A subgroup of patients with childhood-onset obsessive-compulsive disorder (OCD) and/or tic disorders has been identified who share a common clinical course characterized by dramatic onset and symptom exacerbations following group A beta-hemolytic streptococcal (GABHS) infections. This subgroup is designated by the acronym PANDAS (Pediatric Autoimmune Neuropsychiatric Disorders Associated with Streptococcal infections). There are five clinical characteristics that define the PANDAS subgroup: presence of OCD and/or tic disorder; prepubertal symptom onset; sudden onset or abrupt exacerbations (relapsing-remitting course); association with neurological abnormalities (presence of adventitious movements or motoric hyperactivity during exacerbations); and temporal association between symptom exacerbations and GABHS infections. In this subgroup, periodic exacerbations appear to be triggered by GABHS infections in a manner similar to that of Sydenham's chorea, the neurological variant of rheumatic fever.

Rheumatic fever is a disorder with a presumed post-streptococcal autoimmune etiology. The streptococcal pathogenesis of rheumatic fever is supported by studies that have demonstrated the effectiveness of penicillin prophylaxis in preventing recurrences of this illness. A trial of penicillin prophylaxis in the PANDAS subgroup demonstrated that penicillin was not superior to placebo as prophylaxis against GABHS infections in these children, but this outcome was felt to be secondary to non-compliance with treatment, and there was no decrease in the number of neuropsychiatric symptom exacerbations in this group. In a study comparing azithromycin and penicillin, both drugs were completely effective in preventing streptococcal infections - there were no documented titer elevations during the year-long study period for children taking either penicillin or azithromycin. Comparable reductions in the severity of tics and obsessive-compulsive symptoms were also observed. Thus, penicillin was not performing as an "active placebo" as originally postulated, but rather provided effective prophylaxis against Group A beta-hemolytic streptococcal. Both azithromycin and penicillin appear to be effective in eliminating GABHS infections, and reducing neuropsychiatric symptom severity; thus, between-group differences are negligible. Since increasing the "n" to demonstrate superiority of one prophylactic agent over another would be impractical, we have amended the study design to address two issues:

1. To determine if antibiotics prophylaxis against GABHS infections is superior to placebo in prolonging periods of remission among children in the PANDAS subgroup.
2. To determine if antibiotics prophylaxis against GABHS infections is superior to placebo in improving overall symptom severity for obsessive-compulsive symptoms and tics among children in the PANDAS subgroup.

Because penicillin has a narrower therapeutic index and is less expensive than azithromycin, it is the preferable prophylactic agent. Further, penicillin (250 mg orally twice a day) has a long history of providing safe and effective prophylaxis for rheumatic fever and is the first line oral therapy recommended by the American Heart Association. Thus, penicillin has been chosen as the prophylactic antibiotic in the present study. Blister packs are used to increase compliance and to allow for easier documentation of missed doses.