A Low-Risk Score, Forty-Five Minutes From the Nearest Emergency Department
Her MASCC score is the highest the index can give. The disagreement isn't about the score — it's about whether the index's own access assumptions hold for a patient who lives alone down a gravel road with unreliable cell service.
Joan T., 54, moved back to her family's cattle ranch two years ago after her husband's death, forty-five minutes down a gravel road from the nearest town and, more to the point tonight, the nearest emergency department. She lives alone there now, checks on the herd every morning before her coffee, and has cell service reliable enough for texting but not much else once she's out past the second pasture. Her son offered more than once to move back and help out, and she's turned him down every time — the ranch, she says, is the one part of her life that still feels entirely hers, cancer notwithstanding. She was diagnosed with diffuse large B-cell lymphoma four months ago and is on day 8 of cycle 2 of R-CHOP when a temperature she checked twice to be sure read 38.6°C.
She called the on-call oncology line rather than waiting it out, drove herself in, and reports only mild fatigue and some low-grade muscle aches — nothing that reads as a patient who feels as sick as her counts suggest she might be. Her ANC is 310, her exam and chest X-ray show no focal source, blood cultures are pending, and her vitals are unremarkable: BP 118/74, HR 92, no hypotension. Run through the actual index — mild symptoms, no hypotension, no COPD, a hematologic malignancy with no prior fungal infection, no dehydration, outpatient at the onset of her fever, and age under sixty — her MASCC score works out to 26 of a possible 26, the maximum the instrument can award and five points clear of the 21-point threshold Klastersky et al.'s 2000 validation study used to identify patients low-risk enough for oral, outpatient management, and the population Freifeld et al.'s 1999 trial later confirmed do just as well on oral therapy as admitted patients do on IV. What that score doesn't ask, and can't, is how far she'd have to travel if her temperature climbed again at two in the morning with no signal to call anyone. Her cycle 1 course was uneventful, no fever, counts recovering on schedule, which is part of why tonight feels to her like an overreaction rather than the correctly cautious response her actual neutrophil count calls for regardless of how she feels.
In the emergency department, deciding where she spends tonight
Send her home on oral therapy. A MASCC score of 26 is the highest the index goes — five points clear of the low-risk threshold, with nothing left on the instrument to deduct, and Freifeld et al., 1999, is the trial that actually tested this: oral ciprofloxacin plus amoxicillin-clavulanate performed as well as IV therapy in exactly this risk stratum. She's mildly symptomatic, hemodynamically normal, and has no focal source. Admitting a low-risk patient because of her address, not her score, isn't what the evidence supports.
The score isn't in dispute. What I don't think the score accounts for is that both trials assumed a patient who could get back to care quickly if she started to decompensate. Neutropenic sepsis doesn't always announce itself with a slow ramp — it can move from feeling mildly fatigued to hypotensive within a few hours. She's forty-five minutes down a gravel road, alone, with spotty cell service. That's not a comorbidity the index measures, but it's a real constraint on the "appropriate follow-up" every version of this guideline quietly assumes.
I'm not saying the trial data is wrong. I'm saying it was never tested against a patient who might not be able to get back here at all if she needed to.
You're both arguing about her risk level, and I don't think that's actually where you disagree — neither of you thinks she's high-risk tonight. The real question is how much observation answers the access concern without importing the cost of a multi-day admission her score doesn't support. Give her one dose of empiric IV cefepime and keep her twenty-four hours to confirm she's trending down, the way a genuinely low-risk course should.
If she's afebrile and stable in the morning, transition her to the oral regimen with a same-day return plan confirmed — a ride arranged, a working phone number, explicit instructions to come back at the first sign of fever recurrence rather than wait it out on the ranch.
Agreed: admit overnight for one dose of empiric IV cefepime and twenty-four hours of observation; if she remains afebrile and clinically stable, discharge on oral ciprofloxacin plus amoxicillin-clavulanate with a confirmed ride, a working phone number on file, and explicit same-day return instructions.
Not agreed: the infectious disease physician still felt even the overnight admission was probably more caution than a maximum 26-point score warranted, and was concerned about quietly establishing a pattern of treating every rural patient as higher-risk regardless of the number in front of the team. The hospitalist wasn't comfortable discharging same-day regardless of the score, given how little the index actually says about her ability to get back if things changed. Neither position was overruled; the observation-then-discharge plan proceeded as the working compromise, with her risk stratification for the next cycle's fever, if one comes, left as an open question for that visit.