Association of persistent and transient worsening renal function with mortality risk, readmissions risk, length of stay, and costs in patients hospitalized with acute heart failure

Figures & data

Figure 1 Patient selection.

Notes: ^aAn AHF hospitalization is an admission having a primary or secondary diagnosis for any of the following ICD-9-CM heart failure codes: 398.91, 402.x1, 404.x1, 404. x3, 415.0, 416.x, 417.x, 425.x, 428.0, 428.1, 428.2x, 428.3x, 428.4x, 428.9, 429.0, 429.1.
Abbreviations: AHF, acute heart failure; HF, heart failure; LOS, length of stay; SCr, serum creatinine.

Table 1 Patient and hospital characteristics by WRF cohort

	Total (n=55,436)	Non-WRF (n=42,507)	WRFp (n=5,563)	WRFt (n=7,366)	P-value (3 groups)a	P-value (WRFp vs WRFt)b
Age (year), mean (SD)	72.4 (14.3)	72.3 (14.4)	73.5 (13.5)	71.9 (13.9)	<0.0001	<0.0001
Female (%)	53.1	53.2	52.4	53.4	0.4724	0.2411
Race (%)					<0.0001	<0.0001
White	78.7	79.4	76.5	76.3
Black	16.9	16.4	17.5	19.0
Hispanic	1.2	1.2	1.3	1.4
Asian	0.6	0.6	0.9	0.8
Unknown	2.4	2.3	3.5	2.4
Index SCr (mg/dL), mean (SD)	1.4 (1.1)	1.5 (1.2)	1.4 (0.8)	1.4 (0.8)	0.2306	0.1538
Index BUN/SCr ratio, mean (SD)c	20.0 (8.9)	20.1 (8.9)	20.1 (10.0)	19.4 (8.3)	<0.0001	0.0256
Index eGFR (%)d					0.0009	0.733
<60 mL/min/1.73 m^2	52.0	51.4	53.9	54.3
≥60 mL/min/1.73 m^2	48.0	48.6	46.2	45.7
Number of hospital beds (%)					<0.0001	0.0011
<6	0.03	0.02	0.02	0.1
6–99	8.1	8.5	7.5	6.4
100–199	15.7	16.1	14.4	13.9
200–299	18.8	19.0	19.1	17.9
300–499	26.2	26.6	24.9	24.4
500+	31.2	29.8	34.1	37.3
Hospital designation (%)					0.1066	0.7758
Rural	0.2	0.2	0.1	0.1
Urban	99.8	99.8	99.9	99.9
Hospital teaching status (%)					<0.0001	<0.0001
Non-teaching	27.7	28.1	28.6	25.2
Teaching	72.3	71.9	71.4	74.8

Notes:

a The non-WRF cohort is the reference group;

b the WRFt cohort is the reference group;

c missing in 388 patients;

d missing in 28,769 (52%) patients.

Abbreviations: BUN, blood urea nitrogen; eGFR, estimated glomerular filtration rate; SCr, serum creatinine; SD, standard deviation; WRF, worsening renal function; WRFp, persistent WRF; WRFt, transient WRF.

Table 2 Medication utilization and comorbidities by WRF cohort

	Total (n=55,436)	Non-WRF (n=42,507)	WRFp (n=5,563)	WRFt (n=7,366)	P-value (3 groups)a	P-value (WRFp vs WRFt)b
HF-related medication use (%)
ACE inhibitor	42.7	41.3	44.3	49.8	<0.0001	<0.0001
ARB	11.5	11.0	12.1	13.8	<0.0001	0.0064
β-Blocker	77.9	76.6	80.8	83.2	<0.0001	0.0005
Loop diuretic	71.1	67.7	82.5	82.2	<0.0001	0.6309
Charlson comorbidity index score (SD)	2.2 (1.9)	2.2 (1.9)	2.5 (2.0)	2.3 (1.9)	<0.0001	0.0011
Comorbidities (%)
Chronic liver disease and cirrhosis	3.1	3.0	3.9	3.0	0.0015	0.0094
Valvular heart disease	24.1	23.8	24.0	25.7	0.0024	0.0265
Other chronic respiratory conditions	23.4	24.0	20.1	22.3	<0.0001	0.0036
Chronic respiratory failure, primary pulmonary hypertension, and cardiopulmonary obesity	21.6	22.1	18.6	20.6	<0.0001	0.0043
Chronic impaired immune function	14.0	13.7	16.0	13.9	<0.0001	0.0010
Cancer	10.0	10.0	11.0	8.9	0.0003	<0.0001
Sepsis/septic shock	8.0	6.7	13.3	11.5	<0.0001	0.0021
Asthma	8.0	8.0	7.0	8.3	0.0155	0.0063

Notes:

a The non-WRF cohort is the reference group;

b the WRFt cohort is the reference group.

Abbreviations: ACE, angiotensin-converting enzyme; ARB, angiotensin receptor blockers; HF, heart failure; SD, standard deviation; WRF, worsening renal function; WRFp, persistent WRF; WRFt, transient WRF.

Figure 2 Mortality rate by WRF cohort.

Notes: Patients in the WRFp cohort had significantly (P<0.0001) higher rates of mortality at all time points than patients in the non-WRF and WRFt cohorts.
Abbreviations: HF, heart failure; WRF, worsening renal function; WRFp, persistent WRF; WRFt, transient WRF.

Table 3 Statistical modeling summary

Outcome	Comparison	Coefficients/odds ratio
		Index hospitalization		30-day model		180-day model		365-day model
		WRFp	WRFt	WRFp	WRFt	WRFp	WRFt	WRFp	WRFt
Readmissiona (negative binomial model)	WRFp vs WRFt vs non-WRFb			0.216e	0.237e	0.132e	0.168e	0.115e	0.138e
	WRFp vs WRFtc			−0.020		−0.036		−0.031
Mortalityd (logistic regression model)	WRFp vs WRFt vs non-WRFb	7.763e	1.202e	6.537e	1.225e	5.331e	1.269e	4.830e	1.218e
	WRFp vs WRFtc	6.428e		5.427e		4.292e		4.071e
LOSf (GLM model)	WRFp vs WRFt vs non-WRFb	0.211e	0.489e	0.194e	0.475e	0.163e	0.447e	0.149e	0.429e
	WRFp vs WRFtc	−0.273e		−0.274e		−0.278e		−0.273e
Costg (GLM model)	WRFp vs WRFt vs non-WRFb	0.246e	0.481e	0.227e	0.464e	0.195e	0.443e	0.178e	0.425e
	WRFp vs WRFtc	−0.235e		−0.235e		−0.247e		−0.245e

Notes:

a Other significant multivariable predictors for readmission included WRFp, WRFt, atrial fibrillation, BUN/SCr ratio, index year, and admission source;

b the non-WRF cohort is the reference group;

c the WRFt cohort is the reference group;

d other significant multivariable predictors for mortality included WRFp, sepsis, cardiogenic shock, pneumonia, index admission SCr, and index admission BUN/SCr ratio;

e P<0.01;

f other significant multivariable predictors for LOS included WRFp, sepsis, pneumonia, acute cerebrovascular disease, WRFt, and cardiogenic shock;

g other significant multivariable predictors for cost included WRFt, sepsis, cardiogenic shock, pneumonia, emergency admission, geographic region, and index year.

Abbreviations: BUN, blood urea nitrogen; GLM, generalized linear model; LOS, length of stay; SCr, serum creatinine; WRF, worsening renal function; WRFp, persistent WRF; WRFt, transient WRF.

Figure 3 LOS by WRF cohort.

Notes: Patients in the WRFt cohort had significantly (P<0.0001) longer LOS at all time points than patients in the non-WRF and WRFp cohorts. The LOS was significantly longer in the WRFp cohort compared with the non-WRF cohort.
Abbreviations: LOS, length of stay; WRF, worsening renal function; WRFp, persistent WRF; WRFt, transient WRF.

Figure 4 Cost by WRF cohort.

Notes: Patients in the WRFt cohort had significantly (P<0.0001) higher costs at all time points than patients in the non-WRF and WRFp cohorts. The costs were significantly higher in the WRFp cohort compared with the non-WRF cohort. Costs are in 2012 US dollars.
Abbreviations: WRF, worsening renal function; WRFp, persistent WRF; WRFt, transient WRF.

Figure 5 AHF-related hospital readmissions by WRF cohort.

Notes: Patients in the WRFt and WRFp cohorts had a significantly (P<0.0001) higher number of AHF-related readmissions at all time points than patients in the non-WRF. There was no significant difference in readmissions between the WRFt and WRFp cohorts.
Abbreviations: AHF, acute heart failure; WRF, worsening renal function; WRFp, persistent WRF; WRFt, transient WRF.