Subject Data Story Boards

Links

Stage 1B page | CDISC HL7 Project Page | Stage 2 Page

Class 1 – Sponsor Planned Data

Data collected in accordance with the plan.  The Study Design message conveys (among other things) the plan for data collection, the Study Participation message includes information on the subjects in the study, and thus those two messages imply an expected set of data to be collected. The subject data message needs to be able to convey that data.

1.

Study A1234 is complete and Acme Pharmaceuticals now wants to send to the FDA all the observations recorded for each subject during the study as part of their study report submission. Acme uses the CDISC-HL7 subject data message to provide all the recorded observations, as well as all the derived parameters resulting from those observations, as defined by the CDISC SDTM and ADaM standards. The message contains all important relationships, such as the relationship between an observed and planned assessment (or lack thereof), and the relationship between unplanned assessments and other observations (i.e. physical exam finding of jaundice led to a bilirubin measurement). Those observations that were previously reported in a spontaneous adverse event report are not re-submitted, but rather updated and referenced.

2.

FDA runs some form of check to assess if all planned activities were performed. This requires access to the plan held within the Study Design message and needs to allow for all paths to be evaluated at a high level of detail definition (sufficient to allow for a machine to perform the check)

3.

Planned data being captured but study is high risk and therefore FDA requested that data be provided at regular intervals while the trial is running

Class 2

Data collected in connection with ICSRs.  It should be easy to make connections between data in the ICSR and other data conveyed in the Subject Data message.  At minimum, this probably means that the Subject Data message needs to include unique identifiers for ICSRs.

4.

An SAE occurs and is reported in an ICSR. At the time the study report is submitted and the SubjectData is used the SubjectData message should refer to the ICSR content and not repeat the data.

Class 3

5.

Use Case: Physician captures additional subject-level data during the study but does not report it to the sponsor but it is stored in the EHR. This function allows for this data to be reported at some later date. This would require the FDA to request the data from the investigator via the sponsor For example this is seen via reading of the patient narratives by FDA and allows the FDA to realise that additional data may be available Communication between FDA and sponsor is accomplished using RPS mechanism. Communication between sponsor and the healthcare provider is accomplished using ???

6.

Use Case: Sponsor initiates the collection of additional data from investigators and then provides an update to the FDA. All of the data reported would be in addition to the plan.

Class 4

Subject Data may be conveyed in multiple stages, rather than as a monolithic data transfer.  Therefore, the message must be capable of conveying both new and updated data.

7.

See above use cases Use Case: Periodic sending of SubjectData messages (blinded) and then data is unblinded

8.

Use Case: Periodic sending of SubjectData messages and Clarification process results in a data point being changed

9.

Rolling NDA – might be worth having a use case around this

Others

10.

Use Case: Sometimes the original CRF value is changed by the sponsor (could be the investigator in the case of correcting a data entry error). When this happens, the message needs to state the first and final data value. Further information (the full audit trail) would be provided in a different message if requested.

11.

An drug that is marketed in Europe is being evaluated for marketing in the US. A consumer group claims that the drug is associated with a specific adverse event. A reviewer needs to evaluate patients that were treated with the product, and determine if they have experienced the adverse event. This will require the reviewer to evaluate patients that may not have been previously diagnosed as experiencing the adverse event. Planned and unplanned observations and requests for further info (see other use cases)

12.

Use Case: A product is known to cause a particular adverse event. Depending upon the severity of the adverse event, the effect of the adverse event on the patient can range from minor discomfort to disability or death. A reviewer needs to determine how many patients experienced the more severe manifestations of the adverse event. Message needs to provide links to the related data that was used to perform assessment of severity. These links would need to be pre-defined. Need to detail when these links are created and under what circumstances Message needs, links, planned + unplanned and request further data as per previous use cases

13.

Use Case: An drug that is marketed in Europe is being evaluated for marketing in the US. A consumer group claims that the drug causes a specific adverse event. A reviewer needs to evaluate patients that were treated with the product and experienced the adverse event, and determine if these adverse events can be reasonably explained by factors other than the drug, such as high fever, meningitis, treatment with drugs known to cause the adverse event, or pre-existing conditions. In order to determine causality, the reviewer plans to use reasoning similar to that described by Austin Bradford Hill in his paper “The Environment and Disease: Association or Causation (Proceedings of the Royal Society of Medicine, 58 (1965), 295-300.) Similar notes to Case Review 2

14.

Use Case: A study is conducted in order to determine if a product is safe and effective in a sub-population of patients. The inclusion criteria are constructed so that only patients in the sub-population of interest are enrolled in the study. The reviewer wants to ensure that only patients who met the inclusion criteria were enrolled in the study. Requirement on Study Design message to indicate those observations used to support inclusion / exclusion processing

15.

Estimate mean and variance of subject response in a study cell, and functions of these means and variances. A reviewer wishes to estimate the mean and variance of a continuous response variable (e.g. blood pressure) at one or more times (e.g. visit) in one or more study cells, and calculate functions of these means and variances.

16.

Estimate mean survival time for subjects in a study cell A reviewer needs to estimate the mean survival time to an event (e.g. heart transplant) in a study cell. In order to calculate the mean, the reviewer needs to know if the event happened, and if the happened, when the event happened.

17.

Estimate the baseline value of a subject response An analyst want to estimate the pretreatment value of a patient outcome (e.g. blood pressure). Estimation of this value will be based upon one or more values of the attribute in a study cell prior to the study cell containing study treatment, or from patient history data. Being able to link observations to the study design

18.

Test that a function of the data in one or more study cells is equal to, less than, or greater than a constant. Calculate an analysis of covariance for a continuous outcome measure for study cells in the second epoch of the study. The value at visit 3 is the response variable, and the sponsor-defined baseline score is the covariate. Message needs to allow for linking of the data to the study plan.

19.

Test that a function of the data in one or more study cells is equal to, less than, or greater than a constant. Calculate a Mantel-Haenszel test for study cells in the second epoch of the study. The response variable is categorical (e.g. presence or absence of an adverse event, seriousness of an adverse event). Stratification needs to be done by site, age, sex, and race.