Abstract
Introduction:
Treatment burden is emerging as an important patient-centered outcome for older adults with cancer who concurrently manage geriatric conditions. Our objective was to evaluate the contribution of geriatric conditions to treatment burden in older adults with non-muscle invasive bladder cancer (NMIBC).
Methods:
We identified 73,395 Medicare beneficiaries age 66+ diagnosed with NMIBC (Stage <II) in SEER-Medicare (2001–2014). The primary outcome was treatment burden, defined as health system contact days in the year following NMIBC diagnosis. Explanatory variables were the following geriatric conditions: multimorbidity (>=2 chronic conditions), functional dependency, falls, depression, cognitive impairment, weight loss, and urinary incontinence. We used negative binomial regression to model the association between individual geriatric conditions and treatment burden while adjusting for covariates.
Results:
At baseline, 64% had multimorbidity and median 3 conditions (IQR 0–5). Prevalence of other geriatric conditions ranged from 5.9%–15.2%. Adjusted mean health system contact was 8.9 days (95% CI 8.6–9.2). Multimorbidity had the largest effect size (adjusted mean 11.8 contact days (95% CI 8.3–8.8)). Each additional chronic condition conferred a 13% increased odds of health system contact (adjusted IRR 1.132, 95% CI 1.129–1.135). Regardless of number of chronic conditions, rural patients consistently had more treatment burden than urban counterparts.
Discussion:
In this population-based cohort of older NMIBC patients, multimorbidity and rurality were strongly associated with treatment burden in the year following NMIBC diagnosis. These findings highlight the need for interventions that reduce treatment burden due to geriatric conditions among the growing population of older adults with cancer, particularly in rural areas.
Keywords: geriatric oncology, rural disparities, treatment burden, bladder cancer, patient-centered outcomes, multimorbidity
INTRODUCTION
Over 70 million older adults (65 years and older) will be diagnosed with cancer by 2030.1 Across the cancer continuum, older adults are different from their younger counterparts in two important ways.2–4 First, older adults with cancer have geriatric conditions such as functional dependency, multimorbidity, and cognitive impairment, which affect cancer treatment decision-making and healthcare utilization. Second, older adults with serious illnesses like cancer tend to value patient-centered outcomes such as quality of life and maintaining independence over standard oncologic outcomes such as recurrence and survival.5
Treatment burden is emerging as an important patient-centered outcome. Treatment burden encompasses the constant tension between “the work of being a patient” and the resources needed to accomplish that work. This tension has downstream effects on patient and caregiver well-being. Treatment burden was originally described in the context of multimorbidity but is increasingly relevant to older patients with cancer who often manage a substantive workload due to their geriatric conditions.6,7 One of the few studies that evaluated treatment burden in the cancer context found that on average older adults with Stage I lung cancer spent 44 days in contact with the health system, encountered 20 physicians, and received 12 new prescriptions in the year following diagnosis.8 However, that study was notably limited by its lack of consideration of geriatric conditions.
An American Society of Clinical Oncology (ASCO) guideline highlighted the importance of screening for baseline geriatric conditions prior to initiating cancer treatment.9 Assessment of baseline geriatric conditions can facilitate a personalized treatment plan that acknowledges existing treatment burden and avoids excessive patient workload. Defining the relationships between cancer, geriatric conditions, and their cumulative treatment burden is foundational to designing patient-centered care for the growing population of older adults with cancer.
The relationships between geriatric conditions and treatment burden are particularly relevant to bladder cancer, which is the sixth most common cancer in the U.S. and has the highest median age at diagnosis of all cancer sites (73 years). Three-quarters of bladder cancers are non-muscle-invasive (NMIBC) at diagnosis and have a low risk of death.10 Despite this, NMIBC is a burdensome chronic condition with high recurrence rates (30–70%)11 requiring frequent visits for invasive surveillance procedures, weekly bladder instillations, and ambulatory surgery.11 Our prior research demonstrates that patients with bladder cancer have a median of 8 co-existing chronic conditions and frequent multispecialty outpatient visits.12
Our aim was to evaluate the relative contribution of geriatric conditions to treatment burden, defined as health system contact rate in the year following diagnosis, in a population-based cohort of older adults with NMIBC. We hypothesized that geriatric conditions from the ASCO guideline (functional dependence, multimorbidity or the presence of two or more chronic conditions, mobility impairment, depression, dementia, weight loss, and urinary incontinence) would be associated with a higher health system contact rate in the year following NMIBC diagnosis.
PATIENTS AND METHODS
Data
This study utilized the Surveillance, Epidemiology, and End Results (SEER) cancer registry linked to Medicare claims. SEER is a cancer registry consortium and covers approximately 26% of the U.S. population.13 SEER data are compiled by the U.S. National Cancer Institute which works with the North American Association of Central Cancer Registries to compile and harmonize data from selected, representative state cancer registries. SEER has robust quality control measures for its registry data; however, some limitations exist including missing data regarding certain types of cancer therapies and follow up for patients who migrate from SEER regions. For adults age 65 and older diagnosed with cancer in a SEER region, cancer registry data is linked to Medicare claims for healthcare services. Medicare is government sponsored health insurance for older Americans (>=65 years). The SEER-Medicare files were used in compliance with a data-use agreement with the National Cancer Institute. This study was approved by the Geisinger Institutional Review Board.
Cohort
We identified patients age 66 years or older diagnosed with NMIBC (AJCC Stage <II) between January 1, 2001 and September 30, 2014. We selected age 66 in order to have one year of Medicare claims data to determine baseline geriatric conditions. We included patients continuously enrolled in Medicare Parts A and B for at least 12 months prior to and 12 months following NMIBC diagnosis. We excluded patients with missing stage, non-urothelial carcinoma histology, diagnosis made on autopsy or death certificate, those enrolled in Medicare managed care plans, and missing urban/rural or educational attainment or median household income. A total of 73,395 patients met inclusion criteria (CONSORT Diagram, Figure 1) and had complete data for analysis.
Figure 1:
CONSORT Diagram for Cohort Inclusion/Exclusion Criteria
Outcomes
The primary outcome was treatment burden defined as a count of overall health system contact days in the 12 months following NMIBC diagnosis. We totaled the number of dates with claims for outpatient care, emergency department visits, and hospitalization length of stay. We counted multiple claims in one day as a single day of health system contact. For patients who underwent radical cystectomy (n=1150) or had a second cancer diagnosed following NMIBC (n=5374), we calculated healthcare utilization until the event date. We excluded follow up time after these events as they marked a transition to distinct healthcare utilization different from that reasonably attributed to NMIBC.8 For patients who died within the 12 months of follow-up (n=6883), we calculated health system contact through the death date. To minimize the influence of outliers, those with counts beyond the 99th percentile were excluded from analysis (n=716).
The secondary outcome was NMIBC-specific treatment burden defined as a count of days in contact with the health system specific to bladder cancer management. From the larger set of claims identified for the primary outcome, we created a subset with NMIBC-specific diagnosis and procedure codes (Supplemental Table 1). We only counted procedure claims with an associated ICD-9 diagnosis code for bladder cancer.
Explanatory Variables
The primary explanatory variables were the following geriatric conditions selected from the described ASCO guideline: functional status, falls, depression, cognitive impairment, nutrition status, and multimorbidity.9 We included urinary incontinence due to known impact on the quality of life of older adults and direct relevance to NMIBC.14,15 Because of the limited availability of geriatric conditions from SEER-Medicare, we substituted validated claims-based proxies for certain geriatric conditions as follows: probability of functional dependency16, mobility limitations for falls17, dementia for cognitive impairment, and weight loss for nutrition status (Supplemental Table 2).
We established the prevalence of baseline geriatric conditions using claims from the 12 months prior to NMIBC diagnosis. Patients with an ICD-9 code for depression18,19, dementia20, weight loss17,21, or urinary incontinence22 in the 12 months prior to NMIBC diagnosis were defined as having the condition. Multimorbidity is defined as two or more chronic conditions based on the U.S. Department of Health and Human Services definition.23 We first assessed multimorbidity as a dichotomous variable (2 or more chronic conditions versus 0–1 conditions). In order to gain a deeper understanding of the nuances between multimorbidity and treatment burden, we subsequently evaluated the count of baseline chronic conditions. To identify individual chronic conditions, we applied Agency for Healthcare Research and Quality Clinical Classifications Software and Chronic Condition Indicator tools to ICD-9 codes using previously described methods.12,24 Prior cancers were included in the definition of multimorbidity and in the count of chronic conditions.
We defined functional dependency using a validated Medicare claims-based algorithm applied to ICD-9 and durable medical equipment codes to predict the probability of dependency in activities of daily living.16 We defined cutoffs of <5%, 5%–20%, and >20% probability based on other studies.25
Covariates
In multivariable analysis, we controlled for other characteristics including age, sex, race/ethnicity, marital status, urban-rural residence, median household income, educational attainment, geographic region, NMIBC stage and grade, and receipt of NMIBC treatment within 6 months. We treated missing/unknown NMIBC grade as an additional category in each AJCC stage. Median household income and educational attainment were defined using composite measures at the census tract level. Educational attainment was divided into tertiles based on census tract estimates of high school education.26 NMIBC staging was classified according to the American Joint Committee on Cancer (AJCC) staging schema, 8th Edition.27 NMIBC treatment was defined as a claim for transurethral resection of bladder tumor, intravesical therapy, or both with an associated ICD-9 diagnosis code for bladder cancer in the six months following diagnosis date (Supplemental Table 3).
Statistical Analysis
We described the prevalence of explanatory variables and covariates at baseline. We evaluated Poisson, negative binomial, zero-inflated Poisson, and zero-inflated negative binomial distributions for the primary outcome. Poisson and negative binomial distributions were compared based on the log likelihood ratio test, with the negative binomial distribution exhibiting superior fit. The negative binomial was then compared to both of the zero-inflated models using the Vuong test and Akaike Information Criterion; the negative binomial exhibited better fit in both cases. To build the multivariable model, we first evaluated the unadjusted associations of all explanatory variables and covariates of interest with the primary outcome (health system contact days in the year following NMIBC diagnosis) using negative binomial regression models that generate incident rate ratios (IRR). All of the negative binomial evaluations included time offsets to account for observation periods less than one year due to death, radical cystectomy, or diagnosis of a second cancer. Covariates with a p value less than 0.25 were retained for consideration in the multivariable model. In the final multivariable model, education was also removed for not meeting that p value threshold. To aid in model interpretation, adjusted mean outcomes were calculated for all variables included in the final multivariable model. We also conducted unadjusted and multivariable negative binomial regression for the secondary outcome of NMIBC-specific health system contact.
In order to justify our use of count of chronic conditions in lieu of multimorbidity, we conducted a test of linearity between the count of chronic conditions and treatment burden. We identified quintiles of the count of chronic conditions and plotted the natural log of treatment burden for each quintile and performed simple linear regression which yielded an adjusted R-squared 0.92, which constitutes strong evidence of linearity. We proceeded with a second multivariable negative binomial model that substituted the count of chronic conditions for multimorbidity.
All analyses were conducted in SAS version 9.4 (Cary, North Carolina, USA). Reporting of methods and results follows the STROBE guidelines for cohort studies.28 Statistical significance was defined as two-sided p-value <0.05.
RESULTS
At baseline, three-quarters of the cohort were male (76.3%), the majority were White individuals (89.3%), 15.8% lived in rural areas, and 41.8% had 0a low grade urothelial carcinoma at diagnosis; these characteristics are largely consistent with NMIBC demographics.29 Greater than two-thirds received treatment for bladder cancer within 6 months of diagnosis (68%). Approximately two-thirds of the cohort had baseline multimorbidity (64%), and 46.5% had a prior cancer diagnosis. The median number of chronic conditions at baseline was 3 (interquartile range 0–5). The prevalence of other geriatric conditions ranged from 5.9% for weight loss to 15.2% for mobility impairment (Table 1).
Table 1:
Baseline Characteristics of Cohort
| All (n= 73,395) | |
|---|---|
| Age, years (mean, SD) | 77.5 (7.1) |
| Female | 17423 (23.7%) |
| Other/Unknown | 628 (0.9%) |
| Unknown | 5254 (7.2%) |
| Rural | 11600 (15.8%) |
| 4th quartile ($63,000+) | 18687 (25.5%) |
| High Attainment | 65724 (89.6%) |
| West | 29072 (39.6%) |
| 2014 | 3864 (5.3%) |
| Yes | 34106 (46.5%) |
| 1 Unknown Grade | 1832 (2.5%) |
| No | 23480 (32%) |
| None | 23480 (32.0%) |
| High Probability: >20% | 7433 (10.1%) |
| Yes | 11165 (15.2%) |
| Yes | 6056 (8.2%) |
| Yes | 4809 (6.6%) |
| Yes | 4300 (5.9%) |
| Yes (2 or more chronic conditions) | 46985 (64%) |
| Count of chronic conditions, median (IQR) | 3.0 (0, 5) |
| Yes | 5698 (7.8%) |
In patient’s U.S. Census tract.
Abbreviations: SD=standard deviation; AJCC=American Joint Committee on Cancer; TURBT=transurethral resection of bladder tumor(s); IQR=interquartile range
In unadjusted comparisons (Table 2), health system contact rate was higher for Black individuals, rural residence, first quartile of census tract median household income, and Midwest residents. In the initial multivariable model, educational attainment was not significant, likely due to its correlation with median household income, and was therefore excluded from the final multivariable model. There was a monotonic relationship between increasing stage/grade at NMIBC diagnosis and IRR for health system contact. Health system contact rate was significantly higher among individuals with each geriatric condition relative to those without the conditions. Multimorbidity and a high probability of functional dependency (≥20%) had the largest associations with overall health system contact rate.
Table 2:
Unadjusted and Adjusted Associations of Overall Health System Contact Rate with Patient Factors
| Unadjusted IRR (95% Cl) | MODEL 1: MM Adjusted IRR (95% CI)b | p value | MODEL 2: CCC Adjusted IRR (95% CI)C | p value | |
|---|---|---|---|---|---|
| Age | 1.01 (1.01,1.01) | 1.002 (1.000,1.003) | <0.01 | 1.002 (1.001,1.003) | <0.01 |
| Sex | 027 | <0.01 | |||
| Male | Reference | Reference | Reference | ||
| Female | 1.02 (0.998,1.03) | 0.99 (0.97,1.01) | 1.04 (1.02,1.06) | ||
| Race/Ethnicity | <0.01 | <0.01 | |||
| White | Reference | Reference | Reference | ||
| Black | 1.12 (1.08,1.17) | 1.05(1.01,1.09) | 1.07 (1.03,1.11) | ||
| Hispanic | 0.99 (0.95,1.03) | 1.04 (1.01,1.08) | 1.03 (0.99,1.08) | ||
| Asian | 0.84 (0.80,0.88) | 0.85 (0.81,0.89) | 0.90 (0.86, 0.94) | ||
| Other/Unknown | 0.70 (0.65,0.76) | 0.76(0.71,0.82) | 0.75 (0.70, 0.81) | ||
| Married | <0.01 | <0.01 | |||
| Yes | Reference | Reference | Reference | ||
| No | 1.08 (1.06,1.09) | 1.02 (1.01,1.04) | 1.02 (1.01,1.04) | ||
| Unknown | 0.89 (0.86,0.91) | 0.91 (0.89,0.94) | 0.91 (0.88, 0.93) | ||
| Urban-Rural Residence | <0.01 | <0.01 | |||
| Urban | Reference | Reference | Reference | ||
| Rural | 1.32 (1.30,1.35) | 1.20 (1.18,1.23) | 1.19 (1.16,1.21) | ||
| Median Household Incomea | <0.01 | <0.01 | |||
| 4th quartile ($63,000+) | Reference | Reference | Reference | ||
| 1st quartile (<$38,000) | 121 (1.19,1.24) | 1.07 (1.05,1.10) | 1.07 (1.04,1.09) | ||
| 2nd quartile ($38,000–$47,999) | 1.10 (1.08,1.13) | 1.01 (0.99,1.04) | 1.02 (1.00,1.04) | ||
| 3rd quartile ($48,000–$62,999) | 1.03 (1.01,1.05) | 0.99, (0.97,1.01) | 0.99 (0.97,1.01) | ||
| Educational Attainment: tertilesa | |||||
| High Attainment | Reference | ||||
| Low Attainment | 0.97 (0.84,1.12) | N/A | N/A | N/A | N/A |
| Intermediate Attainment | 1.06 (1.03,1.08) | ||||
| U.S. Geographic Region | <0.01 | <0.01 | |||
| Northeast | Reference | Reference | Reference | ||
| South | 1.06 (1.04,1.09) | 0.95 (0.93,0.97) | 0.94 (0.92, 0.96) | ||
| Midwest | 1.46 (1.43,1.50) | 1.30 (126,1.32) | 1.30 (127,1.33) | ||
| West | 1.05 (1.03,1.07) | 1.01 (0.99,1.03) | 1.01 (0.99,1.03) | ||
| AJCC Stage/Grade | <0.01 | <0.01 | |||
| OA Low Grade | Reference | Reference | Reference | ||
| OA High Grade | 1.16 (1.13,1.18) | 1.11 (1.09,1.14) | 1.12 (1.10,1.14) | ||
| OA Unknown Grade | 1.10 (1.07,1.13) | 1.06 (1.03,1.08) | 1.04 (1.02,1.07) | ||
| Ois Low Grade | 1.01 (0.96,1.06) | 1.04 (0.99,1.08) | 1.04 (1.00,1.09) | ||
| Ois High Grade | 1.18 (1.12,1.24) | 1.16 (1.11,1.22) | 1.20 (1.14,1.26) | ||
| Ois Unknown Grade | 1.30 (124,1.35) | 1.22 (1.18,1.28) | 1.24 (120,1.29) | ||
| 1 Low Grade | 1.12 (1.09,1.15) | 1.12 (1.09,1.15) | 1.13 (1.10,1.16) | ||
| 1 High Grade | 1.37 (1.34,1.40) | 1.31 (129,1.34) | 1.35 (1.32,1.37) | ||
| 1 Unknown Grade | 1.37 (1.30,1.43) | 1.36 (1.30,1.42) | 1.34 (129,1.40) | ||
| Bladder Cancer Treatment Within 6 Months | <0.01 | <0.01 | |||
| No | Reference | Reference | Reference | ||
| Yes | 1.26 (1.24,1.28) | 1.24 (122,1.26) | 1.25 (123,1.27) | ||
| Multimorbidity | <0.01 | ||||
| No (0–1 chronic conditions) | Reference | Reference | N/A | N/A | |
| Yes (2 or more chronic conditions) | 2.02 (1.99,2.05) | 1.90 (1.88,1.93) | |||
| Count of Chronic Conditions | 1.142 (1.139,1.145) | N/A | N/A | 1.132 (1.129,1.135) | <0.01 |
| Functional Dependency | <0.01 | <0.01 | |||
| Low: <5% | Reference | Reference | Reference | ||
| Intermediate: 5% - <20% | 1.19 (1.17,121) | 1.13 (1.11,1.15) | 1.08 (1.06,1.10) | ||
| High: ≥20% | 1.41 (1.38,1.45) | 1.28 (124,1.32) | 1.18 (1.15,1.22) | ||
| Mobility Impairment | <0.01 | <0.01 | |||
| No | Reference | Reference | Reference | ||
| Yes | 1.18 (1.16,1.21) | 1.08 (1.06,1.10) | 1.07 (1.05,1.09) | ||
| Depression | <0.01 | 0.25 | |||
| No | Reference | Reference | Reference | ||
| Yes | 1.26 (1.23,1.29) | 1.09 (1.06,1.12) | 1.02 (0.99,1.04) | ||
| Dementia | 0.07 | 0.04 | |||
| No | Reference | Reference | Reference | ||
| Yes | 1.22 (1.18,125) | 0.97 (0.942,1.003) | 0.968 (0.938,0.999) | ||
| Weight Loss | <0.01 | <0.01 | |||
| No | Reference | Reference | Reference | ||
| Yes | 1.30 (126,1.35) | 1.17 (1.13,1.20) | 1.15 (1.12,1.18) | ||
| Urinary Incontinence | 0.01 | 0.01 | |||
| No | Reference | Reference | Reference | ||
| Yes | 1.14 (1.11,1.17) | 1.04 (1.01,1.06) | 1.04 (1.01,1.06) |
Abbreviations: Cl = confidence interval; AJCC = American Joint Committee on Cancer; N/A = Not applicable; IRR = Incidence Rate Ratio.
In patient’s U.S. Census tract.
Model 1 MM: negative binomial multivariable model with multimorbidity (MM) as a dichotomous variable.
Model 2 CCC: negative binomial multivariable model with count of chronic conditions (CCC) in lieu of multimorbidity.
In adjusted analysis (Table 2), rural residence, Midwest residence, bladder cancer treatment, and AJCC Stage 1 were associated with higher rate of health system contact. The geriatric condition of multimorbidity (two or more chronic conditions) had the largest effect size (adjusted IRR 1.90, 95% CI 1.88–1.93, Model 1 MM), consistent with known trends of high treatment burden among older adults with cancer and multimorbidity.12 Moreover in the analysis using count of chronic conditions, each additional chronic condition was associated with a 13% increased odds of overall health system contact (adjusted IRR 1.132, 95% CI 1.129–1.135, Model 2 CCC).
After adjustment, most other geriatric conditions were associated with increased rates of health system contact, including high probability of functional dependency (IRR 1.28, 95% CI 1.24–1.32), mobility impairment (IRR 1.08, 95% CI 1.06–1.10), depression (IRR 1.09, 95% CI 1.06–1.12), weight loss (IRR 1.17, 95% CI 1.13–1.20), and urinary incontinence (IRR 1.04, 95% CI 1.01–1.06). In the adjusted analysis with count of chronic conditions, dementia was associated with a lower rate of health system contact (Table 2, Model 2 CCC).
The adjusted mean days of health system contact for the cohort was 8.9 (95% CI 8.6–9.2). The largest effect size was seen for patients with multimorbidity who had an adjusted mean of nearly 12 days in contact with the health system in the year following diagnosis (Supplemental Table 4, Model 1 MM). As demonstrated in Figure 2, each additional chronic condition was associated with an increase in adjusted mean health system contact days.
Figure 2:
Adjusted Means of Days of Overall Health System Contact in the Year Following NMIBC Diagnosis by Count of Baseline Chronic Conditions
Based on the large effect sizes from rural residence and multimorbidity in the multivariable model of overall health system contact, we calculated adjusted mean days of health system contact for urban/rural status by count of chronic conditions. As demonstrated in Figure 3, for values of the count of chronic conditions ranging from 0 through 11, older NMIBC patients residing in rural areas had consistently higher treatment burden than their urban counterparts (p value for interaction <0.01).
Figure 3:
Adjusted Means of Days of Overall Health System Contact in the Year Following NMIBC Diagnosis by Count of Baseline Chronic Conditions and Stratified by Rural/Urban Status
In adjusted analyses of NMIBC-specific treatment burden, we found higher health system contact rates for women (IRR 1.13, 95% CI 1.09–1.17), Asians (IRR 1.22, 95% CI 1.11–1.33), rural residence (IRR 1.23, 95% CI 1.18–1.29), and Midwest residents (IRR, 1.40, 95% CI 1.33–1.47). Of all combined stages/grades, Stage 1 high grade bladder cancer exhibited the strongest association with the contact rate (adjusted IRR 1.87, 95% CI 1.80–1.94) which is consistent with current NMIBC guidelines that recommend intensive treatment and surveillance guidelines for Stage 1 disease (Supplemental Tables 5 and 6).11 Similar to the analysis of days of overall health system contact, multimorbidity (two or more chronic conditions) had the strongest association with rate of NMIBC-specific contact (adjusted IRR 1.60, 95% CI 1.55–1.64, Model 1 MM). Functional dependency, mobility impairment, dementia, and weight loss were associated with less NMIBC-specific contact. Depression and urinary incontinence were not significant.
DISCUSSION
In this population-based study of older adults with NMIBC, we found that geriatric conditions were associated with higher treatment burden in the year following cancer diagnosis. In particular, the geriatric condition of multimorbidity was significantly associated with higher rates of overall and NMIBC-specific treatment burden. Our study highlights disparities between rural and urban older adults with NMIBC, demonstrating that rural patients have higher rates of overall and NMIBC-specific contact than their urban counterparts. Though overall health system contact rates were higher for older adults with high functional dependency, the NMIBC-specific contact rates were lower.
By 2040, older adults will account for over three-quarters of all people living with a cancer diagnosis.7 Healthcare providers must consider the additive effects of cancer diagnosis and treatment on the work of being a patient with to geriatric conditions. Excessive health system contacts (e.g. office visits) require time and energy, contribute to high treatment burden, and may not align with the goals of this population. For example, in the American Time Use Survey, a single office visit was associated with 35 minutes of travel, 42 minutes of waiting, and 74 minutes receiving services.30 Fifty percent of older adults are accompanied to office visits by caregivers who spend 124 minutes in each encounter. In the current study, we found that geriatric conditions were associated with 6 to 12 days of contact in the year following NMIBC diagnosis, which translates into a heavy burden in terms of time and care coordination for older patients with NMIBC and their caregivers.
Of the geriatric conditions we analyzed, multimorbidity had one of the largest effects on treatment burden. Older adults with multimorbidity have high rates of healthcare utilization, polypharmacy, and fragmented care. Two-thirds of older patients with cancer have pre-existing chronic conditions and have higher rates of multimorbidity than patients without cancer.7,31 Older cancer patients tend to have conditions known to have high treatment burden such as heart failure, chronic obstructive pulmonary disease, and diabetes.7,32,33 The interaction between cancer and multimorbidity creates challenges for patients. A recent systematic review found that medically complex cancer patients feel overwhelmed when self-managing medications and are often unsure who is in charge of treating their other chronic conditions. Rather than a single disease approach to cancer, healthcare for older patients with cancer should address the cumulative burden from all their concurrent geriatric and chronic conditions.
We found lower rates of NMIBC-specific care in geriatric conditions other than multimorbidity, which suggests that urologists may tailor treatment and surveillance intensity for certain geriatric conditions, or functionally dependent older adults may find attending frequent appointments more difficult. These data expand on prior research showing undertreatment of NMIBC in older adults and diminishing returns for NMIBC surveillance. For example, we found underuse of intravesical therapy (approximately one-third). Prior SEER-Medicare analyses have also found that increasing age is associated underuse of intravesical therapy with bacillus Calmette-Guerin (BCG) and that nearly half of older adults with intermediate or high-risk NMIBC do not receive BCG.34,35 For NMIBC surveillance, a microsimulation model found that NMIBC surveillance at age 85 years was associated with <1 QALY gained in contrast to 2–7 QALYs at age 65.36
Reducing or avoiding low value care may be one way to reduce treatment burden in older patients with NMIBC and other early-stage cancers. For example, 0a low grade NMIBC, disproportionately affects older adults over 85 years, is slow-growing, and rarely metastasizes. A prospective observational study of active surveillance for 0a low grade NMIBC showed a reduction in transurethral surgery, an outpatient procedure that still requires general anesthesia.37 Recent studies have suggested a shared decision-making approach to reduce overtreatment or non-surgical treatment for older adults with basal cell carcinoma.38,39 More broadly, aligning healthcare with stated goals and preferences led to lower Treatment Burden Questionnaire scores in a cohort of older adults with multimorbidity in a primary care practice.40
Our findings highlight the impact of rurality on older adults with cancer. We found that older rural patients with NMIBC and multimorbidity had higher treatment burden in the year following cancer diagnosis than urban patients, even after considering healthcare that may have been stacked into a single day. These findings have significant implications for rural residents, who already face logistical challenges to healthcare. One-quarter of older Americans live in rural areas and have higher rates of multimorbidity than those in urban areas.41 Older rural patients with cancer travel long distances to seek healthcare, incurring worrisome out-of-pocket costs.42,43 This population relies on informal social networks for transportation for several reasons including lower overall rates of car ownership and declining driving rates after the age of 75.44 Given this, we were surprised to find that medically complex, rural NMIBC patients had more treatment burden, though we hypothesize that care fragmentation arising from management for non-cancer chronic conditions may contribute. We also found that Midwestern patients had the highest overall health system contact rate of all SEER regions. While a deeper examination of these findings was beyond the current scope, research is needed to understand the additive effects of rurality and regional variation on treatment burden among older adults facing a new cancer diagnosis.
Worsening rural disparities will impact older adults seeking bladder cancer treatment. Urologists conduct the majority of NMIBC diagnosis, treatment, and surveillance; however, urologic care in rural areas is becoming scarce. Two-thirds of U.S. counties do not have a urologist and only 2.4% of all U.S. urologists work in rural counties.45,46 Drive times to obtain urologic care are increasing. A recent study found that less than 13% of Medicare beneficiaries are able to reach a urologist’s office in 30 minutes.47
The current study has several notable strengths. To the best of our knowledge, this is one of the few studies to evaluate the association between geriatric conditions and treatment burden in older adults with cancer. We specifically selected geriatric conditions highlighted in a published ASCO guideline to establish the important role of geriatric assessment as a routine part of cancer care, even for a type of malignancy that is predominantly managed by surgical subspecialists. A second strength is that we utilized a large, nationally representative sample spanning more than a decade of NMIBC treatment. Our large sample size allowed us to have adequate power to assess multiple exposures and adjust for appropriate covariates.
The current study must also be interpreted within some limitations. Due to the nature of the SEER-Medicare dataset, we had to rely on claims-based definitions of geriatric conditions rather than true geriatric assessment. Wherever possible, we applied validated algorithms and consulted with experts to optimize accuracy. Despite this, there is still the possibility that certain conditions such as dementia, depression, and weight loss may be undercounted. Similarly, we would have liked to control for smoking status due to its association with bladder cancer and many common chronic conditions; however, this variable is not reliably available in the SEER-Medicare dataset.
We also note the narrow range of our outcome, i.e., approximately 6–12 days or the equivalent of a healthcare-related visit every 1–2 months. For older adults managing both cancer treatment and geriatric conditions, even a small number of health system contacts requires mobilizing transportation resources, spending additional funds while living on a fixed income, addressing administrative burdens related to care coordination, and managing mental and physical strains for patients and informal caregivers. These findings may encourage healthcare providers to screen for geriatric conditions that put older adults with cancer at higher risk for treatment burden, and then tailor interventions to reduce healthcare-related burdens. Nonetheless, further research is needed to also understand patient and caregiver perceptions of treatment burden, as not all types of healthcare encounters are equivalent. Our treatment burden range of 6–12 days is less than the 44 days of health system contact cited for lung cancer. This difference is expected due to the different types of treatment rendered for NMIBC versus early stage lung cancer which incorporates daily radiation therapy for multiple weeks.8
In this population-based study of older adults with NMIBC, we found that geriatric conditions were associated with higher treatment burden in the year following cancer diagnosis. In particular, multimorbidity and rurality were associated with the largest effects on overall and NMIBC-specific health system contact rate. Our findings highlight the importance of recognizing and managing multimorbidity across the cancer continuum in older adults. These results provide evidence for the need to design efficient healthcare delivery for older adults with cancer that accounts for the whole patient and reduces treatment burden, especially in rural areas where cancer care providers are increasingly scarce.
Several recent technological developments in telehealth may expand availability of urologic procedures to underserved areas and reduce travel needs for rural older adults with bladder cancer. A pilot study of a new tele-cystoscopy program demonstrated the feasibility and accuracy of office cystoscopy performed by trained advanced practitioners.48 Home-based care takes many forms in existing intervention studies ranging from helping patients to enhance self-management of chronic conditions to providing primary care longitudinally in the home.49,50 Most interventions have demonstrated mixed results, and further study is needed to determine the most impactful patient-centered outcomes, effective interventions to address those outcomes, and the implementation strategies to maximize benefit.51 Such interventions in addition to reducing or avoiding low value care, may help older adults with cancer successfully manage their healthcare while mitigating treatment burden.
Supplementary Material
ACKNOWLEDGEMENTS
This study used the linked SEER-Medicare database. The interpretation and reporting of these data are the sole responsibility of the authors. The authors acknowledge the efforts of the National Cancer Institute at the National Institutes of Health; the Office of Research, Development and Information, CMS; Information Management Services (IMS), Inc.; and the Surveillance, Epidemiology, and End Results (SEER) Program tumor registries in the creation of the SEER-Medicare database.
The authors acknowledge Dr. Jennifer Lund for guidance on claims-based frailty measures. We also thank Dr. Frances Lynch, Philip Crawford, and the Mental Health Research Network for guidance on claims-based depression measures. We are grateful to Kirstin Rabinowitz, MPH for administrative assistance over the course of the study.
FUNDING
This work was supported by the National Institute on Aging at the National Institutes of Health (grant number R03AG064382), Yale Claude D. Pepper Older Americans Independence Center (grant number P30AG021342), and the Duke Claude D. Pepper Older Americans Independence Center (grant number P30AG028716). The funders had no involvement in study design, analysis and interpretation of the data, writing of the report, and the decision to submit this article for publication.
ABBREVIATIONS
- AJCC
American Joint Committee on Cancer
- ASCO
American Society of Clinical Oncology
- HCPCS
Healthcare Common Procedure Coding System
- ICD
International Classification of Diseases
- NMIBC
Non-Muscle-Invasive Bladder Cancer
- SEER
Surveillance, Epidemiology, and End Results
Footnotes
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
CONFLICT OF INTEREST
Matthew E. Nielsen serves as a paid consultant to the American College of Physicians High Value Care Task Force and as a consultant/advisor to Grand Rounds for which he is paid via stock options. Tullika Garg served as a paid consultant to WebMD and reports an immediate family member is an employee of DRPLZ and is a stockholder. All other authors report no conflict of interest.
REFERENCES
- 1.Smith BD, Smith GL, Hurria A, Hortobagyi GN, Buchholz TA. Future of cancer incidence in the United States: Burdens upon an aging, changing nation. J Clin Oncol. 2009;27(17):2758–2765. doi: 10.1200/JCO.2008.20.8983 [DOI] [PubMed] [Google Scholar]
- 2.Yabroff KR, Guy GP, Banegas MP, Ekwueme DU. The economic burden of cancer in the United States. Schottenfeld Fraumeni Cancer Epidemiol Prev Fourth Ed. 2017;35(18):169–182. doi: 10.1093/oso/9780190238667.003.0010 [DOI] [Google Scholar]
- 3.Rim SH, Guy GP, Yabroff KR, McGraw KA, Ekwueme DU. The impact of chronic conditions on the economic burden of cancer survivorship: a systematic review. Expert Rev Pharmacoecon Outcomes Res. 2016;16(5):579–589. doi: 10.1080/14737167.2016.1239533 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.National Comprehensive Cancer Network. Older Adult Oncology (Version 2.2020). Published2020. AccessedOctober 5, 2020. https://www.nccn.org/professionals/physician_gls/pdf/senior.pdf
- 5.Fried TR, Bradley EH, Towle VR, Allore H. Understanding the treatment preferences of seriously ill patients. N Engl J Med. 2002;346(14):1061–1066. doi: 10.1056/NEJMsa012528 [DOI] [PubMed] [Google Scholar]
- 6.Sheehan OC, Leff B, Ritchie CS, et al. A systematic literature review of the assessment of treatment burden experienced by patients and their caregivers. BMC Geriatr. 2019;19(1):1–11. doi: 10.1186/s12877-019-1222-z [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Bluethmann SM, Mariotto AB, Rowland JH. Anticipating the “silver tsunami”: Prevalence trajectories and comorbidity burden among older cancer survivors in the United States. Cancer Epidemiol Biomarkers Prev. 2016;25(7):1029–1036. doi: 10.1158/1055-9965.EPI-16-0133 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Presley CJ, Soulos PR, Tinetti M, Montori VM, Yu JB, Gross CP. Treatment burden of medicare beneficiaries with stage i non-small-cell lung cancer. J Oncol Pract. 2017;13(2):e98–e107. doi: 10.1200/JOP.2016.014100 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Mohile SG, Dale W, Somerfield MR, et al. Practical assessment and management of vulnerabilities in older patients receiving chemotherapy: Asco guideline for geriatric oncology. J Clin Oncol. 2018;36(22):2326–2347. doi: 10.1200/JCO.2018.78.8687 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.American Cancer Society. Cancer Facts & Figures 2017. Published2017. AccessedOctober 5, 2020. https://www.cancer.org/research/cancer-facts-statistics/all-cancer-facts-figures/cancer-facts-figures-2017.html
- 11.Chang SS, Boorjian SA, Chou R, et al. Diagnosis and Treatment of Non-Muscle Invasive Bladder Cancer: AUA/SUO Guideline. J Urol. 2016;196(4):1021–1029. doi: 10.1016/j.juro.2016.06.049 [DOI] [PubMed] [Google Scholar]
- 12.Garg T, Young AJ, Kost KA, et al. Burden of Multiple Chronic Conditions among Patients with Urological Cancer. J Urol. 2018;199(2):543–550. doi: 10.1016/j.juro.2017.08.005 [DOI] [PubMed] [Google Scholar]
- 13.Engels EA, Pfeiffer RM, Ricker W, Wheeler W, Parsons R, Warren JL. Use of surveillance, epidemiology, and end results-medicare data to conduct case-control studies of cancer among the us elderly. Am J Epidemiol. 2011;174(7):860–870. doi: 10.1093/aje/kwr146 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Rubin EB, Buehler AE, Halpern SD. States worse than death among hospitalized patients with serious illnesses. JAMA Intern Med. 2016;176(10):1557–1559. doi: 10.1001/jamainternmed.2016.4362 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Garg T, Connors JN, Ladd IG, Bogaczyk TL, Larson SL. Defining Priorities to Improve Patient Experience in Non-Muscle Invasive Bladder Cancer. Bl Cancer. Published online2018. doi: 10.3233/blc-170138 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Faurot KR, Jonsson Funk M, Pate V, et al. Using claims data to predict dependency in activities of daily living as a proxy for frailty. Pharmacoepidemiol Drug Saf. 2015;24(1):59–66. doi: 10.1002/pds.3719 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Chrischilles E, Schneider K, Wilwert J, et al. Beyond Comorbidity. Med Care. 2014;52(3):S75–S84. doi: 10.1097/MLR.0000000000000026 [DOI] [PubMed] [Google Scholar]
- 18.Mental Health Research Network. MHRN List of Resources: List of Diagnoses. AccessedOctober 5, 2020. http://hcsrn.org/mhrn/en/Tools&Materials/MHRNResources/
- 19.Stewart CC, Lu CY, Yoon TK, et al. Impact of ICD-10-CM Transition on Mental Health Diagnoses Recording. eGEMs (Generating Evid Methods to Improv patient outcomes). 2019;7(1):14. doi: 10.5334/egems.281 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Tisminetzky M, Gurwitz JH, Fan D, et al. Multimorbidity Burden and Adverse Outcomes in a Community-Based Cohort of Adults with Heart Failure. J Am Geriatr Soc. 2018;66(12):2305–2313. doi: 10.1111/jgs.15590 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.D’Arcy LP, Stearns SC, Domino ME, Hanson LC, Weinberger M. Is Geriatric Care Associated with Less Emergency Department Use? J Am Geriatr Soc. 2013;61(1):4–11. doi: 10.1111/jgs.12039 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Jang TL, Patel N, Faiena I, et al. Comparative effectiveness of radical prostatectomy with adjuvant radiotherapy versus radiotherapy plus androgen deprivation therapy for men with advanced prostate cancer. Cancer. 2018;124(20):4010–4022. doi: 10.1002/cncr.31726 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Department of Health and Human Services. HHS Initiative of Multiple Chronic Conditions. AccessedOctober 5, 2020. https://www.hhs.gov/ash/about-ash/multiple-chronic-conditions/about-mcc/index.html
- 24.Garg T, Young AJ, O’Keeffe‐Rosetti M, et al. Association between treatment of superficial bladder cancer and 10‐year mortality in older adults with multiple chronic conditions. Cancer. 2018;124(23):4477–4485. doi: 10.1002/cncr.31705 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Cuthbertson CC, Kucharska-Newton A, Faurot KR, et al. Controlling for frailty in pharmacoepidemiologic studies of older adults. Epidemiology. 2018;29(4):556–561. doi: 10.1097/EDE.0000000000000833 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Tan HJ, Litwin MS, Chamie K, Saliba D, Hu JC. Trends in Aging-Related Services During Nephrectomy: Implications for Surgery in an Aging Population. J Am Geriatr Soc. 2017;65(10):2290–2296. doi: 10.1111/jgs.15046 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Amin MB, Edge SB, Greene FL, et al. E. AJCC: Urinary bladder. In: AJCC Cancer Staging Manual, 8th Edition.; 2017. [Google Scholar]
- 28.von Elm E, Altman DG, Egger M, Pocock SJ, Gøtzsche PC, Vandenbroucke JP. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies. J Clin Epidemiol. 2008;61(4):344–349. doi: 10.1016/j.jclinepi.2007.11.008 [DOI] [PubMed] [Google Scholar]
- 29.Nielsen ME, Smith AB, Meyer A, et al. Trends in stage-specific incidence rates for urothelial carcinoma of the bladder in the United States: 1988 to 2006. Cancer. 2014;120(1):86–95. doi: 10.1002/cncr.28397 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Russell LB, Ibuka Y, Carr D. How Much Time Do Patients Spend on Outpatient Visits?: The American Time Use Survey. Patient. 2008;1(3):211–222. doi: 10.2165/1312067-200801030-00008 [DOI] [PubMed] [Google Scholar]
- 31.Pergolotti M, Collins F, Deal AM, et al. Activities, Function, and Health-Related Quality of Life (HRQOL) of Older Adults with Cancer Mackenzi. J Geriatr Oncol. 2018;8(4):249–254. doi: 10.1016/j.jgo.2017.02.009.Activities [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Bohlen K, Scoville E, Shippee ND, May CR, Montori VM. Overwhelmed patients: A videographic analysis of how patients with type 2 diabetes and clinicians articulate and address treatment burden during clinical encounters. Diabetes Care. 2012;35(1):47–49. doi: 10.2337/dc11-1082 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Gallacher K, May CR, Montori VM, Mair FS. Understanding Patients ’ Experiences of Treatment Burden in Chronic Heart Failure Using Normalization Process Theory. Published online2011:235–243. doi: 10.1370/afm.1249.3 [DOI] [PMC free article] [PubMed]
- 34.Richards KA, Lin SW, Chuo CY, et al. Increased Bacillus Calmette-Guérin Treatment Intensity Associated With Improved Outcomes in Elderly Patients With Non–Muscle-invasive Bladder Cancer in United States Clinical Practice. Urology. 2020;145:172–180. doi: 10.1016/j.urology.2020.07.023 [DOI] [PubMed] [Google Scholar]
- 35.Spencer BA, McBride RB, Hershman DL, et al. Adjuvant intravesical bacillus calmette-guérin therapy and survival among elderly patients with non-muscle-invasive bladder cancer. J Oncol Pract. 2013;9(2):92–98. doi: 10.1200/JOP.2011.000480 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Heijnsdijk EAM, Nieboer D, Garg T, Lansdorp-Vogelaar I, de Koning HJ, Nielsen ME. Cost-effectiveness of surveillance schedules in older adults with non-muscle-invasive bladder cancer. BJU Int. 2019;123(2):307–312. doi: 10.1111/bju.14502 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Hurle R, Lazzeri M, Vanni E, et al. Active Surveillance for Low Risk Nonmuscle Invasive Bladder Cancer: A Confirmatory and Resource Consumption Study from the BIAS Project. J Urol. 2018;199(2):401–406. doi: 10.1016/j.juro.2017.08.091 [DOI] [PubMed] [Google Scholar]
- 38.Sreekantaswamy S, Endo J, Chen A, Butler D, Morrison L, Linos E. Aging and the treatment of basal cell carcinoma. Clin Dermatol. 2017;37(4):373–378. doi: 10.1016/j.clindermatol.2019.06.004 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Junn A, Shukla NR, Morrison L, et al. Development of a patient decision aid for the management of superficial basal cell carcinoma (BCC) in adults with a limited life expectancy. BMC Med Inform Decis Mak. 2020;20(1):1–8. doi: 10.1186/s12911-020-1081-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40.Tinetti ME, Naik AD, Dindo L, et al. Association of Patient Priorities-Aligned Decision-Making with Patient Outcomes and Ambulatory Health Care Burden among Older Adults with Multiple Chronic Conditions: A Nonrandomized Clinical Trial. JAMA Intern Med. 2019;179(12):1688–1697. doi: 10.1001/jamainternmed.2019.4235 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41.Centers for Disease Control National Center for Health Statistics. Number of Respondent-Reported Chronic Conditions From 10 Selected Conditions Among Adults Aged 18 and Over. Health, United States, 2017-Data Finder. AccessedOctober 5, 2020. https://www.cdc.gov/nchs/hus/contents2017.htm#Table_039 [Google Scholar]
- 42.Rocque GB, Williams CP, Miller HD, et al. Impact of travel time on health care costs and resource use by phase of care for older patients with cancer. J Clin Oncol. 2019;37(22):1935–1945. doi: 10.1200/JCO.19.00175 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 43.Rocque GB, Williams CP, Ingram SA, et al. Health care-related time costs in patients with metastatic breast cancer. Cancer Med. 2020;(August):1–9. doi: 10.1002/cam4.3461 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 44.Broderick A. The Future of Rural Transportation and Mobility for Older Adults: Current Trends and Future Directions in Technology-Enabled Solutions.; 2018. https://www.giaging.org/documents/180424_CITRIS_rural_mobility_paper_F.pdf
- 45.Cohen AJ, Ndoye M, Fergus KB, et al. Forecasting Limited Access to Urology in Rural Communities: Analysis of the American Urological Association Census. J Rural Heal. 2020;36(3):300–306. doi: 10.1111/jrh.12376 [DOI] [PubMed] [Google Scholar]
- 46.American Urological Association. The State of Urology Workforce and Practice in the United States 2019.; 2019. https://www.auanet.org/common/pdf/%0Dresearch/census/State-Urology-Workforce-Practice-US.pdf.
- 47.Leiser CL, Anderson RE, Martin C, Hanson HA, O’Neil B. Combining Drive Time and Urologist Density to Understand Access to Urologic Care. Urology. 2020;139:78–83. doi: 10.1016/j.urology.2020.02.004 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 48.Lobo JM, Horton B, Jones RA, et al. Blinded Comparison of Clarity, Proficiency and Diagnostic Capability of Tele-Cystoscopy Compared to Traditional Cystoscopy: A Pilot Study. J Urol. 2020;204(4):811–817. doi: 10.1097/JU.0000000000001092 [DOI] [PubMed] [Google Scholar]
- 49.Fisher K, Markle-Reid M, Ploeg J, et al. Self-management program versus usual care for community-dwelling older adults with multimorbidity: A pragmatic randomized controlled trial in Ontario, Canada. J Comorbidity. 2020;10:2235042X2096339. doi: 10.1177/2235042X20963390 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 50.Reckrey JM, Brody AA, McCormick ET, et al. Rationale and design of a randomized controlled trial of home-based primary care versus usual care for high-risk homebound older adults. Contemp Clin Trials. 2018;68(March 2018):90–94. doi: 10.1016/j.cct.2018.03.011 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 51.Smith SM, Wallace E, O’Dowd T, Fortin M. Interventions for improving outcomes in patients with multimorbidity in primary care and community settings. Cochrane database Syst Rev. 2016;3(3):CD006560. doi: 10.1002/14651858.CD006560.pub3 [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.