Understanding Clinical Judgment: When Not to Use FRAX?

The World Health Organization’s FRAX® tool estimates the 10-year probability of a major fracture [1.5.4]. However, its application has specific constraints. Understanding when not to use FRAX is as critical as knowing how to use it for proper clinical assessment [1.3.1].

Quick Summary

The FRAX tool should not be used for patients currently or previously treated for osteoporosis, younger adults, or without careful clinical judgment regarding risk factors not fully captured by the tool, like fall history [1.4.1, 1.5.3, 1.5.5].

Key Points

• Treated Patients: FRAX should not be used if a patient is currently or has previously been on medication for osteoporosis [1.4.1].

• Age Limitations: The tool is not validated for use in pre-menopausal women or men under the age of 40 [1.3.2].

• Dose-Dependent Risks: FRAX uses simple yes/no inputs and does not account for the dosage of glucocorticoids or the quantity of alcohol or tobacco consumed [1.3.1].

• Falls Not Included: A significant limitation is that FRAX does not incorporate a patient's history of falls, a major predictor of fracture risk [1.5.5].

• BMD Specificity: The tool is limited to using only femoral neck (hip) BMD and cannot utilize data from the lumbar spine or other sites [1.3.3].

• Multiple Fractures: The model's risk calculation does not increase with the number of previous fractures a patient has sustained [1.3.1].

• Clinical Judgment is Key: The FRAX score is a tool to aid, not replace, a thorough clinical assessment and professional judgment [1.4.1].

Introduction to the FRAX Tool

The Fracture Risk Assessment Tool (FRAX®) was developed by the World Health Organization to calculate a person's 10-year probability of a hip fracture or a major osteoporotic fracture (hip, clinical spine, forearm, or shoulder) [1.5.4]. It is intended for postmenopausal women and men aged 50 and older who have low bone density (osteopenia) and have not yet taken an osteoporosis medicine [1.5.3, 1.5.5]. The calculation incorporates risk factors such as age, sex, body mass index (BMI), previous fractures, parental hip fracture history, smoking, glucocorticoid use, rheumatoid arthritis, secondary osteoporosis, and alcohol consumption [1.4.4]. Bone mineral density (BMD) of the femoral neck can also be included to refine the prediction [1.5.4].

Core Scenarios: When Not to Use FRAX

While a valuable instrument, FRAX is not universally applicable. Its accuracy and utility are compromised under certain conditions. Good clinical judgment must always supersede a raw score [1.4.1].

1. Patients on Osteoporosis Treatment

The FRAX algorithm was developed and validated using data from untreated patient populations. Therefore, it is not designed to be used for individuals who are currently taking or have previously taken pharmacologic treatment for osteoporosis [1.4.1, 1.4.7]. The tool cannot accurately assess fracture risk in these patients or be used to monitor their response to therapy [1.2.8].

2. Younger Individuals

FRAX is specifically designed for men and women over the age of 40 and is most validated for postmenopausal women and men over 50 [1.3.2, 1.5.3]. It should not be used for children, adolescents, or pre-menopausal women, as the risk models do not apply to these populations [1.2.1].

3. Limitations in Risk Factor Assessment

FRAX uses a binary (yes/no) input for many risk factors, which doesn't account for the dose-response relationship of their impact [1.3.1].

• Glucocorticoids: The tool treats all steroid use over 3 months the same, but fracture risk varies significantly with the dose. A patient on a high dose has a much greater risk than one on a low dose [1.3.3].
• Smoking and Alcohol: FRAX doesn't differentiate between a light smoker/drinker and a heavy one [1.3.1].
• Multiple Prior Fractures: The tool considers a history of a single fracture the same as a history of multiple fractures, even though multiple fractures significantly elevate future risk [1.5.5].

Important Clinical Considerations Not Included in FRAX

Several critical risk factors are explicitly excluded from the FRAX calculation, requiring clinicians to apply their judgment.

• History of Falls: Falls are a major independent risk factor for fractures but are not included in the FRAX model [1.3.1]. A patient with a high propensity for falls has a greater fracture risk than their FRAX score might suggest [1.5.5].
• Lumbar Spine BMD: FRAX only allows for the input of femoral neck (hip) BMD [1.3.1]. A patient may have a normal hip BMD but significant osteoporosis in their lumbar spine, a risk FRAX would not capture [1.3.3].
• Secondary Causes of Osteoporosis: While there is a 'yes/no' input for secondary osteoporosis, it does not have an effect on the risk calculation if a BMD value is entered [1.4.6]. Furthermore, it doesn't account for the impact of multiple combined secondary causes [1.3.1].

Comparison of Fracture Risk Assessment Tools

FRAX is not the only tool available. Other models, like the Garvan Fracture Risk Calculator, may be more appropriate in certain situations, such as for patients with a significant fall risk [1.5.1].

Feature	FRAX® Tool	Garvan Calculator	QFracture®
Primary Inputs	Age, Sex, BMI, 7 Clinical Risk Factors [1.4.4]	Age, Sex, Fractures in last 5 yrs, Falls in last yr [1.6.4]	Age, Sex, BMI, multiple risk factors including falls, cardiovascular disease [1.6.7]
BMD Input	Optional (Femoral Neck only) [1.5.4]	Optional (Femoral Neck) [1.6.4]	Not required [1.6.7]
Includes Falls	No [1.3.1]	Yes [1.6.4]	Yes [1.6.7]
Geographic Scope	Country-specific models [1.5.2]	Australian cohort data	UK cohort data
Output	10-year probability of hip & major osteoporotic fracture [1.5.4]	5 and 10-year probability of hip & any fracture [1.6.4]	1 to 10-year risk of hip or osteoporotic fracture [1.6.7]

Conclusion: A Tool, Not a Replacement for Judgment

Knowing when not to use FRAX is essential for responsible clinical practice. The score is a valuable starting point for conversation and risk stratification but has clear limitations [1.3.2]. It cannot account for every patient's unique risk profile, particularly regarding fall history, dose-dependent risk factors, and multiple previous fractures. The final decision for screening and treatment should always be based on a comprehensive clinical evaluation that uses FRAX as one component among many, not as the sole determinant. For more detailed information, clinicians can refer to guidelines from organizations like the Bone Health & Osteoporosis Foundation.

Frequently Asked Questions

Can FRAX be used for someone already taking osteoporosis drugs like alendronate?

No, FRAX is intended for untreated individuals. It should not be used for patients currently or previously taking osteoporosis medications as the algorithm was not designed to account for treatment effects [1.4.1, 1.4.7].

Is the FRAX score accurate for a 35-year-old woman?

No, the FRAX tool is not intended for use in pre-menopausal women or adults under 40. The risk models are based on data from older populations [1.3.2, 1.5.3].

My patient has fallen three times in the last year. How does this affect their FRAX score?

It doesn't directly affect the score. Fall history is a major limitation of FRAX as it is not an input variable. A clinician must use their judgment to adjust their interpretation of the patient's actual risk upwards [1.3.1, 1.5.5].

What if a patient's spine BMD is very low but their hip BMD is normal?

FRAX only uses femoral neck (hip) BMD. If there is a significant discrepancy where the spine T-score is much lower, FRAX will likely underestimate the true fracture risk. This requires clinical interpretation [1.3.3, 1.4.6].

Does it matter if my patient took a high dose of prednisone versus a low dose?

Clinically, it matters a great deal, but FRAX does not differentiate. It treats any glucocorticoid use over three months as a simple 'yes,' which is a known limitation as risk is dose-dependent [1.3.1].

Why shouldn't I use FRAX to see if osteoporosis treatment is working?

FRAX is a risk prediction tool, not a monitoring tool. It was not designed to measure changes in fracture risk in response to therapy. Bone density scans (DXA) are typically used for monitoring [1.2.8].

Are there alternatives to the FRAX tool?

Yes, other tools like the Garvan Fracture Risk Calculator and QFracture exist. These may be more suitable in certain contexts, for example, as they incorporate fall history as a risk factor [1.6.2, 1.6.7].