Cost-Related Medication Nonadherence and Cost-Reduction Strategies Among Elderly Cancer Survivors with Self-Reported Symptoms of Depression

Abstract

How depression affects the medication cost burden for elderly cancer survivors has not been well studied. This study aims to investigate whether depression is associated with higher rates of cost-related medication nonadherence, and cost-reduction strategies among the elderly cancer survivors. Self-reports from survey files of the 2015 Medicare Current Beneficiary Survey-Medicare database were used to identify elderly cancer patients aged 65 years and older with and without depression. The 2 outcomes were cost-related nonadherence (CRN) and adoption of cost-reduction strategies. Bivariate analysis was used to describe the sample. Multivariable logistic regression was performed to examine the impact of depression on CRN and the use of cost-reduction strategies, after controlling for all other covariates. Among the 3765 elderly cancer survivors identified, 523 (14%) reported depression. In the group with depression, 26% reported CRN compared with 12% of the group without depression; 71% of individuals with depression reported having cost-reduction strategies while 65% of individuals with no depression reported such activity. In adjusted analyses, individuals with depression were significantly more likely to report CRN (adjusted odds ratio, 1.84; 95% confidence interval 1.33–2.54) and cost-reduction strategies (adjusted odds ratio, 1.37; 95% confidence interval, 1.07–1.76). Depression was associated with higher probabilities of both CRN and the adoption of cost-reduction strategies, indicating that depression can exacerbate the medication cost burden for elderly cancer survivors. It is important to detect and manage depression in elderly cancer survivors to reduce CRN and cost-reduction strategies.

Introduction

As costs of prescription drugs and especially cancer drugs continue to soar, cancer survivors are often burdened by medication costs and a significant percentage have difficulty paying for prescriptions.¹ Additionally, given the aging US population and the increased risk of cancer with age, the majority of cancer survivors are age 65 and older.² Many elderly cancer survivors resort to behaviors such as not adhering to medication and adoption of cost-reduction strategies (eg, spending less on basic needs to afford medications, using generic drugs to relieve the medication cost burden). For example, a study using the 2005 Medicare Current Beneficiary Survey (MCBS)-Medicare data found that approximately 10% of cancer survivors reported cost-related nonadherence (CRN), while 6% of cancer survivors reported spending less on basic needs to offset the costs of medications.³ Another study using 2006–2013 National Health Interview Survey data reported 3.6% and 4.6% CRN rates for elderly male and female cancer survivors, respectively.⁴ CRN and other behaviors related to cost burden can result in a wide range of negative consequences including stroke, myocardial infarction, more emergency utilization and hospitalization, and even higher mortality.^5

–8

Depression is a common comorbid condition of cancer. Prevalence of depression among cancer patients is high and is estimated to be 8% to 14% based on a meta-analysis.⁹ Coexisting depression can negatively impact many aspects of cancer survivors' lives including lowering quality of life.^10,11 Additionally, depression also is a very economically burdensome disorder and is the most costly mental health disease.¹² Many studies have found that depression is associated with excess health care costs, with these costs continuing to increase rapidly.^13,14 It has been shown that depression is associated with increased direct health care costs for elderly patients.^15,16 In terms of medication cost burden, previous studies have demonstrated that depression is associated with CRN in elderly populations.^8,17 For example, a study showed that 19% of elderly Medicare beneficiaries with depression reported CRN, while 12% of their counterparts without depression had CRN. In adjusted analyses, depression was still a significant risk factor of CRN (odds ratio [OR], 1.4; 95% confidence interval [CI], 1.1–1.7) in elderly individuals.⁸

Although the government introduced Medicare Part D in 2006 to alleviate Medicare beneficiaries' financial burden from prescription medication costs, medication affordability remains a significant and persistent problem among high-need elderly populations.^17
–19 Because cancer and depression are both high-cost chronic diseases for the elderly population, it is important to study the medication cost burden among depressed cancer survivors; strikingly, little is known about the impact of depression on medication cost burden in this population. Therefore, this study used CRN and cost-reduction strategies, such as spending less on basic needs to afford medications, to characterize this financial burden among elderly cancer survivors and examined the extent to which depression affects it.

Methods

Data source and sample

This study used 2015 survey files of the MCBS-Medicare sponsored by the Centers for Medicare & Medicaid Services. MCBS-Medicare is generated by sampling a nationally representative sample of aged, disabled, and noninstitutionalized Medicare beneficiaries, who are surveyed up to 3 rounds per year for 4 years consecutively. MCBS survey files were used because the data contain comprehensive and detailed information on patient demographics, socioeconomic status, health care utilization, and self-reported health status and symptoms.²⁰

This study received institutional review board exemption status (PA12-0458) from the University of Texas MD Anderson Cancer Center.

Study cohort

The study cohort included community-dwelling elderly cancer survivors aged 65 years and older. Cancer survivors were identified based on beneficiaries' responses to 2 MCBS survey questions. They were asked whether “a doctor (ever) told (you) that (you) had any skin cancer?” and whether “a doctor (ever) told (you) that (you) had any (other) kind of cancer, malignancy, or tumor other than skin cancer?” If a beneficiary responded to one or both of the questions affirmatively, he/she was considered a cancer survivor. There are certainly many definitions for cancer survivors in the literature that can be more accurate.²¹ However, because of data limitations, the stated definition, based on the responses to MCBS questions, was used in this study.

Dependent variables

The 2 dependent variables were CRN and cost-reduction strategies. Many articles in the literature use CRN and cost-reduction strategies as measures to assess the cost burden for patients, so they are well-accepted measures.^{20,22

–27} In the present study, survey respondents were considered to have CRN if they reported ever exhibiting any of the following behaviors: taking smaller doses to make medicine last longer, skipping doses of a prescription to make it last longer, and failing or delaying to fill a prescription because of cost. These questions were for general medications. Respondents were considered to have used cost-reduction strategies if they answered yes or often or sometimes to any of the cost-reduction strategy questions, which included spending less on other basic needs to afford medications, generic drug use, purchasing prescriptions via mail/Internet, asking for prescription samples from a doctor, and comparing pharmacies. These measures have shown 0.6–0.9 test–retest reliability²⁴ and high construct validity.^22,25
–27

Independent variables

Depression

Depression symptoms were defined by 2 questions in the survey: (1) “In the last 12 months, how much of the time did you feel sad, blue or depressed?” and (2) “In the last 12 months, have you had 2 weeks or more when you lost interest or pleasure in the things that you usually cared about or enjoyed?” If a beneficiary responded all of the time or most of the time to the first question, and/or answered yes to the second question, then the beneficiary was considered to have self-reported depression. The combination of these 2 questions was found to have 91% sensitivity and 86% specificity in detecting depression in cancer and palliative care; therefore, it is a good measure of depression presence based on patient self-report.²⁸

Other independent variables

Besides depression status, the other independent variables were based on the expanded Andersen Behavioral Model of Health Services Utilization.²⁹ This model has 5 main constructs: (1) predisposing factors, (2) enabling factors, (3) need factors, (4) personal health practices and use of health services, and (5) external environment factors. In this study, predisposing factors included sex (female/male), age (65–74, 75–84 and ≥85 years), and race/ethnicity (non-Hispanic white, other). Enabling characteristics included marital status (married, other), educational attainment (less than high school, high school, some college, college), household income (<$25,000 and ≥$25,000), and drug coverage type (public, private, none). Need factors included perceived health status (excellent/very good/good, fair/poor); functional status, which is the number of activities of daily living with limitations (none; 1–2; ≥3 limitations); and number of comorbidities, including heart disease, stroke/brain hemorrhage, hypertension, diabetes mellitus, arthritis, mental disorder other than depression, neurological conditions, and lung disease (0–1; 2–3; ≥4 comorbidities). Personal health practices and use of health services included smoking status (current, past, never), body mass index (BMI; underweight or normal, referred to as BMI <25 kg/m2; overweight, referred to as BMI ≥25–29.9 kg/m2; obese/morbid obese, referred to as BMI ≥30 kg/m2).³⁰ External environment was captured by metro status (metropolitan, nonmetropolitan).

Statistical analysis

Bivariate analyses were conducted to compare the characteristics of cancer survivors with and without depression. Prevalence of CRN and cost-reduction strategies by all the independent variables are then reported Finally, 2 separate multivariate logistic regressions were used, controlling for all the independent variables to estimate the odds of CRN and cost-reduction strategies for cancer survivors with and without depression.

All statistical analyses were adjusted for the MCBS complex survey design and performed by using survey sampling and analysis procedures in SAS Enterprise Guide version 6.1 (SAS Institute Inc., Cary, NC) and Stata 14.2 (StataCorp LLC, College Station, TX).

Results

Table 1 presents a description of the sample of elderly cancer survivors by depression status; weighted percentages are reported. Among the 3765 elderly cancer survivors, 523 (14%) reported depression. Compared to those without depression, individuals with depression had significantly higher rates of CRN (14% higher) and cost-reduction strategies (6% higher). Additionally, for predisposing factors, a greater proportion of the depressed group had household income <$25,000 (14% higher) and for enabling factors, education attainment less than high school (12% higher). For need factors, depressed respondents were more likely to have at least 4 comorbid conditions (17% more likely), 1 or 2 functional limitations (15% more likely) or at least 3 functional limitations (20% more likely), and fair/poor self-perceived health status (29% more likely) compared to their counterparts without depression. For factors related to personal health practices and use of health services, depressed individuals were more likely to be obese/morbid obese (9% more likely) (Table 1).

Table 1.

Characteristics of Elderly Cancer Survivors With and Without Depression

Characteristics	Elderly cancer survivors (N = 3765)
	Without depression (N = 3242,Wt% = 86)		With depression (N = 523,Wt% = 14)
	N	Wt%	N	Wt %	P
CRN					<.0001
Yes	351	12	119	26
No	2891	88	404	74
Cost-Reduction Strategies					0.0099
Yes	2068	65	366	71
No	1174	35	157	29
Predisposing
Drug Insurance					0.3018
Public	2347	72	404	75
Private	609	20	80	16
None	286	8	39	9
Income					<.0001
<25,000	997	28	234	42
≥25,000	2245	72	289	58
Sex					0.0012
Male	1652	50	223	41
Female	1590	50	300	59
Age (years)					0.1272
65–74	1026	47	181	51
75–84	1485	37	216	33
≥85	731	15	126	16
Race/ethnicity					0.0563
Non-Hispanic white	3024	93	470	90
Other	218	7	53	10
Enabling
Marital status					0.1068
Married	1862	60	263	55
Other	1380	40	260	45
Educational attainment					<.0001
Less than high school	426	11	128	23
High school	1079	32	181	33
Some college	514	16	80	16
College	1223	40	134	29
Need
Number of comorbidities					<.0001
0–1	913	32	78	16
2–3	1763	51	266	50
≥4	566	17	179	34
Functional status limitation					<.0001
0	2316	75	197	40
1–2	720	20	186	35
≥3	206	5	140	25
Perceived health status					<.0001
Excellent/very good/good	2767	86	299	57
Fair/poor	475	14	224	43
Personal health practices and use of health services
Smoking Status					0.6014
Current	237	8	38	9
Past	1760	54	289	55
Never	1245	38	196	36
BMI					<.0001
Underweight/normal	1161	33	182	34
Overweight	1336	42	179	32
Obese/morbid obese	745	25	162	34
External Environment
Rural status					0.2675
Yes	262	7	47	9
No	2980	93	476	91

BMI, body mass index; CRN, cost-related medication nonadherence; Wt%, weighted percentage.

The prevalence of CRN and cost-reduction strategies stratified by independent variables is presented in Table 2. Depression was significantly associated with CRN and cost-reduction strategies and there were significant subgroup differences.

Table 2.

Prevalence of CRN and Cost-Reduction Strategies, Overall and by Characteristics in Elderly Cancer Survivors

Overall Rates	CRN			Cost-reduction strategies
	14			66
	Without depression	With depression	P	Without depression	With depression	P
Characteristics	Rates (95% CI)	Rates (95% CI)		Rates (95% CI)	Rates (95% CI)
Total rates	12 (10–14)	26 (21–31)	<.0001	65(62–67)	71 (67–76)	0.0095
Predisposing
Drug Insurance
Public	13 (11–15)	25 (20–30)	<.0001	64 (61–67)	71 (65–76)	0.0318
Private	7 (4–9)	20 (9–31)	0.0010	70 (65–75)	74 (62–85)	0.5720
None	16 (11–21)	42 (20–65)	0.0060	60 (54–66)	73 (55–90)	0.2110
Income
<25,000	15 (12–18)	26 (20–33)	0.0002	59 (56–63)	68 (60–75)	0.0473
≥25,000	11 (9–13)	25 (19–31)	<.0001	67 (64–69)	74 (69–80)	0.0198
Sex
Male	11 (9–14)	19 (12–25)	0.0342	66 (63–69)	71 (64–77)	0.2141
Female	13 (11–15)	31 (24–37)	<.0001	63 (60–66)	72 (66–78)	0.0088
Age (years)
65–74	15 (12–19)	34 (26–42)	<.0001	66 (62–70)	73 (66–81)	0.0597
75–84	10 (8–11)	19 (13–26)	0.0005	67 (64–70)	70 (63–77)	0.4107
≥85	8 (6–11)	12 (6–18)	0.1919	56 (51–60)	68 (58–78)	0.0269
Race/ethnicity
White	12 (10–14)	25 (20–30)	<.0001	65 (63–68)	72 (67–76)	0.0299
Other	12 (7–17)	33 (15–51)	0.0037	54 (45–63)	71 (59–82)	0.0232
Enabling
Marital status
Married	11 (9–13)	25 (18–31)	<.0001	68 (65–71)	73 (68–79)	0.0977
Other	14 (11–16)	27 (21–34)	<.0001	60 (56–63)	69 (62–77)	0.0297
Educational attainment
Less than high school	11 (8–14)	27 (17–37)	<.0001	56 (49–62)	68 (59–78)	0.0222
High school	14 (11–17)	22 (15–28)	0.0188	63 (60–67)	69 (61–76)	0.1643
Some college	11 (8–14)	36 (20–52)	<.0001	65 (60–71)	66 (54–77)	0.9615
College	11 (8–15)	24 (15–33)	0.0029	68 (64–72)	80 (72–88)	0.0139
Need
Number of comorbidities
0–1	9 (5–13)	25 (12–37)	0.0029	60 (55–64)	67 (54–80)	0.3238
2–3	13 (10–15)	22 (15–28)	0.0071	66 (63–68)	70 (64–77)	0.1845
≥4	16 (12–19)	33 (24–41)	<.0001	71 (67–75)	75 (67–83)	0.3500
Functional status limitation
0	11 (9–13)	21 (14–28)	0.0016	64 (61–67)	72 (65–79)	0.0542
1–2	17 (13–20)	33 (25–42)	<.0001	67 (63–70)	71 (63–79)	0.3194
≥3	15 (9–21)	22 (13–31)	0.1867	64 (56–72)	71 (62–81)	0.2756
Perceived health status
Excellent/very good/good	11 (9–13)	20 (14–26)	0.001	65 (62–67)	71 (65–77)	0.0649
Fair/poor	19 (14–23)	33 (25–42)	0.0001	64 (58–69)	72 (64–80)	0.0619
Personal health practices and use of health services
Smoking Status
Current	13 (7–20)	39 (17–61)	0.0032	63 (55–71)	83 (70–96)	0.0148
Past	11 (9–13)	23 (18–29)	<.0001	65 (62–69)	72 (66–78)	0.0687
Never	13 (9–17)	27 (18–35)	0.0020	64 (61–67)	68 (60–77)	0.3925
BMI
Underweight/normal	9 (7–11)	22 (14–30)	<.0001	62 (58–65)	73 (65–80)	0.0065
Overweight	12 (9–16)	20 (13–27)	0.0449	67 (64–70)	67 (60–74)	0.9149
Obese/morbid obese	16 (12–19)	35 (26–44)	<.0001	65 (60–69)	75 (66–84)	0.0502
External Environment
Rural status
Yes	15 (9–21)	25 (20–30)	0.0537	62 (50–73)	69 (59–80)	0.3199
No	12 (10–14)	31 (12–49)	<.0001	65 (62–67)	72 (67–77)	0.0143

BMI, body mass index; CI, confidence interval; CRN, cost-related medication nonadherence.

The results of multivariate logistic regression controlling for all independent variables are summarized in Table 3. In terms of CRN, respondents with depression were significantly more likely to report CRN (OR, 1.84; 95% CI, 1.33–2.54) compared with those without depression. Statistically significant associations between CRN and the following characteristics also were identified: drug insurance type, age, functional status, perceived health status, and BMI. For cost-reduction strategies, depressed individuals also were significantly more likely to report spending less on basic needs (OR, 1.37; 95% CI, 1.07–1.76). Other significant characteristics associated with adopting cost-reduction strategies were age, race/ethnicity, marital status, educational attainment, and comorbid conditions.

Table 3.

Adjusted Predictors of CRN and Cost-Reduction Strategies from Multivariate Logistic Regression

	CRN		Cost-saving strategies
	Odds ratio (95% CI)	P	Odds ratio (95% CI)	P
Depression (Ref = No)
Yes	1.84(1.33–2.54)	0.0003	1.37(1.07–1.76)	0.0122
Predisposing
Drug Insurance (Ref = Public)
Private	0.53(0.36–0.79)	0.0016	1.22(0.97–1.53)	0.0875
None	1.58(1.16–2.15)	0.0037	0.95(0.73–1.23)	0.7059
Income (Ref = Male)
≥25,000	0.92(0.69–1.22)	0.5437	1.10(0.91–1.33)	0.3087
Sex (Ref = Male)
Female	1.20(0.91–1.58)	0.2062	1.00(0.83–1.21)	0.9708
Age (Ref = 65–74 years)
75–84	0.51 (0.39–0.66)	<.0001	1.04(0.86–1.25)	0.6945
≥85	0.38 (0.26–0.55)	<.0001	0.71(0.56–0.90)	0.0052
Race/ethnicity (Ref = White)
Other race	0.91(0.59–1.41)	0.6835	0.71(0.52–0.96)	0.0263
Enabling
Marital status (Ref = Married)
Other	1.18(0.88–1.59)	0.2736	0.78(0.65–0.94)	0.0101
Educational attainment (Ref = Less than high school)
High school	1.25(0.86–1.81)	0.2510	1.21(0.91–1.59)	0.1835
Some College	1.31(0.80–2.16)	0.2814	1.30(0.88–1.94)	0.1923
College	1.27(0.83–1.96)	0.2696	1.52(1.11–2.08)	0.0089
Need
Number of comorbidities (Ref = 0–1)
2–3	1.29(0.86–1.94)	0.2264	1.35(1.10–1.64)	0.0037
≥4	1.44(0.94–2.20)	0.7687	1.84(1.41–2.39)	<.0001
Functional status limitation (Ref = 0)
1–2	1.64(1.26–2.13)	0.0002	1.12(0.91–1.39)	0.2782
≥3	1.07(0.69–1.65)	0.7687	1.09(0.81–1.47)	0.5788
Perceived Health Status (Ref = Excellent/very good/good)
Fair/Poor	1.55(1.16–2.09)	0.0034	0.96(0.75–1.23)	0.7389
Personal health practices and use of health services
Smoking Status (Ref = Current)
Past	0.90(0.52–1.57)	0.7069	0.94(0.65–1.38)	0.7679
Never	1.09(0.65–1.85)	0.7380	0.93(0.64–1.36)	0.7218
BMI (Ref = Underweight/normal)
Overweight	1.30(0.96–1.75)	0.0892	1.09(0.92–1.30)	0.3209
Obese/morbid obese	1.45(1.05–2.01)	0.0228	0.97(0.76–1.24)	0.8031
External Environment
Rural status (Ref = No)
Yes	1.18(0.74–1.88)	0.4908	0.89(0.58–1.36)	0.5847

BMI, body mass index; CI, confidence interval; CRN, cost-related medication nonadherence.

Discussion

In this nationally representative population of elderly cancer survivors, the prevalence of self-reported depression was 14.1%, which is similar to the 14% prevalence rate found in a study about depression in adult cancer patients using Medical Expenditure Panel Survey data from 2006–2009.³¹ The latter study used patients' medical files to identify the presence of depression. This value is reasonable because it falls in the range of 8% to 24% estimated from a meta-analysis of depression prevalence among cancer patients assessed by diagnostic interviews and self-report instruments.⁹

In the present study sample, 14% of individuals reported CRN. This CRN rate is somewhat higher than 10%, which was reported in a previous study.³ In a study about elderly cancer survivors using 2005 MCBS data, the reported CRN prevalence was 10.3%.³ The difference is plausible because CRN rates can vary based on CRN definitions and study populations. The definition of CRN in the current study included a delay in filling prescription measure, which became available after Medicare Part D initiation in 2006 and was not available for the aforementioned study.

For cost-reduction tactics, 66% of the sample reported at least 1 of the 5 strategies. It is reasonable. A study examining the cost-reduction strategies of elderly Medicare cancer survivors showed that 6.3% reported spending less on basic needs, 50.9% reported generic drug use, 52.3% reported asking for prescription samples from a doctor, 36.3% reported purchasing prescriptions via mail/Internet, and 22.7% reported comparing pharmacies.³

Results of the present study's multivariate logistic regression showed that depression was associated with higher probabilities of both CRN and cost-reduction strategies. These results indicate that depression can exacerbate elderly cancer survivors' medication cost burden. Indeed, previous research has shown that comorbid depression can cause incremental cost burden for patients. One study found that adult cancer survivors with depression had $913 (107.1%) higher prescription expenditures (including both payer payments and out-of-pocket costs) than those without depression.³¹ Additionally, polypharmacy is common in the geriatric oncology population.³² Because treatment of comorbid depression with antidepressants will increase their medication use and some psychiatric medications are costly,⁸ elderly cancer survivors may become overwhelmed with higher medication costs. Further, most depression medications now are generic³³ and do not normally add a great monetary burden to patients with low co-pays.³⁴

This study has some limitations. First, it uses a cross-sectional design, which may have unmeasured confounding factors than cannot be controlled for; therefore, the results cannot imply causation. Second, CRN and cost-reduction strategies were for general medications; MCBS data do not distinguish among specific medications.

This study has several strengths worth noting. Although some previous studies have reported the prevalence of depression among cancer patients, most of them did not focus on the elderly.³¹ Present study results add to the literature documenting the prevalence of depression, and rates of CRN and cost-reduction strategies for elderly cancer survivors. Also, the findings reveal a strong relationship between depression and both CRN and cost-reduction strategies in elderly cancer survivors, which, to the best of the research team's knowledge, has not been documented previously.

In terms of implications, this study extends the limited existing literature documenting the notable medication cost burden imposed by depression on elderly cancer survivors. Additionally, the findings reinforce the importance of effective depression management in the geriatric oncology setting. There are many depression management choices, including pharmacological, non-pharmacological (eg, psychotherapy) and combinations of pharmacological and psychotherapeutic techniques in this setting.³⁵ Recently, collaborative depression care models have been advocated by multiple studies.^36
–38 In these models, psychiatrists, care managers, and physicians collaborate to provide a systematic approach to manage depression. Currently, little to no research assesses the effectiveness of depression treatment options in decreasing elderly cancer survivors' CRN and other cost-saving behaviors. As such knowledge is crucial for improving their quality and perhaps length of life, further studies are greatly needed.

Conclusions

This study revealed that depression was associated with CRN and cost-reduction strategies, indicating that depression can lead to incremental medication cost burden for elderly cancer survivors. These findings strengthen the importance of diagnosing and treating this very common mental disorder in order to decrease this burden. Also, the results provide compelling evidence to develop and verify efficacious treatments for depression in the geriatric oncology setting.

Footnotes

Acknowledgement

We thank Gary Deyter for editorial assistance.

Author Disclosure Statement

The authors declare that there are no conflicts of interest. The authors received no financial support for the conduct of this study and/or preparation of this manuscript.

References

Bender

. Seeking value as cancer drug costs soar. Cancer Discov, 2013; 3:833–834.

Bluethmann

, Mariotto

, Rowland

. Anticipating the “silver tsunami”: prevalence trajectories and comorbidity burden among older cancer survivors in the United States. Cancer Epidemiol Biomarkers Prev, 2016; 25:1029–1036.

Nekhlyudov

, Madden

, Graves

, Zhang

, Soumerai

, Ross-Degnan

. Cost-related medication nonadherence and cost-saving strategies used by elderly Medicare cancer survivors. J Cancer Surviv, 2011; 5:395–404.

Lee

, Khan

. Gender differences in cost-related medication non-adherence among cancer survivors. J Cancer Surviv, 2016; 10:384–393.

Heisler

, Langa

, Eby

, Fendrick

, Kabeto

, Piette

. The health effects of restricting prescription medication use because of cost. Med Care, 2004; 42:626–634.

Tamblyn

, Laprise

, Hanley

, et al. Adverse events associated with prescription drug cost-sharing among poor and elderly persons. JAMA, 2001; 285:421–429.

Piette

, Wagner

, Potter

, Schillinger

. Health insurance status, cost-related medication underuse, and outcomes among diabetes patients in three systems of care. Med Care, 2004; 42:102–109.

Bambauer

, Safran

, Ross-Degnan

, et al. Depression and cost-related medication nonadherence in Medicare beneficiaries. Arch Gen Psychiatry, 2007; 64:602–608.

Krebber

, Buffart

, Kleijn

, et al. Prevalence of depression in cancer patients: a meta-analysis of diagnostic interviews and self-report instruments. Psychooncology, 2014; 23:121–130.

10.

Yang

, Brothers

, Andersen

. Stress and quality of life in breast cancer recurrence: moderation or mediation of coping?. Ann Behav Med, 2008; 35:188–197.

11.

Holland

, Weiss

. The new standard of quality cancer care: integrating the psychosocial aspects in routine cancer from diagnosis through survivorship. Cancer J, 2008; 14:425–428.

12.

Dieleman

, Baral

, Birger

, et al. US spending on personal health care and public health, 1996–2013. JAMA, 2016; 316:2627–2646.

13.

Olfson

, Marcus

, Druss

, Elinson

, Tanielian

, Pincus

. National trends in the outpatient treatment of depression. JAMA, 2002; 287:203–209.

14.

Greenberg

, Fournier

, Sisitsky

, Pike

, Kessler

. The economic burden of adults with major depressive disorder in the United States (2005 and 2010). J Clin Psychiatry, 2015; 76:155–162.

15.

Katon

, Lin

, Russo

, Unutzer

. Increased medical costs of a population-based sample of depressed elderly patients. Arch Gen Psychiatry, 2003; 60:897–903.

16.

Unutzer

, Patrick

, Simon

, et al. Depressive symptoms and the cost of health services in HMO patients aged 65 years and older. A 4-year prospective study. JAMA, 1997; 277:1618–1623.

17.

Zivin

, Madden

, Zhang

, Soumerai

, Graves

. Cost-related medication nonadherence among beneficiaries with depression following Medicare Part D. Am J Geriatr Psychiatry, 2009; 17:1068–1076.

18.

Naci

, Soumerai

, Ross-Degnan

, et al. Medication affordability gains following Medicare Part D are eroding among elderly with multiple chronic conditions. Health Aff (Millwood), 2014; 33:1435–1443.

19.

Madden

, Graves

, Zhang

, et al. Cost-related medication nonadherence and spending on basic needs following implementation of Medicare Part D. JAMA, 2008; 299:1922–1928.

20.

Centers for Medicare & Medicaid Services. Medicare Current Beneficiary Survey (MCBS). https://www.cms.gov/MCBS Accessed December 2017 .

21.

Feuerstein

. Defining cancer survivorship. J Cancer Surviv, 2007; 1:5–7.

22.

Burcu

, Alexander

, Ng

, Harrington

. Construct validity and factor structure of survey-based assessment of cost-related medication burden. Med Care, 2015; 53:199–206.

23.

Blumberg

, Prager

, Liebmann

, Cioffi

, De Moraes

. Cost-related medication nonadherence and cost-saving behaviors among patients with glaucoma before and after the implementation of Medicare Part D. JAMA Ophthalmol, 2015; 133:985–996.

24.

Pierre-Jacques

, Safran

, Zhang

, et al. Reliability of new measures of cost-related medication nonadherence. Med Care, 2008; 46:444–448.

25.

Safran

, Neuman

, Schoen

, et al. Prescription drug coverage and seniors: findings from a 2003 national survey: where do things stand on the eve of implementing the new Medicare Part D benefit?. Health Aff (Millwood), 2005; Suppl Web Exclusives:W5-152–W5-166.

26.

Safran

, Neuman

, Schoen

, et al. Prescription drug coverage and seniors: how well are states closing the gap?. Health Aff (Millwood), 2002; Suppl Web Exclusives:W253–W268.

27.

Soumerai

, Pierre-Jacques

, Zhang

, et al. Cost-related medication nonadherence among elderly and disabled medicare beneficiaries: a national survey 1 year before the medicare drug benefit. Arch Intern Med, 2006; 166:1829–1835.

28.

Mitchell

. Are one or two simple questions sufficient to detect depression in cancer and palliative care? A Bayesian meta-analysis. Br J Cancer, 2008; 98:1934–1943.

29.

Andersen

. Revisiting the behavioral model and access to medical care: does it matter?. J Health Soc Behav, 1995; 36:1–10.

30.

Centers for Disease Control and

Prevention

. Healthy Weight: About Adult

BMI

. www.cdc.gov/healthyweight/assessing/bmi/adult_bmi/index.html Accessed November 21, 2018 .

31.

Pan

, Sambamoorthi

. Health care expenditures associated with depression in adults with cancer. J Community Support Oncol, 2015; 13:240–247.

32.

Hersh

, Beldowski

, Hajjar

. Polypharmacy in the geriatric oncology population. Curr Oncol Rep, 2017; 19:73.

33.

Ventimiglia

, Kalali

. Generic penetration in the retail antidepressant market. Psychiatry (Edgmont), 2010; 7:9–11.

34.

Bao

, Ryan

, Shao

, Pincus

, Donohue

. Generic initiation and adherence to antidepressant therapy under Medicare Part D. Am J Manag Care, 2013; 19: 989–998.

35.

Winell

, Roth

. Psychiatric assessment and symptom management in elderly cancer patients. Oncology (Williston Park), 2005; 19:1479–1490, discussion 1492, 1497, 1501–1507.

36.

, Kennedy

, Byrne

, et al. Management of depression in patients with cancer: a clinical practice guideline. J Oncol Pract, 2016; 12:747–756.

37.

Sharpe

, Walker

, Hansen

, et al. Integrated collaborative care for comorbid major depression in patients with cancer (SMaRT Oncology-2): a multicentre randomised controlled effectiveness trial. Lancet, 2014; 384:1099–1108.

38.

Duarte

, Walker

, et al. Cost-effectiveness of integrated collaborative care for comorbid major depression in patients with cancer. J Psychosom Res, 2015; 79:465–470.