Renal function and clinical outcomes in survivors of out-of-hospital cardiac arrest

Graphical abstract

Abstract

Background

Comprehensive studies about renal-function changes in the context of out-of-hospital cardiac arrest (OHCA) have been lacking. Therefore, we investigated the impact of renal function on clinical outcomes among patients with OHCA.

Method

This retrospective cohort study enrolled consecutive patients with OHCA between June 2017 and December 2021. Acute kidney injury (AKI) was defined based on the “Kidney Disease: Improving Global Outcomes (KDIGO)” guidelines. AKI recovery was defined as a decrease in serum creatinine below the level determined in the definition of AKI. Clinical outcomes included neurological outcomes and all-cause mortality.

Result

A total of 258 patients were enrolled, including 35 patients with underlying end-stage renal disease (ESRD). Among patients without ESRD, 82.5% developed AKI, of which 31.0% achieved AKI recovery, while 61.0% were discharged with impaired renal function. Multivariable analysis using regression models revealed that unfavorable neurological outcomes at discharge and higher mortality at 2 years were associated with AKI (odds ratio [OR] 7.684, 95% confidence interval (CI) 2.683–22.010, P < 0.001; hazard ratio [HR] 2.159, 95% CI 1.272–3.664, P = 0.004), AKI without recovery (OR 5.275, 95% CI 2.049–13.583, P < 0.001; HR 5.470, 95% CI 3.304–9.862, P < 0.001), and impaired pre-discharge renal function (OR 3.164, 95% CI 1.442–6.940, P = 0.004; HR 2.876, 95% CI 1.861–4.443, P < 0.001). Compared to those without ESRD, patients with underlying ESRD had similar neurological outcomes and mortality.

Conclusion

AKI, AKI without recovery, and impaired pre-discharge renal function were significantly correlated with worse clinical outcomes in OHCA among patients without ESRD, while underlying ESRD did not lead to worse clinical outcomes.

Introduction

Renal function in cases of out-of-hospital cardiac arrest (OHCA) is considered to be strongly associated with clinical outcomes.1, 2, 3, 4, 5, 6, 7, 8, 9, 10 Acute kidney injury (AKI) often develops in the context of OHCA and is a well-known predictor of poor neurological outcomes and mortality.3, 4, 5, 6, 7, 8 Interest has been drawn to the relation between the recovery from AKI and the clinical outcomes in OHCA cases. Some studies have suggested that AKI recovery may be associated with better prognosis,9, 10 but relevant studies have remained limited.

A lower estimated glomerular filtration rate (eGFR) was shown to be correlated with a higher incidence of OHCA in the general population according to an observational study, although it is seldom discussed.¹ A graded association was also observed between decreased eGFR on arrival to the hospital and short-term prognosis in a specific population with OHCA.² Nevertheless, the link between eGFR and the course of patients with OHCA has still been uncertain in view of the insufficient reports. Interestingly, however, patients with end-stage renal disease (ESRD) were found to have better short-term hospital outcomes than patients without ESRD in the context of OHCA,¹¹ although very few studies have examined this. Therefore, we extensively studied the association between renal function and its changes during hospitalization on the clinical outcomes among patients with OHCA in 2 years of follow-up.

Material and methods

Participants

Consecutive patients who presented with OHCA were enrolled at Taipei Veterans General Hospital between June 1, 2017, and December 31, 2021. Patients were included if they were at least 20 years old, had experienced OHCA due to non-traumatic causes, and had achieved return of spontaneous circulation (ROSC) after cardiopulmonary resuscitation (CPR). Patients were excluded if they were under 20 years of age or did not achieve ROSC.

Study design

This study was a retrospective single-center cohort study. All patients were admitted to the critical care unit after ROSC. Comprehensive data were collected by reviewing medical records, including the use of the Utstein template for OHCA items.¹² The study was approved by the Ethics Committee of Taipei Veterans General Hospital (approval number: 2022-09-009AC) and was conducted in accordance with the principles of the Declaration of Helsinki.

Baseline characteristics

The variables associated with resuscitation were documented, including arrest in a public area, shockable rhythm, bystander CPR, whether target temperature management (TTM) was performed, and the duration from first contact until ROSC. The causes of OHCA were recorded if they were attributed to asphyxia or cardiac causes. Patients who received regular hemodialysis (HD) were recognized as having ESRD. Other comorbidities were also recorded, including hypertension, diabetes mellitus (DM), heart failure (HF), prior coronary artery disease (CAD), prior cerebrovascular accident (CVA), and cancer.

Renal function measurement and definition of AKI

For patients without underlying ESRD, the occurrence of AKI during hospitalization, AKI recovery, impaired renal function on arrival and before discharge, and dialysis requirement during hospitalization were recorded. The renal function on arrival and before discharge were presented with eGFR using the four-variable equation proposed by the “Modification of Diet in Renal Disease” (MDRD) study.¹³ eGFR less than 60 ml/min/1.73 m² was defined as impaired renal function; otherwise, renal function was considered as preserved.

AKI was defined as an increase in serum creatinine by ≥ 0.3 mg/dl within 48 h, an increase in serum creatinine to ≥ 1.5 times baseline within 7 days, or urine volume ≤ 0.5 ml/kg/h for 6 h according to the “Kidney Disease: Improving Global Outcomes (KDIGO)” guidelines.¹⁴ The development of AKI was diagnosed if either the criteria for serum creatinine or hourly urine output were fulfilled. If data were available for the serum creatinine level within the last 3 months before the development of OHCA, the level was regarded as the baseline serum creatinine. If such data were not available, the lower value between the estimated serum creatinine level and the first measured serum creatinine during the hospital course was regarded as the baseline serum creatinine level.⁹

The estimated serum creatinine level was assessed according to the KDIGO guidelines.¹⁵ The development of AKI either on arrival or during hospitalization was recorded. The definition of AKI recovery was based on a decrease in serum creatinine below the level determined in the definition of AKI.⁹

Clinical outcomes

The clinical outcomes included neurological outcomes and all-cause mortality. Neurological outcomes were evaluated using the Cerebral Performance Category (CPC) scale at hospital discharge, and unfavorable outcomes corresponded to scores of 3, 4, or 5.¹⁶ All-cause mortality was assessed at discharge, 6 months, 1 year, and 2 years. Survival time was defined as the time from admission to death, with exact death times used in the survival analysis.

Statistical analysis

Statistical analysis was performed using the Statistical Package for Social Sciences (version 21.0, SPSS Inc., Chicago, IL, USA). All data are expressed as the mean ± standard deviation or as a frequency and percentage. Parametric continuous data were compared between different patient groups using the unpaired Student’s t-test, and nonparametric data were compared using the Mann–Whitney test. Categorical variables were analyzed using the chi-squared test. For the neurologic outcome and mortality, the difference in proportions was additionally calculated. Survival analysis was performed using a Kaplan–Meier (KM) curve with significance based on the log-rank test. Baseline clinical factors with a P-value < 0.100 for clinical outcomes were entered into a multivariable analysis. However, the duration from first contact to ROSC was not included as a predictor because the data were not available for all patients. A multivariable logistic regression was used to calculate the odds ratio (OR) to assess the impact of renal function on poor neurological outcomes, while a Cox proportional hazard model was used to calculate the hazard ratio (HR) for its effect on mortality with adjustments for potential confounders. Statistical significance was defined by a two-sided P-value < 0.05.

Results

Baseline characteristics of all patients

A total of 258 patients with OHCA were eligible for enrollment. The mean age of the participants was 67.0 ± 16.2 years, and 61.6% were men. The cardiac arrest took place in public in 45.0% of the cases and at a residence in 55.0%. Furthermore, 31.0% of the patients presented with shockable rhythm, and 39.1% received bystander CPR. After ROSC, 29.5% of patients received TTM. The duration from first contact to ROSC was available for 235 patients, and the average was 33.2 ± 32.0 min. The two leading causes of OHCA were cardiac causes (50.0%) and asphyxia (22.1%) (Table 1).

Table 1.

Baseline characteristics of all patients.

	All (n = 258)
Age, years	67.0 ± 16.2
Male, n (%)	159 (61.6%)
Arrest at public, n (%)	116 (45.0%)
Shockable rhythm, n (%)	80 (31.0%)
Bystander CPR, n (%)	101 (39.1%)
Targeted temperature management, n (%)	76 (29.5%)
First contact to ROSC, mins	33.2 ± 32.0a
Cause of OHCA
Asphyxia, n (%)	57 (22.1%)
Cardiac causes, n (%)	129 (50.0%)
Medical history
End-stage renal disease, n (%)	35 (13.6%)
Hypertension, n (%)	145 (56.2%)
Diabetes mellitus, n (%)	87 (33.7%)
Heart failure, n (%)	33 (12.8%)
Prior coronary artery disease, n (%)	50 (19.4%)
Prior cerebrovascular disease, n (%)	35 (13.6%)
Cancer, n (%)	27 (10.5%)

CPR, cardiopulmonary resuscitation; OHCA, out-of-hospital cardiac arrest; ROSC, return of spontaneous circulation.

There were 35 patients (13.6%) who had a medical history of ESRD and had received hemodialysis. Other comorbidities included hypertension in 145 patients (56.2%), DM in 87 patients (33.7%), HF in 33 patients (12.8%), prior CAD in 50 patients (19.4%), prior CVA in 35 patients (13.6%), and cancer in 27 patients (10.5%) (Table 1). The baseline characteristics of non-ESRD and ESRD patients are presented in Supplementary Table 1.

Renal function after OHCA among patients without ESRD

Among non-ESRD patients, 184 (82.5%) developed AKI after OHCA, and 127 (69.0%) did not recover before discharge. At admission, 169 patients (75.8%) had impaired renal function, decreasing to 136 (61.0%) by discharge (Table 2). AKI patients more frequently had a history of DM (P = 0.032) and longer first-contact-to-ROSC times (P = 0.042) compared to non-AKI patients (Supplementary Table 2). Renal recovery was more common in those with cardiac causes of OHCA (P = 0.013) and prior CAD (P = 0.028). During resuscitation, patients with AKI recovery showed more shockable rhythms (P < 0.001) and shorter first-contact-to-ROSC times (P = 0.044) (Supplementary Table 2). Impaired renal function at discharge was associated with histories of DM (P = 0.009) and HF (P = 0.039), less shockable rhythms (P < 0.001), and longer first-contact-to-ROSC times (P = 0.018) compared to those with preserved renal function (Supplementary Table 2).

Table 2.

Renal function changes in patients without underlying ESRD.

	Patients without ESRD (n = 223)
AKI development, n (%)	184 (82.5%)
AKI without recovery, n (%)	127/184 (69.0%)
AKI recovery, n (%)	57/184 (31.0%)
Renal function on arrival
Impaired renal function, n (%)	169 (75.8%)
Preserved renal function, n (%)	54 (24.2%)
Pre-discharge renal function
Impaired renal function, n (%)	136 (61.0%)
Preserved renal function, n (%)	87 (39.0%)
Dialysis requirement, n (%)	24 (10.8%)
Dialysis required at discharge, n (%)	19/24 (79.2%)
Free from dialysis at discharge, n (%)	5/24 (20.8%)

AKI, acute kidney injury; ESRD, end-stage renal disease.

Clinical outcomes in patients without underlying ESRD

All the patients received follow-up at discharge, but 42 patients (16.3%) were lost to follow-up at 2 years (Supplementary Fig. 1). The mean follow-up duration of all patients was 10.3 ± 20.5 months. Table 3 shows the results of comparisons of the baseline characteristics of the patients according to neurological outcomes at discharge and in-hospital mortality.

Table 3.

Baseline characteristics according to neurologic outcome and in-hospital mortality.

	CPC 3–5 (n = 202)	CPC 1–2 (n = 56)	P-value	In-hospital mortality (n = 165)	Survival to discharge (n = 93)	P-value
Age, years	69.1 ± 15.8	59.3 ± 15.4	<0.001	69.0 ± 16.4	63.4 ± 15.3	0.007
Male, n (%)	117 (57.9%)	42 (75.0%)	0.020	95 (57.6%)	64 (68.8%)	0.075
Arrest at public, n (%)	82 (40.6%)	34 (60.7%)	0.007	62 (37.6%)	54 (58.1%)	0.001
Shockable rhythm, n (%)	42 (20.8%)	38 (67.9%)	<0.001	32 (19.4%)	48 (51.6%)	<0.001
Bystander CPR, n (%)	67 (33.2%)	34 (60.7%)	<0.001	50 (30.3%)	51 (54.8%)	<0.001
TTM, n (%)	55 (27.2%)	21 (37.5%)	0.136	43 (26.1%)	33 (35.5%)	0.111
First contact to ROSC, mins	35.9 ± 34.1a	21.7 ± 17.1b	<0.001	39.4 ± 35.9c	21.4 ± 18.0d	<0.001
Cause of OHCA
Asphyxia, n (%)	53 (26.2%)	4 (7.1%)	0.002	44 (26.7%)	13 (14.0%)	0.018
Cardiac cause, n (%)	83 (41.1%)	46 (82.1%)	<0.001	66 (40.0%)	63 (67.7%)	<0.001
Medical history
Hypertension, n (%)	115 (56.9%)	30 (53.6%)	0.654	95 (57.6%)	50 (53.8%)	0.553
DM, n (%)	73 (36.1%)	14 (25.0%)	0.119	60 (36.4%)	27 (29.0%)	0.232
HF, n (%)	31 (15.3%)	2 (3.6%)	0.020	26 (15.8%)	7 (7.5%)	0.057
Prior CAD, n (%)	39 (19.3%)	11 (19.6%)	0.955	30 (18.2%)	20 (21.5%)	0.517
Prior CVA, n (%)	31 (15.3%)	4 (7.1%)	0.113	26 (15.8%)	9 (9.7%)	0.171
Cancer, n (%)	20 (9.9%)	7 (12.5%)	0.574	17 (10.3%)	10 (10.8%)	0.910

CAD, coronary artery disease; CAG, coronary angiography; CPC, cerebral performance category; CPR, cardiopulmonary resuscitation; CVA, cerebrovascular disease; DM, diabetes mellitus; HF, heart failure; OHCA, out-of-hospital cardiac arrest; ROSC, return of spontaneous circulation TTM; targeted temperature management.;

^an = 189,

^bn = 46,

^cn = 153,

^dn = 82.

Patients with AKI had significantly higher prevalence of unfavorable neurological outcomes (difference, 29%; 95% confidence intervals [CI], 12% to 45%; P < 0.001), in-hospital mortality (difference, 29%; 95% CI, 12% to 45%; P < 0.001), 6-month mortality (difference, 28%; 95% CI, 10% to 47%; P = 0.002), 1-year mortality (difference, 28%; 95% CI, 9% to 47%; P = 0.001), and 2-year mortality (difference, 28%; 95% CI, 9% to 47%; P < 0.001) (Table 4). Patients with AKI without recovery had significantly higher prevalence of unfavorable neurological outcomes (difference, 28%; 95% CI, 15% to 42%; P < 0.001), in-hospital mortality (difference, 65%; 95% CI, 52% to 77%; P < 0.001), 6-month mortality (difference, 60%; 95% CI, 45% to 75%; P < 0.001), 1-year mortality (difference, 56%; 95% CI, 40% to 72%; P < 0.001), and 2-year mortality (difference, 55%; 95% CI, 38% to 72%; P < 0.001) (Table 4). While there was no difference in clinical outcomes between those with impaired and preserved renal function on arrival, patients with impaired pre-discharge renal function had higher prevalence of unfavorable neurological outcomes (difference, 24%; 95% CI, 12% to 35%; P < 0.001), in-hospital mortality (difference, 49%; 95% CI, 38% to 61%; P < 0.001), 6-month mortality (difference, 43%; 95% CI, 30% to 56%; P < 0.001), 1-year mortality (difference, 39%; 95% CI, 26% to 53%; P < 0.001), and 2-year mortality (difference, 39%; 95% CI, 25% to 53%; P < 0.001) than those with preserved pre-discharge renal function (Table 4).

Table 4.

Clinical outcomes in patients without underlying ESRD.

	AKI	No AKI	P-value
CPC 3-5, n (%)	152/184 (82.6%)	21/39 (53.8%)	<0.001
In-hospital mortality, n (%)	124/184 (67.4%)	15/39 (38.5%)	<0.001
6-month mortality, n (%)	125/167 (74.9%)	15/32 (46.9%)	0.002
1-year mortality, n (%)	127/160 (79.4%)	16/31 (51.6%)	0.001
2-year mortality, n (%)	129/156 (82.7%)	16/29 (55.2%)	<0.001
	AKI without recovery	AKI recovery	P-value
CPC 3-5, n (%)	116/127 (91.3%)	36/57 (63.2%)	<0.001
In-hospital mortality, n (%)	111/127 (87.4%)	13/57 (22.8%)	<0.001
6-month mortality, n (%)	112/124 (90.3%)	13/43 (30.2%)	<0.001
1-year mortality, n (%)	113/122 (92.6%)	14/38 (36.8%)	<0.001
2-year mortality, n (%)	115/121 (95.0%)	14/35 (40.0%)	<0.001
	Impaired eGFRadmission	Preserved eGFRadmission	P-value
CPC 3-5, n (n (%)	133/169 (78.7%)	40/54 (74.1)	0.478
In-hospital mortality, n (%)	110/169 (65.1%)	29/54 (53.7%)	0.133
6-month mortality, n (%)	111/153 (72.5%)	29/46 (63.0%)	0.216
1-year mortality, n (%)	113/148 (76.4%)	30/43 (69.8%)	0.381
2-year mortality, n (%)	115/145 (79.3%)	30/40 (75.0%)	0.558
	Impaired eGFRpredischarge	Preserved eGFRpredischarge	P-value
CPC 3-5, n (%)	118/136 (86.8%)	55/87 (63.2%)	<0.001
In-hospital mortality, n (%)	111/136 (81.6%)	28/87 (32.2%)	<0.001
6-month mortality, n (%)	112/132 (84.8%)	28/67 (41.8%)	<0.001
1-year mortality, n (%)	113/129 (87.6%)	30/62 (48.4%)	<0.001
2-year mortality, n (%)	115/127 (90.6%)	30/58 (51.7%)	<0.001

AKI, acute kidney injury; CI, confidence interval; CPC, cerebral performance category; eGFR, estimated glomerular filtration rate; ESRD, end-stage renal disease.

The KM survival curves correspondingly showed higher all-cause mortality at 2 years in patients with AKI (log-rank P = 0.002), AKI without recovery (log-rank P < 0.001), and impaired pre-discharge renal function (log-rank P < 0.001) (Supplementary Figs. 2–4). Logistic regression models revealed that AKI (OR, 7.684; 95% CI, 2.683–22.010; P < 0.001), AKI without recovery (OR, 5.275; 95% CI, 2.049–13.583; P < 0.001), and impaired renal function before discharge (OR, 3.164; 95% CI, 1.442–6.940; P = 0.004) were independently correlated with poor neurological outcomes (Table 5). Cox regression analysis revealed that AKI (HR, 1.868; 95% CI, 1.079–3.231; P = 0.026), AKI without recovery (HR, 7.637; 95% CI, 4.210–13.855; P < 0.001), and impaired renal function before discharge (HR, 3.957; 95% CI, 2.554–6.131; P < 0.001) were independently correlated with in-hospital mortality (Table 5). Cox regression analysis revealed that AKI (HR, 2.159; 95% CI, 1.272–3.664; P = 0.004), AKI without recovery (HR, 5.470; 95% CI, 3.304–9.862; P < 0.001), and impaired renal function before discharge (HR, 2.876; 95% CI, 1.861–4.443; P < 0.001) were independently correlated with 2-year mortality (Table 5).

Table 5.

Multivariate analysis for clinical outcomes.

	CPC 3-5			In-hospital mortality			2-year mortality
	OR	(95% CI)	P-value	HR	(95% CI)	P-value	HR	(95% CI)	P-value
Age, years	1.001	(0.985–1.038)	0.396	0.997	(0.986–1.009)	0.651	1.002	(0.990–1.013)	0.775
Male§	0.950	(0.381–2.370)	0.912	0.855	(0.594–1.231)	0.400	0.928	(0.653–1.317)	0.675
Arrest at public§	0.623	(0.278–1.398)	0.251	0.836	(0.570–1.226)	0.360	0.919	(0.631–1.337)	0.658
Shockable rhythm§	0.202	(0.069–0.590)	0.003	0.574	(0.330–0.999)	0.050	0.431	(0.253–0.732)	0.002
Bystander CPR§	0.446	(0.199–1.001)	0.050	0.659	(0.451–0.963)	0.031	0.660	(0.458–0.952)	0.026
Asphyxia§	1.021	(0.229–4.549)	0.978	0.822	(0.533–1.268)	0.376	1.046	(0.682–1.606)	0.836
Cardiac cause§	0.316	(0.087–1.147)	0.080	0.660	(0.403–1.082)	0.100	0.877	(0.558 –1.379)	0.570
Heart failure§	1.793	(0.340–9.438)	0.491	1.172	(0.699–1.963)	0.547	1.074	(0.654–1.763)	0.777
AKI§	7.684	(2.683–22.010)	<0.001	1.868	(1.079–3.231)	0.026	2.159	(1.272–3.664)	0.004
	CPC 3-5			In-hospital mortality			2-year mortality
OR	(95% CI)	P-value	HR	(95% CI)	P-value	HR	(95% CI)	P-value
Age, years	1.010	(0.978–1.044)	0.543	0.999	(0.987–1.011)	0.861	1.003	(0.991–1.015)	0.636
Male§	1.229	(0.396–3.816)	0.721	0.709	(0.482–1.042)	0.080	0.708	(0.487–1.030)	0.071
Arrest at public§	0.958	(0.363–2.528)	0.931	1.080	(0.720–1.620)	0.709	1.035	(0.702–1.526)	0.861
Shockable rhythm§	0.280	(0.080–0.973)	0.045	0.916	(0.523–1.607)	0.760	0.723	(0.408–1.279)	0.265
Bystander CPR§	0.465	(0.180–1.203)	0.114	0.612	(0.405–0.926)	0.020	0.655	(0.441–0.974)	0.037
Asphyxia§	0.745	(0.103–5.402)	0.771	0.918	(0.577–1.463)	0.720	1.046	(0.662–1.653)	0.848
Cardiac cause§	0.292	(0.062–1.367)	0.118	0.664	(0.400–1.102)	0.113	0.781	(0.483–1.263)	0.314
Heart failure§	5.015	(0.535–47.030)	0.158	1.198	(0.703–2.040)	0.506	1.172	(0.700–1.961)	0.546
AKI without recovery§	5.275	(2.049–13.583)	<0.001	7.637	(4.210–13.855)	<0.001	5.470	(3.034–9.862)	<0.001
	CPC 3-5			In-hospital mortality			2-year mortality
OR	(95% CI)	P-value	HR	(95% CI)	P-value	HR	(95% CI)	P-value
Age, years	1.107	(0.991–1.044)	0.195	1.000	(0.989–1.012)	0.934	1.004	(0.993–1.015)	0.482
Male§	0.923	(0.380–2.237)	0.858	0.800	(0.555–1.154)	0.232	0.835	(0.586–1.191)	0.320
Arrest at public§	0.614	(0.278–1.355)	0.227	0.894	(0.607–1.315)	0.569	0.944	(0.647–1.378)	0.765
Shockable rhythm§	0.310	(0.111–0.867)	0.026	0.844	(0.483–1.476)	0.553	0.641	(0.369–1.113)	0.114
Bystander CPR§	0.422	(0.192–0.926)	0.031	0.579	(0.391–0.857)	0.006	0.610	(0.418–0.892)	0.011
Asphyxia§	0.790	(0.184–3.388)	0.751	0.978	(0.636–1.506)	0.921	1.211	(0.792–1.851)	0.377
Cardiac cause§	0.339	(0.098–1.177)	0.089	0.675	(0.414–1.099)	0.114	0.862	(0.545–1.361)	0.523
Heart failure§	2.038	(0.363–11.459)	0.419	0.809	(0.481–1.361)	0.425	0.820	(0.497–1.353)	0.438
Impaired eGFRpredischarge§	3.164	(1.442–6.940)	0.004	3.957	(2.554–6.131)	<0.001	2.876	(1.861–4.443)	<0.001

AKI, acute kidney injury; CI, confidence interval; CPC, cerebral performance category; CPR, cardiopulmonary resuscitation; eGFR, estimated glomerular filtration rate; HR, hazard ratio; OR, odd ratio;

^§yes vs. no.

When examining the stages of AKI, 57 patients (25.6%) were in stage 1, 62 (27.8%) in stage 2, and 65 (29.1%) in stage 3. Patients in stage 2 or 3 AKI had significantly worse outcomes compared to those without AKI. However, outcomes did not differ between patients with stage 1 AKI and those without AKI (Supplementary Table 3). Unrecovered AKI was associated with higher mortality across all stages and poor neurological outcomes in stage 3 AKI (Supplementary Table 3). The KM survival curves also showed a significant difference in 2-year mortality across AKI stages (log-rank P < 0.001), although no significant difference was observed between patients with stage 1 AKI and those without AKI (Supplementary Figs. 5).

Clinical outcomes of patients with newly diagnosed ESRD

During hospitalization, 24 patients (10.8%) without underlying ESRD received dialysis, including either hemodialysis or continuous veno-venous hemofiltration (CVVH). Among them, only 5 patients (20.8%) were able to discontinue dialysis and recover baseline renal function before discharge (Table 2). However, all 19 patients who were unable to discontinue dialysis died during index hospitalization (Supplementary Table 4).

Clinical outcomes of patients with underlying ESRD

Compared to those without underlying ESRD, patients with underlying ESRD had similar neurological outcomes and all-cause mortality (Table 6).

Table 6.

Clinical outcomes in patients with and without preexisting ESRD.

	All (n = 258)	No ESRD (n = 223)	ESRD (n = 35)	P-value
CPC 3–5, n (%)	202/258 (78.3%)	173/223 (77.6%)	29/35 (82.9%)	0.481
In-hospital mortality, n (%)	165/258 (64.0%)	139/223 (62.3%)	26/35 (74.3%)	0.171
6-month mortality, n (%)	167/234 (71.4%)	140/199 (70.4%)	27/35 (77.1%)	0.412
1-year mortality, n (%)	170/225 (75.6%)	143/191 (74.9%)	27/34 (79.4%)	0.570
2-year mortality, n (%)	173/216 (80.1%)	145/185 (78.4%)	28/31 (90.3%)	0.123

CPC, cerebral performance category; ESRD, end stage renal disease.

Discussion

The main findings of the study comprised the following: (1) patients with OHCA frequently experience renal function impairment, and (2) impaired pre-discharge renal function but not renal function on arrival was a strong predictor of unfavorable neurological outcomes and mortality in patients with OHCA without underlying ESRD. (3) AKI was associated with poorer neurological outcomes and higher mortality in patients with OHCA without underlying ESRD, and (4) patients with OHCA who recovered from AKI had better prognosis than those who did not recover from AKI. (5) Finally, patients with OHCA and underlying ESRD had similar clinical outcomes compared to those without underlying ESRD.

Renal function impairment is common after OHCA.9, 4, 5, 6 In an observational study of 503 patients with cardiac arrest, in which 77% had OHCA, 29.4% of patients encountered AKI within 48 h.³ In another cohort study with 580 patients with OHCA, 48.3% patients developed stage-3 AKI in 48 h.⁴ A single-center cohort study involving 842 cardiac arrest survivors consisting of 69% patients with OHCA revealed an AKI prevalence of 69.8% within 7 days.⁶ However, these studies used creatinine levels at admission as the baseline if no previous data were available when diagnosing AKI, which means that patients with AKI on arrival may have been overlooked.

Two other studies applying the same method to define baseline creatinine as our study reported AKI incidences of 60% and 64% among patients with OHCA in the early period after ROSC.5, 9, 15 There were different definitions and observation periods for AKI in previous studies, and the incidence of AKI might vary. Our study demonstrated a higher rate of AKI (82.5%) due to the consideration of not only AKI on arrival, but also AKI events throughout the whole hospitalization.

Renal dysfunction was reported to be an independent predictor of mortality in the general population17, 18, 19, 20 and is also associated with poor clinical outcomes for patients with OHCA.² The SOS-KANTO 2012 study was a multicenter prospective study involving 5,112 patients with cardiogenic OHCA from Japan. In that study, those with a first-obtained eGFR ≥60 mL/min/1.73 m² had better 3-month survival and neurological outcomes compared to those with eGFR of 45–59, 30–44, and <30 mL/min/1.73 m².² In our study, however, we found that impaired renal function before discharge but not on arrival was independently associated with higher mortality and worse neurological outcomes for patients with OHCA.

However, there were some differences between the SOS-KANTO 2012 study and the present study. First, there were some differences in the causes of OHCA. The cause of OHCA was mainly cardiac causes in the SOS-KANTO 2012 study, while only 50% had cardiac causes in our study. Second, the SOS-KANTO 2012 study applied a different formula for eGFR (the standard formula for Japanese subjects using serum creatinine obtained immediately after hospital arrival at the emergency department).²¹ Our findings suggested that impaired renal function on arrival was not a reliable predictor. Interestingly, our study is the first to report that pre-discharge renal function is a better predictor of clinical outcomes in patients with OHCA.

AKI is well known to be associated with worse clinical outcomes in patients with OHCA.3, 4, 5, 6 In a single-center prospective cohort study with 580 patients with OHCA, a higher 30-day mortality was observed among the AKI population (OR, 1.60, 95% CI: 1.05–2.43, P = 0.03).⁴ In other retrospective research involving 842 patients with cardiac arrest, AKI was significantly associated with higher mortality (HR, 1.35; 95% CI: 1.07–1.71, P = 0.01) and poor neurological outcomes (OR, 2.27; 95% CI: 1.45–3.57, P < 0.001).⁶ In addition, a 10-year observational study with 504 patients who had cardiac arrest demonstrated that the median survival was significantly shorter for patients with AKI than those without AKI (0.07 years vs. 6.5 years, P < 0.001).³ Likewise, in our study, AKI was a solid factor associated with higher mortality and unfavorable neurological outcomes throughout the follow-up period.

Nevertheless, among AKI patients, those who achieved AKI recovery in cardiac arrest appeared to have a better prognosis. Although there is limited research, recovery from AKI was reported to be a potent predictor of better survival and neurological outcomes at discharge among 175 AKI patients in a retrospective multi-center cohort study of OHCA (OR 8.308, 95% CI 3.120–22.123, P < 0.001; OR 36.822, 95% CI 4.097–330.926, P = 0.001).⁹ Likewise, in our study, AKI patients whose renal function had recovered to baseline demonstrated a remarkably better prognosis in terms of survival and neurological outcomes compared to those without AKI recovery. In addition to assessing mortality at time points based on previous studies, we employed a longer observation period of up to 2 years, which is one of the strengths of our study. Furthermore, patients who recovered from dialysis and achieved baseline renal function in our study showed remarkably better survival compared to those who were unable to discontinue dialysis (Supplementary Table 4). These findings may further imply that AKI recovery is a strong predictor of survival in patients with OHCA.

Those with ESRD were reported to have a higher risk of sudden cardiac death,11, 22 and ESRD was listed as one of the contributing variables in the NULL-PLEASE scoring system to predict in-hospital mortality among patients with OHCA.²³ However, patients with OHCA and underlying ESRD did not show higher mortality or worse neurological outcomes in our study. Interestingly, a national database analysis involving over 200,000 patients in Taiwan demonstrated that patients with ESRD had a higher chance of ROSC (adjusted OR, 2.47, 95% CI: 1.90–3.21, P < 0.001) and a better 30-day hospital survival rate (log-rank P < 0.001).¹¹

As a marker of poor prognosis in critical ill patients, AKI itself may also be a contributor to the increased morbidity and mortality. The concept of cross-talk between organs suggests that AKI may impact other organs via several metabolic or humoral pathways.24, 25, 26, 27, 28 The potential mechanisms include uremic toxin accumulation, inflammation, fluid overload, electrolyte disturbance, and acid-base imbalance.²⁴ Thus, our results may provide motivation to pursue kidney-function recovery to achieve preserved renal function before discharge for patients with OHCA to attain better clinical outcomes. However, due to chronic toxin tolerance and vascular compliance training during regular dialysis,11, 29 the clinical outcomes are not necessarily worse in patients with ESRD than those without ESRD.

Limitations

This study has some limitations. First, it was a single-center study with a relatively small study population. However, there are a limited number of similar studies with follow-up durations as long as that in our research, and there are only a few papers discussing the influence of AKI recovery in patients with OHCA. Nevertheless, further studies with larger sample sizes are required. Second, there were many patients who lacked data on previous renal function, and the exact rate of chronic kidney disease in this study population is unknown, which may have interfered with the assessment of AKI. Third, there were some latent confounding factors when calculating mortality. For example, we did not exclude the patients who had a “do not attempt resuscitation” order after cardiac arrest. Fourth, we did not obtain information on hemodynamic status, inotropic drug use, or iodine contrast agents, which may affect renal function after ROSC. Finally, during the latter half of enrollment, the outbreak of the COVID-19 pandemic inevitably affected the medical protocols and the prognoses of patients, especially in emergency departments.30, 31, 32.

Conclusions

For patients with OHCA, there were significant correlations of AKI and impaired pre-discharge renal function with worse clinical outcomes among those without underlying ESRD. Among AKI patients, those who recovered appeared to have a better prognosis than those who did not recover. However, ESRD was not associated with worse clinical outcomes among patients with OHCA.

CRediT authorship contribution statement

Hao-Wei Lee: Writing – original draft, Investigation, Formal analysis, Data curation, Conceptualization. Ming-Jen Kuo: Writing – original draft, Data curation. Pai-Feng Hsu: Writing – original draft, Data curation. I-Hsin Lee: Writing – original draft, Data curation. Chih-Yu Yang: Writing – review & editing, Methodology. Teh-Fu Hsu: Writing – review & editing, Data curation. Chorng-Kuang How: Writing – review & editing, Data curation. Yenn-Jiang Lin: Writing – review & editing, Data curation. Chin-Chou Huang: Writing – review & editing, Supervision, Resources, Project administration, Methodology, Investigation, Funding acquisition, Formal analysis, Data curation, Conceptualization.

Funding

This work was supported by research grants V113C-032, V113EA-012, and V114C-005 from Taipei Veterans General Hospital, Taipei, Taiwan, and research grant NSTC111-2314-B-A49A-509-MY3 from National Science and Technology Council, Taiwan. The funder had no role in the study design, data collection and analysis, decision to publish, or manuscript preparation.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Appendix A. Supplementary material

The following are the Supplementary data to this article: