Slide 1 of 57

Scientific Misconduct--The "F" word-

Stan W. Woollen

Associate Director for Bioresearch Monitoring

Office for the Good Clinical Practices

and

Antoine El Hage, Ph.D.,

Chief

Good Clinical Practices Branch II, DSI

FDA

October 2001

Slide 2 of 57

Questions for the Session

• "FDA Misconduct Focus Areas: What are They and Does the Data Reinforce the Agency Concerns?"

  • Falsification, yes

• "How can Industry/QA Vigilance ÝMinimize Scientific Misconduct?"

  • QA Role=Prevention, detection, correction

Slide 3 of 57

What is Misconduct? FDA's Focus

• Recognizing Research Misconduct

  • Research misconduct means Falsification of data in proposing, designing, performing, recording, supervising or reviewing research, or in reporting research results.
  • Falsification includes acts of omission and commission.

Slide 4 of 57

Types of Falsification

• Acts of omission

  • consciously not revealing all data
    (e.g. reportable adverse events, concomitant meds., etc)

• Acts of commission

  • consciously altering data or Fabricating data (e.g. lab values, BP readings, bogus specimens)

• Data is interpreted broadly

  • individual facts, statistics, tissue samples, items of information, statements made by individuals

Slide 5 of 57

What is Misconduct? FDA's Focus

• Deliberate or repeated noncompliance with the regulations can be considered misconduct, but is secondary to falsification of data.

• Research misconduct does not include honest error or honest differences of opinion.

Slide 6 of 57

Consequences of Fraud

• If fraud takes place in a clinical trial, it places all subjects in that trial at possible safety risk

• Fraud jeopardizes the reliability of submitted and/or published data and undermines the Agencyís mission to protect and promote the public health

Slide 7 of 57

Consequences of Fraud

• Though Fraud is encountered on a relatively rare basis, its impact is broad.

  • Seriously noncompliant CIs are often used by multiple sponsors on multiple trials
  • Small number of investigators can have a broad impact on many applications by many sponsors

Slide 8 of 57

Applications and Sponsors Associated with Violative CIs

Slide 9 of 57

Falsification -A chronic problem-

• Tonyís Tales 1990-2001

  • What types of data were falsified
  • How was data falsified
  • Why was data falsified
  • Who did it
  • How was it detected

• QA- tips for preventing, detecting correcting

Slide 10 of 57

Types of Data Falsified

• EKGs

• Blood Pressure Data

• Physical and lab examinations

• Biological Specimens

• Subject Identities

• Drug Compliance Records

• You name it

Slide 11 of 57

EKGs--a common problem

• Circa 1990

  • Kept stock files of EKGs etc.
  • Made identical/duplicate prints of x-rays with different names
  • Obtained blank copies of lab reports. Used a different numbering system than numbering system used by the lab

• Multiple copy records:

  • No originals of a particular lab report-- all were photocopies

• Dr. S

  • Faked subjects: used names obtained from the obituary column
  • CRF completed after subjects died
  • Had unopened cartons of the investigational drug

Slide 12 of 57

EKGs a common problem

• - 1998

  • Dates were misrepresented on EKGs to coincide with protocol required visits (5)
  • Used 5 year old EKG tracings
  • Dates were changed in 5 records to match the washout periods required by the protocol
  • Five subjects received prohibited medication while on the study
  • Five subjects did not sign consent forms before entering the study
  • Seven subjects EKGs were not done at Visit 1 per protocol
  • 4 subjects had discrepancies between progress notes and what was reported in the CRFs

Slide 13 of 57

How is Data Falsified

• EKGs

  • Continuous strip run on one patient then torn in half and represented as coming from two subjects
  • Preprinted subject identifying information altered or obliterated on existing EKGs
  • Dr. Xerox will see you now

Slide 14 of 57

How is Data Falsified

• Blood Pressure Data

  • Blood pressure measurements and times changed in source documents to match CFR and vice versa

Slide 15 of 57

Blood pressure data questioned during in-house review

94 subjects Data Listing

Slide 16 of 57

Mind if we inspect?

Sponsorís answer--

It was concluded that inconsistencies in the source documentation would make it difficult to trace or verify some of the key information. As a result of our findings, we recommend that the agencyís planned visit to the study sites for June 1997 be CANCELLED.

NDA Withdrawn

Slide 17 of 57

What we see in the submission

Slide 18 of 57

What we find on site

Slide 19 of 57

The explanation

• How about the BP readings?

  • You cannot find them in ìbooksî
  • They are ìrareî but real, and they occur in this country

• Our population is different from the U.S. population and that is why you see consistent BP readings over time (6 months)

Slide 20 of 57

How is Data Falsified

• Physical and/or Lab Exams

  • Documented only by checked blocks in CRF signed by study coordinator and co-signed by PI. No corroborating source documents.
  • Bogus lab results reported

Slide 21 of 57

Clinical Fraud Cases

Fabricated Data

Slide 22 of 57

How is Data Falsified

• Biological Specimens

  • Samples purported to come from a large number of individual subjects actually derived from only a few different subjects.

Slide 23 of 57

The "F"Case

• Robert Fiddes, MD

  • Guilty - Conspiracy to commit an offense against the United States
  • Guilty - Making false statements in matter within jurisdiction of the FDA (18 USC 1001)

• Sentenced - 15 months in jail (see newspaper)

• Fined - $800,000

• Deported

Slide 24 of 57

New York Times May 17, 1999
RESEARCH FOR HIRE: SECOND OF TWO ARTICLES

A Doctor's Drug Studies Turn Into Fraud

By KURT EICHENWALD and GINA KOLATA

Slide 25 of 57

NY Times Allegation Fabrication the undetectable crime?

• "Another study on an antibiotic required that patients have a certain type of bacteria growing in their ear. No problem for Fiddes. He bought the bacteria from a commercial supplier and shipped them to testing labs, saying they had come from his patients' ears."

Slide 26 of 57

NY Times Allegation

• The FDA investigators asked (Fiddes), what evidence of fraud is there in the records reviewed by monitors and the government? What could the watchdogs have seen that would have allowed them to detect his fraud?

Slide 27 of 57

NY Times Allegation

• "Nothing, Fiddes replied. Had it not been for a disgruntled former employee, he would have still been in business."

Slide 28 of 57

Bogus specimens
A new era in detection?

• Chronic Bronchitis Study

  • 84 subjects enrolled, required to produce positive sputum samples to qualify for study
  • Monitor reported that sputum samples were not collected per the protocol (i.e. in presence of the investigator)
  • FDA inspected and sponsor audited to validate data. No direct evidence of a problem with specimen integrity

Slide 29 of 57

Clinical Fraud Cases
detecting the undetectable

• Sponsor performed DNA analysis on sputum and serum samples to put concerns to rest.

• Analysis showed 35 of 84 sets of subject serum and sputum samples did not match

• 26 sputum samples were found to be derived from only 3 subjects

Slide 30 of 57

How is Data Falsified

• Subject Identities

  • Same subject enrolled more than once under two different names and subject numbers
  • Nonexistent subjects created

Slide 31 of 57

Clinical Fraud Cases

Dr. P

• Fabricated laboratory data and eye exams - all subjects
  
• Created data for subjects by using available results from different subjects
  
• Recycled subjects (3x) and invented data (EKG tracings, duplicate lab results)
• Used 2 year old EKG tracings
• Disqualified (Hearing)

Slide 32 of 57

Clinical Fraud Cases

• Dr. K. - 1987

  • 85 cases submitted - ONLY 15 were real subjects
    
  • Faked approximately 70 subjects records
    
  • Never used test article - used marketed products
  • Did not obtain informed consent

• Disqualified

• Sentenced - 400 hours of community service

• Fined - $30,000

Slide 33 of 57

How is Data Falsified

• Drug Compliance Records

  • Returned study drug dumped, 100% compliance in taking test medication recorded in CRF

Slide 34 of 57

Why is Data Falsified? We might be cynical, but we werenít born that way

Reasons why records not available:

• "They were destroyed in a hurricane"

• "They were lost in a boating accident"

  (burglary, robbery or vandalism)
  

• "They were lost in the mail"

• "The mover threw them out"

• "My father-in-law threw them out"

Slide 35 of 57

Why is Data Falsified? We might be cynical but we werenít born that way

• I destroyed the records. You were not informed by the sponsor!!?

• I did not know that the medical charts and CRFs (data) must correspond.

• I do not remember why I changed the dates (visits)

• I care for my subjects, used the "best available local therapy"

• I do not see a problem with re-entering a subject 5 or 6 times (considering them as six different subjects)

Slide 36 of 57

Why is Data Falsified

• Reasons are not always known or clear, and often can only be inferred

  • To qualify ineligible subjects to enroll or continue on the study
  • To please the bosses by filling in the blanks and making the source documents match the CRF
  • For fun and profit

Slide 37 of 57

Why is Data Falsified?

• "Edged into it"- rarely primary intent

• Not enough time or loss of interest

• Staff turnover

• Not enough subjects

• For "good of subject"

• Pressure to publish

• Power & prestige

• Money

Slide 38 of 57

Who Does It??

• Anyone at investigator's site who has access to data

  • Principal Investigator
  • Study coordinator
  • Research Subjects
  • Study Nurse
  • ?????????

Slide 39 of 57

NY Times Allegation
THENURSEDIDIT Maneuver The last refuge of a scoundrel

• "Dr. Fiddes replied that they were going to blame the study nurse for all of the problems, and he was going to say he had no knowledge of what was going on."

Slide 40 of 57

[Pie Chart]

The Blame Game: Who gets the blame?

39% -- Study Coordinator

17% -- Nurse

9% -- Hospital

9% -- Sponsor

9% -- Self

9% -- Office Staff

4% -- Sub-investigator

4% -- Monitor

n (parties blamed) = 23

n (cases) = 20

Slide 41 of 57

QA's Role in Dealing with Misconduct

• Prevention

  • Identify and eliminate/minimize risk factors for misconduct

• Detection

  • Monitor and recognize signs of fraud

• Correction

  • Promptly investigate and report fraud

Slide 42 of 57

Tips for Preventing Fraud

• Make sure all study staff have the necessary resources and support needed to accomplish their tasks

• Don't place needless requirements or unreasonable demands on the site

• Monitor sites closely and pay attention to complaints from site personnel

• Minimize the use of enrollment incentives

Slide 43 of 57

Approaches in Detecting Fraud (Rules?)

• Tip/nose/experience/gut feeling

• Understand the protocol or plan

• Parameters that determine eligibility (inclusion/exclusion criteria)

• Accept no copies - originals

• Identify all source/supporting documents

• Look for mistakes/changes/inconsistencies

• Frequency of changes - identify who made them

• Determine & document if changes are justified

• Challenge suspicious data

• Ask for all information (data) pertinent to the study (CRF, AB, CPF, MV-MR, X-R, EKG, DIA, etc.)

Slide 44 of 57

Tips for Detecting Serious Misconduct

• Get Technical--Read x-rays, EKGs, lab results, don't just inventory

• Fill in the Blanks--Question missing dates, times, information, offer to retrieve records yourself

• Donít be intimidated--tell the emperor he has no clothes

Slide 45 of 57

Tips for Detecting Serious Misconduct

• Donít shoot the messenger--believe the monitor, put the burden of proof on the CI

• Be suspicious of blame shifting--tell CI he/she is totally responsible for the conduct of the study

• Expect Fraud--Start from the assumption the records are bogus and the study is a fraud, and work back

Slide 46 of 57

Tips for Detecting Serious Misconduct

• Cultivate Whistleblowers--establish rapport with study staff, be approachable and available, listen to grievances, observe working conditions

Slide 47 of 57

[Pie chart]

Who Complains?--Everybody!
FY 2000

n = 118

14% -- Sponsors

11% -- Anonymous

9% -- Government Agencies

8% -- Private Citizens

19% -- IRBs

9% -- Health Professionals

3% -- CROs

3% -- Study Coordinators

8% -- Former Employees

5% -- Others

2% -- Universities

2% -- Informants

8% -- Subjects

Slide 48 of 57

What Are They Complaining About? Everything!!

• Failure to follow the protocol (70)

• Falsification (67)

• Informed Consent Issues (55)

• Failure to report adverse events (40)

• Qualifications of persons performing physicals (27)

• Inadequate Records (25)

• Failure to get IRB approval, report changes in research (20)

• Failure to follow FDA regulations (13)

• Charging for the test article (9)

• Drug accountability (7)

• No active IND (7)

• Violations of GLP regs (7)

• Misleading advertisements (5)

• Blinding (3)

• No 1572 (2)

• Monitoring practices (2)

• IRB shopping (1)

Slide 49 of 57

[Pie chart]

Clinical Investigator Inspections*

52%-- VAI

44% -- NAI

3% -- OIA

0.3% --Pending

n = 306

* All inspections, domestic and foreign

Slide 50 of 57

Complaints Inspected and Their Classifications

First Pie Chart Shows:
Inspected = 26%
Not Inspected = 74%
n= 105
 
Second Pie Chart Shows:
OAI = 26%
NAI = 26%
VAI = 52%
n= 27

Slide 51 of 57

[Picture of Uncle Sam]

Correcting Misconduct

How You can help

FDA Wants You to Recognize and Report Misconduct

Slide 52 of 57

FDA Believes Sponsors should Promptly Report...

• Any information they have that any person involved in human subject trials committed research misconduct

• Whenever the sponsor discovers misconduct

• Not just for clinical investigators and not just when a clinical investigator is terminated

Slide 53 of 57

Correcting Misconduct

• Reporting Research Misconduct

  • Name of the person(s)
  • Contact information
  • Specific identity of the affected research
    • IND/IDE #, protocol, study title, and study dates
  • As much information regarding the research misconduct as is available to the sponsor.

Slide 54 of 57

Where to Report Misconduct

Drugs

• Division of Scientific Investigations (HFD-45), Office of Medical Policy, Center for Drug Evaluation and Research, FDA, 7520 Standish Place, Room 103, Rockville, Maryland 20855-2773, (301) 594-0020, fax (301) 594-1204.

Slide 55 of 57

Where to Report Misconduct

Biological Products

• Office of Compliance and Biologics Quality, Division of Inspections and Surveillance, Center for Biologics Evaluation and Research, (HFM-650), FDA, 1401 Rockville Pike, Room 400S, Rockville, Maryland 20852-1448, (301) 827-6221, fax (301) 443-6748.

Slide 56 of 57

Where to Report Misconduct

Medical Devices

• Office of Compliance, Division of Bioresearch Monitoring, (HFZ-310), Center for Devices and Radiological Health, FDA, 2098 Gaither, Room 130, Rockville, Maryland 20850, (301) 594-4718, fax (301) 594-4731.

Slide 57 of 57

Tips for Dealing with Serious Misconduct

Be Prepared--have a system in place to capture, document and deal with complaints of misconduct in a timely fashion. Follow your SOPs!!!

GCP Home | Educational Materials
FDA Home Page | Search | A-Z Index | Site Map | Contact FDA

FDA/Website Management Staff
Web page created by clb 2001-DEC-07.