If Van de Vliet had achieved results in just a fraction of those trials, it would herald the beginning of a new age in medicine.

The NIH monitor for the Gerex trials was Cheryl Gates, just as Dale had said. Her photo was featured along with the introductory description of the trials. Nice‑looking, he thought, probably late thirties, dark hair, dark‑rimmed glasses. She wasn't wearing much makeup in the photo, probably to emphasize how serious she was. Sooner or later, he told himself, he had to find a way to meet her. . . .

He stared at his IBM Aptiva screen a moment longer, overwhelmed at what he was seeing, then got up and walked into the kitchen and made a peanut butter sandwich, whole wheat. It was a rehearsal for the possibly hard times to come. Then he retrieved a Brooklyn Lager from the fridge. It was his day off and the sun was over the yardarm.

He lived on the fourth floor of a brownstone in Yorkville, in New York's East Eighties. The apartment was small, but it was rent stabilized which meant he was paying well under market value—$1,128 a month on a place that probably could go for close to twice that on the open market. He'd lucked into it after he and Jane split—even though they weren't married they'd bought a condo in the West Fifties, and at the breakup they'd switched the mortgage to her name—but the problem now was, how was he going to pay even this piddling rent (not to mention child support for Amy) after he got fired from the Sentinel? That day, he sensed was fast approaching. And if it happened before the book was finished he was just three months away from going back to freelancing. That was how long his "nest egg" would last.

Carrying the sandwich and beer, he walked back to his "office," a corner of the cramped living room that had an Early American desk, and sat down at the frayed chair in front of his IBM.

So here he was, past the first level of security of the NIH site, zeroed in on the Gerex clinical trials. Somewhere here had to be all the data about the patients who had been, and currently were, participating.

He moved on to the results section and opened the first page. Yes.

Then he looked more closely.

Hello, we've got a problem. The patient data he was looking at had only code numbers for names. The categories of trials also were just numbers. Without a key, there was no way to get a single patient name or differentiate Alzheimer's from fallen arches. Then he saw the notice at the top of the page: As part of the NIH policy on privacy, all patient data are aggregated and anonymous.

Shit.