About one-third of the pediatric psychiatric presentations to the ED were due to risk of harm to self/others and only 27.7% had a prior psychiatric diagnosis; 4.6% required psychiatric admission. The presentation for dissociation (27.7%) was also quite common. Often dissociation mimics neurological symptom/s (e.g., pseudo-seizures, dissociative stupor or aphonia), warranting an immediate visit to ED.

In available literature, toxic ingestions/self-harm, aggression or dissociation have been similarly reported as common presentations to pediatric EDs [12,13]. Williams, et al. [14] reported 27-month data from regional EDs across Detroit (n=225, aged 5-18 years). Thirty-eight percent had severe depression, and 52% were judged to be at acute risk of suicide, 16% had psychotic features, and 34% had potential risk of harming others [14]. In another study from US [15], 21.4% presentations were related to mood disorders, 32.5% to anxiety disorders and 41.3% had substance misuse (41.3%) over the four year review period.

Substance use related emergencies were not much represented in this pediatric sample, in contrast to available literature [13], and in contrast to an adult sample during the same period [4]. Adolescent users are non-dependent with no substantial withdrawals, though they may present with road traffic accidents and fights under influence. It is possible that such cases are not being identified as problematic users, emphasizing the need for screening and brief interventions for early users in EDs. No case of child abuse was encountered in the study period. The child sexual abuse with injuries might have been admitted by surgical specialities, with psycho-logical evaluation at later date. A high index of suspicion is required as child abuse may go unrecognized.

Majority (over 70%) had new-onset symptoms, with no psychiatric diagnosis assigned in past. Such cases pose a diagnostic dilemma especially as diverse medical etiologies may also lead to mental or behavioral symptoms. The pediatricians must take a systematic approach to diagnosis and consider need to involve a psychiatrist for an opinion. A formal psychiatric assessment is warranted for behavioral changes such as irritability, withdrawn behavior, self-harm ideation, aggression, muttering/gesturing to self, especially in absence of ‘red flag’ signs (e.g disorientation, tactile or visual hallucinations, fever etc) [12]. The deferred psychiatric diagnosis in nearly 17% patients highlights the diagnostic difficulties, more so in children and adolescents, often requiring multiple, longitudinal assessments.

Limitations of the study include retrospective design, limited generalizability to other settings, and lack of information on subsequent follow-up status. The diagnosis may be provisional in view of need for subsequent evaluations and longitudinal observations. Certain investigations and diagnostic tests may take several days, for which diagnosis was deferred in a few. Further, the study sample is restricted to patients for whom psychiatric team was consulted in ED. Nonetheless, in spite of these limitations, the report provide a large data set of pediatric patients presenting to ED with mental or behavioral symptoms and ICD-10 diagnosis by trained psychiatrists.

The study findings have implications for service delivery aspects. There is a need to train pediatric residents to identify, provide initial management, stabilization and subsequent referral for common psychiatric presentations in ED, especially imminent suicidal risk or violence among children or adolescents in mental health crisis. Additionally, a close liaison is needed between pediatricians and mental health professionals for providing lateral entry points from ED to mental health-care systems.

Acknowledgements: Dr SK Khandelwal, former Professor and Head, Department of Psychiatry for guidance and support.

Contributors: SS, SK: extracted the data, performed the data analysis and contributed to initial manuscript draft; SS, SK, RD: conceptualised the study, interpreted the findings and finalized the manuscript. All authors approved the final version.

Funding: None; Competing interest: None stated.