Impact of Physical and Functional Limitations on Health Care Utilization in Older Cancer Survivors: A Medicare Current Beneficiary Survey

Abstract

Objective: This study examines effects of physical and functional limitations on health care utilization among older cancer survivors, compared with those without cancer and without physical and functional limitations. Method: Medicare Current Beneficiary Survey data from 2008 to 2011 were used. Physical limitations (PL), activities of daily living (ADL), and instrumental activities of daily living (IADL) were measured on a 5-point scale. Propensity score weighting was developed using logistic regressions. Results: Older cancer survivors with physical and functional limitations had higher rate of emergency department visits than those without limitations (PL: 21.8% vs.17%, adjusted odds ratio [aOR]:1.72, 95% confidence interval [CI]: [1.26, 2.35], p < .05; ADL: 25.8% vs.17.4%, aOR: 2.68, 95% CI: [1.86, 3.86], p < .001), and higher cost of hospitalization (IADL: M = US$24,916, SD: 3,877.1). Conclusion: Older cancer survivors with physical and functional limitations had higher health care utilization compared with those without cancer. Addressing complex and unique health care needs in this population will help reduce excess burden on the health care system.

Keywords

older cancer survivors Medicare health care utilization physical limitations functional limitations

Introduction

Longevity among cancer survivors is steadily increasing over the past three decades. About 60% of cancer patients are expected to survive 5 or more years after cancer diagnosis (DeSantis, Siegel, & Jemal, 2015) due to early detection of cancer, advanced medical technologies, newer treatments, and improved cancer follow-up care (Siegel & Miller, 2019; Sweeney et al., 2006). In 2018, the overall 5-year relative survival rate was 68% among Caucasians and 61% among African Americans. The cancer survivor population in the United States is expected to reach 20.3 million by 2026, and older cancer survivors account for almost two thirds of this population (National Cancer Institute, 2018). In addition, the estimated national expenditure of cancer care in 2017 was US$147.3 billion (National Cancer Institute, 2018). As the older cancer survivor population continues to increase, the health care–related cost is also likely to increase.

Older cancer survivors represent a vulnerable population who suffer from prolonged consequences of cancer and its treatments. They often receive fewer optimal cancer treatment options due to their frail conditions which may further deteriorate their overall well-being. They are also more likely to have multiple chronic conditions, thus experiencing additional physical and functional limitations compared with individuals without a history of cancer (de Moor et al., 2013; Hardy, McGurl, Studenski, & Degenholtz, 2010). Physical limitations and functional limitations are two key components related to health care of older cancer survivors. Physical limitations refer to the reduced strength to perform basic physical activities such as walking, lifting, stooping, sitting, and so on (Stineman et al., 2014). Functional limitations refer to the inability to perform essential daily tasks to live a meaningful life at home (Field & Jette, 2007), which include both activities of daily living (ADL) such as using toilet, dressing, bathing, eating, and walking around the house; and instrumental activities of daily living (IADL) such as shopping, managing bills, and participating in events of social life (Katz, 1983; Lawton & Brody, 1970).

While the benefits of building physical capacities in this population have been established in literature (Johnson, Fallon, & Berg, 2019; Stacey, James, Chapman, Courneya, & Lubans, 2015; Van Roekel et al., 2015), not much is known about the effects of physical and functional limitations on health care utilization in this population. Compared with individuals with no history of cancer, cancer survivors are more likely to have functional limitations and poor health conditions (Hewitt, Rowland, & Yancik, 2003). Moreover, physical and functional limitations can be interrelated and affect each other. For instance, more physical limitations lead to increased limitations in ADL and IADL causing more dependence on caregivers to assist with daily tasks. In addition, these limitations may result in negligence of personal health and social isolation, which could further deteriorate physical and mental health among older cancer survivors (Grov, Fosså, & Dahl, 2010).

Furthermore, older cancer survivors are more likely to use supportive medical services and thus, have higher health care utilization than those without cancer. The national estimates of health care expenditure showed older cancer survivors with higher utilization rates of health care services, compared with other chronically ill patients (Bernard, Farr, & Fang, 2011). Older cancer survivors not only had higher overall health care expenditure but also had excess health care expenditure which continued even after many years of cancer diagnosis, compared with individuals without a history of cancer (Guy et al., 2013). Fried et al. demonstrated that functional disability in older adults imposed a substantial burden on government health care services (Fried, Bradley, Williams, & Tinetti, 2001). They also argued that assessment of functional status was the best predictor of health care expenditure in older people. In addition, because older cancer survivors are more likely to access supportive medical services, visit primary care physicians, and consult specialists (Weaver, Rowland, Bellizzi, & Aziz, 2010), more coordination between primary care physicians and other specialists is needed to address their complex health care needs. Effective patient–physician communication and shared decision making are key to manage the health care needs of cancer survivors (Levit, Balogh, Nass, & Ganz, 2013). Implementation of shared care models among the cancer survivor population may improve the health care outcomes in them (Swartz, Chapman, & Doggett, 2017). Providing care for older cancer survivors with physical and functional limitations also exerts a significant burden on the health care system, presenting new challenges and escalating already exorbitant health care cost (Heins, Schellevis, Rijken, van der Hoek, & Korevaar, 2012).

There are only a few population-based studies conducted so far about the effects of physical and functional limitations on health care utilization among older cancer survivors in the United States (Karve et al., 2015; Kurtz, Kurtz, Given, & Given, 2005; Manzano, Luo, Elting, George, & Suarez-Almazor, 2014). The broad picture inference of these studies suggested that the health care utilization tended to escalate with a decrease in physical and functional capacities among older cancer survivors primarily due to increased inpatient, emergency department, and outpatient hospital visits. However, most of these studies are descriptive in nature and not adequately designed to satisfy causal inference criteria. Therefore, prior research may underestimate or overestimate the health care utilization among older cancer survivors. To fill this gap in literature, we examined the effects of physical and functional limitations on health care utilization among older cancer survivors compared with those without cancer in a population-based study, using propensity score analysis method. We hypothesized that older cancer survivors with physical and functional limitations had higher health care utilization compared with those without cancer and to those without limitations.

Data and Method

Study Cohort

Cost and Use files from Medicare Current Beneficiary Survey (MCBS) from 2008 to 2011 were used in this study. MCBS, administered by the Centers for Medicare and Medicaid Services (CMS), is a nationally representative longitudinal, rotating, multistage survey of the entire older Medicare beneficiaries. MCBS conducts three interviews each year over a period of 4 years to track health status changes and health care utilization among older Medicare beneficiaries. During each Fall round of survey, one third of participants are rotated out and replaced by an equivalent number of new participants. In the first interview, baseline information, including demographics, medical history, health status, and information on health care utilization, is collected. During the third interview each year, information from previous years is verified; therefore, longitudinal data are available for up to 3 consecutive years. Various outcomes are assessed for each respondent, but functional status is assessed only in the first 2 years of follow-up. Both cross-sectional and longitudinal survey weights are provided for analytical purposes to account for the complex sampling design and nonresponse in the surveys. The weighted summary statistics represent those of targeted populations (MCBS, 2016).

In this study, we used information on sociodemographic characteristics, health status, physical limitations, and functional limitations (ADL and IADL) collected during each Fall round of survey. Information collected in other rounds on access to health care and health care utilization were also used in this study. We restricted our study cohort to participants aged 65 years or older (n = 23,695 at baseline) and excluded participants who were diagnosed with cancer within 1 year prior to the baseline survey or those who developed new cancer during the study period (n = 821), and those who died during the survey years (n = 1,458) to create a relatively homogeneous study sample to measure the outcomes of health care utilization during the end of life period. Although it could lead to survival bias among cancer survivors and may underestimate the true differences between cancer survivors and those without cancer, our study cohort is more likely to represent long-term cancer survivors. We also included only those participants who had at least one follow-up interview during the study period. In addition, participants enrolled in Health Maintenance Organization (n = 8,736) and those with missing values for functional limitations (n = 99) were also excluded from the final cohort. No imputation strategy was applied in the analyses due to small percentage of missing values. Our final sample consisted of 23,695 participants, representing a weighted total of 110,086,170 older Medicare beneficiaries.

Assessment of Physical and Functional Limitations

MCBS includes measures for self-reported physical limitations and functional limitations (ADL & IADL). Physical limitations include difficulties in stooping/crouching/kneeling, walking ¼ miles, reaching/extending arms above shoulder, lifting/carrying 10 lbs, and writing/handling small objects. Functional limitations include ADL and IADL. ADL limitations include difficulties in bathing, dressing, walking in the house, eating, getting in and out of chair or bed, and using the toilet, and IADL limitations include difficulties in using telephone, shopping, doing light house work, preparing meals, and paying bills. Each domain includes five items and each item is measured on a 5-point scale (1 = no difficulty, 2 = a little difficulty, 3 = some difficulty, 4 = a lot of difficulty, 5 = unable to do it). As the score distribution for each item was positively skewed (i.e., small number of people had a large score of 4 or 5 while the majority had a score of 0 or 1.), we dichotomized each item into a binary variable. Responses to questions with “a lot of difficulty” or “unable to do it” were considered as having a physical limitation or functional limitation for that item. For each domain of limitations, having two or more (out of total five or six) item-wise limitations was considered as having a limitation. This method is consistent with the International Classification of Functioning, Disability, and Health (ICF) created by Institute of Medicine in 1991(Pope & Tarlov, 1991).

Sociodemographic Characteristics and Health Status

MCBS also collects information on sociodemographic characteristics and health status during the face-to-face interviews. Sociodemographic characteristics include age at baseline survey (regrouped as 65-74 years, 75-84 years, and 85 years or older), gender, income (<$15,000, $15,000 - $30,000, and ≥$30,000), and race/ethnicity (regrouped as Caucasian, African-American, and Other).

Consistent with the definition adopted by National Cancer Institute, we defined cancer survivors as those having survived cancer from the time of diagnosis until their end of life (National Cancer Institute, 2015). History of cancer diagnosis was assessed through questions such as “ever told having a non-skin cancer,” “had a cancer past year,” and site (body part/organ) of cancer. Those with non-melanoma skin cancer were not considered cancer survivors in this study because of the nature of noninvasiveness of the skin cancer. Those diagnosed 1 year before the survey and during the survey years were excluded from this study to reduce the bias from the impact of cancer diagnosis and treatment on physical and functional limitations and health care utilization.

The comorbidities were measured based on the number of self-reported medical conditions, such as heart disease, stroke, arthritis, chronic obstructive pulmonary disease (COPD), paralysis/amputation, bone disease, diabetes, hypertension, psychiatric disorder, and neurological disease including Alzheimer’s disease and dementia. These were recorded as yes/no on the questionnaires. Multiple chronic conditions were defined as having two or more of the above chronic conditions, including hypertension. This is similar to the other comorbidity measures, such as modified Charlson’s Comorbidity Index in which hypertension is included (Romano, Roos, & Jollis, 1993).

Outcome Measures

Health care utilization was measured using variables rate of hospitalizations, rehospitalizations, emergency department visits, and 30-day hospital readmissions based on Medicare claims. Specifically, Medicare inpatient claims were searched for hospitalization throughout all survey years to create a longitudinal history of hospitalization for each individual. Then the rehospitalizations were counted since the first hospitalization during the study period. The duration between hospitalizations was assessed using admission and discharge dates on the claims. Similarly, the emergency department (ED) visits were identified using outpatient claims based on the type of services, facility codes, and detailed revenue codes. The duration between ED visits was assessed using claim admission and discharge dates as well. The total medical costs were summed from the Medicare reimbursed cost on the inpatient claims, outpatient claims, and physician claims (Medicare Part B).

Statistical Analysis

We employed propensity score analysis method based on the potential outcome causal inference framework (Rubin, 2004) to obtain a correct estimate of average differences in the prevalence of physical limitations and functional limitations (ADL & IADL) and health care utilization between cancer survivors and people without cancer. We first conducted a logistic regression using the baseline data with cancer status as the dependent variable, and age, race, gender, income level, all comorbidities, and their two-way interactions as independent variables. As suggested by DuGoff, Schuler, and Stuart (2014), the survey stratum and survey weight were also included as predictors in the logistic regression model, but the propensity score model was not weighted (DuGoff, Schuler, & Stuart, 2014). The propensity score (P), that is, the estimated probability of potentially having a cancer diagnosis for both cancer patients and people without cancer, was estimated from the logistic regression given below:

P = \frac{1}{(1 + e^{- (p r e d i c t e d l o g o d d s)})} .

We then calculated a new weight as the product of the original MCBS cross-sectional sampling weight and propensity score weight, where 1/P was for cancer patients, and 1/(1 – P) was for people without cancer. This new weight deducted more on those participants who were in the tails of the propensity score distributions. Next, we multiplied the new weight with the average P for cancer patients or the average 1 – P for people without cancer to correct the outliers in the new weights (also a form of stabilized weight) (Guo & Fraser, 2015). The final new weights were multiplied to the MCBS survey weight and then used for comparing physical and functional limitations between cancer survivors and those without cancer in all the analyses. The sociodemographic characteristics and comorbidities were assessed using weighted mean and standard error (SE) for continuous variables and weighted frequency along with percentage for categorical variables. The group comparisons were analyzed using appropriate t test and Rao-Scott χ² tests.

Furthermore, we assessed potential confounding effect prior to the final outcome model specification using the Mantel-Haenszel stratification analysis. Multivariable logistic regression models incorporating the above calculated new weights and, adjusting for both sociodemographic characteristics and comorbidities, were used to estimate the independent effect of cancer status on physical and functional limitations at baseline. This approach is similar to the double robust method used in propensity score analysis. The adjusted logistic models were also used to estimate the average predicted marginal probabilities of having limitations in each year. Similar process was employed for examining the health care utilization. All analyses were performed in SAS 9.4 (SAS Institute, Inc., Cary, NC) using the new weights to account for the multistage survey design and the appropriate subpopulation (domain) analysis. SAS procedures for survey data such as PROC SURVEYFREQ, SURVEYMEANS, and SURVEYLOGISTIC were used in the analyses. A p value of < .05 was used for statistical significance.

Results

Cancer survivors accounted for 19.3% of total MCBS participants in the study. The description of participants’ sociodemographic characteristics, comorbidities of cancer survivors, and site of cancer by cancer status after weighting by propensity score are presented in Table 1. There was no statistically significant difference among these factors between cancer and noncancer participants after weighting. The weighted mean age of participants was 74 years, with about 58% females among cancer survivors. There were 87% Caucasians and almost 21% of participants had an income of <$15,000 per year. No statistically significant difference was seen in various comorbidities, except for COPD (p = .01) and bone disease (p = .01). Around 89% of cancer survivors had one or more comorbidity, 67% had two or more comorbidities, 42% had three or more comorbidities, and approximately 20% had four or more comorbidities. The prevalence of comorbidities differed by number of comorbidities and types of cancer. Cancer survivors with a history of prostate and breast cancer had higher prevalence of comorbidities (14.8%, p = .0007 and 32.75%, p = .07, respectively) (see supplemental tables).

Table 1.

Demographic Characteristics of Older Cancer Survivors and Those Without Cancer.

Characteristics	Cancer survivors (weighted: 19.3%)	People without cancer(weighted: 80.6%)	p value
Age (M, SE)	73.7 (0.2)	73.6 (0.1)
65-74	59.2	60.4	.52
75-84	29.2	28.0
≥85	11.5	11.6
Gender
Male	41.7	44.3	.20
Female	58.3	55.7	.20
Race/ethnicity
Caucasian	86.7	86.6	.89
African American	7.6	8.0
Other	5.6	5.3
Income (per year)
<US$15,000	20.6	20.8	.98
US$15,000-US$30,000	31.0	30.9
≥US$30,000	48.3	48.3
Multiple chronic conditions (≥2)
Yes	67.4	68.231.8	.65
No	32.5	31.8	.65
Comorbidities
Heart disease	41.8	40.5	.56
Stroke	9.2	9.8	.40
Hypertension	66.6	65.7	.60
Arthritis	57.0	56.1	.56
COPD	18.9	16.0	.01
Paralysis/amputation	3.2	3.4	.80
Bone disease	24.1	21.1	.01
Diabetes	8.9	9.0	.81
Neurological disease	11.3	10.7	.46
Psychological disease	4.5	4.3	.76
Site of cancer
Lung	4.5	N/A
Breast	28.6	N/A
Prostate	21.3	N/A
Colon	10.8	N/A
Kidney	3.0	N/A
Physical limitations (≥2)	21.7	20.0	.16
ADL limitations (≥2)	13.6	12.1	.39
IADL limitations (≥2)	8.9	7.0	.01

Note. Unweighted N = 23,695; Weighed N = 110,086,170. Cancer survivors: Unweighted N = 4,370; Weighed N = 21,310,022. People without cancer: Unweighted N = 19,325; Weighed N = 88,776,148. COPD = chronic obstructive pulmonary disease; ADL = activities of daily living; IADL = instrumental activities of daily living.

Although cancer survivors had slightly higher prevalence of physical and functional limitations at baseline than those without cancer (21.7% vs. 20% for physical limitations, 13.6% vs. 12.1% for ADL, and 8.9% vs. 7.0% for IADL) after weighting (Table 1), these differences were not statistically significant except for IADL (p = .01). Accordingly, when compared with people without cancer, cancer survivors were 20% more likely to have physical limitations (adjusted odds ratio [aOR] = 1.20, 95% confidence interval [CI]: [0.93, 1.54]) and IADL limitations (aOR = 1.20, CI: [0.81, 1.75]), but these were not statistically significant after adjusting for patient’s sociodemographic characteristics and comorbidities (Table 2). On the other hand, older age was significantly associated with higher prevalence of physical limitations (aOR = 1.05, 95% CI: [1.04, 1.06], p < .0001) and functional limitations—ADL (aOR = 1.06, 95% CI: [1.05, 1.07], p < .0001) and IADL (aOR = 1.05, 95% CI: [1.04, 1.07], p < .0001). In addition, participants with higher income and females were less likely to have physical and functional limitations (ADL & IADL). Participants with two or more comorbidities were almost 5 times more likely to have physical limitations (aOR = 4.9, 95% CI: [4.15, 5.76], p < .0001), and almost 3 to 4 times more likely to have functional limitations—ADL (aOR = 4.2, 95% CI: [3.47, 5.06], p < .0001) and IADL (aOR = 3.02, 95% CI: [2.29, 3.94], p < .0001)—compared with those without it. When stratified by types of cancer, lung cancer survivors were more likely to have physical and functional limitations (ADL & IADL) and those with history of renal cancer were almost twice as likely to have ADL (aOR = 1.90, 95% CI: [0.90, 4.02]), although these were not statistically significant.

Table 2.

Adjusted Odds Ratios for the Risk of Having Physical and Functional Limitations After Propensity Score Weighting.

Determinants	Adjusted odds ratio (95% confidence interval)
	Physical limitations	Functional limitations
	Physical limitations	ADL	IADL
Cancer survivors vs. Non-cancer	1.20 [0.93, 1.54]	1.1 [0.85, 1.38]	1.20 [0.81, 1.75]
Age (per year)	1.05 [1.04, 1.06]**	1.06 [1.05, 1.07] **	1.05 [1.04, 1.07]**
Female vs. male	0.73 [0.65, 0.83]**	0.89 [0.79, 1.00]	0.86 [0.73, 1.00]
African American vs. Caucasian	1.13 [0.86, 1.49]	1.07 [0.87, 1.32]	1.32 [1.05, 1.66]
Income
ref:
US$15,000-US$30,000	0.58 [0.51, 0.67]**	0.64 [0.54, 0.74]**	0.64 [0.53, 0.79]**
≥US$30,000	0.33 [0.30, 0.38]**	0.42 [0.36, 0.49]**	0.39 [0.32, 0.47]**
Comorbidity [≥2 vs. <2]	4.9 [4.15, 5.76]**	4.2 [3.47, 5.06]**	3.0 [2.29, 3.94]**
Type of cancer (compared to non-cancer)
Lung	1.38 [0.81, 2.34]	1.68 [0.98, 2.90]	1.54 [0.77, 3.07]
Breast	0.88 [0.63, 1.23]	0.89 [0.61, 1.29]	1.12 [0.68, 1.85]
Prostate	0.68 [0.46, 1.00]*	0.77 [0.52, 1.14]	0.96 [0.56, 1.66]
Colon	1.11 [0.76, 1.62]	1.06 [0.69, 1.63]	1.33 [0.79, 2.23]
Kidney	0.88 [0.42, 1.88]	1.90 [0.90, 4.02]	1.03 [0.43, 2.49]

Note. Models were adjusted for cancer status, age, race, sex, income, comorbidities and type of cancer. CI = confidence interval; ADL = activities of daily living; IADL = instrumental activities of daily living.

p < .05. **p < .0001.

Table 3 shows the health care utilization of participants by cancer status and the presence of physical and functional limitations (ADL & IADL). Among those with physical limitations, older cancer survivors had slightly lower rates of hospitalizations, but had much higher rates of rehospitalizations (49.6% vs. 44.3%), emergency department (ED) visits (21.8% vs. 17.0%), and frequent ED visits within 30 days of a prior ED visit (13.9% vs. 8.2%) compared with those without cancer. Similar results were found among cancer survivors with functional limitations (ADL & IADL). The average health care costs were also higher among cancer survivors compared to individuals with no history of cancer.

Table 3.

Health care Utilization of Older Cancer Survivors by Physical and Functional Limitations After Propensity Score Weighting.

Health-related outcomes	Having limitations		p value	Not having limitations		p value
Health-related outcomes	Cancer survivors (%)	Non-cancer patients (%)	p value	Cancer survivors (%)	Non-cancer patients (%)	p value
Physical limitations
Rate of hospitalizations	29.2	32.7	.16	13.8	11.8	.03
Rate of rehospitalization	49.6	44.3	.11	34.1	34.8	.85
Rate of hospital readmission within 30 days	13.2	14.5	.36	11.7	13.8	.11
Rate of ED visits	21.8	17.0	.01	8.7	8.2	.54
≥2 ED visits	27.9	23.6	.23	24.0	22.1	.55
ED visit within 30 days of prior ED visit	13.9	8.2	.39	11.0	9.4	.43
Average Medicare cost (US$) per hospitalization (M, SD)	22,971 (2,478.4)	20,068 (854.2)	—	15,600 (1,479.7)	17,173 (655.2)	—
ADL
Rate of hospitalizations	32.1	33.6	.66	14.8	13.4	.09
Rate of rehospitalization	50.5	45.2	.19	36.7	36.4	.94
Rate of hospital readmissions within 30 days	14.9	14.6	.83	11.3	13.9	.02
Rate of ED visits	25.8	17.4	.001	9.3	8.9	.58
≥2 ED visits	27.2	23.1	.40	25.0	22.5	.32
ED visit within 30 days of prior ED visit	9.8	8.0	.62	13.2	9.3	.02
Average Medicare cost (US$) per hospitalization (M, SD)	23,711 (2,689.2)	20,982 (1,170.1)	—	16,492 (1,462.5)	17,413 (560.1)	—
IADL
Rate of hospitalizations	30.3	34.9	.33	15.9	14.6	.12
Rate of rehospitalization	45.3	47.7	.61	40.0	37.2	.31
Rate of hospital readmissions within 30 days	12.6	15.1	.26	12.3	13.8	.13
Rate of ED visits	24.2	20.1	.28	10.3	9.2	.13
≥2 ED visits	21.6	26.4	.30	26.5	22.0	.11
ED visit within 30 days of prior ED visit	5.0	8.1	.29	13.7	9.1	.01
Average Medicare cost (US$) per hospitalization (M, SD)	24,916 (3,877.1)	22,699 (1,533.4)	—	17,101 (1,326.5)	17,579 (537.5)	—

Note. ED = emergency department; M = Mean; SD = Standard Deviation; ADL = activities of daily living; IADL = instrumental activities of daily living.

As hypothesized, cancer survivors with physical and functional limitations had an increased health care utilization compared with those without cancer and without limitations (Table 3). For example, among cancer survivors, the rate of hospitalization for physical limitations was 29.2% whereas for those without physical limitations was 13.8%, the rate of rehospitalization was 49.6% versus 34.1%; and the rate of ED visits was 21.8% versus 8.7% among cancer survivors with and without physical limitations. However, among both cancer survivors and non-cancer survivors, there was no statistically significant difference in the rates of readmission within 30 days of a prior hospitalization, having more than two ED visits, or ED visit within 30 days of a prior ED on comparing groups with and without physical or functional limitations.

Table 4 provides adjusted odds ratios for the above comparisons. For example, among those with physical limitations, cancer survivors were 32% more likely to be rehospitalized (aOR = 1.32, 95% CI: [0.96, 1.82], p > .05) and 72% more likely to visit the ED (aOR = 1.72, 95% CI: [1.26, 2.35], p < .05). Similarly, cancer survivors with ADL were almost 3 times more likely to have frequent ED visits (aOR: 2.68, 95% CI: [1.86, 3.86], p < .001). Similar findings were noted among those with IADL.

Table 4.

Adjusted Odds Ratio for Healthcare Utilization by Cancer Status and by Physical and Functional Limitations Status.

Determinants	Adjusted OR for cancer survivors vs. non-cancer patients (95% CI)
	Physical limitations	Functional limitations
	Physical limitations	ADL	IADL
Having limitations
Rate of hospitalizations	0.81 [0.58, 1.11]	0.72 [0.48, 1.09]	0.59 [0.35, 0.98]
Rate of rehospitalization	1.32 [0.96, 1.82]*	1.35 [0.92, 1.96]	1.08 [0.58, 2.00]
Rate of hospital readmissions within 30 days	1.00 [0.77, 1.30]	1.05 [0.77, 1.45]	0.67 [0.36, 1.25]
Rate of ED visits	1.72 [1.26, 2.35]*	2.68 [1.86, 3.86]**	2.29 [1.27, 4.13]*
≥2 ED visits	1.09 [0.71, 1.67]	1.10 [0.66, 1.83]	0.98 [0.51, 1.89]
ED visit within 30 days of prior ED visit	1.43 [0.74, 2.74]	0.93 [0.37, 2.30]	1.36 [0.47, 3.89]
Having no limitations
Rate of hospitalizations	1.19 [0.89, 1.60]	1.19 [0.92, 1.54]	1.16 [0.92, 1.46]
Rate of rehospitalization	1.23 [0.84, 1.82]	0.91 [0.64, 1.28]	0.98 [0.73, 1.33]
Rate of hospital readmissions within 30 days	1.07 [0.75, 1.54]	0.95 [0.68, 1.33]	1.01 [0.77, 1.32]
Rate of ED visits	1.32 [0.99, 1.74]*	1.21 [0.94, 1.56]	1.35 [1.10, 1.65]*
≥2 ED visits	0.97 [0.56, 1.70]	1.04 [0.67, 1.63]	1.08 [0.73, 1.61]
ED visit within 30 days of prior ED visit	1.40 [0.73, 2.67]	1.72 [1.01, 2.94]*	1.55 [0.94, 2.55]

Note. Models were adjusted for cancer status, age, race, sex, income, comorbidities, and type of cancer. OR = odds ratio; CI = confidence interval; ADL = activities of daily living; IADL = instrumental activities of daily living; ED = emergency department.

p < .05. **p < .0001.

Discussion

There was no significant difference in the prevalence of physical and functional limitations between older cancer survivors and those without cancer after weighted by propensity score of cancer status, and also after adjusting for age, sociodemographic characteristics, and comorbidities. There was also no difference in the rates of hospitalization between them. However, cancer survivors were more likely to be rehospitalized and visit emergency departments than those without cancer, particularly if they had either physical limitations or functional limitations (ADL & IADL). Physical and functional limitations exert a considerate burden on cancer survivors due to decreased ability to perform their daily tasks of living (León-Muñoz et al., 2007).

Our results are consistent with previous studies that have examined the factors associated with health care utilization and health care costs in older adults with functional limitations (Fernández-Olano et al., 2006; Heinrich et al., 2008). After carefully considering the propensity of having a cancer diagnosis, our study demonstrated that irrespective of cancer status, physical and functional limitations led to increased health care utilization among older individuals. Bhattacharjee et al. also demonstrated that when compared with functional independence, functional disability in older individuals was associated with increased health care utilization (Bhattacharjee, Gharabei, Kamal, & Riaz, 2017).

In our study, cancer survivors with physical and functional limitations had higher utilization of health care services and required more rehospitalizations, ED visits, and ED visits within 30 days of a prior ED visit. Cancer survivors also had more comorbidities when compared with participants without history of cancer, which may result in increased visits to primary care physicians and specialists. The increased health care utilization can be reduced if these individuals received better coordinated care and counseling on preventive measures of physical and functional limitations during their regular physician visits (Chavan, Kedia, & Yu, 2017; Kedia, Chavan, Boop, & Yu, 2017). The 2013 Institute of Medicine (IOM) report emphasizes the absence of guidance for best practices in cancer survivorship care which causes wide variations in care for these survivors and also focuses on the fragmented health care system (Jacobs & Shulman, 2017; Lee & Johnson, 2013; Levit et al., 2013). A multidisciplinary care infrastructure along with a cross-cutting model of care applicable to all practice settings is required for high-quality cancer survivorship care (Mayer, Nasso, & Earp, 2017; McCabe et al., 2013). The survivorship shared care model in concordance with the Institute of Medicine standards should be implemented by advanced practitioners to provide better survivorship care to older cancer survivors (Thom et al., 2019). Shared care models with unified and holistic clinical approach may help curb the excess health care utilization among cancer survivors (Levit et al., 2013).

Our study also found that the average Medicare cost per hospitalization was higher among older cancer survivors with physical or functional limitations compared with those without these limitations. The burden of health care expenditure was also found to be high among non-elderly cancer survivors with high out of pocket expenses (Guy et al., 2015). A recent year study by Choi and DiNitto (2018) showed that older adults with functional limitations were more worried about increased health care expenditure due to higher ED visits, among other factors. Similar trends of health care utilization were reported in other studies as well (Becker, Kang, & Stuifbergen, 2012; Brown, Riley, Schussler, & Etzioni, 2002; Brown & Yabroff, 2006; Lang et al., 2009). Physical and functional limitations among older cancer survivors led to significantly higher rates of frequent ED visits, compared with those without cancer.

In addition, the higher use of health care among older cancer survivors may hide the variation of cost by cancer stage. For example, some older patients with early stage cancer may have less health care expenditure, whereas patients with late stage of cancer diagnosis may incur significantly higher cost of care due to expensive chemotherapy and radiation therapy. Cost related to care is a major factor in assessing health care expense among these survivors as many of them forgo medical care due to unaffordable cost of care (Kent et al., 2013). However, our administrative data were not able to discern this information.

Our study has demonstrated a stronger impact of physical and functional limitations on the health care utilization of older cancer survivors, supporting its public health relevance. The results from our study can be used to tailor specialized plans for cancer survivors and address their health care needs post-cancer diagnosis and treatment. Health care providers need to understand the unique challenges faced by older cancer survivors in their daily life and develop a better cancer survivorship plan which can deliver optimal quality care to older cancer survivors. Early identification of rehabilitation goals on multidisciplinary level and effective planning for palliative care can help improve the functional status of these survivors. As mentioned earlier, a well-integrated shared care model and multidisciplinary care provider system needs to be in place for more effective and efficient management of cancer survivors (Bazzell, Spurlock, & McBride, 2015). Cost conscious and clinically effective care for these patients may enhance their overall health status and reduce the burden on the health care system.

Strengths and Limitations

One of the main strengths of our study is that MCBS is a national survey representative of the entire older Medicare beneficiary population in the United States. A complete profile of physical and functional limitations experienced by older Medicare population and cancer survivors was obtained through the MCBS data for this study. Another strength of our study is that it is a longitudinal study in which we used follow-up data to determine the impact of functional limitations on health care utilization during the entire study period. Most importantly, we analyzed the data based on the causal inference theory which strengthens our interpretation. We used the propensity score method to account for key confounders and ensure the comparability between cancer survivors and those without any history of cancer, thus avoiding overestimation or underestimation of the differences in health care utilizations between cancer survivors and those without cancer.

The study also has a few limitations. The MCBS questionnaires only contain self-reported information which may lead to recall bias. For those who were diagnosed with cancer more than 1 year prior to the survey, there is no information about the age of cancer diagnosis; therefore, we were not able to identify the length of cancer diagnosis for these patients. There was also no information on the stages of cancer and severity of diseases; therefore, variation in health care cost by different stages of cancer could not be assessed. In addition, we did not include MCBS participants <65 years of age in our analysis, as their eligibility of Medicare coverage is mainly due to disability or end-stage renal disease. Therefore, our results are not generalizable to all cancer survivors. We dichotomized the physical and functional limitations because the distribution of the original score was skewed to the left which may reduce the granularity of the measure.

Furthermore, the health care utilization of older individuals may differ based on the number of comorbidities, history of cancer diagnosis, and type of treatment received for it. In our study, we adjusted for these variables but did not conduct separate analyses by these factors because they may mask the variations in health care utilization among older adults. In addition, propensity score analysis does not account for unmeasured confounders, so the estimates may still be biased. Also, due to the release cycle of MCBS data, the initial data presented in this study are 8 years old. In recent times, improved care plans may be available for these survivors to mitigate some of the issues. However, there is a need to identify and manage them in a more coordinated manner, which, in turn, will reduce the burden on health care in the long run.

Conclusion

As the older adult population continues to grow, we need to devise better care plans for these individuals to avoid excess health care expenditure, especially among the older cancer survivors, which is a rapidly growing population in the United States. As higher prevalence of physical and functional limitations can lead to increased utilization of health care services, coordinated care among primary care physicians, geriatricians, oncologists, and other specialists will be the key to improve the standard of health care among the cancer survivor population. Attending primary care physicians can play a critical role in incorporating assessment of physical and functional status and provide advice on preventive measures to older cancer survivors to improve their overall health and reduce costs associated with excessive health care utilizations.

Supplemental Material

Supplement_material – Supplemental Material for Impact of Physical and Functional Limitations on Health Care Utilization in Older Cancer Survivors: A Medicare Current Beneficiary Survey

Supplemental Material, Supplement_material for Impact of Physical and Functional Limitations on Health Care Utilization in Older Cancer Survivors: A Medicare Current Beneficiary Survey by Prachi P. Chavan, Satish K. Kedia and Xinhua Yu in Journal of Aging and Health

Footnotes

Acknowledgements

We sincerely appreciate the three anonymous reviewers and the editor for their insightful comments that improved the article significantly.

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

ORCID iD

Prachi P. Chavan

Supplemental Material

Supplemental material for this article is available online.

References

Bazzell

Spurlock

McBride

(2015). Matching the unmet needs of cancer survivors to resources using a shared care model. Journal of Cancer Education, 30, 312-318. doi:10.1007/s13187-014-0708-9

Becker

Kang

S. J.

Stuifbergen

(2012). Predictors of quality of life for long-term cancer survivors with preexisting disabling conditions. Oncology Nursing Forum, 39, E122-E131. doi:10.1188/12.ONF.E122-E131

Bernard

D. S.

Farr

S. L.

Fang

(2011). National estimates of out-of-pocket health care expenditure burdens among nonelderly adults with cancer: 2001 to 2008. Journal of Clinical Oncology, 29, 2821-2826.

Bhattacharjee

Gharabei

Kamal

Riaz

(2017). Influence of health and functional status and co-occurring chronic conditions on healthcare expenditures among community-dwelling adults with kidney cancer in the United States: A propensity-score-matched analysis. Clinical Genitourinary Cancer, 15, e357-e368. doi:10.1016/j.clgc.2016.12.017

Brown

M. L.

Riley

G. F.

Schussler

Etzioni

(2002). Estimating health care costs related to cancer treatment from SEER-Medicare data. Medical Care, 40(8 Suppl.), IV-104. doi:10.1097/01.MLR.0000020939.20666.47

Brown

M. L.

Yabroff

. (2006). Economic impact of cancer in the United States. In Schottenfeld

Fraumeni

JF.

(Eds.), Cancer Epidemiology and Prevention. 3rd ed., p. 202. Oxford University Press.

Chavan

Kedia

(2017). Physical and functional limitations in US older cancer survivors. Journal of Palliative Care & Medicine, 7(4). doi:10.4172/2165-7386.1000312

Choi

DiNitto

(2018). Correlates of worry about health care costs among older adults. Journal of Applied Gerontology, 37(6), 763-782. doi:10.1177/0733464816650803

de Moor

J. S.

Mariotto

A. B.

Parry

Alfano

C. M.

Padgett

Kent

E. E.

Rowland

J. H

. (2013). Cancer survivors in the United States: Prevalence across the survivorship trajectory and implications for care. Cancer Epidemiology, Biomarkers & Prevention: A Publication of the American Association for Cancer Research, Cosponsored by the American Society of Preventive Oncology, 22, 561-570. doi:10.1158/1055-9965.EPI-12-1356

10.

DeSantis

Siegel

Jemal

(2015). Cancer treatment & survivorship—Facts & figures by American Cancer Society. Retrieved from http://www.cancer.org/acs/groups/content/@research/documents/document/acspc-042801.pdf

11.

DuGoff

E. H.

Schuler

Stuart

E. A.

(2014). Generalizing observational study results: Applying propensity score methods to complex surveys. Health Services Research, 49, 284-303.

12.

Fernández-Olano

Hidalgo

J. L.

Cerdá-Díaz

Requena-Gallego

Sánchez-Castaño

Urbistondo-Cascales

Otero-Puime

(2006). Factors associated with health care utilization by the elderly in a public health care system. Health Policy, 75, 131-139.

13.

Field

M. J.

Jette

(2007). The future of disability in America. Washington, DC: National Academies Press.

14.

Fried

T. R.

Bradley

E. H.

Williams

C. S.

Tinetti

M. E.

(2001). Functional disability and health care expenditures for older persons. Archives of Internal Medicine, 161, 2602-2607.

15.

Grov

E. K.

Fosså

S. D.

Dahl

A. A.

(2010). Activity of daily living problems in older cancer survivors: A population-based controlled study. Health & Social Care in the Community, 18, 396-406.

16.

Guo

Fraser

M. W.

(2015). Advanced Quantitative techniques in Social Sciences Series: Propensity score analysis: Statistical methods and applications (2nd ed.). Thousand Oaks, CA: SAGE.

17.

Guy

G. P.

Jr. Ekwueme

D. U.

Yabroff

K. R.

Dowling

E. C.

Rodriguez

J. L.

Virgo

K. S.

(2013). Economic burden of cancer survivorship among adults in the United States. Journal of Clinical Oncology, 31, 3749-3757.

18.

Guy

G. P.

Jr. Yabroff

Ekwueme

Virgo

Han

Banegas

. . . Geiger

(2015). Healthcare expenditure burden among non-elderly cancer survivors, 2008-2012. American Journal of Preventive Medicine, 49, S489-S497. doi:10.1016/j.amepre.2015.09.002

19.

Hardy

S. E.

McGurl

D. J.

Studenski

S. A.

Degenholtz

H. B.

(2010). Biopsychosocial characteristics of community-dwelling older adults with limited ability to walk one-quarter of a mile. Journal of the American Geriatrics Society, 58, 539-544.

20.

Heinrich

Luppa

Matschinger

Angermeyer

M. C.

Riedel-Heller

S. G.

König

(2008). Service utilization and health-care costs in the advanced elderly. Value in Health, 11, 611-620.

21.

Heins

Schellevis

Rijken

van der Hoek

Korevaar

(2012). Determinants of increased primary health care use in cancer survivors. Journal of Clinical Oncology, 30, 4155-4160.

22.

Hewitt

Rowland

Yancik

(2003). Cancer survivors in the United States: Age, health and disability. The Journals of Gerontology, Series A: Biological Sciences & Medical Sciences, 58, M82-M91. doi:10.1093/gerona/58.1.M82

23.

Jacobs

L. A.

Shulman

L. N.

(2017). Follow-up care of cancer survivors: Challenges and solutions. The Lancet Oncology, 18, e19-e29.

24.

Johnson

R. S.

Fallon

E. A.

Berg

C. J.

(2019). Correlates of light physical activity among cancer survivors. Psycho-Oncology, 28, 726-734. doi:10.1002/pon.5008

25.

Karve

Lorenzo

Liepa

A. M.

Hess

L. M.

Kaye

J. A.

Calingaert

(2015). Treatment patterns, costs, and survival among Medicare-enrolled elderly patients diagnosed with advanced stage gastric cancer: Analysis of a linked population-based cancer registry and administrative claims database. Journal of Gastric Cancer, 15, 87-104.

26.

Katz

(1983). Assessing self-maintenance: Activities of daily living, mobility, and instrumental activities of daily living. Journal of the American Geriatrics Society, 31, 721-727.

27.

Kedia

Chavan

Boop

(2017). Health care utilization among Elderly Medicare Beneficiaries with Coexisting Dementia and Cancer. Gerontology & Geriatric Medicine, 3, 1-9. doi:10.1177/2333721416689042

28.

Kent

Forsythe

Yabroff

Weaver

de Moor

Rodriguez

Rowland

(2013). Are survivors who report cancer-related financial problems more likely to forgo or delay medical care? Cancer, 119, 3710-3717. doi:10.1002/cncr.28262

29.

Kurtz

M. E.

Kurtz

Given

C. W.

Given

B. A.

(2005). Utilization of services among elderly cancer patients—Relationship to age, symptoms, physical functioning, comorbidity, and survival status. Ethnicity & Disease, 15(Suppl. 2), S2-S17.

30.

Lang

Lines

Lee

Korn

Earle

Menzin

(2009). Trends in healthcare utilization among older Americans with colorectal cancer: A retrospective database analysis. BMC Health Services Research, 9, Article 227. doi:10.1186/1472-6963-9-227

31.

Lawton

Brody

E. M.

(1970). Assessment of older people: Self-maintaining and instrumental activities of daily living. Nursing Research, 19, 278.

32.

Lee

Johnson

(2013). “Fragmented” cancer care system is “chaotic, costly,” IOM report says [Abstract]. Modern Healthcare, 43(37), 8-9.

33.

León-Muñoz

L. M.

López-García

Graciani

Guallar-Castillón

Banegas

J. R.

Rodríguez-Artalejo

(2007). Functional status and use of health care services: Longitudinal study on the older adult population in Spain. Maturitas, 58, 377-386.

34.

Levit

Balogh

Nass

Ganz

(Eds.). (2013). Institute of medicine (IOM) delivering high-quality cancer care: Charting a new course for a system in crisis. Washington, DC: National Academies Press.

35.

Manzano

J. M.

Luo

Elting

L. S.

George

Suarez-Almazor

M. E.

(2014). Patterns and predictors of unplanned hospitalization in a population-based cohort of elderly patients with GI cancer. Journal of Clinical Oncology, 32, 3527-3533.

36.

Mayer

Nasso

Earp

(2017). Defining cancer survivors, their needs, and perspectives on survivorship health care in the USA. The Lancet Oncology, 18, e11-e18. doi:10.1016/S1470-2045(16)30573-3

37.

McCabe

Bhatia

Oeffinger

Reaman

Tyne

Wollins

Hudson

(2013). American society of clinical oncology statement: Achieving high-quality cancer survivorship care. Journal of Clinical Oncology, 31, 631-640. doi:10.1200/JCO.2012.46.6854

38.

Medicare Current Beneficiary Survey. (2016). Retrieved from https://www.cms.gov/Research-Statistics-Data-and-Systems/Files-for-Order/LimitedDataSets/MCBS.html

39.

National Cancer Institute. (2015). Retrieved from https://www.cancer.gov/publications/dictionaries/cancer-terms?cdrid=450125

40.

National Cancer Institute. (2018, April). Cancer statistics. Retrieved from https://www.cancer.gov/about-cancer/understanding/statistics

41.

Pope

A. M.

Tarlov

A. R.

(1991). Disability in America: Toward a national agenda for prevention. National Academies Press.

42.

Romano

P. S.

Roos

L. L.

Jollis

J. G.

(1993). Presentation adapting a clinical comorbidity index for use with ICD-9-CM administrative data: Differing perspectives. Journal of Clinical Epidemiology, 46, 1075-1079.

43.

Rubin

D. B.

(2004). On principles for modeling propensity scores in medical research. Pharmacoepidemiology & Drug Safety, 13, 855-857.

44.

Siegel

Miller

(2019). Cancer statistics 2019. CA: A Cancer Journal for Clinicians, 69, 7-34. doi:10.3322/caac.21551

45.

Stacey

F. G.

James

E. L.

Chapman

Courneya

K. S.

Lubans

D. R.

(2015). A systematic review and meta-analysis of social cognitive theory-based physical activity and/or nutrition behavior change interventions for cancer survivors. Journal of Cancer Survivorship, 9, 305-338. doi:10.1007/s11764-014-0413-z

46.

Stineman

M. G.

Streim

J. E.

Pan

Kurichi

J. E.

Schussler-Fiorenza Rose

S. M.

Xie

(2014). Activity limitation stages empirically derived for activities of daily living (ADL) and instrumental ADL in the U.S. adult community-dwelling Medicare population. PM & R: The Journal of Injury, Function, and Rehabilitation, 6, 976-87; quiz 987. doi:10.1016/j.pmrj.2014.05.001

47.

Swartz

Chapman

Doggett

(2017). Models of shared care. In Korc-Grodzicki

Tew

(Eds.), Handbook of geriatric oncology (pp. 287-299). New York, NY: Demos Medical.

48.

Sweeney

Schmitz

K. H.

Lazovich

Virnig

B. A.

Wallace

R. B.

Folsom

A. R.

(2006). Functional limitations in elderly female cancer survivors. Journal of the National Cancer Institute, 98, 521-529.

49.

Thom

Boekhout

Corcoran

Adsuar

Oeffinger

McCabe

(2019). Advanced practice providers and survivorship care: They can deliver. Journal of Oncology Practice / American Society of Clinical Oncology, 15, e230-e237. doi:10.1200/JOP.18.00359

50.

Van Roekel

Bours

Breedveld-Peters

Meijer

Kant

Van Den Brandt

Weijenberg

. (2015). Light physical activity is associated with quality of life after colorectal cancer. Medicine & Science in Sports & Exercise, 47, 2493-2503. doi:10.1249/mss.0000000000000698

51.

Weaver

Rowland

Bellizzi

Aziz

(2010). Forgoing medical care because of cost: Assessing disparities in healthcare access among cancer survivors living in the United States. Cancer, 116, 3493-3504. doi:10.1002/cncr.25209

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.25 MB