The simple solution to quality record keeping

Day-to-day databases of orthopaedic surgeons are increasingly important not just for record keeping but for research and outcome reports. As a relatively intuitive task, documentation of the details and procedures for entering patient information are often overlooked. Establishing clear rules for data entry can help you avoid common database errors and save time.

Orthopaedic surgeons collect patient data constantly to determine the best treatment options for their patients and evaluate their outcomes. In an increasingly data-driven world, the potential applications of this information are becoming evident. Orthopaedic journals are placing higher value on registry-based research as a midway point between controlled scientific research and real-life clinical practice outcomes (Kaeding, 2018). This has prompted orthopaedic surgeons to question how the data they have collected could be published.

The problem is that research has not always been the initial intention of data collection. Serving more as a record keeping exercise, time consuming administrative processes are likely to have come second place. For these reasons, orthopaedic registries may not be at the level to provide usable and high quality data for research reporting.

In this article, we have put together suggestions to overcome some of the caveats associated with record keeping:

1. Be clear on what information gets entered into which systems

Many orthopaedic practices use multiple databases in parallel to cover different needs. In itself, this is a practical solution to overcome limited software functionality. The issues arise when it is not clear what information should be entered into each program. Overlap in fields creates risk of information duplication, wasting time and resources. Ironically, it also increases the likelihood of losing information, as administrative personnel may wrongly assume data has been entered in another database and not wish to duplicate, thereby resulting in gaps in the records.

2. Establish standard response requirements for free text fields

Lack of standardised rules for filling in free text information is a related but distinct issue. Approaches to filling in free text fields can be inconsistent across individuals and time, as what is considered important varies. This increases the difficulty and time required to interpret and analyse rich qualitative information.

3. Create a unique patient ID that can be commonly used across all systems

The above problems are compounded if there is no clear and consistent patient identification marker across programs. While the person entering the record at the time may know the details of a particular patient’s case, this information can be lost over time. Without a patient ID, it becomes difficult to match records across databases, especially for those with common names. This may also undermine the professionalism of the clinical practice, as the surgeon trying to use the record may have to request or clarify information (which should be known) from the patient.

4. Establish a routine for quality assurance checks, and specify how frequently these should be performed

This is imperative for maintaining data quality over time. Many of the gaps in databases result from minor oversights, which can easily be solved by establishing clear processes and quality assurance checks. Performing data checks by cross-matching, verifying and validating information in the databases ensures a clean and reliable dataset. Performing these checks at specified intervals ensures that inconsistencies or gaps in the data are flagged earlier, and are thus easier to rectify or back-fill. A good quality dataset makes for easier analysis and quicker reporting.

The above suggestions are not just for large administrative teams, even small clinics can establish and benefit from good record keeping practices. All teams experience changes in staff and resources over time, and with it come differences in perspective and interpretation. While this brings potential for innovation and improvement, it can inherently result in inconsistent data entry if there are no clearly documented processes and rationale provided at handover.

References

Kaeding, C. C. (2018). Editorial Commentary: Registries, Prospective Cohorts, and Predictors of Outcomes: Why Bother? Arthroscopy: The Journal of Arthroscopic & Related Surgery: Official Publication of the Arthroscopy Association of North America and the International Arthroscopy Association, 34(8), 2485–2486.