Users debate outsource vs. in-house disaster recovery

January 28, 2002

The fundamental question regarding disaster recovery is not whether to do it, it's how to do it, according to a paper presented Monday at the HIMSS meeting. "If you're not doing disaster recovery planning, the events of Sept. 11 should have convinced

The fundamental question regarding disaster recovery is not whether to do it, it's how to do it, according to a paper presented Monday at the HIMSS meeting.

"If you're not doing disaster recovery planning, the events of Sept. 11 should have convinced you to begin," said Rob Bortz, director of Information Systems Operations at Lehigh Valley Hospital and Health Network in Allentown, PA. "If you have a disaster recovery plan, you should review and test it."

Disaster plans are necessary for everybody, Bortz said. He advised attendees to think about their building being gone and what steps they could take to get back on the air. As a first step, Lehigh recommends commissioning an independent impact analysis.

"We found that recovering from a disaster would cost us $28,000 a day just in overtime expenses," Bortz said. "Doing clinical tasks manually would lead to delays in admissions and discharges, resulting in longer length of stay. With manual billing, there are cashflow issues if you can't get the bills out."

The real question, then, is how to configure disaster recovery - whether to outsource it or contain the plan in-house.

Lehigh Valley Hospital tried both approaches. It chose first to outsource, for several reasons:

?lack of in-house experience
?lower expense ($14,900 per month) at the time
?no upfront capitol costs
?much simpler environment in 1995
?quick outcome, reliable insurance


Over time, however, Lehigh became disenchanted with their outsourcing arrangement. Testing, which required a one-year lead time, was limited to blocks of just a few hours, and vendor hardware was often unreliable. But mostly Leheigh outgrew the vendor, according to Bortz.

"Our environment got complex, vendor costs increased without a corresponding increase in service level, and as the vendor grew it spent less time with customers," he said.

Then two fortunate events happened: Y2K and the purchase of an office building with an unused data center. Both inspired Lehigh to consider disaster recovery insourcing.

"If we hadn't had all that extra hardware from Y2K planning and the new facility, we would not have been able to insource," Bortz said. "We would have negotiated a new contract or changed vendors."

Now, after insourcing, Lehigh controls the testing schedule.

"Our procedures are better because we test more frequently, and therefore our recover times are much better," he said.

Plus, their operating expenses are lower. Bortz estimates Lehigh has saved about $100,000 by moving disaster recovery in-house.