Accuracy of syndrome definitions based on diagnoses in physician claims

Cadieux, Geneviève; Buckeridge, David L; Jacques, André; Libman, Michael; Dendukuri, Nandini; Tamblyn, Robyn

doi:10.1186/1471-2458-11-17

BMC Public Health

Table 1 Characteristics of participating and non-participating physicians

From: Accuracy of syndrome definitions based on diagnoses in physician claims

	Participating physicians (N = 1,098)		Non-participating physicians (N = 2,160)
Physician characteristics	No.	(%)	No.	(%)
Gender:
Female	411	37.4	823	38.1
Male	687	62.6	1,337	61.9
Preferred language:
French	1,006	91.6	1,937	89.7
English	92	8.4	223	10.3
Specialty:
General practice	993	90.4	1,932	89.4
Internal medicine	13	1.2	41	1.9
Pediatrics	62	5.6	102	4.7
General surgery	30	2.7	85	3.9
Geriatrics	0	0	0	0
Type of setting selected:¹
Private clinic	1,060	96.5	2,044	94.6
Community health center	5	0.5	9	0.4
Hospital-based ambulatory clinic	33	3.0	107	5.0
Geographic location of selected setting:^1,3
Urban	921	83.9	1,867	86.4
Rural	177	16.1	293	13.6
	Mean	SD	Mean	SD
Years since licensure	24.2	9.7	22.3	10.5
No. days worked per year¹	157.0	55.0	143.2	59.8
No. patients seen per day worked¹	21.2	13.4	21.0	13.3
Syndrome frequency based on claim ICD-9 code	No. visits ^1,2	Prevalence per 1,000 visits ¹	No. visits ^1,2	Prevalence per 1,000 visits ¹
CDC and DoD ESSENCE ⁴
Fever	80,884	11	160,821	12
Gastrointestinal	162,282	22	309,209	24
Neurological	40,236	5	73,810	6
Rash	126,900	17	224,370	17
Respiratory	911,924	125	1,643,240	126
RODS ⁵
Fever	162,000	22	291,990	22
Gastrointestinal	146,355	20	283,578	22
Neurological	36,344	5	67,344	5
Rash	55,251	8	103,698	8
Respiratory	478,201	65	877,556	67
Influenza-like illness ⁶
Large-group	622,046	85	1,129,782	87
Small-group	32,173	4	61,127	5

¹ As per our study design, for each physician, a single practice location was randomly selected to facilitate the validation process. The information in this table is based in claims generated from the selected practice location during the 2-year study period.
² There were a total of 7,315,994 visits to the 1,098 participating physicians, and 13,010,410 visits to the 2,160 eligible non-participating physicians at the selected practice location during the 2-year study period.
³ We tested the statistical significance (at the p < 0.05 level) of any differences between participating and non-participating physicians using a multivariate logistic regression model where the dependent variable was participation and the independent variables were all characteristics in Table 1. Due to overlap between CDC, RODS, and ILI syndrome definitions, to avoid collinearity, we used separate models for each set of syndrome definitions. As compared to non-participating physicians, participating physicians had been in practice longer (odds ratio (OR)_{per 10 years since licensure}, 1.15; 95% CI, 1.05-1.25), had worked more days (OR_{per 50 days}, 1.18; 95% CI, 1.09-1.28) during the 2-year study period.
⁴ Syndrome case definitions developed and published by the US Centers for Disease Control and Prevention (CDC) in 2003, and used by the US Department of Defense's Electronic Surveillance System for Early Notification of Community-based Epidemics (ESSENCE).
⁵ Syndrome case definitions developed in the context of the University of Pittsburgh's Real-time Outbreak and Disease Surveillance (RODS) system.
⁶ Syndrome case definitions developed in the context of the US Department of Defense's Electronic Surveillance System for Early Notification of Community-based Epidemics (ESSENCE).

Back to article page

ISSN: 1471-2458

Contact us

Submission enquiries: bmcpublichealth@biomedcentral.com
General enquiries: ORSupport@springernature.com