Patient and provider determinants of nephrology referral in older adults with severe chronic kidney disease: a survey of provider decision making

• Published in: BMC nephrology 2011-09, Vol.12 (1), p.47-47, Article 47
• Main Authors: Campbell, Kellie H, Smith, Sandy G, Hemmerich, Joshua, Stankus, Nicole, Fox, Chester, Mold, James W, O'Hare, Ann M, Chin, Marshall H, Dale, William
• Format: Article
• Language: English
• Subjects: Adult; Aged; Aged, 80 and over; Black People - statistics & numerical data; Care and treatment; Chronic kidney failure; Comorbidity; Dementia - ethnology; Female; Health Care Surveys - statistics & numerical data; Humans; Internet; Kidney Failure, Chronic - ethnology; Kidney Failure, Chronic - therapy; Male; Middle Aged; Multivariate Analysis; Nephrology; Nephrology - statistics & numerical data; Patient outcomes; Practice guidelines (Medicine); Predictive Value of Tests; Primary Health Care - statistics & numerical data; Professional Practice - statistics & numerical data; Referral and Consultation - statistics & numerical data; Risk Factors; Severity of Illness Index; White People - statistics & numerical data
• ISSN: 1471-2369; 1471-2369
• DOI: 10.1186/1471-2369-12-47

Although chronic kidney disease (CKD) disproportionately affects older adults, they are less likely to be referred to a nephrologist. Factors that influence the referral decisions of primary care providers (PCPs) specifically for older CKD patients have been incompletely described. Patient factors such as dementia, functional disability, and co-morbidity may complicate the decision to refer an older adult. This study evaluated the role of patient and PCP factors in the referral decisions for older adults with stage 4 CKD. We administered a two-part survey to study the decisions of practicing PCPs. First, using a blocked factorial design, vignettes systematically varied 6 patient characteristics: age, race, gender, co-morbidity, functional status, and cognitive status. CKD severity, patient preferences, and degree of anemia were held constant. Second, covariates from a standard questionnaire included PCP estimates of life expectancy, demographics, reaction to clinical uncertainty, and risk aversion. The main outcome was the decision to refer to the nephrologist. Random effects logistic regression models tested independent associations of predictor variables with the referral decision. More than half (62.5%) of all PCP decisions (n = 680) were to refer to a nephrologist. Vignette-based factors that independently decreased referral included older patient age (OR = 0.27; 95% CI, 0.15 to 0.48) and having moderate dementia (OR = 0.14; 95%CI, 0.07 to 0.25). There were no associations between co-morbidity or impaired functional activity with the referral decision. Survey-based PCP factors that significantly increased the referral likelihood include female gender (OR = 7.75; 95%CI, 2.07 to 28.93), non-white race (OR = 30.29; 95%CI, 1.30 to 703.73), those who expect nephrologists to discuss goals of care (OR = 53.13; 95%CI, 2.42 to 1168.00), those with higher levels of anxiety about uncertainty (OR = 1.28; 95%CI, 1.04 to 1.57), and those with greater risk aversion (OR = 3.39; 95%CI, 1.02 to 11.24). In this decision making study using hypothetical clinical vignettes, we found that the PCP decision to refer older patients with severe CKD to a nephrologist reflects a complex interplay between patient and provider factors. Age, dementia, and several provider characteristics weighed more heavily than co-morbidity and functional status in PCP referral decisions. These results suggest that practice guidelines should develop a more nuanced approach to the referral of older