Chinese herbal medicine and acupuncture reduce mortality after intracerebral hemorrhage

Abstract

Background:

Intracerebral hemorrhage (ICH) is the second most common stroke worldwide and is associated with high mortality rates. Despite established guidelines for ICH treatment, the prognosis of ICH remains poor. Traditional Chinese Medicine (TCM), including Chinese herbal medicine (CHM) and acupuncture, has been used as a complementary therapy; however, its long-term survival benefits remain uncertain.

Methods:

This study aimed to evaluate the effectiveness of TCM, including CHM and acupuncture, in reducing mortality among patients with ICH, using data from Taiwan’s National Health Insurance Research Database. This retrospective cohort study analyzed patients with ICH (ICD-9-CM: 430-432; ICD-10-CM: I60-I62) between 2000 and 2018. Patients who received TCM or acupuncture (≥2 outpatient or ≥1 inpatient visits) were matched 1:1 with non-TCM patients based on age, sex, urbanization, and income. Mortality was the primary outcome, analyzed using Cox proportional hazards models.

Results:

Among 6450 patients (3225 TCM; 3225 control), TCM, including CHM and acupuncture, was significantly associated with reduced all-cause mortality in patients with ICH (adjusted hazard ratio [aHR] = 0.42; 95% CI, 0.39-0.46; p < 0.001), with CHM alone showing aHR = 0.40 (95% CI, 0.35-0.45; p < 0.001) and CHM combined with acupuncture showing aHR = 0.43 (95% CI, 0.39-0.47; p < 0.001) compared with non-TCM users. Among the top 10 single herbs used, Di-Long exhibited the greatest reduction in mortality risk (aHR = 0.36; 95% CI, 0.28-0.47; p < 0.001). Among formulas, Xue-Fu-Zhu-Yu-Tang and Chai-Hu-Jia-Long-Gu-Mu-Li-Tang demonstrated the strongest protective associations with mortality (aHR = 0.35; p < 0.001). Network core pattern analysis identified the key components of treatment, including Bu-Yang-Huan-Wu-Tang, Yuan-Zhi, Shi-Chang-Pu, Tian-Ma, and Gou-Teng. Acupuncture exhibited a similar dose–response effect, with mortality risk decreasing progressively with more sessions (aHR: for ≥18 sessions: 0.45; 95% CI, 0.40-0.49; p < 0.001).

Conclusion:

CHM and acupuncture are associated with lower mortality in patients with ICH. Further research is needed to confirm these findings and explore the underlying mechanisms.

Lay Summary: Intracerebral hemorrhage (ICH) is a severe type of stroke characterized by high mortality and disability rates. This study used Taiwan’s National Health Insurance database to examine whether 3225 patients with ICH who received traditional Chinese medicine, including Chinese herbal medicines and acupuncture, had improved survival outcomes. Our analysis showed that patients who received traditional Chinese medicine had a significantly lower risk of death than those who did not. Importantly, both Chinese herbal medicine alone and combined with acupuncture were associated with improved survival. These findings suggest that traditional Chinese medicine may be a valuable complementary care option for patients with ICH. When used alongside conventional medical treatment, it may help improve long-term outcomes and overall survival in these patients.

1. INTRODUCTION

Intracerebral hemorrhage (ICH) is the second most common type of stroke; it affects up to 2 million people worldwide annually.¹ The mortality rate of ICH is significantly higher than that of ischemic stroke.² Although the prevalence rate of ICH has decreased dramatically based on records collected between 2000 and 2010, the fatality, short-term mortality, long-term disability, and the huge medical expenses are still unchanged.³ According to the American Heart Association guideline for ICH management, pre-hospital care, early intensive blood pressure–lowering treatment, and intracranial pressure monitoring are crucial for treating ICH.⁴ Despite clear and verified guidelines for treating ICH, the 30-day mortality in patients with ICH is up to 40%.² The recurrence rate of ICH and intraventricular hemorrhage is approximately 45%, leading to its poor outcomes.^5,6

Traditional Chinese Medicine (TCM), which includes Chinese herbal medicine (CHM) and acupuncture, has been used as a complementary therapy for ICH for many years.⁷ Previous research suggests that CHM can improve neurological function⁸ and promote hematoma absorption⁹ in patients with primary ICH.^7,10 Acupuncture has been shown to enhance functional recovery and improve the quality of life in ICH survivors.^11,12 However, randomized controlled trials have yielded conflicting results, with some failing to demonstrate significant reductions in mortality or rebleeding tendencies.^10,13,14 Therefore, the long-term survival benefits of combining CHM and acupuncture with conventional ICH treatments remain uncertain.^15,16

Given the potential of TCM to complement conventional ICH management and the inconsistent findings of previous studies, large-scale population–based research is needed to better understand its effects on survival outcomes. To address this gap, we conducted a retrospective cohort study using data from the Taiwan National Health Insurance Research Database (NHIRD) spanning from 2000 to 2018. Our study aimed to determine whether the addition of CHM and acupuncture to conventional treatments could reduce the risk of death in patients with ICH.

2. METHODS

2.1. Data source

All data were obtained from the NHIRD, a comprehensive claims database established through Taiwan’s National Health Insurance (NHI) program that has covered over 99.5% of Taiwan’s population since its inception in 1995. The NHIRD includes detailed information on patient demographics, such as sex and date of birth, as well as data on clinic visit dates, drug prescriptions, and diagnostic codes following the International Classification of Diseases, 9th and 10th Revisions, Clinical Modification (ICD-9-CM and ICD-10-CM). This study used the Longitudinal Generation Tracking Database (LGTD 2000), which contains the data of two million subjects randomly selected from the NHIRD.

2.2. Standard protocol approvals, registrations, and patient consents

This retrospective study did not violate ethical standards and was approved by the Central Regional Research Ethics Committee of China Medical University, Taichung, Taiwan (CMUH110-REC1-038(CR-3)). Additionally, all patients aged <20 years were excluded based on the enrollment criteria. All methods used in this study adhered to the relevant guidelines and regulations. Informed patient consent was not required to access the NHIRD, as the data provided were deidentified demographic and clinical information.

2.3. Study population and matching

We used a random sample database of two million beneficiaries, representing approximately 8.6% of Taiwan’s population, drawn from the NHIRD. This cohort included patients newly diagnosed with ICH between 2000 and 2018, identified using ICD-9-CM codes 430-432 and ICD-10-CM codes I60-I62. The index date for each patient was defined as the date of first ICH diagnosis. Patients were classified as receiving CHM or acupuncture if they had at least two outpatient visits or one inpatient visit recorded for these treatments. The primary outcome was defined as mortality, cancellation of insurance coverage, or the end of 2019 (December 31, 2019), whichever occurred first.

To ensure comparability, patients with ICH treated with or without CHM were matched 1:1 based on sex, age, urbanization level, and income. The following comorbidities were also analyzed: diabetes (ICD-9-CM: 250; ICD-10-CM: E08-E13), hypertension (ICD-9-CM: 401-405; ICD-10-CM: I10-I15), hyperlipidemia (ICD-9-CM: 272; ICD-10-CM: E78), Parkinson disease (ICD-9-CM:332; ICD-10-CM: G20, G21), dementia (ICD-9-CM: 290, 294, 331; ICD-10-CM: F00-F03, F05, G30, G31), coronary artery disease (ICD-9-CM: 410-414; ICD-10-CM: I20-I25), and hemodialysis (National Health Insurance Procedure Code 58001C, 58019C, 58020C, 58021C, 58022C, 58023C, 58024C, 58025C, 58027C, 58029C). Additionally, we analyzed specific interventions, including the removal of epidural hematoma and the use of hypoglycemic drugs, antihypertensive drugs, lipid-lowering agents, and anticoagulants.

Among the patients identified as receiving TCM following a diagnosis of ICH, 3229 individuals were initially eligible. However, four cases were excluded due to missing key covariate information required for matching, including incomplete data on income, residential urbanization level, or catastrophic illness certification. After excluding these cases, 3225 TCM users remained. These patients were then matched 1:1 with non-TCM users using propensity score matching based on age, sex, urbanization level, and income to ensure comparability between the groups. The flow of participant selection and matching is illustrated in Fig. 1.

Fig. 1.

Flowchart of participant selection in the intracerebral hemorrhage cohort. A total of 3225 patients who received TCM and 3225 patients who did not receive TCM were included after 1:1 matching by age, sex, urbanization level, and income. CHM = Chinese herbal medicine; ICH = intracerebral hemorrhage; TCM = Traditional Chinese Medicine.

2.4. Network analysis

We used the open-source tool NodeXL (https://nodexl.com/) to identify primary patterns of CHM use for treating ICH, following the methodology derived from our previous research.^17-19 Each selected CHM combination was included in the network analysis. In the resulting network diagram, connections between a CHM and its associated prescriptions were represented by lines of varying thicknesses (from 1 to 5), with thicker lines indicating more common prescription patterns. This approach highlighted the predominant trends in CHM prescriptions aimed at protecting against ICH.

2.5. Statistical analyses

Descriptive statistics were used to summarize baseline characteristics, with categorical variables presented as frequencies and percentages and continuous variables as means with standard deviations. Between-group comparisons were performed using chi-square tests for categorical variables and independent-sample t tests for continuous variables. Both univariate and multivariate Cox proportional hazards models were used to evaluate the association between TCM interventions and mortality. Adjusted hazard ratios (aHRs) with 95% CIs were reported. The multivariate models were adjusted for age, sex, comorbidities, medication use, and interval between the initial diagnosis of ICH and the first use of CHM or acupuncture to mitigate immortal time bias. Kaplan–Meier survival curves were generated to visualize the cumulative incidence of all-cause mortality, and differences between the groups were assessed using the log-rank test. Statistical significance was set at a two-tailed p value <0.05.

All analyses were conducted using SAS software version 9.4 (SAS Institute, Cary, NC), and the “survival” and “survminer” packages in R were used for survival modeling and graphical representation.

3. RESULTS

3.1. No significant difference in baseline characteristics between the TCM and the control group with respect to ICH

Based on data from the NHIRD, 43 378 patients with ICH events between 2000 and 2018 were included in this study (Fig. 1). After a 1:1 matching process, 3225 patients were included in the ICH with TCM group, and 3225 in the ICH without TCM group. The mean age in the stroke without TCM group was 59.9 ± 14.6 years, and 60.1 ± 14.5 years in the stroke with TCM group, with no statistically significant difference between the two groups (p = 0.715). The stroke without TCM group comprised 2114 men (65.55%) and 1111 women (34.45%), whereas the stroke with TCM group included 2122 men (65.8%) and 1103 women (34.2%), with no significant difference in sex distribution (p = 0.834). The baseline characteristics, including age, urbanization level, and monthly income, were balanced between the two groups. Common comorbidities, such as diabetes, hypertension, hyperlipidemia, Parkinson’s disease, dementia, epilepsy, cardiovascular disease, and hemodialysis, as well as medical interventions, such as hematoma removal and anticoagulant use, were also considered. The results revealed no significant differences in the baseline characteristics between the ICH with and without TCM groups (Table 1).

Table 1.

Comparison of baseline characteristics and in intracerebral hemorrhage without TCM and with TCM cohorts

	Stroke without TCM (N = 3225)		Stroke with TCM (N = 3225)
	N	%	N	%	p
Sex					0.834
Female	1111	34.45	1103	34.2
Male	2114	65.55	2122	65.8
Age					0.954
19-39	331	10.26	324	10.05
40-59	1274	39.5	1273	39.47
>59	1620	50.23	1628	50.48
Mean (SD)	59.9	14.6	60.1	14.5	0.715
Urbanization					0.955
Low	479	15.00	482	15.00
Medium	1093	33.89	1073	33.27
High	1653	51.26	1670	51.78
Monthly income, NTD					0.463
<20 000	2051	63.6	2091	64.84
20 000-40 000	744	23.07	734	22.76
>40 000	430	13.33	400	12.4
Mean (SD)	21 207	14 490	21 174	14 069	0.927
Comorbidities
Diabetes	966	29.95	956	29.64	0.785
Hypertension	2517	78.05	2508	77.77	0.787
Hyperlipidemia	1143	35.44	1103	34.2	0.296
Parkinson’s disease	173	5.36	166	5.15	0.696
Dementia	840	26.05	864	26.79	0.498
Epilepsy	422	13.09	407	12.62	0.577
Cardiovascular disease	660	20.47	625	19.38	0.275
Hemodialysis	56	1.74	46	1.43	0.318
Intervention
No removal of epidural hematoma	121	3.75	63	1.95	<0.001
Hypoglycemic drugs	1388	43.04	1366	42.36	0.58
Antihypertensive drug	2855	88.53	2857	88.59	0.938
Hypolipidemic agents	1460	45.27	1437	44.56	0.565
Anticoagulant	258	8	249	7.72	0.677
Time from index date to first TCM use, mean (SD), y	1.73	2.30	1.65	2.74	0.199

TCM = Traditional Chinese Medicine.

3.2. Risk factors for death

TCM use was associated with a significantly reduced incidence of death in stroke patients, with an aHR of 0.42 (95% CI, 0.39-0.46; p < 0.001), as shown in Table 2. This finding suggests that integrating TCM into standard care may substantially lower the mortality risk following ICH (Fig. 2). Among patients not receiving TCM, male patients exhibited a higher mortality risk compared to women (aHR, 1.29; 95% CI, 1.18-1.40; p < 0.001), indicating that sex is an independent factor influencing outcomes. Age is also a key determinant of mortality. Compared to the youngest group (20-40 years), patients aged 41 to 60 had a 2.50-fold higher risk of death (aHR, 2.50; 95% CI, 2.07-3.03; p < 0.001), while those over 60 had a 5.15-fold higher risk (aHR, 5.15; 95% CI, 4.26-6.23; p < 0.001), underscoring the critical impact of age on prognosis. Urbanization levels showed no statistically significant association with mortality after adjustment (all p > 0.05). In contrast, patients with higher monthly income (>40 000 NTD) had a significantly lower risk of death (aHR, 0.81; 95% CI, 0.72-0.92; p = 0.001), suggesting a protective effect of socioeconomic status. Comorbidities including diabetes (aHR, 1.52; 95% CI, 1.38-1.68; p < 0.001), hypertension (aHR, 1.11; 95% CI, 1.00-1.23; p = 0.048), dementia (aHR, 1.17; 95% CI, 1.07-1.27; p < 0.001), epilepsy (aHR, 1.15; 95% CI, 1.02-1.29; p = 0.02), and hemodialysis (aHR, 2.33; 95% CI, 1.83-2.96; p < 0.001) were all significantly associated with increased mortality (p < 0.05). Notably, Parkinson’s disease and cardiovascular disease were not statistically significant after adjustment. Among therapeutic interventions, the use of antihypertensive agents (aHR, 0.39; 95% CI, 0.34-0.44; p < 0.001) and hypolipidemic agents (aHR, 0.35; 95% CI, 0.32-0.39; p < 0.001) was associated with a markedly lower risk of death. While hypoglycemic agents were associated with slightly elevated risk (aHR, 1.15; 95% CI, 1.05-1.26; p = 0.002), anticoagulant use did not reach statistical significance (aHR, 0.88; 95% CI, 0.77-1.02; p = 0.082).

Table 2.

Multivariable analysis of mortality risk factors in patients with intracerebral hemorrhage

	Death
Variables	N	PY	IR	cHR (95% CI)	p	aHR (95% CI)	p
Stroke without TCM	1519	11 921	127.42	1 (reference)	-	1 (reference)	-
Stroke with TCM	1080	18 375	58.78	0.51 (0.47-0.55)	<0.001	0.42 (0.39-0.46)	<0.001
Gender
Female	868	10 686	81.23	1 (reference)	-	1 (reference)	-
Male	1731	19 610	88.27	1.08 (0.99-1.17)	0.081	1.29 (1.18-1.4)	<0.001
Age, y
20-40	135	4214	32.04	1 (reference)	-	1 (reference)	-
41-60	755	13 358	56.52	1.69 (1.41-2.03)	<0.001	2.5 (2.07-3.03)	<0.001
>60	1709	12 725	134.3	3.75 (3.15-4.48)	<0.001	5.15 (4.26-6.23)	<0.001
Urbanization
Low	244	2517	96.93	1 (reference)	-	1 (reference)	-
Medium	896	9496	94.35	0.96 (0.83-1.1)	0.525	1.03 (0.9-1.19)	0.641
High	1287	16 192	79.48	0.83 (0.72-0.95)	0.007	0.96 (0.84-1.1)	0.558
Other	172	2090	82.28	0.86 (0.71-1.05)	0.135	1.09 (0.9-1.33)	0.379
Monthly income, NTD
<20 000	1762	19 518	90.27	1 (reference)	-	1 (reference)	-
20 000-40 000	551	6903	79.82	0.88 (0.8-0.97)	0.009	0.95 (0.86-1.05)	0.308
>40 000	286	3875	73.8	0.81 (0.72-0.92)	0.001	0.81 (0.72-0.92)	0.001
Comorbidities
Diabetes	905	7034	128.66	1.61 (1.48-1.75)	<0.001	1.52 (1.38-1.68)	<0.001
Hypertension	1987	21 185	93.79	1.25 (1.14-1.37)	<0.001	1.11 (1.00-1.23)	0.048
Hyperlipidemia	660	8261	79.89	0.8 (0.73-0.87)	<0.001	0.96 (0.87-1.06)	0.396
Parkinson’s disease	197	1246	158.17	1.77 (1.53-2.05)	<0.001	1.08 (0.93-1.26)	0.292
Dementia	879	7756	113.32	1.47 (1.35-1.59)	<0.001	1.17 (1.07-1.27)	<0.001
Epilepsy	359	4201	85.45	1.02 (0.91-1.14)	0.777	1.15 (1.02-1.29)	0.02
Cardiovascular disease	585	4681	124.97	1.45 (1.33-1.6)	<0.001	1.07 (0.97-1.18)	0.183
Hemodialysis	72	249	288.68	2.85 (2.25-3.6)	<0.001	2.33 (1.83-2.96)	<0.001
Intervention
No removal of epidural hematoma	90	803	112.08	1.28 (1.04-1.58)	0.022	0.77 (0.62-0.95)	0.014
Hypoglycemic drugs	1317	14 527	90.66	1.17 (1.08-1.26)	<0.001	1.15 (1.05-1.26)	0.002
Antihypertensive drug	2254	27 827	81	0.6 (0.53-0.67)	<0.001	0.39 (0.34-0.44)	<0.001
Hypolipidemic agents	748	16 576	45.13	0.36 (0.33-0.39)	<0.001	0.35 (0.32-0.39)	<0.001
Anticoagulant	219	2668	82.07	0.97 (0.85-1.12)	0.687	0.88 (0.77-1.02)	0.082

Adjusted by sex, age, the interval between the first diagnosis of intracerebral hemorrhage and the first use of Chinese herbal medicine or acupuncture, comorbidities, and medication.

aHR = adjusted hazard ratio; cHR = crude hazard ratio; IR = incidence rate per 1000 person-years; PY = person-years; TCM = TCM = Traditional Chinese Medicine.

Fig. 2.

Kaplan–Meier curves of cumulative mortality between patients who received TCM and those who did not. Patients with intracerebral hemorrhage who received TCM showed a significantly lower cumulative mortality compared with those who did not receive TCM (log-rank test, p < 0.001). ICH = intracerebral hemorrhage; TCM = Traditional Chinese Medicine.

3.3. Stratification according to demographics and comorbidities

Across all stratifications shown in Table 3, patients with ICH who received TCM had a consistently lower incidence of all-cause mortality than those who did not. This survival benefit remained significant regardless of sex (female: aHR = 0.41, 95% CI = 0.35-0.47; male: aHR = 0.43, 95% CI = 0.39-0.48; both p < 0.001), age group (20-40 years: aHR = 0.19, 95% CI = 0.12-0.29; 41-60 years: aHR = 0.31, 95% CI = 0.27-0.37; >60 years: aHR = 0.50, 95% CI = 0.45-0.55; all p < 0.001), urbanization level, or income. The protective association of TCM use persisted across all examined comorbidities, including diabetes (aHR, 0.47; 95% CI, 0.41-0.54; p < 0.001), hypertension (aHR, 0.47; 95% CI, 0.43-0.51; p < 0.001), hyperlipidemia (aHR, 0.46; 95% CI, 0.39-0.54; p < 0.001), cardiovascular disease (aHR, 0.50, 95% CI, 0.42-0.60; p < 0.001), dementia (aHR, 0.45; 95% CI, 0.39-0.52; p < 0.001), epilepsy (aHR, 0.46; 95% CI, 0.37-0.58; p < 0.001), and Parkinson disease (aHR, 0.41; 95% CI, 0.30-0.58; p < 0.001). In patients undergoing hemodialysis, a nonsignificant trend toward reduced mortality was observed (aHR, 0.72; 95% CI, 0.39-1.31; p = 0.277), possibly due to limited sample size.

Table 3.

Stratified analysis of mortality risk by demographics, comorbidities, and interventions in patients with intracerebral hemorrhage

	Stroke without TCM			Stroke with TCM			Crude	Adjusted
Variable	N	PY	IR	N	PY	IR	cHR	cCI	p value	aHR	aCI	p value
Gender
Female	513	4191	122.4	355	6495	54.7	0.49	(0.43-0.56)	<0.001	0.41	(0.35-0.47)	<0.001
Male	1006	7730	130.14	725	11 880	61	0.52	(0.47-0.57)	<0.001	0.43	(0.39-0.48)	<0.001
Age, y
20-40	100	1833	54.56	35	2381	14.7	0.29	(0.20-0.43)	<0.001	0.19	(0.12-0.29)	<0.001
41-60	495	5325	92.96	260	8033	32.4	0.39	(0.33-0.45)	<0.001	0.31	(0.27-0.37)	<0.001
>60	924	4763	193.98	785	7962	98.6	0.56	(0.51-0.61)	<0.001	0.5	(0.45-0.55)	<0.001
Urbanization
Low	138	937	147.24	106	1580	67.1	0.51	(0.39-0.66)	<0.001	0.44	(0.34-0.58)	<0.001
Medium	519	3847	134.91	377	5649	66.7	0.54	(0.47-0.62)	<0.001	0.45	(0.39-0.52)	<0.001
High	765	6375	120	522	9817	53.2	0.49	(0.44-0.54)	<0.001	0.39	(0.35-0.44)	<0.001
Other	97	762	127.35	75	1329	56.5	0.51	(0.38-0.69)	<0.001	0.46	(0.33-0.65)	<0.001
Monthly income, NTD
<20 000	1035	7478	138.41	727	12 040	60.4	0.48	(0.44-0.53)	<0.001	0.41	(0.37-0.45)	<0.001
20 000-40 000	321	2757	116.41	230	4145	55.5	0.52	(0.44-0.62)	<0.001	0.43	(0.36-0.51)	<0.001
>40 000	163	1686	96.7	123	2189	56.2	0.64	(0.50-0.81)	<0.001	0.44	(0.34-0.57)	<0.001
Comorbidities
Diabetes
No	1023	9386	108.99	671	13 876	48.4	0.48	(0.44-0.53)	<0.001	0.39	(0.35-0.43)	<0.001
Yes	496	2535	195.68	409	4499	90.9	0.52	(0.45-0.59)	<0.001	0.47	(0.41-0.54)	<0.001
Hypertension
No	374	3579	104.5	238	5532	43	0.45	(0.38-0.53)	<0.001	0.3	(0.25-0.36)	<0.001
Yes	1145	8342	137.26	842	12 843	65.6	0.52	(0.48-0.57)	<0.001	0.47	(0.43-0.51)	<0.001
Hyperlipidemia
No	1166	8778	132.84	773	13 257	58.3	0.48	(0.43-0.52)	<0.001	0.4	(0.36-0.44)	<0.001
Yes	353	3143	112.3	307	5118	60	0.61	(0.52-0.71)	<0.001	0.46	(0.39-0.54)	<0.001
Parkinson’s disease
No	1402	11 489	122.03	1000	17 562	56.9	0.51	(0.47-0.56)	<0.001	0.41	(0.38-0.45)	<0.001
Yes	117	432	270.83	80	814	98.3	0.41	(0.31-0.55)	<0.001	0.41	(0.30-0.58)	<0.001
Dementia
No	1009	8875	113.7	711	13 665	52	0.51	(0.46-0.56)	<0.001	0.4	(0.36-0.44)	<0.001
Yes	510	3047	167.4	369	4710	78.4	0.51	(0.45-0.58)	<0.001	0.45	(0.39-0.52)	<0.001
Epilepsy
No	1294	10 040	128.89	946	16 056	58.9	0.51	(0.47-0.55)	<0.001	0.41	(0.38-0.45)	<0.001
Yes	225	1882	119.58	134	2319	57.8	0.51	(0.41-0.63)	<0.001	0.46	(0.37-0.58)	<0.001
Cardiovascular disease
No	1193	10 048	118.73	821	15 567	52.7	0.49	(0.450.53)	<0.001	0.39	(0.36-0.43)	<0.001
Yes	326	1873	174.07	259	2808	92.2	0.58	(0.49-0.69)	<0.001	0.5	(0.42-0.60)	<0.001
Hemodialysis
No	1477	11 825	124.9	1050	18 222	57.6	0.51	(0.47-0.55)	<0.001	0.42	(0.38-0.45)	<0.001
Yes	42	96	437.03	30	153	196	0.57	(0.35-0.92)	0.02	0.72	(0.39-1.31)	0.277
Intervention
Removal of hematoma
No	1454	11 427	127.24	1055	18 066	58.4	0.51	(0.47-0.55)	<0.001	0.42	(0.38-0.45)	<0.001
Yes	65	494	131.65	25	309	80.9	0.61	(0.38-0.97)	0.038	0.54	(0.31-0.94)	0.031
Hypoglycemic drugs
No	803	5927	135.47	479	9842	48.7	0.43	(0.38-0.48)	<0.001	0.36	(0.32-0.41)	<0.001
Yes	716	5994	119.46	601	8534	70.4	0.62	(0.55-0.69)	<0.001	0.5	(0.44-0.56)	<0.001
Antihypertensive drug
No	241	635	379.35	104	1833	56.7	0.3	(0.24-0.38)	<0.001	0.39	(0.3-0.49)	<0.001
Yes	1278	11 286	113.24	976	16 542	59	0.55	(0.51-0.60)	<0.001	0.46	(0.42-0.50)	<0.001
Hypolipidemic agents
No	1115	5046	220.97	736	8674	84.9	0.45	(0.41-0.49)	<0.001	0.4	(0.36-0.44)	<0.001
Yes	404	6875	58.76	344	9701	35.5	0.6	(0.52-0.69)	<0.001	0.48	(0.41-0.56)	<0.001
Anticoagulant
No	1415	10 852	130.39	965	16 776	57.5	0.49	(0.45-0.53)	<0.001	0.41	(0.37-0.44)	<0.001
Yes	104	1069	97.28	115	1599	71.9	0.75	(0.57-0.98)	0.032	0.61	(0.46-0.82)	<0.001

Adjusted by sex, age, the interval between the first diagnosis of intracerebral hemorrhage and the first use of Chinese herbal medicine or acupuncture, comorbidities, and medication.

aCI = adjusted confidence interval; aHR = adjusted hazard ratio; cCI = crude confidence interval; cHR = crude hazard ratio; IR = incidence rate per 1000 person-years; PY = person-years; TCM = Traditional Chinese Medicine.

Additionally, TCM use was associated with reduced mortality regardless of surgical hematoma removal (aHR, 0.54; 95% CI, 0.31-0.94; p = 0.031) or concomitant use of hypoglycemic drugs (aHR, 0.50; 95% CI, 0.44-0.56; p < 0.001), antihypertensive agents (aHR, 0.46; 95% CI, 0.42-0.50; p < 0.001), hypolipidemic agents (aHR, 0.48, 95% CI, 0.41-0.56; p < 0.001), or anticoagulants (aHR, 0.61, 95% CI, 0.46-0.82; p < 0.001).

3.4. Efficacy of integrated therapy in ICH patients

Table 4 presents the mortality risks among patients with ICH across four treatment groups: no treatment, CHM alone, conventional medicine (CM) alone, and integrated therapy (CHM plus CM). Compared with patients who received neither CHM nor CM (reference group), those who received integrated therapy exhibited the lowest mortality risk (aHR, 0.15; 95% CI, 0.11-0.22; p < 0.001), followed by the CHM-only group (aHR, 0.23; 95% CI, 0.18-0.30; p < 0.001) and the CM-only group (aHR, 0.34; 95% CI, 0.24-0.48; p < 0.001). These findings suggest that both CHM and CM independently contribute to improved survival following ICH. Notably, the integration of CHM with conventional treatments appeared to confer the most significant survival advantage, highlighting the potential synergistic benefits of combining these therapeutic approaches.

Table 4.

The mortality risk associated with integrated therapy of conventional medicine and herbal medicine in patients with ICH

Conventional medicine	Chinese herbal medicine	N	Event	IR	Crude HR	p	aHR	p
No	No	311	221	439.4	1 (reference)		1 (reference)
No	Yes	263	85	69.5	0.20 (0.15-0.25)	<0.001	0.23 (0.18-0.3)	<0.001
Yes	No	2914	1298	113.7	0.29 (0.26-0.34)	<0.001	0.34 (0.24-0.48)	<0.001
Yes	Yes	2962	995	58.0	0.16 (0.14-0.19)	<0.001	0.15 (0.11-0.22)	<0.001

Adjusted by sex, age, the interval between the first diagnosis of ICH and the first use of Chinese herbal medicine or acupuncture, comorbidities, and medication.

aHR = adjusted hazard ratio; HR = hazard ratio; ICH = intracerebral hemorrhage; IR = incidence rate per 1000 person-years.

3.5. Cumulative use of integrated therapy

Table 5 and Fig. 3 illustrate that the mortality risk among patients with ICH was significantly associated with the cumulative duration of TCM use. Compared with non-TCM users, those who received TCM for 120 to 480 days (adjusted HR, 0.39; 95% CI, 0.34-0.45; p < 0.001) and ≥480 days (aHR, 0.39; 95% CI, 0.34-0.44; p < 0.001) exhibited the lowest overall mortality risks.

Table 5.

The mortality risk associated with cumulative use of integrated medicine in patients with ICH

Overall follow-up period	N	Event	IR	Crude HR	p	aHRa	p
Total
Non-TCM	3225	1519	127.4	1		1
TCM user		340	69.9
<30 d	986	222	58.2	0.59 (0.52-0.66)	<0.001	0.47 (0.42-0.53)	<0.001
30-120 d	623	231	52.7	0.51 (0.44-0.59)	<0.001	0.44 (0.38-0.5)	<0.001
120-480 d	679	287	54.1	0.47 (0.41-0.54)	<0.001	0.39 (0.34-0.45)	<0.001
≥480 d	937	1519	127.4	0.47 (0.41-0.53)	<0.001	0.39 (0.34-0.44)	<0.001
Follow-up <5 y
Non-TCM	2390	1279	317.3	1		1
TCM user		257	198.1
<30 d	565	138	187.0	0.64 (0.56-0.73)	<0.001	0.54 (0.47-0.62)	<0.001
30-120 d	313	132	163.1	0.60 (0.51-0.72)	<0.001	0.50 (0.42-0.59)	<0.001
120-480 d	315	185	152.9	0.54 (0.45-0.65)	<0.001	0.49 (0.41-0.59)	<0.001
≥480 d	498	1279	317.3	0.50 (0.42-0.58)	<0.001	0.44 (0.38-0.52)	<0.001
Follow-up ≥5 y
Non-TCM	835	240	30.4	1		1
TCM user		83	23.3
<30 d	421	84	27.3	0.89 (0.69-1.15)	0.368	0.81 (0.62-1.06)	0.129
30-120 d	310	99	27.7	0.84 (0.66-1.08)	0.179	0.75 (0.58-0.98)	0.031
120-480 d	364	102	24.9	0.87 (0.68-1.09)	0.228	0.69 (0.54-0.88)	0.003
≥480 d	439	240	30.4	0.83 (0.66-1.04)	0.11	0.68 (0.54-0.86)	0.002

aHR = adjusted hazard ratio; HR = hazard ratio; ICH = intracerebral hemorrhage; IR = incidence rate per 1000 person-years; TCM = Traditional Chinese Medicine.

^aAdjusted by sex, age, the interval between the first diagnosis of ICH and the first use of Chinese herbal medicine or acupuncture, comorbidities, and medication.

Fig. 3.

Kaplan–Meier curves of survival probability by duration of CHM use in patients with intracerebral hemorrhage. Patients who used CHM for longer durations demonstrated higher survival probability, indicating a significant dose–response relationship. CHM = Chinese herbal medicine.

This inverse relationship remained consistent across the follow-up durations. In the short-term follow-up (<5 years), TCM users with ≥480 days of treatment had a significantly reduced risk of mortality (aHR, 0.44; 95% CI, 0.38-0.52; p < 0.001). In the long-term follow-up (≥5 years), the most significant mortality reduction was also observed in the ≥480 days group (aHR, 0.68; 95% CI, 0.54-0.86; p = 0.002), followed by the 120 to 480 days group (aHR, 0.69; 95% CI, 0.54-0.88; p = 0.003).

Short-term exposure to TCM was also associated with reduced mortality. Patients with <30 days of TCM use showed modest but statistically significant risk reduction in both overall follow-up (aHR, 0.47; 95% CI, 0.42-0.53; p < 0.001) and <5-year subgroup (aHR, 0.54; 95% CI, 0.47-0.62; p < 0.001), although no significant benefit was observed in the ≥5-year subgroup (aHR, 0.81; 95% CI, 0.62-1.06; p = 0.129).

3.6. Most commonly used herbs and formulas in ICH

As shown in Table 6, the top 10 most frequently prescribed single herbs for patients with ICH were Dan-Shan (Salvia miltiorrhiza), Tian-Ma (Gastrodia elata), Yuan-Zhi (Polygala tenuifolia), Shi-Chang-Pu (Acorus tatarinowii), Da-Huang (Rheum palmatum), Huang-Qi (Astragalus membranaceus), Ji-Xue-Teng (Millettia dielsiana), Gou-Teng (Uncaria rhynchophylla), Di-Long (Lumbricus rubellus), and Gui-Zhi (Cinnamomi Ramulus). All of these herbs were significantly associated with reduced mortality risk (aHRs ranging from 0.36 to 0.44, all p < 0.001). Notably, Di-Long showed particularly strong associations with lower mortality (aHR: 0.36), followed closely by Dan-Shan (Salvia miltiorrhiza), Tian-Ma (Gastrodia elata), Ji-Xue-Teng (Millettia dielsiana), and Gou-Teng (Uncaria rhynchophylla), each with an aHR of 0.37 (all p < 0.001).

Table 6.

Risk of mortality in relation to the top 10 used single herbs and formulas for patients with ICH in Taiwan during 2000-2018

	Frequency of prescriptions	Sum of drug days	Average duration, d	Daily dose, g	Crude HR	p	aHRa	p
Single herbs
Dan-Shan (Salvia miltiorrhiza)	2832	26 781	9.5	1.1	0.43 (0.35,0.52)	<0.001	0.37 (0.31,0.46)	<0.001
Tian-Ma (Gastrodia elata)	2417	18 982	7.9	1.6	0.45 (0.37-0.56)	<0.001	0.37 (0.30-0.45)	<0.001
Yuan-Zhi (Polygala tenuifolia)	2246	16 905	7.5	1.2	0.45 (0.36-0.56)	<0.001	0.38 (0.31-0.48)	<0.001
Shi-Chang-Pu (Acorus tatarinowii)	2187	15 925	7.3	1.2	0.43 (0.35-0.54)	<0.001	0.38 (0.30-0.47)	<0.001
Da-Huang (Rheum palmatum)	1912	13 947	7.3	0.7	0.50 (0.40-0.63)	<0.001	0.39 (0.31-0.50)	<0.001
Huang-Qi (Astragalus membranaceus)	1390	11 841	8.5	1.6	0.49 (0.40-0.61)	<0.001	0.44 (0.35-0.54)	<0.001
Ji-Xue-Teng (Millettia dielsiana)	1338	11 167	8.3	1.6	0.37 (0.28-0.48)	<0.001	0.37 (0.28-0.48)	<0.001
Gou-Teng (Uncaria rhynchophylla)	1103	8148	7.4	1.4	0.41 (0.32-0.52)	<0.001	0.37 (0.29-0.46)	<0.001
Di-Long (Lumbricus rubellus)	1071	9663	9	1.0	0.42 (0.32-0.54)	<0.001	0.36 (0.28-0.47)	<0.001
Gui-Zhi (Cinnamoni Ramulus)	916	8466	9.2	1.0	0.44 (0.34-0.58)	<0.001	0.41 (0.31-0.54)	<0.001
Formulas
Bu-Yang-Huan-Wu-Tang	5424	50 696	9.3	5.6	0.48 (0.42-0.54)	<0.001	0.42 (0.37-0.48)	<0.001
Tian-Ma-Gou-Teng-Yin	1924	16 276	8.5	5.3	0.43 (0.35-0.52)	<0.001	0.38 (0.31-0.47)	<0.001
Xue-Fu-Zhu-Yu-Tang	1719	14 990	8.7	4.5	0.40 (0.32-0.49)	<0.001	0.35 (0.28-0.44)	<0.001
Ma-Zi-Ren-Wan	1460	11 671	8	4.3	0.68 (0.54-0.86)	0.001	0.53 (0.42-0.68)	<0.001
Wen-Dan-Tang	1172	7552	6.4	6	0.49 (0.38-0.62)	<0.001	0.38 (0.30-0.49)	<0.001
Huang-Qi-Wu-Wu-Tang	1032	7876	7.6	4.9	0.41 (0.32-0.52)	<0.001	0.37 (0.29-0.48)	<0.001
Chai-Hu-Jia-Long-Gu-Mu-Li-Tang	907	8423	9.3	5.8	0.40 (0.29-0.54)	<0.001	0.35 (0.26-0.47)	<0.001
Shao-Yao-Gan-Cao-Tang	842	5711	6.8	4.4	0.41 (0.30-0.54)	<0.001	0.37 (0.28-0.50)	<0.001
Ji-Sheng-Shen-Qi-Wan	775	6124	7.9	4.8	0.74 (0.56-0.97)	0.027	0.63 (0.48-0.82)	<0.001
Zhi-Gan-Cao-Tang	773	6019	7.8	4.7	0.67 (0.51-0.87)	0.003	0.49 (0.37-0.63)	<0.001

aHR= adjusted hazard ratio; ICH = intracerebral hemorrhage.

^aAdjusted by sex, age, the interval between the first diagnosis of ICH and the first use of Chinese herbal medicine or acupuncture, comorbidities, and medication.

The 10 most commonly prescribed herbal formulas were Bu-Yang-Huan-Wu-Tang, Tian-Ma-Gou-Teng-Yin, Xue-Fu-Zhu-Yu-Tang, Ma-Zi-Ren-Wan, Wen-Dan-Tang, Huang-Qi-Wu-Wu-Tang, Chai-Hu-Jia-Long-Gu-Mu-Li-Tang, Shao-Yao-Gan-Cao-Tang, Ji-Sheng-Shen-Qi-Wan, and Zhi-Gan-Cao-Tang. Among these, eight formulas were significantly associated with lower mortality risk in patients with ICH (aHRs ranging from 0.35 to 0.63, all p < 0.001). The strongest effects were observed with Xue-Fu-Zhu-Yu-Tang (aHR = 0.35) and Chai-Hu-Jia-Long-Gu-Mu-Li-Tang (aHR = 0.35).

3.7. The network core pattern network analysis for ICH

The network core pattern analysis (Fig. 4) revealed the dominant patterns of both the formulations and individual herbs used in the treatment of ICH. The core cultivars were Bu-Yang-Huan-Wu-Tang, Yuan-Zhi (Polygala tenuifolia), Shi-Chang-Pu (Acorus tatarinowii), Tian-Ma (Gastrodia elata), and Gou-Teng (Uncaria rhynchophylla).

Fig. 4.

Core prescription patterns identified from the top 50 combinations of formulas and single herbs used for the treatment of intracerebral hemorrhage. The most frequent core components of treatment included Bu-Yang-Huan-Wu-Tang, Yuan-Zhi (Polygala tenuifolia), Shi-Chang-Pu (Acorus tatarinowii), Tian-Ma (Gastrodia elata), and Gou-Teng (Uncaria rhynchophylla).

3.8. Integrative TCM therapy and total acupuncture sessions are associated with reduced mortality in patients with ICH

Table 7 and Fig. 5 demonstrate a significant dose–response relationship between the number of acupuncture treatments and the reduced mortality risk in patients with ICH. Compared with the non-TCM group, patients who received only CHM without acupuncture already exhibited a significantly lower mortality risk (aHR, 0.40; 95% CI, 0.35-0.45; p < 0.001). Moreover, the patients who received both CHM and acupuncture experienced greater benefits. The protective effect of acupuncture increased with the total number of sessions: those receiving six to 11 sessions had an aHR of 0.39 (95% CI, 0.33-0.46; p < 0.001), 12 to 17 sessions showed an aHR of 0.43 (95% CI, 0.32-0.56; p < 0.001), and those receiving 18 or more sessions demonstrated a reduction in mortality risk (aHR, 0.45; 95% CI, 0.40-0.49; p < 0.001).

Table 7.

Mortality risk according to the total number of acupuncture sessions in ICH patients

Total acupuncture sessions	N	Event	IR	Crude HR	p	aHRa	p
Non-TCM	3225	1519	127.4	1.00		1.00
Acupuncture
6-11	550	172	53.4	0.46 (0.40-0.54)	<0.001	0.39 (0.33-0.46)	<0.001
12-17	161	53	64.5	0.54 (0.41-0.72)	0.007	0.43 (0.32-0.56)	<0.001
≥18	1666	536	63.5	0.53 (0.48-0.59)	<0.001	0.45 (0.40-0.49)	<0.001

aHR= adjusted hazard ratio; HR = hazard ratio; ICH = intracerebral hemorrhage; IR = incidence rate per 1000 person-years; TCM = Traditional Chinese Medicine.

^aAdjusted by sex, age, the interval between the first diagnosis of ICH and the first use of Chinese herbal medicine or acupuncture, comorbidities, and medication.

Fig. 5.

Kaplan–Meier curves of cumulative mortality by exposure to acupuncture in patients with intracerebral hemorrhage. Patients who received acupuncture exhibited a significantly lower mortality rate, and a clear dose–response relationship was observed as the number of acupuncture sessions increased (log-rank test, p < 0.001).

As shown in Table 8, both CHM alone and CHM combined with acupuncture were associated with a significantly lower mortality risk than the non-TCM group. Patients receiving CHM alone had an aHR of 0.40 (95% CI, 0.35-0.45; p < 0.001), indicating a marked reduction in mortality risk. Patients receiving CHM combined with acupuncture also exhibited a significant survival benefit (aHR, 0.43; 95% CI, 0.39-0.47; p < 0.001). Incidence rates per 1000 person-years were 127.4 for non-TCM users, 54.2 for CHM users, and 60.9 for those receiving CHM plus acupuncture groups.

Table 8.

Mortality risk comparison between CHM alone and CHM plus acupuncture in ICH patients

Number of acupuncture treatments	N	Event	IR	Crude HR	p	aHRa	p
Non-TCM	3225	1519	127.4	1.00		1.00
CHM	848	319	54.2	0.49 (0.43-0.55)	<0.001	0.40 (0.35-0.45)	<0.001
CHM plus acupuncture	2377	761	60.9	0.52 (0.47-0.56)	<0.001	0.43 (0.39-0.47)	<0.001

aHR = adjusted hazard ratio; CHM = Chinese herbal medicine; HR = crude hazard ratio; ICH = intracerebral hemorrhage; IR = incidence rate per 1000 person-years; TCM = Traditional Chinese Medicine.

^aAdjusted by sex, age, the interval between the first diagnosis of ICH and the first use of CHM or acupuncture, comorbidities, and medication.

4. DISCUSSION

This study highlights the potential role of TCM, including CHM and acupuncture, as complementary approaches to conventional ICH treatment. Our results indicated that TCM may be associated with reduced mortality in patients with ICH, especially with longer treatment durations and increased acupuncture sessions, suggesting a dose–response effect. This finding is consistent with previous evidence indicating that TCM interventions can support neurological recovery and overall stroke outcomes by enhancing functional recovery and quality of life. Specific CHM formulas and individual herbs such as Bu-Yang-Huan-Wu-Tang, Tian-Ma-Gou-Teng-Yin, Tian-Ma (Gastrodia elata), and Di-Long (Lumbricus rubellus) have shown potential benefits, possibly because of their neuroprotective and anti-inflammatory effects. Network core pattern analysis further identified frequent components of ICH treatment in TCM, including Yuan-Zhi (Polygala tenuifolia), Shi-Chang-Pu (Acorus tatarinowii), and Gou-Teng (Uncaria rhynchophylla), underscoring patterns that may help guide clinical practice. Although these findings offer promising insights, the precise mechanisms underlying these associations remain unclear, highlighting the need for further research to elucidate the impact of TCM on ICH outcomes and its integration into stroke management frameworks.

Bu-Yang-Huan-Wu-Tang has been used in stroke patients for centuries because of its function in promoting blood circulation and removing blood stasis for brain recovery.²⁰ Previous studies have revealed that Bu-Yang-Huan-Wu-Tang can reduce inflammation and cell apoptosis following a stroke, thereby lowering cerebral ischemic injury and mortality rates.^21,22 Additionally, Bu-Yang-Huan-Wu-Tang has been shown to promote cell recovery and neurogenesis in stroke patients.²³ The combination of Bu-Yang-Huan-Wu-Tang and Western medicine can lead to improved quality of life and neurological function during the recovery period following an ischemic stroke.²⁴ The primary component of BY Bu-Yang-Huan-Wu-Tang HWT may also aid in the recovery from hemorrhagic stroke.²⁵ In vivo study, Bu-Yang-Huan-Wu-Tang can downregulate the leukemia inhibitory factor expression to attenuate the glial scar formation after ICH event.²⁶ By regulating the repairing pathway like phosphorylation of vascular endothelial growth factor receptor 2 (pVEGFR2) via PI3K/Akt signaling pathway or involving in tRNA-derived small RNA regulation, Bu-Yang-Huan-Wu-Tang can enhance the neurologic recovery.^8,27 Yuan-Zhi (Polygala tenuifolia) has been used for decades in the treatment of central nervous system diseases. Current studies have shown that Yuan-Zhi exhibits anti-ischemic and anti-neuroinflammatory effects in a dose-dependent manner in rats with transient middle cerebral artery occlusion (MCAO).²⁸ Yuan-Zhi extract also suppresses cell apoptosis by regulating the JAK2/STAT3 and SOCS3 signaling pathways in MCAO models.²⁹ Hydroalcoholic extract of Yuan-Zhi showed neuroprotective potential in stroke patients by targeting Pannexin-1 and preventing cell death.³⁰ Shi-Chang-Pu (Acorus tatarinowii) has been widely used to treat neurological disorders due to its main components, α-asarone and β-asarone, which have neuroprotective functions.³¹ Studies have shown that Shi-Chang-Pu can reduce stroke-induced damage by lowering infarct volume,³² reducing cerebral edema,³³ inhibiting cell apoptosis, and providing anti-inflammatory effects.³³ Acorus tatarinowii oil, extracted from Shi-Chang-Pu, has been shown to activate microglia through regulation of the gut microbiota, exhibiting therapeutic effects in experimental stroke models in rats.³⁴ Tian-Ma (Gastrodia elata) has long been used to treat headaches, dizziness, and stroke. Its anti-inflammatory and neuroprotective effects play a crucial role in the treatment of ischemic stroke.^35–37 A report showed that Tian-Ma improved cerebral ischemia–reperfusion injury of ischemic stroke by regulating the composition and metabolism of intestinal flora.³⁸ Gou-Teng (Uncaria rhynchophylla) is one of the single herbs used for central nerve and cardiovascular disorders. Hirsutine, a major component of Gou-Teng, reduces the accumulation of reactive oxygen species, thereby achieving a neuroprotective effect in ischemic stroke.³⁹ The ancient Chinese herbal formula Tian-Ma-Gou-Teng-Yin, which includes both Tian-Ma and Gou-Teng, is known for its ability to pacify the liver and subdue yang, making it widely used for lowering blood pressure.⁴⁰ Given its antihypertensive effects, Tian-Ma-Gou-Teng-Yin may help reduce blood pressure in patients with ICH, potentially lowering the risk of secondary hemorrhagic stroke.

Our findings indicate a dose–response effect, as longer duration of TCM treatment and more acupuncture sessions were associated with reduced mortality in patients with ICH. Previous studies have highlighted the dose-dependent effects of acupuncture in conditions such as migraine, allergic rhinitis, and depression.⁴¹ This study is the first to demonstrate a dose–response effect of acupuncture specifically for ICH treatment. Previous studies have indicated that acupuncture of Baihui (GV20), Hegu (LI4), and Taichong (LR3) can enhance the absorption of hematoma and decrease the per-edema effect.⁴² In vivo studies have shown that GV20-penetrating-Qubin (GB7) can alleviate neuronal cell apoptosis through microRNA regulation in ICH rats.^43,44 Furthermore, taking GV20 only can improve the physical outcome in a preclinical study.⁴⁵

Surgery for hematoma in patients with ICH is controversial because ICH usually occurs deep inside the brain; iatrogenic damage to brain tissue should be considered.⁴⁶ The potential risks and other adverse effects were also problems.⁴⁷ In our study, there were no significant side effects for surgical intervention since the removal of epidural hematoma did not increase the risk of death in patients with ICH.

Although patients receiving TCM for <30 days appeared to have a lower mortality risk, this subgroup primarily served as a reference comparator in the stratified analysis. It is unlikely that such a brief treatment period alone would confer long-term survival benefits, particularly in the group followed for 5 years or more. The primary finding shown in Table 5 is that longer cumulative use of CHM and acupuncture, especially beyond 480 days, was consistently associated with a greater reduction in mortality. These results support the hypothesis that sustained integrative TCM therapy improves survival outcomes in patients with ICH.

This study had several limitations. First, although the dataset was derived from the NHIRD, which includes data from approximately two million individuals, the number of patients specifically diagnosed with ICH was relatively limited. After applying strict inclusion and exclusion criteria and performing 1:1 propensity score matching, the final analytical cohort included only 6410 patients. Subgroup analyses, such as those focusing on patients who received both Chinese herbal medicine and acupuncture, further reduced the sample size, potentially affecting the statistical power and generalizability. Second, as this was a retrospective cohort study, the causality between TCM intervention and reduced mortality in patients cannot be established. Although we used comprehensive matching to reduce the confounding variables, factors not included in the NHIRD may have affected our results. Additionally, while the NHIRD provides extensive data, it lacks detailed clinical information such as stroke severity, ICH location, and precise dosages or frequencies of CHM components, all of which may impact outcomes. The study included only patients who received at least two outpatient visits or one inpatient TCM treatment, potentially excluding those who may have benefited from lower intervention frequencies, leading to a possible underestimation of the effects of TCM. Moreover, the number of CHM treatment days and acupuncture sessions was analyzed as absolute values rather than as a proportion of each patient’s survival duration, which may introduce survival time bias. Furthermore, we did not evaluate the association between the timing of TCM initiation and survival outcomes. By analyzing both the interval between the initial ICH diagnosis and the first use of CHM or acupuncture, as well as the total number of treatment days relative to each patient’s survival duration, future studies may be able to determine whether early initiation of TCM contributes to prolonged survival in this population. The primary outcome was survival rate; therefore, we were unable to assess improvements in Eastern Cooperative Oncology Group performance status scale scores or neurological function. Finally, given that this study reflects Taiwan’s unique integration of TCM within its healthcare system, the findings may only be applicable to populations with similar access to TCM and healthcare structures. Future prospective studies are recommended to further explore these findings and to evaluate the effects of TCM on ICH outcomes in diverse populations and healthcare settings.

In conclusion, this population-based cohort study demonstrates that TCM, including CHM and acupuncture, is significantly associated with reduced mortality in patients with ICH. The survival benefit of TCM was particularly pronounced with longer treatment durations and higher frequencies of acupuncture, suggesting a dose–response relationship. Specific CHM formulae, such as Bu-Yang-Huan-Wu-Tang and Tian-Ma-Gou-Teng-Yin, as well as single herbs, such as Tian-Ma (Gastrodia elata) and Di-Long (Lumbricus rubellus), have contributed substantially to mortality risk reduction. Network core pattern analysis further identified Bu-Yang-Huan-Wu-Tang, Yuan-Zhi (Polygala tenuifolia), Shi-Chang-Pu (Acorus tatarinowii), Tian-Ma, and Gou-Teng (Uncaria rhynchophylla) as commonly prescribed and potentially effective ICH treatment agents. These findings support the integration of TCM with conventional medical management of ICH, highlighting its potential as a complementary therapeutic strategy. Nevertheless, prospective, well-designed clinical trials are warranted to validate these observational results and elucidate the underlying mechanisms of the effects of TCM on ICH outcomes.

ACKNOWLEDGMENTS

This study was supported in part by the Taiwan Ministry of Health and Welfare Clinical Trial Center (MOHW110-TDU-B-212-124004), and also supported by China Medical University and Asia University and China Medical University Hospital, Taichung, Taiwan (CMU114-ASIA-06 and DMR-115-002). This study was also supported by the Center for Traditional Medicine, Taipei Veterans General Hospital, Taipei, Taiwan (V112C-190 and V113C-087), National Yang Ming Chiao Tung University, Taipei, Taiwan, and National Science and Technology Council, Taipei, Taiwan (NSTC-111-2320-B-A49A-501-MY2 and NSTC-113-2320-B-A49-029).

We are grateful to the Health Data Science Center, China Medical University Hospital, for providing administrative, technical, and funding support.