Upcoding and Optimal Auditing in Health Care (or The economics of DRG creep)

We present a model of optimal contracting between a purchaser and a provider of health services. We assume that providers can increase demand by increasing quality but can also inflate activity through a manipulative effort (upcoding or DRG creep). We derive and compare the optimal price and audit policy for the purchaser under two scenarios: a) the purchaser can commit to a given audit policy (commitment); b) the purchaser cannot commit to a given audit policy (no commitment). If reported output is additive in quality and manipulation, we find that, if price is exogenously determined, the frequency of an audit is higher under the 'commitment' scenario than under the 'no commitment' one; also, under 'commitment', the degree of manipulation, quality and reported output are higher. If price is endogenous (i.e. it can be optimally chosen by the purchaser), then price is higher under 'no commitment' while the optimal audit policy, the equilibrium quality, manipulation and purchaser's net benefit are identical. If reported output is multiplicative in manipulation and quality, the purchaser sets a higher price under 'no commitment'. Nevertheless, quality and manipulation remain at lower levels, whereas auditing is more intensive than under commitment. The inability to commit now reduces the purchaser's net benefit.