Taming ‘biggish’ data

Introduction

Data can be powerful, it can illustrate an organisation’s health; report progress; justify costs and highlight areas for improvement. However, collecting and managing large amounts of data has its challenges. Spreadsheets and paper records work for simple datasets but become increasingly hard to organise and prone to error as data demands rise.^{1 2} This report aims to highlight some of the tools and technologies available today for making sense of data in the context of a healthcare education centre.

Background

Oxford Simulation Teaching and Research (OxSTaR) is the University of Oxford’s simulation centre. Over 2500 healthcare professionals attend over 250 courses per annum. We needed a robust system to capture, safely store and report data from a rapidly growing portfolio of training and research activity. Design of a new database provided several advantages over our pre-existing methods of data management. Unlike an assortment of spreadsheets and paper files, databases allow for superior organisation of big datasets.¹ They also provide a centralised, single point of access, giving staff more control, simultaneous access to the system and improved security/backup.¹

In addition, we wanted to address the issue that our current systems lacked interoperability. Databases can provide the basis for a connected information technology (IT) system; they are designed to integrate with other applications, extending the versatility and usefulness of a data system.¹ Later in the report, we explore applications for streamlining data capture and another for creating interactive ‘data dashboards’.

Method

We trialled a customised, commercial product alongside a system designed in-house using Microsoft Access. In both cases, we sought a well-designed user interface (UI)³ with the following benefits:

• Intuitive navigation—reducing steps for data processing tasks.

• Automation—pre-populating entries with default values.

• Input control— date-pickers and dropdowns instead of free-text.

• Data validation— prompting users when data are missing or entered incorrectly.

Results

Online documentation and community support for both Access and general database design allowed an otherwise uninitiated simulation technician to create our new data management system (see figure 1). The initial build, including research and design, cost approximately £750 for 40–50 hours of technician time, over 3 months.

Figure 1.

Snapshot of Oxford Simulation Teaching and Research’s database.

As with any new system developed for bespoke needs, staff members using the system—two additional technicians, one administrator and the centre manager—spent a nominal amount of time (1–2 hours each) learning and testing the system as well as providing feedback.

Managing our data in-house with Access proved more useful than outsourcing to the commercial service. Building our own database allowed us to incorporate all our design criteria and integrated well with existing IT systems: located on our secure server, all staff have simultaneous access wherever they may be working. This flexibility is reassuring for continued development.

Using the in-house system, we found a significant reduction in workload. Time spent on data entry halved compared with our old paper and spreadsheet system, the streamlined UI reducing time per entry and preventing error,³ netting a weekly saving of 7.5 hours for our administrator.

In contrast, the commercial product was provided as a tailored, ready to use system. On the surface this seemed to be a straight-forward solution, however, communicating our needs to the development team became very time-consuming, costing our team up to 100 hours over the 6-month trial, twice that of building the better-fitting, in-house system. This led to slow development cycles and a frustrating user experience that, ultimately, never consolidated into a useful product.

Moreover, the commercial product carried a significant ongoing subscription cost whereas Access came installed on our machines, as will be the case for any organisation with a Professional licence for Microsoft Office.

We were able to improve the efficiency of our systems further by integrating a number of applications with our database. A real-time polling app, Meetoo, lets candidates use their mobile device (or spare provided) in place of paper feedback forms. They log in by scanning a quick response(QR) code displayed on a big screen, then click through questions submitting their responses direct to our database. This has improved workflow, eliminating printing and distribution of paper forms, and the error-prone transcription process afterwards.

Early trials for introducing electronic sign-in are promising too. A touchscreen monitor will greet visitors arriving at our centre, asking to confirm attendance—similar to those found in GP surgeries. We can use Access’s form builder to create a welcome screen with prepopulated visitor details to click on, or free-text boxes to fill in. The data are directly submitted to our database, again, reducing time and error compared with printing and transcribing paper forms.

Ambitions to allow guests to sign-in using their National Health Service ID card fell short after we recognised the potential data protection issues of working with another organisation’s security system.

Discussion, conclusions and recommendations

The power of our database is not only saving us time but also enabling a move towards a paperless office, with associated environmental benefits.⁴ Moreover, feedback from our candidates is very positive for the digital experience and we believe that this use of modern technology helps promote OxSTaR’s reputation as a centre for technology enhanced learning.

Access possesses powerful reporting tools. Annual reports, for example, can be preconfigured and then generated for a given period with a single click. In addition, custom reports can be designed to drill-down deep into the data for very specific queries for example, to highlight under-represented attendance for a given area or discipline, or to distinguish the number of unique visitors from repeat attendees.

While Access can stand alone, Microsoft have recently (2014) released dedicated data analytics software, Power BI, which further extends reporting capabilities. The desktop version is free, allowing users to create dynamic dashboards; reports which update in real time, and a subscription allows users to share their interactive dashboards via the web (see online supplementary video).

The ‘do-it-yourself’ approach does require continued support and development, with associated work-hours. However, this time is off-set through efficient data management and, with no middle-man, development is streamlined. Potentially, the data management system we have built may be scaled for use elsewhere. However, this does not necessarily solve the one-size-does-not-quite-fit-all issue we faced when trialling the commercial solution.

If an off-the-shelf product cannot be found to meet your requirements, especially if those are likely to evolve, our experience shows using in-house technical talent and the applications described in this report to build your own, bespoke, scalable system can be successful at relatively low cost.