The Preterm Clinical Network (PCN) Database: a web-based systematic method of collecting data on the care of women at risk of preterm birth

Prior to designing the database the UKPCN identified main principles, intended scope and key features. It was determined that the database should be designed to allow clinicians to easily audit their own practice, with an added facility to combine data for shared audit, service evaluation and research. In order to minimize the additional time burden for already busy clinicians, the design needed to cater for quick and easy entry of an agreed minimal dataset, including risk factors, surveillance methods, interventions and outcomes. The database also needed to allow for the addition of more detailed data, should it be required and when resources were available.

The vision was to collect data prospectively in preterm clinics, with outcomes added later. However, it was decided that the database should also have the flexibility to store historical data. This was considered important as many UKPCN preterm clinics were established and there was a wealth of data already collected for internal audit and service evaluation purposes. Combining these data could immediately provide a valuable source of material for investigation. Flexibility to accommodate data collection for small research projects not warranting a unique database was also considered to be an important component. Security, user-friendliness and accessibility were considered of vital importance to the project, and it was decided that an internet-based platform would be the most suitable.

MedSciNet, a Swedish based company, were contracted to develop the application [11]. The Preterm Clinical Network (PCN) database and a web-based application to access the data was built using the MedSciNet Clinical Trial Framework (CTF), a self-contained environment that enables development, hosting, support and management of individual web-based solutions for clinical trials and studies, quality registries, medical biobanks and other required solutions within the field of academic medicine. The databases conform to relevant FDA, NIH and HL7 standards guidelines and recommendations. Microsoft.NET Framework and Microsoft SQL Server technologies were used for the platform development. A test database was created, piloted and refined after piloting and to incorporate adjustments recommended by the Research Ethics Committee. The data schema is described in Table 1.

Initially, permission was sought through NHS Trust Information Governance departments. However, there was no UK wide system in place for Information Governance regulation for the storage and sharing of clinical data and this proved to be time consuming and laborious. A decision was taken, therefore, to redefine the database as a “Research Database” and apply for Research Ethics Committee (REC) review through the UK Integrated Research Application System (REC Reference 16/ES/0093). REC approval is not mandatory for research databases in the UK, but the process allows for independent review of the protocol, procedures and security of the data, which is acceptable to NHS Trust Information Governance departments and future publishers of papers presenting the findings from projects investigating the data. Although local R & D department approval is not required by the REC, they are informed because they may wish to review their site’s capacity for collaborating in the project. They are also involved with the data sharing agreements which are required before the release of pooled data for approved research projects.

Prospective data collection, i.e. at the time of the clinic appointment, requires the recording of patient identifiers for later follow up of pregnancy outcomes. We were advised by our NHS Trust Information Governance department that written patient consent would be necessary because: a) the data would not be completely anonymous, and b) it is feasible to obtain consent from current patients. Obtaining patient consent also provides the opportunity to seek permission for storage of their baby’s NHS number and potentially long term follow up of the child.

Identifiable data has been kept to a minimum (i.e. initials, date of birth, hospital number, NHS number) and transferred, at registration, to a separate but linked Patient Details Database. On the main database, date of birth is converted to age (at expected date of delivery) and postcode converted to a marker of socio-economic status. Data collection centres outside the UK use alternative socio-economic indicators.

For historically collected data with known pregnancy outcomes Information Governance advised that patient consent would not be required. This was because historical data could be anonymised before transfer into the database, and because it would be impractical to consent women following discharge from maternity care.

Data are entered locally at Data Collection Centres under the supervision and direction of the local Lead collaborators. Figure 1 outlines the flow of data through the project. The single “Clinic Record” page captures the minimal dataset which includes demographics, referral criteria, surveillance methods, test results, preterm interventions and outcomes (Fig. 2). This record has been designed to include the COPOP core outcomes for preterm birth intervention studies [12]. Further information can be entered on additional, non-mandatory forms, e.g. detailed data on risk factors, medical and obstetric history, antenatal and delivery details, and other neonatal outcomes and health utilisation.

The Project Lead, Project Manager and Database Access Committee Chair (named authors of this paper) are all based at the Department of Women and Children’s Health, School of Life Course Sciences, King’s College London, UK. The local lead collaborators are responsible for the delegation of data entry and monitoring duties to local site users, and for the preparation of their own historical data for transfer. The database has an inbuilt data monitoring facility, whereby locally appointed site monitors can monitor data, raise queries and “lock” data. Historical data is prepared for transfer and can be either entered manually, or imported via an Excel spreadsheet template.

Funding for the initial design and development of the database was covered by an NHS Innovations Challenge Prize Fund, won by the Project Lead Professor Andrew Shennan in 2013 for his work in his preterm surveillance clinic. Tommy’s charity, who support the work of the Department of Women and Children’s Health, are contributing towards the ongoing support and maintenance.

Local collaborators can investigate their own data freely, but researchers requiring access to data from other sites are required to submit an application to the Database Access Committee. The PCN Database Project Application Form allows applicants to provide details of their proposed project, and incorporates an “Applicant Agreement” and “Data Requested Form”. Applications are welcomed from UKPCN members and other established research groups. Projects may be based on a variety of study designs, for example: case-controlled studies designed to examine the relationship between risk factors or interventions and outcomes; cross sectional surveys; cohort studies and sample size calculations. The inclusion of COPOP core outcomes will facilitate comparison and combination of PCN data with other studies [13]. Project applications are reviewed by the Database Access Committee, which is comprised of members of the UKPCN, the RCOG Preterm Clinical Study Group [14], and the Guy’s and St Thomas’ Women’s Health Academic Centre’s preterm birth studies patient and public involvement (PPI) panel. The PCN Database Access Committee will scrutinize the quality of the proposal and academic team and will also consider whether further Research Ethics Committee approval is required.

All specific research project timeframes are finite in nature, and most will have relatively short term follow up. Interventions used in pregnancy could have long term, as well as short term, effects on the child and could, potentially, persist into adulthood. A major advantage of the PCN Database is that it also acts as a registry of children who have been born to women at risk, who have undergone specialist preterm surveillance and who may have had preterm interventions, whether born prematurely or not. This offers researchers the possibility of investigating much longer term outcomes of the care of women attending preterm clinics. Patients are given the opportunity to opt in, or out, of longer term follow up, and are asked to indicate their specific consent to the storing of a unique identifier for the child’s health records (in the UK, the NHS number). Additional information collected on the children’s health beyond the initial neonatal outcomes will be determined according to the requirements of the proposed study. The PCN Database Access Committee review will include consideration of the need for further REC approval as well as the method of acquiring the necessary data, e.g. GP or other health records or linked databases, or direct approach to the mother (e.g. for questionnaire completion or face-to-face appointment for developmental assessment of the child). Investigators requesting data for follow up studies involving children beyond 16 years of age will be required to seek individual written consent of the said children. The children will be approached by the PCN Database project team via a communication sent to the most current address held by NHS care records.

Further developments will include using the database for data collection for small intervention studies, such as those managed by the Preterm Trials Consortium, a UK wide survey of maternity care provision for women at risk of preterm birth and patient experience studies. Other plans include future linkage with large population datasets so that comparisons can be made with the wider population, including low risk childbearing women.

Clinicians interested in registering as Data Collection Centres, both from the UK and outside, are invited to contact the corresponding author for more information.

A public area of the website [15] serves as a source of information about the UK Preterm Clinical Network. It is used as a resource for the public, potential new UKPCN members, potential new Data Collection Centres and other health professionals. Useful links and publications resulting from projects using data from the database will be available on this part of the site.