Abstract
BACKGROUND: Attention deficit hyperactivity disorder is considered the most common childhoodneurobehavioral disorder worldwide with well documented adverse consequences in adolescence and adulthood, yet 60-80% of cases go undiagnosed. Routinescreening for the condition is not practiced in most pediatric outpatient services andlittle information exists on factors associated with the condition in developingcountries.
OBJECTIVES: This was a questionnaire based cross-sectional survey whose primary objective was to determine prevalence of attention deficit hyperactivity disorder (ADHD) symptoms in children aged 6-12 years attending the Accidents and Emergency unit of a tertiary care hospital in Nairobi. Secondary objectives were to (i) ascertain if physical injury and poor academic performance were associated with ADHD, (ii) compare diagnostic utility of parent-filled Vanderbilt Assessment Scale (VAS) against Statistical Manual of Mental Disorders-IV (DSM-IV) as the gold reference and (iii) establish if there exists an association between ADHD symptoms cluster and comorbid conditions.
DESIGN/METHODS: The study was undertaken at the paediatric accidents and emergency (A&E) section of the Aga Khan University Hospital (AKUHN) between March and June 2012. AKUHN is a private, not for profit, tertiary health care facility based in Nairobi, Kenya. Paediatrics A&E offers a 24-hour service provided by paediatric residents and senior house officers under the supervision of paediatric registrars. Children aged 6-12 years were enrolled provided guardians demonstrated ability to read and write in English. A written signed informed consent was also required from the primary care provider. Children on methylphenidate, antidepressants or behavioral therapy and those with neurological disorders, hearing and visual impairments or need for emergency care were excluded. Those who consented were clinically evaluated and treated for the ailments that brought them to hospital prior to completion of the self-administered study questionnaire. Sample size was estimated at 240 based on estimated ADHD prevalence of 6% reported by Kashala et al from a neighboring country with similar socio-economic setting as Kenya. Study approval was obtained from the Aga Khan University Hospital Scientific and Ethical Review Committees. Enrolling of children was done after written consent from parents or primary guardians as required by the institutional review board for children under the age of 18 years. It was made clear that recruitment was entirely voluntary and that refusal to participate would not in any way compromise provision of care. Study records were secured in a locked cabinet to safeguard confidentiality. Study was carried out using a two-stage ascertainment procedure. Children were evaluated for eligibility after registration at the reception between 9am to 8pm during week days. A maximum of 10 participants were recruited on any given day to minimize burden in the department and to hopefully capture a wider spectrum of medical conditions. Details about the study were explained to the parents by the principal investigator or the research assistant after patients had been seen by the clinician for the presenting problem. Information necessary for DSM-IV classification was obtained from parents who also completed VAS form. Care providers of study children were requested to complete the risk assessment form with assistance provided as needed. It contained questions about school performance such as repetition of class and average end of term marks which was categorized as; below 25%, 25-50%, 50-75% or above 75%. A grade above 50% was considered as acceptable performance. Only injuries for which medical treatment was sought were considered for inclusion and categorized into burns,fractures and open wounds. Information on causes of injuries was classified under falls, fight, car accident and others. Completion of an assessment form took approximately 15 minutes after which questionnaire was scored and tabulated before providing feedback to parents. Data were entered in Microsoft Excel® and analysis done using STATA®Version 11 (StataCorp). Prevalence of ADHD symptoms was calculated using the number of positive cases as numerator and study population as denominator. Chi square or Fischer’s exact test were used as appropriate to compare categorical variables with P-value below 0.05 considered significant. Wilcoxon test was used for ordinal data. Odds ratios (OR) were used to determine association between ADHD symptoms and categorical variables and 95% confidence interval (CI) to determine precision around individual estimates.
RESULTS: Prevalence of cluster of symptoms consistent with ADHD was 6.3% (95% CI; 3.72-10.33) in 240 children studied. Those affected were more likely to repeat classes than the asymptomatic (OR 20.2; 95%CI 4.02-100.43). Additionally, 67% of the symptomatic had previously experienced burns and 37% post-traumatic open wounds. The odds of having an injury in the symptomatic was 2.9 (95%CI; 1.01-8.42) compared to the asymptomatic. Using DSM-IV as the reference, VAS had a low sensitivity of 66.7% (95%; CI 39.03-87.12) but specificity of 99.0% (95%CI; 96.1-99.2). Its positive predictive value was 83.0% (95%CI; 50.4-97.3) and the negative predictive value 98.0% (CI 95.1-99.1). Positive and negative likelihood ratios were 75(95%CI; 18.3-311.2) and 0.3 (95%; CI 0.21-0.73) respectively. Oppositional defiant disorder symptoms, anxiety, depression and conduct problems were not significantly associated with ADHD cluster of symptoms.
CONCLUSION: A relatively high prevalence of symptoms associated with ADHD was found inchildren visiting the Paediatric Accidents and Emergency department. Symptomaticchildren had also experienced more poor school performance. These findings makea strong case for introduction of a policy on routine screening for ADHD in pediatricoutpatient service in a similar setting. Positive history of injury, especially burns, and poor academic performance is associated with symptoms of ADHD which should trigger need forfurther evaluation for ADHD and appropriate referral. Even though easier toadminister than DSM-IV, Vanderbilt assessment scale has low sensitivity hence itwould not be appropriate for use in ADHD screening. However, in view of its highspecificity and ease of administration, it could be used as an alternative confirmatorytest to determine who among clinically symptomatic patients would require referral to a psychiatrist for further evaluation and management.