Effect of primary care physicians' use of estimated glomerular filtration rate on the timing of their subspecialty referral decisions

Table 3 Percent of primary care physicians with clinically significant improvement in the timing of their subspecialty referral recommendations with the use of estimated glomerular filtration rate (eGFR) by physician and scenario characteristics

Physician characteristics	% of primary care physicians with clinically significant improvement in their subspecialty referral recommendations^*
	Unadjusted	p value	Adjusted^† (95% CI)	p value
Years in practice:
>10 years	44	0.31	50 (32-68)	0.11
≤ 10 years	36		36 (Ref)
Practice type:
Academic	56	0.07	70 (45-87)	0.01
Other	37		37 (Ref)
Percent clinical time:
<80%,	26	0.15	23 (8-49)	0.13
≥ 80%	42		42 (Ref)
Census region:
Midwest	33	0.65	38 (18-64)	0.47
South	38		40 (20-64)	0.56
West	39		41 (20-66)	0.60
Northeast	48		48 (Ref)
Aware of referral guidelines:
Yes	29	0.06	29 (15-48)	0.09
No	45		45 (Ref)
Clinical scenario:
Patient race:
White	57	<0.001	57 (39-73)	<0.001
African American	25		25 (Ref)
Diabetes
Yes	43	0.36	42 (26-61)	0.44
No	35		35 (Ref)

Note. A total of 152 participants with complete data were included in the model
*Clinically significant improvement in subspecialty referral recommendations was defined as present, if primary care physicians selected a serum creatinine-based referral level corresponding to an estimated glomerular filtration rate (eGFR) level of < 30 mL/min/1.73m², but recommended subspecialty referral at level of ≥ 30 mL/min/1.73m² with the use of eGFR.
^†Adjusted for all variables in the table.

ISSN: 1471-2369