Quality assurance is important in all types of clinical research study because assuring the quality insures that the data collected is accurate and reliable, that the participants have been safe and that the research has been conducted ethically. In terms of data quality it is very important to know that the key data points (those that affect the outcome of the study) have been collected correctly and consistently on every occasion throughout the study.

ICH GCP states that “the sponsor is responsible for implementing and maintaining quality assurance and quality controls with written SOPs to ensure that trials conducted and data generated, are documented (recorded), and reported in compliance with the protocol, GCP, and the applicable regulatory requirements”.

The objectives of QA procedures therefore are to assure the accuracy and consistency of study data, from the original observations through to the reporting of results and to ensure that study results are considered valid and credible within the scientific and clinical communities.
Maintaining accuracy and quality should therefore be a continuous and dynamic process. Although study requirements are carefully set forth in detailed documents such as the protocol, it is expected that the accompanying project plans can change during a study. This ongoing process needs to be documented and the revisions communicated to all investigators and support staff.

There are basic processes that help to ensure quality and maintain consistency within a clinical research environment such as standard operating procedures (SOPs). Monitoring is also a quality assurance tool, and there needs to be consideration as to whether the risk and compexity of the research suggest that on-site monitoring would help support the study. Database level checking is also a helpful QA tool and worth building in to all types of research study data management system.

SOPs carefully describe every step of a study to fully explain the process and ensure that they can be repeated the same way by different people, between different participants and at different points in time (see the guidance document on SOPS {link}). SOPs are also important in training new study team members.

Skills sharing and training is a good approach maintaining quality throughout a study. Study specific training with everyone involed with the study should be comprehensively planned. The aim should be to ensure that everyone understands why the study is being undertaken, where their role fits in and how vital everyone’s activities are in ensuring that the study question is being answered accurately, safely and ethically – and that this answer could be impactful for many thousands of future patients. Thereby everyone should feel part of the team, ownership of the study and its processes, and responsible for the quality.

Processes to Improve Data Quality
The general purpose of a research study is to collect data to test the study hypothesis and achieve the study objectives. To assist in preventing problems with data quality, everyone must follow and adhere to an approved, common protocol to assure that data collected can be aggregated for analysis and interpretation.

Case Report Forms (CRFs) and Data Collection
CRFs are used to collect and document all the study data points (and on occastion other study observations) in a standardised manner and at specific time points over the course of a study as set out in the protocol. Forms are usually organised by study visit to assist with data collection and should correspond exactly with the database, and later the data analysis plans. Data collected for a study should be limited to those key data points that will answer the study questions. Some studies attempt to collect data that are nice to have, but are not necessary to answer the study hypotheses. Pre-testing study forms on several participants before a study commences allows the study staff to identify and correct problems, including content or format. It is the responsibility of the site staff to ensure that completed forms are accurate and complete and that subjects are adequately followed up even if treatment is discontinued prematurely.

The typical process is that the key data is recorded as source data. This is the original form of the information as it is captured, and could consist of observations in the clinical notes, x-rays, laboratory results (usually printouts) or clinic diaries. This should happen in real time, by whoever is responsible for the clinical care (or laboratory analysis) and is therefore called the ‘source data’. Later the key data points are transcribed from the source data to the case record form, either by the same person of another designated member of the study team. If the study is monitored, then a percentage of the information recorded in the case record form is verfied against the source data.

This process can be the same whether the CRFs are paper or digital. If electronic data capture is being used then sometimes there is no other source data, and the memory card can be used as the source data. Many studies however will have source data available because it is collected as standard clinical care. Often, in situations where there are no, or insubstantial, clinical notes, then generating source data forms is highly beneficial to all. In many circumstances the source data forms generated for studies in rural developing countries studies have been taken up and become medical note templates, which has given wider benefit to health in those settings.

A risk and complexity assessment of the study will help guide what is approprioate in terms of what is source data in a specific situation, whether source data forms for the study would help improve quality by ensuring the data is recorded somewhere and the accuracy can later be checked. A risk and complexity analysis will also guide what is required in terms of monitoring and data checking. Some data checks can happen at the database level and others require on-site monitoring to check the entries against the source data and also to observe how these data are collected and handled.

On-site monitoring is very beneficial and should be seen as a positive process where the monitors are experienced in the disease area, type of study and the community. Monitoring visits should be about supporting the study staff with a mentoring or supporting approach. However, not every clinical resarch study or indeed trial needs on-site monitoring. Indeed this is clearly stated in ICH-GCP where the guidance says that monitoring should be dependent upon the nature of the study. Therefore many low-risk trials and studies do not require monitoring visits and others very minimal oversight. An appropriate monitoring plan can be put in place by discussion between the sponsor and the investigators (who may be one and the same in some situations). In-house monitoring is a perfectly acceptable approach, where a research centre trains some of their staff to monitor their studies. Another successful approach being used more widely, and developed in East Africa, is reciprocal monitoring where studies teams train one or more of their team members to monitor and they spend a small portion of their time monitoring other people’s studies, either in the same research organisation or perhaps in the same city, district, region or country. This is a highly effective model because everyone gains by sharing experience and knowledge and very high standard monitoring can occur, and of course the costs are minimal, negating the need for outside contractors.

Data Management
The data management plan describes the flow of data from collection through to data analysis. Error identification and resolution should be specified in the plan including steps to derive an analytic database from edited to "clean" records. With electronic data capture becoming more and more popular, it is common to have in-built edit checks to fire queries for out of range or missing data. Electronic data capture now ensures that data can be viewed in real time, but it is important that there is documented process of data validation and verification to ensure their accuracy and integrity. Some of the checks that will need to be done are as follows:

- Ensuring that data generated during the study reflect what is specified in the protocol
- Comparing data in the CRF and data collected in source documents for accuracy
- Ensuring that the data analyzed are the data recorded in the CRF
- Data presented in tables, listings, and graphs correctly match data in the database.
- Data reported in the clinical study report are the data analysed

Choice of software for data management is important and this is covered in the data management guidance notes and in several other places across the Global Health Network (links) including (databox). Software used for data capture, management and storage needs to have certain features such as an audit trail and security. Again this is described elsewhere (link). In brief, tools such as Excel and Access are not robust enough but suitable systems are available to all that are open access and easy to use. Good data management systems are key to your study being able to accurately answer your question and for others, such as publishers and funders to trust your results.

Quality Control (QC) and Quality Assurance (QA) Activities

Quality Correction Procedures
Inevitably, errors, discrepancies and non-compliance do occur over the duration of a study, and the purpose of a Corrective And Preventative Action (CAPA) plan is to ensure that these issues are visible, prioritized, and tracked, and that the root cause is determined and resolved. It also provides a system to track issues of non-conformity that have not been resolved. This process includes not just the person responsible for defining and implementing the corrective action but also how errors identified during the trial will be managed. For instance:

- Data - changes to forms should be documented in ink with a single line drawn through the incorrect item and the correct response noted, along with the date and the initials of the person making the correction. For data stored in a computer, there should be an audit trail of all changes.

- Systematic Errors - Systematic errors may result in changes to the protocol and working document For example, in global studies, it is not uncommon for various labs to have differing units for lab test. If there isn’t the option to choose from a selection of units, the database may need to be updated to accommodate this change. Data should be tracked throughout the study and where repeated systemic errors have been identified, such sites should be retrained before they are allowed to further enrol subjects to the study.

- Detection Processes - Reports and procedures may require modification. For example, if recruitment is slow in a study, the coordinating center may request that sites fax or email their screening logs. A new report may be developed that captures screen failures across sites to better understand if there is a need to change study procedures.

- Protocol Violations - it is often good practice to identify up front, preferably prior to recruitment; what types of protocol violations will constitute serious or non-serious events. Ideally, violations that impact on patient safety or efficacy should be considered serious. Other events that are in breach of GCP, national regulatory requirements or could skew the interpretation of results should also be deemed serious.

Conclusion
Having good quality management processes doesn’t stop with developing a plan but it involves a critical introspective process to continually monitor, evaluate and to improve all study processes. This continual process of improvement tracks and reports metrics for key activities and steps, keeping in mind the adage that "what gets measured gets managed." Other inputs to process improvement include a formal debriefing after study close, which can be really beneficial to others and to the study team.

References

Martin Valenia “Quality Control and Assurance” Applied Clinical Trials Sept 2010

National Institute of Health - National institute of neurological disorders and stroke: Quality Assurance Guidelines April 13
http://www.ninds.nih.gov/research/clinical_research/policies/quality_assurance_guidelines.htm

The WWARN External Quality Assurance team are working closely with 48 laboratories from 26 different countries to improve the analysis of antimalarial drug regimens during clinical trials, and in turn, enhancing the overall quality of the research data captured.

The team are working with laboratories in two different ways; they are providing in vitro, pharmacology and drug quality testing laboratories with certified drug reference standards, and designing a proficiency testing programme to help participating pharmacology laboratories improve the quality of their results. Find out more here.