The cost of data processing
By Art Randall, April 2010
Editor's Note: The is the ninth installment in our year-long 30th anniversary "Pioneers in Healthcare IT" celebration, featuring articles from past issues of Health Management Technology, formerly called Computers in Hospitals. This article appeared in the May 1985 issue. A.M. (Art) Randall was national sales director, McDonnell Douglas Health Systems, health services division.
One of the strongest characteristics that I have observed of data processing in the hospital industry is the emphasis upon cost rather than results. I have heard statements such as "Our system costs only $1 a patient day," while at the same time their days in receivables were in the high 80s. I have heard others that chose systems based upon the low bid by the vendor, with little or no attention given to the personnel cost that would accrue.
Take, for example, a decision to order and install a new computer, based upon the old saw that it "does more and costs less." Most of us still cling to the idea that the computer itself is the largest single cost of data processing, while we have, at the same time, seen DP costs rise overall, and the cost of computers continue a decreasing cost slide that has been in existence for the past 20 years.
Today, costs other than the machine make up the bulk of DP costs, and in most cases up to 70 percent or 80 percent (people, paper, postage, programs, floor space, heating, lighting, security). Now, suppose a new machine that "does more" at 20 percent less is announced. If the total DP current cost is $15,000 per month and the supposed machine represents $4,500 of that amount, then the net savings per month is $900, but only 6 percent of the total.
Because it does more, the buyer wants to take advantage of that "does-more capability" so that the organization begins sending people to school to learn the new machine (lost productivity), rewrite some old programs to take advantage of the new technology (thus ceasing new application development), or incurring dual rentals while converting. What usually happens is that another person will be hired to pick up the slack, and the 6 percent savings becomes, as a result, another increase in the total cost of DP, while little has changed except that DP costs more.
Stories like the foregoing are not just found in the in-house environment, they exist all across the board. Take the shared services of a few companies or organizations, the services of any one of which is the result of offering essentially the same system to all of its users. Yet, one will find in the single client base of its user hospitals costs ranging from an efficient low to a staggeringly inefficient high, and performance equally spread.
What makes the difference? In most cases, use of "charged for" extra options that are offered are not individually justified on the same basis as the original system when it was initially purchased and subsequently installed. The same thing happens when the first in-house computer system is rigorously justified and then, through the passage of time, more memory or another tape drive is added, a little at a time, so that while the original machine cost was justified, most of the current system was never justified.
I would like to see computer selection committees concentrate on "What will it save?" rather than "What will it cost?" And to consider the total cost impact of the various options rather than just the cost of the service or the machine offered by the vendor.
I would also like to see every increase in cost, after the initially justified applications are installed, go through the same rigorous justification process as those first applications were subjected. And if the additional options are really not options at all, I'd like to see the vendor taken to task the same way the hospital is taken to task when it has to justify increases in unplanned expenditures.