Standardised equipment and centralised spirometry for optimal data quality
pharmafile | April 3, 2012 | Feature | Research and Development |
Introduction
When developing the treatments for asthma and COPD, pharmaceutical companies utilise pulmonary function testing (PFT) to determine each compound’s efficacy. The main pulmonary function test utilised in asthma and COPD research is a basic spirometry test.
Developing treatments to cure, prevent or provide palliative effect to sufferers of respiratory disease involves a significant investment, however costs can be decreased by making sure that high quality data is collected during clinical trials. By using centralised spirometry services, sponsors can benefit from increased consistency of processes and reduced data variability, resulting in fewer patients, increased statistical power and reduced costs.
Spirometry testing issues
Spirometry testing involves the co-operation of the subject to give a maximal exhalation in order to obtain good quality data. In 2005, the American Thoracic Society (ATS) and European Respiratory Society (ERS) released guidelines on how to collect and interpret spirometry data. In line with these guidelines, it is vital that pharmaceutical companies decrease data variability to facilitate the detection of true clinical signals in all treatments being tested. Increased data variability can occur for a multitude of reasons, with lack of standardisation due to the use of a decentralised approach cited as a key one. A decentralised model is typically defined as the collection of data across multiple investigator sites using local instrumentation and varying processes.
Problems with non-centralised spirometry
While the majority of manufacturers produce spirometers that meet the ATS/ERS guidelines for devices, there are still varying levels of acceptability within the criteria. Due to this, using equipment from multiple manufacturers increases the risk of data variability, potentially masking a treatment’s effect and efficacy.
In line with the ATS/ERS standards, it is important that feedback about the quality of the collected data is provided to technicians during the measurement. The feedback can take the form of a simple message on data errors, or an elaborate animation designed to encourage the subject to perform a proper measurement. Problems arise in clinical research when the criteria for feedback is different. For example, equipment A may use late time to peak expiratory flow to determine subject effort and equipment B may use a different criteria, resulting in inconsistent assessments.
Reference equations are another problematic area, when using a non-centralised approach. Inclusion of proper subjects in a clinical trial may be established on a reference value from a prediction equation, with the inclusion of improper subjects deemed as a risk to the efficacy of a trial. Subjects who meet a certain percentage of prediction equations will be allowed to be included within a study.
However, there are multiple prediction equations available for spirometry testing from various authors and years of publication, providing great scope for inconsistency. As a result, a trial can be vulnerable to decreased efficacy. Additionally, non-standardised studies often utilise multiple measurement techniques depending on individual technician training, again increasing data variability.
In most non-centralised studies, data is not reviewed for quality either during or at the end of the study. Without continuous data quality review throughout the course of the study, indications of variability or compromised quality can not be detected and solved. In addition, when using a non-centralised approach, data are manually entered into an ECRF application without standard quality statements, leaving it vulnerable to the risk of transcription errors.
The benefits of standardisation and centralisation
A centralised respiratory model is defined by two elements: standardised equipment and centralised spirometry. In order to standardise on equipment, each site receives identical equipment using the same study protocol workflows from a specialist vendor. In that way, the risk of inter-equipment variability is reduced, while protocol violations are minimised using criteria that restrict users to specific time points for data collection. Additionally, the same subject inclusion prediction equations can be programmed into the devices, eliminating erroneous subject inclusion or exclusion. Equipment standardisation also allows for consistent feedback on data quality and standardised training, ensuring that all data are collected utilising the same techniques.
To ensure standardisation, spirometry data are electronically transferred to a central database for review by a centralised OverRead service.
This eliminates any transcription errors associated with manual data entry, while also allowing centralised review of data by a team of expert reviewers based on a combination of the ATS/ERS standards and pharmaceutical company specifications. Proactive review of sites can help reduce large volumes of poor quality data by indicating poor performance at an early stage and highlighting the need for improvement.
Conclusion
In respiratory drug studies, poor data quality produces inconclusive results, wasting a sponsor’s time and money and ultimately preventing the release of a promising compound or, in the worst case, leading to the release of a compound that is harmful to patients. To ensure maximum accuracy of results when developing new compounds, the use of standardised equipment and centralised spirometry provided by a specialist vendor with experience in collecting accurate health outcomes data is recommended.
About ERT
ERT (www.ert.com) is a global technology-driven provider of clinical services and customisable medical devices to biopharmaceutical and healthcare organisations.
By James Sowash – Director, Respiratory OverRead, ERT Inc.
Visit www.ert.com
