Abstract
Purpose
Adjuvant endocrine therapy (AET) reduces breast cancer morbidity and mortality; however, adherence is suboptimal. Interventions exist, yet few have improved adherence. Patient characteristics may alter uptake of an intervention to boost adherence. We examined moderators of the effect of a virtual intervention (STRIDE; #NCT03837496) on AET adherence after breast cancer.
Methods
At a large academic medical center, patients taking AET (N = 100; Mage = 56.1, 91% White) were randomized to receive STRIDE versus medication monitoring. All stored their medication in digital pill bottles (MEMS Caps) which captured objective adherence. Participants self-reported adherence (Medication Adherence Report Scale) at 12 weeks post-baseline. Moderators included age, anxiety, and depressive symptoms (Hospital Anxiety and Depression Scale), AET-related symptom distress (Breast Cancer Prevention Trial Symptom Scale), and AET-specific concerns (Beliefs about Medications Questionnaire). We used hierarchical linear modeling (time × condition × moderator) and multiple regression (condition × moderator) to test the interaction effects on adherence.
Results
Age (B = 0.05, SE = 0.02, p = 0.003) and AET-related symptom distress (B = −0.04, SE = 0.02, p = 0.02) moderated condition effect on self-reported adherence while anxiety (B = −1.20, SE = 0.53, p = 0.03) and depressive symptoms (B = −1.65, SE = 0.65, p = 0.01) moderated objective adherence effects. AET-specific concerns approached significance (B = 0.91, SE = 0.57, p = 0.12). Participants who received STRIDE and were older or presented with lower anxiety and depressive symptoms or AET-related symptom distress exhibited improved adherence. Post hoc analyses revealed high correlations among most moderators.
Conclusions
A subgroup of patients who received STRIDE exhibited improvements in AET adherence. The interrelatedness of moderators suggests an underlying profile of patients with lower symptom burden who benefitted most from the intervention.
Study registration
Keywords: Adjuvant endocrine therapy, Adherence, Symptom burden, Psychosocial intervention, Breast cancer, Symptom profile
Introduction
Hormone-sensitive breast cancer comprises nearly 80% of all breast cancer malignancies [1], for which adjuvant endocrine therapy (AET) treatment is critical [2]. By interfering with the pathway of estrogen and progesterone to breast cancer cells, AET hampers hormone-related tumor growth and prevents downstream risk of recurrence [2]. Guidelines currently recommend up to 10 years of AET receipt for nonmetastatic breast cancer, which can pose challenges to patients due to significant side effects and long-term health risks [2].
Prior evidence suggests up to two thirds of breast cancer survivors do not complete the recommended course of daily AET as prescribed [3]. Side effects of these medications, such as joint pain, hot flashes, sleep disruption, fatigue, sexual dysfunction, and cognitive impairment, are challenging and often result in non-adherence and discontinuation [4–8]. Further, patients contend with elevated risks of developing long-term health issues or AET-related toxicities [9, 10]. These AET-interfering elements have been summarized in the model of determinants of adherence, by Lee and Min [11]. This model characterizes three, person-level domains that impact adherence: sociodemographic/medical characteristics, psychological symptoms, and AET-specific concerns. Prior literature has extensively found evidence for barriers to AET across these domains [5, 6, 12–14]. Such non-adherence presents as a crucial target for intervention due to the link between AET non-adherence or discontinuation and increased cancer-related morbidity and mortality [15].
Within this burgeoning area of research, numerous behavioral interventions have been developed [16–19]. However, there is a lack of evidence for improvement in AET adherence [20–22]. In a recent meta-analysis (2019) [16], null effects were found across seven interventional trials testing effects on AET adherence. Within sub-intervention content analyses, some intervention characteristics, such as those which are multi-faceted (e.g., delivered across multiple modalities), behavioral, and interactive, were found to achieve greater effect sizes for increasing AET adherence compared to standard patient education or information alone [16]. Existing medication adherence programs for other medical populations (e.g., HIV) are also being adapted to target AET adherence, and other intervention trials are currently on-going [18, 19, 23]. While some intervention components may have greater potential to affect AET adherence, few overall effects have been found. As intervention development across this area of the field continues, moderator analyses at a patient-level may offer insights into specific patient subgroups who benefit most from such programs. Notably, a limited number of studies examine both self-reported (e.g., self-report measures) and objective (e.g., electronic pill bottles, pill counts) AET patient adherence simultaneously [21, 22] and none in the context of a behavioral intervention to improve adherence. Prior literature has noted discrepancies between these measures of medication adherence [21], so we evaluated both forms of adherence in the current study.
In this exploratory study, we examined moderators of intervention effects on self-reported and objective AET adherence after a virtual, small-group behavioral intervention (STRIDE) for AET adherence [24]. Results from the parent trial showed the intervention to be acceptable and feasible, yet no changes in self-reported or objective adherence were observed [25]. Patient characteristics and sociodemographics have been shown to alter psychosocial treatment response and uptake. In a study of patients taking oral chemotherapy, those who presented with greater levels of anxiety symptoms or self-reported adherence difficulties (e.g., forgetting, dose alterations) who received a smartphone intervention showed improved objective adherence versus others [26]. Younger women taking AET also often report greater difficulties with toxicity which interferes with adherence [27, 28]. While not examined in the current study, other, structural factors, including geographic resources and socioeconomic status impact AET initiation and adherence beyond the patient-level and are important to recognize [29, 30]. The aim of this study was to examine the three domains of AET adherence determinants. We hypothesized that patient subgroups presenting in greater need may benefit most with regard to AET adherence after a behavioral intervention (Fig. 1). Specifically, patients presenting with factors associated with greater AET difficulties, younger age, greater anxiety and depressive symptoms, greater AET-related symptom distress, and greater AET-specific concerns who received STRIDE would benefit most related to adherence.
Fig. 1.
Conceptual model of AET determinants and condition effect on adherence
Methods
Study design
The current study is an exploratory analysis of a parallel, randomized-controlled trial testing an intervention to improve symptom management and AET adherence at an academic medical center in Boston, Massachusetts, and three affiliated community sites located in suburban communities in the Boston metropolitan area. Eligible patients enrolled between October 2019 and June 2021, with a brief pause in accrual between March 2020 and May 2020 due to the COVID-19 pandemic. The trial was approved by the Dana Farber/Harvard Cancer Center Institutional Review Board (pre-registration #NCT03837496).
Participants
Eligible participants included adult females (≥21 years old) with nonmetastatic (0–IIIB) hormone-receptor-positive breast cancer who were within 1 week–36 months of starting AET, spoke English, and had an Eastern Cooperative Oncology Group performance status ≤2 (i.e., capable of all selfcare). As has been reported previously [25, 31], patients were screened (via adapted National Comprehensive Cancer Network Distress Thermometer items) for AET distress related to symptoms or adherence and were eligible if they reported moderate-severe distress (score of ≥4) on any items [32]. Patients were ineligible if participating in another breast cancer clinical trial, group psychotherapy, were undergoing treatment for another primary cancer, or had uncontrolled psychosis, active suicidal ideation, or cognitive impairment that would interfere with participation.
Procedures
Trained study staff identified potentially eligible patients and received permission to approach the patient. Upon approach, if eligible, the patient provided informed consent, completed a psychosocial assessment, and received a Medication Event Monitoring System (MEMS Caps) pill cap and bottle [33]. Participants were instructed to place their AET medication in the MEMS Caps bottle, which automatically recorded the time/date the bottle was opened as a proxy of medication taken. Participants completed a 1-week trial period of the MEMS Caps prior to randomization to adjust to using the new device and could not view their MEMS Caps data during the study. Study staff connected participants’ MEMS Caps to a reader for data extraction at the end of study participation.
Participants were randomized (unblocked, 1:1) to receive either the STRIDE intervention or a medication monitoring control condition (MedMon). The randomization scheme was generated via random number generator and incorporated stratification by baseline distress level (via Hospital Anxiety and Depression Scale (HADS); high ≥8 vs. low <8, [25, 34]) for balanced representation across groups. The scheme was securely stored and concealed until individual participant randomization. Those randomized to STRIDE received the intervention within the following 12 weeks. This timeframe reasonably allowed for scheduling considerations of participants to complete the six weekly sessions before the next assessment. Approximately 12 weeks post-baseline, all participants completed a psychosocial follow-up assessment. A 24-week follow-up assessment was collected and is not reported here. Because this was a secondary analysis, the sample size was determined from an existing trial and thus findings may be underpowered and should be interpreted as exploratory. As previously published, the sample size was determined to include at least 40 participants per arm who completed study procedures and based on primary outcomes of feasibility and acceptability [25, 31]. Comprehensive study procedures may be found in the study protocol [24].
Group conditions
STRIDE intervention
The STRIDE intervention has been reported in full elsewhere [24, 25, 31]. This intervention took place across six, weekly 1-h sessions in a group of two participants with a clinical psychologist or trained psychology fellow over HIPAA-compliant video conferencing software. Fidelity was monitored weekly. Participants received a program workbook and were expected to practice skills between sessions. Session content was based in cognitive behavioral therapy and motivational interviewing intended to identify and evaluate patients’ thoughts, beliefs, and behaviors related to AET. Skills included weekly relaxation training, cognitive reframing, acceptance strategies, and AET-related side effect management skills. The latter three sessions addressed prominent and impairing AET symptoms.
MedMon condition
All participants randomized to the MedMon condition received standard of care treatment and support from their oncology team for AET-related side effects. Participants were instructed to use the MEMS Caps bottle throughout the active study timeframe.
Measures
Sociodemographic and medical characteristics
Participants reported their age, gender, race, ethnicity, annual income, employment status, and education level. Treatment received for breast cancer, disease stage, and months since AET initiation were collected from the electronic medical record.
Anxiety and depressive symptoms
We used the HADS [34] to assess anxiety and depressive symptoms. Validated for patients with cancer, the HADS consists of 14 items (7 items per subscale) to describe how patients felt over the prior week on a scale of 0 to 3 (subscale range: 0–21) [35]. The HADS subscales had high internal reliability (HADS-Anxiety; α = 0.87; HADS-Depression; α = 0.80).
AET-specific concerns
The Beliefs about Medications Questionnaire–Adjuvant Endocrine Therapy (BMQ-AET) is a 10-item survey that asks patients about agreement with statements about AET across two subscales (concerns and necessity) [36]. Greater scores indicate greater concerns. Only the BMQ-Specific Concerns subscale was included to assess whether concerns about AET were associated with adherence (acceptable reliability; α = 0.708).
AET-related symptom distress
We assessed AET-specific symptom distress using the BreastCancer Prevention Trial Symptom Scale (BCPT) [37]. The BCPT asks patients to rate how much physical and psychological symptoms (e.g., hot flashes, joint pains) related to breast cancer treatment bother them on a scale of 0 (not at all) to 4 (extremely) during the past week. Higher total scores indicate greater AET-related symptom distress. The BCPT had high internal reliability at baseline (α = 0.877).
Objective AET adherence
MEMS Caps data were aggregated across the study timeframe to generate three monthly adherence percentages of daily doses (e.g., within 24 h) taken of all doses prescribed. If participants were prescribed an AET break by a physician within the study timeframe, break days were omitted to reflect doses taken as prescribed. Participants received a paper medication diary as a back-up to the MEMS Caps bottle to record any AET taken independent of MEM Caps. Discrepancies were uncommon and were reconciled at the end of study participation.
Self-reported AET adherence
Participants self-reported AET adherence via the Medication Adherence Report Scale (MARS-5) [38]. The MARS-5 is a 5-item tool to assess medication adherence behaviors, such as “I stop taking my AET medicine for a while.” Patients answer each item on a scale of 1 (Always) to 5 (Never). Notably, internal reliability of the MARS-5 in this study was low (α = 0.546).
Statistical plan
All data were evaluated for normality and outliers. Models controlled for covariates, ovarian suppression (reference: did not receive), and baseline distress (reference: not distressed), due to differences across groups and stratification, respectively. Models were conducted with intention to treat, such that all randomized participants were analyzed. Continuous moderators were tested for effects on objective and self-reported adherence. No baseline characteristics or variables predicted outcome missingness.
Two types of statistical modeling were used to test moderators of adherence response. First, we used hierarchical linear modeling to test moderators of objective adherence trajectories across the first 3 months of the study. A bottom-up modeling building approach evaluated level-one effects (time), level-two effects (covariates, moderators), and cross-level interaction terms (time × condition × moderator). The resulting interaction allowed evaluation of adherence trajectories (e.g., change of adherence overtime as a function of each interaction). Time was tested for random effects and retained if model fit significantly improved via likelihood ratio test (LRT). Three-way interaction terms were probed for simple slopes if the interaction had a p-value of less than 0.15 [39, 40]. The effect of STRIDE on self-reported adherence was tested via multiple regression with a two-way interaction term (condition × moderator).
Post hoc analyses were tested after planned analyses to contextualize results. Within prior literature, patients taking AET have been found to rarely experience only one symptom at a time, thus the moderators in these current analyses may represent a cluster of symptoms experienced by patients taking AET [28]. Therefore, we tested correlations between moderators to aid in the interpretation of results.
Results
Participants were 100 primarily non-Hispanic, White (87%) women with early-stage breast cancer (Stage I: n = 77, 77%). The majority were taking an aromatase inhibitor (n = 60, 60%). Women were on average 56 years of age (SD = 10.94, range = 31–81) and approximately 17.9 months (range = 0.8–34.5) from AET initiation. Eighty-three percent used and returned MEMS Caps pill bottles (Supplemental Fig. 1.). Monthly adherence rates ranged from 88.1% to 91.0%. Fifty-five participants (66.3%) took at least 90% of doses, while 69 (83.1%) took at least 80%. Monthly objective adherence rates did not differ by condition (Table 1). This sample reported an average baseline MARS-5 score of 23.52 (SD = 1.76), which indicated some difficulties taking AET as prescribed and is consistent with prior studies of this population [41]. Self-reported adherence and average objective adherence were correlated (r(92) = 0.48, p < 0.001). As previously reported, session attendance was high (86% completed all sessions) [25]. Participant characteristics may be found in Table 1.
Table 1.
Participant characteristics
| Characteristic (N (%)) | STRIDE intervention (n = 50) | Medication monitoring (n = 50) | Full Sample (N = 100) |
|---|---|---|---|
|
| |||
| Age, years; M (range) | 57.2 (39–81) | 54.9 (31–77) | 56.1 (31–81) |
| Months since AET initiation (AET start to enrollment; M (SD)) | 17.70 (8.87) | 18.13 (8.49) | 17.91 (8.64) |
| Objective adherence | |||
| Month 1 | 93.20% | 88.76% | 91.00% |
| Month 2 | 93.20% | 86.59% | 89.93% |
| Month 3 | 90.99% | 85.10% | 88.08% |
| Gender | |||
| Woman | 50 (100) | 50 (100) | 100 (100) |
| Race | |||
| White | 47 (94) | 44 (88) | 91 (91) |
| Asian | 0 (0) | 4 (8) | 4 (4) |
| Black or African American | 1 (2) | 0 (0) | 1 (1) |
| Other | 1 (2) | 2 (4) | 3 (3) |
| Not reported | 1 (2) | 0 (0) | 1 (1) |
| Ethnicity | |||
| Hispanic or Latino/a | 1 (2) | 2 (4) | 3 (3) |
| Not Hispanic or Latino/a | 47 (94) | 47 (94) | 94 (94) |
| Not reported | 2 (4) | 1 (2) | 3 (3) |
| Education | |||
| 11th grade or less | 0 (0) | 1 (2) | 1 (1) |
| High school graduate/GED | 3 (6) | 5 (10) | 8 (8) |
| Some college/technical school | 9 (18) | 7 (14) | 16 (16) |
| College graduate | 19 (38) | 15 (30) | 34 (34) |
| Master’s degree | 16 (32) | 16 (32) | 32 (32) |
| Advanced professional degree | 3 (6) | 6 (12) | 9 (9) |
| Income | |||
| $25,000-$49,999 | 2 (4) | 5 (10) | 7 (7) |
| $50,000-$99,999 | 11 (22) | 11 (22) | 22 (22) |
| $100,000-$ 149,999 | 9 (18) | 9 (18) | 18 (18) |
| >$150,000 | 27 (54) | 23 (46) | 50 (50) |
| Declined to respond/unknown | 1 (2) | 2 (4) | 3 (3) |
| Employment status | |||
| Full-time/part-time work or student | 33 (66) | 29 (58) | 62 (62) |
| Caring for home or family | 4 (8) | 6 (12) | 10 (10) |
| Unemployed | 1 (2) | 3 (6) | 4 (4) |
| Not working due to illness/disability | 0 (0) | 1 (2) | 1 (1) |
| Retired | 9 (18) | 9 (18) | 18 (18) |
| Other or missing | 3 (6) | 2 (4) | 5 (5) |
| Relationship status | |||
| Married/cohabitating | 38 (76) | 34 (68) | 73 (73) |
| Non-cohabitating relationship | 1 (2) | 2 (4) | 3 (3) |
| Single, never married | 5 (10) | 4 (8) | 9 (9) |
| Divorced/separated | 6 (12) | 7 (14) | 13 (13) |
| Loss of long-term partner/widowed | 0 (0) | 3 (6) | 3 (3) |
| Breast cancer stage | |||
| Stage 0 | 5 (10) | 3 (6) | 8 (8) |
| Stage I | 36 (72) | 41 (82) | 77 (77) |
| Stage II | 6 (12) | 5 (10) | 11 (11) |
| Stage III | 3 (6) | 1 (2) | 4 (4) |
| Type of AET | |||
| Aromatase inhibitor | 28 (56) | 32 (64) | 60 (60) |
| Tamoxifen | 22 (44) | 18 (36) | 40 (40) |
| Primary treatment type | |||
| Surgery only | 10 (20) | 13 (26) | 23 (23) |
| Radiation only | 1 (2) | 0 (0) | 1 (1) |
| Surgery and radiation | 24 (48) | 23 (46) | 47 (47) |
| Surgery and chemotherapy | 2 (4) | 6 (12) | 8 (8) |
| Surgery, chemotherapy, and radiation | 13 (26) | 8 (16) | 21 (21) |
| Ovarian suppression | |||
| Receiving ovarian suppression | 8 (16) | 20 (40) | 28 (28) |
| Not receiving ovarian suppression | 42 (84) | 30 (60) | 72 (72) |
Objective adherence moderators
Model building
The interclass correlation of the intercept-only model was 0.54, and the functional form of adherence was linear (LRT: χ2 = 0.05(1), p = 0.82). Random time effects improved model fit and were retained (LRT: χ2 = 6.97(2), p = 0.03). A time by condition interaction on objective adherence was not significant (B = 0.72, SE = 2.21, p = 0.74). This model was subsequently used to test cross-level interaction effects with the hypothesized moderators, as described below.
Model results
Significant three-way interactions existed for anxiety (B = −1.20, SE = 0.53, p = 0.03; Table 2) and depressive symptoms (B = −1.65, SE = 0.65, p = 0.01), such that the interaction effect of time by condition on adherence differed by levels of each. Individuals with lower anxiety or depressive symptoms who received STRIDE exhibited greater improvements in objective adherence trajectories compared to MedMon or those with greater anxiety (Fig. 2a, b) or depressive symptoms who received STRIDE (Fig. 2c, d). Additionally, a three-way interaction effect was approaching significance in a model testing AET-specific concerns by time by condition (B = 0.91, SE = 0.57, p = 0.12). This effect was probed further and visualized to examine potential patterns within this exploratory analysis at a sub-significant level (Fig. 2e, f). No significant age interaction (B = 0.23, SE = 0.22, p = 0.29) or AET-related symptom distress interaction (B = −0.15, SE = 0.23, p = 0.51) existed for objective adherence trajectories.
Table 2.
Objective adherence moderation models (MEMS Caps, n = 83)
| Parameters | Conditional model: main effects | Interaction model: condition × time | Interaction model: age | Interaction model: anxiety Sx | Interaction model: depressive Sx | Interaction model: symptom Distress | Interaction model: AET concerns |
|---|---|---|---|---|---|---|---|
|
| |||||||
| Regression coefficients (fixed effects) | |||||||
| Intercept | 90.89 (3.89)*** [83.37, 98.43] | 91.52 (4.34)*** [83.12, 99.97] | 93.43 (21.54)*** [52.22, 124.87] | 103.87 (7.99)*** [88.50, 119.20] | 99.12 (6.04)*** [87.52, 110.72] | 87.69 (9.36)*** [69.70, 105.71] | 97.89 (16.31)*** [66.63, 129.31] |
| Time | −1.46 (1.10) [−3.63, 0.70] | −1.83 (1.57) [−4.91, 1.25] | 2.35 (8.36) [−13.82, 18.52] | −6.61 (3.36)+ [−13.12, −0.10] | −3.52 (2.50) [−8.35, 1.32] | −1.31 (4.01) [−9.07, 6.44] | 0.43 (6.73) [−12.58, 13.45] |
| Distress | −2.65 (3.39) [−9.22, 3.91] | −2.65 (3.39) [−9.22, 3.91] | −2.59 (3.53) [−9.34, 4.16] | 0.41 (5.50) [−10.16, 10.97] | −0.39 (3.90) [−7.84, 7.06] | −3.26 (3.58) [−10.09, 3.59] | −3.19 (3.47) [−9.84, 3.46] |
| Ovarian suppression | 0.63 (4.10) [−7.31, 8.57] | 0.63 (4.10) [−7.31, 8.57] | −0.72 (4.67) [−9.67, 8.22] | 1.35 (4.21) [−6.71, 9.41] | 0.82 (4.10) [−7.01, 8.65] | 0.34 (4.17) [−7.62, 8.30] | 0.25 (4.17) [−7.73, 8.23] |
| Condition | 5.71 (3.55)+ [−1.16, 12.59] | 4.46 (5.21) [−5.69, 14.61] | 9.14 (30.30) [−49.17, 67.28] | −15.80 (10.52)+ [−36.01, 4.44] | −14.36 (8.59)+ [−30.88, 2.16] | 1.06 (12.30) [−22.61, 24.73] | 17.93 (20.48) [−21.51, 57.24] |
| Moderator | - | - | −0.03 (0.37) [−0.75, 0.69] | −1.96 (1.14)+ [−4.13, 0.24] | −1.88 (1.02)+ [−3.84, 0.08] | 0.19 (0.39) [−0.56, 0.95] | −0.41 (1.05) [−2.44, 1.62] |
| Condition × time | - | 0.72 (2.21) [−3.60, 5.05] | −12.47 (12.57) [−36.78, 11.84] | 9.79 (4.53)* [1.03, 18.55] | 8.25 (3.68)* [1.14, 15.36] | 3.73 (5.32) [−6.56, 14.02] | −12.54 (8.63)+ [−29.23, 4.16] |
| Condition × moderator | - | - | −0.09 (0.53) [−1.10, 0.93] | 2.75 (1.24)* [0.36, 5.12] | 4.10 (1.51)** [1.20, 7.00] | 0.18 (0.53) [−0.84, 1.21] | −0.94 (1.35) [−3.54, 1.66] |
| Time × moderator | - | - | −0.08 (0.15) [−0.37, 0.22] | 0.66 (0.41)+ [−0.14, 1.46] | 0.36 (0.43) [−0.46, 1.19] | −0.02 (0.17) [−0.35, 0.30] | −0.15 (0.45) [−1.02, 0.71] |
| Condition × time × moderator | - | - | 0.23 (0.22) [−0.19, 0.66] | −1.20 (0.53)* [−2.23, −0.18] | −1.65 (0.65)* [−2.92, −0.39] | −0.15 (0.23) [−0.60, 0.29] | 0.91 (0.57)+ [−0.19, 2.02] |
| Variance components (random effects) | |||||||
| Residual | 124.89 | 124.89 | 124.89 | 124.89 | 124.89 | 124.89 | 124.89 |
| Intercept | 243.87 | 247.15 | 260.94 | 225.59 | 212.29 | 252.44 | 243.07 |
| Time | 37.59 | 38.69 | 39.68 | 34.81 | 32.23 | 39.58 | 35.53 |
MEMS Caps Medication Event Monitoring System Caps, Sx symptoms
95% Confidence Interval depicted in brackets beneath each estimated parameter
= p < 0.001;
= p < 0.01;
= p < 0.05;
= p < 0.15
Fig. 2.
Time × Condition × Moderator effects on the trajectory of adherence across the first 3 months of the study on the MEMS Caps at high and low levels of each moderator. Note: Graphs illustrate that patients who reported lower anxiety symptoms, depressive symptoms, and fewer AET-specific concerns at baseline were randomly assigned to receive the STRIDE intervention and exhibited improvements in objective adherence across the 3 months. Moderators are split at the median of each variable (e.g., higher versus lower) for visualization of the effects found; however, tested models included continuous moderators
Self-reported adherence moderators
No main effects of condition on self-reported adherence were found (B = 0.23, SE = 0.18, p = 0.21). A two-way (condition × age) interaction existed (B = 0.05, SE = 0.02, p = 0.003), such that older individuals who received the STRIDE condition reported greater self-reported adherence at 12 weeks post-baseline. A significant AET-related symptom distress by condition interaction existed (B = −0.04, SE = 0.02, p = 0.02). Individuals with lower reported AET-related symptom distress at baseline who received STRIDE exhibited the greatest increase in self-reported adherence versus those with greater AET-related symptom distress or those who received MedMon. Anxiety symptoms (B = 0.00, SE = 0.04, p = 0.99), depressive symptoms (B = 0.05, SE = 0.05, p = 0.31), or AET-specific concerns (B = −0.04, SE = 0.05, p = 0.37) did not interact with condition on self-reported adherence. Table 3 contains all model estimates.
Table 3.
Self-reported adherence moderation models (MARS-5, N = 100)
| Parameter | Main effects model: condition | Interaction model: age | Interaction model: anxiety Sx | Interaction model: depressive Sx | Interaction model: symptom distress | Interaction model: AET concerns |
|---|---|---|---|---|---|---|
|
| ||||||
| Intercept | 23.78 (0.17)*** [23.44, 24.12] | 24.38 (0.74)*** [22.92, 25.84] | 23.92 (0.28)*** [23.36, 24.48] | 23.99 (0.22)*** [23.56, 24.43] | 23.53 (0.31)*** [22.92, 24.15] | 23.08 (0.54)*** [22.02, 24.14] |
| Distress level | −0.07 (0.17) [−0.42, 0.27] | 0.05 (0.18) [−0.30, 0.40] | 0.21 (0.29) [−0.36, 0.79] | 0.06 (0.21) [−0.35, 0.47] | 0.01 (0.18) [−0.35, 0.37] | −0.04 (0.18) [−0.38, 0.31] |
| Ovarian suppression | 0.24 (0.20) [−0.16, 0.64] | 0.21 (0.21) [−0.21, 0.64] | 0.28 (0.21) [−0.13, 0.68] | 0.25 (0.21) [−0.16, 0.65] | 0.26 (0.20) [−0.14, 0.66] | 0.28 (0.20) [−0.12, 0.68] |
| Condition | 0.23 (0.18) [−0.13, 0.58] | −2.73 (0.99)** [−4.68, −0.77] | 0.25 (0.37) [−0.47, 0.97] | −0.03 (0.31) [−0.64, 0.58] | 1.10 (0.42)** [0.26, 1.93] | 0.82 (0.68) [−0.52, 2.15] |
| Moderator | - | −0.01 (0.01) [−0.04, 0.01] | −0.04 (0.05) [−0.13, 0.05] | −0.06 (0.04) [−0.14, 0.01] | 0.01 (0.01) [−0.02, 0.03] | 0.05 (0.03) [−0.02, 0.11] |
| Condition × moderator | - | 0.05 (0.02)** [0.02, 0.09] | 0.00 (0.04) [−0.08, 0.08] | 0.05 (0.05) [−0.05, 0.16] | −0.04 (0.02)* [0.08, −0.01] | −0.04 (0.05) [−0.13, 0.05] |
MARS-5 Medication Adherence Report Scale—Five Item, Sx symptoms
= p < 0.001;
= p < 0.01;
= p < 0.05
Post-hoc analysis: moderator correlations
Older age was associated with lower anxiety and depressive symptoms and greater AET-related symptom distress and more AET-specific concerns. Additionally, more AET-specific concerns were associated with lower anxiety and depressive symptoms and AET-related symptom distress. Baseline depressive symptoms were positively correlated with anxiety and AET-related symptom distress (Supplemental Table 1).
Discussion
The current study tested exploratory hypotheses of the effects of a virtual, STRIDE on self-reported and objective adherence. Several moderators of adherence were found, including age, AET-symptom distress, and anxiety and depressive symptoms. Neither baseline distress level nor ovarian suppression receipt were found to be significant covariates. Counter to hypotheses, participants presenting with greater distress or symptom profiles who received the STRIDE intervention appeared to decline in adherence (Fig. 2a–f), which may reflect null intervention effects and normative reductions in adherence overtime [3]. Participants presenting with lower distress or symptom profiles who received STRIDE exhibited improved adherence. Most correlations mirrored the directionality of subgroups who conferred the most adherence benefit from STRIDE. For instance, participants who presented with lower anxiety symptoms and those who were older each exhibited greater adherence after STRIDE. When assessed post hoc, these moderators, age and anxiety, exhibited a negative association. Of note, this was not consistent across all moderators, and two associations exhibited an opposite effect (i.e., age and AET-specific concerns; depressive symptoms and AET-related symptom distress).
The current study provides evidence that a behavioral intervention improves short-term adherence among subgroups of patients presenting with nonmetastatic breast cancer at an academic medical center and offers insights into necessary intervention improvement. While STRIDE participants have previously been found to benefit psychosocially, there were no main effects on self-reported or objective adherence within the entire sample [25]. The present study offers evidence that a set of factors, or profile, of less symptomatic participants who received this intervention may more readily respond to the treatment while others with greater symptoms or psychiatric concerns may respond differently. Effects may be the result of distress or symptom burden overwhelming a patient’s availability to fully respond to STRIDE while contending with other demands. Self-reported and objective adherence measures are often relied upon individually, though previous literature has found discrepancies between self-reported adherence versus objective AET adherence metrics [21, 42]. These findings across self-reported and objective adherence support the use of both types of adherence outcomes to detect effects of a behavioral intervention on adherence.
Several interfering barriers to AET adherence exist. Depressive symptoms are one such concern and is a primary barrier to AET adherence and to medication adherence across other medical populations [13, 43–45]. Less is known about the role of adherence-interfering barriers, like depression, on the uptake of a behavioral intervention targeting adherence. To evaluate and address this gap, behavioral interventions may be tailored to simultaneously address both presenting complaints (e.g., psychiatric concerns, age-related challenges) and AET adherence. Prior evidence for this approach exists within studies of patients with HIV. Specifically, in trials testing a cognitive behavioral intervention for simultaneously increasing adherence to anti-retroviral therapy and reducing depressive symptoms, participants who received cognitive behavioral therapy for adherence and depression (CBT-AD) both increased medication adherence and reduced depressive symptoms by post-intervention [46–48]. While the STRIDE intervention was informed by CBT-AD techniques, additional modifications should be considered to specifically target the presenting concerns of participants, including providing proactive referrals to support for distressed patients prior to AET initiation, offering a stepped-care model, and/or providing specific modular content for concerns. Clinically, providers who work with patients taking AET may benefit from approaching efforts to boost adherence holistically. Efforts may include screening for and addressing distress, symptom burden, or treatment-related concerns when specifically working with patients to optimize AET adherence.
Future interventions to improve adherence to AET should recognize and acknowledge interpersonal and structural factors that impact patient outcomes. In the present study, the model of determinants of AET adherence solely considered individual-level factors; however, competing health behavior models (i.e., the Social Ecological Model, the World Health Organization medication adherence model) offer evidence that interpersonal (e.g., social support, relationship with care team), socioeconomic (e.g., insurance coverage, financial burden), and structural (e.g., pharmacy distance) factors are also crucial aspects of treatment adherence (Supplemental Fig. 2) [49]. In prior studies, these multi-level factors have been associated with health outcomes, including morbidity, mortality, breast cancer stage at diagnosis, and adherence to mammogram screenings [50–52]. Additionally, other cognitive behavioral trials have shown differential treatment response as a function of patient factors [53, 54]. Modifications to address multi-level factors should include strategies to increase intervention access (e.g., flexible recruitment strategies outside of hospital settings, providing technology) and adaptations of content to incorporate culturally appropriate tailoring, as is on-going in existing trials [55, 56].
Limitations of the current study include that the present sample was predominately White, non-Hispanic, partnered, and well educated with overall high objective AET adherence (88–91%) over a variable amount of time since initiation. As such, these findings should be taken with caution while generalizing to other populations outside of this urban, academic medical context or at different times during treatment (e.g., AET initiation, multiple years into AET regimen). Additionally, self-reported adherence captured through the MARS-5 relied on adherence perceptions. The MARS-5 exhibited low internal reliability and may reflect the high variability of non-adherence-related behaviors captured as a single construct. Further, poor internal reliability of this self-reported adherence measure points to the current gap that exists for reliable and valid self-reported adherence measures for this population. The MEMS Caps may offer greater accuracy of findings via objective adherence models. Other forms of adherence assessment, such as ecological momentary assessment, may also boost reliability of such metrics. Lastly, this study coincided with the start of the global COVID-19 pandemic, which could have altered the internal validity of study procedures to adjust to public health measures and impacted patients’ willingness to participate or use/return MEMS Caps and bottles.
Conclusions
This study revealed moderated treatment findings of a virtual, behavioral intervention for AET adherence. Those presenting with less severe symptom profiles, including lower anxiety and depressive symptoms, and AET-related symptom distress, as well as older patients, had improved AET adherence after STRIDE. These findings point to the importance of tailoring interventions to specific patient subgroups and the need to incorporate treatment for non-AET-specific concerns, such as psychiatric complaints. The current sample was highly educated and predominantly White/non-Hispanic which limits generalizability of these findings. Further research should explore and address the role of multi-level factors that impact adherence and uptake of treatment, including interpersonal and structural barriers.
Supplementary Material
Funding
This work was supported by the National Cancer Institute (#K07CA211107, PI: Jacobs).
Footnotes
Conflict of interest Ms. Emily A. Walsh, Dr. Kathryn Post, Ms. Katina Massad, Ms. Nora Horick, Dr. Ann H. Partridge, Dr. Elyse R. Park, and Dr. Jennifer S. Temel: no conflicts of interest or financial interests. Dr. Michael H. Antoni: paid consultant for Blue Note Therapeutics and Atlantis Healthcare. Co-inventor of cognitive behavioral stress management, filed as IP with the University of Miami (UMIP-483), which is licensed to Blue Note Therapeutics. Dr. Frank J. Penedo: paid consultant of Blue Note Therapeutics, received payment from the University of Texas San Antonio Health Sciences Center. Dr. Steven A. Safren: Royalties from books with Oxford University Press, Guilford Publications, and Springer/Humana Press. Dr. Jeffrey Peppercorn: Financial compensation from Abbott Labs and reports spousal employment at GlaxoSmithKline. Dr. Joseph A. Greer: Paid consultant for BeiGene, research funding from Blue Note Therapeutics, royalties for an edited book from Oxford University Press. Dr. Jamie M. Jacobs: Prior financial interest in VivorCare, Inc. (reviewed and managed by MGH and Mass General Brigham), consultant for VivorCare, Inc.
Supplementary Information The online version contains supplementary material available at https://doi.org/10.1007/s10549-023-07228-z.
Materials availability Available upon reasonable request from the authors.
Code availability Available upon reasonable request from the authors.
Ethical Approval All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards.
Informed consent Informed consent was obtained from all participants included in the study.
Data availability
Available upon reasonable request from the authors.
References
- 1.Siegel RL, Miller KD, Wagle NS, Jemal A (2023) Cancer statistics, 2023. CA Cancer J Clin 73(1):17–48. 10.3322/CAAC.21763 [DOI] [PubMed] [Google Scholar]
- 2.Burstein HJ, Lacchetti C, Anderson H et al. (2019) Adjuvant endocrine therapy for women with hormone receptor–positive breast cancer: ASCO clinical practice guideline focused update. J Clin Oncol 37(5):423–438. 10.1200/JCO.18.01160 [DOI] [PubMed] [Google Scholar]
- 3.Murphy CC, Bartholomew LK, Carpentier MY, Bluethmann SM, Vernon SW (2012) Adherence to adjuvant hormonal therapy among breast cancer survivors in clinical practice: a systematic review. Breast Cancer Res Treat 134(2):459–478. 10.1007/S10549-012-2114-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Jacobs JM, Walsh EA, Park ER et al. (2020) The patient’s voice: adherence, symptoms, and distress related to adjuvant endocrine therapy after breast cancer. Int J Behav Med 27(6):687–697. 10.1007/s12529-020-09908-2 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Jacob Arriola KR, Mason TA, Bannon KA et al. (2014) Modifiable risk factors for adherence to adjuvant endocrine therapy among breast cancer patients. Patient Educ Couns 95(1):98–103. 10.1016/J.PEC.2013.12.019 [DOI] [PubMed] [Google Scholar]
- 6.Stanton AL, Petrie KJ, Partridge AH (2014) Contributors to non-adherence and nonpersistence with endocrine therapy in breast cancer survivors recruited from an online research registry. Breast Cancer Res Treat 145(2):525–534. 10.1007/S10549-014-2961-3 [DOI] [PubMed] [Google Scholar]
- 7.Vaz-Luis I, Francis PA, Di Meglio A, Stearns V (2021) Challenges in adjuvant therapy for premenopausal women diagnosed with luminal breast cancers. Am Soc Clin Oncol Educ Book 41(41):e47–e61. 10.1200/EDBK_320595 [DOI] [PubMed] [Google Scholar]
- 8.Walsh EA, Chabria R, Vranceanu AM et al. (2022) Understanding pain related to adjuvant endocrine therapy after breast cancer: a qualitative report. Eur J Cancer Care (Engl) 31(6):e13723. 10.1111/ECC.13723 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Hadji P (2008) Menopausal symptoms and adjuvant therapy-associated adverse events. Endocr Relat Cancer 15(1):73–90. 10.1677/ERC-07-0193 [DOI] [PubMed] [Google Scholar]
- 10.Cheung YM, Ramchand SK, Yeo B, Grossmann M (2019) Cardiometabolic effects of endocrine treatment of estrogen receptor-positive early breast cancer. J Endocr Soc 3(7):1283. 10.1210/JS.2019-00096 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Lee JY, Min YH (2018) Relationships between determinants of adjuvant endocrine therapy adherence in breast cancer. BMC Womens Health 18(1):48. 10.1186/S12905-018-0522-3 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Bright EE, Petrie KJ, Partridge AH, Stanton AL (2016) Barriers to and facilitative processes of endocrine therapy adherence among women with breast cancer. Breast Cancer Res Treat 158(2):243–251. 10.1007/s10549-016-3871-3 [DOI] [PubMed] [Google Scholar]
- 13.Mausbach BT, Schwab RB, Irwin SA (2015) Depression as a predictor of adherence to adjuvant endocrine therapy (AET) in women with breast cancer: a systematic review and meta-analysis. Breast Cancer Res Treat 152(2):239–246. 10.1007/S10549-015-3471-7 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Walker HE, Rosenberg SM, Stanton AL, Petrie KJ, Partridge AH (2016) Perceptions, attributions, and emotions toward endocrine therapy in young women with breast cancer. J Adolesc Young Adult Oncol 5(1):16–23. 10.1089/JAYAO.2015.0051 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Hershman DL, Shao T, Kushi LH et al. (2011) Early discontinuation and non-adherence to adjuvant hormonal therapy are associated with increased mortality in women with breast cancer. Breast Cancer Res Treat 126(2):529–537. 10.1007/S10549-010-1132-4/FIGURES/2 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Finitsis DJ, Vose BA, Mahalak JG, Salner AL (2019) Interventions to promote adherence to endocrine therapy among breast cancer survivors: a meta-analysis. Psychooncology 28(2):255–263. 10.1002/PON.4959 [DOI] [PubMed] [Google Scholar]
- 17.Ekinci E, Nathoo S, Korattyil T et al. (2018) Interventions to improve endocrine therapy adherence in breast cancer survivors: what is the evidence? J Cancer Surviv 12(3):348–356. 10.1007/S11764-017-0674-4 [DOI] [PubMed] [Google Scholar]
- 18.Shelby RA, Dorfman CS, Bosworth HB et al. (2019) Testing a behavioral intervention to improve adherence to adjuvant endocrine therapy (AET). Contemp Clin Trials 76:120–131. 10.1016/J.CCT.2018.11.010 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Smith SG, Green SMC, Ellison R et al. (2023) Refining and optimising a behavioural intervention to support endocrine therapy adherence (ROSETA) in UK women with breast cancer: protocol for a pilot fractional factorial trial. BMJ Open 13(2):e069971. 10.1136/BMJOPEN-2022-069971 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Neven P, Markopoulos C, Tanner M et al. (2014) The impact of educational materials on compliance and persistence rates with adjuvant aromatase inhibitor treatment: first-year results from the Compliance of ARomatase Inhibitors AssessmenT In Daily practice through Educational approach (CARIATIDE) study. Breast 23(4):393–399. 10.1016/J.BREAST.2014.02.009 [DOI] [PubMed] [Google Scholar]
- 21.Ziller V, Kyvernitakis I, Knöll D, Storch A, Hars O, Hadji P (2013) Influence of a patient information program on adherence and persistence with an aromatase inhibitor in breast cancer treatment—the COMPAS study. BMC Cancer 13:407. 10.1186/1471-2407-13-407 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Hadji P, Blettner M, Harbeck N et al. (2013) The Patient’s Anastrozole Compliance to Therapy (PACT) Program: a randomized, in-practice study on the impact of a standardized information program on persistence and compliance to adjuvant endocrine therapy in postmenopausal women with early breast cancer. Ann Oncol 24(6):1505–1512. 10.1093/ANNONC/MDS653 [DOI] [PubMed] [Google Scholar]
- 23.Janssen AM, Dam J, Prins J, Buffart LM, de Bruin M (2022) Systematic adaptation of the adherence improving self-management strategy to support breast cancer survivors’ adherence to adjuvant endocrine therapy: an intervention mapping approach. Eur J Cancer Care (Engl) 31(6):e13721. 10.1111/ECC.13721 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Jacobs JM, Rapoport CS, Horenstein A et al. (2021) Study protocol for a randomised controlled feasibility trial of a virtual intervention (STRIDE) for symptom management, distress and adherence to adjuvant endocrine therapy after breast cancer. BMJ Open 11(1):e041626. 10.1136/BMJOPEN-2020-041626 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Jacobs JM, Post K, Massad K et al. (2022) A telehealth intervention for symptom management, distress, and adherence to adjuvant endocrine therapy: a randomized controlled trial. Cancer 128(19):3541–3551. 10.1002/CNCR.3440926 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Greer JA, Jacobs JM, Pensak N et al. (2020) Randomized trial of a smartphone mobile app to improve symptoms and adherence to oral therapy for cancer. J Natl Compr Canc Netw 18(2):133–141. 10.6004/JNCCN.2019.7354 [DOI] [PubMed] [Google Scholar]
- 27.Knobf MT (2006) The influence of endocrine effects of adjuvant therapy on quality of life outcomes in younger breast cancer survivors. Oncologist 11(2):96–110. 10.1634/THEONCOLOGIST.11-2-96 [DOI] [PubMed] [Google Scholar]
- 28.Jing F, Zhu Z, Qiu J, Tang L, Xu L, Xing W (2023) Symptom profiles and related factors among breast cancer patients undergoing endocrine therapy: a latent profile analysis. Cancer Nurs 46(5):E297–E304. 10.1097/NCC.0000000000001125 [DOI] [PubMed] [Google Scholar]
- 29.Camacho FT, Tan X, Alcalá HE, Shah S, Anderson RT, Balkrishnan R (2017) Impact of patient race and geographical factors on initiation and adherence to adjuvant endocrine therapy in Medicare breast cancer survivors. Medicine (Baltimore) 96(24):e7147. 10.1097/MD.0000000000007147 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Haji-Hersi MF, Tilley S, Shelton CA, Lamb N, Kamdem LK (2022) Drug- and patient-related factors are the strongest predictors of endocrine therapy adherence in breast cancer patients. J Oncol Pharm Pract 28(5):1070–1076. 10.1177/10781552211020805 [DOI] [PubMed] [Google Scholar]
- 31.Jacobs JM, Walsh EA, Rapoport CS et al. (2021) Development and refinement of a telehealth intervention for symptom management, distress, and adherence to adjuvant endocrine therapy after breast cancer. J Clin Psychol Med Settings 28(3):603–618. 10.1007/S10880-020-09750-4 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Dabrowski M, Boucher K, Ward JH et al. (2007) Clinical experience with the NCCN Distress Thermometer in breast cancer patients. J Natl Compr Cancer Netw 5(1):104–111. 10.6004/jnccn.2007.0011 [DOI] [PubMed] [Google Scholar]
- 33.Farmer KC (1999) Methods for measuring and monitoring medication regimen adherence in clinical trials and clinical practice. Clin Ther 21(6):1074–1090. 10.1016/S0149-2918(99)80026-5 [DOI] [PubMed] [Google Scholar]
- 34.Zigmond A, Snaith R (1983) The hospital and anxiety and depression scale. Acta Psychiatr Scand 67(6):361–370. 10.1111/j.1600-0447.1983.tb09716.x [DOI] [PubMed] [Google Scholar]
- 35.Bjelland I, Dahl AA, Haug TT, Neckelmann D (2002) The validity of the Hospital Anxiety and Depression Scale: an updated literature review. J Psychosom Res 52(2):69–77. 10.1016/S0022-3999(01)00296-3 [DOI] [PubMed] [Google Scholar]
- 36.Brett J, Hulbert-Williams NJ, Fenlon D et al. (2017) Psychometric properties of the Beliefs about Medicine Questionnaire–adjuvant endocrine therapy (BMQ-AET) for women taking AETs following early-stage breast cancer. Health Psychol Open 4(2):1–8. 10.1177/2055102917740469 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Stanton AL, Bernaards CA, Ganz PA (2005) The BCPT symptom scales: a measure of physical symptoms for women diagnosed with or at risk for breast cancer. J Natl Cancer Inst 97(6):448–456. 10.1093/jnci/dji069 [DOI] [PubMed] [Google Scholar]
- 38.Chan AHY, Horne R, Hankins M, Chisari C (2020) The Medication Adherence Report Scale: a measurement tool for eliciting patients’ reports of nonadherence. Br J Clin Pharmacol 86(7):1281–1288. 10.1111/bcp.14193 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Marshall SW (2007) Power for tests of interaction: effect of raising the Type I error rate. Epidemiol Perspect Innov 4:4. 10.1186/1742-5573-4-4 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40.Wang R, Ware JH (2013) Detecting moderator effects using subgroup analyses. Prev Sci 14(2):111–120. 10.1007/S11121-011-0221-X [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41.Zhao M, Zhao J, Chen J et al. (2022) The relationship between medication adherence and illness perception in breast cancer patients with adjuvant endocrine therapy: beliefs about medicines as mediators. Support Care Cancer 30(12):10009–10017. 10.1007/S00520-022-07411-W [DOI] [PubMed] [Google Scholar]
- 42.Waterhouse DM, Calzone KA, Mele C, Brenner DE (1993) Adherence to oral tamoxifen: a comparison of patient self-report, pill counts, and microelectronic monitoring. J Clin Oncol 11(6):1189–1197. 10.1200/JCO.1993.11.6.1189 [DOI] [PubMed] [Google Scholar]
- 43.Gonzalez JS, Batchelder AW, Psaros C, Safren SA (2011) Depression and HIV/AIDS treatment nonadherence: a review and meta-analysis. J Acquir Immune Defic Syndr 58(2):181–187. 10.1097/QAI.0B013E31822D490A [DOI] [PMC free article] [PubMed] [Google Scholar]
- 44.Hamieh N, Kab S, Zins M et al. (2021) Depressive symptoms and non-adherence to treatable cardiovascular risk factors’ medications in the CONSTANCES cohort. Eur Heart J Cardiovasc Pharmacother 7(4):280–286. 10.1093/EHJCVP/PVAA124 [DOI] [PubMed] [Google Scholar]
- 45.Semahegn A, Torpey K, Manu A, Assefa N, Tesfaye G, Ankomah A (2020) Psychotropic medication non-adherence and its associated factors among patients with major psychiatric disorders: a systematic review and meta-analysis. Syst Rev 9(1):17. 10.1186/S13643-020-1274-3 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 46.Safren SA, O’Cleirigh C, Tan JY et al. (2009) A randomized controlled trial of cognitive behavioral therapy for adherence and depression (CBT-AD) in HIV-infected individuals. Health Psychol 28(1):1–10. 10.1037/A0012715 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 47.Safren SA, Bedoya CA, O’Cleirigh C et al. (2016) Cognitive behavioural therapy for adherence and depression in patients with HIV: a three-arm randomised controlled trial. Lancet HIV 3(11):e529–e538. 10.1016/S2352-3018(16)30053-4 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 48.Simoni JM, Wiebe JS, Sauceda JA et al. (2013) A preliminary RCT of CBT-AD for adherence and depression among HIV-positive Latinos on the U.S.-Mexico border: the Nuevo Día study. AIDS Behav 17(8):2816–2829. 10.1007/S10461-013-0538-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 49.Burkhart PV, Sabaté E (2003) Adherence to long-term therapies: evidence for action. J Nurs Scholarsh 35(3):207. 10.1111/J.1547-5069.2003.TB00001.X [DOI] [PubMed] [Google Scholar]
- 50.Thompson B, Hohl SD, Molina Y et al. (2018) Breast cancer disparities among women in underserved communities in the USA. Curr Breast Cancer Rep 10(3):131–141. 10.1007/S12609-018-0277-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 51.Henderson V, Tossas-Milligan K, Martinez E et al. (2020) Implementation of an integrated framework for a breast cancer screening and navigation program for women from underresourced communities. Cancer 126(Suppl 10):2481–2493. 10.1002/CNCR.32843 [DOI] [PubMed] [Google Scholar]
- 52.Coughlin SS (2019) Social determinants of breast cancer risk, stage, and survival. Breast Cancer Res Treat 177(3):537–548. 10.1007/S10549-019-05340-7 [DOI] [PubMed] [Google Scholar]
- 53.Bouchard LC, Yanez B, Dahn JR et al. (2019) Brief report of a tablet-delivered psychosocial intervention for men with advanced prostate cancer: acceptability and efficacy by race. Transl Behav Med 9(4):629–637. 10.1093/TBM/IBY089 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 54.Walsh EA, Boland CL, Popok PJ et al. (2023) Marital status and perceived stress in men with advanced prostate cancer: a randomized-controlled trial of cognitive behavioral stress management. J Psychosom Res 167:111198. 10.1016/J.JPSYCHORES.2023.111198 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 55.Penedo FJ, Antoni MH, Moreno PI et al. (2018) Study design and protocol for a culturally adapted cognitive behavioral stress and self-management intervention for localized prostate cancer: the Encuentros de Salud study. Contemp Clin Trials 71:173–180. 10.1016/j.cct.2018.06.010 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 56.McKinney NS, Virtue S, Lewis FM et al. (2018) Study protocol: a randomized control trial of African American families fighting parental cancer together. BMC Cancer 18(1):1140. 10.1186/s12885-018-5052-8 [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.
Supplementary Materials
Data Availability Statement
Available upon reasonable request from the authors.