The Bureau of Labor Statistics (BLS) released the October 2026 Consumer Price Index on November 10. Hospital services rose 0.8% for the month, physicians’ services 0.4%, and medical care overall 0.3%. The more consequential detail for reserving is what is missing. BLS states it “did not collect survey data for October 2025 because of a lapse in appropriations,” so the 12-month column in its summary table is blank. Any trend assumption that cites “12-month hospital services CPI” has no current official number behind it.
What was published, and what was not
The September 2026 release is the last complete 12-month reading. It showed hospital services up 5.6% over 12 months (+0.7% for the month), medical care services up 2.5%, physicians’ services up 2.1%, and medical care overall up 1.5%, held down by a 3.3% decline in medical care commodities. October adds monthly moves only.
One month is a weak signal. A 0.8% hospital print follows 0.7% in September, so two consecutive months run well above the 5.6% annual pace, but monthly hospital services data is volatile and the prior 12-month figure is the only full-year evidence we have.
Three substitutes, three answers
Here is how the options spread on hospital services, using only BLS figures and simple arithmetic (our calculation):
- Carry forward September. Keep 5.6% and label it one month stale. Defensible, and the most conservative about what was actually measured.
- Annualize recent months. October alone compounds to about 10.0%; September and October together to about 9.4%. This reads the latest momentum but treats a volatile series as a run rate.
- Approximate the missing month. BLS publishes a geometric-mean method (multiply the index values before and after the gap, take the square root). BLS cautions that results are not official CPI figures, that all parties should agree to the approach, and that confidence is lower for volatile series.
The spread between carrying forward and annualizing is roughly four points on hospital services. For medical care services the gap is smaller: the September 12-month figure is 2.5%, while October’s 0.3% annualizes to about 3.7%.
Who it affects
Self-insured employers with workers’ compensation medical reserves, self-funded health plan sponsors, and captives whose severity trend selections cite a 12-month medical or hospital CPI line. Public entity pools and hospital captives that use hospital services CPI as a cost-of-care proxy are exposed too. Plans using non-CPI trend sources, such as carrier or consultant surveys, are not directly affected, though CPI is often cited alongside them as a cross-check.
Reserve mechanism
The lever is severity trend, applied in two places. First, it projects open-claim medical payments forward to settlement. Second, it sets the expected claim ratio for the most recent accident year in the Bornhuetter-Ferguson and expected-loss methods, where there is little paid data to override the assumption. A higher trend selection raises the immature-year ultimate directly. Older years are less sensitive because their development is anchored to actual emergence. Earlier pieces here show why the choice of index matters even when the data is complete: see hospital services at 6.4% in March, the June services/commodities split, and why CPI understates health plan trend.
Our read
Absent evidence of a break, hospital services inflation has run in the mid-single digits all year, and two strong monthly prints do not by themselves reset that. The defensible default is the September 12-month figure with an explicit sensitivity at the higher annualized rate, not a switch to the annualized number. If the November release in December shows hospital services again above 0.5% for the month, the case for a higher selection strengthens. Whether the 12-month series resumes with the November release is the other thing to confirm.
Documentation
This is a data-quality disclosure. Actuarial Standard of Practice (ASOP) No. 23 covers data quality and ASOP No. 43 covers unpaid claim estimates; both are available from the Actuarial Standards Board. Whether this specific gap triggers a disclosure under either standard is a judgment for your actuary, but a one-sentence statement of which substitute was used is cheap and avoids a later question from an auditor.
What this means for your next review
List every trend assumption in your last reserve report and note which cite a 12-month CPI line. For each, ask your actuary to name the substitute, show the answer under the other two methods, and state the effect on the most recent accident year’s expected claim ratio. See the guide to what is driving your IBNR higher and the workers’ compensation IBNR primer for where those assumptions sit.