Lead investigator: Kelly Emrick, DHSc, PhD, MBA, BSRT(ARRT)Rkellyemrick.comLibraryAbout
Healthcare workflow modeling and algorithm design

See the constraint before you fund the fix.

We model how work actually moves through a service line, then design the algorithms that keep it moving. The pathway shown here is a live model, and you can change it.

Doctorate-levelpanel of healthcare professionals
100+ yearsof combined healthcare leadership
On siteobservation, research, and analysis

Run the model yourself.

Set the arrival rate, then add or remove capacity at any step. The pathway map and the figures update together.

This model is interactive. With JavaScript turned on you can change arrivals and staffing. The figures below show the baseline.

Outpatient visit pathway
10 per hour
Service time variability
Total visit99 min
In care41 min
Waiting58 min
Throughput10/hr
Care 41%Waiting 59%

Registration is the constraint: 83% utilized, with an average wait of 25 min.

Baseline staffing. Add or remove capacity at any step to test a change.

01Constraint

Registration

5 min per patient
1 staff
Utilization83%
Average wait25 min
02Constraint

Prep and screening

12 min per patient
3 staff
Utilization67%
Average wait5 min
03Constraint

Exam

20 min per patient
4 rooms
Utilization83%
Average wait20 min
04Constraint

Release

4 min per patient
1 staff
Utilization67%
Average wait8 min

Illustrative model with assumed service times. It is not calibrated to any site. An engagement model is fit to your own volumes, times, and schedule.

Three kinds of work, one accountable team.

01

Workflow modeling

A quantitative model of how patients, orders, and staff move through a service line, built so leaders can test a change before paying for it.

02

Algorithm design

Decision rules for scheduling, staffing, routing, and prioritization, each documented with its inputs, logic, validation status, and limits.

03

On-site analysis

Direct observation in the department by experienced healthcare leaders, so the model reflects the work as performed, not the work as documented.

Algorithms

Every algorithm ships with its specification.

An executive should be able to see what a model takes in, what it does, how far it has been tested, and where it stops being reliable. The card shown here documents the pathway model at the top of this page.

Outpatient pathway queueing modelSpecification card
InputsArrival rate per hour, staff or rooms at each step, average service time at each step, and service time variability.
LogicEach step is treated as a multi-server queue. Waiting time uses the Erlang C formula with the Allen-Cunneen correction for service time variability.
ValidationIllustrative only. Service times are assumed, and the model has not been calibrated to any site’s data.
LimitsAssumes steady random arrivals and independent steps. It does not represent appointment blocks, rework, shared staff, or downtime.
Method

From the department floor to a model you can govern.

Observe the work on site

Panel members walk the pathway with the people who run it, timing each step and recording the handoffs, interruptions, and workarounds that never reach a report.

Extract and reconcile the data

Timestamps, volumes, and staffing records are pulled from your systems and reconciled against what was observed, so disagreements between the data and the floor are resolved before modeling begins.

Build the model

The pathway is represented as a quantitative model sized to the decision at hand: a staffing change, a schedule redesign, a capital request, or a new routing rule.

Validate against observed performance

Before any scenario is run, the model must reproduce the waits and throughput your department actually recorded. Where it cannot, the gap is reported rather than hidden.

Hand off with governance

Your team receives the model, its specification card, and a named owner, review schedule, and retirement criteria, so the algorithm stays accountable after the engagement ends.

The panel

Doctorate-level healthcare professionals who have run the departments they model.

100+

years of combined healthcare leadership experience across the panel.

Lead investigator

Kelly Emrick, DHSc, PhD, MBA, BSRT(ARRT)R

Healthcare executive, author, researcher, and radiology operations leader. Practice in imaging service line transformation, AI governance, and ancillary services optimization.

Professional profile
Scope a project

Tell us where the work is getting stuck.

Service line
What leadership is seeing
Your starting pointImaging: long waits

A model you can run now: MRI Operations Workflow Analyzer

Tell us the service line and what leadership is seeing, and the panel will reply with a first read.

Start a scoping conversation

Decision support for operations. These models inform staffing, scheduling, and capital decisions. They do not direct clinical care.

Working labs. Every model on this site may be revised or corrected over time. Validate outputs independently before relying on them.

StreamLine Advisors

Site directory

Every resource, organized by practice.

Homekellyemrick.com