Is patient data measurement and recording important in advancing healthcare studies?

The 6-month Audit of Airway Incidents during Anaesthesia ¹ found that only four of the 12 hospitals supplied body mass index kg/m⁻² (BMI) data. When supplying data from our hospital, we also found that less than half of the reports had the height recorded in the anaesthetic electronic medical record (Anaes-MR). This study investigated how electronic record keeping systems could contribute to missing data.

An application for ethics exemption based on low and negligible risk was approved by the Royal Brisbane and Women’s Hospital Ethics Committee (HREC LNR/2019/QRBW/59353 on 11.12.2019). Height and weight were extracted from scanned images of handwritten forms completed in the anaesthetic preadmission clinic (PreAC) and compared with the electronic reports in the Anaes-MR.

Two samples were taken. The first sample period was in November 2018, when 284 records were examined, and the second was in November 2019, when 242 records were examined. In the 2018 sample of the Anaes-MR height was recorded in 26.8% of the cases, and in the 2019 sample height was recorded in 41.7% of cases. The completion rate was higher in the PreAC record, with height recorded in 83.5% (2018) and 77.7% (2019). The relative ratio for the recording of height between Anaes-MR reporting rates to the hospital medical record keeping system (Hosp-MR) was 0.32 in the 2018 sample and 0.54 in the 2019 sample. A similar pattern was noted for the reported percentage for weight. In the 2018 sample, weight was recorded in 46.8% of cases in the Anaes-MR and 92.3% of cases in the PreAC record, with a relative ratio of 0.51. In the 2019 sample, weight was recorded in 66.5% of the Anaes-MR data and 89.7% of the PreAC record data, with a relative ratio of 0.74.

This study shows that the required data to calculate BMI were only recorded in less than half of cases in both the 2018 and 2019 samples. Recording height appeared to be the rate limiting factor, with only 27% of records completed for this parameter in 2018. Weight was recorded in 47% of cases, but as height was missing in 73%, BMI was only able to be calculated in 27% of cases overall. A similar pattern was seen in 2019, with height recorded in 41.7% and weight recorded in 66.5%, which despite a modest improvement, only allowed the calculation of BMI in less than half (41.7%) of cases. In contrast, in the preadmission clinic, height was recorded in approximately 80% and weight in about 90% of cases in both the 2018 and the 2019 samples, which meant calculation of BMI was possible in around 80% of cases for both samples. However, each preadmission clinic consultation was handwritten on a paper form, which was subsequently scanned as an image into the Hosp-MR. As the Hosp-MR stores the data as scanned images of free form handwriting, the task of extracting large amounts of data for height or weight would be challenging. We assert that for any future study, the data would need to be extracted by reading each document and manually transcribing the handwritten data into an electronic format such as an Excel spreadsheet. The latter process we found to be slow, required careful checking and, therefore, propose that it would be unsuitable for extracting large quantities of data for research studies or for managerial purposes without incurring additional labour costs.

A patient’s BMI is regarded by the authors as important for preoperative risk assessment, management of anaesthesia and peri-operative management. The reporting rate for height and weight could potentially be improved by a single fully integrated electronic recording system, in which the data are recorded at the time of the preadmission consultation, when it is also potentially easier to record height and weight. A single integrated electronic recording system has the benefit of avoiding duplication of data entry, thus saving time, cost of labour and the possibility of error during the duplication process. We assert that data quality is important for the official record of anaesthesia, for managerial tasks, as well as for epidemiological research purposes. However, we also accept that for a variety of reasons it might not be possible to determine height and weight in all patients.

Finally, based on the findings in our hospital, we recommend that institutions considering the future purchase of an electronic record keeping system consider a system that stores all data digitally, rather than scanned images of handwritten paper records. We believe that this will assist data extraction, which we assert is important in advancing healthcare studies.