Authors (including presenting author) :
Wong WWY(1), Lim F(2), Cheng A(2), Tong YHA(1), Cheng TH(1), Fung WK(1), Cheng CYM(1), Cheung NT(1)
Affiliation :
(1)Information Technology and Health Informatics Division, Head Office, (2)Department of Oncology, Princess Margaret Hospital
Introduction :
Development of Smart Hospital has been one of the initiatives stated in HKSAR Policy Address 2017. This is aimed at improving the efficiency for provision of care to patients and quality and risk management of care delivery, as well as enhancing clinical big data analytics capabilities for efficient planning on hospital facilities, patient care clinical decision support features and better preparation on urgent disease outbreak. In Hospital Authority (HA), there are numerous types of data, e.g. laboratory, medication, diagnosis, etc., stored in HA Clinical Management System (CMS). In this connection, the exploration of big data analytics application on clinical care in HA is obligated in order to improve the quality of care to patients, and also sculpting the development of Smart Hospital. Aroused by the case that widely reported by media, who was an immunocompromised hepatitis B patient had not been prescribed with antiviral cover, the first used case of big data analytics in clinical care in HA was identified and piloted in various hospitals. That is the generation of alert list of hepatitis B patient on chemotherapy potentially without full coverage of antiviral treatment which would require clinician’s attention.
Objectives :
1. To improve the quality of care for hepatitis B patient on chemotherapy 2. To explore and affirm the potential of big data analytics application in clinical care
Methodology :
With close collaboration with oncology team of Princess Margaret Hospital (PMH), a rule engine to determine potential cases was developed using the data from 1 January 2017. The case list definition was well defined with various types of data, e.g. laboratory, diagnosis & Corp Alert data for Hepatitis B status, medication & procedure data for inclusion criteria of chemotherapy, medical appointment data for grace period exclusion, etc. PMH oncology team provided their expertise on the checking rule refinement during the prototyping phase so that the accuracy of the case list could be enhanced. In addition, the representative(s) of oncology team in piloted hospitals would receive a mail summarizing the number of new case retrieved +/- case retrieved previously but not yet handled. Clinicians could click on the access link in the mail and then review the case details via Clinical Data Analysis Reporting System (CDARS) for further management. Clinicians could also indicate if the case could be “Case closed” because no follow up required, or “Remind me 1 week later” as follow up with patient might be needed. Besides, clinicians could help in tuning the rule engine by indicating if the case is “True positive” or “False positive” so that we could review the checking rules accordingly in later phase.
Result & Outcome :
With the support and endorsement by Coordinating Committee of Oncology, the pilot was first trial run in PMH since April 2018 and further rolled out to oncology department in different hospitals later per individual hospital’s request. After the pilot (data from 1 January 2017 to 31 December 2018), a total of 660 cases were identified. 26.7% of the cases were regarded as “True positive” while the reasons of “False positive” would include seroconversion of hepatitis B status, patient’s choice of follow up by private sector, clinical practice variation on drug prescription, etc. In addition, 15.2% of the cases were chosen to “Remind me 1 week later” by clinicians that meant follow up works was required for those cases. To conclude, this project could help in the provision of better clinical care to hepatitis B patient on chemotherapy to certain extend and affirm that the potential of data analytics application in clinical care should be further explored in different aspects.
Co-location with:
