There's now an ethical SPARK, the science program and autism, I don't know the whole acronym, but it's there and it's a good one. I have a bias. When we evaluated families, we did it in great depth. 13 centers were created around the country, and I think one in Europe. And Cathy Lord, who's a very profound clinician, trained all of the clinicians. She brought them to her own institution, which at that time was Michigan, and they stayed for two days or three days, and taught them how to evaluate the phenotype. And evaluated them by videotaping them and continued to do that after they went back to one of the 13 institutions. They'd be videotaped to make sure they were keeping up with the latest technology. I'm very proud of that, it's called the SFARI collection. In fact, a paper was published.

Cathy and I published a paper together about how things were set up, but then she published a paper on the results from the evaluation. And the coincidence, meaning correspondence, between the scores from the different medical centers – east coast, west coast, middle of the country, Europe, were right on because there was so much attention paid to evaluation of the phenotype. The one thing that was all over the place was the boss, the chief of each section. He was in charge of giving their findings a name and that was all over the spectrum, from autism to autism spectrum to whatever name you can think of. Trying to think of some of the other, the partial, whatever it was, it was a different name. And there was no agreement between Harvard, Yale, UCSF, Washington because the head clinician gave the same phenotype a different name because that was his or her experience. But the emphasis on phenotype is a lesson we'll learn and I hope it comes back.

Today, SPARK gathers phenotypic data by phone, it's not sitting down with the parent or with the child and evaluating them. It's a benefit when they're collecting large numbers – the total was 2,000 families. I think SPARK is up to 50,000 or they want to collect 50,000 on the spectrum. They've already interviewed 100,000 people. But now the question is: how well-characterized are each of these individuals on the spectrum? We know the genetics, but how well do we know the phenotype? That's another one of my concerns, but we'll see, maybe numbers are the answer. The more samples, the more we learn. I think it remains to be seen.

They just had a review of the SPARK program. I sat in on some of that and I thought the review was terrific and pointed out some of the benefits and some of the deficiencies. As a result, there's a new emphasis on informatics and an increasing emphasis on characterizing the phenotype, maybe using external devices to measure blood pressure, pulse rate, perspiration during and after the interview. So, we'll see how that goes.