Inpatient Hospital Prices Drive Spending Variation for Episodes of Care for Privately Insured Patients

By Chapin White, James D. Reschovsky, Amelia M. Bond
National Institute for Health Care Reform, February 2014

When including all care related to a hospitalization — for example, a knee or hip replacement — the price of the initial inpatient stay explains almost all of the wide variation from hospital to hospital in spending on so-called episodes of care, according to a study by researchers at the former Center for Studying Health System Change (HSC) based on 2011 claims data for 590,000 active and retired nonelderly autoworkers and dependents. For example, average spending for uncomplicated inpatient knee and hip replacements ranged across 36 hospitals from less than $17,500 to $37,000 for an episode of care that included all services during the inpatient stay and all follow-up care within 30 days of discharge. The pattern of spending variation for knee and hip replacements held true for other conditions, with hospital inpatient price differences accounting for the vast majority of spending variation rather than differences in spending on physician and other non-hospital services during and after discharge or spending on readmissions. Moreover, hospitals’ case-mix-adjusted relative spending per episode for different service lines — for example, orthopedics and cardiology — tend to be highly correlated with each other. Understanding why spending for episodes of care varies so much among hospitals can help private purchasers accurately target ways to control spending. This study’s findings — inpatient prices drive the bulk of episode-spending variation and hospitals with high spending for one service line tend to have high spending for other service lines — indicate that private purchasers can focus on hospitals’ overall inpatient price levels rather than pursue bundled payments for episodes of care or service-line-specific purchasing strategies.

To Bundle or Not to Bundle…

In Medicare, there is a compelling case for bundled payments — wide variations in post-acute care use are the main factor behind differences between high- and low-spending geographic regions and between high- and low-spending hospitals. Moreover, Medicare patients often have multiple chronic conditions that are complex to manage. But the results of this analysis show that the case for bundled hospital payments for the privately insured is much weaker — post-acute care and other ancillary services account for a relatively small share of overall spending on hospitalization episodes, and they account for almost none of the variation in episode spending from one hospital to another.

Implications for Private Purchasers

It remains to be seen whether, going forward, the tools available to private purchasers — tiered benefits, reference pricing, and so on — can counteract hospitals’ significant market power. Other more dramatic interventions, such as state-based hospital rate setting, or offering a “public option” that uses administered pricing through the state health insurance exchanges are options, albeit unlikely in most states to gain traction.

Rather than relying on proven methods of controlling health care costs through government administered pricing, the Affordable Care Act (ACA) relies on private sector integration of the health care delivery system through entities such as accountable care organizations. One of the mechanisms promoted by ACA is bundled payments – paying a single pre-set amount for all services associated with a single hospital episode of care such as a joint replacement. Will bundled payments adequately control our health care spending?

This study shows that, for private patients, the upper end of the wide variation in spending between hospitals is driven by high prices which permeate all of the hospitals’ service lines. For those services amenable to bundling, such as joint replacement, most of the services provided are related to the specific episode and thus do not vary much between hospitals. Post-acute care and other ancillary services which might otherwise be pared back with bundling are such a small part of the overall services that bundling cannot save money by reducing the volume of services connected to the episode.

Since it is the high prices of the standard services that are a problem, attacking only those services that are bundled will not save much since high prices throughout the rest of the hospital are left undisturbed or even increased some to offset any price concessions for the bundled services.

This article mentions that there is a case for bundled payments under Medicare since these more complex patients have wide variations in acute and post-acute care, providing more flexibility in controlling the volume of services. Prices are already controlled through the Medicare prospective payment system, which, to a limited degree, represents a form of bundling (think DRGs).

Other nations, even if using private insurers, utilize rigid government price setting through one form or another. The authors of this report suggest that state-based hospital rate setting or government administered pricing would not likely gain traction. Yet “bundling” all hospital spending through global hospital budgets (like fire or police department budgets), as part of a single payer system, is precisely what we need. Let’s give it traction.