Shared decision-making and patient control in radiation oncology: Implications for patient satisfaction

Authors

  • Jacob E. Shabason MD,

    1. Department of Radiation Oncology, University of Pennsylvania Health System, Philadelphia, Pennsylvania
    Search for more papers by this author
  • Jun J. Mao MD, MSCE,

    1. Abramson Cancer Center, University of Pennsylvania Health System, Philadelphia, Pennsylvania
    2. Department of Family Medicine and Community Health, University of Pennsylvania Health System, Philadelphia, Pennsylvania
    3. Center for Clinical Epidemiology and Biostatistics, University of Pennsylvania, Philadelphia, Pennsylvania
    Search for more papers by this author
  • Eitan S. Frankel BA,

    1. Department of Family Medicine and Community Health, University of Pennsylvania Health System, Philadelphia, Pennsylvania
    Search for more papers by this author
  • Neha Vapiwala MD

    Corresponding author
    1. Department of Radiation Oncology, University of Pennsylvania Health System, Philadelphia, Pennsylvania
    2. Abramson Cancer Center, University of Pennsylvania Health System, Philadelphia, Pennsylvania
    • Corresponding author: Neha Vapiwala, MD, Department of Radiation Oncology, University of Pennsylvania, 3400 Civic Center Blvd, 4th Fl West Pavilion, Philadelphia, Pennsylvania 19104; Fax: (215) 349-8975; vapiwala@uphs.upenn.edu

    Search for more papers by this author

Abstract

BACKGROUND

Shared decision-making (SDM) has been linked to important health care quality outcomes. However, to the authors' knowledge, the value of SDM has not been thoroughly evaluated in the field of radiation oncology. The objective of the current study was to determine the association between SDM and patient satisfaction during radiotherapy (RT). The authors also explored patient desire for and perception of control during RT, and how these factors relate to patient satisfaction, anxiety, depression, and fatigue.

METHODS

A cross-sectional survey of 305 patients undergoing definitive RT was conducted. Patients self-reported measured variables during the last week of RT. Relationships between variables were evaluated using chi-square analyses.

RESULTS

Among study participants, 31.3% of patients experienced SDM, 32.3% perceived control in treatment decisions, and 76.2% reported feeling very satisfied with their care. Patient satisfaction was associated with perceived SDM (84.4% vs 71.4%; P < .02) and patient-perceived control (89.7% vs 69.2%; P < .001). Furthermore, the perception of having control in treatment decisions was associated with increased satisfaction regardless of whether the patient desired control. Increased anxiety (44.0% vs 20.0%; P < .02), depression (44.0% vs 15.0%; P < .01), and fatigue (68.0% vs 32.9%; P < .01) were reported in patients who desired but did not perceive control over their treatments, compared with those who both desired and perceived control.

CONCLUSIONS

The findings of the current study emphasize the value of SDM and patient-perceived control during RT, particularly as it relates to patient satisfaction and psychological distress. Regardless of a patient's desire for control, it is important to engage patients in the decision-making process. Cancer 2014;120:1863–1870. © 2014 American Cancer Society.

INTRODUCTION

In 2001, the Institute of Medicine identified that patient-centered care is a key component of quality health care delivery. The committee defined patient-centeredness as an approach that involves “providing care that is respectful of and responsive to individual patient preferences, needs, and values and ensuring that patient values guide all clinical decisions.”[1]

The shared decision-making (SDM) process is vital for patient-centered care and for establishing a meaningful and therapeutic physician-patient relationship. In medicine, and particularly in oncology, SDM has been associated with improved outcomes related to patient satisfaction, anxiety, and quality of life.[2-5] Although still relying heavily on physician expertise, SDM is a significant departure from the paternalism traditionally associated with the physician-patient relationship. SDM is a process in which both a physician and a patient share information about a disease, discuss its actual and potential effects on the patient, review the medically appropriate treatment options, and then ultimately reach a consensus together regarding the most appropriate treatment approach for that specific patient.[6-8]

The national importance of SDM is further evident in the finding that the Patient Protection and Affordable Care Act devotes an entire section to establishing a program for SDM. The program's goal is to establish “collaborative processes between patients, caregivers or authorized representatives, and clinicians that engage the patient, caregiver or authorized representative in decision making, provides patients, caregivers or authorized representatives with information about trade-offs among treatment options, and facilitates the incorporation of patient preferences and values into the medical plan.”[9] Furthermore, in a recently published report from the Institute of Medicine, the authors emphasize the importance of patient-centered communication and SDM as a key component to delivering high-quality cancer care. Unfortunately, SDM in oncology is currently suboptimal.[10] Over the past few decades, there have been several studies examining how a patient's role in the decision-making process relates to his or her oncologic treatment. Studies report that patients experience SDM in 9% to 44% of cases.[2, 11, 12]

Furthermore, as medical, surgical, and radiation treatment options become more complex and varied, it will become increasingly important to keep cancer patients fully engaged in the discussions weighing the risks and benefits of all potential treatment modalities. In particular, radiation oncology is a field that is largely unfamiliar to most patients. Thus, there is a significant need to improve patient education about radiotherapy (RT) options and about what to expect during and after RT,[13] a need that can be at least partly addressed through SDM.

SDM has been correlated with important clinical outcomes such as patient satisfaction. Specifically within radiation oncology, there is evidence that enhanced information sharing[14] and decision aid tools[15] lead to enhanced patient satisfaction. However, the endpoints in these studies were satisfaction during the first week of RT[14] or satisfaction with information shared.[15] To our knowledge, there are no data correlating patient-perceived SDM with overall satisfaction with RT.

As such, we sought to assess the prevalence of SDM and the perception of control in treatment decisions among patients receiving RT at a large, urban, academic medical center. We then examined the association between SDM and patient satisfaction during each patient's last week of RT. We also explored the relationship between a patient's desire for and perception of control during RT and satisfaction, anxiety, depression, and fatigue. Findings from the current study should help to guide future changes to the physician-patient interaction in radiation oncology, with the goal of improving patient-centered care.

MATERIALS AND METHODS

Study Population

We conducted a cross-sectional survey study in the Department of Radiation Oncology at the Hospital of the University of Pennsylvania in Philadelphia. Potential participants included patients aged ≥ 18 years who were undergoing RT for a diagnosis of cancer with a Karnofsky performance status of ≥ 60. Patients were excluded if they were receiving palliative RT, had a known brain tumor or abnormal neurologic function, or were unable to understand the requirements of the study. Research assistants obtained permission from the treating radiation oncologists, screened medical records, and approached potential participants about the study during each patient's last week of RT. The survey was distributed during the final week of treatment to ensure that all aspects of the RT process and all the opportunities that existed for SDM throughout a patient's RT course were captured in the survey response. Once informed consent was obtained, each participant was given a self-administered survey. All protocols and surveys were reviewed and approved by the University of Pennsylvania Institutional Review Board and the Abramson Cancer Center Scientific Review and Monitoring Committee.

Variables

We measured patient perception of his or her radiation oncologist's SDM style using a previously used 3-item scale.[16-18] As described in Kaplan et al,[16] patients were asked to rate their physicians' participatory decision style based on a 5-point scale, using the following questions: 1) “If there were a choice between treatments, would your radiation oncologist ask you to help him/her make the decision? (definitely yes to definitely no)”; 2) “How often does your radiation oncologist make an effort to give you some control over your treatment? (very often to never)”; and 3) “How often does your radiation oncologist ask you to take some of the responsibility for your treatment? (very often to not at all).” We tested the reliability of this questionnaire by calculating a Cronbach alpha of these 3 items in our population, which was .7, indicating adequate reliability. Based on the responses, a score ranging from 0 to 100 was calculated by summing the responses, dividing by 15 and then multiplying by 100. An SDM score was then dichotomized into a “yes” or “no” variable using a score of ≥ 70 as a cutoff.

We developed the questions used to evaluate perceived and desired control, satisfaction, fatigue, depression, and anxiety. During the instrument pilot phase, some patients informed us that they did not desire control of their treatment decisions, which led us to develop items for perception and desire for control in decision-making based on the question “How much control do you currently have/would you like to have over decisions regarding your radiation treatment?” We measured these responses on a 5-point rating scale from “a lot” to “none.” Patient satisfaction was measured based on the question “How satisfied are you with the radiation treatments you have received?” We measured these responses on a 5-point rating scale from “very much” to “not at all.” For analysis purposes, responses were dichotomized into “yes” or “no” categories, in which only answers that were in the most positive response option (eg, “very much” satisfied, or “a lot of control”) were placed into the “yes” category for satisfaction or control, and the remaining 4 options were categorized as “no.” Symptoms of anxiety, depression, and fatigue were measured based on the question “Have you experienced any of the following bothersome symptoms during radiation treatment for which you would consider seeking treatment?” We measured these responses in simple “yes” or “no” categories.

Patients self-reported the following demographic data: age, sex, race/ethnicity, marital status, employment status, and education level. Clinical variables such as body mass index, tumor type, stage of disease, and treatment regimen were obtained through chart abstraction.

Statistical Analyses

All analyses were performed with Stata statistical software (version 12.2; StataCorp, College Station, Tex). Standard descriptive statistics were used to report demographic and clinical variables and study outcomes. Relationships between variables including SDM, desired and perceived control, satisfaction, anxiety, depression, and fatigue during the last week of RT were evaluated using chi-square analyses. All analyses were 2-sided, with an α < .05 indicating statistical significance.

RESULTS

Between July 2009 and July 2010, 380 patients were approached for survey enrollment, 324 of whom (85.3%) agreed to participate. Of the 56 patients (14.7%) who declined, 47 (12.4%) did not want to participate in research and 9 patients (2.4%) reported feeling too sick on the day of the survey. Nine subjects withdrew consent, and 10 did not return a completed survey questionnaire, resulting in a final sample of 305 patients and a final response rate of 80.3%.

The demographic and clinical characteristics of the study population are outlined in Tables 1 and 2, respectively. The 305 participants were aged 18 years to 87 years (mean, 59.8 years; standard deviation, 12.0 years). A total of 160 patients were male (52.5%). Two hundred and thirty-one patients (75.7%) were white, 60 patients (19.7%) were black/African American, 8 patients (2.6%) were Asian, 2 patients (0.7%) were Hispanic/Latino, and 4 patients (1.3%) reported “other” race/ethnicity. Sixty patients (19.7%) were diagnosed with prostate cancer, 54 patients (17.7%) with breast cancer, 55 patients (18.0%) with head and neck cancer, 46 patients (15.1%) with gastrointestinal cancer, 41 patients (13.4%) with lung cancer, and 49 patients (16.1%) with other types of cancer. Patients with all stages of cancer were included, specifically 73 patients (23.9%) with stage I, 73 patients (23.9%) with stage II, 77 patients (25.3%) with stage III, and 58 patients (19.0%) with stage IV disease. Of note, the patients with stage IV disease were not receiving palliative RT and were treated with curative intent. Approximately one-half of the patients were currently receiving or previously had received chemotherapy. Approximately 50% of the patients previously had undergone or were planning to undergo surgery as part of their therapeutic regimen.

Table 1. Demographic Characteristics of Participants by Shared Decision-Making, Control, and Satisfaction (N=305)
 No. Participated (% of Total)aHave Shared Decision-Making, No.%bPcHave Control of Treatment, No.%bPcVery Satisfied With Treatment, No.%bPc
  1. a

    Numbers may not add up to 100% due to missing data.

  2. b

    Indicates the percentage of patients with shared decision-making, control, or who were very satisfied within a specific category. Percentages were calculated based on the number of patients who answered each respective section of the survey.

  3. c

    Derived using the chi-square test.

  4. d

    Bold type indicates statistical significance.

  5. e

    A total of 19.7% of patients were reported as black/African American, 2.6% were reported as Asian, 7% were reported as Hispanic/Latino, and 1.3% were reported as “other.”

Total305 (100%)9031.3 9832.3 22776.2 
Age, y   .393  .047d  .180
<5592 (30.2)2730.0 3740.2 6672.5 
55-65118 (38.7)3935.8 4034.5 9581.9 
>6595 (31.2)2427.0 2123.3 6672.5 
Sex   .101  .617  .285
Male160 (52.5)5335.6 5334.2 12278.7 
Female145 (47.5)3726.6 4531.5 10573.4 
Race/ethnicity   .795  .152  .558
White231 (75.7)6931.7 6970.4 17477.0 
Non-whited74 (24.3)2130.0 2929.6 7273.6 
Education attainment   .639  .237  .730
≤High school88 (28.9)2530.9 3440 6877.3 
Some college or technical school150 (49.2)4229.4 4630.9 11174.0 
≥College67 (22.0)2335.9 1828.12 4871.6 
Employment status   .369  .769  .217
Not currently employed157 (51.5)5033.8 5032.7 11474.0 
Employed part or full time140 (45.9)3828.8 4734.3 10980.2 
Marital status   .677  .112  .136
Not currently married106 (34.8)3029.7 4038.8 7471.2 
Married/partnered199 (65.3)6032.1 5829.7 15378.9 
Table 2. Clinical Characteristics of Participants by Shared Decision-Making, Control, and Satisfaction (N=305)
 No. Participated (% of Total)aHave Shared Decision-Making, No.%bPcHave Control of Treatment, No.%bPcVery Satisfied With Treatment, No.%bPc
  1. Abbreviations: BMI, body mass index; GU, genitourinary.

  2. a

    Numbers may not add up to 100% due to missing data.

  3. b

    Indicates the percentage of patients with shared decision-making, control or who were very satisfied within a specific category. Percentages were calculated based on the number of patients who answered each respective section of the survey.

  4. c

    Derived using the chi-square test.

  5. d

    Bold type indicates statistical significance.

Total305 (100%)9031.3 9832.3 22776.2 
BMI   .166  .768c  .064c
Normal (<24.9)120 (39.3)4136.9 4135.3 8775.7 
Overweight (25-29.9)106 (34.8)3030.6 3231.1 8782.9 
Obese (≥30)79 (25.9)1924.0 2531.7 5368.0 
Cancer type   .028d  .38  .372
Prostate60 (19.7)1326.0 1935.9 4481.5 
Breast54 (17.7)1220.3 1932.2 4880.0 
Head/neck55 (18.0)2548.1 1935.9 4076.9 
Gastrointestinal46 (15.1)1432.6 920.5 3069.8 
Lung41 (13.4)923.7 1742.5 2665.0 
GU/skin/other49 (16.1)1737.0 1530.6 3979.6 
Disease stage   .571  .171  .447
I73 (23.9)2028.6 2838.9 5575.3 
II73 (23.9)1926.8 2027.8 5780.3 
III77 (25.3)2332.9 1925.3 5876.3 
IV58 (19.0)2137.5 2229.3 3867.9 
Chemotherapy   .767  .092  .269
No145 (47.5)4331.6 5337.6 11278.9 
Yes158 (51.8)4530.0 4428.4 11373.4 
Surgery   .731  .681  .143
No152 (49.9)4631.9 4731.5 11879.7 
Yes152 (49.9)4330.1 5033.8 10872.5 
Hormonal therapy   .947  .825  .667
No253 (83.0)7431.1 8032.4 18675.6 
Yes51 (16.7)1530.6 1734 4078.4 

Ninety patients (31.3%) reported experiencing SDM, 98 patients (32.3%) perceived control in treatment decisions, and 227 patients (76.2%) reported feeling very satisfied with their radiation oncology care. Among patients who perceived control of radiation treatments, 53.2% also experienced SDM, and of those who experienced SDM, 55.6% also perceived control. Similarly, of the patients who did not perceive control, 79.4% also did not experience SDM, and of those who did not experience SDM, 77.8% also did not perceive control. The relationships of a variety of demographic and clinical factors were analyzed to determine whether there were any factors linked with increases in SDM, perceived control, and patient satisfaction. Patients who had a head and neck malignancy experienced more SDM compared with patients with other disease sites (P = .028). Patients of a younger age (aged < 55 years) perceived more control of treatment decisions compared with patients of other age groups (P = .047). Otherwise, there were no significant differences noted with regard to SDM, perceived control, or satisfaction based on a variety of demographic or clinical factors (Table 1 and Table 2, respectively).

SDM and perceived control of treatment decisions were each found to be independently associated with an increase in patient satisfaction. Specifically, 84.4% of patients who experienced SDM reported being very satisfied with their radiation treatments, compared with only 71.4% of patients who did not experience SDM (P < .02). The perception of having control in treatment decisions was associated with a similar trend of increased satisfaction (89.7% vs 69.2; P < .001) (Fig. 1). Importantly, a patient's perception of control in treatment decisions was associated with an increase in satisfaction regardless of whether the patient actually preferred control. Specifically, patients who did not desire control over their treatments but who did perceive a sense of control offered by their physicians were more satisfied with their treatments compared with patients who did not perceive control (100% vs 72.8%; P < .02). Similar trends in satisfaction were noted in patients who desired control of their treatments (87.5% vs 46.2%; P < .001) (Table 3).

Figure 1.

Shared decision-making (SDM) and perceived control of radiation treatments are associated with an increase in patient satisfaction. There was a significant association noted with patients feeling very satisfied with their radiation treatments if they experienced SDM (84.4% vs 71.4%; P < .02) or perceived control over their treatment regimen (89.7% vs 69.2%; P < .001). The P values were based on chi-square analyses.

Table 3. Percentage of Patients Satisfied Based on Preference For or Perception of Control
 Perceived ControlDid Not Perceive ControlPa
  1. a

    Derived using the chi-square test.

Preferred control87.5%46.2%<.001
Did not prefer control100%72.8%<.02

Further analysis revealed that patients who desired, but did not perceive, control of their radiation treatments were more likely to experience certain detrimental symptoms that were severe enough for patients to consider seeking additional treatment for them. In particular, these patients experienced an increase in self-reported anxiety (44.0% vs 20.0%; P < .02), depression (44.0% vs 15.0%; P < .01), and fatigue (68.0% vs 32.9%; P < .01) compared with individuals who both desired and perceived control (Fig. 2).

Figure 2.

Patient-perceived control in treatment decision-making and associated detrimental symptoms are shown. Patients who specifically desired control over their treatment decisions, but did not perceive this control, experienced significantly more anxiety (44.0% vs 20.0%; P < .02), depression (44.0% vs 15.0%; P < .01), and fatigue (68.0% vs 32.9%; P < .01) compared with patients who did perceive a sense of control in their treatment decisions. The P values were based on chi-square analyses.

DISCUSSION

There is a growing trend for patients diagnosed with cancer to desire a more active role in their treatment decisions. SDM takes into account physician expertise as well as patient beliefs and values to jointly develop a medically acceptable treatment plan. With this evolving patient-provider relationship comes a significant need for the medical community to develop tools and training methods for both physicians and patients to adapt to this method of decision-making.[19, 20]

In the current study, comprising a group of radiation oncology patients with diverse demographic and clinical characteristics, only approximately one-third of patients experienced SDM or perceived a sense of control in their radiation treatments. These numbers are consistent with prior studies examining SDM in the oncology setting.[11] It is interesting to note that SDM and perceived control are distinct variables, in which SDM focuses more on the joint decision-making process of a physician and patient together coming to a mutually acceptable decision and patient control is more simply a measure of how much control a patient perceives in his or her treatment decisions. Patients who experienced SDM or who perceived a sense of control were more likely to report feeling very satisfied with their radiation treatments. Importantly, this increase in satisfaction was observed in patients who perceived control, regardless of whether they expressed a desire for this control. This increase in patient satisfaction confirms other reports in which improved information sharing and patient participation led to improved patient satisfaction.[2, 5, 14, 21]

In addition, patients with cancer can perceive a loss of control in their lives, which can potentially lead to debilitating psychological effects.[22] Other groups have reported that SDM is correlated with a decrease in various psychologically detrimental symptoms, such as patient anxiety.[2] Although we did not observe an increase in anxiety or depression in patients who did not experience control or SDM (data not shown), we did find a significant increase in patient anxiety, depression, and fatigue in those who entered the treatment process with a desire for control, but who did not actually experience a sense of control in guiding their treatment decisions.

The ideal and highest level of SDM in the oncology decision-making process involves a patient deciding with his or her physician between different, but equally effective, treatment modalities. For example, a woman with early-stage breast cancer can decide between mastectomy and breast conservation therapy (lumpectomy and RT), each of which has similar clinical efficacy.[23] However, a thorough discussion with the oncology team is important to address the different risks and benefits of each treatment option. Such a clinical scenario has been studied in depth by Whelan et al,[5] who randomized patients with early-stage breast cancer to the use of a specific decision aid tool during consultation compared with standard consultation. The decision tool described the differences and associated side effects of mastectomy and breast conservation therapy. Patients randomized to the arm using the decision aid tool had better knowledge regarding treatment options and less decision conflict, and they were more satisfied.[5]

The decision for a patient to undergo RT is the first of many opportunities for SDM. Depending on the stage and type of malignancy, there may be several possible radiation regimens using different radiation types and different fractionation schemes (number of and dose per radiation treatment). One important example is the treatment of prostate cancer, in which there have been and are numerous ongoing studies evaluating different radiation fractionation schemes that can range from as short as 2 weeks to as long as 9 weeks.[24] In fact, van Tol-Geerdink et al found that offering patients a choice in their radiation dose with decision aids reviewing the risks and benefits of each therapy improved patient knowledge, risk perception, and satisfaction of information compared with patients who did not have a choice of their radiation dose.[15] As various radiation regimens become more established as acceptable alternatives to standard regimens, and thus more widely available, it will become incumbent on radiation oncologists to thoroughly explain the potential risks and benefits of each regimen and to help patients select a mutually acceptable treatment plan.

Lastly, while undergoing RT, patients see their radiation oncologists weekly for symptom management related to the acute toxicity of radiation. These weekly visits offer opportunities for ongoing SDM, because the care teams can actively engage a patient in the management of his or her symptoms. For example, pain control should involve a discussion of the risks and benefits of standard opioid analgesics, as well as the option of potentially effective forms of alternative medicine, such as acupuncture. Because the choice of cancer treatment regimens may often be dictated by the extent of disease and available therapeutics, involving patients in decisions regarding symptom management is an opportunity to give them some sense of control in decision-making.

Importantly, during all of the steps of treatment decision-making for a patient undergoing RT, it is paramount that physicians not allow patients to dictate care to a level that is either potentially unsafe or not within a physician's comfort zone. Hence, SDM should take into account both physician and patient preferences.

There are several potential limitations to the current study. First, this is a cross-sectional study, which can only identify the association between measured variables. Without the temporal relationship, we cannot for example determine whether SDM or patient control actually leads to patient satisfaction. It is conceivable that patients who are more satisfied with their treatment may retroactively claim more responsibility for the treatment decisions then those who are dissatisfied with their treatment.[25] This limitation also applies to other measured variables. Second, our measures for control, satisfaction, fatigue, anxiety, and depression are brief and may be subject to response bias. Furthermore, the time point of our survey during the last week of RT could introduce some bias, because the last week of RT is often when side effects are the most severe and may affect how a patient responds to the survey questions. However, we chose this time point to capture all aspects of SDM in the treatment process and not introduce further recall bias by distributing the survey at a follow-up appointment. A prospective longitudinal cohort study may more accurately measure how SDM throughout the course of RT may impact satisfaction and other outcomes. Last, this study was conducted at an urban academic cancer center, and therefore may not be generalizable to patients in other settings.

Despite the above limitations, the current study surveyed a diverse group of patients with a very high participation rate. To our knowledge, this is the first study to date to evaluate the relationship between both patient-perceived SDM and control with satisfaction with RT. We found that less than one-third of patients experienced SDM, which highlights the critical need for both physician-targeted and patient-targeted methods to improve the participatory decision-making process. These findings also emphasize the value of SDM and patient-perceived control during RT, particularly as it relates to patient satisfaction, anxiety, depression, and fatigue. Irrespective of a patient's desire for control, it is important for physicians to engage patients in the decision-making process and to allow patients to believe that they have some control in their cancer care. The previously mentioned Institute of Medicine Report entitled “Delivering High-Quality Cancer Care: Charting a New Course for a System in Crisis,” provides excellent guidance in the field.[10] Specifically, the authors identify broad methods to improve patient-centered communication and SDM including “…(1) making more comprehensive and understandable information available to patients and their families, (2) developing decision aids to facilitate patient centered communication and shared decision making, (3) prioritizing clinician training in communication, (4) preparing cancer care plans, and (5) using new models of payment to incentivize patient-centered communication and shared decision making.”[10] In addition to these helpful suggestions, another important method to improve SDM in radiation oncology is to educate physicians about both the prevalence and relevance of SDM in patient care. Once radiation oncologists become more aware of the issue, they will hopefully be more cognizant to incorporate SDM in their daily practice. Future research may in addition identify both patient and physician barriers to SDM, and in turn test the feasibility and efficacy of methods to address these barriers to improve patient satisfaction and clinical outcomes.

FUNDING SUPPORT

Supported in part by the Penn Integrative Oncology Fund.

CONFLICT OF INTEREST DISCLOSURES

Dr. Mao is a recipient of the National Institutes of Health/National Center for Complementary and Alternative Medicine 5K23 AT004112-5 award.

Ancillary