Review studies hospital system crash
An external review of a network "glitch" that crashed systems and disrupted patient appointments at Palmerston North Hospital will be held, with findings reported to the MidCentral District Health Board.
The outage occurred at 1.53am on February 13 after the Storage Area Network (SAN) experienced a disk controller failure, which compromised five disks.
Thirty-three outpatient appointments were rescheduled, two joint operations were cancelled and rebooked and a pregnant woman was transferred to Wellington Hospital.
Images were stored locally on machines and read within the department and staff had to physically transport them between departments. Four referral forms were lost during the outage but copies were retrieved from referrers. No images were lost.
The outage was described as "serious" by MidCentral chief executive Murray Georgel and took four days to fix, although the cause is not known.
Patient safety and clinical effectiveness director Muriel Hancock said an external review using specialist technicians would help provide answers.
An internal review of the health board's response to the problem would be provided to the board in April.
"In terms of recovery, the most significant impact appears to have been on medical imaging," Ms Hancock said.
"Most of the corrective actions to come through in the review of our response and recovery are around having timely access to manual systems [pens and paper]."
Inspire Net technical specialist Dave Mill, who was not involved with the MidCentral SAN, said it was common for disks to fail in IT, but a SAN was set up to allow for a certain amount of disk failures.
"The SAN could be set up in a way that you lose one disk and there is no impact to the data or anything else, or it could be set up so you can lose many disks," he said.
"There is no real limit, it depends on how much money you want to spend and what kind of technology you want to use. What you can have sometimes is a SAN with 20 disks in total, for example, and it might be set up in a way where you might have a problem on four of those disks and there is no impact.
"But the moment you go over that number . . . you've got massive problems."
Mr Mill said there was typically no best practice to choosing the limit but it was not normal for multiple disks to fail at once.
He said anything from temperature in the room to a lack of maintenance could cause enough disks to fail so an entire SAN would crash.
General manager Mike Grant said external technicians would undertake the review to ensure best practice guidelines were met.
"Because we run an IT shop with a capacity that is generalist, in relation to the review, we need to bring in specialist subject matter experts to align us with best practice," he said.
The MidCentral SAN is due for replacement and tenders are being evaluated.
- Manawatu Standard
What is the purpose of speed cameras?Related story: Hundreds caught by new speed camera