Background
Cerebrovascular disease is a common cause of morbidity and mortality in patients with end-stage renal disease (ESRD). Patients with ESRD have advanced atherosclerosis of the cerebral vasculature [
1]. Stroke rates for ESRD patients are markedly higher than for the general population [
2]. Intracerebral hemorrhage (ICH) is a type of stroke that involves bleeding within the brain tissue itself. It is well known that a bleeding tendency is a major concern among ESRD patients. In addition, the routine administration of heparin during hemodialysis may increase the risk of bleeding. ICH is more common in dialysis patients than in the general population [
2], and is a devastating and life-threatening condition that requires urgent intervention. The onset of ICH is usually sudden and, unlike ischemic stroke symptoms, ICH symptoms tend to appear without warning. ICH in dialysis patients has a poor prognosis and a very high mortality rate, despite advances in diagnosis and treatment [
3,
4]. It is important to identify dialysis patients at risk for developing ICH and the prognostic factors after ICH.
ICH is more common among Taiwanese than Caucasians [
5]. The incidence and prevalence rates of ESRD are also high in Taiwan [
6]. However, little is known about the annual incidence of ICH in the Taiwanese dialysis population. There are also few nationwide studies analyzing the predictors of ICH and the effect of comorbidities on survival following ICH in patients on chronic dialysis. In the study, we utilized a large data set from the National Health Insurance Research Database (NHIRD) to identify predicting factors for intracerebral hemorrhage in dialysis patients as the primary objective. The secondary objective was to report the clinical outcomes and identify the independent predictors of ICH mortality in these patients.
Discussion
This nationwide study was designed to investigate the risk factors of ICH occurrence and predictors of mortality following ICH in prevalent dialysis patients using the Taiwan National Health Insurance (NHI) database. We found a high incidence rate of ICH among chronic dialysis patients. Male gender, middle age, hypertension, and history of old stroke were independent factors of incident ICH. Diabetes mellitus and malignancy were significant independent predictors of mortality after ICH. Long-term outcome of ICH survivors was much poorer among the group with a history of prior stroke than among those without a history of prior stroke.
The present findings confirm the high rate of ICH among dialysis patients compared with the general population reported in a prior study [
2]. We also showed that hypertension and male gender were independent predictors of hemorrhagic stroke among patients on chronic dialysis. These observations were consistent with recent study by Yoo et al., who evaluated predictive variables of ICH in dialysis patients using the United States Renal Data System (USRDS) database [
8]. Similar results were also observed in previous study by Seliger’s [
9], in which increased ICH rate was associated with a raised mean arterial pressure of at least >10 mmHg. Thus, hypertension is a risk factor for ICH for both the general population and dialysis patients [
10]. While higher blood pressure has relevance to incidental ICH, the effect of it on all-cause mortality seems to be different. We did not find an association between hypertension and all-cause mortality after ICH in patients undergoing chronic dialysis. Our finding that a history of hypertension is not an independent predictor of death among chronic dialysis patients is in agreement with the work of Tetri and colleagues, who investigated the factors associated with outcome after ICH in the general population [
11]. However, the observation of a null association between elevated blood pressure and poor outcome of dialysis patients should be interpreted with some caution. We did not differentiate patients using antihypertensive medications from those without treatment in this analysis, and it seems probable that the effect of hypertension is confounded by the use of these drugs.
This study found that previously diagnosed diabetes predicted mortality. Diabetes was also demonstrated to be an independent determinant of death in the general population suffering from ICH [
12]. The exact reason why diabetes is associated with the risk of mortality following ICH remains unclear. The effect of DM on survival may be explained by the severity of hemorrhagic events. Diabetes induces structural and functional changes in the microvasculature that lead to fibrinoid necrosis of blood vessel walls and microaneurysm formation. It is possible that these changes in vessel walls make diabetic patients more likely to develop larger sized hematomas than non-diabetics. There is evidence showing that patients with an elevated glucose level due to diabetes were associated with hemorrhages of a larger size compared to patients with a normal glucose level [
13], and that the volume of ICH was a strong predictor of mortality after primary ICH [
14,
15]. In addition, patients with uncontrolled diabetes and a high glucose level may also have a higher risk of infectious or cardiovascular complications during hospitalization.
Moreover, we found that a previous medical history of stroke is strongly associated with both a higher incidence of ICH and higher mortality risk following ICH among dialysis patients. The risk of ICH was reported to increase several times after ischemic infarct in the general population [
16,
17]. It is well established that dialysis patients have a much higher incidence of cerebral ischemic events. Therefore, it is prudent to avoid over-anticoagulation in these patients and to monitor them more closely when providing antithrombotic drugs to prevent recurrent ischemic infarction. A complex series of pathophysiological events are initiated in the brain after ischemic injury. It has been reported that matrix metalloproteinases-2 (MMP-2), MMP-3 and MMP-9, which are capable of digesting basal lamina and extracellular matrix components and lead to subsequent blood brain barrier breakdown and hemorrhagic transformation, were abnormally upregulated in the ischemic region and peri-infarct area [
18‐
20]. Increased expression of MMP-9 of the brain tissue was also observed in patients suffering from hemorrhagic stroke [
20]. Moreover, the increased activities of inducible matrix proteases may sustain for several months or years after brain injury [
21]. It is possible that physiologic and pathologic responses to stroke cause chronic damage to vessel wall components and pose a sustained risk of subsequent ICH among dialysis patients.
The present study showed the in-hospital mortality rate among dialysis patients hospitalized with ICH was 36.15%. This result correlated with previous reports from other Asian countries. In a study from Korea, the 30-day mortality rate among 102 ESRD cases of spontaneous ICH was 53.9% [
22]. Another study from Japan using registered data showed that the death rate of dialysis patients after cerebral hemorrhage was 27.9% [
23]. A recent study from Taiwan reported that the mortality rate within 30 days after spontaneous ICH in the general population was 19.8% [
24]. Thus, the outcome of ICH is poor among patients on chronic dialysis, and can be worse than in the general population.
Whether gender has an impact on survival after stroke remains controversial. It has been suggested that estrogen may exert neuroprotective effects after brain injury by way of stimulating insulin-like growth factor-I synthesis, lessening inflammatory processes, scavenging free radicals, and stimulation of intrinsic anti-apoptotic pathways [
25,
26]. A recent study demonstrated a significant survival advantage for females over males over 3-year period after primary ICH [
27]. Another study in Denmark also revealed a female superiority in survival time after stroke [
28]. However, we did not find such an association between female gender and lower all-cause mortality rates after hemorrhagic stroke among chronic dialysis patients.
Peritoneal dialysis (PD) was claimed to have some benefits, such as minimal anticoagulation and improved cardiovascular or intracranial stability as a consequence of smaller changes in serum osmolality. There is also evidence showing that intracranial pressure increased significantly in patients receiving conventional intermittent hemodialysis [
29]. So patients on PD were expected to experience fewer acute cerebral events compared with those on hemodialysis (HD), and some authors suggest a survival advantage with the use of PD instead of HD for chronic dialysis patients with cerebral hemorrhage [
30,
31]. In the USRDS study by Yoo mentioned above, use of peritoneal dialysis as long-term renal replacement therapy was associated with lower risk of ICH compared to use of hemodialysis [
8]. However, there is a contradictory result. A recent work showed an even worse prognosis for patients receiving PD after ICH [
4]. Our analysis did not support the association of PD as a dialysis modality with less incidental ICH or decreased mortality following ICH among chronic dialysis patients.
The strength of our study lies in its large population size, and high likelihood that nearly all relevant patients were collected from the database. The limitations of this study include a lack of laboratory data, and a lack of information about the causes of death, because they were not included in the Taiwan NHI database. We used ICD-9 code 431 to identify cases of ICH, so patients with isolated subarachnoid hemorrhage (ICD-9 code: 430) would not be included in our analysis. We could not affirmatively determine the duration and severity of comorbidities, which may influence outcome of our subjects, in the databases-derived study. We did not have access to detailed information about ICH and dialysis, so the severity and location of the ICH, blood pressure values, and dialysis adequacy cannot be assessed. Surgery was required in patients with large hematoma formation or refractory elevated intracranial pressure. Operation may have an influence on short-term mortality or long-term disability in our study subjects but we did not analyze the proportion of patients receiving surgical treatment after ICH. Moreover, we did not evaluate drug treatment in ESRD population, such as antiplatelet or anticoagulants, which may be associated with incidence of ICH. We also did not differentiate ischemic stroke from hemorrhagic stroke in determining comorbidities. Further large prospective studies with detailed evaluations are needed to confirm our findings.
Competing interests
The authors declare that they have no competing interests.
Authors’ contributions
CYL and FMS: Draft the article and provide intellectual content of critical importance to the work described. HAC, JJW, CCW and CCC: Design, analysis and interpretation of data. CCC and YJL: Final approval of the version to be published. All authors read and approved the final manuscript.