-held unless noted. I = Immediate Family Member, Inst = My Institution. Relationships may not relate to the subject matter of this manuscript. For more information about ASCO's conflict of interest policy, please refer to www.asco.org/rwc or jop.ascopubs.org/site/misc/ifc.xhtml.

Caitriona B. O’Neill

Employment: Novartis

Coral L. Atoria

No relationship to disclose

Eileen M. O’Reilly

No relationship to disclose

Martin C. Henman

No relationship to disclose

Peter B. Bach

Leadership: Exam Works Stock or Other Ownership: Exam Works Honoraria: The American Journal of Managed Care, America’s Health Insurance Plans, Barclays, Express Scripts, Goldman Sachs, McKinsey and Company, National Comprehensive Care Network, Association of Community Cancer Centers Consulting or Advisory Role: Foundation Medicine Travel, Accommodations, Expenses: America’s Health Insurance Plans, Barclays, The American Journal of Managed Care, Express Scripts, Goldman Sachs, McKinsey and Company, National Comprehensive Cancer Network, Association of Community Cancer Centers

Elena B. Elkin

No relationship to disclose

References

  • 1.American Cancer Society : Cancer Facts and Figures 2012 http://www.cancer.org/acs/groups/content/@epidemiologysurveilance/documents/document/acspc-031941.pdf
  • 2.DeSantis CE Lin CC Mariotto AB, etal: Cancer treatment and survivorship statistics, 2014 CA Cancer J Clin 64:252–271,2014 [DOI] [PubMed] [Google Scholar]
  • 3.Murphy SL, Xu J, Kochanek KD: Deaths: Final data for 2010 Natl Vital Stat Rep 61:1–117,2013 [PubMed] [Google Scholar]
  • 4.Peppercorn JM Smith TJ Helft PR, etalAmerican Society of Clinical Oncology : American society of clinical oncology statement: Toward individualized care for patients with advanced cancer J Clin Oncol 29:755–760,2011 [DOI] [PubMed] [Google Scholar]
  • 5.McIllmurray M: Hanks G, Cherny NI, Christakis NA, et al.The medical treatment of cancer in palliative care, Oxford Textbook of Palliative Medicine 513–525,2010. Oxford: Oxford University Press [Google Scholar]
  • 6.Fojo T, Grady C: How much is life worth: Cetuximab, non-small cell lung cancer, and the $440 billion question J Natl Cancer Inst 101:1044–1048,2009 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Weeks JC Catalano PJ Cronin A, etal: Patients’ expectations about effects of chemotherapy for advanced cancer N Engl J Med 367:1616–1625,2012 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Näppä U Lindqvist O Rasmussen BH, etal: Palliative chemotherapy during the last month of life Ann Oncol 22:2375–2380,2011 [DOI] [PubMed] [Google Scholar]
  • 9.Earle CC Landrum MB Souza JM, etal: Aggressiveness of cancer care near the end of life: Is it a quality-of-care issue? J Clin Oncol 26:3860–3866,2008 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Earle CC Neville BA Landrum MB, etal: Trends in the aggressiveness of cancer care near the end of life J Clin Oncol 22:315–321,2004 [DOI] [PubMed] [Google Scholar]
  • 11.Emanuel EJ Young-Xu Y Levinsky NG, etal: Chemotherapy use among Medicare beneficiaries at the end of life Ann Intern Med 138:639–643,2003 [DOI] [PubMed] [Google Scholar]
  • 12.Ho TH Barbera L Saskin R, etal: Trends in the aggressiveness of end-of-life cancer care in the universal health care system of Ontario, Canada J Clin Oncol 29:1587–1591,2011 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Bach PB, Schrag D, Begg CB: Resurrecting treatment histories of dead patients: A study design that should be laid to rest JAMA 292:2765–2770,2004 [DOI] [PubMed] [Google Scholar]
  • 14.Warren JL Klabunde CN Schrag D, etal: Overview of the SEER-Medicare data: Content, research applications, and generalizability to the United States elderly population Med Care 40:IV-3–IV-18,2002 [DOI] [PubMed] [Google Scholar]
  • 15.Klabunde CN Potosky AL Legler JM, etal: Development of a comorbidity index using physician claims data J Clin Epidemiol 53:1258–1267,2000 [DOI] [PubMed] [Google Scholar]
  • 16.Hassett MJ O’Malley AJ Pakes JR, etal: Frequency and cost of chemotherapy-related serious adverse effects in a population sample of women with breast cancer J Natl Cancer Inst 98:1108–1117,2006 [DOI] [PubMed] [Google Scholar]
  • 17.Hardy D Liu C-C Cormier JN, etal: Cardiac toxicity in association with chemotherapy and radiation therapy in a large cohort of older patients with non-small-cell lung cancer Ann Oncol 21:1825–1833,2010 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Du XL, Osborne C, Goodwin JS: Population-based assessment of hospitalizations for toxicity from chemotherapy in older women with breast cancer J Clin Oncol 20:4636–4642,2002 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Bach PB: Limits on Medicare’s ability to control rising spending on cancer drugs N Engl J Med 360:626–633,2009 [DOI] [PubMed] [Google Scholar]
  • 20.Brooks GA Li L Sharma DB, etal: Regional variation in spending and survival for older adults with advanced cancer J Natl Cancer Inst 105:634–642,2013 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Kantarjian H, Zwelling L: Cancer drug prices and the free-market forces Cancer 119:3903–3905,2013 [DOI] [PubMed] [Google Scholar]
  • 22.Sandler A Gray R Perry MC, etal: Paclitaxel-carboplatin alone or with bevacizumab for non-small-cell lung cancer N Engl J Med 355:2542–2550,2006 [DOI] [PubMed] [Google Scholar]
  • 23.Kolodziej M Hoverman JR Garey JS, etal: Benchmarks for value in cancer care: An analysis of a large commercial population J Oncol Pract 7:301–306,2011 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Hassett MJ Rao SR Brozovic S, etal: Chemotherapy-related hospitalization among community cancer center patients Oncologist 16:378–387,2011 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Krzyzanowska MK Treacy J Maloney B, etal: Development of a patient registry to evaluate hospital admissions related to chemotherapy toxicity in a community cancer center J Oncol Pract 1:15–19,2005 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Jacobson JO, Mulvey TM: Time to focus on inpatient safety: Revision of the American Society Of Clinical Oncology/Oncology Nursing Society chemotherapy administration safety standards J Clin Oncol 30:1021–1022,2012 [Google Scholar]
  • 27.Slavin MA, Thursky KA: Outpatient therapy for fever and neutropenia is safe but implementation is the key J Clin Oncol 31:1128–1129,2013 [DOI] [PubMed] [Google Scholar]
  • 28.Schnipper LE Smith TJ Raghavan D, etal: American Society of Clinical Oncology identifies five key opportunities to improve care and reduce costs: The top five list for oncology J Clin Oncol 30:1715–1724,2012 [DOI] [PubMed] [Google Scholar]
  • 29.McCabe MS Bhatia S Oeffinger KC, etal: American Society of Clinical Oncology statement: Achieving high-quality cancer survivorship care J Clin Oncol 31:631–640,2013 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Lyman GH, Lyman CH, Agboola O: Risk models for predicting chemotherapy-induced neutropenia Oncologist 10:427–437,2005 [DOI] [PubMed] [Google Scholar]
  • 31.Hendricks AM, Loggers ET, Talcott JA: Costs of home versus inpatient treatment for fever and neutropenia: Analysis of a multicenter randomized trial J Clin Oncol 29:3984–3989,2011 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 32.Teuffel O Cheng S Ethier MC, etal: Health-related quality of life anticipated with different management strategies for febrile neutropenia in adult cancer patients Support Care Cancer 20:2755–2764,2012 [DOI] [PubMed] [Google Scholar]
  • 33.Goldberg RM Tabah-Fisch I Bleiberg H, etal: Pooled analysis of safety and efficacy of oxaliplatin plus fluorouracil/leucovorin administered bimonthly in elderly patients with colorectal cancer J Clin Oncol 24:4085–4091,2006 [DOI] [PubMed] [Google Scholar]
  • 34.O’Shaughnessy J Miles D Vukelja S, etal: Superior survival with capecitabine plus docetaxel combination therapy in anthracycline-pretreated patients with advanced breast cancer: Phase III trial results J Clin Oncol 20:2812–2823,2002 [DOI] [PubMed] [Google Scholar]
  • 35.Ramsey SD Martins RG Blough DK, etal: Second-line and third-line chemotherapy for lung cancer: Use and cost Am J Manag Care 14:297–306,2008 [PubMed] [Google Scholar]
  • 36.National Comprehensive Cancer Network : NCCN clinical practice guidelines in oncology. http://www.nccn.org
  • 37.Dufresne A Pivot X Tournigand C, etal: Impact of chemotherapy beyond the first line in patients with metastatic breast cancer Breast Cancer Res Treat 107:275–279,2008 [DOI] [PubMed] [Google Scholar]
  • 38.Helft PR: Necessary collusion: Prognostic communication with advanced cancer patients J Clin Oncol 23:3146–3150,2005 [DOI] [PubMed] [Google Scholar]
  • 39.Temel JS Greer JA Muzikansky A, etal: Early palliative care for patients with metastatic non-small-cell lung cancer N Engl J Med 363:733–742,2010 [DOI] [PubMed] [Google Scholar]
  • 40.Greer JA Pirl WF Jackson VA, etal: Effect of early palliative care on chemotherapy use and end-of-life care in patients with metastatic non-small-cell lung cancer J Clin Oncol 30:394–400,2012 [DOI] [PubMed] [Google Scholar]
  • 41.Holcomb CN Graham LA Richman JS, etal: The incremental risk of coronary stents on postoperative adverse events: A matched cohort study Ann Surg doi: 10.1097/SLA.0000000000001246 [e-pub ahead of print on April 17, 2015] [DOI] [PubMed] [Google Scholar]
  • 42.Jarosek SL Virnig BA Chu H, etal: Propensity-weighted long-term risk of urinary adverse events after prostate cancer surgery, radiation, or both Eur Urol 67:273–280,2015 [DOI] [PubMed] [Google Scholar]
  • 43.Vaara ST Pettilä V Kaukonen KM, etalFinnish Acute Kidney Injury Study Group : The attributable mortality of acute kidney injury: A sequentially matched analysis Crit Care Med 42:878–885,2014 [DOI] [PubMed] [Google Scholar]
  • 44.Greenland S: Concepts and pitfalls in measuring and interpreting attributable fractions, prevented fractions, and causation probabilities Ann Epidemiol 25:155–161,2015 [DOI] [PubMed] [Google Scholar]
  • 45.Brooks GA, Jacobson JO, Schrag D: Clinician perspectives on potentially avoidable hospitalizations in patients with cancer JAMA Oncol 1:109–110,2015 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 46.Austin PC Mamdani MM Stukel TA, etal: The use of the propensity score for estimating treatment effects: Administrative versus clinical data Stat Med 24:1563–1578,2005 [DOI] [PubMed] [Google Scholar]
  • 47.Potosky AL Warren JL Riedel ER, etal: Measuring complications of cancer treatment using the SEER-Medicare data Med Care 40:IV-62–IV-68,2002 [DOI] [PubMed] [Google Scholar]

Articles from Journal of Oncology Practice are provided here courtesy of American Society of Clinical Oncology

RESOURCES