AI systems in clinical practice

Decision support systems
LISA
Clinical information and decision support system for collaborative care in childhood acute lymphoblastic leukaemia

developed by clinical domains keywords
A collaboration between three departments at Cancer Research UK: the Children's Cancer Group at St Bartholomew's Hospital, the Advanced Computation Laboratory and Information Systems. Cancer, childhood acute lymphoblastic leukaemia, shared care, collaborative care Clinical information system, decision support system
location commissioned status
Royal London Hospital, Great Ormond Street Hospital (London), (planned) Addenbrookes Hospital (Cambridge), Royal Manchester Childrens' Hospital, Middlesex Hospital, Diana, Princess of Wales Children's Hospital (Birmingham). 2004 To enter clinical use 2004/ under clinical evaluation
description
LISA's main components are:
  • A centralised Oracle database holding all patient information (drug schedules, blood and toxicity results, doses prescribed etc.) and accessible by health professionals from different sectors in different locations.
  • A web-based decision support module, implemented using PROforma guideline development technology, designed to provide advice about dose adjustments in the treatment of acute childhood lymphoblastic leukaemia.

Background and summary description of decision support system

Acute Lymphoblastic Leukaemia (ALL) is the most common paediatric malignancy. Over 95% of the 320 children diagnosed with ALL each year in the UK are enrolled into Medical Research Council (MRC) trials, and their treatment is determined by the research protocol for the trial. Treatment of ALL splits into three general phases: induction of clinical remission, consolidation of remission, maintenance therapy which typically lasts 2-3 years.

LISA is primarily concerned with providing support in the maintenance period during which drug dose decisions have to be made weekly. The mainstay of treatment during this period is the regular administration of two oral chemotherapy agents: 6-mercaptopurine (MP) which is administered daily, and meth-otrexate (MTX) which is given weekly. Doses have to be continually monitored and adjusted as responses to these drugs vary significantly from patient to patient. Decision support is considered useful in this period as dose errors have occurred in practice.

The dose adjustment rules from the MRC trial protocol were implemented using the PROforma guideline modeling language and recommendations are provided in a clinical setting by the TALLIS PROforma-based enactment engine.

At any time during the maintenance period, a patient can be in one of eight "states", each of which represents a drug dose combination of MP and MTX. LISA uses PROforma decision making capabilities to control transitions between states. A 9th option allows a clinician to prescribe an alternative not defined in the protocol. PROforma bases its dose adjustment recommendations on five main data inputs defined in the protocol: current state, current platelet and absolute neutrophil count of the blood result on which the decision is being based, number of weeks that the patient has been at the current state and number of weeks that the patient has tolerated treatment.

 

[Click for full size image.]
LISA screen

In the above dose calculator screen from Lisa (legible on the full size image), blood results data have been entered for a patient. The LISA decision support module now calculates and displays the recommended next dose, along with the reasons for the recommendation. A doctor or specialist nurse can now either confirm the recommended dose for administration or (using the Choose dose drop-down menu) select an alternative drug/dose combination.

references
Bury J, Hurt C, Roy A et al. A quantitative and qualitative evaluation of LISA, a decision support system for chemotherapy dosing in childhood Acute Lymphoblastic Leukaemia. Medinfo. 2004;2004:197-201.

[PubMed]   [paper - CRUK]

" Objectives: To assess the acceptability to clinicians of a webbased decision support system designed to assist with dosage adjustments during maintenance therapy for childhood Acute Lymphoblastic Leukaemia (ALL), and to evaluate the potential impact of the system on decision-making and dosage calculations. Design: Balanced-block crossover experiment with simulated cases; questionnaire study and semi-structured interviews. Participants: 36 clinicians with differing experience in the management of ALL. Interventions: Subjects were asked to decide on appropriate levels of chemotherapy dosing for 8 simulated cases, 4 using the LISA decision support system, 4 using conventional paper-based records and guidance. Main outcome measures: Number of protocol-consistent dosage decisions made; time taken to manage each case; accuracy of dosage calculations; subjects’ opinions as to whether or not they would use the system in practice. Additional outcome measures: Functions subjects would like to see in an idealised system; subjects’ satisfaction with the implementation of the functions provided by LISA; qualitative data on issues subjects felt would impact upon the successful deployment of the system. "
Hurt C, Fox J, Bury J, Saha V. Computerised advice on drug dosage decisions in childhood leukaemia: a method and a safety strategy. To appear in the proceedings of the 9th Conference on Artificial Intelligence in Medicine in Europe (AIME-03), Cyprus, 18 – 22 October.

[Paper - CRUK/ACL]

" ... The trial protocol specifies that following initial treatment there is a 2-3 year maintenance period during which drug dosage decisions are made weekly according to a set of pre-defined rules. These rules are complex, and there is a significant frequency of error in clinical practice, which can lead to patient harm. We have built a web-based decision support system (called LISA) to address this problem. The dose alteration rules from the MRC protocol were for-malised in the PROforma guideline modeling language as a state transition problem, and dose adjustment recommendations are provided into the clinical setting by a PROforma enactment engine. The design and implementation of the decision support module, the safety issues raised and the strategy adopted for resolving them are discussed. System safety is very likely to become a major professional challenge for the medical AI community and it can be addressed, in this case, with relatively straightforward techniques. "
J. P. Bury C. Hurt, C. Bateman et al. LISA: A Clinical Information and Decision Support System for Collaborative Care in Childhood Acute Lymphoblastic Leukaemia. Proceedings of the annual AMIA Annual Symposium, 2002.

[Paper - CRUK/ACL]

" The treatment of a child with Acute Lymphoblastic Leukaemia (ALL) requires the collaboration of multiple providers across different organisations. We describe the implementation of a clinical information system for supporting collaborative care in the management of children with ALL. The system integrates the provision of patient data and clinical information with protocol-based, patient-specific decision support. The approach illustrated here should find applicability in the management of other diseases requiring collaborative care across institutions. "
Medical Research Council. UK Acute Lymphoblastic Leukaemia Trial ALL97 (version 1.1 revised November 1999) (2000).

[MRC ALL Guideline]

The dose adjustment rules modelled in PROforma and implemented in LISA are defined in Appendix B of this MRC ALL trial protocol.

contact links

John Fox (then Advanced Computation Lab. Cancer Research UK), Professor of Engineering Science, University of Oxford

 bullet  LISA demonstration [OC]  bullet  Cancer Research UK  bullet  Advanced Computation Laboratory
acknowledgements
Chris Hurt, Advanced Computation Lab. Cancer Research UK; Jonathan Bury, Academic Unit of Pathology, University of Sheffield Medical School; Vaskar Saha, Cancer Research UK Children's Cancer Group, St Bartholomew's Hospital, London.
Entry on archive: 10 June 2003
Last main update: 11 June 2003
Search this site
 

 

Privacy policy User agreement Copyright Feedback

Last modified:
© Copyright OpenClinical 2002-2011