Osteoporotic Vertebral Fracture Prevalence Varies Widely
Bibliographic record
Abstract
To The Editor, In a recent issue of JBMR, we read with interest “Osteoporotic Vertebral Fracture Prevalence Varies Widely Between Qualitative and Quantitative Radiological Assessment Methods: The Rotterdam Study” by Oei and colleagues1 and “Comparative Analysis of Radiology of Osteoporotic Vertebral Fractures in Women and Men” by Lentle and colleagues,2 as well as the Commentary that follows these two articles by Pawel Szulc.3 Oei and colleagues point out the different methods for the radiological assessment of vertebral fractures, and comment on the lack of a gold standard. They used existing radiological images and two different types of analysis to distinguish between significant and mild vertebral deformities. Lentle and colleagues, on the other hand, comment on cross-sectional longitudinal observations from the Canadian Multicentre Osteoporosis Study. They too compared two methods for the osteoporotic vertebral fracture assessment of existing and lateral spine radiographs and concluded that defining vertebral fractures by mABQ (modified algorithm-based qualitative approach) is preferred to the GSQ (Genant semiquantitative technique) for clinical assessments. The authors of both studies used large groups of patients and undertook a time- and resource-intensive analysis.1, 2 Pawel Szulc further comments that the identification of vertebral fractures is a major health problem and suffers from misleading and inaccurate terminology used in radiological reports. In New Zealand, we took a different approach to our patient analysis. We used existing images of patients, available electronically through a shared repository imported to HCAS (Healthcare Analytic Systems). The Canterbury District Health Board has legislative obligations to monitor the delivery and performance of services, as well as to provide information to the Ministry of Health, including identifying information essential for the purposes for which the information is sought. This is designed to assess whether patients with fractures are receiving appropriate care. We then used a clinical health ontology: SNOMED CT (https://www.snomed.org/). SNOMED is a comprehensive and precise clinical health terminology product that enables us to search and codify terms and small sentences within clinical notes and radiological reports; it incorporates words or a combination of words such as anterior, wedge, fracture, vertebra, closed, thoracic, lumbar, or collapse. We used this software to extract those individuals with clearly identified thoracic and lumbar vertebral fractures (https://www.health.govt.nz/nz-health-statistics/classification-and-terminology/new-zealand-snomed-ct-national-release-centre/snomed-ct). These were then visually reviewed online and appropriate therapy and or investigation instituted. Using this approach, we identified all vertebral fractures: Approximately 30% of patients were already on some form of treatment or were being investigated; 30% were not on any treatment and were followed up by the Fracture Liaison Service with appropriate investigations, radiology, and treatment; another approximately 30% were deemed not to have osteoporotic vertebral fractures based on a number of additional clinical factors or because of wrong classification. This approach was endorsed by New Zealand's Health Information Standard Organisation (https://www.health.govt.nz/our-work/ehealth/digital-health-sector-architecture-standards-and-governance/health-information-standards) as the clinical terminology to be used across the health and disability sector within New Zealand. The use of digital images and clinical notes in our health system has streamlined the approach to diagnosing and treating those people with vertebral fractures without the use of significant additional resources. All authors state they have no conflicts of interest.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.014 | 0.072 |
| Meta-epidemiology (narrow) | 0.000 | 0.001 |
| Meta-epidemiology (broad) | 0.002 | 0.001 |
| Bibliometrics | 0.004 | 0.006 |
| Science and technology studies | 0.001 | 0.002 |
| Scholarly communication | 0.004 | 0.003 |
| Open science | 0.002 | 0.002 |
| Research integrity | 0.006 | 0.007 |
| Insufficient payload (model declined to judge) | 0.005 | 0.002 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".