The antibiotic approach to treating PANDAS — Pediatric Autoimmune Neuropsychiatric Disorders Associated with Streptococcal Infections — is dividing the medical community because it challenges a fundamental assumption: that psychiatric symptoms in children should be treated with psychiatric drugs rather than antibiotics. The short answer is that a standard 14-day course of antibiotics for an active strep infection is widely accepted, but the real controversy erupts over long-term prophylactic antibiotic use, which some physicians prescribe for one to two years or even until age 18. One study found a 61% overall reduction in neuropsychiatric symptom flares during antibiotic prophylaxis, and a striking 94% reduction in strep-triggered exacerbations specifically. Yet the National Institute of Mental Health maintains there is not enough evidence to recommend long-term antibiotic use for PANDAS or PANS, and many pediatricians and psychiatrists remain deeply skeptical of the entire framework. Consider a seven-year-old who develops sudden, severe OCD and emotional volatility two weeks after a strep throat infection. Her parents watch her personality change almost overnight.
A PANDAS-informed physician prescribes antibiotics and sees the psychiatric symptoms recede within weeks. A conventional psychiatrist, seeing the same child, might prescribe an SSRI and cognitive behavioral therapy without ever testing for strep. Both doctors believe they are acting in the child’s best interest, and therein lies the fault line. First described in 1998 by Dr. Susan Swedo and colleagues at the NIMH, PANDAS remains one of the most polarizing diagnoses in pediatric medicine nearly three decades later. This article examines why the antibiotic approach remains so contentious, what the research actually shows about prophylactic antibiotics and IVIG therapy, how a recent American Academy of Pediatrics report has made insurance coverage even harder to obtain, what biomarker research may eventually settle the debate, and which states have passed legislation mandating coverage for PANDAS and PANS treatment.
Table of Contents
- Why Is the Antibiotic Treatment for PANDAS So Controversial Among Doctors?
- What Does the Evidence Actually Show About Long-Term Antibiotics for PANDAS?
- IVIG Therapy and the Battle Over Advanced PANDAS Treatments
- How the 2024 AAP Report Changed the Insurance Landscape for PANDAS Families
- Biomarker Research and the Search for Diagnostic Certainty
- State Insurance Mandates Are Slowly Changing the Treatment Landscape
- Where the PANDAS Debate Goes From Here
- Conclusion
- Frequently Asked Questions
Why Is the Antibiotic Treatment for PANDAS So Controversial Among Doctors?
The divide comes down to a question of mechanism. PANDAS proposes that a strep infection triggers an autoimmune response in which the body’s own antibodies attack the basal ganglia — a brain region involved in movement and behavior — producing sudden-onset OCD, tics, food restriction, and a constellation of other neuropsychiatric symptoms including emotional instability, sleep disturbances, school deterioration, sensory abnormalities, and enuresis. If this mechanism is real, then treating and preventing strep infections with antibiotics is a logical intervention, much the way long-term penicillin prophylaxis prevents recurrences of rheumatic fever. Skeptics counter that the evidence base is too thin, that the diagnostic criteria are too subjective, and that prescribing years of antibiotics to children based on a contested hypothesis risks antibiotic resistance and medicalizes normal behavioral variation. The acute treatment piece generates little argument. When a child with PANDAS symptoms tests positive for strep, a standard 14-day course of penicillin, amoxicillin, azithromycin, cefdinir, or Augmentin is indicated just as it would be for any strep infection. Where physicians diverge sharply is on what comes next.
PANDAS-informed doctors, drawing on rheumatic fever prophylaxis models endorsed by the American Heart Association, prescribe ongoing antibiotics — typically Augmentin at around 400 milligrams per day, azithromycin at 250 milligrams per day, or penicillin at 250 milligrams twice daily — to prevent future strep exposures from retriggering the autoimmune cascade. The duration has no established consensus, but advocates generally recommend one to two years after the last flare, with continuation until age 18 in severe cases. Compare this to rheumatic fever, where prophylactic antibiotics have been standard practice for decades and no one questions the approach. The difference is that rheumatic fever’s autoimmune mechanism was established through decades of research and broad medical consensus, while PANDAS has been mired in debate since its formal description in 1998. Dr. Swedo has estimated that PANDAS children may represent as much as 25% of children diagnosed with OCD and tic disorders, including Tourette syndrome. If that figure is even approximately correct, the implications are enormous — it would mean a significant subset of children currently receiving psychiatric treatment might benefit more from an infectious disease approach.
What Does the Evidence Actually Show About Long-Term Antibiotics for PANDAS?
The most frequently cited statistic in favor of prophylactic antibiotics is the study showing a 61% overall reduction in neuropsychiatric symptom exacerbations during antibiotic prophylaxis, with a 94% reduction in exacerbations specifically triggered by Group A Streptococcal infections. These are compelling numbers on their face. However, the evidence base remains small, and the NIMH’s official position — that there is not enough evidence to recommend long-term antibiotic use — reflects the reality that large, well-designed randomized controlled trials have not been completed. The studies that do exist tend to involve modest sample sizes and variable methodologies, making it difficult to draw the kind of definitive conclusions that would shift mainstream medical opinion. The limitation that skeptics emphasize most is the difficulty of establishing causation. Children’s OCD and tic symptoms can wax and wane naturally, and strep infections are extraordinarily common in school-age children.
Establishing that a given neuropsychiatric flare was caused by a specific strep infection — rather than coinciding with one — requires the kind of rigorous, large-scale prospective research that has proven difficult to fund and execute. If a child improves on prophylactic antibiotics, was it the antibiotics that helped, or would the symptoms have remitted on their own? This is not a trivial objection. However, parents and clinicians who have watched children experience dramatic, repeatable cycles of strep infection followed by psychiatric crisis and then improvement on antibiotics find the pattern too consistent to dismiss as coincidence. Prevalence estimates illustrate the diagnostic uncertainty. Conservative figures suggest 1 in 200 children in the United States may be affected by PANDAS, while a 2023 retrospective study of primary care populations published in Frontiers in Pediatrics estimated the annual incidence at just 1 in 11,765 children ages 3 to 12. That enormous gap reflects not just methodological differences but the fundamental challenge of identifying a condition that many physicians do not believe exists or do not know to look for. Underdiagnosis almost certainly plays a significant role, but the absence of a definitive diagnostic test means that overdiagnosis remains a legitimate concern as well.
IVIG Therapy and the Battle Over Advanced PANDAS Treatments
Beyond antibiotics, intravenous immunoglobulin therapy has become another flashpoint in the PANDAS debate. IVIG works by modulating the immune system, and the rationale for its use in PANDAS is straightforward: if the disorder is autoimmune in nature, then immune-modulating therapies should help. Early placebo-controlled trials showed that IVIG can decrease symptom severity by more than 60% and shorten the illness course. An Italian cohort study found that 29 of 34 children treated with IVIG experienced symptom reduction or complete disappearance lasting at least one year — a result that, if replicable, would represent a remarkable treatment outcome for a psychiatric condition. But the picture is not uniformly positive. A major double-blind randomized controlled trial found a 24% decrease (plus or minus 31%) in OCD scores on the Children’s Yale-Brown Obsessive Compulsive Scale for the IVIG group, compared to a 12% decrease (plus or minus 27%) for placebo.
That difference was not statistically significant. The wide standard deviations suggest enormous variability in individual responses, which could mean that IVIG works very well for a subgroup of patients but not for others — or it could mean the effects are largely attributable to placebo response and natural symptom fluctuation. Until researchers can identify which patients are most likely to respond, the treatment will remain controversial, expensive, and difficult to access through insurance. The practical reality is that IVIG is costly, often running thousands of dollars per infusion, and many families must fight protracted insurance battles to obtain coverage. For families whose children are in acute psychiatric crisis, the combination of medical uncertainty and insurance resistance creates a particularly cruel situation. They can see their child suffering, they know a treatment exists that has shown promise, and they are told the evidence is insufficient.
How the 2024 AAP Report Changed the Insurance Landscape for PANDAS Families
The American Academy of Pediatrics published a clinical report on PANS in late 2024 or early 2025 that has become the single most contentious document in the PANDAS treatment debate. Insurance companies rapidly seized on the report to deny coverage for IVIG and other treatments, even though the AAP explicitly stated that the report was not intended to serve as a clinical guideline. The distinction between a clinical report and a clinical practice guideline is significant in medical policy — guidelines carry more authority and undergo a more rigorous evidence review process — but insurers have not respected that distinction when issuing denials. The impact has been measurable. Based on a review of nearly 100 PANS and PANDAS insurance appeals filed since November 2024, appeal success rates dropped from approximately 80% to roughly 60% after the report’s release.
For families already navigating a medical system that often does not recognize their child’s condition, a 20-percentage-point decline in appeal success rates translates directly into children going without treatment. The tradeoff here is stark: the AAP’s caution about the evidence base, while scientifically defensible, has had the real-world consequence of empowering insurers to deny care that many treating physicians believe is medically necessary. The PANDAS community has pushed back forcefully. In July 2025, the PANDAS Network and allied organizations published a formal rebuttal letter noting that the AAP report excluded over 50 critical publications from 2018 to 2025 and did not disclose which experts were consulted during its preparation. The PANDAS Physicians Network issued a separate urgent appeal calling for retraction or revision of the report. Whether the AAP will respond to these calls remains to be seen, but the episode illustrates how a single publication — even one that explicitly disclaims guideline status — can reshape treatment access for thousands of families.
Biomarker Research and the Search for Diagnostic Certainty
Much of the controversy surrounding PANDAS treatment could be resolved if a definitive diagnostic biomarker existed. Without one, the diagnosis depends on clinical judgment — the pattern of abrupt symptom onset following strep infection, the presence of characteristic neuropsychiatric symptoms, and the exclusion of other causes. This leaves enormous room for disagreement between physicians who are primed to recognize the pattern and those who are not. Researchers have made progress on several fronts. Studies have identified antineuronal antibodies in the serum of PANDAS patients and documented inflammatory changes in the basal ganglia through neuroimaging — findings consistent with the autoimmune hypothesis. A particularly promising development is the identification of GlcNAc-specific IgG2 autoantibodies, described by Kirvan and colleagues in 2023, that appear to arise only in conditions like rheumatic fever, Sydenham’s chorea, and PANDAS — not in uncomplicated strep infections.
If validated, this biomarker could distinguish children whose immune systems mount a pathological response to strep from those who clear the infection without neuropsychiatric consequences, potentially transforming diagnosis from a clinical judgment call into a laboratory test. The Cunningham Panel, which measures levels of several autoantibodies, has shown that 100% of patients in one study had out-of-range results. However, its diagnostic utility remains debated. A test that is positive in all patients studied may indicate genuine pathology, but it may also reflect issues with the reference ranges or the selection of study participants. The warning for families is this: a positive Cunningham Panel result does not guarantee a PANDAS diagnosis, and a negative result does not rule one out. Until biomarker research matures further, the diagnosis will continue to rest on clinical criteria, and the treatment debate will persist.
State Insurance Mandates Are Slowly Changing the Treatment Landscape
Frustrated by the pace of medical consensus, PANDAS advocates have turned to state legislatures to mandate insurance coverage. California enacted AB 2105 on January 1, 2025, requiring coverage for PANS and PANDAS diagnosis and treatment, including IVIG. Colorado’s mandatory coverage for large group plans began the same day, with individual and small group plans following on January 1, 2026, covering antibiotics, immunomodulating medicines, plasma exchange, and IVIG. Virginia Governor Youngkin signed HB 1641, mandating coverage for policies issued on or after January 1, 2026.
Washington passed HB 2196 with bipartisan support, requiring health plans to cover IVIG when other treatments have been exhausted. These legislative victories are meaningful for families in those states but leave a patchwork of coverage nationally. Thirty-five states still have no laws guaranteeing PANS or PANDAS treatment coverage. A family in California can obtain IVIG for their child with relative ease; a family with the same child and the same diagnosis in a state without a coverage mandate may face repeated denials and financial ruin attempting to pay out of pocket. This geographic lottery in treatment access is one of the more troubling aspects of the current landscape.
Where the PANDAS Debate Goes From Here
The path forward likely depends on two developments: larger and more rigorous clinical trials, and validated diagnostic biomarkers. The GlcNAc-specific IgG2 autoantibody research represents perhaps the most promising avenue toward a blood test that could objectively identify PANDAS, which would undercut the primary objection that the diagnosis is too subjective to justify aggressive treatment. Meanwhile, the growing body of state insurance mandates is creating a de facto standard of care through legislation rather than medical consensus — an unusual dynamic that reflects the intensity of parental advocacy and the frustration of clinicians who treat these patients daily.
The broader term PANS, proposed by Dr. Swedo and colleagues in 2012 to encompass non-strep triggers, may ultimately prove more durable than PANDAS itself, as it acknowledges that the autoimmune mechanism can be set off by infections other than strep. For the brain health community, the PANDAS story is a case study in how slowly medicine moves when a new paradigm challenges established diagnostic categories — and how much suffering that slowness can cause when children are caught in the gap between emerging science and institutional caution.
Conclusion
The antibiotic approach to PANDAS treatment is dividing doctors because it sits at the intersection of infectious disease, immunology, and psychiatry — three fields that do not always speak the same language. The evidence for acute antibiotic treatment of strep in PANDAS patients is uncontroversial. The evidence for long-term prophylactic antibiotics is promising but incomplete. The evidence for IVIG is mixed, with some studies showing dramatic improvement and others failing to achieve statistical significance.
The AAP report has made insurance access harder, state legislatures are stepping in where medical consensus has stalled, and biomarker research may eventually provide the diagnostic clarity that resolves the debate. For families navigating this landscape, the practical reality is difficult. Finding a physician who is knowledgeable about PANDAS, securing a proper diagnosis, obtaining insurance coverage, and managing long-term treatment all require persistence and advocacy that no family should have to provide but many must. The most important step is awareness — knowing that sudden-onset OCD, tics, or personality changes in a child following an infection may have an autoimmune explanation, and that treatment options exist even if the medical establishment has not yet reached consensus on them.
Frequently Asked Questions
What is the difference between PANDAS and PANS?
PANDAS specifically refers to neuropsychiatric symptoms triggered by Group A Streptococcal infections and was first described by Dr. Susan Swedo in 1998. PANS, proposed in 2012, is a broader term that encompasses the same sudden-onset neuropsychiatric syndrome but includes non-strep triggers such as other infections. A child with PANS may have the same symptoms as a child with PANDAS but without a confirmed strep connection.
How common is PANDAS in children?
Estimates vary widely. A conservative figure suggests 1 in 200 children in the United States may be affected, while a 2023 retrospective study found an annual incidence of just 1 in 11,765 children ages 3 to 12. Dr. Swedo has estimated that PANDAS patients may account for up to 25% of children diagnosed with OCD and tic disorders. The wide range reflects ongoing diagnostic uncertainty and likely underdiagnosis.
How long do children typically take prophylactic antibiotics for PANDAS?
There is no established consensus on duration. Advocates generally recommend one to two years after the last symptom flare, with continuation until age 18 in severe cases. Common regimens include Augmentin at approximately 400 milligrams per day, azithromycin at 250 milligrams per day, or penicillin at 250 milligrams twice daily. The NIMH has stated there is not enough evidence to recommend long-term antibiotic use.
Does insurance cover PANDAS treatment?
It depends on where you live. As of early 2026, California, Colorado, Virginia, and Washington have enacted laws mandating some level of PANS and PANDAS treatment coverage, including IVIG in most cases. However, 35 states still have no laws guaranteeing coverage. The 2024 AAP clinical report has made insurance denials more common, with appeal success rates dropping from approximately 80% to 60% since its release.
What is the Cunningham Panel and should my child get one?
The Cunningham Panel is a blood test that measures levels of several autoantibodies associated with PANDAS. In one study, 100% of PANDAS patients had out-of-range results. However, its diagnostic utility is debated in the medical community. A positive result does not definitively confirm PANDAS, and a negative result does not rule it out. It may be one useful piece of information among many, but it should not be treated as a standalone diagnostic tool.
Can PANDAS affect adults or is it only a childhood condition?
PANDAS is defined as a pediatric condition, with onset typically occurring between ages 3 and puberty. However, some researchers and clinicians have observed similar autoimmune neuropsychiatric presentations in adolescents and adults following infections. The broader PANS framework does not have a strict age cutoff, and there is growing interest in whether the same autoimmune mechanisms may operate beyond childhood, though research in adult populations remains very limited.
