Use of electronic health data to identify patients with moderate-to-severe osteoarthritis of the hip and/or knee and inadequate response to pain medications

Lu, Yi; Ganz, Michael L.; Robinson, Rebecca L.; Zagar, Anthony J.; Okala, Sandra; Hartrick, Craig T.; Johnston, Beth; Dorling, Patricia; Slim, May; Thakkar, Sheena; Berger, Ariel

doi:10.1186/s12874-023-01964-y

Table 3 Performance of predefined algorithms to identify patients with moderate-to-severe OA of the hip and/or knee (Step 2, total N = 490 with available claims data)

From: Use of electronic health data to identify patients with moderate-to-severe osteoarthritis of the hip and/or knee and inadequate response to pain medications

Algorithm	TP	FP	TN	FN	Sensitivity	Specificity	NPV	PPV	Accuracy
1. ≥ 1 code for any OA-related surgical procedures	263	0	77	150	0.64	1.00	0.34	1.00	0.69
2. ≥ 2 medical encounters on different days resulting in diagnosis of a relevant psychiatric comorbidity (i.e., anxiety, depression), with at least two encounters for psychiatric comorbidities both occurring within 180 days of any OA diagnosis	87	14	63	326	0.21	0.82	0.16	0.86	0.31
3. ≥ 2 medical encounters resulting in diagnosis of relevant psychiatric comorbidities, with at least two encounters for the psychiatric comorbidities within 90 days of any OA diagnosis	83	12	65	330	0.20	0.84	0.16	0.87	0.30
4. ≥ 2 prescriptions for opioids within 90 days of any OA diagnosis	66	5	72	347	0.16	0.94	0.17	0.93	0.28
5. ≥ 2 prescriptions for oral NSAIDs (including COX-2 inhibitors) within 90 days of any OA diagnosis	1	0	77	412	0.00	1.00	0.16	1.00	0.16
6. ≥ 2 prescriptions for topical NSAIDs (including COX-2 inhibitors) within 90 days of any OA diagnosis	10	2	75	403	0.02	0.97	0.16	0.83	0.17
7. ≥ 1 administration of HA or IA corticosteroids	331	3	74	82	0.80	0.96	0.47	0.99	0.83
8. ≥ 2 administrations of HA or IA corticosteroids at least 90 days apart	191	1	76	222	0.46	0.99	0.26	0.99	0.54
9. ≥ 1 administration of a nerve block ± 30 days of any OA diagnosis	191	3	74	222	0.46	0.96	0.25	0.98	0.54
10. ≥ 1 claim for a mobility aid including walking cane, walker, and wheelchair, or crutch ± 30 days of any OA diagnosis	79	5	72	334	0.19	0.94	0.18	0.94	0.31
11. ≥ 1 code for physical and/or occupational therapy ± 30 days of any OA diagnosis	175	12	65	238	0.42	0.84	0.21	0.94	0.49
12. OA-related costs > 30% mean OA-related costs for the cohort	147	0	77	266	0.36	1.00	0.22	1.00	0.46
13. OA-related costs > 50% mean OA-related costs for the cohort	0	0	77	413	0.00	1.00	0.16	N/A	0.16
14. ≥ 2 codes for X-ray examinations in a 1-year period	295	10	67	118	0.71	0.87	0.36	0.97	0.74
Combined algorithm (using OR function)
Combinations of all algorithms with performance ≥ 0.98 PPV	360	6	71	53	0.87	0.92	0.57	0.98	0.88
Combinations of all algorithms with performance ≥ 0.95 PPV	372	13	64	41	0.90	0.83	0.61	0.97	0.89
Combinations of all algorithms with performance ≥ 0.90 PPV	378	21	56	35	0.92	0.73	0.62	0.95	0.89
Combinations of all algorithms with performance ≥ 0.85 PPV	378	25	52	35	0.92	0.68	0.60	0.94	0.88
Combinations of all algorithms with performance ≥ 0.80 PPV	378	26	51	35	0.92	0.66	0.59	0.94	0.88

Abbreviations: COX-2 Cyclooxygenase-2, FN False negative, FP False positive, HA Hyaluronic acid, IA Intraarticular, NPV Negative predictive value, NSAIDs Nonsteroidal anti-inflammatory drugs, OA Osteoarthritis, PPV Positive predictive value, TN True negative, TP True positive

Back to article page

ISSN: 1471-2288

Contact us

General enquiries: journalsubmissions@springernature.com

BMC Medical Research Methodology

Contact us