Inflammatory bowel diseases (IBD), including Crohn’s disease (CD) and ulcerative colitis (UC), are chronic inflammatory diseases of the gastrointestinal tract. Approximately 25% of patients first develop symptoms under 20 years of age, while only 4%–10% are diagnosed before the age of 6 years (Rosen et al., 2015; Afzali and Katz, 2018). The incidence of pediatric IBD has risen rapidly in recent years, and it has emerged as a global disease in developed and developing nations (Benchimol et al., 2011; Benchimol et al., 2017; Piovani et al., 2020). Symptoms of IBD include gastrointestinal symptoms such as abdominal pain, diarrhea, and blood in stool, and non-gastrointestinal symptoms such as arthritis and autoimmune liver disease, which lead to slow growth, delayed puberty, and decreased quality of life. In addition, psychiatric comorbidities, particularly depression and anxiety, are common in pediatric IBD patients (Loftus et al., 2011; Keerthy et al., 2016).

The treatment goals of IBD are to induce and maintain clinical remission, achieve mucosal healing, and reduce the need for surgeries and hospitalizations (Xiao et al., 2014). Medications, including corticosteroids, 5-aminosalicylates, immunomodulators, and biological agents, are the primary means to achieve these treatment goals. However, non-adherence to physician-prescribed medications is common in children with IBD (Hommel et al., 2009; Spekhorst et al., 2016). Previous studies have shown that poor adherence to medication can lead to disease recurrence or progression, resulting in the need to step-up therapy, reducing health-related quality of life, and increasing treatment costs (Gray et al., 2012; LeLeiko et al., 2013; Pearce and Fleming, 2018; Plevinsky et al., 2019; Collyer et al., 2020). Therefore, understanding and improving medication adherence in children with IBD is critical for optimizing treatment and improving treatment outcomes. However, very little is known on this subject in China.

Oral medications are a mainstay of IBD treatment, non-adherence to oral medications is regarded as a pervasive problem in the management of IBD (Greenley et al., 2012). Therefore, we focus our discussion and analysis on the oral medication adherence in children with IBD. The first step in understanding adherence is quantifying its incidence. This can be performed by self-report, objective measures, or mixed indicators; the most commonly used is the patient-self-reported medication adherence scale because of its effectiveness, reliability, convenience, and cost-effectiveness. However, there are literatures showing that subjectively reported adherence (e.g., self-report, parent-report) can lead to overestimates of medication adherence due to recall bias and social desirability (Wu et al., 2013; Spekhorst et al., 2016). Then, some scholars have found that subjectively and objectively reported adherence were consistent and correlated (Wu et al., 2013; Selinger et al., 2019). Selinger et al. (2019) suggested the Medication Adherence Report Scale (MARS) and Visual Analog Scales (VAS) could be used as simple and effective screening tools for medication non-adherence in patients with IBD in clinical practice. The 4-item Medication Adherence Report Scale (MARS-4) is a widely used adherence-specific self-reported questionnaire that has been translated into several languages and has been shown to have good reliability and validity in assessing medication adherence in patients with different diseases (Horne et al., 2009; Jacobsen et al., 2009; Jeganathan et al., 2017). However, a Chinese version of the MARS-4 is not currently available. Therefore, in this study, we sought to translate the English version of the MARS-4 into Chinese, evaluate its psychometric properties, and explore the predictors of non-adherence to anti-inflammatory bowel disease medication in a group of children with IBD in China.

IBD is a chronic disease requiring lifelong treatment regimens to maintain remission. Unfortunately, poor adherence to oral medication is quite prevalent. Objective studies reveal medication non-adherence rates up to 93% in pediatric patients with IBD (Oliva-Hemker et al., 2007; Hommel et al., 2009; Ingerski et al., 2010; Spekhorst et al., 2016). The assessment and promotion of oral medication adherence is very complex. MARS could be used as a simple and effective screening tools for medication non-adherence in patients with IBD in clinical practice (Selinger et al., 2019). To our knowledge, this study is the first to demonstrate the reliability and validity of the Chinese version of the MARS-4 in assessing medication adherence in Chinese children with IBD. The study shows that the internal consistency reliability, test-retest reliability, and structural validity of the scale are relatively good, indicating that it can be used to measure medication adherence in Chinese children with IBD. However, it is worth mentioning that Item 2 has an item-total correlation below 0.5, this result is consistent with that of a previous study (Jacobsen et al., 2009). The reason for this may be that fewer patients changed the dose. Further, patients had access to a WeChat group with their children’s doctors, allowing them to get clear and timely answers when they had any questions regarding the medication use of their children at home, rather than arbitrarily changing the drug dosage themselves. This may be the reason why item 2 has a lower item-total correlation. Despite this, when this item was deleted, the Cronbach’s ⍺ only has a very small improvement with 0.006, and the internal consistency reliability of the Chinese version of the MARS-4 remains good. Based on this, MARS-4 was used in this study to assess children’s medication adherence. We found that nearly 70% of them forgot to take medication, while few changed the medication dosage, which were similar to the findings of (Ingerski et al., 2010), (Schurman et al., 2011). The most commonly reported medication adherence barrier was forgetfulness, for which some scholars have adopted a variety of reminder systems, both traditional, such as pill boxes, and novel, such as email and telephone alerts, have been considered as useful behavioral interventions for unintentional non-adherence (Bermejo et al., 2010). Therefore, in future research, we can adopt medication reminder systems, which may be able to promote children’s medication adherence.

This is the first study to investigate medication adherence and analyze the predictors of non-adherence to medication in Chinese children with IBD. Our results showed that about a quarter of children with IBD have poor medication adherence, which is higher than the rate in the United States (14.61%–17.47%) (Greenley et al., 2015), but similar to that in South Korea (31.2%) (Lim et al., 2020), lower than that in Jordan and the United Kingdom (36.2%) (Alsous et al., 2020), and much lower than that in Japan (43.0%) (Tanaka et al., 2021). Non-adherence to medication in children with IBD is common worldwide. To our surprise, parent reported non-adherence (31.37%) was higher than Child self-reported non-adherence (25.00%) in this study, however, the difference being not statistically significant. This was different from the findings of (Alsous et al., 2020), (Reed-Knight et al., 2011), but (Ingerski et al., 2010). Furthermore, it revealed that children reported lower scores for unintentional non-adherence and tended to have higher scores for intentional non-adherence, compared with parent-reported results, but the differences between them were not statistically significant. This may be why parents reported a higher proportion of non-adherence than children reported outcomes, children may gained an advantage in intentional non-adherence over the advantage parents gained in unintentional non-adherence. On the other hand, it may also be due to the small sample size, which needs to be increased in the future.

In this study, we found that longer disease duration (3 years and above) and parents with poor knowledge about the disease were independent risk factors for poor medication adherence, while other factors, such as age, family economic level, medication frequency, and treatment with monotherapy or multiple medications did not show a significant effect, similar to and different from previous findings. Kitney et al. (2009) also found that medication non-adherence in children was significantly associated with longer disease duration, and Lim et al. (Tanaka et al., 2021) showed that patients’ knowledge about previously prescribed medications taken was related to good adherence. However, several studies have reached different conclusions. Greenley et al. (2010) showed that children whose regimen involved >1 daily medication administration reported poorer adherence than those whose regimen involved 1 or <1 daily medication administration time, and children on monotherapy reported better adherence than those on multiple medications. LeLeiko et al. (2013); Kitney et al. (2009) found that the medication adherence of children aged ≥15 years was poorer than that of children aged <15 years. In our study, we did not find the same, possibly due to the small sample size. For example, in this study, only eight of the 119 children with IBD were treated with multiple medications, which was not sufficient for a statistically significant analysis, and needed to be re-evaluated with a larger sample size.

In this study, the total score of the parental CCKNOW was only 8.0 (5.0-10.0), which is highly consistent with the findings of (Colombara et al., 2015). It revealed that the parental knowledge about IBD was sub-optimal and still needs to be improved. In this study, we found that parental knowledge about IBD might be a predictor for medication adherence in children with IBD, and the innovation of the current research lies in that we further analyzed the correlation between the different categories of parental knowledge and medication adherence in children with IBD. To our surprise, both drug and complication knowledge categories were significantly associated with medication adherence in children with IBD, which had not been explored in previous studies. Previously, some scholars (Quan et al., 2003; Colombara et al., 2015) also found that the IBD patients appeared to even more lack knowledge about drugs and complications, but did not explored the correlation with medication adherence. In addition, a number of previous studies have shown that providing knowledge about medication and disease to patients effectively increased treatment adherence (Robinson et al., 2001; Tae et al., 2016; Tiao et al., 2017; Lim et al., 2020; Vernon-Roberts et al., 2020). Interventions including parent education seminar (Vernon-Roberts et al., 2020), Pharmacist Counseling (Tiao et al., 2017), self-management training (Robinson et al., 2001) and so on. However, only (Reed-Knight et al., 2011) revealed that neither adolescents’ nor parents’ report of medication knowledge were significantly associated with medication adherence in children with IBD. It might be due to the small sample size and different knowledge assessment tools. Therefore, in future research, we should incorporate knowledge about IBD, especially the drug knowledge and complication knowledge as key components of implementing targeted strategies to improve medication adherence in children with IBD.

The power of health insurance status as a predictor of medication adherence has not been previously studied, but it could be a risk factor for medication adherence. The actual cost of medicines varies depending on the type of health insurance, which may affect medication adherence in children with IBD, particularly when taking expensive medicines. Ediger et al. (2007) had previously reported the medication cost was the most prominent issue, with 25% of patients reporting that cost often made it difficult for them to take their medication regularly. However, health insurance status was not a significant predictor in this study, and was not significantly different between the adherent and non-adherent groups, this could be due to a lack of statistical power as the small sample size. It may also be that oral medication, such as mercaptopurine and mesalazine, are not as expensive as biologics, so the type of health insurance has no obvious impact on patients’ cost and thus on medication adherence. Similarly, (Nahon et al., 2011) found that deprivation was not a predictor of medication non-adherence in IBD patients due to the ease of access medication in France where IBD patients qualify for payment exemption. The above studies were conducted in adults with IBD, data in children are limited. Current researches are more focused on exploring whether non-adherence will lead to increased medical costs (Kane and Shaya, 2008; Higgins et al., 2009; Elkjaer et al., 2010), while relatively limited research reports on whether cost, such as drug prices and health insurance types, are predictors of non-adherence in IBD patients. However, this potential risk factor deserves more time for further investigation with different medicine, such as biologics, in future studies.

This single-center, cross-sectional study was conducted at a children’s hospital. However, the demographics of these patients reflect only the epidemiology of IBD in Chinese children, meaning that the results may not be generalizable to other nationalities. In addition, this study only used a subjective adherence assessment tool with no objective tool for more comprehensive information, which needs to be supplemented in our next step for more exploration. Further, relatively small sample size may not have provided sufficient statistical power to identify some factors; we will expand the sample size in the future. However, the usual rule of thumb for sample size in binary logistic regression of 5–10 cases per variable was met in our study, and we adopted forward stepwise logistic regression analysis. In addition, about one-third of the children had intentional dose omissions, while about one-quarter experienced self-discontinuation behavior in our study, which suggests that it is necessary to analyze the potential causes of children’s intentional non-adherence. In the future, we will perform further studies to address these limitations.

We report the Chinese version of the MARS-4 as a useful medication adherence tool in children with IBD of China and further describe the prevalence and predictors of medication non-adherence in these populations. Poor medication adherence was common in children with IBD in China, and subjective forgetting to take medication was the main barrier. Disease duration and the level of parental knowledge about IBD were independent predictors of medication non-adherence. Future studies should be carried out to focus on the interventions of disease-related knowledge education, especially drug and complication knowledge education, and medication reminders to promote medication adherence in children with IBD. In addition, identifying the adherence threshold necessary for medication efficacy will be of great value in clinical practice, it will provide a basis for when to initiate interventions.