Aidi Injection, a Traditional Chinese Medicine Injection, Could Be Used as an Adjuvant Drug to Improve Quality of Life of Cancer Patients Receiving Chemotherapy: A Propensity Score Matching Analysis.

BACKGROUND: Clinical research has paid increasing attention to quality of life (QoL) in recent years, but the assessment of QoL is difficult, hampered by the subjectivity, complexity, and adherence of patients and physicians. According to previous studies, QoL in cancer patients is related to performance status (PS) and influenced by chemotherapy-related toxicity. Aidi injection, a traditional Chinese medicine injection, is used as an adjuvant drug to enhance effectiveness of chemotherapy. The study aims to investigate whether Aidi injection could improve QoL by improving PS and reducing toxicity caused by chemotherapy.
METHODS: A retrospective cohort study was performed at the First Affiliated Hospital of Anhui Medicine University. Data of consecutive patients diagnosed with cancers between January 2014 and June 2017 were retrieved from the electronic medical record system. After a 1:1 propensity score match, patients were then divided into 2 groups based on the therapies used, that is, Aidi injection combined with chemotherapy and chemotherapy alone, and the PS, chemotherapy-related toxicity, and combined medication information were compared. The effect of different dosages of Aidi injection on patients was further explored.
RESULTS: A total of 3200 patients were included in this study. Aidi injection combined with chemotherapy exhibited significantly benefit in PS ( P < .001, odds ratio [OR] 3.4, 95% confidence interval [CI] 2.4-4.8) compared with chemotherapy alone after adjusting for the factors that affect PS. The improvement rate of PS in the Aidi group was significantly higher than in the control group across the stratification of gender, age, tumor type, TNM stage, body mass index, nodal metastasis, prior chemotherapy, chemotherapy regimens, other Chinese tradition medicines, and chemotherapy cycle. Meanwhile, Aidi injection used synchronously with chemotherapeutic drugs could decrease the incident rate of damage to liver and kidney function, myelosuppression, and gastrointestinal reactions caused by chemotherapy.
CONCLUSION: It was indicated that the integrative approach combining chemotherapy with Aidi injection, especially with the conventional dosage of Aidi injection, had significant benefit on QoL in cancer patients.

Introduction

Cancer is a major public health problem and has emerged as a leading cause of death globally. Although significant progress has been made in research into the cause of cancer, effective strategies to inhibit the carcinogenic process still lag behind.[1] By 2020, the world population is expected to have increased to 7.5 billion; of this number, approximately 15 million new cancer cases will be diagnosed.[2] After being diagnosed, chemotherapy is the most frequently used way to control disease and to extend survival.[3] However, chemotherapy-induced toxicities critically cause various complications, increasing the risk of chemotherapeutic failure.[4] Based on these problems, clinical workers are actively looking for complementary and alternative therapies for enhancing effectiveness and relieving symptoms of chemotherapy. In China, there has been a long history of using traditional Chinese herbal medicine (TCM) as one kind of adjuvant medicine, combined with antitumor drugs in practice for the treatment and cure of cancers.[5] Undoubtedly, patients would seek these treatments when faced with cancers regardless of whether one is a supporter of TCM or not.

Aidi injection is an adjuvant TCM injection commonly used in treatment of lung cancer,[6] gastric cancer,[7] colorectal cancer,[8] liver cancer,[9] pancreatic cancer,[10] malignant lymphoma,[11] and esophageal cancer.[12] The ingredients of Aidi injection are extracts from Renshen (Radix Ginseng), Huangqi (Astragaloside), Ciwujia (Eleutherococcus senticosus), and Banmao (Cantharidin), all of which are important Chinese herbal medicines, with ability to inhibit cancers and improve immunity.[13-16] Previous studies had reported that Aidi injection has the effects of suppressing tumor metastasis, inducing apoptosis of cancer cells, and inhibiting overexpression of drug-resistant proteins induced by chemotherapy.[17,18] In addition, the effects of radio-sensitization and altering the expression profiles of microRNAs may be important antitumor mechanisms of Aidi injection.[19,20]

In recent years, quality of life (QoL) has been recognized as a key component in clinical trial evaluations.[21,22] Indeed, the vast majority of clinical trials now report QoL as either primary or secondary endpoints. Therapies that improve QoL in cancer patients are of increasing interest, especially in patients with advanced cancer in whom the scope for improving survival has proved to be limited. Because of the various restrictions in the assessment of QoL in the real-world medical environment, it is important to find surrogate endpoints of QoL as a preliminary assessment.[23] Previous research had demonstrated that QoL would be affected by physical function, which is a key component of QoL and is usually assessed objectively by performance status (PS).[24,25] In addition, a clinical research study had reported the QoL was related to PS and affected by chemotherapy-related toxicity, and that assessment of those indicators might lead to more effective care for patients.[26]

After analyzing the progress of research Aidi injection, we found that there is lack of research based on real-world data[27] with a large sample size to prove the effect of Aidi injection on QoL, which might be meaningful for clinical decision making. In order to supplement the evidence-based medical evidence, a population-based retrospective cohort study was conducted to investigate whether use of Aidi injection as an adjuvant to chemotherapy might improve QoL in patients with different types of cancer. PS was used as a surrogate endpoint to assess QoL, with assessment of the effect of Aidi injection on relieving chemotherapy-related toxicity as supplementary evidence. The dosage of Aidi injection is determined according to the doctor’s advice, which is as follows: The available dosage for all patients is 20 to 100 mL per day and the conventional dosage for adults is 50 to 100 mL per day; 0.9% sodium chloride injection or 5% glucose injection are used for the dilution and it is used synchronously with chemotherapeutic drugs.

Methods

Population

A retrospective cohort study was conducted at the First Affiliated Hospital of Anhui Medical University. Data of patients diagnosed with different types of cancer between January 2014 and June 2017, retrieved from the electronic medical system at the hospital, determined the sample size of this study. Each patient’s information on chemotherapy during the period of study and the data of outpatient reexamination after discharge until January 2018 was followed up. The following inclusion and exclusion criteria were adopted for the study. Inclusion criteria: (1) aged 18 years and older, (2) having complete basic information, and (3) receiving at least 1 cycle of chemotherapy. Exclusion criteria: (1) having second primary malignancy, (2) combining multiple therapies or changing chemotherapy methods, (3) missing data, and (4) using Aidi injections 3 months before enrollment or stopping use of Aidi injection during the chemotherapy period.

The information on covariates adjusted for confounding or used for stratification during the baseline period was collected. To control the influence of confounding factors and nonrandom assignment of patients on the results, a logistic regression model was constructed and used as the propensity score. Patients were then propensity score matched 1:1 into the Aidi group and the non-Aidi group. Covariates used for matching were as follows: gender, age, body mass index (BMI), baseline ECOG (Eastern Cooperative Oncology Group) score, tumor type, TNM stage, node metastasis, distant metastasis, prior chemotherapy, prior radiotherapy, comorbidity, chemotherapy regimens, other TCMs, and chemotherapy cycle. The propensity score matching (PSM) algorithm used a 1:1 match with a maximum distance of 0.02. No replacement was allowed, and patients were matched only once. The distribution of propensity score was shown by a mirror histogram, and the symmetrical figure proved the match was sufficient. The balances of matched covariates were evaluated with standardized differences, and less than 10% of differences were considered matched sufficiently.

Ethics

The study was approved by the institutional review board of the First Affiliated Hospital of Anhui Medicine University (PJ2017-01-08), and the need for informed consent was waived.

Chemotherapy-Related Toxicity

The indices of renal function (serum creatinine, SCr) and liver function (alanine aminotransferase, ALT and aspartate aminotransferase, AST) were assessed from routine blood analyses before chemotherapy and 1 week after chemotherapy. The cutoff values of SCr, ALT, and AST were determined according to the reference ranges of markers. Abnormal values (value >1.5 times the upper limit) before treatment were defined as hepatic insufficiency or renal insufficiency. These indexes were normal before treatment but abnormalities after treatment (value >1.5 times the upper limit) were considered as injury caused by chemotherapy. Nausea, vomiting, diarrhea, constipation, and allergy were recorded per day according to the progress notes. Myelosuppression was assessed by the changes in counts of white blood cells (WBC), absolute neutrophil count (ANC), and platelets (PLT) before and after chemotherapy. Myelosuppression was divided into 0 to IV grades according to Common Terminology Criteria for Adverse Events Version 4.0 (CTCAE V4.0). WBC: 0, ⩾4.0 × 109/L, I, 3.9-3.0 × 109/L; II, 2.9-2.0 × 109/L; III, 1.9-1.0 × 109/L; IV, <1.0 × 109/L. ANC: 0, ⩾2.0 × 109/L, I, 1.9-1.5 × 109/L; II, 1.4-1.0 × 109/L; III, 0.9-0.5 × 109/L; IV, <0.5 × 109/L. PLT: 0, ⩾100 × 109/L; I, 99-75 × 109/L; II, 74-50 × 109/L; III, 49-25 × 109/L; IV, <25 × 109/L.

Performance Status

ECOG score was used to assess the PS in patients, which measures PS on a 5-point scale with 0 being “fully active, able to carry on all pre-disease performance without restriction” and 5 indicating that the patient is deceased. The evaluation of ECOG score in patients was performed as follows; ECOG score on day 1 of the chemotherapy cycle and at the end of chemotherapy treatment course were collected, and the changes of scores were considered as changes of PS in patients A decline in ECOG score was defined as improvement in PS. The invariable score was defined as stability and the increase in score was defined as deterioration in PS.

Combined Medication Information

The combined medication information was analyzed according to the prescription of each patient between 2 groups. The association rule model based on the Apriori algorithm was conducted to explore the rules of combined medication. The most used 14 medicines were included in the model. The intensity of association was divided into strong (link >50%), medium (link 25%-50%), and weak (link <25%).

Statistical Analysis

The categorical variables among the Aidi group and control group were evaluated by chi-square or Fisher’s exact test and continuous variables were analyzed by t test.

Logistic regression analysis was conducted to explore the independent influence factors associated with PS. A multivariate analysis was performed with statistically significant factors from the univariate analysis (P < .05). The difference in improvement rate of PS among the patients with different characteristics in the Aidi group was also compared by logistic regression analysis.

The stratified analysis of difference in improvement rate of PS between the Aidi group and the non-Aidi group was shown in the form of forest plots. When the 95% confidence interval (CI) upper and lower limits of odds ratios (ORs) were >1, that is, in the forest map, the 95% CI horizontal line did not intersect the invalid vertical line, and the horizontal line fell to the right side of the invalid line, it could be considered that the improvement rate of PS in the Aidi group is greater than that in the control group. The stratified analysis of association between the dosage of Aidi injection and QoL was shown in tabular form. Taking the non-Aidi group as the reference, the effects of different dosages of Aidi injection on PS were compared in the subgroup.

All analyses described above were performed using SPSS version 22.0 software (IBM Corp, Armonk, NY, USA).

Results

Patient Characteristics

A total of 2122 Aidi injection users and 2950 non-Aidi injection users among cancer patients identified between January 2014 and June 2017 (Table 1) were enrolled in our study. A PSM was then performed and 5072 cancer patients were 1:1 matched, comprising an Aidi and non-Aidi group. A total of 3200 cancer patients were eventually included in the analysis, and the baseline characteristics after PSM are shown in Table 2. After matching, covariates were well balanced with no significant differences in demographic or tumor-related variables between the 2 groups. Propensity score with mirror histograms and standardized differences before and after matching were shown in Figures 1A-B and 2.

Table 1.

Characteristics of Patients Before Propensity Score Matching.

	Aidi Injection Used
	No (N = 2950)		Yes (N = 2122)
Variable	n	%	n	%	P [a]
Gender					<.001
Female	1168	39.6	995	46.9
Male	1782	60.4	1127	53.1
Age, years
⩽60s	1350	45.8	1090	51.4	<.001
>60	1600	54.2	1032	48.6
Mean ± SD	59.7±10.8		59.2 ± 11.6		.296
Body mass index, kg/m2
<18.5	439	14.9	346	16.3	.146
18.5-24	1962	66.5	1419	66.9
>24	549	18.6	357	16.8
Tumor type					<.001
Gastric cancer	1482	50.2	1022	48.2
Lung cancer	636	21.6	372	17.5
Breast cancer	350	11.9	326	15.4
Colorectal cancer	270	9.2	245	11.5
Cardias and esophageal cancer	119	4.0	87	4.1
Hepatobiliary cancer	34	1.2	24	1.1
Ovarian cancer	59	2.0	46	2.2
TNM stage					.163
I-II	444	15.1	350	16.5
III-IV	2506	84.9	1772	83.5
Node metastasis					<.001
Negative	1639	55.6	1325	62.4
Positive	1311	44.4	797	37.6
Distant metastasis					<.001
Negative	1771	60.0	1454	68.5
Positive	1179	40.0	668	31.5
Prior chemotherapy					<.001
No	1742	59.1	1041	49.1
Yes	1208	40.9	1081	50.9
Prior radiotherapy					<.001
No	2792	94.6	2096	98.8
Yes	158	5.4	26	1.2
ECOG score					.038
0-1	2822	95.7	2003	94.4
⩾2	158	5.4	119	5.6
Comorbidity
Hypertension	385	13.1	268	12.6	.659
Renal insufficiency	33	1.1	66	3.1	<.001
Hepatic insufficiency	495	16.8	375	17.7	.406
Diabetes	136	4.6	111	5.2	.311
Chemotherapy regimen					<.001
Platinum and fluorouracil based	1370	46.4	1006	47.4
Platinum and pemetrexed based	215	7.3	166	7.8
Irinotecan and fluorouracil based	266	9.0	201	9.5
Platinum and taxane based	681	23.1	463	21.8
Cyclophosphamide and adriamycin based	157	5.3	157	7.4
Targeted therapy and others regimen	261	8.8	129	6.1
Other Chinese medicine
ShenqiFuzheng injection	827	28.0	756	35.6	<.001
Glycopeptide injection	764	25.9	645	30.4	<.001
Kushen injection	457	15.5	292	13.8	.087
Shenmai Injection	271	9.2	168	7.9	.113
DiyuShengbai tablet	161	5.5	143	6.7	.058
Chemotherapy cycles					<.001
1-3	1777	60.2	1402	66.1
>3	1173	39.8	720	33.9

Abbreviation: ECOG, Eastern Cooperative Oncology Group.

Boldfaced P values are statistically significant.

Table 2.

Characteristics of Patients After Propensity Score Matching.

	Aidi Injection Used
	No (N = 1600)		Yes (N = 1600)
Variable	n	%	n	%	P a
Gender					.414
Female	706	44.1	729	45.6
Male	894	55.9	871	55.4
Age, years
⩽60	791	49.4	784	49.0	.805
>60	809	50.6	816	51.0
Mean ± SD	60.7 ± 10.4		60.5 ± 11.1		.8085
Body mass index, kg/m2
<18.5	252	15.8	260	16.2	.665
18.5-24	1105	69.1	1082	67.6
>24	243	15.2	258	16.1
Tumor type					.101
Gastric cancer	798	49.9	815	50.9
Lung cancer	288	18.0	294	18.4
Breast cancer	212	13.2	238	14.9
Colorectal cancer	154	9.6	108	6.8
Cardiac and esophageal cancer	86	5.4	77	4.8
Hepatobiliary cancer	20	1.2	23	1.4
Ovarian cancer	42	2.6	45	2.8
TNM stage					.825
I-II	274	17.1	278	17.4
III-IV	1326	82.9	1322	82.6
Node metastasis					.689
Negative	979	61.2	990	61.9
Positive	621	38.8	610	38.1
Distant metastasis					.244
Negative	1054	65.9	1085	67.8
Positive	546	34.1	515	32.2
Prior chemotherapy					.524
No	830	51.9	812	50.7
Yes	770	48.1	788	49.9
Prior radiotherapy					.351
No	1582	98.9	1576	98.5
Yes	18	1.1	24	1.5
ECOG score					.317
0-1	1504	94.0	1517	94.8
⩾2	96	6.0	83	5.2
Comorbidity
Hypertension	197	12.3	204	12.8	.709
Renal insufficiency	24	1.5	24	1.5	.999
Hepatic insufficiency	262	16.4	256	16.0	.773
Diabetes	73	4.6	75	4.7	.866
Chemotherapy regimen
Platinum and fluorouracil based	770	48.2	773	48.2	.585
Platinum and pemetrexed based	83	5.2	96	6.0
Irinotecan and fluorouracil based	148	9.3	136	8.5
Platinum and taxane based	392	24.5	413	25.8
Cyclophosphamide and adriamycin based	110	6.9	102	6.4
Targeted therapy and others regimen	96	6.0	80	5.0
Other Chinese medicine
ShenqiFuzheng injection	503	31.4	497	31.1	.819
Glycopeptide injection	424	26.5	433	27.1	.719
Kushen injection	255	15.9	268	16.8	.534
Shenmai Injection	123	7.7	123	7.7	.999
DiyuShengbai tablet	50	3.1	65	4.1	.154
Chemotherapy cycles
1-3	1049	65.6	1055	65.9	.823
>3	551	34.4	545	34.1

Abbreviation: ECOG, Eastern Cooperative Oncology Group.

Figure 1.

Mirror histograms (A) before propensity score matching and (B) after propensity score matching.

Figure 2.

Standardized differences.

As shown in Table 3, the incidence rate of damage to renal and liver function, nausea, vomiting, and diarrhea in the Aidi group was significantly lower than that in the control group both in short and long chemotherapy cycles. As shown in Table 4, the combination of Aidi injection and chemotherapy significantly decreased the incidence rate of chemotherapy-related leucopenia, neutropenia, and thrombocytopenia compared with chemotherapy alone (P < .001, P = .033, and P < .001, respectively).

Table 3.

Chemotherapy-Related Toxicity.

	Aidi injection used
	No (N = 1600)		Yes (N = 1600)
Toxicity	n	%	n	%	P [a]
Short cycle (1-3)	1049	65.6	1055	65.9
AST elevation	114	10.9	80	7.6	.009
ALT elevation	162	15.4	96	9.1	<.001
CR elevation	30	2.9	16	1.5	.035
Nausea, vomiting	463	43.9	367	34.8	<.001
Diarrhea	93	8.9	52	4.9	<.001
Constipation	69	6.6	54	5.1	.154
Allergy	41	3.9	38	3.6	.711
Long cycle (>3)	551	34.4	545	34.1
AST elevation	99	17.9	74	13.6	.046
ALT elevation	114	20.7	83	15.2	.019
CR elevation	49	8.5	30	5.5	.030
Nausea, vomiting	293	53.2	227	41.7	<.001
Diarrhea	74	13.4	42	7.7	.002
Constipation	51	9.2	39	7.2	.205
Allergy	55	6.4	26	4.8	.254

Abbreviations: AST, aspartate aminotransferase; ALT, alanine aminotransferase; CR creatinine.

Boldfaced P values values are statistically significant.

Table 4.

Chemotherapy-Related Hematological Toxicity.

		Grade
Type	Group	0, n (%)	I, n(%)	II, n (%)	III, n (%)	IV, n (%)	P [a]
Leucopenia							<.001
	Aidi group	1221 (76.3)	166 (10.4)	174 (10.8)	38 (2.4)	1 (0.1)
	Control group	1110 (69.4)	232 (14.5)	218 (13.6)	35 (2.2)	5 (0.3)
Neutropenia							.033
	Aidi group	1248 (78.0)	153 (9.6)	149 (9.3)	47 (2.9)	3 (0.2)
	Control group	1207 (75.4)	183 (11.4)	169 (10.6)	32 (2.0)	9 (0.6)
Thrombocytopenia							<.001
	Aidi group	1448 (90.5)	97 (6.1)	48 (3.0)	7 (0.4)	0 (0.0)
	Control group	1365 (85.3)	133 (8.3)	77 (4.8)	17 (1.1)	8 (0.5)

Boldfaced P values are statistically significant.

The number of patients with decreased ECOG score (ie, improved PS) for those who underwent Aidi injection combined with chemotherapy was 137 while it was 48 for those who underwent chemotherapy alone, with a significant difference. The patients with increasing ECOG score (ie, worse PS) in the Aidi group were significantly less than those in the control group. Overall, there was a significant improvement of PS in patients in the Aidi group (Table 5).

Table 5.

Changes in Performance Status.

	Aidi Injection Used
	No (N = 1600)		Yes (N = 1600)
Performance Status	n	%	n	%	P [a]
Deterioration	59	3.7	6	0.4	<.001
Stabilization	1493	93.3	1457	91.1	<.001
Improvement	48	3.0	137	8.5	<.001

Boldfaced P values are statistically significant.

The results of the univariate analysis are shown in Table 6. Patients with node metastasis (P < .001, OR 0.4, 95% CI 0.3-0.6), distant metastasis (P < .001, OR 0.3, 95% CI 0.2-0.4), worse baseline ECOG score (P = .045, OR 0.4, 95% CI 0.1-1.0), and hepatic insufficiency (P < .001, OR 0.2, 95% CI 0.1-0.5) were significantly associated with worse PS. On the contrary, Aidi injection (P < .001, OR 2.9, 95% CI 2.1-4.1) and long chemotherapy cycles (P < .001, OR 2.1, 95% CI 1.5-2.8) were the influence factors associated with better PS. Patients with lung cancer, breast cancer, cardia and esophageal cancer, and ovarian cancer had better improvement rate of PS compared with patients with gastric cancer. According to the results of multivariate analysis, Aidi injection was a significant independent factor of improving PS after adjusting for the confounding factors (P < .001, OR 3.4, 95% CI 2.4-4.8). Other independent factors were tumor types, node metastasis, distant metastasis, baseline ECOG score, hepatic insufficiency, and chemotherapy cycle.

Table 6.

Logistic Regression Model With Odds Ratios (ORs) and 95% CI of Improvement Rate of Performance Status.

	Univariable			Multivariable
Variable	Crude OR	95% CI	P	Adjusted OR	95% CI	P
Gender
Female	[reference]
Male	1.2	0.9-1.6	.294
Age, years
⩽60	[reference]
>60	0.8	0.6-1.2	.291
Body mass index, kg/m2
<18.5	[reference]
18.5-24	0.8	0.5-1.2	.264
>24	1.3	0.8-2.1	.303
Tumor type
Gastric cancer	[reference]			[reference]
Lung cancer	3.1	2.1-4.6	<.001	3.7	2.3-5.9	<.001
Breast cancer	3.0	1.9-4.6	<.001	3.1	1.9-5.1	<.001
Colorectal cancer	1.4	0.7-2.6	.371	0.3	0.1-1.0	.060
Cardiac and esophageal cancer	2.7	1.4-5.1	.002	2.5	1.3-4.8	.007
Hepatobiliary cancer	2.2	0.7-7.3	.200	2.6	0.7-9.1	.135
Ovarian cancer	5.8	3.1-10.9	<.001	6.8	3.4-13.7	<.001
TNM stage
I-II	[reference]
III-IV	1.0	0.7-1.4	.848
Node metastasis
Negative	[reference]			[reference]
Positive	0.4	0.3-0.6	<.001	0.4	0.3-0.6	<.001
Distant metastasis
Negative	[reference]			[reference]
Positive	0.3	0.2-0.4	<.001	0.2	0.2-0.4	<.001
Prior chemotherapy
No	[reference]
Yes	0.9	0.7-1.3	.683
Prior radiotherapy
No	[reference]
Yes	2.2	0.9-5.8	.094
ECOG score
0-1	[reference]			[reference]
⩾2	0.4	0.1-1.0	.045	0.3	0.1-0.9	.024
Hypertension
No	[reference]
Yes	1.4	0.9-2.1	.121
Renal insufficiency
No	[reference]
Yes	1.1	0.3-3.5	.918
Hepatic insufficiency
No	[reference]			[reference]
Yes	0.2	0.1-0.5	<.001	0.5	0.2-1.0	.037
Diabetes
No	[reference]
Yes	0.9	0.4-1.9	.793
Chemotherapy regimen
Platinum and fluorouracil based	[reference]			[reference]
Platinum and pemetrexed based	2.3	1.3-4.0	.003	1.2	0.6-2.3	.589
Irinotecan and fluorouracil based	1.0	0.5-1.8	.928	1.3	0.7-2.5	.444
Platinum and taxane based	1.6	0.8-3.0	.161	1.7	0.9-3.4	.127
Cyclophosphamide and adriamycin based	1.9	1.1-3.4	.019	1.1	0.6-2.0	.870
Targeted therapy and others regimen	1.8	1.2-2.5	.002	1.2	0.8-1.8	.353
Other Chinese medicine
ShenqiFuzheng injection
No	[reference]
Yes	1.1	0.8-1.5	.477
Glycopeptide injection
No	[reference]
Yes	1.0	0.7-1.4	.965
Kushen injection
No	[reference]
Yes	1.0	0.7-1.5	.975
Shenmai injection
No	[reference]
Yes	1,0	0.6-1.8	.878
DiyuShengbai tablet
No	[reference]
Yes	1.8	1.0-3.5	.067
Chemotherapy cycles
1-3	[reference]			[reference]
>3	2.1	1.5-2.8	<.001	2.8	2.1-3.9	<.001
Aidi injection used
No	[reference]			[reference]
Yes	2.9	2.1-4.1	<.001	3.4	2.4-4.8	<.001

Abbreviation: ECOG, Eastern Cooperative Oncology Group.

Boldfaced P values are statistically significant.

In a stratified analysis (Figure 3), the result indicated that Aidi injection led to a durable improvement in PS across almost all subgroups Stratified by gender, age, TNM stage, node metastasis or not, prior chemotherapy or not, baseline ECOG score, chemotherapy regimens, other Chinese medicines and different chemotherapy cycles, the Aidi group showed higher improvement rate of PS than that in the control group. In subgroups of patients with hepatobiliary cancer, distant metastasis and comorbidity (hypertension, renal insufficiency, hepatic insufficiency, and diabetes), the improvement rate of PS showed no statistical difference between the 2 groups. However, the ORs from almost all subgroups suggested that cancer patients receiving chemotherapy might benefit from Aidi injection.

Figure 3.

Stratified analysis.

The difference in improvement rate of PS among the patients with different characteristics in the Aidi group is shown in Table 7. Aidi injection was shown to be less effective in patients with advanced age (>60 years), node metastasis, distant metastasis, worse baseline ECOG score compared to patients with younger age (⩽60 years), no node metastasis, no distant metastasis and better baseline ECOG score. In addition, Aidi injection had better effects for patients with lung cancer, breast cancer, cardia and esophageal cancer and ovarian cancer compared to patients with gastric cancer, colorectal cancer, and hepatobiliary cancer. It was worth noting that Aidi injection combined with platinum and pemetrexed or taxane should be more beneficial to patients in comparison with other chemotherapy regimens. From the perspective of treatment cycles, patients with long chemotherapy cycles might have a higher improvement rate of PS.

Table 7.

Odds Ratios (ORs) and 95% CI of Performance Status Benefit Associated With Aidi Injection for Patients With Different Characteristics.

	Aidi User (N = 1600)
	Crude OR			Adjusted OR
Variable	Crude OR	95% CI	P [a]	Adjusted OR	95% CI	P [a]
Gender
Female	[reference]			[reference]
Male	1.4	1.0-2.0	.064	1.1	0.7-1.7	.798
Age, years
⩽60	[reference]			[reference]
>60	0.8	0.5-1.1	.121	0.5	0.4-0.8	.004
Body mass index, kg/m2
<18.5	[reference]			[reference]
18.5-24	0.8	0.5-1.2	.264	1.0	0.6-1.7	.932
>24	1.3	0.8-2.1	.303	1.5	0.8-2.9	.242
Tumor type
Gastric cancer	[reference]			[reference]
Lung cancer	2.6	1.6-4.2	<.001	2.4	1.3-4.4	.004
Breast cancer	2.6	1.5-4.2	<.001	2.1	1.0-4.2	.041
Colorectal cancer	1.9	0.9-3.9	.080	1.5	0.7-3.4	.290
Cardiac and esophageal cancer	2.9	1.4-6.0	.005	3.3	1.4-7.3	.004
Hepatobiliary cancer	1.6	0.4-7.0	.537	2.1	0.4-9.8	.367
Ovarian cancer	6.4	3.1-13.2	<.001	6.9	2.8-17.2	<.001
TNM stage
I-II	[reference]			[reference]
III-IV	1.0	0.6-1.6	.948	1.5	0.8-2.6	.188
Node metastasis
Negative	[reference]			[reference]
Positive	0.4	0.3-0.6	<.001	0.4	0.3-0.7	<.001
Distant metastasis
Negative	[reference]			[reference]
Positive	0.2	0.1-0.4	<.001	0.2	0.1-0.3	<.001
Prior chemotherapy
No	[reference]			[reference]
Yes	1.0	0.7-1.4	.924	0.9	0.6-1.3	.583
Prior radiotherapy
No	[reference]			[reference]
Yes	2.9	1.1-7.8	.039	1.0	0.7-1.6	.874
ECOG score
0-1	[reference]			[reference]
⩾2	0.4	0.1-1.2	.109	0.2	0.1-0.8	.024
Hypertension
No	[reference]			[reference]
Yes	0.7	0.4-1.3	.235	1.1	0.6-2.3	.687
Renal insufficiency
No	[reference]			[reference]
Yes	0.5	0.1-3.4	.447	0.5	0.1-4.3	.499
Hepatic insufficiency
No	[reference]			[reference]
Yes	0.3	0.1-0.6	<.001	0.5	0.2-0.8	<.001
Diabetes
No	[reference]			[reference]
Yes	0.4	0.1-1.4	.158	0.5	0.1-1.9	.336
Chemotherapy regimen
Platinum and fluorouracil based	[reference]			[reference]
Platinum and pemetrexed based	3.0	1.6-5.6	<.001	2.4	1.1-5.4	.037
Irinotecan and fluorouracil based	1.1	0.5-2.4	.731	1.2	0.5-2,7	.662
Platinum and taxane based	2.4	1.1-4.9	.021	2.9	1.3-6.5	.010
Cyclophosphamide and adriamycin based	2.3	1.2-4.5	.012	1.2	0.5-2.6	.728
Targeted therapy and others regimen	2.0	1.3-3.0	.002	1.4	0.9-2.4	.160
Other Chinese medicine
ShenqiFuzheng injection
No	[reference]			[reference]
Yes	1.1	0.8-1.5	.477	1.2	0.8-1.9	.389
Glycopeptide injection
No	[reference]			[reference]
Yes	1.0	0.7-1.4	.965	1.1	0.7-1.7	.756
Kushen injection
No	[reference]			[reference]
Yes	1.0	0.7-1.5	.975	1.0	0.6-1.7	.899
Shenmai injection
No	[reference]			[reference]
Yes	1.0	0.6-1.8	.878	1.4	0.7-2.8	.365
DiyuShengbai tablet
No	[reference]			[reference]
Yes	1.8	1.0-3.5	.067	1.6	0.4-4.0	.328
Chemotherapy cycles
1-3	[reference]			[reference]
>3	2.2	1.5-3.1	<.001	3.0	2.0-4.6	<.001
Dosage
No Aidi use	[reference]			[reference]
Less than conventional dosage (20-50 mL per day)	2.3	1.2-4.4	.014	2.8	1.4-5.6	.004
Conventional dosage (50-100 mL per day)	3.1	2.2-4.4	<.001	3.6	2.5-5.1	<.001

Abbreviation: ECOG, Eastern Cooperative Oncology Group.

Boldfaced P values are statistically significant.

Effect of Different Dosage of Aidi Injection

According to the instructions, the conventional dosage of Aidi injection for adults is 50 to 100 mL per day. Yet in our research, a subset of patients had used Aidi injection at less than conventional dosage in actual clinical use (20-50 mL). To understand the association between the dosages of Aidi injection and QoL, we further classified patients into three subgroups (conventional dosage group, less than conventional dosage group and non-Aidi group) according to the actual dosage of Aidi injection. Compared with non-Aidi injection, we found that the adjusted OR for PS is 2.8 (95% CI 1.4-5.6; P = .004) among the patients who used Aidi injection at less than the conventional dosage and 3.6 (95% CI 2.5-5.1; P < .001) in the conventional dosage group, which suggested that conventional dosage of Aidi injection might have had a better effect on PS (Table 7). According to this result, a subgroup analysis was then performed to further investigate the effect of different dosage on patients with different characteristics. As showed in Table 8, in almost all subgroups, the conventional dosage of Aidi injection showed more benefit in PS on patients compared to the other 2 groups. Besides, in comparison with the non-Aidi group, using the conventional dosage of Aidi injection decreased the incidence rate of all types of chemotherapy-related toxicity. However, using Aidi injection with less than conventional dosage had a poor effect on reducing incidence rate of chemotherapy-related toxicity (only showing significant difference in leukopenia compared with the non-Aidi group) (Table 9). In addition, the proportion of patients with improved PS was higher in the conventional dosage group, which was statistically different from the other 2 groups (Table 9). These results revealed that conventional dosage of Aidi injection had a better effect in improving PS and decreasing toxicity, that is, it was more beneficial to QoL of cancer patients.

Table 8.

Odds Ratios (ORs) and 95% CI of Performance Status Benefit Associated With Different Dosage of Aidi Injection in the Stratified Analysis.

	Non-Aidi Use (N = 1600)	Aidi Use
		Less Than Conventional Dosage (n = 183)		Conventional Dosage (n = 1417)
	OR	OR (95% CI)	P [a]	OR (95% CI)	P [a]
Variable
Gender
Female	[reference]	2.0 (0.6-5.9)	.238	2.2 (1.7-2.9)	<.001
Male	[reference]	1.2 (1.1-1.6)	<.001	1.5 (1.2-1.8)	<.001
Age, years
⩽60	[reference]	1.5 (1.2-1.9)	<.001	2.0 (1.6-2.6)	<.001
>60	[reference]	1.7 (0.6-4.5)	.297	1.6 (1.2-2.0)	<.001
Body mass index, kg/m2
<18.5	[reference]	4.4 (0.8-25.3)	.094	2.9 (1.5-5.0)	<.001
18.5-24	[reference]	2.6 (1.3-5.4)	.009	1.5 (1.2-1.8)	<.001
>24	[reference]	1.3 (0.9-4.3)	.252	2.1 (1.4-3.2)	<.001
Tumor type
Gastric cancer	[reference]	6.0 (2.4-15.0)	<.001	2.1 (1.4-2.9)	<.001
Lung cancer	[reference]	1.1 (0.2-4.9)	.912	1.4 (1.0-1.9)	.029
Breast cancer	[reference]	1.6 (1.1-2.3)	<.001	1.9 (1.3-2.8)	.002
Colorectal cancer	[reference]	1.4 (0.9-3.3)	.792	1.8 (1.1-3.6)	<.001
Cardiac and esophageal cancer	[reference]	12.1 (0.9-143.0)	.057	3.0 (1.0-8.9)	.042
Hepatobiliary cancer	[reference]	1.3 (0.6-9.7)	.891	0.8 (0.3-1.8)	0.591
Ovarian cancer	[reference]	2.2 (1.6-3.3)	.042	2.5 (1.1-5.6)	.026
TNM stage
I-II	[reference]	2.1 (0.4-10.2)	.357	1.7 (1.1-2.5)	.009
III-IV	[reference]	1.3 (1.1-1.7)	<.001	1.8 (1.5-2.2)	<.001
Node metastasis
Negative	[reference]	1.6 (1.4-1.9)	<.001	1.7 (1.4-2.1)	<.001
Positive	[reference]	0.9 (0.1-7.4)	.936	1.9 (1.3-2.8)	<.001
Distant metastasis
Negative	[reference]	1.3 (1.2-1.6)	.018	1.8 (1.5-2.2)	<.001
Positive	[reference]	1.3 (0.2-11.0)	.782	1.4 (0.9-2.2)	.145
Prior chemotherapy
No	[reference]	1.4 (0.5-3.8)	.476	1.7 (1.3-2.1)	<.001
Yes	[reference]	3.7 (1.5-9.2)	.004	1.9 (1.5-2.5)	<.001
ECOG score
0-1	[reference]	1.4 (1.3-1.7)	<.001	1.8 (1.5-2.1)	<.001
⩾2	[reference]	0.9 (0.3-1.6)	.496	1.2 (0.5-2.6)	.733
Hypertension
No	[reference]	1.4 (1.2-1.9)	<.001	1.9 (1.6-2.3)	<.001
Yes	[reference]	2.0 (0.4-9.7)	.395	1.3 (0.9-1.9)	.181
Renal insufficiency
No	[reference]	1.4 (1.3-1.7)	<.001	1.8 (1.5-2.1)	<.001
Yes	[reference]	0.4 (0.2-0.8)	.797	0.8 (0.2-2.8)	.731
Hepatic insufficiency
No	[reference]	1.4 (1.3-4.7)	.008	1.8 (1.5-2.2)	<.001
Yes	[reference]	1.1(0.9-2.9)	.594	1.4 (0.8-2.6)	.265
Diabetes
No	[reference]	1.5 (1.3-1.8)	<.001	1.9 (1.6-2.2)	<.001
Yes	[reference]	0.6 (0.4-1.9)	.395	0.8 (0.4-1.6)	.484
Chemotherapy regimen
Platinum and fluorouracil based	[reference]	3.5 (1.4-8.7)	.006	1.8 (1.4-2.4)	<.001
Platinum and pemetrexed based	[reference]	6.7 (1.0-46.0)	.054	2.2 (1.1-4.2)	.018
Irinotecan and fluorouracil based	[reference]	3.2 (0.6-17.6)	.185	1.2 (1.0-1.8)	.042
Platinum and taxane based	[reference]	2.8 (1.1-4.9)	.045	3.2 (1.1-9.2)	.033
Cyclophosphamide and adriamycin based	[reference]	1.7 (1.3-3.4)	.037	2.5 (1.2-5.6)	.019
Targeted therapy and others regimen	[reference]	0.5 (0.1-4.1)	.542	1.6 (1.2-2.1)	.002
Other Chinese medicine
ShenqiFuzheng injection
No	[reference]	1.4 (1.2-1.7)	<.001	1.7 (1.4-2.1)	<.001
Yes	[reference]	1.8 (0.8-4.2)	.152	1.9 (1.4-2.6)	<.001
Glycopeptide injection
No	[reference]	2.4 (1.1-5.0)	.026	1.7 (1.4-2.1)	<.001
Yes	[reference]	2.1 (0.6-7.5)	.275	1.9 (1.3-2.6)	<.001
Kushen injection
No	[reference]	1.9 (0.9-4.1)	.083	1.8 (1.5-2.2)	<.001
Yes	[reference]	5.5 (1.3-22.3)	.018	1.5 (1.1-2.2)	.038
Shenmai injection
No	[reference]	1.5 (1.2-1.7)	<.001	1.7 (1.4-2.0)	<.001
Yes	[reference]	3.2 (1.6-7.2)	.042	3.9 (1.4-10.9)	.009
DiyuShengbai tablet
No	[reference]	1.3 (1.2-1.7)	.011	1.7 (1.5-2.1)	<.001
Yes	[reference]	1.7 (1.2-4.8)	.023	2.5 (1.1-7.2)	<.001
Chemotherapy cycle
1-3	[reference]	1.3 (0.4-3.8)	.631	1.7 (1.4-2.2)	<.001
>3	[reference]	4.1 (1.7-9.6)	.001	1.8 (1.4-2.3)	<.001

Abbreviation: ECOG, Eastern Cooperative Oncology Group.

Boldfaced P values are statistically significant.

Table 9.

Effects of Different Dosage of Aidi Injection on Chemotherapy-Related Toxicity and Performance Status.

	Non-Aidi Group	Aidi Group
	(N = 1600), n (%)	Less Than Conventional Dosage (n = 183), n (%)	Conventional Dosage (n = 1417), n (%)
Toxicity
AST elevation	213 (13.3)	27 (9.6)	127 (8.9)[a,b]
ALT elevation	276 (17.3)	31 (16.9)	152 (10.7)[a,b]
CR elevation	79 (4.9)	8 (4.4)	36 (2.7)[b]
Nausea, vomiting	756 (47.2)	67 (36.6)	527 (37.0)[a,b]
Diarrhea	167 (10.4)	18 (8.2)	76 (5.4)[a,b]
Constipation	120 (7.5)	16 (8.7)	77 (5.3)[a]
Allergy	96 (6.0)	8 (4.4)	56 (4.0)[a]
Leucopenia	490 (30.6)	41 (22.4)[a]	338 (23.9)[a]
Neutropenia	393 (24.6)	43 (23.4)	305 (21.5)[a]
Thrombocytopenia	235 (14.7)	18 (9.8)	134 (9.4)[a]
Performance status
Deterioration	59 (3.7)	4, (2.2)	2 (0.2)[a,b]
Stabilization	1493 (93.3)	167 (91.2)[a]	1290 (91.0)[a]
Improvement	48 (3.0)	12 (6.6)[a]	125 (8.8)[a]

Abbreviations: AST, aspartate aminotransferase; ALT, alanine aminotransferase; CR, creatinine.

P < .05 compared with non-Aidi group.

P < .05 compared with Aidi group with less than conventional dosage.

To explore the real clinical combined medication rules of Aidi injection and determine whether there was a difference in medication between the 2 groups, the prescriptions of patients during chemotherapy were collected. The 9 most commonly used kinds of Western medicines and 5 kinds of traditional Chinese medicines are listed in Table 10. Glucocorticoid, antiemetic drugs, painkillers, stomach drugs, antiallergic drugs, and Chinese medicines, which were mainly extracted from ginseng were the most commonly used drugs. Notably, there is no statistical difference between the two groups, which further showed that the difference in QoL between Aidi users and non-Aidi users could be attributed to the Aidi injection. As shown in Figure 4A and B, the highest frequency (except Aidi injection) of joint use were dexamethasone plus metoclopramide plus tropisetron both in the Aidi group and non-Aidi group (the stronger the link is in the figure, the higher the frequency of combined use).

Table 10.

Combined Medication Information.

	Aidi Injection Used
	No (N = 1600)		Yes (N = 1600)
Drug	n	%	n	%	P
Dexamethasone	1126	70.3	1112	69.5	.589
Metoclopramide	1049	65.5	1037	64.8	.656
Tropisetron	911	56.9	905	56.6	.830
Pantoprazole sodium	706	44.1	703	43.9	.915
Omeprazole	231	14.4	229	14.3	.920
Clonazepam	179	11.2	163	10.2	.360
Cimetidine	159	9.9	149	9.3	.549
Diphenhydramine	185	11.6	191	11.9	.742
Tramadol	143	8.9	139	8.7	.803
ShenqiFuzheng injection	503	31.4	497	31.1	.819
Glycopeptide injection	424	26.5	433	27.1	.719
Kushen injection	255	15.9	268	16.8	.534
Shenmai injection	123	7.7	123	7.7	.999
DiyuShengbai tablet	50	3.1	65	4.1	.154

Figure 4.

Analysis of association rules of combined drug use of (A) Aidi users and (B) non-Aidi users.

Discussion

In past years, clinical research attached increasing importance to QoL, and a large number of studies investigated QoL as an outcome of cancer treatment. Yet due to the restriction of the subjectivity, complexity, and adherence of patients and physicians, it is difficult to assess QoL in real-world treatment.[23] Much research had indicated that PS is associated with QoL from different perspectives. As Moningi and Walker[25] reported, patients with the worse PS are correlated with worse QoL. It was reported in another study that higher QoL is significantly associated with better PS, and in the lower the QoL, the lower the PS will be.[24] Apart from PS, several reports had suggested that QoL and chemotherapy-related toxicity are significantly correlated.[28,29] Based on this research, PS served as a major surrogate endpoint in the present study to assess QoL in cancer patients, with the incidence rate of chemotherapy-related toxicity supplied for supplementalevidence. The findings of our study suggested that Aidi injection combined with chemotherapy could decrease the incidence rate of chemotherapy-related toxicity and improve PS in patients as compared with chemotherapy alone, that is, Aidi injection could improve QoL in cancer patients.

In fact, PS refers to a patient’s ability to perform activities and functions that meet basic physical needs in the face of illness and is attempt to quantify cancer patient’s general well-being and activities of daily life, which are affected by various symptoms and immune functions.[28-30] Our study has shown that Aidi injection could decrease the incidence rate of various symptoms caused by chemotherapy, including nausea, vomiting, and so on (Tables 3 and 4). This might be due to the ability of Aidi injection to support body energy, clearing away heat and toxicity. Additionally, immune function damage is a serious toxicity, including lower antitumor and anti-infective immunity induced by chemotherapy, and is a major factor influencing PS. According to several studies, Aidi injection could significantly increase the percentage of CD3+ cells, CD4+ T cells, CD8+ T cells, NK cells, and the CD4+/CD8+ T cells ratio of peripheral blood and reverse the Th1/Th2 shift. These results revealed that Aidi injection could enhance cellular immunity.[31,32] In addition, many studies have also shown that ginseng, Astragaloside and Cantharidin have immune regulation functions, which indirectly confirms that Aidi injection might enhance PS through improving immune function.[14-16]

Generally speaking, confounding factors are the major problem inherent in observational studies, which can affect persuasiveness. Addressing these known and unknown confounders was carefully considered in the study design: PSM was conducted to group patients[33] and the chemotherapy-related toxicities were then compared after PSM. Damage to liver and renal and gastrointestinal reactions that occurred during chemotherapy in the Aidi group were significantly less than that in the control group, as shown in Table 3. The 2 groups had statistically significant difference in the incidence rate of leukopenia, neutropenia, and thrombocytopenia, particularly leukopenia and thrombocytopenia (P < .001, P = .033, and P < .001, respectively). This finding is probably because of Astrgaloside,[16] one of the main ingredients of Aidi injection, which can promote the proliferation of bone-marrow stem cells. Apart from PSM, stratification and multivariate modeling were conducted to adjust the impact of confounding factors. The PS benefit of using Aidi injection remained consistent after adjustment for a large number of factors including age, gender, TNM stage, and other variables (Table 6). Regardless of gender, age, TNM stage, node metastasis, prior chemotherapy, baseline ECOG score, chemotherapy regimens, other Chinese medicines as well as different chemotherapy cycles, the PS benefit of Aidi injection was confirmed in the stratified analysis. By adjusting the model, the probability of false results is likely to be low, most interference of confounding factors can be eliminated, and the validity of our results can increase. Besides, this study suggests that patients taking Aidi injection in conventional dosages might experience a higher improvement rate of PS and lower incidence rate of chemotherapy-related toxicity compared to the non-Aidi group and the group receiving lower dosages. The benefit of Aidi injection for patients was further supported by this finding.

Our results showed that patients with gastric cancer had a lower improvement rate of PS in comparison with the patients with lung cancer, breast cancer, cardiac and esophageal cancer, as well as ovarian cancer. This is probably because 70.16% patients with gastric cancer in our study had gastric surgery before enrollment, which might result in various symptoms, including loss of body weight and malnutrition.[34] The resulting symptoms could cause significant deterioration in their PS.[35] Yet from the subgroup of gastric cancer in the stratified analysis, the improvement rate of PS in the Aidi group was significantly higher than that in the control group (OR = 3.7, 95% CI 1.9-7.2). This suggests that patients with gastric cancer could benefit from Aidi injection, although gastric cancer is a factor associated with worse PS. This benefit might be attributed to inclusion of Ciwujia (Eleutherococcus senticosus) among the main ingredients of Aidi injection,[8] which has been shown to replenish the vital essence and alleviate poor appetite.[36] These results revealed that patients with gastric cancer could benefit from using Aidi injection compared with non-Aidi-users.

Hepatic insufficiency caused by chronic liver disease or liver metastasis, a poor prognostic factor for cancer patients, was found in 518 patients in the study group.[37-41] In our study, the result of presence of hepatic insufficiency substantially inhibited any gains (OR = 0.5, 95% CI 0.2-1.0) in PS compared with normal liver function. Because of the protective effect of Astragaloside on liver injury,[16,42,43] Aidi injection reduced the incidence rate of liver injury (Table 3). Still, patients using Aidi injection showed no advantages in the improvement rate of PS in comparison with the patients not using Aidi injection in the hepatic insufficiency subgroup (Figure 3). Besides, in the Aidi group, patients with hepatic insufficiency show no significant difference compared with those with normal liver function (OR = 0.5, 95%CI 0.2-1.8, P < .001) (Table 7). Hepatic insufficiency, which affects the detoxification and metabolic functions of the liver, might also alter the effect of Aidi injection. Unfortunately, the specific reasons still need further exploration because of the complex effective components of TCM. Given this problem, we are making preparation to conduct further relevant study on Aidi injection and liver injury, to supplement clinical evidence.

In this study, to explore the information of the combined medication of Aidi injection, the prescriptions of patients receiving chemotherapy were collected (Table 10). By conducting the association rule model,[44] we found there was no significant difference in the drug combination between Aidi users and non-Aidi users (Figure 4A and B). Thus, we further demonstrated that the difference in QoL between the 2 groups was the effect of Aidi injection. Based on this model, thousands of combined medication rules were revealed and it was difficult to compare the impact of these rules of drug use on QoL. In a follow-up study, we will pay more attention to the most effective combination of Aidi injection. Yet our research has shown Aidi injection as the main factor on QoL and demonstrated the clinical effect of the combined medication of Aidi injection, which might be meaningful for clinical rational drug use.

The strengths of the present study consist of the large sample size and the fact that we had collected more detailed information, inclusive of routine blood work, blood biochemistry and combined medication information before and after treatment. Its size and the fact that it was extracted from real-world data provide a robust profile of PS for these patients in conditions of usual clinical practice. The findings of this study suggested a positive causal link between Aidi injection and the QoL of cancer patients receiving chemotherapy. It is reasonably assumed that this integrative approach might be recommended to patients.

Limitations of our study were a result of the retrospective design. First and foremost, the findings were limited by potential and unrecognized confounding factors, for example, drinking and smoking. Yet these are considered unimportant factors determining patients’ entry into the cohort groups. Second, our data were collected from EMR database in hospital, where some problems remained. Because of the mobility of patients in large hospitals and limited time of follow-up, the outcome of survival time could not be explored in this study. Third, patients with longer hospitalization days (eg, longer chemotherapy cycle) have a better chance to use Aidi injection. Although we strive to conquer this problem by matching the baseline characteristics and apply logistic regression model to adjust all of the covariates in 2 groups, we cannot overstate the findings. However, to the best of our knowledge, this study is still the largest observational study to investigate the effectiveness of Aidi injection for patients with different types of cancer who receiving chemotherapy on the basis of real-world data, which might provide robust evidence.

Conclusion

This retrospective cohort study showed that Aidi injection as adjuvant to chemotherapy might improve the QoL in cancer patients by improving PS and reducing incident rate of chemotherapy-related toxicity. In addition, because of excellent effect of Aidi injection on patients with different types of cancer, it should be recommended to apply in clinical work.