As part of our continuous programme of incorporating further datasets into our repository to enhance our support for research and innovation, we are delighted to announce new data related to prescriptions on discharge from hospital as well as improved community (non-hospital) prescribing data.
Hospital Discharge Prescribing
Since 2021, a new system (Hospital Electronic Prescribing and Medicines Administration – HEPMA) has been rolled out across the NHS Lothian region, allowing individual-patient prescription data to be recorded electronically for the first time in hospitals. We have begun to incorporate these new data into repository. Once fully integrated, researchers will be able to see more directly how conditions were treated during a hospital admission and at discharge.
As an example benefit, community prescribing data have already been available electronically, but there would be a gap in information where patients were admitted to hospital. As a result, researchers would have to assume that any treatment changes were made by hospital clinicians. This might not have been true. Additionally, short courses of treatment (e.g. antibiotics for infections) would not be present in community records, because the whole supply of medications would have been provided by hospital.
Furthermore, in initial trials with the Edinburgh Cancer Informatics team, they were able to check on when patients were given opiate pain medication (such as morphine) and for how long: information which is unavailable through community prescribing data alone.
However, since a significant level of quality checking is required before the entire dataset is research ready, we have initially focused on medications prescribed on discharge from hospital. These data are now available for research and are aligned with community prescribing data to make the connections between them more intuitive for researchers.
We will continue to develop our understanding of these hospital-prescriptions data and will make more available in future repository updates.
Community Prescribing
Our community prescribing dataset holds information on prescriptions dispensed by community pharmacists in the Lothian region and paid for by the NHS. Due to the source records platform - managed by Public Health Scotland - moving to a new coding system, our pharmacy-related prescription data for the Lothian region has recently been limited to prescriptions issued before May 2023. Now, with the new system operational, researchers can access data up to August 2024 through our service.
Among our next steps, we plan to explore incorporating structured dosage data to offer greater insights to the prescription information. Currently, researchers usually make assumptions about the amount of medication people are taking based on its stated strength, the quantity dispensed, and frequency of repeat prescription. Accessing dosage data would allow daily doses and duration of treatments to be known accurately, which would greatly assist medication-related studies, such as those looking at possible side effects and other outcomes.
Standard variable sets
We are trialling the production of standard variable sets, where researchers will access the most research-useful group of variables within a dataset, but will have to justify optional variables which are either not seen as useful for most research purposes, or are more sensitive. (A variable is a category of information, such as medicine name, prescription date, etc.) Currently, we’re focusing on Hospital Discharge Prescribing and Community Prescribing (detailed above) for this trial. The result will be a simpler data-selection process: only a single justification is needed for the standard variables in each dataset, and not for each individual variable.
Helpfully, for researchers unfamiliar with these datasets, this means that they are less likely to overlook potentially helpful variables from their selection.
Accessing our Metadata Catalogue
Visit our Metadata Catalogue to see the changes for yourself. If you don’t already have access, please Connect with Us.
