Why validation matters
NDTMS datasets are validated against the current specification before they are accepted. A submission that fails validation is not counted until it is corrected and resubmitted, so validation errors are the single most common cause of last-minute submission stress for treatment services.
The good news is that the errors recur. The same handful of issues account for the large majority of failed records, and once a service understands them, most can be prevented rather than fixed under deadline pressure.
Missing or overdue TOPs
By far the most common issue is TOPs that are absent when they should be present — a start TOP not recorded at the beginning of an episode, or a review or exit TOP that fell due and was never completed.
Fix it by running an overdue-TOP report ahead of the deadline and clearing the backlog while clients are still in contact. Prevent it by prompting keyworkers when a TOP falls due rather than relying on memory.
Episode dates that do not line up
Validation frequently flags episodes whose start or discharge dates are inconsistent — a discharge before the recorded start, an intervention dated outside its episode, or an episode left open long after the client actually left.
Fix it by reconciling episode start and discharge dates against the actual admission and discharge events in the case record. Prevent it by closing episodes as part of the discharge process, not as a separate data task weeks later.
Duplicate or inconsistent client identifiers
Where the same client appears under more than one identifier — often after moving between services or being re-referred — records can duplicate or fail to link, distorting counts and breaking the treatment journey.
Fix it by reconciling identifiers before submission and merging genuine duplicates. Prevent it with a consistent client-matching process at referral so the same person is recognised across services.
Orphaned or misattached SIRs
Sub Intervention Reviews that are attached to the wrong intervention, or left incomplete after an intervention has ended, are a recurring validation problem because a SIR must relate correctly to its intervention.
Fix it by checking that each active client's SIRs are linked to the correct intervention and that none are left open past the intervention's end. Prevent it by having the system enforce the intervention-to-review relationship.
Specification updates not yet reflected locally
When OHID updates the dataset specification, field mappings and validation rules change. Services relying on spreadsheet macros or a static export template often discover the change only when a submission fails against the new rules.
Fix it by updating your mappings to the current specification. Prevent it by using a system where specification updates are maintained for you, so your export always matches the rules in force. GreenShoots validates every required field against the current specification before you export, so most of the errors above are caught while records can still be corrected.