ISCD CBDT cheat sheet

DXA measures areal bone mineral density (aBMD, g/cm²), not volumetric density; aBMD = BMC (g) ÷ projected area (cm²), so larger bones read falsely higher and smaller bones falsely lower.

T-score = (patient BMD − young adult reference mean) ÷ young adult reference SD; Z-score = (patient BMD − age-matched mean) ÷ age-matched SD. WHO osteoporosis threshold is T-score ≤ −2.5; osteopenia is −1.0 to −2.5.

Z-score, not T-score, is the preferred index in premenopausal women, men under 50, and children; a Z-score ≤ −2.0 is defined as "below the expected range for age" and warrants workup for secondary causes.

Each manufacturer's reference database is device- and often software-version-specific; T-scores and Z-scores must be computed from the same manufacturer's normative data as the scanner, never mixed across brands.

Precision is expressed as CV (%) = (SD of repeated measurements ÷ mean) × 100; the least significant change (LSC) at 95% confidence = 2.77 × precision CV for a single measurement comparison.

In vivo lumbar spine precision is typically about 1% CV and total hip about 1.5–2% CV; femoral neck and Ward's region are less precise and should not drive monitoring decisions alone.

Monitoring intervals are set from LSC and expected rate of change: with 1% spine precision (LSC ≈ 2.8%), a 3% annual loss reaches detectability in roughly 1–2 years, whereas stable patients need 2 years or more.

Effective dose is low: roughly 1–10 µSv per DXA site, comparable to a few hours to a day of background radiation and far below a chest radiograph; shielding is not routinely required for the operator.

Daily quality control uses a spine phantom scanned before patient studies; the measured BMD must stay within the manufacturer's control limits (commonly ±1.5% of baseline), and out-of-limit results trigger recalibration or service before clinical use.

Correct positioning: spine AP supine with hips and knees flexed over a block to flatten lumbar lordosis, mid-sagittal line centered, scan from about T12 to L5; hip with the femur internally rotated 15–25° to bring the femoral neck parallel to the table.

Invalid vertebrae for analysis: any vertebra with a T-score more than 1.0 unit different from an adjacent evaluable vertebra is excluded, as are vertebrae with focal structural change, fracture, vertebroplasty cement, or dense surgical hardware; report the minimum of two evaluable vertebrae.

The hip region of interest used for diagnosis is the femoral neck or total hip, choosing the lower of the two; Ward's triangle and the greater trochanter are not diagnostic sites.

The distal one-third radius (33% radius) is the forearm site of choice, used when spine and hip are not evaluable or in hyperparathyroidism and extreme obesity exceeding the table weight limit; it is the preferred site in primary hyperparathyroidism.

Bilateral hip scanning is recommended at baseline; for follow-up, scan the same hip and same region of interest, because side-to-side BMD differences can exceed the LSC.

In vivo soft-tissue artifacts that raise aBMD: aortic calcification, degenerative osteophytes, facet sclerosis, and overlying metal; artifacts that lower it: laminectomy, prior wide surgical resection, and severe scoliosis off-axis positioning.

Body composition from DXA divides the field into bone mineral, fat mass, and lean soft tissue; the three-compartment model assumes constant hydration of lean tissue, so edema or dehydration biases fat and lean estimates.

Vertebral fracture assessment (VFA) uses a lateral spine acquisition with morphometric measurement of vertebral heights; a reduction of 20% or more in anterior, middle, or posterior height defines a prevalent vertebral deformity.

FRAX estimates 10-year probability of major osteoporotic fracture and hip fracture from clinical risk factors with or without femoral neck BMD; treatment is generally considered at 10-year major osteoporotic probability ≥20% or hip probability ≥3%.

DXA uses two X-ray photon energies (typically about 40 keV and 70–100 keV, or equivalent kVp switching); the low- and high-energy attenuation ratio separates bone mineral from soft tissue without a water bath.

Beam geometry: pencil-beam scanners use a single detector with a narrow fan and slower scan; fan-beam scanners use a multi-detector array, scan faster but deliver a slightly higher dose and require stricter centering to avoid magnification error.

Cross-calibration is required when patients move between scanners or after major hardware/software changes; a phantom cross-calibration with a regression equation converts BMD values, and T-scores are never cross-calibrated.

Hounsfield units from CT are not interchangeable with DXA aBMD; QCT measures true volumetric BMD (mg/cm³) and is less affected by degenerative artifact but uses higher dose and lacks a WHO T-score framework.

Quantitative ultrasound of the calcaneus measures speed of sound and broadband ultrasound attenuation; it predicts fracture but is not used to diagnose osteoporosis by WHO criteria or to monitor therapy.

Radiation protection: use the lowest collimation consistent with the region, collimate to the anatomy, keep the operator behind the barrier during acquisition, and apply the ALARA principle; the technologist's annual occupational limit is 50 mSv effective dose (20 mSv averaged over 5 years).

Bone densitometry is contraindicated during pregnancy unless the study is clearly justified and approved; DXA of the spine and hip is generally avoided, and VFA in particular should be deferred.

The forearm scan is performed on the non-dominant arm unless that side has a fracture, hardware, or severe deformity; the patient sits with the forearm pronated and the ulna flat on the table, and the region from the ulnar styloid to the distal third is analyzed.

A DXA report must include the scanner make and model, software version, sites and regions analyzed, BMD in g/cm², T-score and Z-score with the reference database named, and the WHO classification when applicable.

Osteoporosis definition: a skeletal disorder of compromised bone strength predisposing to increased fracture risk; bone strength reflects both bone density and bone quality, so BMD alone does not capture all risk.

Fragility fracture is a fracture from a fall from standing height or less; a prior fragility fracture of the hip or spine raises subsequent fracture risk independently of BMD and can justify treatment even with a T-score above −2.5.

Secondary causes to screen for: glucocorticoid excess, hyperparathyroidism, hyperthyroidism, vitamin D deficiency, malabsorption, hypogonadism, hypercortisolism, multiple myeloma, and chronic kidney disease; basic labs include calcium, 25-OH vitamin D, PTH, TSH, and alkaline phosphatase.

Glucocorticoid-induced osteoporosis: a dose of prednisone ≥5 mg/day for ≥3 months is a risk factor; treatment thresholds are lower and fracture risk rises early, so BMD monitoring is recommended within 6 months of starting and then annually.

Pharmacologic classes: bisphosphonates (antiresorptive, first-line), denosumab (RANKL inhibitor, antiresorptive, requires continued dosing because stopping causes rapid rebound loss), teriparatide and abaloparatide (anabolic), romosozumab (sclerostin inhibitor, anabolic with antiresorptive effect).

Bisphosphonate monitoring: DXA at baseline and generally every 1–2 years; a decline in BMD or a new fracture on therapy prompts reassessment of adherence, secondary causes, and possible switch to an anabolic agent.

Atypical femoral fracture and osteonecrosis of the jaw are rare adverse events associated with long-term bisphosphonate use; a drug holiday is considered after roughly 3–5 years of oral therapy in patients whose fracture risk has fallen.

Osteoporosis in men is often secondary; testosterone deficiency, alcohol excess, glucocorticoids, and malabsorption are common contributors, and treatment uses the same agents as in postmenopausal women.

Error taxonomy: accuracy (systematic bias from calibration) versus precision (random variability across repeats); only precision can be improved by technologist technique, while accuracy requires phantom calibration and service.

Short-term precision is measured by scanning the same patient or phantom 2–3 times with repositioning between scans; the root-mean-square SD across subjects is preferred to a single-subject SD for estimating precision error.

Least significant change (LSC) = 2.77 × precision error at the 95% confidence level; a change smaller than the LSC cannot be distinguished from measurement noise and must not be reported as real bone loss.

The 95% confidence interval for a single measurement is approximately ±1.96 × precision SD; for comparing two measurements the interval widens to ±2.77 × precision SD, which is why LSC uses the 2.77 multiplier.

Quality control log content: date, phantom BMD, mean and SD of prior measurements, control chart limits, technologist initials, and any corrective action; Shewhart rules (e.g., two consecutive points beyond 2 SD) flag drift.

Common technical errors that inflate spine BMD: scanning the wrong level (including T12 or L5 with ribs or iliac crest), off-axis rotation, and failure to flex the hips; each produces a false elevation that can mask osteoporosis.

Common technical errors that deflate BMD: incorrect soft-tissue baseline placement, patient motion, and analyzing a vertebra that is partially outside the scan field; reanalysis must follow the manufacturer's specified labeling of the same vertebral levels.

The ISCD position is that BMD testing be performed and interpreted by technologists and clinicians who have completed accredited training and that reports follow a standardized format; certification bodies maintain separate technologist and clinician examinations.

Bone densitometry results should be interpreted with the clinical context: a normal BMD does not exclude fracture risk, and a low BMD in a young person requires evaluation for secondary causes before a diagnosis of osteoporosis is made.