Statewide, at least 75 firms, some with long waiting lists, offer ABA in a variety of styles. The cost -- often $50,000 a year or more for a single child -- has been covered primarily by taxpayers up to now.
In October, California passed a law requiring private health insurers to include coverage for ABA in their policies. More than two dozen other states have similar requirements, the result of relentless campaigning by parents and their advocacy groups.
Yet for all the belief and investment in ABA, important questions remain about its effectiveness and how best to use it.
Researchers have established that the therapy can improve behavior, language and intelligence test scores. But no scientist using the gold standards of medical research has reproduced the success rate reported by Lovaas, who died last year.
Moreover, nobody can explain why some children respond to treatment and others make little progress, even after thousands of hours. It is also unclear which elements of ABA are responsible for the gains, how many hours a week are needed and for how long, and whether children who receive the therapy fare any better in adulthood than those who do not.
Complicating the picture is that scientists and providers themselves have taken the therapy in a variety of new directions. Although the principle remains the same -- long hours, day after day, in early childhood -- there has been little research on how various approaches compare.