Basic Data Collection Elements in Cancer Clinical Trials

1
Basic Data Collection Elements in Cancer
Clinical Trials

• Julia Challinor, RN, PhD
• University of California, San Francisco
• INCTR Annual Conference
• 10-12 December 2005
• Chennai, India

2
U.S. National Cancer Institute Cancer Therapy
Evaluation Program (CTEP) General Guidelines

• Do not abbreviate
• For example drug names, can be misinterpreted
• Use methotrexate not MTX
• Do not use brand names
• Generic names are standardized
• Extraneous information is not helpful
• If using only injectable drugs, do not include
  instructions for oral doses

(http//ctep.cancer.gov/handbook/append_16.html,
retrieved 3/8/04)
3
CTEP

• Use consistent notation
• Use EITHER qid OR Q6h do not use both
• Spell out the word units
• A u can look like a zero
• Never put a zero after a decimal point
• Use 2 NOT 2.0
• A decimal point can be missed and result in a
  10-fold dose increase

(http//ctep.cancer.gov/handbook/append_16.html,
retrieved 3/8/04)
4
CTEP

• If using units that are less than the number one,
  then a zero should precede the decimal point
• 0.50mg should be used, NOT .50mg
• Otherwise the decimal might be missed and result
  in a 10-fold dose increase
• Calculation of body weight should be specified
  and formula used for calculation should be
  included
• Identify if actual, ideal or lean body weight is
  used to calculate drug dosage

(http//ctep.cancer.gov/handbook/append_16.html,
retrieved 3/8/04)
5
CTEP

• Specify total number of days a drug is given and
  include cycle days when treatment occurs.
• Specify contiguous treatment days and
  non-contiguous treatment days
• Specify cycle (or course) duration
• Clearly identify the duration of the
  administration of the drug
• If given more than once a cycle, clearly describe
  the cycle days when the drug is given

(http//ctep.cancer.gov/handbook/append_16.html,
retrieved 3/8/04)
6
CTEP

• Be clear about total dose per treatment course
• Use total dose as function of body weight or
  surface area
• If appropriate, describe administration starting
  days and times.
• Using a 24 hour clock notation, i.e. 1430 avoids
  errors using am and pm.

(http//ctep.cancer.gov/handbook/append_16.html,
retrieved 3/8/04)
7
Vocabulary

• Toxicity
• Toxicity is NOT clearly defined by regulatory
  organizations
• Toxicity has been described as an adverse event
  that has an attribution (the relationship to
  investigational agent) of possible, probable or
  definite
• If the study specifies using a specific Adverse
  Event criteria (i.e. NIH) do NOT change the
  criteria during the course of data collection

(http//ctep.cancer.gov/handbook/append_16.html,
retrieved 3/8/04)
8
Adverse Event (AE)

• Definition
• A negative experience encountered by an
  individual during the course of a clinical trial,
  that is associated with the drug. An AE can
  include previously undetected symptoms, or the
  exacerbation of a pre-existing condition. When
  an AE has been determined to be related to the
  investigational product, it is considered an
  Adverse Drug Reaction.
• Ginsberg, D, ( 2002) The Investigators Guide to
  Clinical Research, 3rd ed., p 283

9
Adverse Event (AE) Terminology

• Grading is from 1 to 5
• Grade 1 Mild AE
• Grade 2 Moderate AE
• Grade 3 Severe AE
• Grade 4 Life-threatening or disabling AE
• Grade 5 Death related AE

10
Adverse Event Grading Example
Adverse Event Short Name Grade 1 Grade 2 Grade 3 Grade 4 Grade 5
Neutrophils/ granulocytes (ANC/AGC) Neutro-phils ltLLN-1500/mm3 ltLLN-1.5x109/L lt1500-1000/mm3 lt1.5-1.0x109/L lt1000-500/mm3 lt1.0-0.5x109/L lt500/mm3 lt0.5x109/L Death
ANC is Absolute Neutrophil Count LLN is Lower
Limit of Normal
11
Potential Problems

• Terms must be defined in study PROTOCOL
• Examples
• Date of Registration
• Is this the first day the patient was seen?
• Is this the date they were given a medical record
  number?
• Is this the date they were enrolled in the study?

12
Potential Problems

• Date of Informed Consent
• Is this the date the patient/parent signed the
  consent?
• Date of Enrollment
• Is this the date the patient signed the consent?
• Is this the date the patient began therapy?

13
Potential Problems

• Date of Birth
• This is NOT the age of the child
• The date of birth MUST be consistent
• In the U.S. it is standard to use month/day/year
• In Europe and Latin America it is standard to use
  day/month/year
• Ideally, the data form or case report form will
  be designed with the format CLEARLY indicated
• DOB (DD/MM/YYYY)
• Best to use all 4 numbers for year, i.e. 1953, or
  2004

14
Potential Problems

• Date Abandoned care
• Missed one appointment?
• Missed two appointments?
• Refused treatment at diagnosis?
• Went to a different center for treatment?
• Disappeared for 2 months and then returned?

15
Potential Problems

• Subject Identifiers (subject ID)
• Many hospitals have their own patient
  identification numbers
• Using the hospital patient identification number
  as the subject ID can provide a path back to the
  patient for verification of data entry if a
  future problem or question arises.

16
Classification

• Disease classification should be standardized
  among protocol collaborators
• Once a classification system has been selected,
  all participants should use identical diagnostic
  procedures as outlined in the study PROTOCOL

17
Empty Fields

• If a data field is EMPTY then it is assumed that
  the data entry was incomplete and the event did
  NOT occur!
• ALL fields should have data input
• If the data is not applicable then it should be
  noted as N/A (or by the chosen notation for not
  applicable)
• If the data is unknown then it should be noted
  as UNK (or by the chosen notation for unknown)

18
History Log

• If there is a need to deviate from the case
  report form fields for ANY reason, there must be
  a history log that explains the reasoning.

19
Problems

• Physicians using their own style to make
  clinical notes
• Language is not uniform
• Patient information detailed is different from
  physician to physician
• Nutritionist uses different abbreviations for
  medical terms than nursing or physicians.

20
Problems

• Just because the protocol states that a
  patient/subject should receive a specific
  chemotherapy/medication on a specific date within
  their cycle does NOT mean they received it!
• the medication on that date
• the dose specified on the protocol
• the chemotherapy/medication AT ALL
• e.g., physician orders are not the source
  document for chemotherapy received by patients in
  reality

21
Problems

• NO white out
• NO colored pens
• NO pencil
• NO invented vocabulary
• NO acronyms or unique site specific abbreviations

22
Problems

• Healthcare staff does not understand or believe
  in the importance of data collection
• Too little time for careful data entry
• Patient demands too high for good data collection
• Charts are missing or moved
• Temporary staff working on the oncology unit
• Have not received training regarding clinical
  research protocol based treatment basics

23
Data is important, it helps us get from here
to.there!