Julia, meanwhile, tackled the spaghetti of IV and monitoring lines on Kathleen's bed. ''Pick one thing and back out" the line from the tangle, said M.J., urging Julia to first fish out the two lines that measure heart function. Because Kathleen was dehydrated, they needed to know -- immediately -- whether she was in imminent danger of heart failure. Not waiting for Julia, M.J. grabbed one of the lines only to discover that there was no electrical cable in the room to attach it to the monitor.

''Did you have three cables when you set up the room?" M.J. asked. Julia said she did, but M.J. shook her head, replying coolly, ''Actually, there was probably only one." Julia started to respond, but held her tongue. She turned back to her patient, then went to fetch more cables.

When she returned, Julia connected the line from Kathleen's pulmonary artery to the monitor, but M.J. was immediately puzzled by the heart rhythm that appeared on the screen. The peaks and valleys didn't seem to match the racing heart of the woman in the bed. Then M.J. figured it out: They were looking at the heart rhythm of the last patient in the room. Julia had forgotten to erase the ''brick," the memory bank in the monitor that holds each patient's information. The mistake was small and correctable, but irritating to M.J. because it showed a lack of planning.

''If you don't do things before, it's just one more thing you have to do when you're really, really busy," she reminded her student.

Still, M.J. was sympathetic, knowing that lots of people make mistakes in an emergency: The operating room team, for example, hadn't kept track of how much blood Kathleen had received during surgery.

Forty-five minutes had passed since Kathleen's arrival, and the nurses were finally getting a clearer picture of her heart's performance. The information on the screen above her bed suggested dangerous dehydration: Her heart was beating fast, 134 beats per minute, but producing only about two-thirds the normal volume of blood. There wasn't enough fluid to pump.

''Damn," said Dr. Jose Ramos, the resident directing Kathleen's care. ''Let a whole liter [of fluid] flow in." He had been reluctant to give too much fluid to Kathleen, fearing that it could accumulate and compress her weak lungs.

M.J. was frustrated, too, realizing that Kathleen had been growing more dehydrated the whole time she was in the room, and, in part, she faulted Julia. She tried to turn frustration into a teaching moment, asking Julia what she should have done first for Kathleen's heart. But when Julia suggested a drug to strengthen Kathleen's heartbeats, M.J. cut her off: ''If there's no volume to pump, it's not enough. Always check the stroke volume before prescribing."

Any pretense that Julia was in charge of Kathleen's care had given way to the reality that M.J. was calling the shots -- and not just among the nurses. She told Ramos that Kathleen would need a nasogastric tube inserted up her nose, both for feeding and for draining fluid from her stomach, and she sent the resident to get it. When Ramos sheepishly returned empty-handed, M.J. chided him good-naturedly -- ''Oh, give me a break" -- and sent a nurse to get the tube.

Julia remained stoic -- and busy -- through the afternoon as Kathleen's vital signs gradually stabilized, but she was disappointed in herself. ''I wasn't happy with how I handled the first 40 minutes," she admitted later. ''I didn't have a good grasp of my patient."

Still, her basic self-confidence was undiminished. ''Smart people learn from their own mistakes," she said. And, she knew, ICU nurses and doctors inevitably make mistakes, given the frailty of the patients and the complexity of care. Even at a top hospital like Mass. General, the goal isn't perfection, but to prevent errors from harming patients.

So, with her first solo shift as an ICU nurse tentatively slated for April 4, Julia felt she was ready. ''There will be people to help me," she said. ''I think I'm pretty realistic about myself. I know I don't know a lot of things."