If I may ask: does impact on outcomes/patients factor in or is that
analysis left to individual departments?




---
Nikita Bernstein
CEO, Founder

[log in to unmask] | 609.462.6715
www.jomi.com | Journal of Medical Insight

<https://jomi.com>

On Tue, Aug 4, 2015 at 10:46 AM, Harman, Susan E. <[log in to unmask]> wrote:

> A number of you were interested in the replies I got to my question.  I
> was thinking about ebooks, etc. and most people responded about journals,
> but there isn't that much difference really.  Most people used some form of
> cost per usage, but often had different thresholds.  Also the decisions
> would be a little different for academic, large hospital, small hospital,
> etc.  Very interesting info!
>
> Following are excerpts from replies, with identifying informant removed:
>
> <<<<
> Unfortunately we have a stable budget, as journal prices increase, so
> we're going to have to make some cuts. We've decided that for this first
> round, we're going to eliminate any titles that have a cost-per-use that is
> higher than the average ILL fees (even though we get most things through
> freeshare). So, we're looking at about $20. Taking into consideration that
> there are staff time that isn't included in that cost, as well as copyright
> fees, so we might find that we cut too much this year, and add something
> back. We'll see.
>
> It's an experiment, and I think that the more years of data we have, the
> better decisions we'll be able to make. We want to see if we get pushback,
> which I don't think we will because our document delivery/ILL response rate
> is so fast.
>
> That being said, there are some titles that we may keep despite high
> cost-per-use because the usage is really high (even if the price is too) or
> because of incredibly vocal/supportive users. There's a bit of politics
> involved as well as data-driven decision-making!  So, we aren't using hard
> and fast rules, as much as cost-per-use guidelines.
> >>>>
>
> <<<<
> I do a rough cost-per-use calculation AND look at the total number of uses
> MINUS usage for filling ILLs.  If usage for our own staff is over $15/use
> or so, AND is less than 5-7 in a year, it's definitely a candidate for
> cancellation, as I could legally fill that level of need via ILL.  If usage
> is higher than that, I'm more lenient about the cost-per-use, as that
> exceeds fair use guidelines, so I'd have to pay commercial fees of $30+ per
> article.  BUT, if the journal costs $1000 a year, and is used 15 times,
> it's still cheaper for me to pay $30/article ($450) than to subscribe.
>  It's of course all complicated by the near impossibility of picking
> title-by-title any more, as we're all stuck with the Big Bundles With No
> Substitutions problem....
> >>>>
>
> <<<<
> When my library physically closed in 2013, I had a report run of usage
> statistics by journal title. For me, any of the titles that were used less
> than 10 times within 6 months were discarded & subscriptions canceled.
> >>>>
>
> <<<<
> I think the correct amount varies from institution to institution. . .
>
> We were thinking that if a title is used at all during the year, it was
> worth it.  Zero usage needed to be looked at, but, for us, that usually
> appears in bundled titles, so I don't worry about it.
> >>>>
>
> <<<<
> Usage is only one piece of data I use to make decisions to drop - and I
> figure cost per usage.  It's helpful to have historical use data over 2 or
> more years to see if the lack of use is consistent.
>
> I track my use data in a spreadsheet through the year and have Excel
> figure the CPU for me. What is the tipping point of CPU? That can vary  -
> but if it is approaching twice the cost of an ILL fee it's a contender for
> dropping.
> >>>>
>
> <<<<
> [F]or me the answer has to do with how much money is in the budget.  When
> our budget was flush, we picked an arbitrary number that allowed everything
> to balance.  With a more constrained budget, we have made decisions based
> on cost per use and reasonable availability.  To elaborate, if reasonably
> available, we would say that the cost per use would need to be less than
> the "real" cost of getting it by Interlibrary Loan.  The real cost includes
> both staffing cost to locate + any fees the lending library charges + any
> collateral cost of not having it for a "vocal" user.
> >>>>
>
> <<<<
> At our libraries we calculate cost per use for all our electronic
> journals. . . . .. if it costs less than $20 per use, I will keep it. I try
> to balance how much an interlibrary loan article could potentially cost vs.
> keeping the subscription.
>
> This goes for individual titles and bundled titles . . . Last year I ended
> up cancelling [2 collections] - and just bought individual subscriptions
> for the most used titles in each.....even with adding 34 new titles, I
> still saved a significant amount of money.  My rule of thumb for new titles
> is to give each at least two full years to become established with users
> before considering whether to keep permanently or not.
> >>>>
>
> <<<<
> [A]t the liberal arts college where I previously worked, we were starting
> to base our decisions for journals by comparing usage against other
> resources. We determined cost/use for all the journals in the collection,
> and the bottom third (including those with zero usage) were ones we would
> flag for liaisons to review with their departments.
>
> What "review" meant for a given year and department depended on the budget
> for that year, whether we thought we needed to reallocate funds between
> departments, etc. Some years it was "you need to cut $X; these are the
> lowest use titles that we'll cut by default unless you tell us to cut
> others," other years it was "you don't have to cut this year, but review
> these low-use titles to see if there are other titles that would better
> meet your department's needs."
> >>>>
>
> <<<<
> We've used a combination of cost per use and impact factor here -- to the
> satisfaction of our faculty and other heavy users.
> >>>>
>
> <<<<
> We've looked at a ton of things,. . . but for what's it's worth, we've
> found that cost-per-use to be the most accurate stat for our use. It tends
> to provide the best justification for decisions, since it's a money stat.
>
> It's also the best thing to show to the "passionate defenders." It's one
> thing to say that a resource is expensive (most users don't have a good
> handle on costs), or that it doesn't receive much use (users often think
> they use things more than they do, or they think they will use them more in
> the future), but when you can tell someone that their favorite journal cost
> $250 per article view last year, or that the database they want costs
> $50,000, most will say "woah - I don't need it that badly."
>
> You still have to look at other factors though. Resources that are
> specific to smaller programs/disciplines are never going to receive the
> same level of usage that bigger or more general resources will, so that has
> to be taken into account. If it's a journal or other borrowable resource,
> how easy will it be to acquire from another library if you don't get it or
> decide to cancel it? Usage per month can be helpful as well. If all of the
> usage falls within one short period of time each year, that often indicates
> that only a specific article/component of the resource is being used, which
> means you might be able meet that need in other ways besides a direct
> subscription.
>
> As for hard numbers, in my world, anything less than 10 uses a year is
> grounds for immediately cutting, regardless of the size of the
> program/speciality, or the cost-per-use. At that point, it's more
> cost-effective for the institution as a whole to fill that need via a
> combination of ILL and personal subscriptions for the few users using the
> resource than it is to carry an institutional subscription.
> >>>>
>
> <<<<
> Our budget has remained the same for years now, which means that we always
> have to cut things.  Here's how we go about it.
>
> First, we gather all the usage statistics.  Everything we have is now
> online, so I use the Linkout statistics and the ones from the providers.
> Next, we rank them by their usage.  After that, we usually add in the price
> we pay for each title and calculate the price per download.
> We may compare the usage of the title with usage for the last couple of
> years as well.
>
> We have a ballpark $ amount that we have to cut, so we use that to decide
> where to draw the line.  If something has had only a few uses, that's easy
> to add to the list to cut.  And, if there are enough of them to add up to
> our target $ amount, then we can stop there.  But we usually have to cut
> more than that.  At that point, what we usually do is to look at the titles
> with moderate usage, say in the teens for the year.  If this is an
> expensive title with a cost per download that's substantially more than we
> would pay to buy the article from the publisher, then we may select it to
> be cut.  The idea is, we can order articles via ILL until we reach the
> copyright limit, then purchase any others we need and still be saving money
> on that title.
> >>>>
>
> Thanks for everyone's replies and interest.
>
>
> Susan E. Harman, MSLS
> Medical Staff Library
> University of Maryland Medical Center Midtown Campus
> 827 Linden Ave.
> Baltimore, MD 21201
> 410-225-8383
> Fax 410-225-8119
> [log in to unmask]
>
>
>
>
> This e-mail and any accompanying attachments may be privileged,
> confidential, contain protected health information about an identified
> patient or be otherwise protected from disclosure. State and federal law
> protect the confidentiality of this information. If the reader of this
> message is not the intended recipient; you are prohibited from using,
> disclosing, reproducing or distributing this information; you should
> immediately notify the sender by telephone or e-mail and delete this e-mail.
>