Between 2013 and 2018, Medicare's total sepsis discharges grew from 556,562 to 806,561. Over that same period, the average payment per sepsis discharge fell relative to all other DRGs, converging almost entirely with the hospital-wide average. More volume, less premium per case. That combination is unusual for a genuine disease outbreak and worth examining closely.

The Definition Change Arrived, and the Billing Followed

In 2015 and 2016, the medical community adopted Sepsis-3, a revised clinical definition that broadened the criteria for diagnosing sepsis. The billing data tracks almost perfectly with that timeline.

DRG 871, the workhorse sepsis code covering cases without mechanical ventilation but with major comorbidities, grew from 403,847 discharges in 2013 to 607,584 in 2017, an increase of more than 200,000 cases in four years. Total Medicare payments for that single DRG reached $7,579,521,838 in 2018. For context, that's one billing code, one payer, one year.

The volume trajectory is steep and consistent. But the payment-per-case story runs in the opposite direction, which is where the analysis gets complicated.

A Premium That Disappeared as Volume Grew

In 2013, the average sepsis discharge paid $12,472, compared to $10,245 across all DRGs. That $2,227 premium, 21.7% above the hospital-wide average, reflected the genuine clinical intensity of sepsis cases: longer stays, more resources, higher acuity.

By 2017, that premium had collapsed to $346 per discharge, or 3.1%. By 2018, it was 2.8%. The all-DRG average payment rose from $10,245 to $11,697 over that period, while the average sepsis payment barely moved, going from $12,472 in 2013 to $12,029 in 2018.

When a diagnosis category's payment premium shrinks as its volume grows, it typically means the new cases being added are lower-acuity than the original population. The cases that drove the 2013 sepsis premium, genuinely resource-intensive infections, were still there. What changed is what got added around them.

Some States Moved Faster Than Others

The geographic variation reinforces the billing-behavior interpretation. DC recorded 3,323 sepsis discharges in the 2013-2015 period and 5,787 in 2016-2018, a 74.1% increase. Nevada went from 13,538 to 22,202 discharges, a 64% increase, ranking third among states for the fastest adoption of expanded coding. Nebraska rose 56.6%.

These aren't states with notably different infection rates or population health profiles. They're states where hospital coding practices shifted quickly after the definition change. The variation across states suggests the increase reflects administrative and clinical documentation decisions, not a regional disease pattern.

For Medicare, the financial exposure is direct. Even with the per-case premium shrinking, total payments for DRG 871 alone grew by more than $2.6 billion between 2013 and 2018. Auditors and recovery audit contractors have flagged sepsis as a high-risk billing category for years. The data here shows why: the volume increase is large enough that even a modest share of cases that don't meet clinical criteria represents hundreds of millions of dollars in payments.

Does the volume increase reflect genuine clinical capture of previously undercoded patients, or primarily upcoding of lower-acuity cases into sepsis DRGs? The shrinking payment premium points toward the latter, but the two explanations aren't mutually exclusive, and the data alone can't assign proportions. What it can say is that a definition change and a billing surge arrived together, and the cases added after 2015 look systematically less resource-intensive than the ones that came before.