Case Study 05: Healthcare telephony
Hospital Phone Program
Between 900 and 1,000 phones across three hospitals and five doctors' offices, cut over on time
- 01SurveyEvery site walked and documented: phones, ports, switches, power, the hours it could not be touched.
- 02InstallNew phones placed at their desks in the window each site agreed to, with the old system still answering.
- 03CutoverSite by site, in the window each site could afford, with the old system kept alive until the new one answered.
- 04VerifyEvery device installed and tested in place. A phone that was not verified was not counted as done.
8 sites cut over. 0 missed cutover dates.
01: ANALYSIS
Eight sites that could not stop answering the phone.
- The organizations
- Three hospitals and five doctors’ offices. One technology partner, one program.
- What existed
- Aging phone systems at every site, each with its own layout and its own hours it could not be touched.
- Who it affected
- Clinical staff and patients, for whom a dead line is a safety issue.
- The constraint
- No site could lose service during business hours. Patient data set the security requirements on every network segment the phones would touch.
The analysis, in one picture
Eight sites, one rule.
Eight sites: three hospitals and five doctors' offices on aging phone systems, and 900 to 1,000 phones, each test-called before it counted.
Clinical sites
- 3 hospitals
- 5 doctors' offices
Aging phone systems. A dead line is a safety issue.
The program
- 900 to 1,000 phones
Each one test-called at its desk before it counted.
02: THE PROBLEM
Someone has to own the order, not just the install.
- The gap
- Eight healthcare sites on aging phone systems, each with its own hours and its own layout. Phones placed out of order, or in the wrong window, would have meant dead lines in clinical hours and a rollback.
- Why a vendor install alone would not fix it
- A partner installs what the order says. Someone has to own the order: the survey, the sequence, the windows, and the definition of done at each site.
- What success had to look like
- Every site cut over on its agreed date, every device verified in place, and no clinic without a working line at any point.
Stated as a performance gap
Eight healthcare sites on aging phone systems. The target: 900 to 1,000 phones installed and verified, each cutover on the date agreed with the site, with no loss of service during business hours.
03: THE DESIGN
One site, one window, one cutover. Verify before moving on.
The approach
I led the program and did engineering work on it, with the technology partner doing the installation at scale. The plan ran in phases with milestones and risk checks, the same way the three-state election rollouts had: survey everything, cut over one site at a time, verify before moving on.
01
Survey before order
Each site documented before any equipment was ordered: handset counts, ports, switch capacity, power, and the hours it could not be interrupted.
02
The old system stays until the new one answers
Each site kept its old system live until the new one was answering, so there was never a moment with no system taking calls.
03
One site, one window, one cutover
Each site cut over in a window it agreed to, so no clinic went dark during its hours.
04
Verified means tested in place
Done was defined as installed and verified on the desk, not shipped or racked. The count went to the client as it happened.
What each stage had to do, stated so it could be checked
- Survey
- Every site walked and documented before anything is ordered: phones, ports, switches, power, and the hours it cannot be touched.
- Cutover
- A site moves to the new system on its agreed date, with a working line throughout business hours.
- Device
- A phone is installed at its location, registers, and completes a test call before it is counted.
04: OUTCOME
On time. Every device verified.
What was delivered
Between 900 and 1,000 phones installed and verified across three hospitals and five doctors’ offices, and cutover on time at every site.
Attribution, stated plainly
The program plan, the sequence, the site windows and the definition of done were mine, with engineering work alongside. The partner did the installation at scale. What the work caused was a cutover with nothing missed; the equipment itself was the partner’s and the vendor’s.
05: EVALUATION
Verified in place at cutover; measured once it has run.
| Level | Evidence | Status |
|---|---|---|
| Level 1: Reaction | Site contacts signed off each cutover. A short survey of site staff after the first quarter is the next measure. | Planned |
| Level 2: Learning | Not a training deliverable. Site staff were shown the new handsets at cutover, and a one-page job aid followed. It belongs to the client, so it is not shown here. | Planned |
| Level 3: Behavior | Every device installed and verified with a test call; every site cut over on its date, counted as it happened and reported to the client. That is acceptance, not use: whether site staff work the new features rather than routing around them is the behavior measure, taken with the first-quarter survey. | Validated |
| Level 4: Results | Support tickets about telephony and call quality, before and after, per site. Counted once the new system has run a full quarter. | Planned |
06: WHAT I WOULD DO DIFFERENTLY
Count the tickets before the first phone.
The baseline
Telephony tickets and dropped-call complaints were not counted per site before the program. They would have made the before-and-after a figure rather than a feeling.
The handover
A one-page job aid per site type, written during the survey rather than after cutover, would have cut the first week’s questions.
A phone that is not verified is not installed.