1. Performing a Successful Supportive Care Clinical Trial Jennifer Temel, MD

2. Supportive Care Clinical Trials 1. Identify the problem Why are you thinking about doing a trial? 2. Develop your intervention What do you propose to do to fix the problem? 3. Clarify your population Who will you include in the trial? 4. Define the detail of the intervention Where will it take place? 5. Determine the timing and extent of your measures When will you administer PROs? 6. Finalizing your study procedures How will you administer intervention and collect data?

3. Identify the Problem (Why)  What is the problem you are trying to fix, solve, improve, enhance, etc?  As clinicians, trust your instincts about the deficiencies in care you see in the patient population you care for.  Reinforce these ideas by reviewing the scholarly literature.

4. Identifying the Problem Experienced by Patients with Lung Cancer  As a fellow caring for patients with metastatic lung cancer, I felt as if their physical and psychosocial needs were not being adequately addressed.

5. Develop the Intervention (What)  Now that you have identified a problem, which is supported by the scholarly literature, how do you propose to fix it?  At this stage, you need to have a sense of the primary components of your intervention but you will continue to refine the details as you proceed further in the intervention development.

6. Develop the Intervention  Often what you are proposing to do in supportive care research is so new or novel that you must start with a pilot study.  Not only does a pilot study allow you to assess the feasibility and acceptability of your intervention but also allows you to work out the kinks.

7. Developing an Intervention for Patients with Lung Cancer  Since palliative care focuses on patients’ physical and psychological symptom burden, maybe they could help fix the problem.  However, palliative care only cared for patients in the hospital which was too late to substantially impact patients’ experience throughout their course of illness.

8. Developing the Palliative Care Intervention 1. Palliative care clinicians see patients only in the hospital. 2. Palliative care clinicians do not have a clinic. 3. Palliative care clinicians do not have a schedule to see patients in clinic. 4. The oncology clinic does not have rooms for palliative care clinicians to see patients in clinic. Therefore, we began with a pilot study to allow us to develop the clinic and work out the kinks before starting the RCT.

9. Clarify Your Study Population (Who)  Patients with any cancer? only lung cancer? only NSCLC ? only metastatic NSCLC?  Who do you think would benefit most from the intervention?  Is it feasible to do a study in different cancer populations i.e. breast and lung?  Do you have sufficient numbers of patients to do only a subset of cancer patients?  Do you even have a database to KNOW the number of patients seen with metastatic breast cancer?

10. Clarifying the Population in the Palliative Care Study 1. I was a lung cancer clinician so it was easiest for me to get buy in from other lung cancer clinicians. 2. The MGH Cancer Center is divided into Disease Centers and it is more difficult to open studies across different centers. 3. We saw many patients with metastatic NSCLC and they were sick so I hoped palliative care could make them better. 4. I wanted to examine end of life care outcomes.

11. Define the Detail of the Intervention (Where)  Where will the intervention take place? In the clinic? In the infusion room? In radiation oncology? Over the telephone?  What barriers do you need to overcome to make the location of the intervention possible?  What will you do if patients miss their appointments?  Where will you see patients to collect their patient reported measures?

12. Defining the Detail of the Palliative Care Intervention  We have significant space limitations so many palliative care visits happened in the infusion room (complex scheduling!!).  Sometimes palliative care clinicians missed the patient (i.e. chemo cancelled) so they were permitted to do telephone visits.  It takes patients 15 minutes to complete PROs and we often did not have space to complete them in exam rooms which impacts the format of administration (i.e. do you want patients wandering around with your tablet computers?)

13. Determine the Timing and Extent of Your Measures (When)  Administering PROs is time consuming for your patients and research staff.  If you ask people to complete them too often – you will increase missing data.  TIME PROs WITH PATIENTS SCHEDULES  If you ask people to complete too many measures – you will increase your missing data.  PUT WHAT YOU CARE ABOUT MOST FIRST IN THE PACKET  What window will you allow to complete? What if patients have no appointment in the window?  Take into account your staffing and sample size when determining the frequency of administration of PROs.

14. Timing and Extent of Measures in Palliative Care Study  We first administered questionnaires at baseline, 12, 18 and 24 weeks but had too much missing data at 18 weeks so now only do three time points.  Our primary outcome measure (QOL) is first in the packet.  We had added a coping measure – but patients complained - so we removed it.  We allow a three week window to try to have patients complete during regularly scheduled visits.

15. Finalizing Your Study Procedures (How)  How will you identify eligible patients?  Will you rely on referrals from physicians or will you screen patients?  If you have a longitudinal intervention, how will you know when the patients are due for a study visit?  If you have longitudinal PROs, how will you know when the patients are due to complete them?  Who will consent patients for the study?

16. Finalizing Study Procedures for the Palliative Care Study  In early trials, we relied on clinicians to identify patients.  Now, we screen new patients lists and track patients for eligibility AND send email reminders AND put reminders on paper charts.  Clinicians still consent their patients for the study.  Study staff tracks patients to determine when they are due to see palliative care and which appointments are in window for completing PROs.

17. Determining Your Study Design  If you can randomize 1. Parallel 2. Cross-over 3. Adaptive  If you cannot randomize 1. Non-randomized controls (e.g. historical) 2. Pre-post design 3. Time series There is an article posted on website on clinical research designs Hopkins course on Design and Interpretation of Clinical Trials

18. What if you want to collect data from others?  Family caregivers:  Sometimes multiple family caregiver accompany patients to visits.  Sometimes family caregivers do not attend all visits which can be problematic if you are collecting longitudinal data.  Physicians:  Simplify data collection as much as possible.  If obtaining qualitative data, use existing meeting structures.

19. Continue to Revisit Your Study Plan Study Plan Intervention (what) & Outcome measures (what and when) Your study population (who) The location of your intervention (where) Your study procedures (how)

20. Dyspnea Intervention for Patients with Lung Cancer 1. Why? Patients with metastatic lung cancer frequently have dyspnea but do not have access to behavior therapy. 2. What? Behavioral interventions are beneficial but not feasible due to lack of trained clinicians. We trained NPs in infusion room to do brief CBT in a pilot study. 3. Who? Patients with metastatic lung cancer with dyspnea. 4. Where? Infusion room (no space in clinic and patients are a captive audience). 5. When? Just pre and post intervention measures as longitudinal measures will have missing data and effect of intervention should be immediate. 6. How? Reviewed upcoming clinic lists at weekly oncology meeting to identify patients with dyspnea and then RA enrolled.

21. Summary  One of the most gratifying aspect of conducting supportive care research is allowing your clinical experience and judgment to guide your research questions.  Conducting successful supportive care trials requires careful attention to details, which are often not necessary, for example, when conducting treatment clinical trials in oncology.  Beginning with pilot studies to establish feasibility and work out challenges is often the best strategy with a new or novel intervention.