Fever on Postoperative Day Two: Treat or Watch
A fever on the second day after major surgery, in a patient who otherwise looks well, tests a piece of teaching — that early atelectasis explains postoperative fever — against a literature that has actually moved past it.
Harold B., a 66-year-old retired postal supervisor, underwent an open sigmoid colectomy two days ago for diverticular disease complicated by a walled-off abscess that was drained intraoperatively. His recovery so far has otherwise tracked as expected — ambulating with assistance yesterday evening, tolerating sips of clear liquids, pain reasonably controlled on his current regimen. Overnight his temperature rose to 38.6°C, and this morning's white count is 13,200/µL, up from 10,800 on postoperative day one. The surgical team's overnight note already raises the possibility of starting empiric antibiotics.
Almost everything that would make this fever worrying is a thing his exam this morning does not show. His abdomen is soft, with expected incisional tenderness but no rebound or guarding, no wound erythema or drainage, and his lungs are clear apart from mild bibasilar crackles that clear partially with coughing — unimpressive findings, not the picture of an evolving surgical site infection or pneumonia. That pattern matters against a specific piece of teaching worth naming directly: postoperative day one to two fever has long been attributed reflexively to atelectasis, but Engoren's own study, examining chest films and fever curves together in a large postoperative cohort, found no reliable association between the two — atelectasis on imaging did not predict fever, and fever did not predict atelectasis. His own baseline before this admission was genuinely healthy for his age — well-controlled hypertension on a single agent, no diabetes, no chronic pulmonary disease, and an active retirement that still has him gardening most mornings — which matters because it removes several of the more common reasons an early postoperative fever might have a real infectious explanation in an older or more comorbid patient. What the literature instead increasingly attributes to this exact window is a genuine, IL-6-mediated surgical stress response — tissue trauma itself, independent of any infection, driving a measurable fever and leukocytosis that characteristically peaks around this postoperative day and resolves on its own over the following forty-eight hours, a pattern his own vitals and exam so far look far more consistent with than with an evolving wound or pulmonary process.
Surgical ward round, postoperative day 2
I'd rather start empiric coverage now than wait — he had an abscess drained intraoperatively, this is exactly the population where an early surgical site infection or anastomotic problem is a real risk, and a new fever with a rising white count on day two is enough for me to want to be covered rather than watch it.
I understand the instinct, but his exam doesn't actually support that read — soft abdomen, clean wound, no localizing signs. I'd point out that the old teaching linking early fever to atelectasis has real evidence against it now; Engoren's study specifically found no reliable relationship between the two on chest imaging and fever curves together. What this looks like instead is the expected IL-6-driven surgical stress response, which peaks right around now and resolves on its own.
I take the abscess history seriously — but 'he had an infection source drained' isn't the same finding as 'his exam shows a new one starting,' and right now the exam doesn't show that.
I'd frame the actual decision rule around the exam and the trend, not the calendar day itself. A benign abdomen and stable-to-improving vitals are real reassurance regardless of whether this is postoperative day two or day five — the day-of-surgery framing is only ever a rough proxy for how likely a given fever is to be infectious, not a rule on its own.
So: hold antibiotics, but set an explicit trigger — any wound change, abdominal exam change, or hemodynamic drift over the next twenty-four hours moves us straight to workup and treatment, no further debate needed at that point. That gives the surgical team's real concern a concrete answer without treating on suspicion alone today.
Agreed: antibiotics held, with an explicit, written trigger for starting them — any new wound erythema/drainage, abdominal exam change, or hemodynamic drift over the next twenty-four hours — rather than an open-ended “watch and see” that would leave the surgical resident's real concern unaddressed.
His fever and white count were reassessed the following morning: temperature down to 37.4°C, WBC trending back toward 11,000/µL, exam unchanged — consistent with the surgical stress response resolving as expected, and no antibiotics were ultimately started.