Evaluation of Adherence to Dual-Energy X-Ray Absorptiometry Scan Ordering for Persons Living with HIV at a Tertiary Medical Center

To the Editor:

Aging-related comorbidities, such as osteoporosis, cause significant morbidity in persons living with HIV (PLWH). ¹ PLWH are at higher risk for the development of osteoporosis, diagnosed by either a history of fragility fracture or dual-energy X-ray absorptiometry (DEXA) measurement of bone mineral density (BMD). ^2,3 Both antiretroviral treatment (ART) and interactions between HIV and the individual affect BMD. ^4,5 HIV Medicine Association (HIVMA) recommends DEXA screening for all HIV-infected postmenopausal women and men aged ≥50 years. ⁶ There have been no published reports evaluating adherence with this guideline. This study investigates the DEXA ordering practices of providers at a large urban tertiary care center.

PLWH aged ≥50 years from a single academic urban outpatient HIV clinic were included. To qualify, all individuals had at least one clinic visit in the year 2018 and were virally suppressed on ART. A convenience sample was selected. Unblinded chart review assessed if at any time a DEXA scan was (1) ordered, (2) ordered by an HIV provider, (3) completed, and (4) abnormal. In addition, it was noted if individuals were on a tenofovir disoproxil fumarate (TDF)-containing regimen at time of chart extraction.

This project was exempt from Institutional Review Board approval as it was done for a quality improvement initiative in the clinic.

A total of 307 successive charts were reviewed. Men comprised 83.4% (256) of the population; median age was 59 years (range, 50–85). Of the 55 (18%) DEXA scans ordered, HIV providers ordered 16 (29%). Forty-three (78%) PLWH completed their DEXA scans.

These DEXA scans diagnosed 13 (30%) PLWH with osteoporosis, 21 (49%) with osteopenia, 8 (19%) with normal BMD. One (2%) was uninterpretable due to prosthetic hips. At time of chart review, 40 (13%) patients were on TDF-containing regimens of which four (10%) had DEXAs ordered.

Logistic regression examined the relationship among age, gender, and TDF usage with DEXA scan ordering. Increase in age and female gender was associated with an odds ratio of 1.050 (p = 0.023) and 5.95 (p < 0.001), respectively, of having a DEXA scan ordered. Current TDF usage was not associated with ordering (p = 0.534).

Adherence to DEXA scan screening as indicated by HIVMA is low at an academic urban tertiary care center. When performed, there is a high rate of BMD abnormalities. These findings align with Horberg et al., who found that a set of primary care quality metrics for PLWH, such as vaccination and sexually transmitted infections screening, were met <50% of the time. ⁷

Our study had limitations. A convenience sample of a single clinic was used, which may not reflect the national practices. In addition, use of chart review is limited to documentation in the medical record. DEXA scans in other electronic health record systems would not be captured.

Finally, of the patients who did undergo DEXA scans, almost all had evidence of BMD. This may reflect selection bias, indicating that patients with higher risk factors (i.e., frailty) may end up being screened. ⁸

Only 18% of eligible PLWH have been referred for DEXA. Of those who underwent DEXA screening, nearly all (93%) had abnormal BMD. Further study is needed to emphasis the importance of DEXA screening within clinical practice and to determine barriers to its implementation.