Description of Organization
Good Health Cardiac Clinics is an Ontario outpatient cardiology organization that collaborates with cardiologists to build and operate specialized cardiac diagnostic clinics, currently in Guelph and Markham with expansion into Cambridge planned. Services include cardiology consultation, electrocardiography, ambulatory cardiac monitoring, echocardiography, exercise stress testing, stress echocardiography and ambulatory blood-pressure monitoring. Delivering these services requires close coordination between referrals, scheduling, administrative staff, cardiovascular technologists, sonographers, physicians, equipment, rooms and reporting workflows. As the network grows, Good Health is focused on preserving rapid patient access while building scalable, data-informed operating systems.
Problem area
A meaningful portion of the work required to deliver outpatient cardiac care occurs before the patient arrives: scheduling, confirmations, reminders, instructions, intake, medication and history collection, inbound questions, rescheduling, failed-contact follow-up, chart preparation and correction of incomplete information. As patient volume grows, the scalability challenge is preventing pre-visit administrative labour from growing at the same rate, while recognizing that aggressive automation can reduce information quality, exclude patients who need assistance, worsen patient experience, or simply shift work to check-in and clinical staff.
Good Health is already evaluating and implementing third-party patient-communication and intake automation, including AI-enabled tools, so this capstone should not recreate a specific vendor product. The central problem is to independently determine the true current cost-to-serve, identify where administrative effort is concentrated, quantify what automation actually removes, and design the optimal future-state operating model: 'touchless where appropriate, human by exception', in which routine patients become appointment-ready with minimal staff involvement and patients who need assistance are identified early. The project remains operational rather than clinical; automated tools must not diagnose, triage clinical risk or make treatment decisions.
Main objectives
- Define and measure the current pre-visit cost per appointment-ready patient using an activity-based costing approach, including staff minutes, contact attempts, rework and technology costs.
- Identify which patient journeys and appointment types generate disproportionate administrative burden.
- Determine which activities can be removed, simplified, automated or shifted to patient self-service without degrading readiness, accessibility or experience.
- Evaluate the actual operational impact of existing or pilot automation tools, including whether work is eliminated or displaced elsewhere in the patient journey.
- Develop escalation rules that define when automated outreach should stop and human intervention should begin, and prototype the optimized workflow and its cost/exception-management model.
- Model financial and staffing impact at current and future volumes, including patients supported per administrative FTE and avoided incremental administrative staffing.
Scope of work
Phase 1 - Current-state process mapping. Map the patient journey from booking to arrival for selected appointment types; document staff touchpoints, communication channels, handoffs, repeated work and common failure modes; identify where incomplete information creates rework at check-in or downstream.
Phase 2 - Activity-based costing and baseline metrics. Estimate staff time and loaded labour cost for major pre-visit activities using system data, contact records, sampling and structured time studies; add technology costs and the cost of failed contacts, rework and no-shows; establish baseline metrics including cost per appointment-ready patient, minutes and contacts per patient, intervention rate and patients per administrative FTE.
Phase 3 - Segmentation and burden analysis. Analyze burden by appointment type, clinic, channel and other non-discriminatory operational characteristics; identify low-, medium- and high-touch patterns without using the model to restrict access to care; assess accessibility and digital-literacy needs so automation does not remove necessary human support.
Phase 4 - Automation evaluation and future-state design. Evaluate current/pilot automation against baseline metrics to determine which tasks are genuinely removed, reduced or shifted; compare combinations of SMS/digital intake, automated voice, reminders, self-service and human follow-up without locking to one vendor; define explicit escalation rules.
Phase 5 - Prototype and economic model. Prototype a future-state workflow, cost calculator, exception queue and/or patient-readiness process on de-identified or simulated data; model current-volume and growth scenarios (capacity released, technology cost, FTE avoidance, sensitivity to assumptions); validate with staff and, where approved, patient-experience measures.
Deliverables
- Report
- Presentation
- Resources
- New protocols/procedures
- Activity-based cost model and baseline scorecard; end-to-end current-state pre-visit process map; patient-journey segmentation; a vendor-agnostic automation evaluation framework; and a functional or high-fidelity prototype of the supporting cost/exception-management tool, with a business case and sensitivity analysis.
Frequency
- Bi-weekly
Skills and training required
Industrial engineering, process improvement, service design or operations management; activity-based costing and business analysis; data analysis (Python, R, SQL, Excel, Power BI); human factors, UX/service design or behavioural research; the ability to evaluate automation and AI tools critically; process mapping, stakeholder interviewing and change-management thinking. Healthcare experience is helpful but not required; Good Health provides workflow and privacy orientation.
Resources required
De-identified or aggregate appointment and workflow data (booking and confirmation status, appointment type, cancellations and no-shows, communication attempts, pre-visit completion); workflow documentation, scripts and operating procedures; administrative roles and labour-cost assumptions; controlled outputs from current or pilot automation tools where sharing is permitted; access to leadership, administrative leads and frontline staff; time-study opportunities that avoid unnecessary exposure to personal health information; synthetic test cases. Standard University computing and analytical software (Python, R, SQL, Excel, Power BI) plus prototyping/workflow-design tools are sufficient; no production EMR or automation-system access is required.
NDA or a commercialization agreement for this project?
Yes