No quick fix

By the first week of August, it was clear to Glaser that something was wrong with the EMR system. On Friday, Aug. 6, after the system had experienced outages and slowdowns almost every day for two weeks, Glaser realized the status quo was no longer acceptable. He assembled two key groups for a meeting -- the people who touch each of the organization's systems on a daily basis and the IS people who work on the hospital floors.

"I wanted to hear from a technical point of view everything that had happened," Glaser recalls. "I also asked the people on the front lines, What does it look like for the doctors? How disruptive is this? I needed to know what it feels like to live with this."

What Glaser heard wasn't reassuring. The doctors were growing restless. The outages compromised their ability to treat patients. And there wasn't a quick fix looming. Glaser's IS staff identified multiple reasons for the disruptions. The largest single culprit: A server that supported the EMR system kept crashing. Other times, it seemed as if the EMR database had outstripped its capacity. Still other times, an unpatched memory leak in Internet Explorer was the cause of the outages.

"It was clear that there was no way we were going to fix this in one week," Glaser says. "We had to prepare for a series of lousy weeks."

Glaser quickly came up with a game plan. The technical people would continue trying to diagnose and fix the problem, focusing solely on this task. Other projects, such as compliance with the Health Insurance Portability and Accountability Act, would be put on hold. Glaser would provide cover if anybody asked questions about these temporarily neglected tasks. He also brought in a team of IBM Corp. consultants to help diagnose the problem.

In addition, all new EMR system deployments in outpatient physician practices would be put on hold, and a new working group, headed by Glaser's deputy Mary Finlay, was created to approve all application and infrastructure changes. Glaser, meanwhile, would face the music.

He was the one who had to explain to CEO James Mongan and the rest of the leadership exactly what was going on and what was being done about it. While these administrators made it clear that they were not happy, they didn't panic, giving the CIO time to fix the problem. Glaser also met with the physician leadership at Partners' hospitals and explained what was going on. "They don't want to know what you are doing on a day to day basis," he says. "But they want to know that you are taking it seriously."

On Aug. 11, Glaser sent out an e-mail to the entire Partners medical community. In it he acknowledged the recurrent slowdowns and said that fixing the underlying problem was his top priority. He tried to be as frank as possible. "You don't sugarcoat anything, even if it means that you have to look at yourself in the mirror and not like what you see," he says in hindsight. "If you screwed up, you need to be able to say so."