SEARCH

SEARCH BY CITATION

Keywords:

  • Juvenile idiopathic arthritis;
  • Caregivers;
  • Treatment behaviors;
  • Parent Adherence Report Questionnaire

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. AUTHOR CONTRIBUTIONS
  8. Acknowledgements
  9. REFERENCES

Objective

Health care providers in juvenile idiopathic arthritis (JIA) might refer to caregivers' self-report of children's treatment-related behaviors to assist in clinical decisions. However, caregivers may believe that they are adhering to treatment even though they have a different understanding of recommendations than that intended by the medical team. We examined whether caregiver recall of children's JIA treatment matched actual recommendations at baseline and 3, 6, 9, and 12 months.

Methods

A total of 235 primary caregivers were recruited from rheumatology clinics at 2 pediatric university–based teaching hospitals in Canada. Using the Parent Adherence Report Questionnaire, caregivers indicated whether their child was prescribed medications and/or exercises. Medical charts were reviewed to determine the prescribed treatment. Level of agreement between both sets of data was then examined.

Results

A total of 175 caregivers provided complete data. Mean age of the children was 10.2 years (68.6% girls); 44.6% were diagnosed with oligoarthritis. Kappa coefficients for medication represented substantial to almost perfect agreement beyond chance, with better levels of agreement at 12 months (κ = 0.81, 95% confidence interval [95% CI] 0.68, 0.94) than at baseline (κ = 0.61, 95% CI 0.47, 0.76). Kappa coefficients for exercise represented slight to moderate agreement beyond chance, with better agreement at 12 months (κ = 0.44, 95% CI 0.24, 0.63) than at baseline (κ = 0.27, 95% CI 0.08, 0.47).

Conclusion

Weaker agreement for the exercise regimen raises concern that caregivers may pay less attention to exercise recommendations or that these recommendations may not be easily understood.


INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. AUTHOR CONTRIBUTIONS
  8. Acknowledgements
  9. REFERENCES

Juvenile idiopathic arthritis (JIA) is a chronic inflammatory disease of the joints and occasionally other systems, with onset in children younger than age 16 years (1). The treatment of children with JIA typically involves a multidisciplinary approach (2). Physicians invariably prescribe medications to counter the inflammation and pain; physical therapists develop exercise programs to improve range of motion, strength, and function; and occupational therapists may use exercises and splinting techniques for specific joints. Suboptimal adherence to the medical regimens in JIA can have individual (3) and social (4) consequences and may also have a negative effect on clinical decision making (4, 5).

Determining the extent of adherence and factors that contribute to adherence is not without its challenges (6, 7). Researchers have suggested examining children's levels of self-care behaviors instead of focusing on the extent to which their treatment-related behaviors coincide with actual recommendations (8). Levels of self-care behaviors are assessed using self-report questionnaires and refer to the respondent's perception of the extent to which the treatment plan was carried out as prescribed (3, 9–16). This methodologic approach does not require making a comparison of reported self-care behaviors with actual recommendations. However, what caregivers (or children) are reporting may reflect what they believe was prescribed as opposed to what was actually recommended by the medical team, and this may be problematic.

Although no study in JIA has reported the mismatch between what caregivers (or children) believe the treatment was compared with what health care providers actually prescribed, this has been well described in other chronic diseases. In a study of type 1 diabetes, for example, caregivers and/or children recalled recommendations that were never made and few accurately remembered those that were made shortly after a followup clinic visit (17). Children with cystic fibrosis were found to be ingesting medications in a manner not prescribed by their physician (18) and caregivers reported being unsure of whether a formal diet and chest physiotherapy had been prescribed (16). In a study of children with asthma, a substantial proportion of children (51.7%) were exposed to allergens that they had been advised to avoid (19). In another study involving children with asthma, 13 of 49 children were using their metered dose inhaler so poorly as to minimize the likelihood of any delivery of the medication to their lungs (20). Recent research by Modi and Quittner (21) involving children with asthma and cystic fibrosis further supports this earlier work by identifying significant disagreement between physicians and caregivers with regard to prescribed treatment. The authors found that physicians and caregivers of children with cystic fibrosis disagreed on 17% of the medication prescribed, with greater disagreement (31%) for children with asthma.

In the studies described above, concordance between caregiver recall and actual recommendation was determined either by using correlation coefficients (20) or by calculating the numbers/percentages of children who followed recommendations as prescribed (16, 17–19, 21). The correlation coefficient is not a measure of concordance or agreement. Rather, it is a measure of the strength of linear associations between continuous scores, that is, the degree to which 2 sets of scores covary (22). A more useful statistic for measuring the extent of agreement between caregiver recall and actual prescriptions is the kappa statistic for categorical data (22).

The purpose of the present study was to examine the extent of agreement between caregiver recall of their child's treatment regimen for JIA and actual recommendations stipulated by various health care professionals at baseline and at 3, 6, 9, and 12 months following baseline. We used the kappa statistic, which provides a measure of the extent of agreement beyond chance agreement (23). It is necessary to have an accurate measure of the extent of this agreement because health care providers may refer to caregivers' self-report of children's treatment-related behaviors to assist in clinical decisions. In addition, the longitudinal nature of this study affords the opportunity to determine whether agreement strengthens or weakens over time.

PATIENTS AND METHODS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. AUTHOR CONTRIBUTIONS
  8. Acknowledgements
  9. REFERENCES

Sample and procedures.

Data were derived from the JIA Longitudinal Study, which aims to document treatment adherence, disease course, pain, and disability, as well as health-related costs and quality of life (10, 24). Participants were recruited from the rheumatology clinics at 2 pediatric university-based teaching hospitals in Canada (Montreal Children's Hospital, McGill University Health Centre and British Columbia's Children's Hospital in Vancouver) between September 2001 and November 2003. Primary caregivers of children were approached by a project coordinator while attending routine followup clinic visits with their child if they fulfilled criteria for a diagnosis of JIA (1), were fluent in either English or French, and were going to be followed in the clinic for the duration of the study. The caregiver who accompanied the child at his or her clinic visit provided informed consent for participation. This study was approved by the institutional review boards at both hospitals.

At the time of enrollment into the cohort, caregivers were asked to complete a questionnaire packet, which included self-report measures, under the supervision of the project coordinator. Caregivers who were unable to complete the measures in the available time were given a stamped addressed return envelope. Of the completed questionnaire packets, 90% were completed at home and returned by mail, and the remaining 10% were completed in the clinic.

Clinical information on JIA onset type, active joint count, and date of diagnosis was obtained from the patient's chart. Within the context of the JIA Longitudinal Study, doctors and therapists were required to complete flag sheets that inquired about treatment prescriptions (medications, exercises, splinting) at each appointment. These flag sheets, placed in the patient's medical chart, were reviewed at baseline and at 3, 6, 9, and 12 months' followup to determine the prescribed treatment plan.

Measures.

Demographics.

Baseline information on the child's ethnicity, caregiver marital status, and family income level was obtained from the General Information Questionnaire completed by caregivers. This questionnaire was developed by the research team for the JIA Longitudinal Study.

Medical characteristics.

Patients were classified by pediatric rheumatologists into the following 7 specific disease-onset subtypes using the International League Against Rheumatism classification system (1): oligoarthritis, polyarthritis with presence of rheumatoid factor, polyarthritis with absence of rheumatoid factor, systemic arthritis, enthesitis-related arthritis, psoriatic arthritis, and unclassified. The child's treating rheumatologist determined the number of active joints by the sum of swollen joints, and those not swollen, that exhibited warmth or tenderness with pain or limitation on movement. Active joint count has been shown to be associated with disease severity (25). In our sample, the correlation between active joints and disease severity, as measured by the Childhood Health Assessment Questionnaire, was 0.48 (r = 137, P < 0.0001). Disease duration and age at disease onset were recorded in years.

Caregiver recall of treatment.

As part of regular clinic visits, providers spend time with caregivers/children explaining the treatment plan. Additionally, nursing staff follow up on prescribed treatments. In this study, caregivers' reports of whether medication and exercise were recommended as part of their child's treatment plan at baseline and at 3, 6, 9, and 12 months followup were determined using a yes/no format. These reports stem from the Parent Adherence Report Questionnaire (PARQ), a validated questionnaire that also assesses child ability in relation to different treatment components, family distribution of treatment responsibilities, errors in taking medication, caregivers' perceptions of the helpfulness of therapies, and potential barriers to treatment (10). Instructions for completion of the PARQ are prefaced with the phrase “in the past 3 months.” The PARQ has been shown to have satisfactory construct validity and test–retest reliability (10).

Statistical analysis.

Summary statistics, including means and standard deviations for continuous variables and frequencies and proportions for categorical variables, were used to describe the study sample. The kappa statistic was used to measure agreement beyond chance between caregiver reports of the treatment plan and actual medication and exercise recommendations, as indicated on the flag sheets found in the child's medical chart. A priori kappa values >0.81 were taken to represent almost perfect agreement; values from 0.61 to 0.80 represented substantial agreement; values from 0.41 to 0.60 represented moderate agreement; values from 0.21 to 0.40 represented fair agreement; and values <0.21 represented slight agreement (26). Only 20 patients were prescribed splints as part of their treatment plan. Analysis, therefore, focused only on the medication and exercise regimens.

RESULTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. AUTHOR CONTRIBUTIONS
  8. Acknowledgements
  9. REFERENCES

Description of study sample.

As shown in Figure 1, 144 of 157 caregivers who were approached at the Montreal site consented to participate, and 91 of 97 caregivers who were approached at the Vancouver site consented to participate, for a total of 235 participants. Reasons for declining participation included having no time to devote to the study, lack of interest in the study, and simultaneous participation in other studies. Of the 235 who consented to participate, 182 completed and returned their baseline questionnaires. Of these 182, 7 patients had not been prescribed a treatment plan over the initial year of followup because their disease was in complete remission; thus they were also excluded. Therefore, 175 participants represented the final sample for analysis (113 from Montreal and 62 from Vancouver).

thumbnail image

Figure 1. Flow chart of study sample.

Download figure to PowerPoint

At 3, 6, 9, and 12 months there were 25, 5, 19, and 14 dropouts, respectively. There were no statistically significant differences between participants who returned baseline questionnaires and those who did not with respect to children's sex (χ2 = 1.51, P = 0.22), age (t = −0.87, P = 0.38), disease duration (t = −1.31, P = 0.19), age at disease onset (t = 0.53, P = 0.60), and disease onset type (χ2 = 6.19, P = 0.40). There was, however, a statistically significant difference in active joint count (t = −3.88, P = 0.0001), with children participating in the study having more active joints than those not participating (mean ± SD 1.83 ± 3.75 versus 0.58 ± 1.12). Children at the Montreal site had a longer disease duration (P = 0.025) and a lower number of active joints (P = 0.027) than children at the Vancouver site. In addition, more girls (P = 0.004) and a higher number of children with a French or European cultural background (P < 0.001) were recruited at the Montreal site.

Baseline data for child, family, and disease characteristics for the full sample are presented in Table 1. The majority of the children were of Canadian background. The children (55 boys and 120 girls) had a mean age of 10.2 years (median age 11 years), with 77 (44%) being older than 11 years. The majority (76.4%) of participants resided in 2-parent families, with 24.4% reporting a family income between $55,000 and $74,999. Most children were diagnosed with oligoarthritis (44.6%). The mean age at disease onset was 6 years, with 89 (50.9%) children diagnosed with JIA before 5 years of age. Eighty-eight children had a disease duration >3 years. The range for active joint count was 0–29, with 84 children (48%) having no active joints. The sample mean of 1.8 active joints may, in part, reflect the relatively high proportion of patients with oligoarthritis. It may also be indicative of our aggressive approach to therapy that focuses on the eradication of joint disease severity.

Table 1. Descriptive characteristics of the sample at baseline*
CharacteristicValue
  • *

    Values are the number (percentage) unless otherwise indicated. Number of children for each variable does not always equal 175 due to missing data. RF = rheumatoid factor.

Age, mean ± SD (range) years10.2 ± 4.4 (2.0–18.0)
Sex 
 Male55 (31.4)
 Female120 (68.6)
Cultural background 
 French Canadian65 (38.7)
 English Canadian62 (36.9)
 European24 (14.3)
 Other17 (10.1)
Family configuration 
 2 parents113 (76.4)
 1 parent25 (16.9)
 Other (includes shared custody)10 (6.8)
Annual income in Canadian $ 
 <35,00024 (18.9)
 35,000–54,99923 (18.1)
 55,000–74,99931 (24.4)
 75,000–94,99921 (16.5)
 ≥95,00028 (22.1)
Diagnostic categories 
 Oligoarthritis78 (44.6)
 Polyarthritis (RF+)7 (4.0)
 Polyarthritis (RF−)29 (16.6)
 Systemic arthritis16 (9.1)
 Enthesitis-related arthritis18 (10.3)
 Psoriatic arthritis17 (9.7)
 Unclassified10 (5.7)
Age at disease onset, mean ± SD (range) years6.1 ± 4.1 (1.01–15.6)
Disease duration, mean ± SD (range) years4.1 ± 3.6 (0.08–15.6)
Active joint count, mean ± SD (range)1.8 ± 3.8 (0.0–29.0)

Treatment recommendations over the 1-year period.

By physician report, the percentage of children prescribed medication varied over time, but remained >60%: 75.4% at baseline, 74.2% at 3 months, 72.4% at 6 months, 65% at 9 months, and 68.6% at 12 months. According to the caregivers, a somewhat higher percentage of children (always >70%) were prescribed medications: 81.6% at baseline, 75.6% at 3 months, 73.1% at 6 months, 72.3% at 9 months, and 72.3% at 12 months.

By physiotherapist report, the majority of children were initially prescribed exercises, but this percentage dropped steadily over the study period: 77.5% at baseline, 72.7% at 3 months, 68.8% at 6 months, 61.3% at 9 months, and 54.9% at 12 months. Caregivers reported similar numbers to the therapists, except at the 12-month period when substantially more children had been prescribed exercises: 73.5% at baseline, 69.1% at 3 months, 66.4% at 6 months, 59.6% at 9 months, and 66.3% at 12 months.

Extent of agreement between caregivers and health care providers.

Kappa coefficients representing the extent of agreement between caregiver recall and actual recommendations for medication and exercise, separately, are reported in Table 2. The percentage of observed total agreement is also provided for both treatment components. The obtained kappa values for medication ranged from 0.61 to 0.87, representing substantial to almost perfect levels of agreement. The level of agreement was better at 12 months than at baseline. The percentage of observed total agreement for medications ranged from 86% to 95%. These values represent the percentages of caregivers who correctly reported that either no medication was prescribed or that a medication regimen was recommended.

Table 2. Kappa statistics (κ) and percentage of observed agreement (OA%) between caregiver and health care provider reports of medication and exercise regimens*
 BaselineMonth 3Month 6Month 9Month 12
  • *

    Reported statistics calculated on available data, ranging from 74 to 168.

Medication     
 κ (95% CI)0.61 (0.47, 0.76)0.87 (0.76, 0.97)0.82 (0.71, 0.94)0.68 (0.52, 0.84)0.81 (0.68, 0.94)
 OA%86.995.093.285.791.9
Exercise     
 κ (95% CI)0.27 (0.08, 0.47)0.34 (0.13, 0.56)0.19 (−0.02, 0.40)0.54 (0.35, 0.74)0.44 (0.24, 0.63)
 OA%73.673.365.678.472.7

The obtained kappa coefficients for exercises ranged from 0.19 to 0.54, representing slight to moderate agreement beyond chance. As observed for the medication regimen, the level of agreement was better at 12 months than at baseline. The percentage of observed total agreement ranged from 66% to 78%. Of interest is the almost similar percentage of observed total agreement for both baseline and 12 months, even though the kappa statistics for these 2 periods represent different levels of agreement beyond chance.

DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. AUTHOR CONTRIBUTIONS
  8. Acknowledgements
  9. REFERENCES

In this study of children with JIA from 2 Canadian academic pediatric rheumatology centers, levels of agreement beyond chance between caregiver recall and actual prescriptions for medication were substantial to almost perfect. Levels of agreement beyond chance for prescribed exercise regimens, however, were only slight to moderate. This finding raises important questions regarding these differences.

Unlike previous published studies (16–21) that examined caregiver recall of treatment regimens in other chronic diseases using correlation analysis or percentage of agreement to determine group relationships, our study used the kappa statistic. Agreement, therefore, referred to the extent to which responses in treatment regimens between caregivers and health care providers were interchangeable. This addresses the study question of whether caregivers and health care providers produce the same response. A feature of the kappa coefficient is that it compares the observed agreement with the agreement that would be expected to occur by chance (23). Examining the degree of concordance between caregiver and physiotherapist demonstrated that the obtained kappa values were low despite high levels of observed agreement. This provides us with confidence that the results reported herein are adjusted for chance agreement.

An interesting finding to emerge from our data was the stronger level of agreement at 12 months than at baseline for the medication component. This improvement in agreement might be a reflection of caregivers becoming more aware of the importance of paying attention to medication recommendations. However, if this were the case, one would have expected to see a steady and sustained increase in levels of agreement over time. Still, agreement for medication remained considerably higher than that for exercise and was consistently above 0.60. Variability in levels of agreement was especially pronounced for exercises, with month 6 showing the lowest level of agreement and month 9 showing the highest level of agreement. Perhaps the physiotherapists were not specific regarding the continuation of previously prescribed treatment. For example, at a followup clinic visit, caregivers (and children) could have been told to continue with the same exercise routine even if there were no signs of active disease. If the physiotherapist was not clear in communicating that the same exercise routine was being prescribed again, caregivers (and children) may have left the office believing that no treatment was recommended. Reviews of the literature on childhood treatment adherence by Di Matteo (27) and Rapoff (4) emphasize the importance of good communication among caregiver, child, and health care provider in ensuring optimal and sustained adherence. More recently, De Civita and Dobkin (6) suggested that researchers reconceptualize this triadic partnership (caregiver/child/health care provider) by emphasizing the medical team as opposed to only one provider. This takes into account the reality that children and caregivers must interact with different health care providers to ensure optimal management of their treatment plan. The likelihood of misunderstanding recommendations in the absence of coordinated care increases and may, therefore, have an impact on the extent of adherence. Modi and Quittner (21) recently conceptualized misunderstanding in the caregiver-provider relationship with regard to treatment as a potential barrier to adherence. Future research should examine whether the extent of agreement is a barrier to adherence over time.

The kappa agreement provides further reliability data for the PARQ measure that deals specifically with the question of whether medication and exercises were part of the child's treatment plan. Nonetheless, there are possible limitations to our findings that merit consideration. First is the question of how representative the study population is to the total clinic population. Only a relatively small number of children were approached, albeit the great majority of those approached were successfully recruited. Also, considerably fewer children were approached in Vancouver than in Montreal. In both centers there was a tendency not to approach patients when the diagnosis had only very recently been made or when the families were thought to be under a lot of psychosocial stress. This reluctance not to approach patients who might be particularly stressed was perhaps more marked in Vancouver than Montreal. We can only speculate that had these families been included, agreement levels would have been lower given the potential negative effect of stress on caregiver recall of actual recommendations. What we do know is that adherence is affected by familial risk factors (4), but the extent to which stress impacts caregiver recall of treatment recommendations remains to be investigated.

Our comparative analysis showed that the 2 populations recruited from 2 different rheumatology clinics (Montreal and Vancouver) were more similar than dissimilar on a number of characteristics, namely, child age, family configuration, family annual income, JIA type, and age at disease onset. In addition, there is the question of how representative our final sample is when compared with the original sample. Project coordinators at both study sites made several attempts to retrieve questionnaires from caregivers either by contacting them by telephone or by asking them directly about the missing questionnaires at followup clinic visits. Although the attrition rate over the 1-year period may raise concern of whether the study sample is still representative, it is important to note that with the exception of active joints, no statistically significant differences were noted between children retained in the study and those with incomplete data on child and family characteristics. At the very least, the information derived herein allows one to qualify any conclusions as well as to set limits on the extent of generalizability.

Second, caregivers' recall of medication and exercise regimens was matched with recommendations as recorded by the health care providers in children's medical charts. One could not rule out the possibility that, in some cases, disagreement may reflect the reality that recommendations were missing from the medical forms. The practice of recording all recommendations in medical charts may not always be easy for busy health care providers. However, our study incorporated structured forms (i.e., flag sheets) that inquired about details regarding the prescribed treatment, which were completed by health care providers at each clinic visit. The project coordinators at both study sites reviewed the flag sheets to ensure that health providers were complying with the established protocol. This may have minimized the occurrence of missing data. Despite our efforts to measure unbiased caregiver recall, answering the question on whether the child was prescribed medication and/or exercises may have heightened caregivers' awareness of their child treatment plan. One could therefore argue that caregiver recall within the time frame of this study was better than what could be expected in reality (i.e., when not being assessed). If this were the case, then this study has demonstrated that even under the best of circumstances (i.e., being part of such a study) caregiver recall is not perfect and that errors can occur, especially for the exercise treatment plan.

Third, caregivers were asked whether their child was prescribed a medication and an exercise regimen using a yes/no format. They were not asked about the specific details of each treatment component in terms of the type, frequency, duration, and timing, as well as the dose for medication. Studies involving children with cystic fibrosis have reported that accurate recall of specific details related to the medication regimen might be a problem for some caregivers (16, 28). Whether or not caregivers agreed with the physician regarding the type, frequency, and dose for each type of medication prescribed is worthy of future research, but unfortunately was beyond the limits of our data set.

Finally, 44% of the children in our sample were older than 11 years. Research indicates that children with JIA can assume responsibility for their treatment plan (10, 29, 30). It is therefore reasonable to assume that caregivers and children (especially young teenagers) may have shared responsibility for the overall treatment plan and that this sharing may have changed over the 1-year period. However, for all of the children in our sample, caregivers attended each clinic visit and therefore were aware of their child's prescribed treatment plan even though they may not have actively supervised each treatment their child performed. Still, future research should consider how treatment responsibilities are distributed between children and caregivers.

Notwithstanding these limitations, our study is the first to demonstrate that caregivers' recall of treatment for JIA may not always coincide with actual recommendations over a 1-year period. This knowledge is important because the extent of recall agreement may influence adherence over time. Health care providers working with this patient population should therefore apply caution when making decisions about treatment based on self-reports of treatment-related behaviors. This appears to be especially true for exercise recommendations. When treatment involves more than 1 component, the onus lies on the entire medical team to ensure that the caregiver (and, if age appropriate, the child) leaves the clinic with an accurate understanding of what needs to be accomplished by the next clinic visit. A second implication of our findings underscores the importance of having a standard of comparison in place. Recommendations must therefore be accurately and consistently reported in medical charts. In doing so, the practice of comparing self-reports with actual recommendations will undoubtedly prove to be helpful in resolving misunderstandings with the prescribed treatment plan before these become a major barrier to optimal care. In this regard, it is important to emphasize that this study may, in fact, represent the best-case scenario where documentation was indeed high and probably better than in most practice settings.

AUTHOR CONTRIBUTIONS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. AUTHOR CONTRIBUTIONS
  8. Acknowledgements
  9. REFERENCES

Dr. De Civita had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study design. De Civita, Ehrmann Feldman, Dobkin, Malleson, Duffy.

Acquisition of data. De Civita, Ehrmann Feldman, Malleson, Duffy.

Analysis and interpretation of data. De Civita, Ehrmann Feldman, Meshefedjian, Malleson, Duffy.

Manuscript preparation. De Civita, Ehrmann Feldman, Dobkin, Meshefedjian, Malleson, Duffy.

Statistical analysis. De Civita, Meshefedjian.

Acknowledgements

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. AUTHOR CONTRIBUTIONS
  8. Acknowledgements
  9. REFERENCES

We gratefully thank Karen Duffy, Rosie Scuccimarri, Hamid Alenezi, Sarah Campillo, and Paul Daney for their participation in the study; Michele Gibbon, Oliva Ortiz-Alvarez, and Eman Yousif for patient recruitment and data management; and the caregivers and children who agreed to participate in this study.

REFERENCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. AUTHOR CONTRIBUTIONS
  8. Acknowledgements
  9. REFERENCES
  • 1
    Petty RE, Southwood TR, Baum J, Bhettay E, Glass DN, Manners P, et al. Revision of the proposed classification criteria for juvenile idiopathic arthritis: Durban, 1997. J Rheumatol 1998; 25: 19914.
  • 2
    Duffy CM. Patient centred-care and the team approach. In: SzerI, KimuraY, MallesonP, SouthwoodT, editors. Arthritis in children and adolescents: juvenile idiopathic arthritis. Oxford: Oxford University Press; 2006. p. 30314.
  • 3
    Sturge C, Garralda ME, Boissin M, Dore CJ, Woo P. School attendance and juvenile chronic arthritis. Br J Rheumatol 1997; 36: 121823.
  • 4
    Rapoff MA. Adherence to pediatric medical regimens. New York: Kluwer Academic/Plenum; 1999.
  • 5
    Matsui DM. Children's adherence to medication treatment. In: DrotarD, editor. Promoting adherence to medical treatment in chronic childhood illness: concepts, methods, and interventions. Mahwah (NJ): Lawrence Erlbaum Associates; 2000. p. 13552.
  • 6
    De Civita M, Dobkin PL. Pediatric adherence as a multidimensional and dynamic construct, involving a triadic partnership [published erratum appears in J Pediatr Psychol 2004;29:following 404]. J Pediatr Psychol 2004; 29: 15769.
  • 7
    De Civita M, Dobkin PL. Pediatric adherence: conceptual and methodological considerations. Child Health Care 2005; 34: 1934.
  • 8
    La Greca AM, Bearman KJ. If “an apple a day keeps the doctor away,” why is adherence so darn hard? J Pediatr Psychol 2001; 26: 27982.
  • 9
    Davis CL, Delamater AM, Shaw KH, La Greca AM, Eidson MS, Perez-Rodriguez JE, et al. Parenting styles, regimen adherence, and glycemic control in 4- to 10-year-old children with diabetes. J Pediatr Psychol 2001; 26: 1239.
  • 10
    De Civita M, Dobkin PL, Ehrmann-Feldman F, Karp I, Duffy CM. Development and preliminary reproducibility and validity of the Parent Adherence Report Questionnaire: a measure of adherence in juvenile idiopathic arthritis. J Clin Psychol Med Settings 2005; 12: 112.
  • 11
    Glasgow RE, McCaul KD, Schafer LC. Self-care behaviors and glycemic control in type 1 diabetes. J Chron Dis 1987; 40: 399412.
  • 12
    Law GU, Kelly TP, Huey D, Summerbell C. Self-management and well-being in adolescents with diabetes mellitus: do illness representations play a regulatory role? J Adolesc Health 2002; 31: 3815.
  • 13
    Ott J, Greening L, Palardy N, Holderby A, DeBell WK. Self-efficacy as a mediator variable for adolescents' adherence to treatment for insulin-dependent diabetes mellitus. Child Health Care 2000; 29: 4763.
  • 14
    Schafer LC, Glasgow RE, McCaul KD, Dreher M. Adherence to IDDM regimens: relationship to psychosocial variables and metabolic control. Diabetes Care 1983; 6: 4938.
  • 15
    Wysocki T, Greco P, Harris MA, Bubb J, White NH. Behavior therapy for families of adolescents with diabetes: maintenance of treatment effects. Diabetes Care 2001; 24: 4416.
  • 16
    Gudas LJ, Koocher GP, Wypij D. Perceptions of medical compliance in children and adolescents with cystic fibrosis. J Dev Behav Pediatr 1991; 12: 23642.
  • 17
    Page P, Verstraete DG, Robb JR, Etzwiler DD. Patient recall of self-care recommendations in diabetes. Diabetes Care 1981; 4: 968.
  • 18
    Rusakow LS, Miller T, McCarthy CA, Gershan WM, Splaingard ML. Unsuspected nonadherence with recommended pancreatic enzyme administration in patients with cystic fibrosis. Child Health Care 1998; 27: 25964.
  • 19
    Leickly FE, Wade SL, Crain E, Kruszon-Moran D, Wright EC, Evans R 3rd. Self-reported adherence, management behavior, and barriers to care after an emergency department visit by inner city children with asthma [article online]. Pediatrics 1998; 101: E8. URL: http://pediatrics.aappublications.org/cgi/content/full/101/5/e8.
  • 20
    Celano M, Geller RJ, Phillips KM, Ziman R. Treatment adherence among low-income children with asthma. J Pediatr Psychol 1998; 23: 3459.
  • 21
    Modi AC, Quittner AL. Barriers to treatment adherence for children with cystic fibrosis and asthma: what gets in the way [article online]. J Pediatr Psychol 2006; 31: 84658. URL: http://jpepsy.oxfordjournals.org/cgi/content/full/31/8/846.
  • 22
    Ottenbacher KJ. An examination of reliability in developmental research. J Dev Behav Pediatr 1995; 16: 17782.
  • 23
    Feinstein AR, Cicchetti DV. High agreement but low kappa. I. The problems of two paradoxes. J Clin Epidemiol 1990; 43: 5439.
  • 24
    Ehrmann Feldman D, Duffy C, De Civita M, Malleson P, Philibert L, Gibbon M, et al. Factors associated with use of complementary and alternative medicine in juvenile idiopathic arthritis. Arthritis Rheum 2004; 51: 52732.
  • 25
    Singh G, Athreya BH, Fries JF, Goldsmith DP. Measurement of health status in children with juvenile rheumatoid arthritis. Arthritis Rheum 1994; 37: 17619.
  • 26
    Landis JR, Koch GG. The measurement of observer agreement for categorical data. Biometrics 1977; 33: 15974.
  • 27
    Di Matteo MR. Practitioner-family-patient communication in pediatric adherence: implications for research and clinical practice. In: DrotarD, editor. Promoting adherence to medical treatment in chronic childhood illness: concepts, methods and interventions. Mahwah (NJ): Lawrence Erlbaum Associates; 2000. p. 23758.
  • 28
    Ievers CE, Brown RT, Drotar D, Caplan D, Pishevar BS, Lambert RG. Knowledge of physician prescriptions and adherence to treatment among children with cystic fibrosis and their mothers. J Dev Behav Pediatr 1999; 20: 33543.
  • 29
    Degotardi PJ, Revenson TA, Ilowite NT. Family-level coping in juvenile rheumatoid arthritis: assessing the utility of a quantitative family interview. Arthritis Care Res 1999; 12: 31424.
  • 30
    Hayford JR, Ross CK. Medical compliance in juvenile rheumatoid arthritis. Arthritis Care Res 1988; 1: 1907.