AHRQ on Bundled Payments

The fee-for-service method of compensating health care providers has taken much of the blame for excessive utilization and spending.  It is not surprising therefore that several different payment methods have been and are being explored.  One of these is bundled payments and an Agency for Healthcare Research and Quality report reviews the research on this form of payment to see what effect it might have on spending, utilization and quality.  (AHRQ Report)  The Agency first defines bundled payment as a method in which payment to providers is based on the predetermined expected costs of a grouping of related health care services.  In that sense it is similar to an episode-based payment, but distinct from a capitation form of reimbursement.  A bundled payment may cover multiple provider types or only one type and may vary in the time period captured in the payment.  The theory behind bundled payments is that providers will be more careful with service use, since more services raise the cost of providing services and reduce profitability.  If multiple providers are covered, there should also be an incentive for better coordination.  Potential downsides include underuse of appropriate services, avoidance of risky patients, increasing the number of “bundles” and upcoding to higher-paid bundles.

The authors found 58 relevant studies, only one of which was any kind of randomized trial.  Most were observational.  Twenty different bundled payment schemes were studied, most in the US and most by public payers such as Medicare.  There is little evidence about the use of or effect of bundled payment in the private health insurance market.  The studies did not provide any meaningful evidence on what design or contextual features might be a factor in the effectiveness or lack thereof in a bundled payment mechanism.  In terms of effect on spending and utilization, most studies found a significant effect, with reductions of ten percent or less in spending and from 5% to 15% in utilization.  The quality effects were more uncertain and potentially troubling, with no consistent improvement and in some cases a  reduction in performance on some quality measures.  Some studies also reported that the use of bundled payment did appear to result in a shift in utilization to other care settings.  All-in-all, as with most current “innovations”, there is little real evidence to support a meaningful effect on outcomes.