PIWI Rules for Older Adults After Ground-Level Falls
Key Takeaways
- In older adults aged 65 years or older who underwent head CT after a ground-level fall with mild traumatic brain injury in a prospective multicentre cohort, 4.4% had significant acute traumatic intracranial hemorrhage and 0.3% underwent urgent neurosurgical intervention.
- Visible forehead-scalp impact, GCS below baseline, focal neurological deficit, signs of basal skull fracture, acute confusion, vomiting, loss of consciousness, and headache were identified as bedside predictors associated with significant hemorrhage.
- PIWI 1 and PIWI 2 each showed 100% sensitivity for significant acute traumatic intracranial hemorrhage, with specificities of 43.6% and 25.3%, respectively.
- In this derivation cohort, PIWI 1 and PIWI 2 would have reduced CT use by 41.7% and 24.2%, and antiplatelet and anticoagulant use were not associated with higher hemorrhage risk in bivariate analysis.
In the Dubucs et al prospective multicentre cohort study of PIWI head CT rules in older adults after ground-level falls, investigators enrolled adults aged 65 years or older across five university-affiliated emergency departments in France, Monaco, and Slovakia from 1 July 2023 to 30 June 2025. Eligible patients presented within 24 hours of a ground-level fall, had suspected or confirmed mild traumatic brain injury—defined as head trauma with GCS 13 to 15, usually based on reported head impact or an external sign of head trauma—and underwent head CT; patients taking antiplatelet or anticoagulant therapy were not excluded. Predictors were recorded before CT ordering with assessors blinded to outcome, and the primary outcome was significant acute traumatic intracranial hemorrhage defined as a Neuroimaging Radiological Interpretation System (NIRIS) score greater than 1. Derivation combined bivariate analysis, logistic regression, recursive partitioning, interobserver reliability assessment, and 500-sample bootstrapping with complete-case analysis and no imputation to produce bedside rules from history and physical examination features.
Significant acute traumatic intracranial hemorrhage occurred in 72 of 1,620 patients (4.4%, 95% CI [3,5]); urgent neurosurgical intervention occurred in 5 of 1,620 patients overall (0.3%) and in 5 of 72 hemorrhage cases (6.9%, 95% CI [2,15]). In the predictor and rule-performance findings for PIWI 1 and PIWI 2, both rules had 100% sensitivity (95% CI [95,100]), with specificities of 43.6% (95% CI [41,46]) for PIWI 1 and 25.3% (95% CI [23,28]) for PIWI 2. Logistic regression identified eight predictors and had an area under the receiver operating characteristic curve of 0.91, and scalp impact versus no visible head impact had an odds ratio of 2.8 (95% CI [1,5]) in bivariate analysis. All candidate variables except amnesia showed good interobserver reliability, although agreement testing was feasible in only 69 patients (4.3% of the cohort); neither antiplatelet nor anticoagulant use was associated with increased risk in bivariate analysis, and the two rules sorted patients into lower- and higher-risk groups within the derivation cohort.
Interpretation remains limited to patients who underwent head CT during the emergency visit, making selection bias possible, and local CT interpretation without centralized independent rereading leaves room for misclassification. The NIRIS-based outcome definition had not been validated in older populations, and delayed bleeding after the initial scan or neurosurgical intervention beyond the 7-day follow-up window could have been missed. The authors also noted that the observed 72 events fell short of the 80 anticipated in sample-size planning, which could increase overfitting risk.
Bedside history and physical examination features, including visible head impact location, were associated with significant acute traumatic intracranial hemorrhage after ground-level falls in older adults. Within this cohort, the derived PIWI rules were highly sensitive, and PIWI 1 relied on history plus examination whereas PIWI 2 may offer an examination-based alternative when symptoms or fall history are harder to assess. The authors indicated that outside validation is still required, but the reported signal was that simple bedside findings tracked with clinically important intracranial bleeding in this scanned emergency department population.
Clinician Questions
Which older adults do the PIWI rules apply to after a ground-level fall?
The PIWI derivation cohort included adults aged 65 years or older who presented within 24 hours of a ground-level fall, had mild traumatic brain injury defined by GCS 13 to 15, and underwent head CT in five emergency departments in France, Monaco, and Slovakia. Older adults who were evaluated after a fall but were not scanned were not part of the cohort, so applicability does not automatically extend to all older fall patients seen in emergency care.
How was significant acute traumatic intracranial hemorrhage defined in the PIWI derivation study?
Significant acute traumatic intracranial hemorrhage was defined as a head CT finding with a NIRIS score greater than 1, including lesions such as epidural hematoma, subdural hematoma, parenchymal hematoma, parenchymal contusion greater than 0.5 mL, mild or moderate hydrocephalus, midline shift, intraventricular hemorrhage, or diffuse axonal injury. That endpoint focuses the rules on radiographically important bleeding, and the authors noted that NIRIS has not been validated specifically in older populations.
Why were two PIWI rules derived instead of one after mild traumatic brain injury in older adults?
PIWI 1 incorporated eight predictors from fall history and physical examination, whereas PIWI 2 used six predictors from physical examination alone and excluded loss of consciousness and headache. The authors framed PIWI 2 as a pragmatic alternative when fall history or symptoms are difficult to assess, which can occur in older adults with cognitive impairment or unclear histories.