Background/Purpose: Gout is a common inflammatory arthritis with significant impact on both patients and health care systems. Despite ACR/EULAR management guidelines and gout quality indicators (QI) developed to improve outcomes, standard of care is often suboptimal and infrequently measured. We evaluated the standard of care of in a cohort of Veterans Affairs (VA) gout patients, using QI that include medication, laboratory and counseling criteria.

Methods: During a 4 year period, VA administrative data was used to identify gout outpatients (ICD 9 codes: 274.xx and at least 2 related visits). QIs assessed were QI 1: patients with creatinine clearance < 60 ml/min, initial allopurinol dose to be < 300 mg/day; QI 2: uric acid (UA) within 6 months of allopurinol start; QI 3: if on colchicine CBC, CPK done within 6 months; QI 4: counseling on gout specific diet, weight loss and alcohol consumption. For QI 4, natural language processing (NLP), a technique that analyzes large amounts of text to identify key words and/or phrases, was used to extract dietary, alcohol and weight loss counseling data from electronic medical records. Data collected were socio- demographics, comorbidities [HTN, DM, CVD, hyperlipidemia, obesity and chronic kidney disease (CKD)] and number of rheumatology outpatient visits for gout. QI compliance versus non-compliance was compared using chi-square analyses.

Results: Of 2,280 gout patients, 2,260 (99.1%) were male, with mean age of 66.8 years. Comorbidities were common: HTN in 2,102 (92.2%); Obesity in 1,578 (69.2%); CVD in 1,384 (60.7%); hyperlipidemia in 1,170 (51.3%); DM in 1,075 (47.1%); and CKD in 748 (33.8%). Most 1,587 (69.6%) had at least one rheumatology outpatient visit, and more than half received specific gout therapy: colchicine was dispensed to 1424 (62.5%) and allopurinol to 1,336 (58.6%) patients. Compliance with each QI was as follows: QI 1: 92.1%; QI 2: 44.8%; QI 3: CBC monitoring 70%; but only 6.3% for both CBC and CPK (only 1 patient had CPK without CBC). For QI 4, there was counseling for weight loss in 1008 (44.2%), diet in 390 (17.1%), alcohol in 137 (6.0%) and 51 (2.2%) had counseling on all 3 elements. Of those on new allopurinol prescriptions, target serum uric acid (< 6 mg/dl) was achieved in 64 (30.1%) patients within one year. Compared with non-compliant patients, patients compliant with QI 2 had more rheumatology visits (3.5 vs. 2.6; p< 0.001), while those compliant with QI 3 (CBC) were older (67.3 vs 64.9 years p<0.001) and had more CKD (p<0.001), DM (p<0.001) and CVD (p<0.001). Patients with DM, obesity, and hyperlipidemia were more often counseled on diet compared to those without comorbidities.  Alcohol counseling was less frequent in gout patients with hyperlipidemia (p= 0.024) and DM (p=0.012) compared to patients without comorbidities.

Conclusion: In our study cohort, compliance with QI for uric acid and CPK monitoring were subpar. In gout patients, specific dietary counseling appeared to be directed by other comorbidities. NLP proved a valuable tool for evaluating dietary QI in patients with gout.

---

« Back to 2012 ACR/ARHP Annual Meeting

ACR Meeting Abstracts - https://acrabstracts.org/abstract/natural-language-processing-in-the-evaluation-of-gout-quality-indicators/