Impact of renal impairment on outcomes of intracerebral hemorrhage: a systematic review and meta-analysis

Abstract

Background

Chronic kidney disease (CKD) is linked to a number of cardiovascular complications, including intracerebral hemorrhage (ICH). However, the potential association between CKD and the outcomes of ICH are still inconclusive. This review aimed to clarify the relationship between CKD and ICH outcomes, such as mortality, functional disability, and length of hospitalization.

Methods

A comprehensive literature search was done in MEDLINE, Scopus, and Cochrane Central Register of Controlled Trials (CENTRAL), Google Scholar, EMBASE databases from inception until November 2023. Observational studies examining adult ICH patients with CKD were included. Risk of bias was evaluated using Newcastle–Ottawa Scale, and data synthesis was performed by random-effects meta-analysis.

Results

Twenty studies were included, encompassing a broad range of CKD and ICH patients. The meta-analysis demonstrated a significant association between CKD and adverse ICH outcomes. CKD patients had significantly higher mortality rates (pooled OR = 2.21; 95%CI 1.88–2.59; I² = 98.7%) and functional disability (pooled OR = 1.51; 95%CI 1.30–1.75; I² = 97.2%). CKD also showed potential associations with increased duration of hospital stay, though this outcome was less consistently reported.

Conclusion

Our findings showed that CKD was linked to higher rates of mortality, and functional disability in patients with ICH, and was associated with extended hospitalization. Our results underscore the need for a comprehensive, multidisciplinary approach to managing this complex patient population. Further research should elucidate the underlying mechanisms of this association and inform strategies for improving patient outcomes. The integration of nephrological and neurological care may significantly benefit the management and prognosis of ICH in patients with CKD.

Supplementary Information

The online version contains supplementary material available at 10.1186/s40001-025-03379-5.

Introduction

Intracerebral hemorrhage (ICH) is a life-threatening cerebrovascular event characterized by the spontaneous rupture of blood vessels in the brain, resulting in the accumulation of blood within the brain parenchyma [1]. ICH is a major cause of mortality and long-term disability worldwide, with incidence rates that vary widely across different geographical locations and ethnic groups [2]. Despite advances in medical and surgical interventions, the overall prognosis of ICH remains dismal, and survivors are often left with significant functional impairment [3]. Chronic kidney disease (CKD), is a pervasive and escalating health issue [4–6]. The link between CKD and cerebrovascular diseases, including ICH, became a focus of extensive research. The pathophysiological mechanisms of this association are complex and include hypertension, platelet dysfunction, impaired coagulation, and inflammation, which all play crucial roles in the onset and progression of ICH [7].

The recent studies have demonstrated a significant association between CKD and poor outcomes in patients with ICH, including higher mortality rates and worse functional outcomes, but with inconsistent results [8–10]. The heterogeneity in observed outcomes may be due to differences in study designs, populations, definitions of CKD and outcomes, and adjustment for potential confounding factors. Given the substantial burden of both ICH and CKD, it is crucial to gain a comprehensive understanding of the impact of CKD on ICH outcomes to inform patient management strategies and prognosis, especially by the nursing personnel. The goal of this systematic review and meta-analysis is to summarize available evidence on the association between CKD and outcomes of ICH.

Methods

Eligibility criteria

Inclusion criteria were as follows: (a) observational studies; (b) adult patients diagnosed with ICH; (c) studies that reported on the association between renal impairment and outcomes of ICH. The primary outcome was the association between renal impairment and mortality in patients with ICH. Other outcomes were functional outcomes and length of hospital stay.

Exclusion criteria were as follows: (a) case reports, editorials, reviews, and non-original studies; (b) studies without a clear definition of renal impairment; (c) studies without available data on outcomes of ICH.

Search strategy

A systematic literature search was conducted in MEDLINE, Scopus, and Cochrane Central Register of Controlled Trials (CENTRAL), Google Scholar, EMBASE databases from inception till November 2023 for studies published in any language (no language restrictions has been kept). We have also manually searched the reference lists of selected studies and relevant reviews. The search strategy included the Medical Subject Headings (MeSH) and text words related to “intracerebral hemorrhage”, “renal impairment”, “chronic renal disease”, “chronic kidney disease”, and “mortality”. The detailed search strategy has been provided in Supplementary appendix.

Study protocol registration

We have registered our protocol before start at PROSPERO, the number is CRD42023441519 and no deviations made from the registered protocol.

Study screening process

The initial screening of titles and abstracts of all the records retrieved from the databases were independently performed by two authors based on the pre-specified eligibility criteria. The full texts of eligible articles were reviewed independently by two authors. Any disagreements in the study selection process were resolved through discussion or, if required, third-party arbitration.

Data extraction

Two authors independently extracted data from the selected studies using a standardized data extraction form. The extracted data included: first author’s name, publication year, study design, country of origin, sample size, patients’ characteristics, definition of renal impairment, ICH outcomes, and adjustments for potential confounding factors.

Risk of bias assessment

The risk of bias was examined using Newcastle–Ottawa Scale (NOS) for observational studies. It consists of three parameters: the selection domain which gives score from 0 to 4 stars, the comparability domain gives score from 0 to 2 stars, and the ascertainment of outcome or exposure gives score from 0 to 3 stars. Studies which are scoring 7 stars or above were identified as lower risk of bias studies. Any disagreements were resolved through discussion or, if required, third-party arbitration [11]. We quantified agreement as Cohen’s kappa for the overall judgement dichotomised as low risk (≥ 7 stars) vs high risk (< 7 stars). The Cohen’s kappa value was 1.0 and the overall agreement was 100% indicating full agreement between both reviewers.

Data synthesis and analysis

The effect measures were reported as pooled odds ratios (OR) for dichotomous outcomes with 95% confidence intervals (CIs). Random-effects model was employed using inverse variance method to pool the data given the anticipated clinical heterogeneity [12]. Forest plot was depicted to show the visual representation of pooled estimate and other important findings like heterogeneity related statistics. Heterogeneity among studies was evaluated using the I² statistic, with values above 75% indicating substantial heterogeneity. We have performed sensitivity analysis by excluding one study at a time to examine the robustness of the results.

Subgroup analysis was performed based on the pre-decided time of outcome assessment and quality of studies. For mortality and functional outcomes, we organised studies by the timing of outcome ascertainment using anchor bins that reflect common clinical follow-up points and reporting conventions: (i) in-hospital (within hospital discharge) (ii) 1 month, (iii) 3 months and (iv) 12 months. When authors reported timepoints as “in-hospital/discharge,” “28–30 days,” “90 days/3 months, or “12 months/1 year,” outcomes were mapped to the corresponding bin. Publication bias assessment was done using funnel plots and Egger’s test. All the reporting was done as per PRISMA guidelines [13].

Results

Search results

In primary screening across the databases identified 1265 studies. After deduplication, 78 full-text articles were retrieved, and underwent secondary screening. Finally, 20 studies that satisfied the eligibility criteria were included (Fig. 1) [8–10, 14–30].

Fig. 1.

Search strategy

Characteristics of the included studies

Most studies were from the United States followed by China and Japan. Most studies were conducted as secondary data analysis and prospective studies, with sample sizes varying widely from 101 to 328,728 participants. Renal impairment was commonly defined as a glomerular filtration rate (GFR) below 60 mL/min/1.73 m, though some studies use end-stage renal disease or ICD-9 codes as criteria. Functional impairment was frequently assessed using tools like the modified Rankin Scale (mRS) with varying definitions. Outcome assessment times ranged from in-hospital to 12 months post-admission. Mean ages of the patients ranged from 54.2 to 74.6 years, and hematoma volumes, when reported, ranged from 7.6 mL to 62.1 mL. Overall, 9 out of 20 studies had a low risk of bias (Table 1). Most studies had lower scores because of the Comparability (limited covariate adjustment; e.g., no ICH severity measure) and Outcome/Exposure (mortality ascertained from administrative codes without validation or unclear follow-up), with fewer in Selection (single-centre or non-consecutive sampling).

Table 1.

Studies included in meta-analysis

Author and year	Location	Study design	Sample size	CKD definition	Tool and definition of functional impairment	Time of outcome assessment	Mean age (in years)	Mean/median hematoma volume (in mL)	Risk of bias# (NOS score)
Beuscher 2020	Germany	Observational cohort study	368 after PSM	GFR below 60 mL/min/1.73 m	mRS 4–6	12 months	74 in CKD patients and 72 in non-CKD patients	12.2 in CKD patients and 13.1 in non-CKD patients	Low (7)
Cutting 2014	USA	Prospective	411	GFR below 60 mL/min/1.73 m	NA	In-hospital outcomes	61.4	11.2	Low (7)
Fukuda-Doi 2021	Japan	Post-hoc analysis of RCT	974	GFR below 60 mL/min/1.73 m	mRS 4–6	1 month	66 in patients with GFR < 60; 65 in GFR 60–89 and 58 in GFR > 90	8.1 in patients with GFR < 60; 9.6 in GFR 60–89 and 10.9 in GFR > 90	Low (8)
Khatri 2019	USA	Secondary analysis of data	328,728	ICD-9 codes	Based on discharge destination	In-hospital outcomes	68.2 in patients with renal failure; 68.6 in patients without renal failure	NR	Moderate (5)
Kim 2013	Korea	Retrospective	1943	End stage renal disease	NA	1 month	62.3	NR	High (3)
Li 2022	China	Secondary data analysis	85,167	GFR below 60 mL/min/1.73 m	NA	In-hospital outcomes	62.9	NR	High (3)
Lui 2023	USA	Secondary analysis of data	211,266	End stage renal disease	Based on discharge disposition	In-hospital outcomes	67.4	NR	Moderate (4)
Luo 2022	USA	Secondary data analysis	2682	ICD-9 codes recorded in the MIMIC-III database	NA	12 months	66.5	NR	High (3)
Miyagi 2015	Japan	Prospective	203	GFR below 60 mL/min/1.73 m	mRS 5–6	3 months	74.6 in patients with GFR < 60; 65.2 in GFR 60–89 and 61 in GFR > 90	11.9 in patients with GFR < 60; 9 in GFR 60–89 and 11.8 in GFR > 90	Low (7)
Molshatzki 2011	Israel	Prospective	128	GFR below 60 mL/min/1.73 m	NA	12 months	71.7	NR	Low (7)
Ovbiagele 2014	USA	Secondary data analysis	113,059	GFR below 60 mL/min/1.73 m	NA	In-hospital outcomes	72.5 in CKD patients and 67.3 in non-CKD patients	NR	Low (8)
Rhoney 2012	USA	Retrospective	101	GFR below 90 mL/min/1.73 m	NA	In-hospital outcomes	61 in CKD patients and 56 in non-CKD patients	NR	High (3)
Saeed 2015	USA	Secondary data analysis	614,454	ICD-9 codes	Discharge disposition	In-hospital outcomes	68 in renal failure patients and 69 in patients without renal failure	NR	Moderate (6)
Sakamoto 2014	Japan	Retrospective	366	Requirement of hemodialysis	NA	In-hospital outcomes	64 in patients with renal failure; 69 patients without renal failure	NR	Low (7)
Shimoyama 2013	Japan	Retrospective	507	GFR below 60 mL/min/1.73 m	NA	15 days	60.5 in CKD patients and 70 in non-CKD patients	10.6 in CKD patients and 7.6 in non-CKD patients	Moderate (6)
Tung 2021	Thailand	Secondary data analysis	195,650	ICD codes	NA	In-hospital outcomes	63.75	NR	High (2)
Wang 2023	Taiwan	Secondary data analysis	4036	GFR below 60 mL/min/1.73 m	NA	1 month	61.4	NR	Low (7)
Xu 2016	China	Retrospective cohort	302	End-stage renal disease	NA	3 months	54.2 in CKD patients and 56.6 in non-CKD patients	62.1 in CKD patients and 30.4 in non-CKD patients	High (3)
You 2016	China	Prospective	365	GFR below 60 mL/min/1.73 m	mRS ≥ 3	3 months	64.1	10 in patients with GFR < 60; 10.9 in GFR 60–89 and 9.6 in GFR > 90	Low (8)
Zheng 2016	China	Secondary data analysis	2623	GFR below 60 mL/min/1.73 m	mRS ≥ 3	3 months	59 in patients with GFR < 60; 70 in GFR 60–89 and 71 in GFR > 90	NR	High (3)

CKD Chronic kidney disease, GFR Glomerular filtration rate, ICD International Classification of Diseases, mRS modified Rankin Scale, NA Not applicable, NR Not reported, PSM Propensity score matching, RCT randomized controlled trial, USA United States of America

^#Newcastle Ottawa Scale (NOS) score: 0–3 = high risk of bias; 4–6 = moderate risk of bias; 7–9 = low risk of bias

Mortality

A total of 19 studies provided information on the difference in mortality outcome between CKD and non-CKD patients. The pooled OR was 2.21 (95%CI 1.88–2.59; I² = 98.7%), signifying a significant differences in the mortality rates between the two groups (p < 0.001) (Fig. 2). Funnel plot was slightly asymmetrical (Fig. 3). However, Egger’s test had non-significant p-value (p = 0.07), indicating the lack of publication bias.

Fig. 2.

Forest plot for mortality outcome

Fig. 3.

Funnel plot for mortality outcome

Subgroup analysis was performed based on the time of outcome assessment (Fig. 4). Outcomes assessed after 3 months had the highest pooled OR of 3.62 (95%CI 1.55–8.47) when compared with other time points. However, mortality rate was significantly higher in CKD group irrespective of the time of outcome assessment. Subgroup analysis based on the risk of bias (Fig. 5) did not show significant difference in mortality between high risk and low risk of bias studies.

Fig. 4.

Subgroup analysis for mortality outcome based on timing of outcome assessment

Fig. 5.

Subgroup analysis for mortality outcome based on quality of studies

Functional disability

A total of 10 studies provided information on the differences in functional disability between CKD and non-CKD patients. The pooled OR was 1.51 (95%CI 1.30–1.75; I² = 97.2%), signifying evidence of significant differences (p < 0.001) (Fig. 6). Funnel plot was slightly asymmetrical (Fig. 7). However, Egger’s test had nonsignificant p value (p = 0.17), indicative of no publication bias.

Fig. 6.

Forest plot for functional disability

Fig. 7.

Funnel plot for functional disability

Discussion

This review, involving 20 studies and a large, heterogeneous patient cohort, affirms the significant association between CKD and poor outcomes in patients with ICH. Our findings underline a robust association of CKD with higher mortality and increased functional disability in ICH patients. Our results are consistent with the previous systematic reviews on CKD and acute stroke [31, 32]. This acknowledges the intertwining pathophysiology of the kidneys and brain in shaping the disease course and outcomes.

It is important to note that CKD patients had approximately twice the odds of death compared to non-CKD counterparts following an ICH event. This result shows the crucial role of renal function in prognosis of ICH. CKD and ICH share common risk factors such as diabetes mellitus, hypertension, and vascular inflammation, but the relationship between these two conditions appears to extend beyond shared risk factors [33]. From a nephrological perspective, the uremic environment, characteristic of CKD, may directly and indirectly impact cerebral homeostasis [34]. Accumulation of uremic toxins and subsequent systemic inflammation may possibly cause oxidative stress, endothelial dysfunction, and disruption of the blood–brain barrier [35]. This can potentially exacerbate cerebral injury caused by ICH and impair neurological recovery. In addition, CKD has been linked in few of the evidences to platelet dysfunction, impaired coagulation, and a state of chronic inflammation, all of which are detrimental to the brain’s integrity and resilience to injury [36, 37]. These pathophysiological changes can further aggravate ICH, thus creating a vicious cycle of brain- kidney damage.

From a neurological standpoint, brain and kidneys share unique vascular properties. Both organs have low-resistance, high-flow vascular beds that are susceptible to damage from hypertension and other vascular events [38]. Autoregulatory dysfunction, a common occurrence in CKD, can lead to cerebral hypoperfusion, predisposing patients to lacunar infarcts and white matter lesions [39]. These pathological changes can increase the vulnerability of brain to ICH and influence the functional outcomes of the ICH event.

Further, our meta-analysis indicates a considerable impact of CKD on functional disability among ICH survivors. The observed association can be attributed to various factors, such as the high prevalence of neurological and cognitive complications in CKD patients [40]. CKD-related metabolic and vascular changes can increase the vulnerability of the brain to damage and impede its recovery following an ICH event [34]. Moreover, CKD and ICH share several risk factors such as hypertension and diabetes, which independently increase the risk of disability [33].

Interestingly, the impact of CKD on mortality was highest when assessed after 3 months. It implies that the detrimental effect of CKD on ICH is not confined to the acute phase but persists or even amplifies over time. This observation might be due to the cumulative impact of CKD-associated complications or the progressive nature of CKD itself. Furthermore, equal distribution of the effect of CKD on mortality among studies with high and low risk of bias ascertains the validity and reliability of our findings.

However, our study has limitations. First, the presence of considerable heterogeneity among the included studies could have influenced the estimates. Although we attempted to address this through a random-effects model and subgroup analyses, residual heterogeneity persisted. This discrepancy might be due to variability in the study populations, definition of CKD, timing of outcome assessment, and adjustments for potential confounding factors. Second, despite our thorough search strategy, the potential of publication bias cannot be ruled out. Furthermore, due to the limited number of studies, publication bias assessment was not feasible for functional disability. Future studies are warranted to provide more robust evidence regarding this association.

The clinical implications of our findings are profound. For healthcare providers, the management of ICH patients with CKD presents significant challenges that require a multidisciplinary approach, encompassing nephrological and neurological care. Clinicians should be vigilant in monitoring CKD patients for neurological symptoms. Moreover, aggressive management of modifiable CKD complications might attenuate the risk of poor ICH outcomes.

The findings of our meta-analysis underscore the need for a paradigm shift in the management of ICH in CKD patients. The conventional silo-based approach, where nephrologists manage CKD and neurologists manage ICH, may not suffice in addressing the complex needs of these patients. A more integrated approach, with cross-disciplinary collaboration and shared decision-making, can optimize patient outcomes. Nursing intervention in the multidisciplinary team is crucial, as their keen observations and patient management skills can significantly influence outcomes. Their role in implementing tailored rehabilitation plans, administering prescribed medications, and educating patients about disease management and lifestyle modifications is paramount. Multidisciplinary teams, where nephrologists, neurologists, and nursing staff jointly oversee patient care, from the acute management of ICH to long-term rehabilitation and CKD care, would significantly improve outcomes in this vulnerable population of patients. Future research should focus on identifying effective strategies to improve the prognosis of ICH in CKD patients, potentially through targeting the shared pathophysiological mechanisms. Furthermore, large scale longitudinal studies are needed to evaluate the effect of CKD management on ICH outcomes.

Nurses play a critical role in managing ICH and CKD patients, and their continuous patient observation, timely identification of symptom changes, delivery of essential care, and patient education on lifestyle modifications are imperative to ensure favorable outcomes. Public health initiatives should focus on prevention strategies targeting the shared risk factors for CKD and ICH, such as blood pressure control, glycemic management, and lifestyle modifications. Such preventive measures can have a substantial impact on the burden of CKD and ICH, given the increasing prevalence and high mortality rates associated with these conditions.

Conclusion

Our review demonstrates a significant association between CKD and poor outcomes in ICH, including higher mortality and functional disability. Further studies are needed for a more comprehensive understanding the importance of the brain-kidney axis in shaping the clinical course of ICH. Integrated care models and therapeutic strategies would improve the outcomes of CKD patients with ICH with low to moderate level evidence certainty.

Author contributions

JZ: conceptualization. JZ, FS and SL: methodology, systematic search, data extraction, formal analysis, software, visualization, and JZ: manuscript draft and revision.

Funding

None.

Data availability

Data can be obtained from the corresponding author upon reasonable request.

Declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.