Study design and setting
This is a retrospective, observational study based on consecutively electronic health records from West China Hospital Sichuan University between December 2010 and July 2019 and the First of People’s Hospital of Longquanyi District between January 2017 and October 2020. This study was approved by both West China Hospital and the First of People’s Hospital of Longquanyi District Institutional Review Board. All authors had no access to information that could identify individual participants during or after data collection.
Participants
Patients through the following were included for further analysis: (1) radiologically confirmed case of acute primary (non-traumatic) ICH, (2) no history of previous ICH, (3) follow-up data available at 90 days.
Patients were excluded if they met the critical exclusion criteria: (1) traumatic ICH, (2) acute ischemic stroke, (3) subarachnoid hemorrhage, (4) subdural hemorrhage, (5) first computed tomography (CT) digital imaging and communication not available in electronic medical records, (6) unavailable hemoglobin concentration at admission.
Patients younger than 50 were defined as young adult patients, and older than 50 were defined as elderly patients.
Hemoglobin on admission
We used hemoglobin on admission as the exposure factor in this study. As per the guideline of WHO12, the diagnosis of anemia in men was based on a hemoglobin of less than 130 g/L; in women, it was less than 120 g/L. Multiple guidelines recommend the hemoglobin threshold for red blood cell transfusion is 70–80 g/L for adults. Thus, we defined patients with hemoglobin concentration < 80 g/L as severe anemia and 80–120/130 g/L as mild to moderate anemia. We also defined patients with hemoglobin concentration > 160 g/L as high hemoglobin.
Outcomes
The primary outcome was mortality at 90 days. Secondary outcomes were mortality at 30 days, 1 year, and the longest follow-up.
We retrieved the whole death records up to August 15, 2021, for further analysis via the Household Registration Administration System from the Department of Public Security, Sichuan Province, China. China’s law demands that once a citizen dies, the head of household, relatives, dependents, or neighbors shall report the death registration to the household registration authority and cancel the household registration within 1 month. This system has accurate death records. Therefore, it was negligible for the rate of loss to follow up this study.
Statistical analysis
Continuous variables were presented as mean ± standard deviation (SD), and categorical variables were reported with frequencies and percentages to describe the baseline characteristics of enrolled patients. The missing values have been replaced by mean values.
Univariable logistic regression analysis was performed to assess the association of baseline characteristics with favorable outcomes. Confounders included in the multivariable regression models were derived from previous relative studies and clinical expertise, including age, gender, smoking, alcohol abuse, hypertension, diabetes, size of the hematoma, infratentorial hematoma, intraventricular hematoma, Glasgow Coma Scale (GCS). We used restricted cubic spline functions to analyze the log-linearity assumption for continuous characteristics. Then, factors associated with a value of P < 0.10 in univariate analyses were implemented into the multivariable logistic regression model. We also performed subgroup comparisons between patients with low hemoglobin concentration and normal hemoglobin concentration at admission by age stratification.
Kaplan–Meier analyses and log-rank tests were used to examine all-cause mortality after ICH amongst different subgroups of hemoglobin concentration. Cox’s proportional hazards model was used to adjust for major potential confounders and estimate hazard ratios (HR) for survival.
P values less than 0.05 were set as the threshold for statistical significance, and all P values were 2-sided. R statistical software (version 4.1.3; Foundation for Statistical Computing) was used to perform all statistical analyses.
Ethical statement
This is a retrospective study based on medical records and has been granted an exemption from requiring written informed consent by the ethics committee of West China Hospital (No. 20211701) and the ethics committee of the First of People’s Hospital of Longquanyi District (No. 2022004). This study complied with the guidelines for human studies and was conducted ethically in accordance with the World Medical Association Declaration of Helsinki.