Abstract
Introduction Communicating the risks associated with a procedure is critical to the principle of informed consent. While clinicians typically use risk rates that are appropriate to the patient being consented, when there is a lack of published data or the patient has rare or complex disease, this can be challenging. GOSH has developed PICTURE, a clinical informatics platform allowing efficient analysis of EHR data from a flexibly defined cohort of previous patients. In this study we explore how PICTURE can produce provider-specific information about the prevalence of complications.
Methods We used PICTURE to define a cohort of patients who had a repair of defect of interventricular septum with a recorded procedure of either OPCS-4 K11.2 or K11.9. Based on this cohort, we used the frequency analysis methods within PICTURE to compute the prevalence of diagnoses and procedures recorded for patients post-operatively.
Results PICTURE provided data for n=315 cardiac repair surgery patients who had their repair between May 2019 and September 2022. This cohort contained 44.1% Female (55.9% Male) patients with a mean age of 0.40 years (IQR 0.20-0.84 years). Results showed that the two most common newly acquired post-operative diagnoses were ICD-10 I45.1 ‘Other and unspecified right bundle-branch block’ (25.1%) and ICD-10 J90 ‘Pleural effusion, not elsewhere classified’ (11.1%). The analysis also showed the following rates for specific complications of interest ICD-10 I33.0 ‘Acute and subacute infective endocarditis’ (0.6%), ICD-10 I38 ‘Endocarditis, valve unspecified’ (0.6%), ICD-10 I44.7 ‘Left bundle-branch block, Unspecified’ (0.3%), OPCS-4 X33.7 ‘Autologous transfusion of red blood cells’ (0.6%).
Conclusions This study has shown how a simple application in the PICTURE platform can generate a range of insightful information about potential risks associated with a cardiac surgery. With appropriate outcomes, it would also be possible for PICTURE to provide clinicians with data regarding potential benefits to a patient.
Acknowledgements for funding or support This work is supported by the NIHR GOSH BRC. The views expressed are those of the author(s) and not necessarily those of the NHS, the NIHR or the Department of Health. This work is supported by the Great Ormond Street Hospital Children’s Charity.