Defining patient-centred real-life clinical pathways during chemotherapy


Year:

Session type:

Theme:

Elaine Dunwoodie1,Karl Baker2,Angelina Kurniati3,Owen Johnson3,Jane Wolstenholme4,Jose Leal4,Christopher Price5,Chris Twelves6,Geoff Hall6
1University of Leeds,2X-lab Ltd,3School of Computing, University of Leeds,4Health Economics Research Centre, University of Oxford,5Nuffield Department of Primary Care Health Sciences, University of Oxford,6Leeds Institute of Cancer & Pathology, University of Leeds

Abstract

Background

It is unclear how best to deliver the potential benefits of routinely collected electronic health records (EHRs) to inform enhanced service delivery and facilitate clinical research. Driven by the desire to establish the potential impacts of a patient self-test home blood count monitoring device, we aimed to (i) establish reproducible methods of data mining EHRs, and (ii) use the information so derived to define and quantify patient pathways during chemotherapy.

Method

Data were extracted from a single integrated EHR (PPM, Patient Pathway Manager) at a large UK Cancer Centre.  Data on chemotherapy administration were linked to records of patient contact with the hospital. Patient EHRs were included if they had (i) a diagnosis of breast cancer and received adjuvant epiribicin and cyclosphosphamide [EC90] chemotherapy or (ii) colo-rectal cancer and received palliative oxaliplatin and infusional 5-fluorouracil [OxMdG] chemotherapy, and (iii) the first diagnosis of cancer was between January 2004 and February 2013.

Results

Software and a Markov model were developed to identify a schematic of patient pathways through chemotherapy. 67,529 events were used to generate pathways for 535 patients with breast cancer, and 420 patients with colo-rectal cancer.  Only 27 (5%) of those with breast cancer and 26 (6%) with colo-rectal cancer completed 6 cycles of chemotherapy without unplanned hospital contact; there were 474 and 329 pathway variants respectively. Over 6 chemotherapy cycles, 169 (31.6%) patients with breast cancer and 190 (45.2%) patients with colo-rectal cancer were admitted to hospital. 

Conclusion

Pathways of patients on chemotherapy are more complex than assumed by expert opinion.  Modelling these pathways has multiple uses such as demonstrating a clear route to identifying unmet needs and potential improvements in care and outcomes, care and data quality assurance, facilitating quantification of innovation impact and communicating with stakeholders.