Closing Non-Performing Sites

setting poor sites out to pasture

According to a 2009 study by Morrison and Pfizer, roughly 30% of sites that are opened fail to enroll.  Despite careful site selection criteria and thorough feasibility, dud sites are opened in most trials (although unintentionally and we do go to great measures to avoid this). In my career I have only worked on three studies where every single site enrolled at least one patient.  It is more common to have a handful of sites that recruit the bulk of your patients and then between 10-20% of the sites who never randomize a subject (about 75% of these will offer you Screen Fail subjects…oh boy).

Non-performing sites are a drain on resources and will have a negative impact and cost to your study.

So let’s talk costs of non-recruiting sites.  There is a huge expense to your timeline when a site fails to perform or under-performs because inevitably you are now spending more time to recruit and pick up the slack at other sites.  It is a large drain on resources to start and carry a non-performing site; you still have to monitor, supply this site, contact this site, train this site, perform budgeting and accruals for this site, and provide additional support and cheer-leading for a nonperforming site.  Almost assuredly, those tied-up resources would have a better return on investment if re-directed to other study activities or to support the sites that are making more significant contributions.  Finally there is an opportunity cost because you likely have selected this particular site in lieu of another site that may have been a better recruiter.

Since non-performing sites cost your study, it may be worthwhile to close them, even before the end of the trial. Cut your losses and limit your disappointment.  Your company may choose to close a site that has enrolled subjects but this is much more rare.  This is more likely to happen because of a change in PI, departure of qualified staff required to perform assessments, or some other irreconcilable issue.

Rotten Apple
Don’t let one bad apple spoil the bunch…

Now, I am beating up on the non-performers so I will pause to say that sometimes sites don’t perform and it is not anything that they can control.  Perhaps the inclusion exclusion criteria are too stringent, if they don’t have access to the right patients, if a more attractive study is recruiting nearby, and so on.  However, for the purposes of this discussion, let’s instead call out on some more avoidable reasons that sites fail to recruit: competitive trials at the study site (your PI is cheating on you!), failure to negotiate an appropriate budget, insufficient staffing and resources, no commitment to an Investigative Site Recruitment Plan, uncoordinated search through clinic charts or database, out-of-sight / out-of-mind, inexperience in “pitching” the trial to potential participants, inattention by the PI/failure to prioritize the trial, etc.

Many site contracts are written now so that a start-up fee is partially or wholly refundable to the sponsor if enrollment does not occur within a pre-defined timeframe.  As a cautionary tale, beware of the risk of losing goodwill at your “nice” sites when you close your “naughty” sites.  The community can be tight (and chatty) so conduct yourself accordingly, and exercise tact and professionalism.  If you tread carefully, you may be able to salvage the relationship and appoach this site for future feasibility (assuming you would even want to…past performance is usually the best indicator of future capabilities…).  As an added bonus, you will avoid the ripple effect of negative morale, concern, or fear at your other sites; these can be very de-motivating and halt/delay your study recruitment.

For more information on the tasks required to close a site please refer to a previous blog post on close-out visits.

Are there other hidden costs to keeping a non-performing site open? I would love to hear your comments or you can send me an email to [email protected] – thanks!

You may also like…from The Lead CRA archives:

About The Author

The Lead CRA

Nadia started The Lead CRA blog in 2007. She is now lead author for ClinOps Toolkit. Nadia is currently working as a Clinical Program Manager at a small specialty pharmaceutical company in the San Francisco Bay Area. You can reach Nadia via email at [email protected] anytime.


  • Ankit

    April 9, 2014

    What are the common challenges which a CRA can face during site monitoring visit?

  • Anonymous

    April 10, 2012

    This is good information. You are right. Non preforming sites are a drain. Maybe the site wants to enrol but can’t get it together. Should we help? How can we help? When is it time to cut our loss and close the site? Do we set up parameters at start up: If no subjects enroled by month three close the site. Just thinking!

Leave A Response