Use of Databases in Post-Marketing Surveillance

Document Sample
Use of Databases in Post-Marketing Surveillance Powered By Docstoc
					  Is FDA failing us
Or are we failing FDA?




     Jonathan C. Javitt, M.D.,M.P.H
     Senior Fellow, Potomac Institute for Policy Studies
     Former Chair, Health Subcommittee, President’s
     Information Technology Advisory Committee,
Don’t take my word for it
•Our [FDA’s] approach to risk
management includes finding ways to
help make sure that drugs we review
are used safely once they’re approved.
•Too often, the drugs, devices, and
other products we regulate are
involved in preventable adverse events.
•As many as 20% of Americans have
experienced some kind of medical
error.
•Preventable errors and complications
involving prescription drugs alone are
responsible for thousands of deaths,
millions of hospitalizations, and
billions of dollars in additional health
care costs each year, not to mention the
unnecessary suffering.

          Mark McClellan, M.D., Ph.D.

                                           Jonathan C. Javitt, M.D., M.P.H.: NEI/FDA Endpoints Meeting. November 28, 2006
A seemingly vicious circle

   Too few NME’s will be approved this year
   The cost of bringing new drugs to market is far
    too high and takes too long
   Congress and the public have labeled FDA’s drug
    safety process as fundamentally flawed
   The more pressure that is brought to bear on FDA
    to improve drug safety, the more risk averse the
    agency becomes
   Nobody ever got fired from FDA for witholding
    approval on safety grounds



                                      Jonathan C. Javitt, M.D., M.P.H.: ASENT 2008
The other side of the coin

   The real threat to American lives is not dangerous
    drugs, but rather errors in the use of generally
    safe and effective drugs
   More Americans are harmed by their failure to
    receive a needed drug than will ever be harmed
    by administration of a harmful drug
   The problem lies in the disconnect between
    postmarketing surveillance, pharmaceutical risk
    management, and clinical practice




                                       Jonathan C. Javitt, M.D., M.P.H.: ASENT 2008
Today’s question

   What is the best way to ensure public safety once
    drugs are introduced to the marketplace?


 Yesterday’sanswer  anecdotal reporting
 Today’s answer  database analysis
 Tomorrow’s answer  sentinel systems




                          Jonathan C. Javitt, M.D., M.P.H.: NEI/FDA Endpoints Meeting. November 28, 2006
A Unified Sentinel System: transforming a
vicious circle into a virtuous one
                               A Sentinel System

                               Continuously monitors
                               all available sources of
                               medical information in
                               order to identify:
                               1. Common medical
                                    errors
                               2. Drug adverse events
                                    and interactions
                               3. Failure to comply
                                    with care
                               4. Threatening clinical
                                    parameters
                               5. Opportunities to
                                    improve care




                                Jonathan C. Javitt, M.D., M.P.H.: ASENT 2008
GAO Recognizes the problem

However the proffered solution
will add enormous cost to the
regulatory process and provide
far less benefit than might
otherwise be achieved

GAO does realize, however, that
anecdotal reporting databases for
drug complications are
inherently flawed and that
population surveillance is needed
                   Jonathan C. Javitt, M.D., M.P.H.: ASENT 2008
Drug Safety
Health Plan Records Seen as Cost-EffectiveW ay to Identify Risks of Drugs on Market



        Analyzing large health plan databases to
        detect drug side effects would be a more cost-
        effective way to monitor drug safety than the
        risk-minimization plans proposed in a Senate
        bill (S. 484), a former head of the Food and
        Drug Administration said recently.

        Mark McClellan, the former commissioner of
        FDA, said that he supports S. 484, the drug-
        safety legislation sponsored by the chairman
        and ranking member of the Senate Health,
        Education, Labor, and Pensions Committee,
        calling it "the right foundation to build on."
        However, he and others who spoke March 21
        at a Capitol Hill briefing cast doubt on whether
        the legislation makes use of the best type of
        postmarket drug-safety surveillance



                                                           Jonathan C. Javitt, M.D., M.P.H.: ASENT 2008
As Safety concerns mount, sponsors have
three basic alternatives

10,000   – 100,000 person
safety trials prior to market
approval Cost prohibitive
Construction of ad-hoc
registry programs for each
newly launched drug
          Extremely expensive

Deploy    Sentinel
Database systems in order
to monitor drug safety and
utilization $.50 pmpm


                                Jonathan C. Javitt, M.D., M.P.H.: NEI/FDA Endpoints Meeting. November 28, 2006
The drug safety paradigm must change

While reporting databases can, and should be used
 to report rare, life-threatening events, moving to
 active surveillance systems enables us to:

   Actively keep drugs safe in the marketplace
   Improve patient safety by identifying those who
    would benefit from therapies they are not
    receiving
   Potentially launch new drugs much more quickly
    into a monitored environment



                          Jonathan C. Javitt, M.D., M.P.H.: NEI/FDA Endpoints Meeting. November 28, 2006
            The Same Framework that
            addresses the most critical
               problems confronting
                 patient care, also
             addresses the challenges
                facing postmarket
               surveillance and drug
                      approval




Jonathan C. Javitt, M.D., M.P.H.: NEI/FDA Endpoints Meeting. November 28, 2006
        The Active Health Care Engine®

 Hospital           Lab                Pharmacy                                   Provider                         RN Notes
  Claims           Results              Claims                                    Claims




                                                                                                                16 million
                                                                                                                 covered
                                                                                                                   lives


                                          Communication Portal                  Letters
                         MD / RN
                         Review
     Clinical
      Rules          Clinical Review
                         System
      C                                                                                                    7. Action by Doctor
                         Automated
                                                                                                               and Patient


4. Apply Medical       5. Create                                    6. Communicate
   Intelligence    Recommendations                                Recommendations
                                                                 Jonathan C. Javitt, M.D., M.P.H.: NEI/FDA Endpoints Meeting.   November 28, 2006
Can deployed Sentinel Systems save lives
and save cost?



    100,000 member midwestern managed care plan
    Selected all members with at least one physician visit
     and one pharmacy claim in preceeding 12 months
    20,000 members selected per group
    Random allocation to Sentinel system with mailed
     clinical prompts or to watchful waiting
    Study period of 12 months




                             Jonathan C. Javitt, M.D., M.P.H.: NEI/FDA Endpoints Meeting. November 28, 2006
Implementation of Recommendations


 Adoption            Care Consideration Description

   86%      Improve a subtherapeutic INR
   73%      Normalize high INR
   63%      Monitor for toxicity in amiodarone use
   58%      Stop NSAID in anticoagulation
   56%      Stop either Viagra or nitrates
   50%      Monitor INR with interacting drugs
   46%      Monitor INR in Coumadin use
   44%      Stop Wellbutrin in patients with seizures
   42%      Monitor LFTs in patients on glitazones
   38%      Monitor for toxicity in methotrexate use
   37%      Stop metformin in patient with renal insufficiency.

                                    Jonathan C. Javitt, M.D., M.P.H.: NEI/FDA Endpoints Meeting. November 28, 2006
Randomized Prospective Study:

Approximately 1,000 interventions issued in the intervention
group and a similar number observed in the control group
                           Control                      Study                     Difference
Total Admissions          1052                     936                                         11%*

Admits/1000               47.55                    42.1                                        11%*
Hospital Bed-Days total   4089                     3483                                       15%*
ALOS                      3.89 days                3.72 days                                 4.5%*
Days per 1000             184.8                    156.9                                     15%**
Total paid                $31.3 MM                 $29.6 MM                                5.5%**
Inpatient paid            $4.6 MM                  $4.1 MM                                   11%**

                                                                               *                p<.03
 Overall cost saving of 6% - 8%                                               **               p<.005
                                  Jonathan C. Javitt, M.D., M.P.H.: NEI/FDA Endpoints Meeting. November 28, 2006
                Sentinel Systems in
             Postmarketing Surveillance

   Reinventing the concept of drug safety
    – Keeping drugs safe in the marketplace through early detection
      of adverse events
       • Enables early intervention
       • Enables early implementation of monitoring
       • Enables early labeling changes
    – Keeping drugs safe in the marketplace by avoiding interactions
      and proactive notifying caregivers
    – Keeping drugs safe by ensuring clinical monitoring
    – Keeping patients safe by increasing compliance with therapy
    – Keeping patients safe by identifying those in need of
      unprescribed therapy



                                 Jonathan C. Javitt, M.D., M.P.H.: NEI/FDA Endpoints Meeting. November 28, 2006
Sentinel Systems will actually Increase
appropriate sales of lifesaving drugs

   Half of those who would benefit from beta
    blockers post MI are not receiving them

   Half of those who qualify for ACE inhibitors under
    the HOPE trial criteria do not receive them

   And so on…




                                       Jonathan C. Javitt, M.D., M.P.H.: ASENT 2008
Keeping Drugs Safe in the Marketplace

   When Rezulin was approved for the treatment of
    diabetes, the possibility of liver failure was well
    recognized and all involved were confident that with
    proper laboratory monitoring patient safety could be
    assured
   When the first (predictable) cases of liver failure were
    reported, FDA determined that only 9% of physicians
    were following the rezulin label and obtaining LFT’s.
    Therefore the drug was withdrawn on the grounds
    that safety could not be maintained in the
    marketplace




                             Jonathan C. Javitt, M.D., M.P.H.: NEI/FDA Endpoints Meeting. November 28, 2006
Frequency of pharmaceutical risk management
recommendations:



                                                         Patients               Triggering
Clinical Issue                                           (N)                    (%)

Need for LFT’s on Rezulin                                155                    13%

Need renal monitoring on Metformin                       329                    19%

Need for INR’s on Coumadin                               153                    7%

Need for LFT’s on Statins                                931                    10%


 In other words, 87% of physicians in this sample were
 monitoring their Rezulin patients (not 9% as FDA thought) and
 those who were not monitoring frequently complied with a single
 reminder letter.
                              Jonathan C. Javitt, M.D., M.P.H.: NEI/FDA Endpoints Meeting. November 28, 2006
Proven Results

   Sentinel systems are currently workable and
    deployable
   Sentinel systems can play a direct role in
    pharmaceutical risk management
   Sentinel systems suggest action that Physicians will
    accept and implement
   The costs of sentinel systems are more than offset
    by savings in health care




                            Jonathan C. Javitt, M.D., M.P.H.: NEI/FDA Endpoints Meeting. November 28, 2006
Are Sentinel Systems merely a Dream?

   25 Million Americans already say NO!
    – However, these systems were engineered only for the
      purpose of patient safety and care improvement
   The next phase is the RHIO –
       Regional Health Information Exchange
    – First Gateways: 20 million lives this year
    – Indiana: 5+ million lives
    – New England: 5+ million lives

    – RHIOs are being engineered primarily for transportability of
      health information, driven by health systems economy, with
      patient safety as a secondary opportunity and postmarket
      surveillance not even a major topic of conversation.


                                                   Jonathan C. Javitt, M.D., M.P.H.: ASENT 2008
The RHIO environment

   Draws together all known diagnoses, procedures,
    admissions, labs, and increasingly other findings
    under a common data umbrella.
   Privacy and security issues have been legislated
    and ironed out.
   Provides the ideal environment for post-market
    surveillance, in addition to patient safety
   The problem is the business model to support
    RHIOs – right now they are grant funded




                                       Jonathan C. Javitt, M.D., M.P.H.: ASENT 2008
Business Case for RHIOS

   Today’s case
    – 1 in 5 diagnostic tests is performed simply because the
      results of previous tests are not available during the time
      frame of the medical decision
    – 1 in 8 hospital admissions is triggered by events resulting
      from failure of information flow in the outpatient environment
   Tomorrow’s case
    – New drugs will come to the market sooner
    – Drugs will be less likely to be withdrawn from the market
      because of unexpected and unmanageable toxicities
    – Patients will be more likely to be offered critical, lifesaving
      drugs



                                                    Jonathan C. Javitt, M.D., M.P.H.: ASENT 2008
The future opportunity

   Reallocate current resources spent on episodic
    reporting to supporting RHIO initiatives
   Create a structure for earlier monitored drug
    launches in a RHIO environment with user fees
    paid by the sponsors
    – Instead of requesting a 10,000 person premarketing safety
      study, the safety results from 100,000 patients can be
      obtained in a RHIO environment at far lower cost
   Replace single product risk management
    programs with population based programs that
    work through and simultaneously support the
    evolving RHIO structure


                                               Jonathan C. Javitt, M.D., M.P.H.: ASENT 2008
Cost Effectiveness

   Sentinel systems cost approximately
     $10 per person / year to implement
   Cost represents < 0.5% of overall
     health care costs
   Direct savings exceed costs at least
    10-fold, not including the benefits to
    industry and FDA of safer drugs,
    maintained in the marketplace
   Extrapolation of these implications to
     the nation suggest that a very modest level of
     IT investment (i.e. not universal EHR’s) could
     readily save the country in excess of $70 Billion/yr


                            Jonathan C. Javitt, M.D., M.P.H.: NEI/FDA Endpoints Meeting. November 28, 2006
Think about the potential for doing better, more real-world clinical
   studies at a much more affordable cost. For example, post-
   approval studies could be much cheaper and easier to do, using
   clinical protocols administered over the internet via electronic
   records, with patients selected and results recorded directly
   from electronic medical records. Low-cost clinical trials to be
   performed more quickly and easily in real-world medical
   settings. And we can use the signals we obtain from our better
   surveillance system to work with health care providers, funding
   agencies, and health care payers to target these studies more
   effectively.

                                 Mark McClellan, MD, PhD
                                           Testimony to the
                         President’s Information Technology
                                Advisory Committee -- 2003
                                Jonathan C. Javitt, M.D., M.P.H.: NEI/FDA Endpoints Meeting. November 28, 2006

				
DOCUMENT INFO
Shared By:
Stats:
views:59
posted:4/23/2009
language:English
pages:26