Background/Purpose: The 2017 ACR guideline for glucocorticoid-induced osteoporosis (GIOP) recommends DXA testing in patients aged ≥40 years on chronic glucocorticoids. We performed a quality improvement project to measure and improve provider adherence to this guideline in an academic adult rheumatology practice.

Methods: This project was led by adult and pediatric rheumatology fellows at our institution as part of our annual quality improvement curriculum. Our primary aim was to increase the percentage of patients aged ≥40 on prednisone with a DXA scan in the last 2 years from 15% to 30%. A secondary measure was the percentage of visits containing prednisone orders in which a DXA scan was ordered. Additional analyses were performed by patient age, sex, and ordering provider. Data were obtained with Slicer Dicer, an analytic data tool within our electronic health record (EHR), and validated by manual review of charts over a two-week period in 10/2019. Data were analyzed using descriptive statistics. We conducted these cycles using Plan-Do-Study-Act (PDSA) methodology: (1) combined pediatric and adult rheumatology grand rounds describing the project, (2) presentation on GIOP by an endocrinology expert, and (3) data feedback on DXA scan ordering rates and distribution of EHR tools for easily identifying the last DXA scan.

Results: The percentage of patients aged ≥40 on prednisone with a DXA scan in the last 2 years increased from 15.2% among patients seen in 7/2019 to a maximum of 22.9% by 12/2019 after PDSA 1, was 18.7% after PDSA 2 in 2/2020, then declined despite PDSA 3 with onset of COVID19 (Figure 1). Among all patients seen from 7/1/19 – 4/30/20 (n = 1629), the highest percentage was in women aged ≥ 65 (31.6%, n = 443), followed by women aged 40 – 65 (18.2%, n = 752), men aged ≥ 65 (16.1%, n = 211), and men aged 40 – 65 (15.7%, n = 223).

The percentage of visits containing prednisone orders in which a DXA scan was ordered increased from 8.2% in 7/2019 to a maximum of 12.4% after PDSA 2 in 2/2020, then declined with onset of COVID19 (Figure 2). Among all visits with prednisone orders from 7/1/19 – 4/30/20, DXA scan ordering rates varied among 30 providers from 0.0% – 29.2% (IQR 4.3% – 18.4%). Compared with the 2018 – 2019 academic year, DXA scan ordering rates increased among fellows throughout the project (Figure 3).

The manual chart review included 93 patients and 82 visits. Chronic prednisone use (≥ 2.5mg daily for ≥ 3 months) occurred in 65.6% of patients and 59.6% of visits. The rest were on less prednisone or only took short courses. The manual chart review also demonstrated that our search for patients with a DXA scan in the last 2 years only identified scans done at our institution; 29% of all DXA scans occurred outside our institution.

Conclusion: We have not yet met our primary aim of increasing the percentage of patients aged ≥ 40 on prednisone with a DXA scan in the last 2 years from 15% to 30%; initial improvements were negated with onset of COVID19. However, we succeeded in validating a method for tracking this data over time, enabling ongoing efforts to improve adherence to the 2017 ACR GIOP guidelines. Furthermore, the large variability in DXA scan ordering rates and substantial improvement among fellows leading this project suggest that change is possible.

Figure 1: Percentage of patients with a DXA scan in the last 2 years (line) among patients aged ≥40 on prednisone seen in rheumatology clinic each month (bars).

Figure 2: Percentage of visits with DXA scan ordered (line) among rheumatology visits with a prednisone order and patient aged ≥40 each month (bars).

Figure 3: Percentage of visits with DXA scan ordered by provider type

« Back to ACR Convergence 2020

ACR Meeting Abstracts - https://acrabstracts.org/abstract/a-quality-improvement-project-to-improve-bone-densitometry-ordering-in-adult-patients-on-chronic-glucocorticoids/