P124 Disease modifying anti-rheumatic drug (DMARD) therapy monitoring and evaluation of patient progress as part of standard clinical care in rheumatology patients. (23rd April 2022)

• Record Type: Journal Article
• Authors: Luqmani, Raashid A; Escudero, Alexa; Driver, Krystie; Mirza, Maria; Jackman, John; Manderson, Sarah; Soni, Anushka
• Is Part Of: Rheumatology. Volume 61(2022)Supplement 1
• Publication Date: 2022-04-23
• Subjects: Rheumatism -- Periodicals; Rheumatology -- Periodicals; 616.723005
• DOI: 10.1093/rheumatology/keac133.123
• Languages: English
• ISSNs: 1462-0324

Abstract

Abstract: Background/Aims: Review of blood results for monitoring DMARDs is performed manually in our department and subject to human error with up to 200 starters per month. In the last 6 months at least three patients have had abnormalities which were missed, highlighting a significant clinical risk. The aim of the project was to develop, test and implement an automated algorithm to detect and act on significant blood abnormalities in patients starting DMARD therapy. Methods: We created a system to automatically analyse laboratory results supplied by our laboratory. We imported data into a specifically designed database to provide analysis based on BSR guidelines. The data is immediately ready for review by a clinician, grouped as: normal, mildly abnormal (exceeding the upper or lower limit of normal), missing (individual values not recorded), or abnormal (outside the threshold values); records with consecutive results showing a trend towards abnormality are also recorded. The clinician has the responsibility to review all records and endorse them after taking appropriate action where required. Results: We tested the system on two cohorts of 100 and 227 sets of blood tests respectively and confirm that it is more rapid and robust than a manual process. We compared each of record with a manual evaluation of the data on a spreadsheet. We identified more abnormalities using the new system than were found on manual inspection (29% vs 10%, Chi square P < 0.001). The manual system took 3 hours to prepare the data for evaluation; the new system performed the same task in under 1 minute. Table 1 shows the summary data for 100 test records, assessed manually and then using the algorithm, followed by real data using 570 records from 160 patients initiating DMARD. Subsequent manual inspection of the test records defined as abnormal by the algorithm confirmed that the new system had correctly identified abnormalities in every case. Conclusion: We have developed a blood monitoring system for patients initiating DMARDs, capable of processing up to 10, 000 results per session. The system is practical, more efficient and more accurate than a manual process for DMARD monitoring. Disclosure: R.A. Luqmani: None. A. Escudero: None. K. Driver: None. M. Mirza: None. J. Jackman: None. S. Manderson: None. A. Soni: None.