This appendix describes the process by which patients are automatically selected for Partial Source Data Verification (PSDV).
You can set RDC to automatically select patients for 100% SDV based on an initial count, auto-select rate, or both. This quota of patient data to be verified can be specified in your study site patient SDV plan.
In order for auto-selection to work, the following conditions must be met:
A study site patient SDV plan must be published in which an initial patient count, an auto-select rate, or both are specified. For more information, see "Using Source Data Verification".
The Pending Patient Updates job must be scheduled, at an interval of your choosing, to check for newly eligible patients and to process those patients marked for auto-selection. For more information, see Schedule Pending Patient Updates in the Oracle Clinical Administrator's Guide.
For more information, also see Setting Up Partial Source Data Verification in Oracle Clinical Creating a Study.
Conceptually, SDV auto-selection uses patient pools that correspond to a patient's current SDV selection status and eligibility status. You can see the current Selection and Eligibility statuses for each patient in the published version of the SDV plan.
You can also see current patient statuses in a draft version of the patient SDV plan if there is one in progress and you also have a published patient SDV plan. The draft version lists all patients assigned to the site, while the published version lists only those patients already selected for SDV.
Patient pools and their corresponding patient statuses are as follows:
Exclusion Pool — Contains patients explicitly excluded by the patient SDV plan, which may have been excluded individually or through an import list. These patients may or may not be eligible. The Selection status in the patient SDV plan can be either Excluded or Import Excluded.
Manual Pool — Contains patients explicitly selected in the RDC Onsite SDV Plan, either individually or in an import list. These patients may or may not be eligible. The Selection status in the patient SDV plan can be either Selected or Imported.
Newly Eligible Pool — Contains patients that became eligible since the last time auto-selection was executed. For a patient to be included in this pool, SDV Eligibility must show a valid date. The Selection status in the patient SDV plan can be null, an Exclusion Pool status, or a Manual Pool status.
Initial Pool — Contains patients which have been auto-selected for SDV as one of the first patients to have become eligible at the site. The number of patients to be included in this pool is specified in the patient SDV plan as the Number of Initial Patients. For a patient to be included in this pool, SDV Eligibility must show a valid date. The Selection status in the patient SDV plan is Initial.
Auto-selected Pool — Contains patients which have been auto-selected based on the auto-select rate specified in the published patient SDV plan. For a patient to be included in this pool, SDV Eligibility must show a valid date. The Selection status in the patient SDV plan is Auto-Selected.
Discard Pool — Contains SDV-eligible patients that went through auto-selection, but were not selected. The Discard pool is revisited if the auto-select rate is increased while the study is in progress. There is no information in the patient SDV plan that can certainly distinguish a discarded patient from a newly eligible patient that is not in the Exclusion or Manual pools because in both cases the Selection status is null and a valid eligibility date is present. While not always accurate, you can assume a patient is discarded if it has a null Selection status and the eligibility date precedes the last run of the Pending Patients Updates job.
Note:
Depending on your methodology for populating patient eligibility dates, you could have a patient become eligible after the last run of the Pending Patient Updates job with a date in the past.Active SDV Pool — Contains patients in the combined Auto-selected and Initial pools, plus those patients in the Manual pool which are SDV-eligible. These patients are further called out in the patient SDV plan with a value of Active in the Active SDV? column of the patient SDV plan. The Active SDV Patients table in the patient SDV plan reports the total count of Active SDV patients, broken down by selection type. These are the patients that are retrieved in RDC Onsite when a search for Patients Requiring 100% SDV is conducted. Patients in this pool must have a valid SDV Eligibility date. Patients in the Active SDV pool can have a Selection status of Auto-Selected, Initial, Selected, or Imported.
Auto-selection runs whenever a patient SDV plan is published and each time the Pending Patient Updates job runs. RDC auto-selection processes the newly eligible patients that are not already manually selected or excluded. Therefore, take note that when you use a combination of manual and automatic selection, the effective percent of patients selected for SDV exceeds the auto-select rate you specify.
Patients are processed in order by their SDV eligibility date. If there are multiple patients with the same eligibility date, the processing order is based on the Oracle RDBMS row ID for the patient. The first patients are added to the Initial Pool until the initial count is reached, if specified. Patients are then selected in a round-robin fashion, based on the auto-select rate specified in the plan. The round-robin cycle number is derived as 100 divided by the auto-select rate. For example, if the auto-select rate is 20%, every 5th patient is selected for SDV. Fractional cycle numbers are rounded down to a whole number.
It is possible for a once eligible patient to become un-eligible. In this case, the patient Eligibility Date and Selection status are updated to null.
When patient eligibility is reversed, auto-selection compensates the Initial or Auto-selected pool as needed. If the patient was in the Initial pool, auto-selection looks at both the Discard and Auto-selected pools to select the patient with the earliest SDV Eligible date. If a patient is selected from the auto-select pool, then that pool is compensated with a patient from the discard pool.
If all data for a patient is soft-deleted or hard-deleted, all eligibility fields for the patient (Eligible Date, Ineligible Date, Proc Vers) are nulled. If the patient was formerly SDV eligible, this change triggers the auto-selection process described in Patient Eligibility is Reversed.
If the patient is manually selected for, or excluded from, SDV then the selection status remains as is.
You can publish a new patient SDV plan version with an increased or decreased initial count or auto-selection rate. When you do this, RDC adjusts the Initial and Auto-selected pools as needed.
If the Initial pool needs to be increased, compensation selection(s) are made from the combined Discard and Auto-selected pools.
If the Auto-selected pool needs to be increased, compensation selection(s) are made from the Discard pool.
If the Initial pool needs to be decreased, patients with the most recent SDV eligibility date are removed from the Initial pool and added to the Discard pool and their Selection status changes to null.
If the Auto-selected pool needs to be decreased, patients with the most recent eligibility dates are removed from the Auto-selected pool and added to the Discard pool. Their Selection status changes to null.
When a patient's study site is updated, the Pending Patient Updates job automatically updates the SDV plans at the old and new sites as follows:
If there is a patient SDV plan in place at the old site, the patient is removed from the old site's plan.
If there is a patient SDV plan in place at the new site, the patient is added to the draft SDV plan for the new site with a null Selection status.
If the new site has a published SDV plan that includes an initial count or auto-select rate and the patient is eligible, the patient is processed for auto-selection at the new site.
When Patient Transfer is used to transfer data from one patient ID to another (Oracle Clinical, Conduct, Security), eligibility and SDV plans may be affected as follows:
For the target patient ID, SDV eligibility field values are transferred along with data. This may trigger auto-select processing as described above.
For the source patient ID, all data is soft-deleted, and processing is performed as described above, unless this is an ongoing transfer. This is to prevent a subsequent operation from transferring artificially nulled eligibility data.
When the patient ID for one or more RDCIs is updated, the source patient is treated as if the CRF has been deleted. A check is made to see if all data for the patient is deleted and, if so, the patient is processed as described above. For the target patient, there is no immediate change. However, if the new patient requires 100% SDV, this CRF appears in RDC as requiring SDV.
If you have an eligibility validation procedure which populates a patient's eligible date field, ineligible date field, or both, based on a date field in the CRF, an update to the field will result in an update to the eligibility date. However, if you use the DCI date for this purpose, the date change is not detected.