Abstract
Objective:
The aim of this study was to investigate medication persistence in Turkish children and adolescents with attention-deficit/hyperactivity disorder (ADHD). The effects of sociodemographic characteristics, symptom severity of ADHD, comorbidity, and treatment-related factors influencing medication persistence in children diagnosed with ADHD were studied.
Methods:
Medication persistence over a continuous 12 month period was evaluated for 877 children and adolescents between 6 and 18 years of age, who were diagnosed with ADHD for the first time and started to receive medication. Medication persistence was determined according to whether or not taking the prescribed medication continued for 12 months after the initiation of treatment. Whereas the symptom severity of ADHD was assessed by using the Turgay Diagnostic and Statistical Manual of Mental Disorders, 4th ed. (DSM-IV)-based Child and Adolescent Behaviour Disorders Screening and Rating Scale-Parents Form (T-DSM-IV-S), perceived medication efficacy after the first treatment was evaluated by the Clinical Global Impressions-Improvement Scale (CGI-I).
Results:
In this study, medication persistence over a continuous 12 month period occurred at a rate of 30.2% (n=265) in the subjects studied. The hierarchical regression analysis conducted in this research revealed that younger age, higher hyperactivity/impulsivity symptom severity, use of long-acting methylphenidate, addition of another ADHD medication, addition of other psychotropic medications, absence of side effects, and perceived medication efficacy were associated with successful medication persistence over a continuous 12 month period.
Conclusions:
Understanding the factors that affect medication persistence in ADHD may improve treatment efficacy and symptom control, while minimizing future risks.
Introduction
A
ADHD should be considered a chronic disorder, and the treatment plan should consist of psychopharmacological and/or behavioral therapy (Pliszka 2007). It is stated that stimulant and nonstimulant medications are effective in ADHD treatment when used separately, together, or with behavioral therapies (Pliszka 2007). However, these medications can only improve symptoms when children and their parents adhere to and comply with the prescribed regimen (Barner et al. 2011). Even though it is known that not taking medication at the required time and dosage adversely affects the course of ADHD, medication persistence reportedly ranges between 36% and 86.8% in children with ADHD, and discontinuation of treatment is common (Adler and Nierenberg 2010).
Some influential factors were noted to improve medication persistence: younger age (Miller et al. 2004; Sanchez et al. 2005; Faraone et al. 2007; Bokhari et al. 2008), male gender (Miller et al. 2004; van den Ban et al. 2010), not living with parents (Atzori et al. 2009; Palli et al. 2012), lower symptom severity of ADHD (Thiruchelvam et al. 2001; Faraone et al. 2007), absence of comorbid mental disorders (Toomey et al. 2012; Bahmanyar et al. 2013), receipt of initial prescription from a psychiatrist (Miller et al. 2004), absence of side effects (Toomey et al. 2012), addition of another psychotropic medication (Palli et al. 2012), type and efficacy of medication (Marcus et al. 2005; Sanchez et al. 2005; Christensen et al. 2010; Toomey et al. 2012). Poor family functionality and concerns regarding stigmatization (Hamrin et al. 2010) are reported to reduce medication persistence.
Although medication persistence in ADHD patients is a relatively frequently studied field, assessment methods differ among studies. Moreover, medication persistence is affected by local factors peculiar to every country, such as cultural characteristics, differing health systems of various countries, and varying maintanance of records among health systems. The aim of this study was to investigate medication persistence in Turkish children and adolescents with ADHD. The sociodemographic characteristics, symptom severity of ADHD, comorbidity, and treatment-related factors influencing medication persistence over a continuous 12 month period were studied.
Methods
In order to conduct the study, permission was obtained from the Ethics Committee of Sakarya University, Faculty of Medicine with no 71522473.050.01.04/57.
Study population
The study population consisted of 877 children and adolesncents between 6 and 18 years of age, visiting the child psychiatry outpatient clinic of Sakarya University Training and Research Hospital between January 10, 2009 and January 10, 2012, who were diagnosed with ADHD for the first time and met the inclusion criteria of the study.
The inclusion criteria for subjects in the study were determined as follows. Medication for ADHD was first administered to the patient at least 12 months prior to the time of the study, and the patient was evaluated in terms of efficacy and side effects of treatment after the initiation of medication. Subjects who had received treatment for ADHD before, who received ADHD treatment from another clinic after initiation of medication, who were >18 years of age at the evaluation date, and whose treatment was discontinued by the clinicians were not included in the study. Subjects whose file data were not sufficient or whose evaluation scores were not calculated appropriately were excluded from the study. Subjects who were adopted, living in an institution, or were not living with their parents were not included in the study in order to compose a homogeneous group in terms of familial factors. Subjects diagnosed with mental retardation (determined via clinical evaluation or intelligence tests), pervasive developmental disorders, psychosis, or substance addiction or abuse were not included in the study. Because other medications can interfere with the medication treatment of ADHD, or ADHD medications may be discontinued by clinicians during the course of bipolar disorder, subjects with bipolar disorder were excluded from the study. Between January 10, 2009 and January 10, 2012, 2171 children and adolescents received medication for ADHD at the child psychiatry outpatient clinic of the Sakarya University Training and Research Hospital. Of these 2171, 425 had previously received medication for ADHD, and 398 were ineligible because of the exclusion criteria. Of the 1348 who met the study criteria, 195 were excluded from the study because their families were not reached by phone. A total of 276 for whom medication was switched by clinicians were excluded from the study because of the lack of sufficient data about the treatment efficacy and side effects after each medication switch.
Medication persistence
Medication persistence was determined according to whether or not taking the prescribed medication continued for 12 months after the initiation of treatment. Because the length of the school year is 9 months and the summer holiday is 3 months in Turkey, a treatment gap >3 months was defined as treatment discontinuity.
Assessment measures
The symptom severity of ADHD was assessed Turgay Diagnostic and Statistical Manual of Mental Disorders, 4th ed. DSM-IV-based Child and Adolescent Behaviour Disorders Screening and Rating Scale-Parents Form (T-DSM-IV-S). The T-DSM-IV-S, developed by Turgay (1994) based on DSM-IV diagnostic criteria, includes 9 items regarding attention deficit, 6 items regarding hyperactivity, 3 items regarding impulsivity, 8 items regarding oppositional defiant disorder, and 15 items regarding conduct disorder. The symptoms were scored by assigning a severity estimate for each symptom on a four point Likert-type scale (0=none, 1=little, 2=much, and 3=very much).
Clinical Global Impressions-Improvement Scale (CGI-I) was used to assess the improvement in global symptoms (Guy 1976). Scores ranged from one (very much improved) through seven (very much worse).
Procedures
In Turkey, every child <18 years of age has health insurance compensated by the state. Only child and adolescent psychiatrists can prescribe an ADHD medication. If there is no child and adolescent psychiatrist in a state, only a psychiatrist is allowed to treat ADHD in children. Appointments to the child psychiatry clinic were made via telephone or Internet from the central system, on the request of the parents. The subjects, having been diagnosed and monitored, had an appointment before each interview.
In our child psychiatry outpatient clinic, parents of all children are required to complete a sociodemographic form before the first interview. The sociodemographic form is filed by the child psychiatrist after checking for deficiencies. In subjects with a prediagnosis of ADHD, one of their parents is required to complete the T-DSM-IV-S after the first interview in order to support the diagnosis. Diagnoses of ADHD and comorbid mental disorders were determined according to the DSM-IV (American Psychiatric Association, 1994). In the treatment of ADHD, short-acting methylphenidate, long-acting methylphenidate (OROS methylphenidate), and atomoxetine are used in Turkey. After the initiation of medication, efficacy of the treatment is recorded by asking questions of the parents in accordance with the CGI-I scale during the first interview (within 1 month after the treatment). Information about the side effects of the drugs was elicited spontaneously by parental report. During every assessment of the subjects, clinicians recorded medication usage in each subject's file. The evaluation notes of the clinician are retained in files in our clinic. Information about medication persistence and the medication gap was obtained from the paper files that were systematically held by the clinicians. Initially, in the formation of these records, medication usage was obtained simply by parental reports. Subsequently, electronic records of the prescriptions were reviewed.
After permission was obtained from the ethics committee for this research, sociodemographic characteristics and clinical data on the subjects were collected retrospectively from the files in our institution, using a form prepared by the researchers. The families were called by phone, and oral permission regarding participation in the study was acquired. Moreover, they were questioned concerning whether or not the follow-up visit to a diagnosis of ADHD was continued in another clinic, or if medication was discontinued. Reasons for medication discontinuation were questioned using a semistructured interview developed by the clinicians.
The educational levels of the parents were rated as being low (illiterate, primary or secondary school graduate) or high (high school or college graduate). The relationship between parents was rated as togetherness of the biological parents or separation of the biological parents (parents divorced or one of the parents deceased). Data related to a history of mental disorders in the family were collected in consideration of the existence of a diagnosis in the family established by a psychiatrist or child psychiatrist.
Statistical analysis
SPSS v.17.0 for Windows (SPSS, Inc. Chicago, IL) was used to analyze the data. Descriptive statistics were conducted for means and standard deviations of continuous variables and percentages of categorical variables. Hierarchical logistical regression analysis was performed to identify predictors associated with medication persistence. At the sociodemographic level, explanatory variables included age, gender, presence of chronic health conditions, low educational level of the mother, low educational level of the father, togetherness of parents, and a history of psychiatric disorders in the family. At the clinical level, the explanatory variables included T-DSM-IV-S inattention, hyperactivity/impulsivity, oppositional defiant disorder, conduct disorders subscale scores, and presence of at least one comorbid mental disorder. The treatment-related variables included type of medications, addition of another ADHD medication, addition of other psychotropic medications, presence of side effects with first drug use, and the CGI-I score. The results were presented as an odds ratio with a confidence interval of 95%. The level of statistical significance was accepted as p≤0.05.
Results
Evaluation of the data revealed that 30.2% (n=265) of children and adolescents with ADHD continued to use medication 12 months after the initiation of the prescription, and 69.8% (n=612) discontinued the medication. Of the subjects whose medication was switched (n=276), 36.6% (n=101) continued ADHD medication 12 months after the initiation of the prescription. Table 1 illustrates sociodemographic characteristics of children who continued and discontinued treatment, and Table 2 illustrates clinical and treatment-related characteristics.
ADHD, attention-deficit/hyperactivity disorder.
ADHD, Attention-deficit/hyperactivity disorder; T-DSM-IV-S, Turgay Diagnostic and Statistical Manual of Mental Disorders, 4th ed. (DSM-IV)-based Child and Adolescent Behavior Disorders Screening and Rating Scale-Parents Form; CGI-I, Clinical Global Impressions-Improvement Scale.
As a result of hierarchical regression analysis, it was determined that younger age, higher T-DSM-IV-S hyperactivity/impulsivity score, addition of another ADHD medication, addition of other psychotropic medications, absence of side effects, and lower CGI-I scores were associated with successful medication persistence over a continuous 12-month period. Medication persistence was greater for long-acting methylphenidate than for atomoxetine and short-acting metylphenidate. No statistically significant relationship was determined between medication persistence and gender; chronic health condition; educational levels of parents; togetherness of parents; mental disorder history in the family; at least one mental disorder comorbidity; or T-DSM-IV-S inattention, oppositional defiant disorder, or conduct disorder scores. Table 3 summarizes the results of hierarchical regression analysis performed to determine predictors of medication persistence over a 12 month period.
Model χ2=173.096, p<0.001.
ADHD, attention-deficit/hyperactivity disorder; T-DSM-IV-S, Turgay Diagnostic and Statistical Manual of Mental Disorders, 4th ed. (DSM-IV)-based Child and Adolescent Behavior Disorders Screening and Rating Scale-Parents Form; CGI-I, Clinical Global Impressions-Improvement Scale.
In the cases in which treatment was discontinued (n=612), medication persistence failed because of the inability to obtain an appointment, or other health system problems (19.4%, n=119); recovery from symptoms (17.3%, n=106); side effects of medications (12.3%, n=75); ineffectiveness of medications (12.1%, n=74); parents who did not want to give medication to their children (11.3%, n=69); family problems (8.3%; n=51); children who did not want to use medication or attend follow-up sessions (7.8%, n=48); dissatisfaction with the health service (4.1%, n=25); and other reasons (7.4%, n=45).
Discussion
This study investigated medication persistence over continuous 12-month periods of time in Turkish children and adolescents with ADHD. The effects of sociodemographic characteristics, ADHD symptom severity, comorbidity, and treatment-related factors were examined.
Reportedly, establishing an ADHD diagnosis at younger ages increases medication persistence (Miller et al. 2004; Sanchez et al. 2005; Gau et al. 2006; Faraone et al. 2007; Bokhari et al. 2008; Palli et al. 2012; Bahmanyar et al. 2013). The findings related to the age of onset for treatment in this study support other studies conducted in this field. Studies conducted with respect to medication persistence in ADHD have obtained contradictory results in terms of the role of gender. Some studies have reported that medication persistence is higher in boys (Miller et al. 2004; van den Ban et al. 2010), and some report that it is higher in girls (Barner et al. 2011). However, in most studies conducted in this field, no difference was found between genders (Barbaresi et al. 2006; Faraone et al. 2007; Gau et al. 2008; Toomey et al. 2012) consistent with the findings of this study related to gender. The literature has limited information regarding the effects of chronic health conditions on medication persistence in children with ADHD. Consistent with the study conducted by Toomey et al. (2012), this study determined that no relationship existed between medication persistence and chronic health conditions in children and adolescents.
In the treatment of ADHD, one of the most important factors affecting treatment and medication persistence is symptom severity and comorbid behavioral problems. Because the majority of the studies conducted on medication persistence in ADHD rely on health system records, the knowledge about the relationship between symptom severity and medication persistence is limited. Higher baseline symptom severity reportedly increases treatment persistence (Thiruchelvam et al. 2001; Faraone et al. 2007). Studies also exist that did not find any relationship between ADHD symptom severity and discontinuation of treatment (Toomey et al. 2012), or determined that treatment persistence decreased as the symptom severity increased (Gau et al. 2008). Specifically, this study revealed that although no relationship was determined to exist among medication persistence and inattention, oppositional defiant disorder, and the conduct disorder scores of parental assessments, higher hyperactivity/impulsivity scores increased medication persistence. Ibrahim (2002) reported that an excess of externalizing behavioral problems perceived by parents increased the willingness of parents to ensure medication persistence. However, in some studies, it was determined that oppositional defiance symptoms adversely affected medication persistence (Thiruchelvam et al. 2001). In this study, no relationship was found between medication persistence and scores for oppositional defiant disorder and conduct disorders. Differences in assessment scales, study patterns, and average ages of subjects may explain varying results in studies.
According to Bahmanyar et al. (2013) and Toomey et al. (2012), being diagnosed with another mental disorder constitutes a risk in terms of medication persistence. A study by Atzori et al. (2009) did not find any relationship between being diagnosed with at least one mental disorder and treatment continuity. In this research, diagnosis of a comorbid mental disorder in children with ADHD was determined not to be related to medication persistence. However, the fact that comorbid diagnoses were not established by structured interviews in this study may affect the results in terms of a low rate of mental disorder diagnoses, such as mood and anxiety disorders, which frequently accompany ADHD. In the future, findings in this field should be supported with diagnoses determined by structured interviews and follow-up studies.
In this study, medication persistence was greater for the use of long-acting methylphenidate than for the use of atomoxetine and short-acting methylphenidate. In some studies, medication persistence were also determined to be greater for long-acting methylphenidate (Marcus et al. 2005; Christensen et al. 2010; Lawson et al. 2012; Palli et al. 2012). Christensen et al. (2010) reported that using stimulants was associated with medication persistence to a greater degree than using atomoxetine; however, Barner et al. (2011) determined that using atomoxetine improved medication persistence. The discrepancy may be explained by the fact that different forms of stimulant drugs were used in these studies. In our study, using comorbid ADHD medications and other psychotropic medication increased persistence. Multiple medications prescribed for comorbid medical conditions were found to increase persistence in children with ADHD (Barner et al. 2011; Palli et al. 2012). However, the fact that some comorbid medication treatments were added to the treatment regimens after a long follow-up time, and that a low rate of occurrence of comorbid medications prescribed to subjects in this study existed, decreased the reliability of such findings.
Hamrin et al. (2010) reported that concerns about side effects of medications, the presence of side effects on using medication, the continuation of symptoms after the initiation of medication, or an insufficient level of efficacy perceived by the parents adversely affected medication persistence, and caused discontinuation of medications. In this study, it was determined that 12.3% of subjects who discontinued treatment terminated the follow-up the side effects of treatment; whereas 12.1% terminated the follow-up because of inefficacy of the medication. Conversely, the results of this research support previous studies conducted on efficacy as perceived by the parents (Thiruchelvam et al. 2001; Gau et al. 2006; Faraone et al. 2007; Toomey et al. 2012) and the presence of side effects (Gau et al. 2008; Toomey et al. 2012) in medication persistence.
This study revealed that the educational levels of parents, a history of mental disorders in the family, and the togetherness of the parents did not affect medication persistence. Similarly, in some studies, no relationship was found among the income of the family or the educational level of the parents and medication persistence (Gau et al. 2006; Toomey et al. 2012).
In ADHD treatment, medication persistence is important to symptom control and prevention of future risks (Biederman et al. 2009). In this study, 30.2% of cases diagnosed with ADHD were observed to exhibit continued medication persistence after 12 months. This rate was lower than the results of other studies on medication persistence (Faraone et al. 2007; Winterstein et al. 2008; Atzori et al. 2009; Adler and Nierenberg 2010). The differences among health systems and health system records in various countries may result in significant differences in the results of studies. In this study, the inability to obtain an appointment or other health system problems was the most common reason for medication discontinuity. This might negatively affect the rate of medication persistence.
Limitations
This study has certain limitations resulting from its retrospective approach. Therefore, effects of medication persistence on symptom control could not be evaluated. In this study, diagnosis of ADHD and comorbid mental disorders were determined by means of clinical evaluation. Diagnoses of certain mental disorders, such as the depression and anxiety disorders frequently accompanying ADHD, were found at a lower rate than expected because of the non-application of structured evaluations, and the effect of each individual comorbid mental disorder on medication persistence could not be evaluated. Moreover, another limitation of this study was the determination of medication persistence by means of information obtained from paper files that were supported by phone interviews. The data led to the conclusion that in ADHD treatment, changes in medications may affect medication persistence. Because of the lack of sufficient data about the treatment efficacy and side effects after each medication was switched, subjects whose medication was changed by clinicians could not be evaluated. Obtaining information about the side effects spontaneously by parental reports may also cause some side effects to be overlooked.
Conclusions
This study revealed that medication persistence in Turkish children with ADHD may be affected by younger age, higher hyperactivity/impulsivity scores, use of long-acting methylphenidate, addition of another ADHD medication and addition of other psychotropic medications, the absence of side effects, and the efficacy level perceived by the parents.
Clinical Significance
Medication persistence in children and adolescents with ADHD is a significant concern. The findings in this study regarding the type of medication may help to influence the choice of medication in clinical practice. This study revealed that combined medications increased medication persistence in children and adolescents with ADHD, suggesting that combined medications may improve treatment efficacy and symptom control. The study findings may also contribute to the comprehension of the reasons and risk factors for discontinuation of medication in ADHD. Additionally, this research emphasized that crucial modifications in the health system are needed for improving medication persistence.
Footnotes
Acknowledgment
We thank Dr. Ümit Tural for help with the statistical analysis.
Disclosures
No competing financial interests exist.