If MAAC88 < 1.15*PC89 then MAAC89 = Max(MAAC88+0.5*(1.15*PC89-MAAC88), If HCPCS = OP code then MAAC89 = = Min(MAAC89, If MAAC89 < 1.15*PC90 then MAAC90 Max(1.15*PC90, 1.01*MAAC89); = Else MAAC90 = 1.01*MAAC89; If HCPCS = OP code then MAAC90 = Min(MAAC90, H. Allowed charges for 1990 for each record are the minimum of the updated prevailing charge, the customary charge (defined as the updated actual charge for 1989), or the actual charge for 1990. II. FIRST-ORDER (PREBEHAVIOR) CONVERSION FACTOR This section specifies how to calculate the first-order conversion factor described under section 4001 of the bill. The formulas here are an algebraic representation of the fee schedule specified in section 4001. The initial (prebehavior) budget neutral conversion factor (CF1) for the reference fee schedule can be obtained algebraically from the adjusted BMAD III data as follows: CF1 = sum of 1990 allowed amounts for all affected services under prior payment policies (that is, in the absence of implementation of section 4001, but including the MEI freeze specified in section 4002), divided by the sum over all affected services of (RV(i,k) * GPCI(j) * service frequency). RV(i,k) denotes the relative value of service i by specialty k, which is equal to the sum of the physician work component and the practice expense component (W(i) + E(i,k)), as specified in appendix A. GPCI(j) denotes the geographic practice cost index for locality j, which is equal to the weighted sum of cost indexes for physician work (WGPCI) and for practice expense (EGPCI), as specified in appendix C. That is, GPCI(j) may be written as: (WGPCI(j)*W(i)/RV(i,k)) + (EGPCI(j)*E(i,k)/RV(i,k)). Hence, the first-order reference fee schedule amount for each affected service i and locality j may be written as: This is equivalent to the following, which is a more convenient formulation for obtaining the conversion factor: RFS1(i,j,k) = CF1 * [W(i)*WPGCI(j) + E(i,k)*EGPCI(j)]. III. SECOND-ORDER (POSTBEHAVIOR) CONVERSION FACTOR This section explains how to make the second-order adjustment called for in section 4001(a)(5) to achieve budget neutrality for 1990. The final (postbehavior) budget neutral conversion factor (CF2) is equal to: CF2 = ADJ * CF1, so that the final reference fee schedule amounts will be: RFS2(i,j,k) = ADJ * CF1 * [W(i)*WPGCI(j) + In the expression for RFS2, ADJ is a volume adjustment factor obtained by iterative simulations from the BMAD III data, as described below. For 1990, the bill specifies that payment rates are to be set by the usual customary, prevailing, and reasonable criteria, but that prevailing rates for the affected services are to be adjusted as follows: In this formula, RFS2 is set at whatever level necessary (via the value set for ADJ) to achieve new Medicare payment amounts for all services (not only affected services) that equal, in the aggregate, what payments would have been had the prior prevailing charge (PC90) been used instead. Medicare payment rates under both prior prevailing charges and those set in this bill are the smallest of the customary, actual, and prevailing charges for each service. Because the bill continues current limits on actual charges-which are defined relative to prevailing charges-both actual and prevailing charges might change relative to prior law. Aggregate payments made by Medicare will change for two reasons-changes in payment rates for affected services, and changes in the volume of all services induced by the payment rate changes. The adjustment factor (ADJ) is set so that changes due to new payment rates are just offset by changes due to volume responses, in the aggregate. In the simulation model, the value of ADJ (initialized to a value of 1) is reduced if estimated new aggregate payments exceed prior law payments, and ADJ is increased if estimated new aggregate payments are below prior law payments. This is an iterative process repeated until equality is achieved. More precisely, the simulation model must specify that for each service i in locality j, physicians' allowed amounts (A) and Medicare receipts (R) under prior law are: = A(i,j) Min(customary, prevailing, actual) charges; and, if the claim is assigned: R(i,j) = A(i,j); where S(i,j) denotes the actual (submitted) charge. New prebehavior allowed amounts (new A(i,j)) and receipts (new R(i,j)) under the bill are calculated analogously, but new prevailing and new actual charges must be used for the calculation where appropriate (that is, for affected services). Postbehavior values for allowed amounts and receipts are obtained as follows. First, sum up prior law (A0, R0) and prebehavior new law (A1, R1) allowed amounts and receipts for all services, separately for each practice in the data set. Then calculate postbehavior new law allowed amounts and receipts (A2, R2) for each practice as follows: For 1991 and later years, X = 0.375 if (R1-R0) is greater than or equal to zero; X = 0.555 if (R1-R0) is less than zero. However, for 1990 these values for X are divided by 2, to account for the delayed implementation date of these payment changes and for the likelihood that behavioral responses to payment changes lag behind implementation of the changes. The effect of this formulation is partially to offset the first-order impact of the payment change, both for gaining and losing practices. The behavioral parameters used in this formulation were obtained by regression analysis of Medicare claims for physicians' services in Colorado, subsequent to an abrupt and substantial realignment of prevailing charges in the late 1970's ("Volume Responses to Exogenous Changes in Medicare's Payment Rates,' Technical Memorandum, August 1988, U.S. Congressional Budget Office.) Initially, ADJ is set equal to 1; the resulting values for A0 and for A2 are summed over all practices. If the sum of A2 is greater than the sum of A0, then the value of ADJ is reduced until the sums are equal for A2 and A0. Conversely, if the initial sum of A2 is less than the sum of A0, the value of ADJ is increased until the sums are equal. |