Improving Central Monitoring and Data Management
Monitoring of clinical trials is crucial to ensure patient safety and data integrity, however monitoring activities can be very resource-intensive.
Our clinical trial monitoring toolkit provides a range of resources designed to help the clinical trials community improve how they design, implement and report monitoring activities.
How to use the toolkit
This toolkit contains research publications, guidance documents, training materials and templates to encourage the efficient conduct of clinical trial monitoring. Some material will need adapting to your SOPs. Please acknowledge our work as indicated with each tool.
Click the headings below to find more information and links to useful resources.
Data cleaning
Data cleaning addresses problems with data such as incomplete, invalid or inconsistent data. When data are entered, most databases have some automated checking of data and flagging of problems. On a regular basis or maybe before data monitoring committee (DMC) meetings, central trial team members run checks on the participant data and query any strange or required values with sites. Before any interim or final analysis, these processes will be repeated. These are all data cleaning activities. They happen often in the course of a trial. The main action is sending out data clarification requests.
Central monitoring
This is the monitoring performed in a location away from the investigator research site and often at a CTU/Sponsor office. Central monitoring is looking to centrally identify any issues with trial conduct such as inadequate processes or procedures not being followed through a lack of clarity in the protocol or active fraud.
Looking through centrally held data by site, to discover odd patterns or features in the site’s data (e.g. missing treatment data) or unacceptable data activity (e.g. digit preference in white blood cell level), during the trial, at times specified in the trial’s trial monitoring plan, is best called ‘central monitoring’. This may result in data queries to sites or may provoke dedicated communication with sites or an on-site monitoring visit.
The central monitoring should be targeted at risks to the trial patients or the trial. Central monitoring results are an indicator of the quality of a trial and show due diligence. Any issues found during central monitoring should be followed up by contacting the site and may also result in actions such as the delivery of (re)training or the making of an on-site visit.
Central monitoring need only be repeated periodically, the period depending on trial parameters such as the duration of treatment and recruitment rate and on the assessment of risk. Sometimes central monitoring is done across sites, comparing data between sites to show differences. In some instances, this may be done across trials run from the same organisation. Central monitoring can include review of trial management data such as records of protocol deviations.
On-site monitoring
This is the monitoring performed at research sites at which the clinical trial is being conducted, via a physical visit by appropriately trained individuals from the Sponsor and/or its delegated representatives. It requires access to medical records and other source documents of trial participants for the purposes of protecting the rights, safety and well-being of patients, Source Data Verification (SDV)/Source Data Review (SDR), to confirm the accuracy of data transcription, compliance with the protocol, GCP and applicable regulatory requirements and verification of the existence of participants.
Remote monitoring: This is the remote evaluation performed by appropriately trained individuals from the Sponsor and/or its representatives, at a location remote from the investigator research site and which replicate some on-site activities. It may include documentation being sent to the central office (with appropriate encryption and consent for any document including patient identifiers) to enable a number of checks to be performed. Remote monitoring can also be conducted by review of site self-completed monitoring checklists, telephone/video monitoring calls with screen sharing or those performing monitoring activities having direct access to trial participants’ electronic medical records and electronic site master files.
Metric
Metrics are numeric measurements, mostly obtained and calculated from data held in the trial database, that are used to evaluate a sites’ risk or performance.
Threshold
Metrics are compared with acceptability thresholds to highlight and assess potential or actual risks and/or under performance.
Trigger
The comparison between metric values and acceptability thresholds can also be called a trigger.
Love SB, Yorke-Edwards V, Diaz-Montana C, et al. Making a distinction between data cleaning and central monitoring in clinical trials. Clinical Trials. 2021;18(3):386-388.
https://doi.org/10.1177/1740774520976617
Data cleaning and clinical trial monitoring are intertwined to the possible detriment of clinical trial conduct. This paper concludes that “it is important to correctly define data cleaning and central monitoring in order to communicate the conduct of a trial, to ensure adequate risks mitigation and to ensure that the data are appropriately corrected.”
The paper contains definitions of data cleaning and clinical trial monitoring and a table noting some similarities and differences.
This monitoring handbook was written by the UKCRC Registered Clinical Trials Units (CTU) Network Task and Finish Monitoring group and reviewed by the Medicines and Healthcare products Regulatory Agency
(MHRA). It is aimed at academic trialists undertaking monitoring activities such as on-site, remote and central monitoring. It is intended as a resource to be used to support initial training and as a point of reference thereafter.
The handbook provides general information about monitoring, which should be supported by training on CTU/Sponsor specific SOPs, guidance documents and templates. This handbook provides information on the theory of monitoring, tips on conduct and real-life examples that may help to explain and support possible monitoring approaches and their application.
The UKCRC Registered CTU Network Task and Finish Monitoring group has prepared a set of four training module webinars to provide context and understanding of the key clinical trial monitoring principles, together with practical examples.
As well as an introduction to monitoring, there is a session on on-site monitoring, one on remote monitoring and one on central monitoring. Including doing the exercises, each session should take about an hour to complete.
- An introduction to the principles of monitoring
Emma Laing (NHS Blood & Transplant CTU)
This module is an introduction to the monitoring activities in clinical trials. It covers the purpose, explains risk-based monitoring, touches on the different types of monitoring (on-site, remote and central), discusses the frequency of monitoring and explains the roles and responsibilities of a person involved in clinical trial monitoring. - On-site monitoring
Emily Dight (Warwick Clinical Trials Unit) and Brian Finucane (Cambridge Clinical Trials Unit)
This module is an introduction to on-site monitoring. It covers what data should be checked, how to prepare conduct and conclude an on-site monitoring visit and how to communicate with the site. - Remote monitoring
Patricia Rafferty (Northern Ireland Clinical Trials Unit)
This module is an introduction to remote monitoring. It describes the variety of methods and their advantages/disadvantages, covers how to select the best method for your trial, and describes the preparation and completion of remote monitoring. - Central monitoring
Heather Cook (Exeter Clinical Trials Unit) and Sharon Love (MRC Clinical Trials Unit at UCL)
This module is an introduction to central monitoring. It explains what central monitoring is, why its useful, how to conduct central monitoring, how you select the frequency, what to do if you find something worrisome and how it links to external trial committees.
Love SB, Yorke-Edwards V, Ward E, et al. What is the purpose of clinical trial monitoring? Trials. 2022;23:836.
https://doi.org/10.1186/s13063-022-06763-2
The sources of information on clinical trial monitoring often do not give information in an accessible language and do not give detailed guidance. In a three-step process, we synthesised materials from sources that describe clinical trial monitoring into principles of monitoring. In this paper, the principles are clearly expressed in lay terms.
Hsieh S-F, Yorke-Edwards V, Murray ML, Diaz-Montana C, Love SB, Sydes MR. Lack of transparent reporting of trial monitoring approaches in randomised controlled trials: A systematic review of contemporary protocol papers. Clinical Trials. 2023;20(2):121-132.
https://doi.org/10.1177/17407745221143449
Monitoring should be briefly described in protocols. This is a systematic review of 811 published protocol papers for randomised controlled trials, extracting information on whether they mentioned clinical trial monitoring and if they do, what information they give.
Love SB, Armstrong E, Bayliss C, et al. Monitoring advances including consent: learning from COVID-19 trials and other trials running in UKCRC registered clinical trials units during the pandemic. Trials. 2021;22:279.
https://doi.org/10.1186/s13063-021-05225-5
This paper describes informed consent and monitoring without going on-site. These advances were provoked by the COVID-19 pandemic.
Love SB, Yorke-Edwards V, Lensen S, et al. Monitoring in practice – How are UK academic clinical trials monitored? A survey. Trials 2020;21:59.
https://doi.org/10.1186/s13063-019-3976-1
This paper describes the results of a survey to find out how clinical trial monitoring is carried out in the UKCRC registered clinical trials units.
Yorke-Edwards V, Diaz-Montana C, Murray ML, Sydes MR, Love SB. Monitoring metrics over time: Why clinical trialists need to systematically collect site performance metrics. Research Methods in Medicine & Health Sciences. 2022;0(0).
https://doi.org/10.1177/26320843221147855
Monitoring metrics are used at distinct timepoints to make decisions on the quality of the data. This paper investigates whether we can learn from looking at metric levels across the duration of the trial.
The paper explains relevant background information and provides an example of how to look at clinical trial monitoring metrics as outlined in SWAT 167. Monitoring metrics have been used for the last two decades without much work looking at how they perform. This method needs to be followed in multiple trials to increase the understanding of metrics and further develop of the use of metrics.
Wyman Engen N, Huppler Hullsiek K, Belloso WH, et al. A randomized evaluation of on-site monitoring nested in a multinational randomized trial. Clinical Trials. 2020;17(1):3-14.
https://doi.org/10.1177/1740774519881616
This paper reports on a trial which randomised sites to on-site or no on-site monitoring, to evaluate the value of on-site monitoring.
Cragg WJ, Hurley C, Yorke-Edwards V, Stenning SP. Dynamic methods for ongoing assessment of site-level risk in risk-based monitoring of clinical trials: A scoping review. Clinical Trials. 2021;18(2):245-259.
https://doi.org/10.1177/1740774520976561
This scoping review found 30 papers describing the methods for central monitoring.
Written by the UKCRC Registered Clinical Trials Units Network Task and Finish Monitoring group, this document contains information about remote monitoring and advice for minimising the burden to site staff.
Sydes MR, Wong WK, Bakhai A, Joffe N, Love SB. Protecting blinded trials in electronic hospital systems. Clinical Trials. 2022;19(2):231-233.
https://doi.org/10.1177/17407745211069985
With the increasing use of electronic health record systems in hospitals, care needs to be taken to avoid inadvertent unblinding of hospital staff or participants of trials. This paper gives a table of actions for those running blinded trials to ensure the trial stays appropriately blinded.
- A collaboration is writing a protocol deviation framework to form the basis of a common infrastructure for collecting, processing and reporting protocol deviations.
- A collaboration is writing a template risk assessment template containing the essential contents along with prompts.
- A collaboration is creating a metrics spreadsheet, indicating what you should think about when creating metrics to mitigate the risks in your trials. The spreadsheet will also give examples of metrics and their associated metrics and contain case studies for 4 trials giving the metrics they used and why.