Diagnostic prediction models for spinal fractures in individuals with spinal pain or trauma: a systematic review and meta-analysis
Bibliographic record
Abstract
Background: Multivariable diagnostic models are often used to identify spinal fractures in patients with spinal pain and/or trauma. However, their performance and clinical utility remain uncertain. We aimed to evaluate the performance of diagnostic models for detecting spinal fractures in individuals with spinal pain and/or trauma. Methods: In this systematic review and meta-analysis, we searched MEDLINE, EMBASE, and Web of Science on April 15, 2024 and May 27, 2024 for relevant work published since database inception. The first search included only studies on spinal pain and the second additionally included spinal trauma studies, following a protocol adjustment during screening. A search update was performed on May 19, 2025. An expert librarian assisted in developing the search strategy, which was limited to work published in English, Italian, and Dutch. We also performed backward and forward citation tracking. We included studies that developed and/or externally validated multivariable diagnostic prediction models for spinal fractures. Two independent reviewers screened studies for eligibility, extracted data using the CHARMS checklist, and assessed the risk of bias using the PROBAST. The certainty of evidence was evaluated using the GRADE approach. The protocol was registered in PROSPERO, CRD42024539898. Findings: We included 27 studies encompassing 34 diagnostic models. All models showed an overall high risk of bias, while the concerns about their applicability varied due to the frequent use of spinal injuries as the outcome instead of explicitly addressing spinal fractures. Meta-analyses of ten studies that externally validated the Canadian C-spine Rule in adults presenting with trauma to emergency departments or trauma centres demonstrated, with very low certainty of the evidence, excellent sensitivity (0.999; 95% CI 0.976-1), an high area under the curve (0.850; 95% CI 0.720-0.970), and a low specificity (0.188; 95% CI 0.063-0.443). We estimated a pooled non-statistically significant positive likelihood ratio of 1.230 (95% CI 0.978-1.548) and a negative likelihood ratio of 0.007 (95% CI 0.001-0.082) for the same model. Other models for traumatic cervical fractures and osteoporotic fractures showed promise but lacked external validation or sufficient reporting on calibration and discrimination measures (with low to very low certainty of the evidence). No models for thoracolumbar fractures were deemed ready to be used clinically. Interpretation: Although the Canadian C-spine Rule shows potential for screening traumatic cervical fractures, the very low to low certainty of the evidence limits confidence in its accuracy and appropriateness for clinical use. We did not identify any externally validated models suitable for clinical use regarding osteoporotic or traumatic fractures of the thoracolumbar spine, and traumatic fractures of the cervical spine in non-emergency settings. Future research with rigorous methodological and statistical approaches should aim to fill these knowledge gaps. Funding: None.
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 | 0.043 | 0.104 |
| Meta-epidemiology (narrow) | 0.004 | 0.002 |
| Meta-epidemiology (broad) | 0.021 | 0.064 |
| Bibliometrics | 0.015 | 0.012 |
| Science and technology studies | 0.001 | 0.001 |
| Scholarly communication | 0.005 | 0.003 |
| Open science | 0.004 | 0.002 |
| Research integrity | 0.003 | 0.003 |
| Insufficient payload (model declined to judge) | 0.004 | 0.000 |
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".