 |
 |
|
|
|
Major depressive disorder (MDD) in adolescents is recognized as a serious psychiatric illness with extensive acute and chronic morbidity and mortality. The period prevalence rate is estimated to reach 6% to 8% and is characterized by a recurrence rate of 60% to 80% by the end of adolescence (1). Antidepressant medications have become an important component in managing adolescent depression. Double-blind trials of tricyclic antidepressants have failed to show any greater efficacy than placebo, and concerns have been raised about the risks of using these treatments in adolescents, who appear to be particularly sensitive to the side effects of these medications (2,3). Newer antidepressants, such as the selective serotonin reuptake inhibitors (SSRIs), have a lower side effect burden and may potentially be better tolerated in the adolescent population, and therefore more effective, than the older medications (4). Several double-blind, placebo-controlled trials have studied the efficacy of SSRIs in adolescent depression (5–7). The largest study involved 275 patients and demonstrated significantly higher remission rates for subjects in the paroxetine group, compared with imipramine or placebo (7). Much of the research in child and adolescent depression has focused on the effectiveness and efficacy of antidepressants (8). Less attention has been paid to side effects and their impact on compliance and quality of life for these patients. To date, no study has specifically examined these issues in this age group. Therefore, we studied the impact of adverse effects from antidepressants on quality of life to better inform clinicians and researchers on the consequences of prescribing antidepressant treatments and to potentially aid in the development of new antidepressants. Our current understanding of the impact of potential anti- depressant side effects on adolescents with depression is limited to studies that only examine the frequency and severity of adverse effects. Several studies using SSRIs have reported the frequency of side effects in adolescent patients (7,9,10). However, reporting the frequency of occurrence alone does not truely represent the impact of side effects, as was clearly illustrated by Boulos and others, who evaluated side effects in 15 adolescents and young adults with depression treated with fluoxetine (9). The most common side effects were dry mouth, nausea, sweating, tremor, and decreased appetite. However, the most severe side effects were not the most frequent. For example, restlessness occurred in only 27% of the subjects, but all subjects ranked it as severe, compared with tremor. The latter side effect occurred in 40% of the subjects but was ranked severe by only 1 out of 6 subjects. These data illustrate the importance of measuring the impact, rather than just the frequency, of side effects. Understanding the impact of side effects is further complicated by the role that age and history of depression may play in altering a person’s subjective experience. For example, if a side effect resembles a symptom of depression, patients with a history of depression may report greater impact. There are no data comparing assessments of the impact of side effects made by adolescent patients suffering from depression with assessments from healthy adolescents or from adults who also have a history of depression. Finally, although clinicians who treat adolescents with depression regularly assess the impact of side effects, it is not known whether physician impressions of the impact of side effects are similar to adolescent perceptions. Therefore, this study evaluates the subjective experience of antidepressant side effects in adolescents with a history of depression and compares it with the perceptions of healthy adolescents and of adults with a history of depression, and also with the judgement of the clinicians treating adolescent patients. MethodsQ-sort Methodology Producing the Q-sort The number of Q-cards was limited to 40, because with 40 Q-cards force ranked into 7 slots, the Q-sort best simulated a normal distribution. One author developed the model for the Q-sort: each Q-card is force ranked into 1 of 7 slots with semantic tags that explain the ranking. On one end is the side effect with the most impact (1 card), labelled as “–3”; on the other is the side effect with the least impact (1 card), labelled as “+3.” Side effects with neutral impact (labelled as “0”) are placed in the centre slot. The number of cards placed in each slot is designed to best approximate a quasi-normal distribution. Subjects
Procedures Once subjects were identified and found eligible for the study, the research coordinator obtained informed consent. At the initial appointment, each subject (excluding clinician subjects) completed a demographics questionnaire, the Slosson Oral Reading Test (SORT) (15), the SCID-IV, and the HDRS. Subjects were then asked to complete the Q-sort. Initially, they were asked to review all 40 Q-cards and divide them into 3 piles: those with most impact, those with least impact, and those in the neutral category. The subjects were then asked to choose Q-cards for each slot, alternating between the negative and positive slots. Each subject was asked to start by choosing the Q-card with the most impact followed by the Q-card with the least impact. They then repeated the exercise, selecting the next 4 Q-cards with the most impact. This procedure continued until 12 cards remained from the original 40 cards, which were placed in the neutral (centre) slot. After completing the Q-sort, the subjects completed the Scale for Medication Compliance (SMC) and the Visual Analogue Scale for Side Effect–Related Health Status (VAS Health Status). The SMC was developed for this study to measure the level of compliance for any given side effect. Subjects were asked to rate their level of compliance if they were to experience daily the side effect that they rated as having the most impact on their lives. The scale used a 10-cm line with points marked 0%, 10%, 20%, and so on, up to 100%. Subjects could mark any place on the line. Their response mark was measured according to the distance from 0%; that is, if the response was at 23 mm from 0%, the score was recorded as 23%. The VAS Health Status scale was also developed for this study to evaluate how subjects viewed their health status if given a hypothetical side effect. Subjects were asked to rate how they would describe their life if they were to experience daily the side effect they ranked as having the most impact, the least impact, or a neutral impact. The scale used a 9-cm line similar to the SMC, with 10 points marked from 1 (“near death”) to 10 (“perfect health”); subjects were asked to choose a point that reflected their side effect–related health status. Because in this case the left anchor point was set at 1 (“near death”), a response at 23 mm from the left anchor was recorded as 3.3. Statistical Analyses For the secondary analyses, we used the t-statistic to compare the mean ratings for compliance and side effect–related health status. Finally, we used Fisher’s exact test to evaluate the differences in the reasons given by each group for choosing the side effect with the most impact. ResultsSubjects Ranking of Side Effects The mean rankings for the top 5 side effects were numerically similar across all groups, with the only statistically significant differences listed above. We also compared the mean scores of the adolescents with depression and the other groups for any side effect that had a numerical difference greater than 1. This numerical difference would represent a mean difference of 1 category or greater in the Q-sort. Sexual dysfunction was the only side effect, other than anxiety, sleepiness, and hair loss, that had a numerical difference greater than 1 in mean scores. This numerical difference was found between adolescents with depression (mean 0.14, SD 1.36) and clinicians (mean –1.11, SD 0.96; t = 3.28, df 38, P = 0.002). Table 2 lists the 5 side effects ranked as having the most impact by each of the groups. All 4 subject groups ranked syncope and vomiting as 1 of the 5 side effects with the most impact. Adolescents with depression ranked vomiting, syncope, confusion, hair loss, and anxiety as having the most impact.
Compliance With Medications Judgement of Health Status
Reasons for Rating a Side Effect as Having the Most Impact Paired comparisons showed significant differences between the proportion of adolescents with depression (14%) and healthy adolescents (56%, Fisher’s exact P = 0.03) who gave “medical concerns” as the reason for rating a side effect as having the most impact on their daily lives (Table 4).
DiscussionIn this study, we found both similarities and differences between adolescents with depression and the other 3 groups in the profile of opinions regarding the impact of side effects. The mean rankings for 2 side effects—vomiting and syncope—were homogeneous: all 4 groups ranked these as 1 of the 5 side effects with the most impact. Given this similarity across all groups, health professionals need to address these 2 side effects when prescribing medications, and researchers need to consider their impact when developing new drugs. However, the mean ratings of side effects that adolescents ranked as having the most impact on their lives also showed significant differences, especially in regard to hair loss and anxiety. Hair loss had less negative impact on adults with depression, compared with adolescents with depression. This result may represent less concern on the part of adults regarding their physical appearance. Adolescents with depression rated anxiety as having more impact, compared with clinicians and healthy adolescents. One explanation for this finding is that the group with a history of depression has experienced anxiety as part of their illness. Therefore, this side effect may signify to them the onset of a new episode of depression. In contrast, clinicians may view anxiety as an easily managed side effect of a medication, and healthy adolescents may view anxiety as the feeling they might experience when writing an exam or giving an oral report. Of interest is the difference in clinicians’ ability to judge the impact of sexual side effects. Clinicians in this study rated sexual dysfunction as having more impact than did adolescents with depression. In fact, for adolescents with depression, the mean ranking for sexual dysfunction was in the positive range, whereas for clinicians, it was in the negative range. Adolescents may not see this side effect as having significant impact because they are not sexually active, or perhaps the impact of this side effect in this population is not as important as clinicians may anticipate. Alternatively, clinicians may base their opinion on real-world experience, whereas adolescents with depression are asked a hypothetical question. Compared with adolescents with depression, healthy adolescents were more likely to give medical concerns as the reason for ranking a side effect as having the most impact on their daily lives. Adolescents with depression were more concerned with functional impairments from side effects. One possible explanation for this observation may be that healthy adolescents have experienced few psychiatric symptoms and are generally healthier. Healthy adolescents may be less aware of, and therefore less concerned about, the impact of functional impairments from both physical and emotional perspectives. Compared with adolescents with depression, adults more frequently listed functional impairment as the reason for ranking a side effect as having the most impact. This may represent the more severe consequences for adults of impaired ability to function in daily life, compared with adolescents. For example, if adult patients have difficulties functioning in their jobs, this will likely lead to significant financial difficulties. In contrast, adolescents with similar impairments at school will still be supported by their parents or guardians. Clinicians’ reasons for rating a side effect as having the most impact were divided between medical concerns and functional impairment. Ideally, this may represent the clinicians’ need to understand the impact of side effects on all aspects of patient well-being, both functional and medical. Q-sort methodology is used to measure and compare opinions among groups. It was developed to measure subjectivity and yet allow researchers to use more traditional statistical methods in their analyses. Because it is a measure of subjectivity, there are no built-in validators for data collected with this methodology. For example, subjects are limited by the force rank, and therefore, when a side effect is rated as negative, it may simply represent the limitations of the force rank rather than a true negatively valued opinion. However, when we examined the mean VAS Health Status ratings for all subjects, we found significant differences in the mean ratings for the side effect ranked as having the most impact, the least impact, and a neutral impact. This finding likely represents a meaningful variability in the perceived impact of the side effects ranked in the 3 categories. Further confirming this interpretation is the observation that the total mean rating for VAS Health Status for the side effect ranked as having the most impact was similar to the health status ratings from other research examining the impact of severe depression (16). In a study by Schaffer and others, the mean health utility score of subjects with severe depression was 0.31 (with 0 being “near death” and 1 being “perfect health”). This finding is very similar to the mean health status score for the side effect ranked as having the most impact (mean 3.2) from this study. Therefore, the impact of side effects could have as much impact on patients’ daily lives as severe depression itself. Finally, compliance with medications is a major problem in clinical practice. The differences in mean ratings for compliance in the 3 planned comparisons did not reach statistical significance, but there were numerical differences. Clinicians judged the compliance rates of adolescent patients to be lower than the rates reported by the adolescents themselves. There are several possible explanations for these differences: 1) clinicians may misjudge their patients’ level of tolerance for side effects; 2) not wanting to be perceived as “bad” patients, adolescents may give an inflated rate of compliance owing to social desirability; and 3) the question was hypothetical and therefore may not be translated into the real world. This study has several limitations. First, the sample was small (n = 82). However, the standard number per subject group for Q-methodology has been approximately 20 per group (11). Second, we asked clinicians to give general information on adolescent patients rather than information on a specific individual under their care. In other words, we did not study clinician–patient pairs, a direct evaluation that might have provided a more comprehensive indication of the agreement between clinicians and their patients with regard to the impact of side effects. Finally, Q-sort methodology involves hypothetical situations. It is possible that adolescents might rank the impact of side effects differently if they were experiencing them at the time of ranking. Notwithstanding this concern, there are no other methods to evaluate the impact of side effects that patients may not have experienced. The current Q-sort is a reasonable approach to quantifying subjectivity in the context of a potentially hypothetical situation. This study examined the impact of side effects on adolescents with depression, compared with adults with depression, healthy adolescents, and the judgement of clinicians. We found both similarities and differences in the groups. Further research in this area might address the reliability of clinicians’ judgement for specific patients and the reliability of parents’ judgement regarding the impact of side effects (given that parents frequently act as medical informants for their adolescent children). Researchers might also consider directly measuring the impact (and not just frequency) of side effects in clinical trials of antidepressants in adolescents with depression. Funding and SupportThis project was supported by The Physicians’ Services Incorporated Foundation Grant #99-09R. AcknowledgementsThe authors thank Dr M Sanford for his assistance in recruiting subjects, C MacDonald for her research assistance, and M Katie and E Duke for their statistical support. References1. American Academy of Child and Adolescent Psychiatry. Practice parameters for the assessment and treatment of children and adolescents with depressive disorders. J Am Acad Child Adolesc Psychiatry 1998;37(10 Suppl):63S–83S. 2. Geller B, Reising D, Leonard HL, Riddle MA, Walsh BT. Critical review of tricyclic antidepressant use in children and adolescents. J Am Acad Child Adolesc Psychiatry 1999;38:513–6. 3. Hazell P, O’Connell D, Heathcote D, Henry D. Efficacy of tricyclic drugs for treating children and adolescents depression: a meta analysis. BMJ 1995;310:897–901. 4. Emslie GJ, Walkup JT, Pliszka SR, Ernst M. Nontricyclic antidepressants: current trends in children and adolescents. J Am Acad Child Adolesc Psychiatry 1999;38:517–28. 5. Simeon JG, Dinicola VF, Ferguson HB, Copping W. Adolescent depression: a placebo-controlled fluoxetine treatment study and follow-up. Prog Neuropsychopharmacol Biol Psychiatry 1990;14:791–5. 6. Emslie GJ, Rush AJ, Weinberg WA, Kowatch RA, Hughes CW, Carmody T, and others. A double-blind, randomized, placebo-controlled trial of fluoxetine in children and adolescents with depression. Arch Gen Psychiatry 1997;54:1031–7. 7. Keller MB, Ryan ND, Strober M, Klein RG, Kutcher SP, Birmaher B, and others. Efficacy of paroxetine in the treatment of adolescent major depression: a randomized, controlled trial. J Am Acad Child Adolesc Psychiatry 2001;40:762–72. 8. Ambrosini PJ. A review of pharmacotherapy of major depression in children and adolescents. Psychiatr Serv 2000;51:627–33. 9. Boulos C, Kutcher S, Bardner D, Young E. An open naturalistic trial of fluoxetine in adolescents and young adults with treatment-resistant major depression. J Child Adolesc Psychopharmacol 1992;2:103–11. 10. Colle L, Belair J, DiFeo M, Weiss J, LaRoche C. Extended open-label fluoxetine treatment of adolescents with major depression. J Child Adolesc Psychopharmacol 1994;4:225–32. 11. McKeown B, Thomas D. Q methodology. In: Quantitative applications in the social sciences. No. 66. Thousand Oaks (CA): Sage Publications; 1988. p 83. 12. Day JC, Bentall RP, Warner S. Schizophrenic patients’ experiences of neuroleptic medication: a Q-methodology investigation. Acta Psychiatr Scand 1996;93:397–402. 13. Hamilton M. A rating scale for depression. J Neurol Neurosurg Psychiatry 1960;14:203–12. 14. First MB, Spitzer RL, Gibbon M, Williams JB. Structured Clinical Interview for Axis I DSM-IV Disorders Patient Edition (SCID-I/P. Version 2). New York: Psychiatric Research Institute; 1994. 15. Slosson RL. Slosson Oral Reading Test (SORT). East Aurora (NY): Slosson Educational Publications; 1963. 16. Schaffer A, Levitt AJ, Hershkop SK, Oh P, MacDonald C, Lantcot K. Utility scores of symptom profiles in major depression. Psychiatric Research 2002;110:189–97. AuthorsManuscript received November 2002, revised, and accepted May 2003. Preliminary data previously presented at the Canadian Psychiatric Association Conference; 1999; Toronto (ON). 1. Fellow, Department of Psychiatry, Sunnybrook and Women’s College Health Sciences Centre, University of Toronto, Toronto, Ontario. 2. Psychiatrist-in-Chief, Department of Psychiatry, Sunnybrook and Women’s College Health Sciences Centre, University of Toronto, Toronto, Ontario. 3. Director, Department of Research Design and Biostatistics, Sunnybrook and Women’s College Health Sciences Centre, University of Toronto, Toronto, Ontario. Address for correspondence: Dr AH Cheung, Sunnybrook and Women’s College Health Sciences Centre, Department of Psychiatry, 2075 Bayview Avenue, Toronto, ON M4N 3M5 e-mail: amy.cheung@sw.ca
1 | 2
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||