Inflammatory bowel disease (IBD) is a chronic relapsing and debilitating condition, and in many of the cases, there is no cure. It can first occur at any time from early childhood till late adulthood, with 20% occurring before the age of 20 years. The disease is potentially well controlled with patients experiencing a normal lifestyle with ongoing use of medical (pharmacological) therapy and, in select cases, surgical therapy. From adult studies, it has been reported that patients who are nonadherent to treatment are up to 5.5 times more likely to experience a relapse in symptoms (Kane, Huo, Aikens, & Hanauer, 2003). In addition, nonadherence to treatment generates a 12.5% greater annual healthcare cost than adherent patients (Higgins, Rubin, Kaulback, Schoenfield, & Kane, 2009).

Background

Adherence refers to the consistency between patient behavior and the medical advice given to them after collaboration regarding patient values and preferences (Sewitch et al., 2003). Good adherence involves taking medication at the correct dose and frequency, as well as following other lifestyle recommendations (e.g., diet, substance use) that also help promote good health (Bermejo et al., 2010). Medication nonadherence is associated with a threefold likelihood relapse in symptom control (Feagins, Iqbal, & Spechler, 2014), underscoring the important preventive role that the medication offers. Nonadherence is not always intentional (i.e., a deliberate decision not to take medication as directed), and can occur unintentionally (i.e., forgetting, not fully understanding, financial barriers; Horne, Parham, Driscoll, & Robinson, 2009). Previous reviews identified convenient medication regimens and good patient-doctor relationship (Lakatos, 2009), as well as focusing on modifiable factors such as anxiety and medication attitudes, are considerations for intervention (Selinger, Robinson, & Leong, 2011).

Examining nonadherence in IBD cohorts is complicated by the drastic changes in treatment seen in recent years. A substantially greater number of patients are now treated with biological medications (e.g., infliximab, adalimumab; Yu et al., 2018). This change may reduce the burden associated with daily medication administration, as doses are administered every 2-8 weeks depending on the drug, rather than on a day-to-day basis, as is typically necessitated by oral medications (e.g., 5-ASA, azathioprine). Although some biological medications are self-administered via subcutaneous injections (e.g., adalimumab, certolizumab), others such as infliximab and vedolizumab require a switch to facility-administered intravenous infusions.

Several studies have highlighted the potential impact this change may have on medication adherence among IBD patients. As Bucci et al. (2017) recently found, patients undergoing biological therapies appear more likely to be adherent compared with those taking daily oral treatments. Rates of "low" adherence, according to the Morisky Adherence Scale (Morisky, Green, & Levine, 1986), were considerably less for intravenous (4.4%) and injectable biological medications (6.3%) compared with oral steroids (40%), oral 5-ASA (37.5%), and immunomodulators (28%). Along the same lines, facility-administered biological treatments are associated with greater adherence compared with self-administered biological medications (Selinger et al., 2013; Wentworth et al., 2018).

Despite the apparent adherence benefit of biological therapies, findings to date suggest that rates of nonadherence are still problematic among these patients. A review by Lopez, Billioud, Peyrin-Biroulet, and Peyrin-Biroulet (2013) across 13 studies revealed pooled adherence rates of 82.6% among infliximab and adalimumab cohorts. A more recent 24-month retrospective cohort study (N = 365) found pooled adherence rates across biologics to be 66%, with the highest rate for the facility-administered vedolizumab (83%) and the lowest for certolizumab pegol (50%; Wentworth et al., 2018). In comparison, another study examining infliximab adherence within a pediatric cohort found that 91.4% of patients were adherent (Vitale, Greenley, Lerner, Mavis, & Werlin, 2015).

As with conventional, nonbiological treatments, adherence is essential for achieving optimal patient outcomes. Nonadherence has been associated with a greater likelihood of acute infusion reactions, reduced efficacy due to greater antidrug antibody formation, and greater likelihood disease flare-ups (Baert et al., 2003; Fidder et al., 2009; Govani et al., 2018). While much of the extant literature is based on patients adhering to daily oral medication regimens, these findings are nonetheless relevant, given the use of concomitant immunomodulator therapy (e.g., daily azathioprine) among patients undergoing biological treatments (Khanna et al., 2015), and the low adherence rates associated with immunomodulator use in adult (Bucci et al., 2017; Mantzaris et al., 2006) and pediatric (Oliva-Hemker, Abadom, Cuffari, & Thompson, 2007) IBD cohorts alike.

In a recent comprehensive systematic review of the IBD adherence literature, Lenti and Selinger (2017) identified modifiable (e.g., psychological distress, doubts about medication, low trust in physician) and nonmodifiable (e.g., low socioeconomic status, minority patients, religion, pregnancy) predictors of nonadherence. However, Lenti and Selinger argued that little progress had been made with regard to interventions, and more rigorous methodology was needed. On the basis of their comprehensive review of the literature, Lenti and Selinger concluded that adherence is associated with a complex interaction of modifiable and nonmodifiable factors and that patient-tailored and multidimensional interventions are required. Recommendations in relating to adolescents versus adult IBD cohorts were not provided and hence are the basis for this paper.

Adolescence is commonly defined as the ages between 10 and 17 years, with adulthood beginning to emerge at the age of 18 years (Arnett, 2000). This article reviews the differences and similarities between adolescent and adult cohorts in relation to the rates, predictors, and consequences of nonadherence. Modifiable factors such as barriers, the social and medical setting, and interventions to improve adherence in IBD are also considered, with recommendations for healthcare professionals working in this area. Although this is the definition of adolescent and adult used in this paper, some studies reviewed may vary by +/-2 years.

Methods of Adherence Measurement

Medication adherence can be measured by subjective (e.g., self-reported during interview), objective (e.g., pill counts), or biological (e.g., bioassay/blood draw) methods (Hommel, Davis, & Baldassano, 2009). Standardized subjective self-report scales have been used such as the Medication Adherence Report Scale (e.g., Ediger et al., 2007) or the Morisky Medication Adherence Score (Morisky et al., 1986). However, it should be noted that the interpretation of scores vary, in that sometimes any items endorsed are classified as nonadherent (e.g., Cross, Cheevers, Rustgi, Langenberg, & Finkelstein, 2012) or there is a threshold such as 80% (e.g., Gifford et al., 2013). Objective pill counts have been undertaken involving manually monitoring prescription dates, dosing instructions, and remaining pills in bottles (e.g., Hommel, Hente, Herzer, Ingerski, & Denson, 2013), or using Medical Event Monitoring System pill bottles that electronically record the time and date of each time the bottle is opened (e.g., Greenley, Gumidyala, et al., 2015). Biological assays such as blood or urine tests can also objectively monitor adherence through medication levels in the body (Gifford et al., 2013; Hommel et al., 2009). See Lenti and Selinger (2017) for a detailed review of the advantages and disadvantages of adherence measure methods.

Rates of Adherence

Adult IBD Cohorts

Jackson, Clatworthy, Robinson, and Horne (2010) systematically reviewed medication adherence in IBD among adults and estimate that self-reported nonadherence rates (using self-reports, interviews) are between 55% and 70%. Younger adults have been found to be significantly less adherent (Cerveny, Bortlik, Kube na, et al., 2007; Shale & Riley, 2003) and have significantly greater odds (1.04-1.5, 3.64 in women; Campos, Portela, Sousa, & Sofia, 2016; D'inca et al., 2008; Ediger et al., 2007; Moshkovska et al., 2009) of medication nonadherence. Lenti and Selinger (2017) argue that this finding is counterbalanced by studies that find no age difference among adults. However, many of these studies with nonsignificant age findings focused on individuals with IBD who are adults only (e.g., Bernal et al., 2006; Bokemeyer et al., 2007; Lopez San Roman, Bermejo, Carrera, Perez-Abad, & Boixeda, 2005).

Adolescent IBD Cohorts

Adolescents report higher rates of adherence than the rate of adults described above, with 65%-90% of adolescents reporting perfect adherence (Greenley et al., 2012; Greenley, Stephens, Doughty, Raboin, & Kugathasan, 2010; LeLeiko, Lobato, Hagin, Hayes, et al., 2013; LeLeiko, Lobato, Hagin, McQuaid, et al., 2013; Reed-Knight, Lewis, & Blount, 2013). When comparing older adolescents with younger adolescents, younger adolescents report greater rates of >80% adherence (67%-84%) than older adolescents (15%-25%; LeLeiko, Lobato, Hagin, McQuaid, et al., 2013). Kamperidis et al. (2012) found in a sample of 138 patients that adolescents were over fourfold more likely to be nonadherent to thiopurines than adults. Individuals earlier in adolescence are more adherent than individuals later in adolescence (Kitney et al., 2009).

Comparing adults and adolescents with IBD, adolescents tend to be approximately 10%-20% more adherent, with this observation being more pronounced in early adolescence. This difference may be related to the fact that parents and caregivers have more of an influence in general on their children in all aspects of life when they are younger when compared with when they are older. As parental involvement and supervision are reduced, it is likely that adolescents lose a key protective factor with regard to nonadherence. Given that IBD may be diagnosed in middle age, it is also possible that these individuals do not have the same level of familiarity with long-term medication usage compared with adults who were diagnosed during adolescence, where parental involvement may help establish good medication-taking habits from an early age.

Risk Factors for Poor Adherence

Adult IBD Cohorts

Risk factors for nonadherence can be found in patient demographics and social surroundings, as well as the medical setting. Although largely not modifiable (Selinger et al., 2011), these influences are nonetheless illustrative of the range of challenges IBD patients must navigate in adhering to a medication regimen. Being busy with work (Kim et al., 2016; Shale & Riley, 2003) and higher perceived stress (Tabibian et al., 2015) have both been found to increase the risk of nonadherence. Further, a low education level has been found to nearly double the risk of nonadherence (Lakatos et al., 2010). Individuals with previous IBD surgeries are 48% less likely to be nonadherent (Pittet et al., 2014) and African Americans are less likely than White Americans to be adherent to medication, independent of employment, socioeconomic status, income, health insurance, and marital status (Nguyen et al., 2009).

With regard to social influences, Mountifield, Andrews, Mikocka-Walus, and Bampton (2014) found low social support doubles the risk of nonadherence, even after controlling for demographic correlates of nonadherence (i.e., age, gender, employment, pill burden, quality of life, and disease type). Another potential social factor in medication adherence may lie in the perceived social stigma experienced by IBD patients. This is evidenced by the findings of Taft, Keefer, Leonhard, and Nealon-Woods (2009), which found social stigma to account for 5% of the variance in adherence to medication, after controlling for demographic variables and illness characteristics (e.g., remission status, flare ups, flare-up duration). Similarly, qualitative findings suggest the social embarrassment of taking multiple pills and fatigue of long-term medication use are potential barriers to adherence (Devlen et al., 2014). Research by Horne et al. (2009) examined IBD patient attitudes regarding maintenance therapy and found attitudes of skepticism (low perceived need, high concerns), indifference (low need, low concerns), and ambivalence (high perceived need, high concerns) to predict nonadherence, independent of clinical and demographic factors. Struggles with mental health, such as anxiety (Nahon et al., 2011) and depression (Goodhand et al., 2013; Long et al., 2014; Nahon et al., 2012; Shale & Riley, 2003) also impact adherence. However, it should be noted that adult IBD patients report a range of different strategies to aid with adherence, including using daily routine as cues (78%), a medication dispenser (17%), calendar or checklist (6%), reminder from a family member (6%), and blister pack from the pharmacy (3%; Ediger et al., 2007).

The relationship a patient has with medical professionals influences medication adherence. The patient-doctor relationship has been shown to be an important predictor of medication adherence. A study by Nguyen et al. (2009) of 235 IBD patients found that greater trust in their physician was associated with a decrease in nonadherence, and for each half a standard deviation increase of trust, there was a 41% increase in odds of being adherent. Similarly, having a negative relationship with medical professionals is associated with over a five-fold risk of low adherence (Mountifield et al., 2014). The frequency of clinical visits and absenteeism from IBD appointments has also been shown to be associated with nonadherence (Mountifield et al., 2014; Tae et al., 2016). Specifically, Tae et al. (2016) found that patients who visit 3 months apart are 30-fold less likely to be adherent than patients who attend monthly or less. With regard to appointment absenteeism, Mountifield et al. (2014) found that patients who missed IBD appointments were 4.4-fold more likely to be nonadherent. Experiencing visible bleeding as part of abdominal symptoms, being overburdened with nine or more tablets (Kawakami et al., 2014), three-times daily dosing (Shale & Riley, 2003), or four or more concomitant medications with mesalamine (Kane, Cohen, Aikens, & Hanauer, 2001) also predict nonadherence. Research also suggests that patient knowledge and information-seeking are related to medication nonadherence. Tae et al. (2016) found that low medication knowledge is associated with a 5.6-fold increased risk of nonadherence. After taking into account potential confounders and nonadherence risk factors, Pittet et al. (2014) found active information-seeking patients (i.e., seeking out information on IBD, treatments, research, and development) were 2.43 more likely to be nonadherent.

It is more common that individuals unintentionally (38%-81%) are nonadherent than intentionally (8%-39%; Bernal et al., 2006; Cerveny, Bortlik, Vlcek, Kubena, & Lukas, 2007; Sewitch et al., 2003). Unintentional nonadherence is self-reported as due to being the weekend (16%; Bernal et al., 2006), working days (10%; Bernal et al., 2006), not being home (10%; Bernal et al., 2006), forgetting (5%-60%; Ediger et al., 2007; Horne et al., 2009; Lopez San Roman et al., 2005; Sewitch et al., 2003), or just being careless at times about taking medication (12%-38%; Lopez San Roman et al., 2005; Sewitch et al., 2003). In contrast, intentional reduce-dose nonadherence is self-reported as being due to feeling well and no need to continue (11%-64%; Cerveny, Bortlik, Kube na, et al., 2007; Cerveny, Bortlik, Vlcek, et al., 2007; Ediger et al., 2007; Lopez San Roman et al., 2005; Sewitch et al., 2003), adverse side effects (9%-33%; Cerveny, Bortlik, Kube na, et al., 2007; Cerveny, Bortlik, Vlcek, et al., 2007; Ediger et al., 2007; Lopez San Roman et al., 2005; Sewitch et al., 2003), reducing dose by oneself without the doctor (11%-32%; Ediger et al., 2007; Horne et al., 2009), reducing the dose to avoid running out before appointment (18%-22%; Cerveny, Bortlik, Kube na, et al., 2007; Cerveny, Bortlik, Vlcek, et al., 2007), reduce dose to reduce medication costs (25%; Ediger et al., 2007), deciding to skip doses (17%; Horne et al., 2009), not refilling their prescription in time (11%-15%; Cerveny, Bortlik, Kube na, et al., 2007; Cerveny, Bortlik, Vlcek, et al., 2007), unpleasant associations with medication (13%; Ediger et al., 2007), uncertainties about the effectiveness of medication (12%; Ediger et al., 2007), or pregnancy or breastfeeding in women (4%-10%; Cerveny, Bortlik, Vlcek, et al., 2007). Regarding increase-dose intentional nonadherence (i.e., self-medicating), self-reported reasons include suspected worsening of disease (80%; Bernal et al., 2006), fear of hospitalization (12%; Bernal et al., 2006), and fear of being on sick leave (10%; Bernal et al., 2006).

Adolescent IBD Cohorts

Unintentional nonadherence in adolescents is often referred to as barriers to adherence (e.g., forgetting; Hommel & Baldassano, 2010). Semistructured clinical interviews (Medication Adherence Measure [MAM]; Zelikovsky & Schast, 2008), which assess perceived barriers to medication adherence (e.g., "Ran out/didn't fill," "Not feeling well"), reveal that adolescents with IBD report an average of 2.6-2.9 distinct barriers to adherence (Hommel, Denson, & Baldassano, 2011; Ingerski, Baldassano, Denson, & Hommel, 2010). Commonly reported barriers include forgetting (85%-94%; Gray, Denson, Baldassano, & Hommel, 2012; Hommel & Baldassano, 2010), not being home (43%-75%; Gray et al., 2012; Hommel & Baldassano, 2010), medication interfering with an activity (34%-69%; Gray et al., 2012; Hommel & Baldassano, 2010), regimen complexity (63%; Hommel & Baldassano, 2010), difficulty swallowing pills (56%; Hommel & Baldassano, 2010), and running out due to not refilling a prescription in time (15%; Gray et al., 2012). The time of day can also act as a barrier, and adolescents are least likely to miss a dose during lunchtime at school (10%) and dinner time (12%; Ingerski et al., 2010). Barriers to adherence have also been found to interact with other predictors. As noted above in Gray et al. (2012), anxiety/depressive symptoms moderate the barriers-adherence relationship, with high anxiety/depression adolescents reporting poorer adherence in relation to barriers, compared with low anxiety/depression adolescents. Barriers to adherence have also been shown to mediate the relationship between both attention problems and conduct problems with parent-reported adherence, underscoring the potential significance of behavioral functioning in identifying youth at-risk of nonadherence (Reed-Knight et al., 2013). Peer victimization is associated with lower adherence, and prosocial support moderates its impact (Janicke et al., 2009). As such, several publications have recommended the development of problem-solving skills among adolescents with IBD as a means of maneuvering barriers and optimizing adherence (Greenley, Gumidyala, et al., 2015; Hommel, Denson, et al., 2011; Hommel et al., 2012).

There are several factors that can influence the adherence of adolescents. Kim et al. (2016) compared the attitudes of adolescents aged 10-20 years and adults aged 30-50 years in a sample of 287 individuals with IBD. Although the overall risk of nonadherence due to age was unremarkable in this sample, the differences in attitudes between groups revealed attitudes that risk nonadherence. Compared with adults, adolescents were significantly more likely to endorse indifferent or skeptical attitudes toward medication and less likely to hold accepting or ambivalent attitudes. With regard to risk of nonadherence, indifferent or skeptical attitudes towards medication predicted nonadherence.

Knowledge about medication also affects adherence. Greenley, Stephens, et al. (2013) assessed the level of knowledge regarding nutritional supplements in IBD among 68 adolescent IBD patients. Comparing those who did not know and those who knew specifically how a multivitamin helped them, a large effect size was found on adherence with those having specific knowledge of the medication. Another consideration is whether the adherence is reported by the adolescent or parent, as Greenley et al. (2010) found that the relationship between adolescent and maternal reported adherence was not strong. Regarding intentional and unintentional nonadherence in adolescents, less research has been undertaken compared with adults (Schurman, Cushing, Carpenter, & Christenson, 2011). In a sample of 100 adolescent IBD patients, Schurman et al. (2011) found that the rate of unintentional nonadherence (73%) was higher than intentional (35%), and that the two kinds of nonadherence were not significantly correlated. Intentional nonadherence predicted 8%-10% additional variance in disease severity beyond age and gender alone.

The approach to risk factors for poor adherence differs between adults and adolescents. In adults, the emphasis is placed upon developing a positive and trusting physician-patient relationship, rather than demographic factors. Trust, regular contact, understanding the purpose of the medications, and integrating with an established routine are very important. Most frequently, when nonadherence occurs, it is unintentional and less-well integrated with the patient's routine. For adolescents, the terminology for unintentional nonadherence is barriers (e.g., Hommel & Baldassano, 2010). Assessing and adjusting barriers help to manifest support and better medication habits. As the literature thus far indicates, the best results are seen when adolescents understand and feel accepting toward their medications, and when their environment is conducive toward establishing and maintaining successful and lasting medication adherence habits.

Recent Diagnosis

Adult IBD Cohorts

For adults, there are mixed results to whether time since diagnosis impacts adherence rates. In a sample of 491 patients with IBD, D'inca et al. (2008) found that patients with a longer diagnosis (>=5 years; 15%) were more likely to be adherent than patients with a shorter diagnosis (<5 years; 24%). However, it should be noted that Cerveny, Bortlik, Kube na, et al. (2007) did not find a significant difference when comparing at 10 years or greater time, irrespective of Crohn disease, ulcerative colitis, or medication type.

Adolescent IBD Cohorts

In predicting prescription medication adherence, Reed-Knight, Lewis, and Blount (2011) found a weak negative influence of time since diagnosis in a block that explained 20% of the variance. Similarly, Kitney et al. (2009) found more adherent adolescents with IBD tended to have a shorter disease duration compared with nonadherent adolescents. In regard to specific medications, LeLeiko, Lobato, Hagin, McQuaid, et al. (2013) found that in IBD individuals prescribed 5-ASA medication, newly diagnosed individuals were more likely to be adherent than longer term diagnosed individuals (77% vs. 62%). However, this difference was not found in 6-MP medication users. Although these findings suggest a common theme toward greater adherence in newly diagnosed individuals, this finding may be masked by patient characteristics.

Greenley, Karazsia, et al. (2015) closely monitored the trajectory of adherence among 96 adolescents with IBD using an objective method of an electronic bottle cap that records the time and date of medication doses. Using group-based trajectory modelling, a two-group solution was established. Individuals who were high adherers maintained a trajectory of high adherence over time. In contrast, individuals who are mild nonadherers initially progressed to poorer adherence over time. As such, adherence tends to worsen over time only if a good and sustainable regimen is not established early.

The opposite trend is found in the literature between adults and adolescents in time since diagnosis. Adults can be slower to develop good medication adherence. In contrast, adolescents tend to become less adherent over time if they are not supported to establish good adherence habits early.

Impact of Nonadherence

Adult IBD Cohorts

Nonadherence to medication is associated with impaired health outcomes for individuals with IBD (Selinger et al., 2011). Comparing low adherers with medium/high adherers (N = 110), Trindade, Ehrlich, Kornbluth, and Ullman (2011) found that low adherers were significantly more likely to report higher disease activity on the Harvey Bradshaw Index (HBI). Low adherence to medication is also associated with up to a 3.5-fold increased risk of symptom relapse (Bhatt, Patil, Joshi, Abraham, & Desai, 2009; Feagins et al., 2014; Robinson, Hankins, Wiseman, & Jones, 2013; Tae et al., 2016). However, it should also be noted that among individuals who are adherent to the mesalazine medication, a change in regimen is associated with a 3.5-fold greater risk of symptom relapse than persisting with the same regimen (Robinson et al., 2013).

Adolescent IBD Cohorts

There is comparatively less research for adolescents with IBD. In a study of 100 parent/adolescent dyads, Schurman et al. (2011) found that volitional (i.e., intentional) nonadherence predicted an additional 10% (for parents and 8% for adolescents) of the variance in IBD disease severity beyond age and gender. The consequences of nonadherence to medication in both adults and adolescents is a worsening of IBD symptoms and risk of relapse. Consequently, consistent adherence to medication plays an important role in symptom control. A summary of the differences in adolescent and adult non-adherence-related evidence can be seen in Supplemental Digital Content 1 (available at http://links.lww.com/GNJ/A58).

Implications for Practice

There are multiple promising strategies that may enhance medication among IBD patients, and nurses are likely to play a central role in implementing these. As suggested by Greenley, Kunz, Walter, and Hommel (2013), educational approaches in particular are a promising strategy for enhancing adherence in adults and adolescents. Nurses are an invaluable conduit for improving patient knowledge about their condition and medication. The involvement of nurses in adherence promotion strategies may range from simple and informal recommendations during clinic visits (e.g., medication reminder strategies, pill boxes) to more complex, resource-intensive multicomponent interventions (e.g., Cook, Emiliozzi, El-Hajj, & McCabe, 2010; Cross et al., 2012).

There are clear benefits to patient adherence, with an additional 26% of individuals with chronic health conditions achieving a good health outcome when adherent (Dimatteo, Giordani, Lepper, & Croghan, 2002). For individuals with IBD specifically, nonadherence is associated with increased disease activity (Trindade et al., 2011), and an approximately threefold increased risk of relapse (Bhatt et al., 2009; Feagins et al., 2014; Tae et al., 2016). Atreja, Bellam, and Levy (2005) developed a simple model of medication adherence that uses the mnemonic "SIMPLE" to guide healthcare providers toward effective interventions that promote adherence. Based on this mnemonic, recommendations to enhance adherence are provided and split across adolescents and adult patient groups (see Table 1).

TABLE 1. Summary of the SIMPLE Mnemonic of Adherence Integrated With IBD Adherence Research

This review provides an IBD-specific evidence-based review of the literature on treatment adherence, with a further specific focus on practical adherence-promoting solutions across the life span for health providers working in this area. Supporting individuals with IBD to optimize their adherence has the potential to avoid costly flares/relapses that cause discomfort, time in in- or outpatient services, and hospital or insurance resources (Wan, Kozma, Slaton, Olson, & Feagan, 2014), which carries benefits for all. A key limitation of research on this subject is the heterogeneity of available data. For example, the measurement of adherence and the interpretation of adherence scores often vary between studies. Such inconsistencies make it difficult to compare findings across studies and reach comprehensive conclusions. Furthermore, there is considerably less evidence available for adolescents than for adults.

Future research should focus on areas such as barriers for adults and looking at intentional versus unintentional nonadherence in adolescents. This study identified that psychological disorders (e.g., anxiety/depression) can moderate the effect of barriers for adolescents (Gray et al., 2012), and that interventions that include a cognitive-behavioral therapy component helps adults (Cook et al., 2010; Elkjaer et al., 2010; Moshkovska et al., 2011). As such, future research could explore the interaction of thinking styles, psychological disorders, and coping strategies on medication adherence.

Conclusion

Nonadherence to medication for individuals with IBD is estimated to occur in 65%-90% of adolescents (Greenley et al., 2010, 2012; LeLeiko, Lobato, Hagin, Hayes, et al., 2013; LeLeiko, Lobato, Hagin, McQuaid, et al., 2013; Reed-Knight et al., 2013) and 55%-70% of adults (Jackson et al., 2010). Overall there has been less research with respect to adherence in adolescents compared with adults; however, there appears to be greater focus on unintentional nonadherence (barriers) in adolescents than in adults and more promising interventions. Although less is spent on medication when one is nonadherent, the cost of complications associated with nonadherence on the healthcare system is significantly higher (Wan et al., 2014).

Adherence is important for individuals with IBD as nonadherence is associated with increased disease activity (Trindade et al., 2011), and an approximately threefold increased risk of relapse (Bhatt et al., 2009; Feagins et al., 2014; Tae et al., 2016). Modifiable factors are considered the best target to help increase adherence (Selinger et al., 2011). IBD treatment has changed significantly with the widespread use of biological therapies, which may reduce patient burden due to their infrequent administration through intravenous infusions or self-injections. Nonetheless, research has reported suboptimal adherence rates (Lopez et al., 2013; Wentworth et al., 2018), especially for self-administered biologics (adalimumab, certolizumab). Furthermore, the practice of combined biological/immunomodulator therapy means that many patients undergoing biological therapies are subject to the same challenges and barriers surrounding daily medication adherence as those on nonbiological daily oral medication regimens. Adolescents (Greenley, Gumidyala, et al., 2015; Lenti & Selinger, 2017) and women considering pregnancy (de Lima, Zelinkova, Mulders, & van der Woude, 2016) are particularly responsive to intervention. Health providers supporting patients with IBD can use the "SIMPLE" mnemonic (Atreja et al., 2005), which promotes establishing good communication, information, simple regimen, overcoming barriers, focusing on the modifiable, and multimethod evaluation.

Acknowledgments

The authors thank Mr Stephan Moller, Miss Sarina Cook, and Dr Sally Woodhouse for their editorial assistance in bringing this paper to fruition.

REFERENCES