The prognostic value of the Charlson comorbidity index in aged patients with intracerebral hemorrhage

Intracerebral hemorrhage (ICH) accounts for 6.5 to 19.6% of all strokes [1, 2], and it is significantly more common in the elderly population [3], accounting for approximately 25% of all strokes [4]. As the population continues to age, the prevalence of ICH in increasingly aged patients will increase accordingly [5]. Despite numerous advances in stroke management and neurocritical care, ICH remains the most devastating cerebrovascular disease subtype with significant rates of disability and mortality. The 1-year survival from ICH is approximately 40%, and older age is associated with an increased risk of 30-day death [3, 6].

Over the past few decades, great advances have been made in neurocritical care that may help improve the outcome of cerebral hemorrhage. For these aged ICH patients, stratification is of great necessity to appropriately allocate limited medical resources and counsel both patients and their families in goals of cure. However, considering the high morbidity and mortality of ICH in the elderly population and the heavy burden it may impose on families and society, it is vital that these decisions are based on accurate and repeatable information. Therefore, it is essential to reevaluate previously validated scores, such as the Charlson Comorbidity Index (CCI), to assess whether their prognostic ability can improve with medical advances [7].

Comorbidities are common in ICH patients, especially in elderly individuals, [5, 8] and they are widely considered as one of the factors affecting the outcomes of stroke [9‐13], but few studies have focused on hemorrhagic stroke. The CCI accounts for multiple comorbidities by creating a sum score weighted according to the presence of various conditions, which was originally developed against 1-year death, and its validity as a predictor of other outcome measures, either in the short term or in-hospital death, has not been tested in ICH. Therefore, the aim of this study was to determine whether a patient’s prehemorrhage medical comorbidities, as assessed by the CCI, affect ICH short-term outcomes. The results of this study may be helpful to obtain a better understanding of ICH in the elderly age group and to assist in the choice of best clinical decision making.

A total of 248 patients were analyzed in the study. Table 1 shows the frequency of CCI categories in our ICH cohort. All our samples are older than 70 years old, of which 28.2% are over 80 years old. The most frequent comorbidities were pulmonary disease found in 80.6% of patients, followed by cerebrovascular disease (62.5%) and diabetes (15.7%, without complications in 13.7%). Other frequent conditions were congestive heart failure (6.5%), ulcer disease (5.6%), and hemiplegia (5.6%).

Figure 1 gives the distribution of CCI scores. The mean CCI score was 4.41 (standard deviation 1.54), with a median of 4 (25th to 75th percentile: 3-5). The range was 2-12.

Overall, 17 patients (8%) died during hospitalization, and 127 patients (51.2%) had a poor outcome after 30 days of ICH. Table 2 shows the distribution of the demographic and clinical characteristics of the study participants who were grouped into poor or favorable outcomes based on the 30-day mRS score. The poor outcome group included higher CCI scores [(4.93 ± 1.57) vs (3.92 ± 1.33), p < 0.001], lower admission GCS [(8.23 ± 3.32) vs. (13.48 ± 1.68), p < 0.001], higher ICH volume [(58.62 ± 47.20) vs. (24.32 ± 21.52), p < 0.001], and a higher proportion of IVH (72.7 vs. 34.6%, p < 0.001). SAH (41.5 vs. 22.2%, p = 0.032) and finger-like projection (64.2 vs. 24.1%, p < 0.001) rates were higher in the poor outcome group in aged lobar hemorrhage patients.

In multivariate analysis, as shown in Table 3, the CCI score (odds ratio (OR) 1.642, 95% CI 1.254 ~ 2.150, P < 0.001), admission GCS (OR 0.505, 95% CI 0.408 ~ 0.626, P < 0.001), and ICH location in infratentorial region (OR 3.492, 95% CI 1.489 ~ 5.690, P = 0.002) independently predicted 30-day mRS outcome. The CCI score (OR 1.408, 95% CI 1.147 ~ 1.910, P = 0.003), admission GCS score (OR 0.818, 95% CI 0.720 ~ 0.930, P = 0.002) and ICH location in infratentorial region (OR 5.058, 95% CI 1.200 ~ 10.305, P = 0.022) independently predicted in-hospital mortality.

In our study, 248 elderly ICH patients aged over 70 years were retrospectively enrolled, and the results showed that the CCI score was independently associated with poorer short-term functional outcome and higher in-hospital mortality after adjusting for the components of the ICH score and sex.

Comorbidities, as measured by the CCI, are widely considered one of the factors affecting the outcomes of stroke [9‐12], but few studies have focused on hemorrhagic stroke. Bar et al. found that comorbid medical conditions, as measured by the CCI, independently affect functional outcomes at 12 months after ICH [12]. However, Bar et al. investigated the age of patients over a wide range instead of a limited range. The comorbidity pattern in older ICH patients can be very different from that in younger patients, both in terms of the number and type of conditions. Compared with previous studies, we found that the incidence of pulmonary disease (80.6% vs. 11.1%) and cerebrovascular disease (62.5% vs. 23.5%) was particularly high in the elderly population. In addition, pulmonary [18‐20] and cerebrovascular [21, 22] diseases have been proven to play an important role in the prognosis of ICH patients. A previous study focused on functional outcome after 1 year of ICH [12], and our study demonstrates that CCI, as a sum score weighted according to the presence of various comorbidities, also has an impact on the short-term prognosis of elderly ICH patients. Therefore, we consider that more intensive care and medical resources might be needed to improve the prognosis of elderly patients with higher CCI scores.

At the same time, our study found that neither IVH nor hematoma volume was an independent factor affecting the prognosis of elderly ICH patients. However, as components of the ICH score, ICH volume and IVH have been demonstrated to be independent predictors of prognosis in ICH patients [23, 24]. One possible explanation is that brain atrophy is more common in elderly individuals, which leaves more room for compensation in the brain (Fig. 2). Neither hematoma volume nor IVH has a significant impact on intracranial pressure in elderly patients and thus has a small impact on patient prognosis. At the same time, as a result of “immunosenescence”, elderly patients will experience an overall decline in the protective immune response [25], and secondary brain damage from hematoma may also be reduced. In addition, lobar intracerebral hemorrhage in elderly patients is most often caused by cerebral amyloid angiopathy (CAA) [26]. In our study, the proportion of patients with lobar hemorrhage was 43.1%. According to the Edinburgh criteria, 22.4% of them were high-risk CAA-related ICH patients, and 30.8% were moderate risk [27]. CAA is a common small vascular disease of the brain caused by progressive deposition of amyloid beta protein in the pia meningeal and cortical arterial walls and cortical capillaries. Patients with CAA-related ICH have a lower mortality rate than those with cerebral hemorrhage from other causes [28]. In previous clinical studies on the prognosis of ICH, CAA-related ICH has not been distinguished from hypertension-related intracerebral hemorrhage. As CAA-related ICH accounts for a high proportion in elderly patients, the prognostic factors traditionally associated with ICH may not be applicable to our elderly CAA-related ICH.

Our study has several limitations. First, as a retrospective study, a potential source of selection bias lies within the patient population itself. Second, several CCI factors, such as peripheral vascular disease and ulcer disease, may have been omitted, as they may not be routinely collected in the emergent setting and are difficult to assess in the routine daily medical exam. In addition, it is significant that subconscious bias towards elderly sicker patients may lead to less aggressive care. For ICH patients with high CCI scores, families may have lower treatment expectations and physicians may be more conservative in their treatment.

Overall, for elderly patients with ICH, the CCI score, reflecting the presence of various comorbidities, might be an independent factor related to the patients’ short-term outcomes in terms of in-hospital mortality and 30-day prognosis, while the characteristics of the hematoma itself, such as ICH volume and presence of IVH, seem to have a reduced effect on patient short-term outcomes.