Her Glucose Only Spikes in the Afternoon, and That's the Whole Clue
Her fingersticks trace the exact shape of her prednisone dose working through the day. The real question isn't whether to treat the hyperglycemia — it's whether to match the insulin to that shape or default to a flat, all-day coverage that doesn't.
Yvonne T., a 58-year-old woman, teaches middle school choir and was admitted two days ago for a moderate COPD exacerbation, started on daily oral prednisone 40mg each morning alongside her inhaled bronchodilator therapy, with a planned taper over the next two weeks as her breathing improves. She has no personal or family history of diabetes, and her admission glucose was unremarkable. By hospital day two, her afternoon and evening fingersticks were consistently running 220–280 mg/dL, while her fasting morning glucose stayed closer to 130–150 — a pattern, not a random elevation, and one that tracks closely with when her single morning prednisone dose is doing most of its metabolic work.
Corticosteroid-induced hyperglycemia has a recognizable shape when the dosing is once-daily in the morning: prednisone's insulin-resistance effect peaks in the afternoon and evening hours after administration and largely clears by the next morning, which is exactly the pattern her fingersticks are showing. That timing detail matters for more than just explaining the numbers — it's a real argument for matching the insulin's own action profile to the pattern rather than defaulting to a flat, all-day basal coverage that doesn't track when the actual risk is concentrated.
Matching the insulin to the hour the steroid actually bites
I'd start our standard weight-based basal-bolus protocol. Sliding-scale-only regimens have repeatedly underperformed scheduled basal-bolus dosing in hospitalized patients since Umpierrez’s RABBIT-2 trial — they react to an already-high number rather than anticipating the pattern — and a protocolized basal-bolus approach is well studied and reliable across a broad range of steroid regimens, not just once-daily morning dosing like hers.
For her specific dosing pattern, I'd use NPH insulin rather than a standard long-acting basal analog. NPH's intermediate duration and defined peak line up with the afternoon-to-evening rise her fingersticks are already showing, far better than a flat 24-hour basal profile does — it's a genuine pharmacokinetic match, not just a generic insulin choice.
I'd also flag that a flat, high-dose basal analog sized to cover her afternoon peak risks real overnight and early-morning hypoglycemia, since her fasting numbers are already near-normal and the steroid effect has largely worn off by then — NPH's own waning profile overnight is actually the safer match here, not a compromise.
Whichever insulin you choose, I want the taper protocol written down today, not improvised in two weeks. The most common preventable harm I've seen in steroid-induced hyperglycemia isn't inadequate control now — it's a patient sent home on a fixed insulin dose that never gets reduced as the prednisone taper brings her own insulin resistance back down, and she has a real hypoglycemic event a week from now that nobody anticipated.
I'd propose a proportional reduction schedule tied explicitly to her prednisone taper days, built into her discharge instructions in writing, so this isn't left to her or a covering physician to figure out reactively once she's already home.
NPH insulin started with morning dosing, correctional rapid-acting insulin added for mealtime coverage, and a written proportional taper protocol built into her discharge plan tied explicitly to each prednisone dose reduction. Her afternoon and evening glucose values improved substantially by hospital day four.
Not agreed: whether she needs a formal outpatient endocrinology follow-up given no prior diabetes history, or whether her primary care physician can manage the taper independently using the written protocol. The endocrinologist recommended a single follow-up visit near the taper's end to confirm insulin was discontinued appropriately; the hospitalist felt the written protocol was sufficient on its own. Left to be decided at discharge planning.