Medication adherence in leukemia children receiving home-based treatment and its related factors

ABSTRACT

Objectives:

This study aimed to investigate home-based medication adherence in children with acute leukemia and its related factors.

Methods:

We examined 132 children with acute leukemia in a tertiary pediatric hospital in Chongqing. A general questionnaire, the MMAS-8 (the eight-item Morisky Medication Adherence Scale), SEAMS (Self-efficacy for Appropriate Medication Use Scale), and a multifactorial logistic regression model were used to analyze the factors related to drug adherence of the children.

Results:

54.55% of patients showed good medication adherence, while 50.76% either forgot a dose or administered an incorrect dose. The average score on the Self-Efficacy for Appropriate Medication Use Scale (SEAMS) was 32.47.61. The logistic regression analysis revealed that a SEAMS score, the occupation of caregivers, and the patient’s age were predictors of medication adherence among pediatric leukemia patients (P < 0.05).

Conclusions:

The home-based medication adherence in children with acute leukemia was not favorable. Patients with low SEAMS scores, farmers serving as caregivers, and children under the age of three should arouse more attention. It is expected to strengthen the confidence of patient families in medication by emphasizing the development of their relationships with healthcare professionals. Awareness of breakthroughs in leukemia home-based medication management systems by utilizing Internet technology.

KEYWORDS:

• Child
• leukemia
• medication adherence
• related factors

Introduction

Leukemia in children is a severe, malignant hematologic disease with the highest incidence rate among childhood cancers. In China, the average incidence of infant leukemia each year is 42.9 per million [1]. Chemotherapy has a more pronounced effect on children with leukemia than on adults. The 5-year OS (overall survival) and EFS (event-free survival) of children with ALL (acute lymphoblastic leukemia) in China exceed 90% and 80%, respectively [2]. Oral medication plays a crucial role in the treatment of acute leukemia [3], as it influences the therapeutic effect, therapeutic response, and disease recurrence [3,4]. A study published in Blood, a journal of the American Society of Hematology [5], revealed that approximately 25% of adolescent ALL patients did not take their maintenance medication after disease remission. In addition, the study found that the risk of recurrence increases as drug adherence decreases. This study examined home-based medication adherence in children with acute leukemia and analyzed its related factors to provide a reference for targeted intervention strategies.

Materials and methods

Investigation

Between September and December of 2021, 132 patients with acute leukemia were recruited from the hematologist-oncologist clinic of a tertiary pediatric hospital in Chongqing. The inclusion criteria: a diagnosis of acute leukemia according to international guidelines for childhood leukemia; complete remission and continuous maintenance treatment following initial prescriptive treatment; one or more orally administered medications; and voluntary participation. Exclusion criteria: the patient’s caretaker doesn’t want to take part; either the patient or the caretaker has a severe mental illness or has been experiencing huge emotional stress recently.

Investigative tools

General questionnaire

Researchers reviewed relevant literature and developed their own general questionnaire to take the potential related factors into account. The demographic data included the patient’s age and sex; the caregiver’s age, sex, education, marital status, etc. Disease-related information such as when the patient began maintenance treatment, error rates related to drug administration, such as missed dose and wrong dose (missed dose or wrong dose/ total number of medications taken*100%), etc.

MMAS-8 (The eight-item Morisky Medication Adherence Scale)

Morisky et al. created the Morisky Medication Adherence Scale (Cronbach’s alpha coefficient of 0.83) in 2008 [6]. A total of eight questions were based on this scale. The first seven questions required a ‘Yes’ or ‘No’ response, with a negative response yielding 1 point and a positive response yielding 0 point. Question 5 was scored in reverse order. Question 8 used a 5-point Likert scale, with the following five options available: (1) never; (0.75) rarely; (0.50) occasionally; (0.25) frequently; and (0) always. The MMAS-8 score ranges from 0 to 8, with higher scores indicating greater patient adherence. A score of ≥7 indicates good adherence, whereas <7 indicates poor adherence.

SEAMS (The self-efficacy for appropriate medication use scale)

The SEAMS scale (Cronbach’s alpha coefficient of 0.934 and retest confidence of 0.932) was developed by Risser in 2007 and then translated by Dong Xiaofang et al. in 2015 [7]. It consists of 13 items and is divided into two domains: self-efficacy in uncertain situations (five items) and self-efficacy in difficult situations (eight items). A 3-point Likert scale with the following responses was used: (1) lack of confidence, (2) some confidence, and (3) extreme confidence. An overall score falls between 13 and 39 points.

Data collection

A total of 132 caregivers participated in this study. In the outpatient clinic, they signed PIC (patient informed consent) forms. Five to ten minutes were then spent independently completing the questionnaires. All 132 questionnaires were valid and complete.

Statistical techniques/methods

In order to ensure the accuracy of our data and analyses, we entered the questionnaire data using version 24.0 of the SPSS software. In a univariate analysis, data not obeying normal distribution were represented by the median (quartile-spacing), while data obeying normal distribution were represented by mean ± standard deviation and frequencies and percentages were expressed as count data. To eliminate the interaction of multiple factors, the significant factors indicated by univariate analysis were used as independent variables, and the drug adherence of children served as the dependent variable. Multivariate logistic regression model was performed. P < 0.05 indicated significant difference.

Results

Subject information

In this group, the average age of caregivers was (35.36 ± 8.23) years; 66 cases (50%) were college graduates; 41 farmers (31.06%) were the major labor force, while 35 caregivers (26.52%) were unemployed; There were 93 mothers (70.45%), 31 fathers (23.48%), and 8 other relationship ties (6.06%). The average age of the patients was (6.72 ± 3.87) years; 75 were male (56.82%) and 57 were female (43.18%); and fifty percent were enrolled in kindergarten or primary school. 118 patients (89.39%) were diagnosed with acute lymphocytic leukemia; 90.15% of patients had medical insurance. 52.27% of patients came from households with an annual income of more than 30,000 CNY. The typical maintenance treatment duration was four months [3,8]. Details are shown in Table 1.

Table 1. Subject information.

Demographic characteristics	n	Ratio (%)
Relationship between caregivers and patients
Father	31	23.48%
Mother	93	70.45%
Other relationship	8	6.06%
Caregiver’s age (years)
<30	25	18.94%
30–39	63	47.73%
≥40	42	31.82%
Caregiver’s education level
Primary school	23	17.42%
Secondary school	65	49.24%
College and above	44	33.33%
Caregiver’s occupation
Worker	13	9.85%
Civil service/Corporate staff	13	9.85%
Farmer	41	31.06%
Other	30	22.73%
Unemployed	35	26.52%
Patient’s sex
Male	75	56.82%
Female	57	43.18%
Patient’s age (years)
< 3	11	9.85%
3∼6	60	45.45%
6∼12	40	30.30%
>12	21	15.90%
Patient enrolled in school
Yes	66	50%
No	66	50%
Household annual income (CNY)
<30,000	63	47.73%
30,000 ∼80,000	41	31.06%
>80,000	28	21.21%
Mode of payment for medical expenses
Self-funded	13	9.85%
Medical insurance	119	90.15%
Medication data
Number of missed dose/wrong dose	67	50.76%

Home-based medication of children with leukemia

Patients self-efficacy scores

The scores are shown in Table 2, and the Self-efficacy total score was (32.47 ± 6.13) points.

Table 2. Self-efficacy score of leukemia patients.

Indicator	Mean ± Standard Deviation
The doctor changes your medicine*	2.70 ± 0.51
You need to take several types of drugs every day*	2.73 ± 0.46
You take medication more than once per day*	2.75 ± 0.45
You are not at home*	2.66 ± 0.54
You are busy*	2.63 ± 0.57
You experience adverse effects*	2.21 ± 0.74
No one reminds you*	2.52 ± 0.61
Medication process is complicated*	2.55 ± 0.58
Your daily schedule is disrupted*	2.48 ± 0.61
You don’t believe in taking the medication*	2.25 ± 0.72
You’re not sure how to take the medication*	2.27 ± 0.74
You have other illnesses (chills or a cold)*	2.14 ± 0.76
The drug you bought is different from the one you are used to taking*	2.56 ± 0.54
SEAMS total score	32.47 ± 6.13

*No confidence = 1, some confidence = 2, very confident = 3.

Medication adherence of children with leukemia taking home-based medicine

It was demonstrated that the Morisky score corresponded to actual medication statistics. 60 (45.45%) patients were scored less than 7 points, indicating poor drug adherence; 72 (54.55%) patients were scored 7 or more points, indicating good adherence. The drug compliance score was not representative of a normal distribution. The median (quadratic spacing) analysis yielded a value of 7.00 (5.75, 7.75) points. According to the statistics, 67 cases (50.76%) had missed a dose, while 5 patients (3.79%) received an incorrect dose.

Analysis of factors related to medication adherence

Univariate analysis results of medication adherence

The univariate analysis of medication adherence revealed a significant correlation with the SEAMS score, caregiver’s occupation, whether or not the patient is enrolled in school, and patient’s age (P < 0.05). According to the data, 67 cases (50.76%) had either missed a dose (62 patients) or taken an incorrect dose (5 patients). The univariate analysis of a patient’s missed dose or wrong dose revealed a significant correlation with the SEAMS score and the caregiver’s literacy (P < 0.05). The findings of the univariate analyses are shown in Table 3.

Table 3. Univariate analysis of medication adherence and error.

Demographic characteristics	Morisky score		χ^2	P value	Missed dose/wrong dose		χ^2	P value
	<7	≥7			Yes	No
Relationship between caregivers and patients
Father	12	19	0.998	0.464	14	17	2.002	0.321
Mother	43	50			47	46
Other	5	3			6	2
Caregiver’s sex
Male	15	24	1.092	0.296	18	21	0.469	0.493
Female	45	48			49	44
Caregiver’s age (years)
< 30	14	11	1.407	0.236	16	9	2.168	0.141
≥ 30	45	60			50	55
Caregiver’s education level
High school and below	43	45	1.237	0.266	50	38	3.880	0.049*
Above high school	17	27			17	27
Caregiver’s occupation
Farmer	24	17	4.105	0.043*	24	17	1.440	0.230
Other	36	55			43	48
Patient’s sex
Male	33	42	0.148	0.700	37	38	0.141	0.707
Female	27	30			30	27
Patient’s age (years)
< 3	9	2	6.400	0.011*	7	4	0.796	0.372
≥ 3	51	70			60	61
Patients enrolled in school
Yes	36	30	4.400	0.036*	35	31	0.273	0.601
No	24	42			32	34
Patient’s disease
Acute Lymphoblastic Leukemia	54	64	0.043	0.836	62	56	1.418	0.234
Other	6	8			5	9
Annual household income (CNY)
≤ 30,000	34	29	3.523	0.061	36	27	1.966	0.161
>30,000	26	43			31	38
Mode of payment for medical expenses
Self-funded	31	45	2.785	0.248	37	39	1.855	0.396
Medical insurance	24	19			25	18
Other	5	8			5	8
Information	Median	Morisky score		P value	Missed dose/wrong dose		P value
		<7 points	≥7 points		Yes	No
Duration of intermittent treatment (M)	4 [2, 9]	5 [2,10.25]	3 [2, 7]	0.191	4 [2, 8]	4 [2, 9]	0.987
SEAMS score	34 [28, 38]	31 [25, 36]	36 [30,39]	0.001	36 [31,39]	30 [25.5,36.5]	0.001

Multivariate analysis of factors related to Morisky scores of leukemia patients

SEAMS score, caregiver’s occupation, patient’s age, and enrollment status indicated by univariate analysis were used as independent variables, and children’s medication adherence was taken as dependent variable. Logistic regression was performed, and the results showed that a low SEAMS score, a caregiver who was a farmer, and a patient who was younger than three years old were risk factors for medication adherence in children. The values of the independent variables are presented in Table 4, and the results of the analysis are presented in Table 5.

Table 4. Values of independent variables.

Independent variable	Value
Caregiver’s education level
High School and below (Primary school, Middle school, High school)	1
Above high school (College, Undergraduate, Masters and above)	2
Caregiver’s occupation
Farmer	1
Not a farmer	2
Patient’s age
≤ 3 years old	1
>3 years old	2
Patients enrolled in school
Yes	1
No	0

Table 5. Multivariate logistic regression analysis on Morisky scores.

Incorporated indicators		B	S.E.	Wald	P	OR	95% of CI value
							Lower	Upper
Step 1	Seams score	.100	.031	10.388	.001	1.105	1.040	1.174
	constant	−3.046	1.019	8.931	.003	.048
Step 2	Patient’s age	.091	.031	8.550	.003	1.095	1.030	1.164
	Seams score	1.568	.820	3.658	.056	4.797	.962	23.923
	Constant	−5.785	1.828	10.017	.002	.003
Step 3	Caregiver’s occupation	.090	.031	8.220	.004	1.094	1.029	1.164
	Patient’s age	1.633	.823	3.935	.047	5.120	1.020	25.702
	Seams score	.820	.405	4.089	.043	2.270	1.026	5.023
	Constant	−7.262	1.998	13.207	.000	.001

Multivariate analysis of missed and wrong doses

The SEAMS score and the literacy level of the caregiver indicated by univariate analysis were used as independent variables, and the patient’s missed dose and incorrect dose were taken as dependent variables. Multivariate logistic regression was performed, and the results showed that a low SEAMS score and a secondary school education as the highest level of education for the caregiver were risk factors for missed or incorrect doses in leukemia patients. The results of the analysis are shown in Table 6.

Table 6. Multivariate logistic regression analysis of missed or wrong doses in children with leukemia.

Incorporated indicators		B	S.E.	Wald	P	OR	95% of CI value
							Lower	Upper
Step 1	Seams score	−.101	.031	10.349	.001	.904	.850	.961
	Constant	3.324	1.047	10.085	.001	27.779
Step 2	Seams score	−.104	.032	10.559	.001	.901	.846	.959
	Caregiver’s education level	−.803	.397	4.093	.043	.448	.206	.975
	Constant	4.507	1.243	13.146	.000	90.648

Discussion

Poor medication adherence

The study revealed that 45.45% of patients had poor medication adherence; up to 50.76% of them missed (50 patients) or took the incorrect (5 patients) dose; 12 patients discontinued medication due to observed side effects. Numerous studies have demonstrated that oral chemotherapy and/or prophylactic antibiotics are frequently prescribed to non-adherent cancer patients. More than 44% of adolescent cancer patients did not adhere to oral chemotherapy, and more than 48% did not adhere to oral antibiotics for prevention [8–10]. Approximately 25% of ALL adolescent patients missed their maintenance medication after disease remission, according to a study published in Blood, a journal of the American Hematology Society [5]. Patients who did not adhere to oral antibiotic preventative drugs had significantly lower survival rates than those who did [11]. Medical personnel should always place a high priority on medication adherence; one way to improve adherence is to encourage both the children and their families.

Analysis of the factors related to drug adherence

The results of the multivariate analysis revealed three risk factors of leukemia patients taking home-based medication, including a low SEAMS score, a farmer as a caregiver, and a patient younger than 3 years old.

SEAMS (The self-efficacy for appropriate medication use scale)

Self-efficacy is the expectation, perception, confidence, or belief in one’s ability to successfully implement the course of action necessary to attain a specific objective. Self-efficacy is a significant component of an individual’s personality structure and plays a crucial role in human psychological and behavioral development [12]. Self-efficacy is also a significant predictor of drug compliance. Several studies suggest that rational drug self-efficacy is a key factor for improving drug compliance in patients with chronic diseases [13], consistent with our findings. The SEAMS reflects the widespread belief that children should receive the utmost care at home; therefore, children with leukemia must take their medication at home under the supervision of a caregiver. Continuous advancements in chemotherapy have contributed to an increase in the survival rate of children with leukemia. It has increased the confidence of patient families in oral medication. The top three leading indicators were ‘You take medication more than once per day’ [(2.75 ± 0.45) points], ‘You need to take several types of drugs every day’ [(2.73 ± 0.46) points], and ‘The doctor changes your medicine’ [(2.70 ± 0.51) points]. The lowest three indicators were ‘You have other illnesses (chills or a cold)’ [(2.14 ± 0.76) points], ‘You experience adverse effects’ [(2.21 ± 0.74) points], and ‘You’re not sure how to take the medication’ [(2.25 ± 0.72) points]. Lack of experience in managing other diseases and adverse reactions, lack of timely access to medical support when experiencing adverse reactions, and uncertainty regarding drug usage have also been shown to affect a patient’s drug adherence. In light of these considerations, it is crucial to strengthen relationships between patient families and health personnel. This will encourage families to seek more support from their health care provider, while also encouraging health personnel to provide personalized services to patients.

Caregivers occupation and educational level

High school was the highest level of education in 95.17% of farmers in Liu Zhiyuan’s survey. In this study, there was a higher risk of noncompliance with medication if the caregiver was a farmer [14]. Similarly, the multivariate analysis in our study indicated that the risk of a missed or incorrect dose of medication increased if the highest level of education of the caregiver was secondary school. It may be suggested that patients whose caregivers are farmers are more likely to have poor drug adherence. Our study’s findings align with those of ACHARYA [15] and Moster et al. [16]. A higher level of education suggests a greater capacity to independently perceive and learn information. There are numerous oral drugs for leukemia, some for long-term use and others with varying frequencies (daily, weekly, three times a week, etc.). The dose and frequency are adjusted based on the child’s condition and observed changes.

In this study, 66.66% of the caregivers had completed only secondary school. Medical personnel must concentrate on finding new ways to make leukemia treatment at home more manageable, particularly for caregivers with limited literacy.

Patients age

The results indicated that children younger than 3 years old were more likely to not adhere to their medication regimens. This is due to their inability to accurately describe the effects of medication on their body and provide feedback to their caretakers, as well as their inability to take drugs orally. Younger children are more susceptible to adverse drug reactions, and chemotherapy affects their growth and development because their vital organs are not fully developed. To improve drug adherence, both primary and secondary caregivers must actively participate in the home medication management of children. Families could change their attitudes toward medication with additional support.

According to this study, children receiving home-based treatment for leukemia who were younger than 3 years old, had a low SEAMS score, and had a caretaker who worked as a farmer were at a higher risk for poor drug adherence. Due to the limitations of this study, our research was conducted at a single pediatric hospital. Extrapolating the results was conducted with caution. In addition, the factors influencing medication adherence were not exhaustive; future research may include psychological conditions of patients and their families, etc. The study revealed that the medication adherence of leukemia patients receiving home-based treatment was subpar. Due to research constraints, only one hospital for children with leukemia was studied. Moreover, the factors that influence drug compliance are not exhaustive.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This research was funded by Chongqing Medical University Affiliated Children's Hospital, Chongqing Science and Health Joint Medical Youth Project [grant number 20220NXM029]; Chongqing Science and Health Joint Medical General Project [grant number 2022MSXM084] and Construction and application of Integrated “monitoring-evaluation-intervention” intelligent management platform for home medication of children with malignant tumors [grant number ZHYX202225].