WHAT WAS HALAMKA’S ROLE?
Wow. All I can say regarding your cover story “All Systems Down” [Feb. 15, 2003] is “Duh!” I can’t believe your coverage of CareGroup’s CIO John Halamka makes him out to be a reluctant hero. The guy is a CIO for goodness sake. If anyone should know you don’t make an octopus out of an extension cord, it should be him.
The real heroism he showed was bringing in Cisco with its CAP program. It knew right off the bat what was happening there and took appropriate steps to fix it.
All of it, however, seems to come down to money. There was plenty of money for workstation cycles, for applications and so on, but where was the money for the backbone? A network has to be considered as a whole, from its patch panels all the way to its workstations and everything in-between.
This CIO’s IT staff also bears some of the blame. I’ll bet anything people there knew about these issues for a long time but never pushed hard enough to get them resolved. Again, it all seems to come down to money.
Funny how there was no expense spared in calling Cisco and installing the appropriate gear once things went down. IT management is always like that?you can’t budget for upgrades, but you can sure pour money on a burning fire.
This is my personal opinion only, and does not reflect the views of KPMG in any way.
Michael L. Mansfield
Senior Editor Scott Berinato responds: Michael Mansfield is correct on one count: “It all comes down to money.” But he incorrectly supposes money abounded for everything but the network. At an institution that, around the time of the network outage, reported a $26 million loss as phenomenally positive news, there was never enough money for any element of IT, including staffing, another culprit in Mansfield’s estimation.
He also credits Cisco and the CAP program. But as many other readers pointed out, the vendors and integrators never seem to have raised red flags when they were selling Halamka hardware and services along the way, even though they would have immediately recognized a 1996 architecture as outmoded and, ultimately, dangerous.
In fact, Mansfield seems to be looking for heroes and villains, and I believe one of the most compelling elements of the story is that there are neither. Just a lot of human beings.
I found your article very interesting. But it does leave some lingering questions. The seven hops in networking to prevent data from losing its way is a very elementary requirement. I have a difficult time believing that someone from the technical group did not know and could not identify this problem right away.
How was the communication within the IT department?
How well-liked was John Halamka within the IT department?
How open was Halamka to human communication before the shutdown?
The article stated that Halamka was an emergency room physician. That hardly qualifies him as a CIO. How much did he know about IT before the shutdown?
I wonder if the shutdown has its root cause in poor human communication.
Harold W. Niekamp
Beacon Technology Solutions
Berinato responds: Harold Niekamp’s questions focus on internal forces causing Beth Israel Deaconess’s shutdown?specifically a lack of communication and a lack of deep IT knowledge. I don’t believe either of those factors were at the root of the crisis. Halamka is well-regarded for both his IT knowledge and his communication skills. The piece instead focused on external cultural factors contributing to the problem. A culture of merging networks on shoestring budgets, and also a culture of focusing on next-generation health-care IT applications, while the “utility” aspect of technology was treated just that way?as a utility you run and don’t think about.
BEST PRACTICES: GOOD AND BAD
If Michael Schrage’s gloomy if generally accurate assessment of the state of best practices in IT (“Worst Practice,” Feb. 15, 2003) has a fault, it’s that it lets CIOs off the hook far too easily. Especially the appallingly high number of them who, either through inexperience or intellectual laziness, do not distinguish between the simplistic best practices of water-cooler anecdotes, focus groups and advice columns and those that are the outgrowth of empirically derived, methodologically rigorous research.
CIOs should recognize that best practices are rarely simple; more often they are multidimensional and interdependent, and thus take both time and consistent application in order for their full value to be realized. In a decade of research, we have found that the strongest predictor of a company’s ability to capture the benefits of best practices is its level of commitment to addressing the need for sustained change across all dimensions of the IT function?people, processes, and the information IT generates and uses in the way it goes about selecting, deploying and managing technology itself.
The fact that the companies identified by my company’s research methodology vary widely in size, industry and operating model validates the long-held assertion that bona fide best practices are replicable and universally valid. Nowhere is this more true than in the IT function, where the combination of rapidly declining costs for almost all types of technologies, increasing power and improved functionality allows even the smallest organizations to leverage best practices that were once only available to the largest organizations.
President, The Hackett Group
One example of the problem with best practices concerns a major retail chain that developed a “best practices” approach to opening new stores. It applied this methodology to a new store in Latin America, down to the specifications for electrical systems, only to find the systems run on a different voltage than the U.S.-made systems.
Along with best practices, one should also be cautious with benchmarks, another popular KM concept. It may be useful to know current industry norms, but it may be dangerous to make decisions based solely on them.