Infusion Pump Fleet Replacement: Age Is Not a Risk Score
You are the HTM director for a 410-bed community hospital. The fleet includes 520 syringe infusion pumps used across med-surg, ICU, and oncology. The CMMS shows 318 pumps (61%) are past the 7-year design life, and the original manufacturer has announced that service parts and software support end in 14 months. Your biomed team has 3.5 FTE supporting 4,900 active devices, 1,140 pump PMs per year, and an average of 34 pump repair work orders per month. Last year 11 pumps were simultaneously out of service for more than 48 hours, and two units were borrowed from a sister hospital during a respiratory-season surge. The vendor quotes $6,900 per replacement pump plus $240,000 for a wireless drug-library platform. The CFO has approved $1.1 million, roughly 160 pumps, and has instructed you to "replace the oldest ones first." Nursing leadership wants the wireless platform purchased now.
1. The CFO's rule is to replace the oldest pumps first. Explain why age alone is an inadequate prioritization rule, and describe the data and scoring model you would present instead.
Read the worked answer
Age is a proxy, not a measure of risk. Two pumps installed the same month can have very different duty cycles, battery histories, and exposure to harsh cleaning agents, and a 6-year-old pump in a low-acuity unit may pose less risk than a 5-year-old pump in oncology running continuous infusions. Replacing by age also ignores the failure modes that actually harm patients: unexpected shutdown mid-infusion, battery failure during transport, inaccurate delivery, and alarm failure.
I would build a scoring model from data that already exists or can be collected in two weeks:
- Asset record: install date, software/firmware version, battery replacement date, whether the unit has dose-error-reduction software.
- Work order history (24-36 months): failures per 10,000 pump-hours, mean time to repair, parts cost, downtime hours, repeat repairs for the same fault.
- Support risk: OEM end-of-service date, parts availability, third-party repair options, recall history.
- Clinical criticality: ICU, oncology, and OR weighted above med-surg; utilization (pump-hours per day); backup ratio (pumps per bed).
- Nursing-reported problems from unit walk-rounds, which often surface issues the CMMS misses.
A defensible weighting might be clinical risk 30%, failure probability 25%, parts/support risk 20%, downtime and backup coverage 15%, cost 10%. Output three tiers: replace now, retain with defined spares and monitoring, and monitor. The model is reproducible, updates quarterly, and ties each dollar to patient risk rather than a calendar.
2. With $1.1 million and a hard 14-month parts deadline, lay out a defensible replacement and bridging strategy, including what happens to the pumps that are not replaced.
Read the worked answer
I would convert the end-of-service date into a project schedule, not a budget line.
- Secure a written bridge agreement now. Before spending on hardware, negotiate with the OEM (or qualify an ISO 13485 third-party repair vendor) for parts and repair coverage through the retirement of the last legacy pump. Without this, the 360 un-replaced pumps become unrepairable and the plan fails in month 15.
- Buy in two tranches against the scoring model. Tranche one replaces the highest-risk 160 units identified by the model, not the oldest 160. Place the order early enough to absorb lead time.
- Build a deliberate buffer. Hold roughly 4-6% of the fleet (20-30 pumps) as dedicated spares rather than stretching every pump into service. The 11-simultaneous-outage event is a coverage problem, not a repair problem.
- Control cannibalization. If dead units are stripped for parts, do it under documented configuration control in the biomed shop only, never returning a modified pump to patient use.
- Budget the consumables. Batteries and power cords drive a large share of failures; fund a battery replacement cycle inside the same request.
- Track monthly: PM completion, pumps out of service more than 24 hours, rental/borrow events, and repair turnaround. Report those to the safety and capital committees.
- Return with a year-two request for the remaining units, using year-one data to show the failure and downtime trend. A phased plan with evidence is more fundable than a single $3.5 million ask.
3. Nursing leadership wants to spend part of the $1.1 million on the wireless drug-library platform and only 60 new smart pumps. How do you respond?
Read the worked answer
I would not simply agree or refuse. The request is likely solving a real problem, so I would first ask what problem: programming errors, pump location, interoperability with the EHR, or reporting on library compliance. Some of those can be addressed on the current fleet at far lower cost.
If we proceed, the clinical and technical case is strong: one standardized platform, one formulary library, one training program, one PM procedure. Buying 60 pumps plus a platform creates two fleets, two libraries that must be maintained in parallel, two sets of biomed competencies, and a migration problem when the other 460 units are replaced. It also spends capital on pumps we plan to retire and delivers the safety benefit to only a fraction of patients.
The platform is also not a pump purchase. It requires validated wireless coverage in every care area, pharmacy ownership of the library build and change control, EHR/BCMA integration, a cybersecurity review (MDS2, supported operating system, patching process), and interface testing. The $240,000 quote is the entry point, not the total cost.
My recommendation: fold the platform into a full-fleet, multi-year replacement strategy, with the first tranche sized to a single platform. If nursing needs an interim pilot, confine it to one unit with pre-agreed success metrics and a documented plan to migrate the remaining fleet. Put the decision in front of the value analysis or product evaluation committee so it is a shared clinical and financial decision.