Healthcare Asset Data Quality Scoring: How Struktive Scores Biomedical Equipment Records
The seven scoring dimensions, clinical criticality weighting, and what a score of 80+ means for CMMS import readiness
Healthcare asset quality scoring is more complex than DC or mining scoring because clinical criticality changes the stakes. A ventilator with a missing serial number is a higher-priority gap than a desk fan with the same issue. This post explains how Struktive's healthcare scoring model works and what the scores mean for CMMS import readiness.
Key Takeaways
- Healthcare quality scoring weights clinical criticality — records for life-critical equipment are scored more strictly than records for non-clinical assets.
- The seven scoring dimensions are: manufacturer completeness, model completeness, serial number completeness, location completeness, GMDN classification, UDI cross-reference, and maintenance history.
- A score of 80+ indicates the record is ready for CMMS import. A score below 60 indicates the record requires manual review before import.
- The score is not a measure of how 'good' the original data was — it is a measure of how complete and reliable the normalised output is.
Why Healthcare Scoring Is Different
Data quality scoring for healthcare asset registers is more complex than scoring for data centre or mining equipment registers. The reason is clinical criticality.
In a data centre, a server with a missing serial number is a data quality gap — it makes asset tracking harder and DCIM import messier. In a healthcare setting, a ventilator with a missing serial number is a patient safety risk. Recall notifications cannot be delivered. Mandatory maintenance schedules cannot be enforced. Regulatory reporting is incomplete. The consequences of the same data quality gap are categorically different depending on what the asset is.
Struktive's healthcare scoring model accounts for this by applying a criticality multiplier to each record. Life-critical equipment — ventilators, infusion pumps, defibrillators, anaesthesia machines — is scored more strictly than non-clinical assets. A ventilator needs a higher score to be considered import-ready than a desk fan.
The Seven Scoring Dimensions
Healthcare quality scores are calculated across seven dimensions.
Manufacturer completeness (15 points) — is the manufacturer field populated, and does it match a known manufacturer in the GMDN or GUDID database? A confirmed manufacturer scores 15. An unrecognised manufacturer name scores 8. A blank manufacturer field scores 0.
Model completeness (15 points) — is the model field populated, and does it match a known model for the confirmed manufacturer? A confirmed model scores 15. A partial match (abbreviation that was expanded) scores 12. An unrecognised model scores 6. A blank model field scores 0.
Serial number completeness (20 points) — is the serial number field populated, correctly formatted, and unique? A confirmed serial number scores 20. A serial number that was reformatted during normalisation scores 16. A serial number that appears to be a duplicate scores 8. A blank serial number scores 0. For life-critical equipment, the serial number dimension is weighted at 25 points.
Location completeness (15 points) — is the location field populated and structured? A fully structured location (site, building, floor, department, room) scores 15. A partially structured location scores 8. A free-text location that could not be parsed scores 3. A blank location scores 0.
GMDN classification (15 points) — has the record been assigned a GMDN code? A GMDN code confirmed via UDI lookup scores 15. A GMDN code assigned via machine learning classification scores 10. No GMDN code scores 0.
UDI cross-reference (10 points) — has the record been cross-referenced against FDA GUDID? A successful UDI lookup scores 10. A partial match (manufacturer confirmed but model not found) scores 5. No UDI data available scores 0.
Maintenance history (10 points) — is there any maintenance history associated with the record? A record with at least one maintenance event scores 10. A record with no maintenance history scores 0. This dimension is informational — it does not block CMMS import, but it affects the overall data quality picture.
Try Struktive on your own data
Upload a raw asset CSV and get back a normalised, DCIM-ready file in minutes. No account required.
Score Interpretation
| Score range | Interpretation | Recommended action |
|---|---|---|
| 90–100 | Excellent | Import ready. No manual review required. |
| 80–89 | Good | Import ready. Minor gaps (typically maintenance history or UDI) can be completed post-import. |
| 70–79 | Acceptable | Import ready with caveats. Review flagged fields before import. |
| 60–69 | Marginal | Manual review recommended before import. Gaps may affect compliance reporting. |
| 50–59 | Poor | Manual review required. Do not import until critical gaps are resolved. |
| Below 50 | Not import-ready | Record requires significant remediation. Import will create incomplete or inaccurate CMMS records. |
The Criticality Multiplier
For life-critical equipment, Struktive applies a criticality multiplier that raises the minimum import-ready threshold from 70 to 80. The multiplier also increases the weight of the serial number dimension from 20 to 25 points, reflecting the importance of serial number completeness for recall management and regulatory reporting.
The criticality classification is based on GMDN category. Life-critical GMDN categories include: ventilators (34938), defibrillators (35956), infusion pumps (13089), anaesthesia machines (17069), and patient monitors (57987). The full list of life-critical GMDN categories is available in the Struktive documentation.
Using Scores to Prioritise Remediation
The quality score is most useful as a prioritisation tool. Rather than reviewing every record manually, focus remediation effort on:
- Life-critical equipment with scores below 80 — these are the highest-priority gaps from a patient safety and compliance perspective.
- Any equipment with a blank serial number — serial number is the most common gap and the most impactful for recall management.
- Records with unrecognised manufacturer or model names — these will fail CMMS import and need to be resolved before the import can proceed.
The gap report that Struktive produces alongside the normalised dataset lists all records below the import-ready threshold, grouped by gap type and sorted by criticality. This report is the starting point for the manual review phase of any healthcare asset normalisation project.