Guest Column | July 21, 2023

Save Time And Money By Using AI To Compare CRO Proposals

By Audrey Rossow

I have been asked often in my career, “How do you select a CRO?” or “What are the best CROs?” There is no single or simple answer to such a complex question. Sponsor companies outsource millions of dollars to their CRO partners, and so the identification, selection, and ongoing oversight should be considered carefully and managed proactively. Further, CROs deserve and desire a true partnership, not a transactional relationship.

When it comes to finding and contracting with a CRO, here is what I believe we tend to do as an industry:

  • Ask coworkers, friends, and colleagues.  
  • Conduct an internet search.
  • Read articles or annual CRO report ratings.
  • Work with whom we have used before.
  • Work with whom we have heard of before — like a “big name” organization.

These are all somewhat non-scientific ways to select a partner with whom you are trusting millions of dollars and the care of your asset — and  potentially the future success of your company. We have regulations and guidances on many aspects of clinical research and on oversight of vendors, yet none on the selection of CROs. There must be a better way!

What Matters Most

I have worked at small and midsize sponsor companies and CROs, as well as a few patient recruitment vendors. I also spent eight years as an independent consultant. During that time, I’ve encountered three recurring themes to consider when selecting the best CRO partner for your organization.

  1. Compatibility — How do you identify and select a CRO partner that is right-sized for your organization?
  2. Capability — How do you ensure you have the partner with the right expertise in the critical areas for success?
  3. Communication — How rigorous is your request for proposal (RFP) process? The RFP is the foundation upon which the partnership with the CRO is built. It must be all-encompassing, very detailed, and express the depth and breadth of what you need very clearly. Included in the RFP should be these key modules:
  • Assumptions – for example the number of sites, the location of the sites (i.e., in what countries), how many local versus central IRB/ECs we expect to have, the anticipated number of subjects screened, enrolled and completed.
  • Timeline/Milestones
  • Roles and Responsibilities (R&R)
  • Questions and Answers
    • Applicable to every CRO:
  • When was your company founded?
  • Are you privately or publicly held?
  • In which countries/cities do you have a legal presence physically vs. virtually?
  • What is your current employee turnover rate?
  • Please provide at least two references from a company similar in size to the sponsor company.
  •  Relevant to a particular trial:
    • Please describe your experience in this phase and therapeutic area.
    • In which countries have you run a similar study?
    • Please provide redacted information for 3-5 sites you would propose for this study, including city, state, practice type as well as past performance on a recent similar study.

Case Study #1: Selecting A CRO The Old Way

The old-fashioned way is for the sponsor company to send out the RFP in a combination of Word and Excel files along with the protocol and other relevant documents, then wait for the CROs to send back their proposals, budgets, and additional relevant documents. Then, in my experience, I would spend weeks sifting through all the data and trying to compare them across CROs to understand costs, capabilities, and the more subjective qualities each company brought to the table. Subjective questions may include:

  1. How do you ensure proper site feasibility when there is a dearth of appropriate sites?
  2. In what way do you "risk share" regarding milestone achievement?
  3. Are you willing and able to start certain activities "at risk" – e.g., ICF templates, site budgets and contracts, EDC build – if the protocol is not final?

As an example, in my role as the head of clinical operations at a small biotech company,  I sought to obtain competitive bids for an early phase drug-drug interaction study from three CROs, then compare the budgets and services of those three CROs and make a recommendation to executive leadership. I paid a consultant $200 per hour to send the RFPs and upon receipt, to analyze and cross-compare, by functional areas, the proposed costs and capabilities. It took the consultant approximately 50 hours to complete the task, which added up to $9,000. This cost included a slide presentation with her findings and final recommendation. Had I as the employee done this work, it would have taken a little more than one week— if I did absolutely nothing else during that time.

Case Study #2: A Better Way With AI

Now, I use AI-based technology to make this task much easier. There are a few AI-enhanced vendors in the marketplace that can help you quickly create, send, and receive RFPs online to more efficiently select vendors, such as CROs, imaging labs, spirometry vendors, etc. Such platforms provide AI-based analytics, making rapid apples-to-apples RFP comparisons possible. They are available by paid subscription and are life sciences specific.

With these AI-enhanced platforms, you can:

  1. choose from a library of CRO or vendor RFP templates,
  2. create RFPs within minutes from these best practice templates,
  3. use AI-based analytics (side-by-side RFP comparisons) to compare costs, and
  4. remain ICH E6 R2 compliant.

For example, not long after case study #1, I obtained a bid on a rare respiratory disease across six CROs and two studies (a Phase 2 and an open label extension). Based on the math in case study #1 above, that would have cost me $40,000 to $50,000 to outsource the task or taken me about four to six weeks to complete on my own with all its complexities. Instead, I personally researched, vetted, and then requested an AI-enhanced vendor to work with the identified CROs and produce the necessary data and analytics. When vetting the vendors, I chose to pursue one built by a veteran ClinOps professional. Within minutes of receiving the vendor-provided data, I could see that the top five areas of cost divergence were in pass-through costs, project management, regulatory, site monitoring, and TMF management. Thanks to the vendor data, I could then focus my efforts on working through those five areas to determine the true differences in cost and expertise.

Digging In Further

For the more qualitative considerations, the AI vendor RFP included a Q&A module that assesses such facets as:

  • Therapeutic area expertise
  • Geographic area expertise
  • Site identification and selection
  • Vendor management and QA
  • Contracting/Finance
  • Compliance/Regulatory
  • Timeline management
  • Risk management and risk sharing
  • Team composition
  • TMF considerations

Compared to the traditional manual process, AI-enhanced vendor comparisons reduced the time it took for me to compare multiple CRO proposals and budgets for my clinical trials. I estimate that I saved eight weeks in vendor selection time and 10 percent of the total contract value with vendors. By applying rapid and powerful AI behind the scenes, I could focus on the functional area line items that were most disparate, rather than having to analyze every single line item from project management to data cleaning to clinical study report. (Some CROs found the platform to be restrictive in terms of not being able use their own RFP templates, but the AI vendor pivoted to being able to upload the CROs’ templates and integrate them into the system.)  With the data visualization tools plus the interactive questions and answers section, I reduced the cycle time between revisions to and from the CROs and got the most accurate comparisons. Importantly, executive management supported my exploration of AI-assisted vendors from the get-go because the services were relatively inexpensive and could be applied to more than one study. Finally, thanks to the data analytics and accompanying graphics, I presented a coherent and succinct proposal via slide presentation to management, which ultimately used the information to choose the right CRO for us.

About The Author:

Audrey RossowAudrey Rossow is the US-based Director of Clinical Operations at Precirix,  a Belgian company dedicated to extending and improving the lives of cancer patients by designing and developing novel radiopharmaceuticals, using camelid-derived single-domain antibodies labeled with radioisotopes. She has over 30 years’ experience in pharmaceutical and biotech clinical drug development, in many therapeutic areas, Phases 1 through 3. Her expertise is in project management and clinical operations. She is passionate about site engagement and support, patient recruitment and retention, and sponsor oversight of their CROs.  You can reach her at