Skip to main content

Table 1 The methodology applied to address problems in care for chronic lymphocytic leukaemia, the intensive care and diabetes

From: How can we discover the most valuable types of big data and artificial intelligence-based solutions? A methodology for the efficient development of the underlying analytics that improve care

 

CLL Problem 1

CLL Problem 2

ICU Problem 1

ICU Problem 2

Diabetes

Clinically relevant problem

Variations in treatment response to 1st and 2nd line

Imperfect algorithms for identifying newly diagnosed, high-risk CLL patients

Identifying patients with ineffective efforts at risk of poor outcomes

Diagnosing catheter related bloodstream infections (CRBSI)

Unknown variation in response to treatment with SGLTs + GLPs

Assess data for development

– NGS data available

– Follow-up probably sufficient

– Large variation in treatments

– NGS data available

– Follow-up sufficient

– Monitoring & EHR data available

– Sufficient sample size, sufficient follow-up, limited missing data

– EHR & biosignal data available & continued prospectively

– Limited missing data anticipated

– EHR data available from secondary care

– Large amounts of missing follow-up data

Identify critical barriers for successful development and implementation

–

P: Newly diagnosed CLL patients without treatment indication

I: Analytics that identify high risk patients followed by treatment with ibrutinib

C: Stratification using clinical symptoms without receiving treatment

O: Costs, LYG, QALYs

Barriers:

– Site-specific validation required

– Reimbursement of novel treatment

P: Patients on assisted mechanical ventilation

I: Identify patients at risk of poor outcomes with analytics and intervene to avoid ineffective efforts

C: Care in which ineffective efforts are not identified

O: Mortality, LOS, costs, LYG, QALYs

Barriers:

– Availability of monitor that identifies ineffective efforts

– Site-specific validation

P: Patients with central venous catheter

I: Early identification of CRBSI, catheter removal & antibiotics

C: Late identification of CRBSI, catheter removal & antibiotics

O: Mortality, LOS, costs, LYG, QALYs

Barriers:

– Varying prevalence of CRBSI

– Integration of analytics in an EHR

– Site-specific validation

–

Economic Evaluation

–

Benefits: 0.13 QALYs, + €89,985

Benefits: − 3% mortality, 0.21 QALYs, − €264 [58]

Benefits: − 0.5% mortality, + 0.06 QALYs, − €886

–

Continue development

Not feasible. Sample size too small and large variations in prescribing practices

Not feasible. High costs of treatment offset benefits gained

Feasible. Invest in research into the effectiveness of intervention and the price of the analytics [58]

Feasible. If the target market extends beyond Greece the impact of the prevalence of CRBSI on benefits should be considered

Not feasible. Small sample size and large amount of missing follow-up data

  1. CLL chronic lymphocytic leukaemia, ICU intensive care unit, NGS next generation sequencing, SGLTs sodium glucose transporter-2 inhibitors, GLPs glucagon-like peptide-1 agonists, CRBSI catheter related bloodstream infection, EHR electronic health record, LOS length of stay, LYG life years gained, QALY quality-adjusted life years gained