Welcome! Pragmatic Clinical Studies. David Hickam, MD, MPH Program Director Clinical Effectiveness Research. David Hickam, MD, MPH

Transcription

1 Pragmatic Clinical Studies David Hickam, MD, MPH Program Director Clinical Effec2veness Research June 23, 2015 Welcome! David Hickam, MD, MPH Program Director Clinical Effectiveness Research 2

2 In this Session PCORI s approach to evidence gaps and developing new clinical evidence Key features of the funding announcements for PCORI s Pragma2c Clinical Studies Methodological issues/standards Lessons we have learned PCORI priority topics 3 What is Evidence-based Information? Clinical evidence: Valid data about the outcomes experienced by pa2ents who receive medical care. The popula2on is well defined. The clinical interven2ons are well defined. We have informa2on about the most important outcomes (both benefits and harms). Compara2ve effec2veness Star2ng point is the choices people make about the op2ons for managing a disease. These choices inform the focus of new research. The research compares the benefits and harms associated with each op2on.

3 Perspectives on Comparative Effectiveness Research Compara2ve Effec2veness Research should be a public good that: Gives health care decision makers pa2ents, clinicians, purchasers and policy makers access to the latest open and unbiased evidence- based informa2on about treatment op2ons Informs choices and is closely aligned with the sequence of decisions pa2ents and clinicians face First Steps in Developing New Comparative Effectiveness Research Understand the choices made by pa2ents and clinicians Which clinical op2ons are realis2cally available to pa2ents? Define the important pa2ent sub- groups Recognize dispari2es and their sources Define the outcomes (benefits and harms) that are important to pa2ents Benefits Harms

4 New Comparative Effectiveness Research Must Address Important Evidence Gaps Assess the available evidence about important outcomes Systema2c reviews Evidence gaps that are important to decision makers Design a study that can feasibly close the evidence gap If the gap is not important, the research will not be useful Timeline of PCORI Pragmatic Studies Initiative First funding announcement in February 2014 First funded projects in mid Compe22ve LOI s Deadline past for current (fourth) announcement Next LOI deadline fall 2015 Emphasis on priority clinical topics Inves2gator- ini2ated topics are also considered 8

5 Details About Funding Project dura2ons of up to 5 years Up to $10M in total direct costs Expect to fund approximately 20 projects/year in What Are We Talking About? Pragma2c Clinical Studies are intended to provide informa2on that can be directly adopted by healthcare providers. Mostly conducted in rou2ne clinical segngs Large, because the expected differences in effec2veness may be small, yet important or different in pa2ent subgroups Less intrusive to rou2ne clinical prac2ce Some2mes called Large Simple Trials 10

6 Traditional Randomized Controlled Trials Study sample tends to be homogeneous, highly mo2vated (and therefore more adherent), rela2vely free of comorbid condi2ons Research tends to take place in specialized research segngs Research protocols are ojen strict and do not represent typical clinical prac2ce 11 What is a Pragmatic CER Study? Answers a prac2cal, real world compara2ve effec2veness research ques2on. Assesses whether two or more op2ons differ in effec2veness when administered as they are in real life Project is conducted in a clinical segng that is as close as possible to a real world segng. The methodological approach (including study design, outcome measures, and follow- up) is as simple as possible without sacrificing scien2fic rigor.

7 Justification for the Design Elements of a Large Pragmatic Study Suggest reviewing pragma2c explanatory con2nuum indicator summary (PRECIS) tool Consider tradeoffs Eligibility criteria Flexibility of interven2on Range and types of outcomes Follow up intensity Adherence Etc. Source: A pragma.c explanatory con.nuum indicator summary (PRECIS): a tool to help trial designers. Thorpe, et al. CMAJ 2009; 180:E47- E57. Research Activities Not Supported in this PFA Studies of decision aids Efficacy trials Evidence syntheses Cost- effec2veness analysis Research that aims to compare the overall costs of care between two or more alterna2ves and use the results to determine the preferred alterna2ve

8 Study Design Compare the benefits and harms of at least two approaches known to be effec2ve for a par2cular clinical condi2on Drugs Devices Procedures Behavior modifica2on Integra2ve services Delivery- system interven2ons Projects can address comparisons of discrete clinical services or complex interven2ons RCTs or large prospec2ve observa2onal studies 15 Focus of Studies Outcomes meaningful to pa2ents: Morbidity Mortality Symptoms Func2onal status Quality of life Absenteeism from work or school 16

9 Studies Must Have an Impact Focus on a CER ques2on that is important to pa2ents and decision makers Address an evidence gap that has been substan2ated by exis2ng rigorously conducted systema2c reviews or specifically emphasized by a professional society s evidence- based clinical prac2ce guideline Endorsement by relevant pa2ent organiza2ons, clinician organiza2ons, payer/purchaser consor2a, and/or life sciences industry representa2ves as being a cri2cal ques2on one that, if adequately answered, would substan2ally improve decision making 17 Study Populations Examine diverse popula2ons receiving care in real- world segngs Have strong interest in and support for the study by host delivery systems and clinical care segngs Specify broad and simple eligibility criteria that will allow wide generaliza2on of results, while amending appropriately to any ethical concerns of excess risk in some pa2ent subgroups 18

10 Studies Need to Launch Quickly Compare interven2ons that are either known to be efficacious, or are commonly in use, and can be implemented in real- world segngs Capacity to collect efficiently pa2ent- centered outcomes periodically during follow- up Provide preliminary evidence of the poten2al for efficient recruitment, high par2cipa2on rates, and appropriate oversight by local or centralized Ins2tu2onal Review Boards (IRBs), including plans for streamlining or waiving individual informed consent in cases of low- risk interven2ons Plan for sharing de- iden2fied data 19 Strategies for Engagement with Clinical Partners Iden2fy and engage with major pa2ent and stakeholder organiza2ons that would implement study findings as well as with exis2ng local communi2es of pa2ents and care providers to formulate the research ques2ons, design the study, help monitor progress, and disseminate the findings Minimize disrup2on to par2cipants daily rou2nes (e.g. minimize par2cipant visits intended solely for study- assessment purposes; capture PROs during office visits, electronically, or via phone) 20

11 More Strategies Design the study so that the conduct can be, as seamlessly as possible, integrated with rou2ne clinic/office opera2ons Use efficient methods to obtain par2cipant consent while s2ll mee2ng ethical and legal requirements Capitalize on the exis2ng electronic health records and other computerized informa2on to iden2fy and recruit eligible pa2ents, monitor study conduct and pa2ent safety, and collect study outcomes informa2on If data standardiza2on and interoperability across study sites has not already been accomplished, develop methods that will enhance the standardiza2on of data that are accessed from different electronic health record systems. 21 How Large? Sample size large enough to enable precise es2mates of the rates of benefits and harms of the clinical services being compared Support tes2ng of hypotheses related to poten2al differences in effec2veness in relevant pa2ent subgroups (heterogeneity of treatment effects) Studies should be large enough and long enough to Capture the relevant outcomes Allow examina2on of possible differences in effec2veness in key pa2ent subgroups At least 2,000 pa2ents More than 45,000 in several trials Must include a broad and diverse popula2on 22

12 Human Studies Requirements Applicants should provide preliminary evidence of the poten2al for: efficient recruitment high par2cipa2on rates appropriate oversight by local or centralized Ins2tu2onal Review Boards (IRBs) Intensity of oversight and complexity of informed consent procedures should be closely related to the degree of risk from study par2cipa2on. 23 Engagement in PCS Applicants must partner with relevant pa2ent, clinician, and other stakeholder organiza2ons Partners must strongly endorse the proposed study and be involved with research teams throughout the conduct of the study Research is more relevant Findings are more likely to be disseminated and implemented 24

13 Partnerships Expecta2on that applicant researcher have partnered with relevant pa2ent organiza2ons, specialty professional organiza2ons, healthcare systems, insurers, and/or employer purchasers in preparing applica2ons Partner finaliza2on and endorsement is essen2al for labeling high priority When partnerships should be established 25 Methodological Issues In the case of randomized trials, adherence to current best prac2ces for conduc2ng pragma2c trials: Standardized inclusion/exclusion criteria Proper randomiza2on Techniques to minimize poten2al for missing data Appropriate safety monitoring (including establishment of a data- safety monitoring board, or discussion of why such a board is unnecessary) 26

14 The Application Process 27 First Steps Register and create an account in PCORI Online Submit Lemer of Intent (LOI) Full Applica2ons by Invita2on Only 28

15 LOI Evaluation Criteria Importance and relevance of the topics to PCORI priori2es Other topics are considered but preference will be given to the PCORI Priority Topics The likelihood of meaningful change in pa2ence outcomes and/or healthcare prac2ces The clarity and credibility of applicants responses to the LOI ques2ons, as well as jus2fica2on of the need for a large pragma2c study Prior relevant research experience and programma2c fit and balance, taking into considera2on whether the par2cular proposal fills a gap in the porpolio of proposals with certain characteris2cs including disease category, topics, priority popula2on, methodologies, and other variables 29 Budget Will vary widely among studies based on Study topic and design Needs for recruitment and/or primary data collec2on Required length of follow- up Analy2c complexity 30

16 Lessons We Have Learned New op2ons are less desirable than an op2on that is commonly available PCORI is not looking to fund new or untested interven2ons 31 Pitfalls Compara2ve Efficacy Usual Care Placebo as comparison Tests efficacy (or compara2ve efficacy) within a 2ghtly protocol- controlled research segng (as opposed to more real- world, pragma2c CER) Formal cost- effec2veness analysis in the form of dollar- cost per quality- adjusted life- year to compare two or more alterna2ves Direct cost analysis Natural history of disease Instrument development 32

17 PCORI Funded Studies Cancer Care Stroke Back Pain Improving Healthcare Systems PCORI Priority Topics PCORI Priority Topics Medical versus invasive procedures for asymptoma2c caro2d artery disease Surgical op2ons for hip fracture in the elderly Pelvic floor mesh implants 34

18 Questions? Thank You! David Hickam, MD, MPH Program Director Clinical Effectiveness Research