The Hip That Cannot Be Measured
An 80-year-old woman with severe kyphoscoliosis and a prior right total hip arthroplasty is referred for DXA. She is 148 cm tall and weighs 44 kg. On the table, she cannot lie flat without pain, and her scoliosis makes positioning difficult. The left hip is scanned; the femoral neck region is partly obscured by hardware artifact from a contralateral procedure and the technologist notes the left hip is 'not evaluable.' The spine scan shows severe degenerative change throughout, with only L1 considered evaluable. The printed report shows a femoral neck T-score of −2.9 derived from a region the technologist says is invalid.
1. What are the valid diagnostic options when neither the spine nor the hip can be reliably measured?
Read the worked answer
When the spine and hip are both non-evaluable, the ISCD approach is:
- Forearm (33% radius) as an alternative site, using the appropriate reference database for the non-dominant forearm, when spine and hip cannot be used. This is the standard fallback.
- If forearm is also not feasible, the diagnosis may rest on clinical grounds - prior fragility fracture, secondary causes, and other risk factors - rather than a DXA T-score.
- A vertebral fracture assessment (VFA) or lateral spine imaging can identify prevalent vertebral fractures, which change management independent of T-score.
What is not acceptable is reporting a T-score from a region the technologist has identified as invalid. The printed −2.9 at the femoral neck must be discarded and the reason documented. The report should state that hip and spine are non-evaluable, describe why, and present whatever valid data exist (e.g., L1 alone, with a caveat that a single vertebra is insufficient for diagnosis).
2. How should the report be structured, and what do you tell the referring clinician about treatment?
Read the worked answer
Report structure:
- State clearly which sites were attempted and which are non-evaluable, with reasons (scoliosis, hardware artifact, positioning).
- Do not present invalid T-scores as findings. If they appear on the manufacturer printout, explicitly disown them in the interpretation.
- Report the forearm result if obtained, with the correct reference database.
- Report L1 with a note that a single evaluable vertebra is insufficient for diagnostic classification.
- Recommend VFA/lateral imaging if not already done.
Treatment: This patient has multiple strong clinical indications independent of DXA - age 80, low BMI (20.1), likely prior fracture history (to be confirmed), and severe kyphoscoliosis. Treatment should not be withheld because the DXA is uninterpretable. Conversely, the invalid −2.9 should not be used to justify a particular drug class or to label her 'severe osteoporosis' in a way that drives therapy selection. The clinician should treat on clinical risk (e.g., FRAX with BMD omitted, or clinical judgment), and the DXA report should support that rather than obstruct it.