The use of Central Laboratories (CLs) is very widespread in clinical trials. There are two main reasons for using CLs in clinical trials (see Table 1): high requirements for the quality of the lab data in trials1-3, and inability to organize required tests in local laboratories.
A central laboratory is an institution that is exclusively responsible for laboratory assessments and provides services from conducting laboratory assessments and compiling lab test reports, to contracting courier services for delivering lab kits and biosamples from/to medical institutions where diagnostics and treatment of patients is performed4.
Regardless of the important role (and rather high cost) CLs play in clinical trials, it is not so much the objective factors, but rather the subjective ones quite often that influence its selection. Having interviewed more than 50 sponsors we have found the following leading four motives in CL selection:
- “Good recommendation from a person we know.”
- “One our colleague worked with this laboratory.”
- “We know a good specialist in this laboratory.”
- “We had positive experience working on a different study some time ago.”
We should particularly emphasize a situation when, before the start of a study, the sponsor performs a project-specific audit of its pre-selected CL. This approach is more adequate since it provides an opportunity to obtain the most reliable information about the lab. However, it requires time and additional expenses. Also, to have a truly objective optimization of the CL selection process, it would be necessary to audit several laboratories.
Even though some companies offer a selection of CLs for trials, the criteria of objective and competitive selection of CLs are not formulated and are very rarely discussed. Having many years of experience in this area, we have developed an algorithm for CL selection.
Modern CLs are able to provide laboratory test results of similar quality. There are no laboratories insured from mistakes, but no laboratory worth its salt could be expected to provide incorrect results that would amount to a statistically significant level. The availability of various certificates and licenses that prove the quality of work does not necessarily guarantee high quality. And if the project audit of a lab is performed by auditors who don’t possess professional laboratory knowledge, it does not ensure the high performance of the lab during the study.
The general concept of our algorithm for selecting a CL is provided in Fig. 1.
For the initial contact (Stage 1), CLs can be pre-selected using your own database, references, and even an internet search. Our experience shows it’s sufficient to contact three to five CLs. The general rule is that, the more complex the tests that are required, then the more labs should be approached. Even assuming that we have a CL in which we trust, it is necessary to collect the info from at least two more labs. On the other hand, more than five such contacts is an excessive labor-intensive work and evidence of poor planning. For a preliminary contact, a short request about the available tests — with the indication of the study geomix — and general interest in the study is enough. At this stage, we would not recommend approaching a non-replying CL for the second time; we suggest considering “no reply” as absence of interest in the project. According to our experience, it takes no more than a week to gather initial information from all labs.
The second contact (Stage 2) is the most difficult and demanding. The initial readiness to participate in a study may not necessarily prove to be the case later on. It is important to get the CL’s feedback on the following issues: study design and logistics, Quality Systems, and some study specific questions. We recommend developing standard forms for collection of all necessary information in a single step.
We use two questionnaires: Project-Specific Pre-qualification questionnaire and General Lab Pre-qualification questionnaire. These questionnaires together contain approximately 30 questions covering all needed aspects. Any laboratory that can act as a CL can provide all requested information pretty fast (in two to three weeks).
Selection of a CL requires a cross-functional effort with participation of Quality Assurance, Project Management, and Lab Specialists. Our experience shows that the best way is to appoint a Lab Specialist as responsible for contacting potential CLs. There is no need to say that the Lab Specialist will coordinate all activities with a dedicated Project Manager and Auditor. Thus, all documents and questionnaires are prepared by Lab Specialist and are then reviewed and confirmed by all parties involved.
When potential vendors send out the completed questionnaires along with the copies of requested documents and budget, it becomes possible to perform a comparative analysis of the collected information (Stage 3) by creating a Project-Specific Lab Profile for each CL. This form contains the answers of the CL to the questionnaires together with Lab Specialist and (internal) auditor comments. Finally the CL’s capabilities regarding the project are assessed using the simple scale for each aspect of evaluation: Acceptable, Partially Acceptable (corrective actions can be implemented), or Not Acceptable.
The key point is that the evaluation of the CL for each aspect is made by Lab Specialist and auditor independently for more objective evaluation. The importance of different aspects of evaluation may change depending on the particular clinical trial. If a standard safety panel is envisaged in the trial, the most important aspect is evaluation of Quality Systems and IT, and the auditor has the leading role. On the other hand, in case of complicated Pharmacodynamic or Pharmacokinetic studies, the main aspect is the professional opinion of Lab Specialist, especially in those cases when assessments are related to the validation of a method or approach that is new for the CL.
An essential part of the profile is the lab budget and evaluation of the adequacy of the lab’s costs. When comparing budgets, it should be taken into account that the main elements of the budget can be compiled and calculated in different ways by different CLs, so the transparency of calculations is the major requirement for the budget.
While compiling a profile, Lab Specialist may ask additional questions and request clarifications regarding the budget and other provided information. The process of communication between Lab Specialist and the CL representative reflects the ability of a potential vendor to maintain communication with the sponsor at the later stages of CT; this ability is also being evaluated by the Lab Specialist. We have had situations when CLs that provided completed questionnaires on time later demonstrated lack of communication when we sent additional questions. This resulted in their being excluded from the list of potential vendors, because not obtaining answers in a timely manner during a clinical trial almost always has a negative impact on the results. The time of the follow-up communication must take no more than one month, which includes obtaining all possible clarifications, if necessary.
As a result, we develop quite a full picture about each potential vendor as well as the results of professional evaluation of their capabilities in reference to each key aspect of the project.
Based on such project-specific lab profiles, we recommend preparing the final document containing comparative evaluation of each CL’s ability to conduct the trial. We outline 10 general sections including the comparative evaluation of budgets:
- Quality Systems
- Ability to meet protocol specifications
- Logistics Services
- Data Management and IT Services
- Additional services
- Budget and Reliability of cost estimates
- Willingness to cooperate
- Previous experience (if available)
- Additional project-specific aspects
- General conclusion on lab acceptability
For CLs, the “price to quality” criterion generally reflects either the probability to lose a slightly greater number of biosamples than would otherwise be expected due to logistic or managerial inadvertence, or the likelihood of spending more effort or more time solving various organizational and managerial issues. With any CL, a certain percentage of biosamples will be lost for whatever reason (i.e. a percentage of undelivered laboratory data within the clinical trial).
According to the data provided (very reluctantly) by CLs themselves, about 2% to 9% of biosamples collected within a CT may not be processed, for a variety of reasons. This factor should be taken into account at the stage of statistical planning, especially when the clinical trial has primary or important secondary endpoints grounded up on the lab test results.
In all cases, we recommend selecting a backup CL that can be used in case of any unsolvable problems or conflicting views when negotiating the contract between the client and the primary CL, or any other force majeure. A brief summary justifying the selection is made for both the primary and backup CLs.
As far as the importance of CL selection criteria is concerned, it can be almost guaranteed that laboratories offering excessively favorable financial terms will also wind up offering a number of unpleasant surprises. Thus, the financial proposals of laboratories cannot and should not be the only criterion for determining a CL’s appropriateness for a certain project, and the final decision on selecting a laboratory should be made based on the integral assessment of all criteria.
The final CL and back-up CL choice is agreed on by the dedicated Lab Specialist and Auditor and then approved by the Project Manager. This way we believe a balance between qualification marks of a potential vendor is reached. According to our experience, the information analysis and selection of primary and back-up CLs takes a week at the most.
If needed, the CL choice document is approved by the trial sponsor. If the sponsor has questions in regard to how the evaluation scores of the laboratories’ activities have been estimated, complete project-specific profiles may be provided for all reviewed CLs. This practice allows the sponsor not only to familiarize itself with any information content it considers relevant, but also to structure this information with no loss of quality, and with much less expenditure in time and effort. According to our experience, the entire process of the selection and approval of a CL for a trial takes from four to six weeks, depending on the planned scope and complexity of laboratory assessments and the sponsor’s specific requirements. The final CL selection results and the sponsor’s approval of these results is a signal for starting contract negotiations with the selected lab.
Our suggested approach to the selection of a CL is in reality a kind of remote audit and is based upon the following principles:
- Comparative analysis of several potential vendors’ proposals made on a competitive basis;
- Impartial assessment, achieved through independent expert evaluation of documents performed by qualified Lab Specialist and Auditor;
- CL selection, based on the best combination of the following factors:
b. Availability of a satisfactory quality control system in the laboratory, including a set of regulatory documents (licenses, certificates, etc.);
c. Adequacy of the logistic solutions offered by the CL (including terms of transportation of the biosamples and proposed duration of their stability);
d. Acceptability of the budget;
e. Satisfactory level of laboratory management, also proven by previous cooperation.
Based on our experience of cooperation with different laboratories, it seems that our established process allows selection of a CL within limited timelines and at a rather low cost, meeting the requirements of all the selection process parties and minimizing the impact of the subjective factors on the final result. At least, we never had to do such a complex and laborious move as replacing a CL when a trial was in progress. That replacement however has occurred in some trials we conducted in which our sponsors had contracted CLs directly without our assistance. In all those cases, we contend, the main reason for this difficult and stressful procedure was the domination of subjective approaches in selecting a central laboratory.
- Clinical laboratory improvement amendment: www.cms.gov/ Regulations-and-Guidance/Legislation/CLIA/index.html
- CLIA related Federal Register and Code of Federal Regulation Announcements wwwn.cdc.gov/clia/docs/CMS-2226-F.htm
- ISO 15189; 2007 Medical laboratories — Particular requirements for quality and competence
- F. Leão, Jr., “The Local Central Lab Model. With globalization of trials comes the difficulties of sample logistics. Enter the central laboratory model,” Applied Clinical Trials Publish date: Apr 1, 2008
Veronika Khokhlova, Ph.D. is a senior laboratory specialist at PSI CRO AG and has 15 years experience as a senior research associate. She can be reached at email@example.com.
Andrey Karelin, Ph.D., Dr.Sci. is director laboratory support services at PSI CRO AG and a graduate of Moscow State University. He can be reached at firstname.lastname@example.org.
Maxim Belotserkovsky, M.D., Ph.D., M.B.A. is the head of medical affairs at PSI CRO AG and a board-certified physician.
He can be reached at email@example.com.
Akhil Kumar, M.D. is a medical director at PSI CRO AG and board-certified physician in medical oncology and hematology.
He can be reached at firstname.lastname@example.org