In Q3 we cut infusion pump rentals 28% (about $190k annualized) after tagging 1,100 assets with passive RFID and enforcing a 24-hour return-to-clean standard; average search time fell from 18 minutes to under 5. What dashboard metrics or par-level formulas have moved the needle for you, especially on fleets over 3,000 devices?
But > under 5. What dashboard metrics or par-level formulas have moved the needle for Agree — on a 3,400-device fleet, a “return-to-clean compliance by unit” tile (redline 90%) plus an “idle >12h at location” count that pages transport cut rentals about 20%. For par, we use 95th percentile 7-day rolling demand per unit + 1 float/20 staffed beds, rebalanced weekly; anything >110% of par triggers auto-redistribution before renting. If your 24-hour standard is working, try slicing compliance by hour and unit to catch the night-shift lag that hides pumps.
Quick example: “par-level” — we switched to a rolling P95 of hourly in-use per unit (last 8 weeks) plus a 1-device transfer buffer, and paired it with a rental-deflection alert that flags any request when ready-to-serve already covers the next 24h; that trimmed pump rentals 18% on a 3.2k fleet… Caveat: it only works if return-to-clean gets a mandatory scan, not just passive reads-do you have that in place?
Big win for us: an “idle >12h” heatmap off RFID last-seen to surface floors where pumps stall after discharge, plus a live ready-to-use count against current census to head off rentals. Small caveat — passive tags sometimes miss elevator rides, so we added a quick decon barcode scan as source of truth; @c_jensen90 do you vary the threshold nights vs weekdays?
What moved the needle for us was tracking ‘request-to-bedside’ P90 and ‘dirty-to-available’ median by shift and weekday vs weekend, then flagging care areas that miss both so transport and sterile processing can swarm the handoff… Minor caveat: RFID last-seen can lag on low-traffic wings, so we added a 10 a.m./10 p.m; sweep to refresh tags — @lisa_m57, curious if your dashboard breaks out nights.
One thing that moved the needle for us was a ‘ready-to-use ratio’ by care area and shift — ready pumps divided by in-use beds — trended hourly with a redline (we use 0.10), and if it dips twice the system auto-opens a rebalance task and tracks time-to-restore; it’s a credit score for pumps. Caveat: RFID read gaps can lie, so we add a trust score that downranks counts older than 90 minutes or from readers with downtime over 5% to avoid overcorrecting. Curious if your ‘return-to-clean’ workflow also logs restore time, or would that be easy to tie into transport?
We put a “rental gate” tile on the dashboard showing Redeployable Now = clean room count + units with >60 min no‑movement, and dispatch can’t approve a rental unless that number is zero. That plus a simple “two calls before rent” check-in with the two highest-surplus units cut pump rentals about 20% in six weeks. Small caveat: we had to tune the no‑movement threshold by service line (ICU about 30, med‑surg about 90) or it got noisy — does your ADT feed let you segment like that, @OP?