2009 Southern Regional Meeting Abstracts

Purpose of Study: It is well known that renal insufficiency is more prevalent in patients with congestive heart failure (CHF) and is an independent prognostic factor in diastolic and systolic dysfunction. However, the predictive factors for development of CHF in the setting of chronic kidney disease (CKD) are uncertain.
Methods Used: We analyzed data from the Kaiser Permanente Georgia CKD cohort comprised of enrollees between 1995-2007 who had a minimum of 2 estimated glomerular filtration rates (eGFR’s) < 90 ml/min/1.73 m2 that were more than 60 but less than 365 days apart. This analysis included only patients with an eGFR between 10-59 ml/min. CHF was defined by the presence of a CHF ICD-9 billing code at the time of cohort inception.
Summary of Results: Of the 14,922 eligible patients, 6.3% had a diagnosis of CHF. Those with CHF were older [68.4 versus 61.1 years (p<0.0001)], and were more likely to be male [484 versus 457 patients (p<0.0001)], had more advanced kidney disease [eGFR 44.5 versus 49.4 mL/min/1.73 m2 (p<0.0001)], HTN [99.0% versus 73.0% (p<0.0001)], diabetes [39.2% versus 22.1% (p<0.0001)], CAD [44.9% versus 9.3% (p<0.0001)], and PVD [12.5% versus 3.5% (p<0.0001)]. Among patients with a claim-based diagnosis of CKD, the likelihood of being diagnosed with CHF was similar across GFR strata. However, among patients with a CKD diagnosis, the likelihood of a CHF diagnosis increased with worsening GFR.
Conclusions: Among CKD patients, CHF is more likely among patients with advanced kidney disease and cardiovascular risk factors. The likelihood of CHF is increased as GFR declines among patients without a claim-based CKD diagnosis.

Table 1: Likelihood of Claim-based CHF diagnosis among Patients Lacking Claim-based CKD diagnosis

GFR	OR (95% CI)	CI
50-59	ref
40-49	1.32 (1.09-1.59)	P = 0.0032
30-39	1.87 (1.48-2.36)	P < 0.0001
20-29	1.81 (1.21-2.72)	P = 0.0040
10-19	1.26 (0.51-3.06)	P = 0.6086