Attention deficit hyperactivity disorder (ADHD) is the most commonly diagnosed neurobehavioral disorder of childhood, representing one-third to one-half of referrals to child mental health services 1–3. In an extensive review of community studies on childhood ADHD worldwide, prevalence estimates ranged from 1 to 18% 4. In an Egyptian study carried out in Alexandria (2000) 5, the prevalence of ADHD among primary school students was 7.48%. In a more recent study carried out in Cairo (2008) 6, 7.9% of school children were diagnosed with ADHD.
The impact of childhood ADHD behavior on the family has compounding effects on the physical, emotional, and psychological well-being of the child, the family, and the entire community. It is associated with disturbances in family functioning, disturbed parent–child relationships, reduced parenting self-efficacy, and increased levels of parenting stress and parental psychopathology including depression, anxiety, and drug abuse 7,8.
Although medication is a first-line treatment for ADHD children, there are several important limitations to an exclusively pharmacological approach in the treatment of ADHD 9. The results of the Multimodal Treatment Study of children with ADHD suggested that only combined behavioral–pharmacological treatment resulted in improved social skills and improved parent–child relationships, including a reduction in harsh and ineffective parenting 9,10.
Parents play a unique role in ADHD management as the primary caregivers for their children 11. Behavioral parent training (BPT) is one vehicle through which psychosocial assistance can be provided. It involves working with parents to improve their parenting behaviors using behavior modification techniques in order to increase positive outcomes with their children and therefore treat ADHD. They are taught how to identify and manipulate the antecedents and consequences of child behavior, target and monitor problematic behaviors, reward prosocial behavior, and decrease unwanted behavior through planned ignoring, time out, and other nonphysical discipline techniques 12,13.
The efficacy of BPT in treating ADHD has been evaluated in several studies 12,14–17. Pelham and Fabiano 15 reviewed the psychosocial treatment literature on ADHD and concluded that BPT was an evidence-based treatment for ADHD. Children with ADHD have shown improvements in several important areas, most remarkably, parent ratings of problem behaviors and observed negative parent and child behaviors. It also resulted in improvements in parental reports of stress and social behavior 12,14,15. Danforth et al.16 studied the effects of group BPT. Results showed that BPT reduced children’s hyperactive, defiant, and aggressive behaviors, improved parenting behavior, and reduced parent stress. Lees and Ronan 17 studied the effectiveness of a parent training program in the functioning of families of children with ADHD in New Zealand, and they recorded an improvement in parent reports of child behavior, reduced stress and depression levels for most parents, and better parent–child relationships.
The augmented benefits of adding a parent enhancement component to standard BPT aiming to cope with interpersonal distress were evaluated in different studies. This addition produced improvements in parent communication, problem-solving skills, self-reported depressive symptoms, self-esteem, and consumer satisfaction 12,14.
Considering the huge burden of ADHD on victims and their families, effective mental health interventions are greatly needed. This work aimed to evaluate the efficacy of a predesigned culturally sensitive psychosocial intervention program for parents of ADHD children.
Participants and methods
The study was carried out at the Smouha Health Insurance Child Mental Health Clinic for School Students affiliated to the Health Insurance Organization, Alexandria, Egypt. A one-group pre–post test design was used to evaluate the impact of the psychoeducational intervention program on children with ADHD and their parents.
All parents of children with ADHD registered at initiation of the study as well as those of new cases presenting to the above-mentioned clinic during the entire period of the study (10 months period during 2008–2009) were invited to participate in the study. Although 58 parents accepted to join the intervention program, only eight parents dropped out (13.8%). Total of 50 mothers participated in the study.
Data collection tools and techniques
All mothers willing to participate in the program were subjected to a predesigned structured interview to collect family-related and child-related sociodemographic data. In addition, the parental ADHD-related knowledge questionnaire (10 items) was designed on the basis of a literature review and covered parents’ knowledge of the nature of ADHD, course, comorbidity, etiology, and effective management techniques. Three responses for each item were either ‘Yes’, ‘No’, or ‘I do not know’, and were assigned scores of 2 for the right response, 1 for the ‘I do not know’ response, and 0 for the wrong one. The total score of the questionnaire was calculated by summing the scores of the 10 items. The knowledge level was categorized as unsatisfactory knowledge (<50%), moderately satisfactory knowledge (50–75%), and highly satisfactory knowledge (>75%).
All mothers were also subjected to the following psychological tools.
The Arabic version of Conners’ Rating Scale 18
The scale consists of 10 questions identifying ADHD and assessing its severity on a four-point Likert scale. The total score ranges from 0 to 30. Children with scores above 15 are considered to have ADHD 18,19.
Home situations questionnaire 20
Home Situations Questionnaire (HSQ) was originally designed to assess the number and severity of behavior problems of ADHD children in home situations. It consists of 16 items describing everyday home situations. Parents first choose problem situations and then rate the degree of problem severity on a Likert scale from 1 to 9. Parent ratings result in three scores: total number of problem situations; total score of problem severity; and average problem severity.
An Arabic version of this scale was developed and checked through forward/backward translation by qualified interpreters. Then, it was subjected to a pilot study including 10 participants. For the best use of the scale in our culture, some wording of a few items was modified. In addition, the severity score of individual items was collapsed to range from 1 to 3. The internal consistency of the Arabic version of HSQ as measured by Cronbach’s α was 0.87.
The Arabic version of Depression Anxiety Stress Scales 21
The Depression Anxiety Stress Scales (DASS) is a set of three self-report scales designed to measure the negative emotional states of depression, anxiety, and stress. Each of the three DASS scales includes 14 items 22. Parents were asked to use four-point severity/frequency scales to rate the extent to which they have experienced each state over the past week. Scores for depression, anxiety, and stress are calculated by summing the scores for the relevant items. The Arabic version of DASS has been shown to have high internal consistency and to yield meaningful discriminations in a variety of settings 21. Scores of DASS subscales were interpreted according to the following:
The Parenting Scale 23
It is a 30-item self-report scale is designed to measure dysfunctional discipline practices. The measure includes three subscales: laxness, over reactivity, and verbosity. Responses are made using a seven-point scale. Because of its adequate psychometric properties, it was chosen as one of the assessment instruments in the evaluation of community-based programs to improve the parenting skills of parents 24. It was validated by Harvey et al.25 for use with parents of children with ADHD.
An Arabic version of this scale was developed and checked by forward/backward translation by qualified interpreters. Then, it was subjected to a pilot study including 10 participants, and minor wording changes were carried out accordingly. In addition, one modification was applied for the best use of the scale in our culture, where each item received only two responses. Participants’ responses on individual items were categorized into ineffective and effective responses, receiving scores of 2 and 1, respectively. The internal consistency of the Arabic version of the Parenting Scale as measured by Cronbach’s α was 0.78.
A culturally sensitive psychosocial intervention program was constructed with the following objectives: improving parents’ knowledge of ADHD; increasing parents’ understanding of the reasons behind deviant child behavior; improving parents’ positive parenting practices; helping parents practice stress management techniques; helping parents acquire problem-solving skills; and providing social support for parents of children with ADHD.
Mothers of 50 children with ADHD were assigned to groups, each of five to eight parents. Mothers of mentally retarded ADHD children were not included. They received a total of eight sessions on a weekly basis, 45–60 min per session. Sessions of the program included introduction and review of ADHD, understanding deviant child behavior, positive parenting practices (three sessions), managing learning problems among ADHD children, stress management and problem solving, and termination of the program. Training was conducted through a collaborative approach using different training methods including illustrations, vignettes, role playing, brain storming, and group discussions. Evaluation of the impact of the program was carried out using a pre–post test design. Post-testing was performed twice: immediately after program completion and 2 months later using the Parenting Scale, the Arabic version of DASS, HSQ, the Arabic Version of Conners’ Rating Scale, and the parental ADHD-related knowledge questionnaire.
The study was approved first by the ethical committee of the High Institute of Public Health. Then, approval of the North West Delta division of the Health Insurance Organization was obtained for carrying out the study in Smouha Child Mental Health Clinic for School Students affiliated to the Health Insurance Organization in Alexandria.
The objectives of the study were explained to mothers of children with ADHD before program initiation. Only mothers who provided oral consent were included in the study.
The Statistical Package for Social Sciences (SPSS, version 11.5; SPSS Inc., Armonk, New York, USA) was used for data analysis. χ2-test and Fisher’s exact test were used as tests of significance. The one-sample Kolmogorov–Smirnov test was used to test for normality of the distribution. The Friedman test was used as the nonparametric equivalent of a one-sample repeated-measures design. The Wilcoxon signed-rank test is a nonparametric statistical test used to compare two related samples or repeated measurements on a single sample. The McNemar test was used as a nonparametric test for two related dichotomous variables. For all statistical analyses, P-values of 0.05 or less were considered significant.
The age of all participants ranged between 21 and 40 years (mean±SD=33.44±2.7). The majority of them were aged 31 years to less than 41 years (82%), 80% were married, and 20% were divorced. University graduates accounted for the highest percentage (40%), followed by secondary school graduates (34%). A minority of mothers had primary educational certificates (8%). The majority of them were housewives (82%) and belonged to low to middle socioeconomic strata (44 and 40%, respectively).
The differences between the mean total scores recorded by mothers of children with ADHD on Conners’ Rating Scale before and after the program were statistically significant (χ2=61.50, P<0.001). The total mean scores decreased significantly at post-test 1 compared with pretest (24.62±3.76 pretest, 16.66±5.92 post-test 1). Although such mean scores showed a significant increase at post-test 2, they were still significantly lower than the pretest (24.62±3.76 pretest, 16.66±5.92 post-test 1, 19.78±7.18 post-test 2). A similar pattern was found on all items of the Conners’ Rating Scale (Table 1).
As shown in Table 1, the mean total scores on HSQ decreased significantly at post-test 1 compared with the pretest. However, although the mean total scores increased significantly at post-test 2 compared with post-test 1, they remained significantly lower than the pretest (
±SD=29.98±8.21 pretest, 19.50±9.0 post-test 1, 23.76±11.70 post-test 2). Although no significant differences were found between the three phases of the intervention program in terms of the number of problem situations (χ2=3.92, P>0.05), there was a statistically significant difference between the number of problem situations recorded at pretest and post-test 1 (
±SD=13.36±2.63 pretest, 12.24±3.24 post-test 1, and 12.66±3.26 post-test 2).
Figure 1 and Table 2 show mothers’ ADHD-related knowledge before and after program implementation. Program participants recorded statistically significant higher total mean knowledge scores at both post-test 1 and 2 compared with the pretest (χ2=89.63, P<0.001,
±SD=10.78±3.33 pretest, 17.32±2.37 post-test 1, and 17.24±2.33 post-test 2). Percentages of correct answers for each item increased significantly in post-test 1 and remained the same in post-test 2 compared with those in the pretest. Four items had 100% correct responses after the program (Table 2).
As indicated in Table 3, differences between Parenting Scale total scores, and laxness, over reactivity, and verbosity subscales mean scores before and after the program were statistically significant (χ2=62.09 for each of total, laxness, and over reactivity subscales scores, χ2=51.56 for verbosity subscale, P<0.001). There were significant decreases in the mean total and subscale scores at post-test 1 compared with the pretest. Although the mean scores increased significantly at post-test 2 compared with post-test 1, these scores were still significantly lower than the pretest.
Similarly, significant differences between the total mean scores at pretest and post-test 1 and 2 on DASS subscales were found (χ2=28.42 for the depression subscale, χ2=22.96 for the anxiety subscale, χ2=57.05 for the stress subscale, P<0.001). There were significant decreases in the mean scores for the three subscales at post-test 1 compared with the pretest. Mothers’ mean scores reported at post-test 2 showed a significant increase compared with their scores at post-test 1. However, scores reported at post-test 2 remained significantly lower than those reported at the pretest (Table 3).
The acquisition of parental knowledge of ADHD and behavior management has been considered as an important element in the overall treatment of childhood ADHD. The results of the present work highlight one of the important unmet needs of parents with ADHD children. Only minority of the sample (8%) had highly satisfactory knowledge of ADHD. This may be partly attributable to the study setting and socioeconomic profile of the study sample. The study was carried out in a health insurance clinic that is characterized by a high rate of attendance, where doctors have very limited time to offer educational services for parents. Furthermore, the majority of parents attending the clinic had a low level of education, which limits their chances for self-education. Expectedly, the concern of both doctors and parents focuses on treatment options rather than on other issues such as causes of the problem.
The unsatisfactory knowledge of parents of children with ADHD in this work stresses the need for provision of educational services. Group participation in a parent training program is one vehicle toward satisfying that need. One of the most beneficial effects of the present psychosocial intervention was the significant improvement in mothers’ knowledge of ADHD that extended 2 months after program termination. Our findings were consistent with those of Weinberg 26, who reported a significant increase in the knowledge of American parents following parent training for ADHD.
The present work showed a significant improvement in parenting practices. Although this improvement decreased significantly 2 months following the program compared with that recorded immediately after program termination, parenting practices remained significantly better than those reported before program initiation.
In partial agreement with our results, Bradley et al.27 reported significant changes in the total mean scores as well as individual scores of parents of ADHD children following a psychoeducational program. The gains in positive parenting behaviors were maintained at the 1-year follow-up. The maintained effect of the program may be because of the booster session that followed termination of the program by 1 month. This highlights the importance of providing consolidation sessions to such programs to maintain their positive effects.
In agreement with our results, two Australian studies and an American one, applying the same tool as that used in our study, showed a significant decrease in mothers’ scores, that is, improvement in parenting practices following parent training 16,28,29. Similar to our results, Dean et al.28 found a significant decrease from postintervention to follow-up, but the improvement from pregroup to follow-up was still significant.
Further beneficial effects of parent training have been reported for core ADHD symptoms, cooccurring oppositional, and other disruptive behaviors 16,27,28,30–32. Our findings indicated that the severity of ADHD symptoms, severity of problem situations, and number of problem situations among ADHD children improved significantly after the intervention program. Although such improvements were not maintained 2 months following program termination, the severity of symptoms and severity of problem situations were still significantly lower than those reported before program initiation.
In agreement with our findings, Huang et al.30 concluded that a parent training program could effectively enhance the ability of Taiwanese parents to manage their ADHD children and reduce symptoms. Similar to our findings, parents reported only a slight post-training decrease in the total number of problem situations at home. Considering the chronic nature of ADHD, professionals and parents strive to decrease the severity of symptoms in different settings rather than eliminating the problem in one or more settings.
In Canada, Bradley et al.27 implemented a psychoeducational parenting program and found that the intervention group reported significant changes in child behavior compared with controls. Follow-up after 1 year showed that this improvement was not maintained. Perhaps one of the limitations of the present work was the short duration of follow-up. The impact of the current parent intervention program after 1 year with provision of consolidation sessions is an important research question that has to be examined in further research.
Outcome data of an American study carried out by Danfort et al.16 showed that parent training reduced children’s hyperactive, defiant, and aggressive behaviors. As assessed by the HSQ, effect size was medium for the number of problem settings and large to very large for severity of noncompliance, which is consistent with the findings of our work.
It is well documented that the mothers of children with ADHD experience elevated levels of stress in their parenting roles compared with mothers of normal children. Elevated parenting stress is believed to disrupt the parent–child relationship and negatively affect parenting practices. The links between elevated parenting stress, disruptions to the parent–child relationship, and parenting practices argue for treatment programs that reduce parenting stress in the families of children with ADHD 32–34.
In the present work, parents’ psychological stress was addressed through sessions that focused on stress management and problem solving. A significant decrease in mothers’ experience of stress as well as of depressive and anxiety symptoms was found. This decrease was significant immediately following program termination. Although symptom severity and stress level showed a significant increase 2 months after program termination, it remained significantly lower than that at baseline.
Dean et al.28 assessed depression, anxiety, and stress symptoms using DASS among Australian parents of ADHD children following a parenting program. Their findings showed a significant decrease in depression and stress from pregroup to postgroup, but no change from postgroup to follow-up, indicating maintenance of change over 6 months. The results for anxiety were similar to those of the present work; the decrease in anxiety from pregroup to postgroup was significant and there was no significant difference from post to follow-up. However, the decrease in anxiety from pregroup to follow-up was significant. Partial inconsistency in the results of the two studies in terms of maintenance of improvement may be attributed to differences in culture, sex, and socioeconomic level of the samples included. In addition, all participants of our program were only mothers, the parent who is more likely to carry the main burden of child rearing and more likely to have depressive symptoms.
Further evidence of support is derived from other studies that reported significant reductions in parenting stress, depression, as well as significant improvements in parenting style, and increase in participants’ perceived social support following parent training programs 14,16,31–33.
The role of social support provided to parents attending such programs is of great benefit not only because they obtain more information about their children’s problems and feel much better about their problems but also because such support may play a role in improving parents’ psychological well-being.
It is notable that, except for the impact of the program on parents’ knowledge, which remained the same 2 months following program termination, the improvement observed in all other aspects evaluated by the intervention program was not maintained 2 months later compared with that reported immediately after program termination; however, this improvement was still significant compared with baseline data. One explanation is that knowledge can be easily attained and better maintained than building skills and practices. This raises the need for booster sessions to help parents consolidate learnt skills and find solutions to problems that emerge during practice and in different stages of child development.
Conclusion and recommendations
The present work has developed a culturally sensitive psychosocial intervention program for parents of children with ADHD that proved to be effective in reducing the severity of ADHD symptoms and problem situations among ADHD children as well as improving parents’ ADHD-related knowledge, discipline practices, and their psychological well-being.
Accordingly, the following recommendations are suggested:
- Establishing a parental services unit as a subdivision of mental health clinics, aiming at delivering specialized services for parents of children with ADHD.
- Effective multimodal intervention programs should include psychosocial intervention programs for parents as an integral component of a patient management plan that address the multiple needs of children with ADHD and their parents.
- Providing booster sessions to consolidate the beneficial effects of parents’ psychosocial intervention programs.
Conflicts of interest
There are no conflicts of interest.
1. .Increasing prevalence of parent-reported attention-deficit/hyperactivity disorder among children – United States, 2003 and 2007.MMWR Morb Mortal Wkly Rep2010;57:1339–1343.
2. .ADHD: clinical practice guideline for the diagnosis, evaluation, and treatment of attention-deficit/hyperactivity disorder in children and adolescents.Pediatrics2011;128:1007–1022.
3. Cormier E.Attention deficit/hyperactivity disorder: a review and update.J Pediatr Nurs2008;23:345–357.
4. Faraone SV, Sergeant J, Gillberg C, Biederman J.The worldwide prevalence of ADHD: is it an American condition?World Psychiatry2003;2:104–113.
5. Abo-Rass NT.Attention-deficit/hyperactivity disorder among primary school pupils in Alexandria [dissertation]2000.Alexandria:Alexandria University.
6. Farid MN, Sabbour SM, Osman MH.Prevalence and risk factors of attention deficit hyperactivity disorder among school children.Egypt J Commun Med2008;26:27–41.
7. Johnston C, Mash EJ.Families of children with attention deficit/hyperactivity disorder: review and recommendations for future research.Clin Child Fam Psychol Rev2001;4:183–207.
8. Salmeron PA.Childhood and adolescent attention-deficit hyperactivity disorder: diagnosis, clinical practice guidelines, and social implications.J Am Acad Nurse Pract2009;21:488–497.
9. Chronis AM, Jones HA, Raggi VL.Evidence-based psychosocial treatments for children and adolescents with attention-deficit/hyperactivity disorder.Clin Psychol Rev2006;26:486–502.
10. .National Institute of Mental Health Multimodal Treatment Study of ADHD follow-up: 24-month outcomes of treatment strategies for attention-deficit/hyperactivity disorder.Pediatrics2004;113:754–761.
11. Dobie C, Donald WB, Hanson M, Heim C, Huxsahl J, Karasov R, et al..Institute for Clinical Systems Improvement. Diagnosis and management of attention deficit hyperactivity disorder in primary care for school-age children and adolescents2012:9th ed..Bloomington:The Institute;79.
12. Chronis AM, Chacko A, Fabiano GA, Wymbs BT, Pelham WE.Enhancements to the behavioral parent training paradigm for families of children with ADHD: review and future directions.Clin Child Fam Psychol Rev2004;7:1–27.
13. Efron D, Hazell P, Anderson V.Update on attention deficit hyperactivity disorder.J Paediatr Child Health2011;47:682–689.
14. Lee Pc, Niew W, Yang H, Chen VC, Lin K.A meta-analysis of behavioral parent training for children with attention deficit hyperactivity disorder.Res Dev Disabil2012;33:2040–2049.
15. Pelham WE, Fabiano GA.Evidence-based psychosocial treatments for attention-deficit/hyperactivity disorder.J Clin Child Adolesc Psychol2008;37:184–214.
16. Danforth JS, Harvey E, UlaszeK WR, McKee TE.The outcome of group parent training for families of children with attention-deficit hyperactivity disorder and defiant/aggressive behavior.J Behav Ther Exp Psychiatry2006;37:188–205.
17. Lees DG, Ronan KR.Parent management training for solo mothers of children diagnosed with attention deficit hyperactivity disorder: an effectiveness and multiple baseline evaluation. Palmerston North, New Zealand: Massey University 2005. Available at: http://126.96.36.199/Library/items/parent-management-train-solo-moms-adhd_05.pdf
. [Accessed 5 September 2012].
18. El-Dafrawi MH, Mahfouz R, Ragab LA.Reliability and validity study of a rating scale for ADHD in Egyptian children and adolescents. Proceedings of the 3rd International Egyptian Congress of Psychiatry; 31 May–2 June 1990; Cairo, Egypt.
19. Conners CK.Manual for Conners’ Rating Scales1989.New York:Multi-Health Systems Inc.
20. Altepeter TS, Breen MJ.The Home Situations Questionnaire (HSQ) and the School Situations Questionnaire (SSQ): normative data and an evaluation of psychometric properties.J Psychoeduc Assess1989;7:312–322.
21. Taouk M, Lovibond P, Laube R.Psychometric properties of an Arabic version of the Depression Anxiety Stress Scales (DASS). Report for New South Wales Transcultural Mental Health Centre2001.Cumberland Hospital, Sydney:Transcultural Mental Health Centre.
22. Lovibond SH, Lovibond PF.Manual for the Depression Anxiety Stress Scales1995.Sydney:Psychology Foundation.
23. Arnold DS, O’Leary SG, Wolff LS, Acker MM.The Parenting Scale: a measure of dysfunctional parenting in discipline situations.Psychol Assess1993;5:137–144.
24. Irvine AB, Biglan A, Smolkowski K, Ary DV.The value of the Parenting Scale for measuring the discipline practices of parents of middle school children.Behav Res Ther1999;37:127–142.
25. Harvey E, Danforth JS, Ulaszek WR, Eberhardt TL.Validity of the parenting scale for parents of children with attention-deficit/hyperactivity disorder.Behav Res Ther2001;39:731–743.
26. Weinberg HA.Parent training for attention-deficit hyperactivity disorder: parental and child outcome.J Clin Psychol1999;55:907–913.
27. Bradley Sj, Jadaa D, Brody J, Landy S, Se Tallett, Watson W, et al..Brief psychoeducational parenting program: an evaluation and 1-year follow-up.J Am Acad Child Adolesc Psychiatry2003;42:1171–1178.
28. Dean C, Myors K, Evans E.Community-wide implementation of a parenting program: the South East Sydney Positive Parenting Project.Australian e-Journal for the Advancement of Mental Health (AeJAMH)2003;2:12. Available at: http://pandora.nla.gov.au/pan/24225/20040403-0000/auseinet.flinders.edu.au/journal/vol2iss3/dean.pdf
. [Accessed 5 September 2012].
29. Burke S, Soltys M, Trinder M.A preliminary evaluation of the Together Parenting Program – a stand alone component of the Exploring Together Program.Australian e-Journal for the Advancement of Mental Health (AeJAMH)2008;7:10. Available at: http://pandora.nla.gov.au/pan/24225/20081003-0015/www.auseinet.com/journal/vol7iss1/burke.pdf
. [Accessed 5 September 2012].
30. Huang HL, Chao CC, Tu CC, Yang PC.Behavioral parent training for Taiwanese parents of children with attention-deficit/hyperactivity disorder.Psychiatry Clin Neurosci2003;57:275–281.
31. Thorell LB.The Community Parent Education Program (COPE): treatment effects in a clinical and a community-based sample.Clin Child Psychol Psychiatry2009;14:373–387.
32. Chacko A, Wymbs BT, Wymbs FA, Pelham WE, Swanger-Gagne MS, Girio E, et al..Enhancing traditional behavioral parent training for single mothers of children with ADHD.J Clin Child Adolesc Psychol2009;38:206–218.
33. Treacy L, Tripp G, Baird A.Parent stress management training for attention-deficit/hyperactivity disorder.Behav Ther2005;36:223–233.
34. Barlow J, Coren E, Stewart-Brown S.Parent-training programmes for improving maternal psychosocial health.Cochrane Database Syst Rev2003;Issue 4:CD002020.