District Hospital Partners, L.P. v. Sebelius, 973 F. Supp. 2d 1

Background

Plaintiffs (186 hospitals) challenge HHS Secretary Sebelius’s methodology for setting the ‘‘fixed loss threshold’’ component of the Medicare IPPS outlier payment threshold for FFYs 2004–2006, alleging it was arbitrary and capricious and resulted in underpayments.
The fixed loss threshold is set prospectively so projected aggregate outlier payments approximate 5–6% of DRG-related payments; Secretary models historical charge data (inflated forward) and multiplies by hospital-specific cost-to-charge ratios (CCRs) to project costs.
CMS issued the June 9, 2003 Outlier Correction Rule to curb “turbo‑charging” (abusive, rapid charge increases); that Rule changed how CCRs were calculated (allowing tentatively settled cost reports, ending defaulting to statewide averages, and creating limited reconciliation).
Plaintiffs contend CMS’s inflation factors and CCRs still relied on historical, turbo‑charged data (including data from 123 high‑charging hospitals) and failed to account adequately for the Outlier Correction Rule’s effects, thus overstating projected costs and raising the fixed loss thresholds.
The district court reviewed the administrative record under the APA’s arbitrary-and-capricious standard and granted summary judgment for the Secretary, finding her methodology reasonable and adequately explained for FFYs 2004–2006.

Issues

Issue	Plaintiff's Argument	Defendant's Argument	Held
Whether charge inflation factors were arbitrary because they included data from turbo‑charging hospitals and failed to reflect post‑rule declines	Inclusion of turbo‑charger data (FFY 2000–2002) overstated inflation; Secretary should have excluded 123 worst hospitals or used more recent partial‑year data	Secretary reasonably used available historical data, had no reliable newer universe for FFY 2004, later used more recent part‑year data for FFY 2005, and inclusion/exclusion of suspicious points is a rational choice	Court upheld CMS: inclusion of the data and chosen inflation methods were reasonable and sufficiently explained
Whether CMS erred by using approximated "latest tentatively settled" CCRs that still pre‑dated the Outlier Correction Rule	Approximated tentative CCRs remained outdated and failed to capture predictable CCR declines post‑rule	Using tentatively settled CCRs reduced lag and better reflected hospital‑specific data; approximations were adequately explained and preferable to speculative projections	Court found CMS’s use and approximation of tentatively settled CCRs reasonable under APA
Whether CMS should have adjusted for elimination of defaulting to statewide averages or modeled its behavioral effects	Failure to explicitly explain adjustment for ending statewide‑average default and to model its broader behavioral impact understated the effect on thresholds	CMS incorporated Outlier Correction Rule effects in simulations (and elimination simply restored use of actual CCRs); modeling complex behavioral knock‑on effects would be speculative and unnecessary	Court held CMS adequately accounted for the policy change and permissibly declined to model speculative behavior changes
Whether CMS should have accounted for potential reconciliation (post‑payment adjustments) when setting thresholds	Reconciliation would reduce CCRs for some hospitals; failing to adjust overstates CCRs and thresholds	Reconciliation effects were unpredictable ex ante (uncertain number of hospitals and resource limits); CMS attempted limited projections for FFY 2004 and reasonably omitted reconciliation for later years	Court upheld CMS’s predictive judgments and explained approach as reasonable and not arbitrary

Key Cases Cited

County of Los Angeles v. Shalala, 192 F.3d 1005 (D.C. Cir.) (statutory 5–6% outlier projection rule and standard for prospective threshold-setting)
Motor Vehicle Mfrs. Ass'n v. State Farm, 463 U.S. 29 (U.S.) (arbitrary-and-capricious standard for agency rulemaking)
Bell Atl. Tel. Cos. v. F.C.C., 79 F.3d 1195 (D.C. Cir.) (including suspicious but relevant data points may be rational)
Mt. Diablo Hosp. v. Shalala, 3 F.3d 1226 (9th Cir.) (agency choice between imperfect datasets is permissible)
National Wildlife Federation v. EPA, 286 F.3d 554 (D.C. Cir.) (issues not raised in administrative comments are waived)
Cablevision Sys. Corp. v. F.C.C., 649 F.3d 695 (D.C. Cir.) (agency must not rely on factors outside its statutory mandate and must consider important aspects of a problem)
West Virginia v. E.P.A., 362 F.3d 861 (D.C. Cir.) (deference to agency predictive judgments balancing model complexity and data limits)
Oceana, Inc. v. Gutierrez, 488 F.3d 1020 (D.C. Cir.) (agency predictions based on expertise and past performance entitled to deference)