Patient, physician, encounter, and billing characteristics predict the accuracy of syndromic surveillance case definitions

Table 1 Physician characteristics associated with accuracy of syndrome definitions based on physician claims (OR >1.00 means the encounter characteristic increased the PPV of the syndrome definition, OR < 1.00 means the encounter characteristic reduced the PPV)

	No. visits with a syndrome-positive physician claim						Bivariate regression analysis			Multivariate regression analysis¹
Physician characteristics	Syndrome-positive in the chart (N = 2,967)		Syndrome-negative in the chart (N = 1,363)		Total (N = 4,330)		OR	95% CI	P value	OR	95% CI	P value
	No.	%	No.	%	No.	%
Gender:
Female	1,164	39.2	523	38.4	1,687	39.0	Ref.	Ref.	Ref.	Ref.	Ref.	Ref.
Male	1,803	60.8	840	61.6	2,643	61.0	0.97	(0.83, 1.12)	0.64	1.13	(0.96, 1.33)	0.13
Preferred language:
French	2,743	92.5	1,253	91.9	3,996	92.3	Ref.	Ref.	Ref.	Ref.	Ref.	Ref.
English	224	7.5	110	8.1	334	7.7	0.93	(0.69, 1.25)	0.63	0.94	(0.69, 1.26)	0.66
Specialty:
General practice	2,721	91.7	1,246	91.4	3,967	91.6	Ref.	Ref.	Ref.	Ref.	Ref.	Ref.
Pediatrics	203	6.8	75	5.5	278	6.4	1.24	(0.88, 1.77)	0.22	0.83	(0.57, 1.20)	0.33
Internal medicine or general surgery	43	1.5	42	3.1	85	2.0	0.46	(0.31, 0.69)	< 0.001	0.59	(0.35, 0.98)	0.04
	Mean	SD	Mean	SD	Mean	SD
Years since licensure (per 5 years)	22.9	9.2	23.7	9.6	23.1	9.4	0.95	(0.92, 0.99)	0.02	0.96	(0.92, 1.00)	0.04

¹ Multivariate analysis adjusted for all physician characteristics in Table 1, all patient characteristics in Table 2, and all encounter characteristics in Table 3.

ISSN: 1471-2458