DRAFT ECMAD Weight Calculation Documentation

Maryland Health Services Cost Review Commission DRAFT ECMAD Weight Calculation Documentation
Market Share Subgroup Meeting – November 12, 2014 For questions or comment please contact: Nduka Udom [email protected] 1 Overview: The Maryland Health Services Cost Review Commission (HSCRC) uses equivalent case mix adjusted discharges (ECMADs) as a method for quantifying inpatient and outpatient hospital volume into a single measure. This document outlines the steps taken to calculate the weight of an outpatient visit within the context of a hospital’s ECMAD count. Background: ECMADs are utilized in several HSCRC methodologies including calculations related to the Reasonableness of Charge (ROC) methodology, Global Budget demographic adjustment, and market share adjustments. A hospital’s ECMAD count includes case mix adjusted inpatient discharges as well as equivalent outpatient case mix adjusted discharges which is a value imputed from a hospital’s outpatient total charges and unit charge. While inpatient cases can be attributed to a single DRG, which groups up to 30 diagnosis and procedures codes into a single category, and calculation of their weights reflecting relative complexity and resource demands have been used for many years, creating a similar measure in the outpatient setting is more complex. as a single outpatient visit can contain up to 45 Current Procedural Terminology (CPT) codes and therefore 45 Ambulatory Procedure Groups (APGs), which require a more complex . complex weighting methodology in the outpatient setting. Overview of the Calculation: Outpatient weights for ECMADs are calculated by first separating outpatient visit records by principal APG Type into groups and then creating weights within those groups based on the normalized distribution of expected charges for each outpatient visit record about the group average. Within certain groups, specific APGs are excluded and trim points are applied to the resulting distributions before creating weights. These details are outlined in greater detail in the section below. A. Group and Assign Outpatient Records a Principal APG Type Step 1: Group Data The outpatient data is grouped using APG grouper version 3.8. An APG is identified for every CPT that is coded in the record. Each record can contain up to 45 APGs. Step 2: Exclude Observation Cases If any outpatient visit record has Observation Rate Center number of units greater than 23, the entire record is excluded from the outpatient weight assignment calculation. Step 3: Assign Principle Record Type For the records that have more than one APG, a principal APG Type is assigned. In order to do this, a hierarchy based on APG Type is applied. This is appropriate as each CPT (up to 45 per record) is linked to an APG and each APG is linked to an APG Type. The hierarchy of the assigned APG Type is as follows: Type 1: Mental Health/Substance Abuse 2 Type 2: Significant Procedures Type 3: Medical Visit Type 4: Ancillary Type 5: Incidental Type 6: Drug Type 7: Durable Medical Equipment Type 8: Unassigned For example, if a record contained both an APG classified as APG Type 1 and an APG classified as APG Type 4, the entire record would be assigned the principle APG Type 1, as that APG Type is superior in the hierarchy. Additional examples of how the APG Type is assigned are in Table 1 below. Table 1: Assignment of Overall APG Type APG 1 APG 2 APG 3 APG 4 APG APG APG APG Principle TYPE 1 TYPE 2 TYPE 3 TYPE 4 APG Type Record A 6 6 9 390 2 2 2 4 2 Record B 295 408 9 4 4 Record C 491 401 408 628 5 4 4 3 3 Record D 403 403 410 400 4 4 4 4 4 *Note: APG Type of ‘0’ is assigned to any record that contains only zeros or blanks instead of APGs classified as APG Types 1‐8. B. Assign Weights to Records with Principal APG Type 2 (Significant Procedure) *Steps 4‐7 only apply to outpatient records with principal APG Type 2. Step 4: Exclude APGs with Consolidation, Packaging, and Termination Flags The 3M EAPG Grouper is applied to the first 45 APG fields of each outpatient record with principal APG Type 2 to identify APGs with Consolidation, Packaging and Termination flags. If any one of the following four fields is flagged (= “1”), the corresponding APG (not the whole record) is excluded: 



Patient Terminated Procedure Discount Flag Packaging Flag Same Significant Procedure Consolidation Flag Clinical Significant Procedure Consolidation Flag 3 Step 5: Exclude Some Non‐Significant Procedure APGs Non‐significant procedure APGs (not whole records) (APG= 0, 999) are excluded. Step 6: Create and Apply ‘Singleton’ Weights to Remaining Records A table of “singleton expected charges” is created from the average charge per APG of records containing only one type 2 APG. This table is used to determine the statewide expected charge for each APG. These expected charges per APG are used to determine an expected charge per outpatient record. The distribution of expected charge per outpatient record is then pseudo‐normalized (three iterations only) about the actual mean statewide charge per outpatient record to yield an adjusted charge per outpatient record. A charge adjustment factor is then calculated for each hospital. The charge adjustment factor is calculated from each hospital’s ratio of overall mean adjusted outpatient charge/overall mean expected outpatient charge. The adjusted charge for each record of a given hospital is divided by this factor and the distribution of these values are then pseudo‐normalized (three iterations only) about the actual mean hospital charge per outpatient record to yield the weight for the outpatient record. Step 7: Assign Weights to the Infusion Records A specific weighting methodology is applied to records with a high APG of infusion (APG=110, 111). These records are segregated from the rest of the records with principal APG Type 2, and only a specific subset of APGs present on the record are included in the calculation of expected charge: infusion (APG=110, 111) and drugs (APG= APGs 430‐444, 460‐465). The expected charge for the outpatient record is then calculated by summing the expected charges for all infusion and drug APGs on the record. Then the sum of the expected charges for each record is multiplied by the ratio of the infusion average charges/ expected average charges to arrive at the “adjusted expected charges” to yield the weight. Finally, a trim point is applied to limit the impact of outlier records. If the total charges for a record exceeds the greater of: (weight*2) or (weight+$10,000), the total charges for that record will be set to the trim point. D. Assign Weights to Records with Principal APG Type 3 (Medical Visits) *Step 8 is only applied to outpatient records with principal APG Type 3. Step 8: Find Medical Visit APG and Calculate Weights Records with principal APG Type 3 only have one medical visit APG. Nonmedical APGs (APG = 999 or < 500) are excluded from the record. The remaining APG is a medical visit APG. Only medical visit APGs are utilized to calculate the expected charge of the outpatient record. The distribution of expected charge per outpatient record is then pseudo‐normalized about the actual mean statewide per outpatient record in order to arrive at the adjusted charge per outpatient record. A charge adjustment factor is then calculated for each hospital. The charge adjustment factor is calculated from each hospital’s ratio of overall mean adjusted outpatient charge/overall mean expected outpatient charge. The adjusted charges are divided by this factor and the distribution of these values 4 are then pseudo‐normalized (three iterations only) about the actual mean hospital charge per outpatient record to yield the weight for the outpatient record. If the total charge for case is greater than the trim point for the medical APG, the included charge will be based on the trim point. (Trim point = the greater of: (weight *2) or (weight + $10,000)). E. Assigning Weights to Records with Other Principal APG Type Step 9: Find Other Visit APG (Type 0, 1, 4‐8) and Calculate Their Relative Weights The Type 0, 1, 4‐8 records are treated as if they have only one APG in the record regardless of the number of APGs found in the record. The first recorded APG is treated as the only APG in the record. For visits assigned an APG Type of 0, the APG for the record is set at 0. An expected is calculated for the identified APG in the record. The distribution of expected charge per outpatient record is then pseudo‐normalized about the actual mean statewide per outpatient record in order to arrive at the adjusted charge per outpatient record. A charge adjustment factor is then calculated for each hospital. The charge adjustment factor is calculated from each hospital’s ratio of overall mean adjusted outpatient charge/overall mean expected outpatient charge. The adjusted charges are divided by this factor and the distribution of these values are then pseudo‐normalized (three iterations only) about the actual mean hospital charge per outpatient record to yield the weight for the outpatient record. If the total charge for case is greater than the trim point for the APG, the included charge will be based on the trim point. (Trim point = the greater of: (weight *2) or (weight + $10,000)). F. Merge the Datasets with Weights and Normalize the Weights to an Overall Statewide Average Weight of 1.00 Step 10, Calculated Statewide Normalized Weights for each High Weight APG: The overall average statewide expected charge for all records is calculated using the entire outpatient record dataset. Each high APG weight is calculated as the expected charge divided by the average Statewide expected charge. Next, calculate the Statewide Case Mix Index (CMI) for all records. Calculate the ratio of 1 to the statewide CMI as an "Adjustment Factor." Multiply each record's High APG weight by the "Adjustment Factor" to normalize the statewide CMI to 1.00. The weight is then rounded to the nearest 0.000001. G. Calculate ECMAD as the High APG weight multiplied by a factor (calculated as the average outpatient charge divided by the average inpatient charge). Step 11, Calculated Statewide Average Charge per Case from the Included Cases Used to Determine the Inpatient Weights from the Inpatient Dataset The Statewide Average Charge per outpatient record is determined from the outpatient dataset in F above. The ECMAD Adjustment Factor is then calculated as Statewide Average Charge per outpatient record divided by Statewide Average Charge per inpatient case. ECMAD is defined as the normalized weight from Step 10 multiplied by the ECMAD Adjustment Factor. 5

Download Report