Printing Tips

CMS established two programs to monitor the accuracy of payments made in the Medicare Fee-for-Service (FFS) program: The Comprehensive Error Rate Testing (CERT) program and the Hospital Payment Monitoring Program (HPMP). The national paid claims error rate is a combination of error rates calculated by the CERT program and HPMP; the CERT program represents approximately 60% of the payments upon which the error rate is calculated while the HPMP represents the remaining 40%. The CERT program calculates the error rates for Carriers, Durable Medical Equipment Regional Carriers (DMERCs), and Fiscal Intermediaries (FIs). HPMP calculates the error rate for the Quality Improvement Organizations (QIOs). More information on the differences between Carriers/DMERCs/FIs/QIOs may be found in later sections of this report.

Strong outcome-oriented performance measures are a good way to assess the degree to which a government program is accomplishing its mission and to identify improvement opportunities. This May 2006 Report describes the performance measurement process for Carriers/DMERCs/FIs/QIOs.

The Department of Health and Human Services (DHHS), Office of Inspector General (OIG) produced Medicare FFS error rates from 1996 to 2002. The OIG designed a sampling method that estimated only a national FFS paid claims error rate (the percentage of dollars that Carriers/DMERCs/FIs/QIOs erroneously allowed to be paid). To better measure the performance of the Carriers/DMERCs/FIs/QIOs and to gain insight about the causes of errors, CMS decided to calculate a number of additional rates. The additional rates include provider compliance error rates (which measure how well providers prepared claims for submission) and paid claims error rates (which measure how accurately Carriers/DMERCs/FIs made coverage, coding, and other claims payment decisions) for specific contractors, service types, and provider types. CMS began producing error rates and estimates of improper payments in November 2003.

CMS calculated the Medicare FFS error rate and improper payment estimate for Carriers/DMERCs/FIs/QIOs for this report using a methodology approved by the OIG. This methodology includes:

• CERT randomly selecting a sample of 116,417 claims submitted to Carriers/DMERCs/FIs during the reporting period.
• HPMP randomly selecting a sample of 39,821 acute care inpatient hospital discharges.
• Requesting medical records from the health care providers that submitted the claims in the sample.
• Where medical records were submitted by the provider, reviewing the claims in the sample and the associated medical records to see if the claims complied with Medicare coverage, coding, and billing rules, and, if not, assigning errors to the claims.
• Where medical records were not submitted by the provider, classifying the case as a no documentation claim and counting it as an error.
• Sending providers overpayment letters/notices or making adjustments for claims that were overpaid or underpaid.

CMS calculated error rates in this report by reviewing claims that providers submitted during specific reporting periods. CMS has accelerated the reporting periods of both the CERT program and the HPMP. Beginning with this report, the CERT reporting period has been accelerated by three months and the HPMP has accelerated its schedule by six. The CERT acceleration process has, as a side effect, forced the exclusion of three months worth of data that would normally have been included. More information is available in the section titled "The CERT Program". The following table outlines the reporting periods to date for improper payment reports as well as the changes planned for upcoming reports.

To promote consistency in improper payment reporting across federal agencies, the IPIA requires agencies to follow a number of methodological requirements when calculating error rates and improper payment estimates. One requirement is the use of gross figures when reporting improper payment amounts and rates. A gross improper payment amount is calculated by adding underpayments to overpayments. Unless labeled otherwise, figures in this report are gross figures; historical figures that were originally reported as net numbers have been converted for consistency.

The IPIA also requires the inclusion of denied claims in the sample. The CERT program includes denied claims in its sample for both the May and November reports. The HPMP samples denied claims only for the November report. Therefore, the HPMP denied claims data from the November 2005 report was used to make calculations for this May report. For more information please see "Two Measurement Programs: CERT and HPMP".

This report shows that 5.1% of the dollars paid nationally did not comply with one or more Medicare coverage, coding, billing, and payment rules. Projected overpayments were $11.9 B and the underpayments were $1.2 B. Thus, gross improper payments were projected as $13.1 B (i.e., $11.9 B plus $1.2 B).

The following table displays the error rates and improper payment amounts for the Medicare FFS Program for this reporting period.

This report also describes the other error rates in order to provide the most specific information available to target problem areas. Other error rates include error rates by specific contractor, error rates by service type, and error rates by provider type.

When comparing contractors, services, or provider types, it is important to note that the highest error rate does not necessarily indicate the highest projected improper payments. For example, the reported error rate is higher for chiropractic services than for E&M services, but the projected improper payments associated with claims submitted for E&M are higher than those for chiropractic services. Therefore, efforts focused on reducing improper payments may focus on E&M services despite the higher error rate in chiropractic services.

One of the performance goals for CMS is the reduction of improper payments made under the FFS program to 5.1% or less by the November 2006 reporting period. The findings in this report indicate that CMS has made progress toward its November 2006 goal. 

CMS is working with the QIOs to implement the following efforts to lower the paid claims error rate:

1. Using the First Look Analysis Tool for Hospital Outlier Monitoring (FATHOM) that generates state-specific hospital billing reports to help QIOs analyze administrative claims data and target interventions with hospitals,
2. Increasing and refining one-on-one educational contacts with providers found to be billing in error,
3. Developing projects with the QIOs addressing state-specific admissions necessity, coding concerns, and billing, as well as, conducting surveillance and monitoring of inpatient payment error trends by error type,
4. Distributing FATHOM generated hospital-specific reports to hospitals,  
5. Providing targeted education to hospitals with high numbers of medically unnecessary admissions,
6. Developing and distributing QIO-specific payment error cause analyses, and
7. Conducting national training on the use of FATHOM reports in compliance efforts.
8. Developing a reporting tool that provides monthly updates to error rates.

CMS is working with each Carrier/DMERC/FI to develop a plan that addresses the cause of the contractor’s errors, the steps the contractor will take to fix the problems, and other recommendations that will ultimately lower the error rate.

CMS is working with the CERT contractors to:

1. Reduce the lag time between the end of a reporting period and the production of the CERT report for that period, thereby providing Carriers/DMERCs/FIs with more timely error rates. CMS has accelerated the sampling and review process; beginning with this May report the interval between the last sampled claim for a report and it's publication has been reduced from 11 months to 8 months.  
2. Perform a small area variation analysis to produce maps of the United States that display CERT error rates and improper payment amounts geographically (available at www.CMS.HHS.gov/cert ).
3. Reduce the no documentation errors by:
   • Having CERT contractors make direct contact with every provider that has not provided a medical record or other requested information.
   • Developing a monthly newsletter to explain the importance of CERT and how the CERT program operates.
   • Sending the monthly newsletter to all Carriers/DMERCs/FIs for redistribution to their providers.
   • Providing a website ( www.CERTprovider.org) to help providers understand the importance of providing an address from which CERT can obtain the provider’s medical records.
   • Encouraging providers to use www.CERT provider.org to correct address errors in CERT records.
4. Decrease the insufficient documentation errors by: 
   • Improving the processes of requesting and receiving medical records. For example, the CERT Documentation Contractor uses fax servers to capture images of incoming faxes. In addition, they manually image all hardcopy medical records they receive.
   • Modifying the medical record request letters to clarify the components of the record needed for CERT review and to encourage the billing provider to forward the request to the appropriate location. Although the May and November 2006 reports will be partially effected, the full impact of this change will not be seen until the November 2007 report.  
   • Encouraging Carriers/DMERCs/FIs to educate providers about the importance of submitting thorough and complete documentation, including signing all plans of care, etc.

Reporting Periods
   Table :
Contractor Type Error Rates
   Table : Error Rates and Projected Improper Payments by Contractor Type
Other Error Rates
   Table :
Two Measurement Programs: CERT and HPMP
   Table : Error Rates Available in this Report
Weighting and Determining the Final Results
   Table : Summary of Inclusion vs. Exclusion
National Medicare FFS Error Rate
   Table : Error Rates and Projected Improper Payments by Contractor Type
   Table : National Error Rates by Year
Paid Claims Error Rate by Error Type
   Table : Summary of Error Rates by Category
   Table : Type of Error Comparison for 2005 and May 2006
No Documentation Errors
   Table : Top 20 Services with No Documentation Errors: Carriers/DMERCs/FIs/QIOs
Insufficient Documentation Errors
   Table : Top 20 Services with Insufficient Documentation: Carriers/DMERCs/FIs
Medically Unnecessary Services
   Table : Top 20 Medically Unnecessary Service: Carriers/DMERCs/FIs/QIOs
Incorrect Coding
   Table : Top 20 Services with Incorrect Coding Underpayment Errors: Carriers/DMERCs/FIs
   Table : Impact of One Level E&M (Top 20)
Other Errors
   Table : Top 20 Other Errors: Carriers/DMERCs/FIs/QIOs
Paid Claims Error Rate by Contractor Type
   Chart : Paid Claims Error Rates by Contractor Type
   Chart : Projected Improper Payments by Contractor Type
Carrier-Specific Error Rates
   Table : Error Rates and Improper Payments: Carriers
DMERC-Specific Error Rates
   Table : Error Rates and Improper Payments: DMERCs
FI-Specific Error Rates
   Table : Error Rates and Improper Payments: FIs
QIO-Specific Error Rates
   Table : Error Rates and Improper Payments: QIOs
Error Rates by Type of Service
   Table : Top 20 Service Types with Highest Improper Payments: Carriers
   Table : Top 20 Service Types with Highest Improper Payments: DMERCs
   Table : Top 20 Service Types with Highest Improper Payments: FIs
   Table : Top 20 Service Types with Highest Improper Payments: QIOs
Paid Claim Error Rates by Provider Type
   Table : Error Rates and Improper Payments by Provider Type: Carriers
   Table : Error Rates and Improper Payments by Provider Type: DMERCs
   Table : Error Rates and Improper Payments by Provider Type: FIs
   Table : Error Rates and Improper Payments by Provider Type: QIOs
Error Rates by Type of Service
   Table : Top 20 Service Type Error Rates: Carriers
   Table : Top 20 Service Type Error Rates: DMERCs
   Table : Top 20 Service Type Error Rates: FIs
   Table : Top 20 Service Type Error Rates: QIOs
Error Rates by Type of Error
   Table : Error Rates for Each Cluster by Type of Error: Carriers
   Table : Error Rates for Each Cluster by Type of Error: DMERC
   Table : Error Rates for Each Cluster by Type of Error: FI
   Table : Error Rates for Each Cluster by Type of Error: QIO
Paid Claims Error Rate by Service Type
   Table : Paid Claims Error Rates by Service Type: Carriers
   Table : Paid Claims Error Rates by Service Type: DMERCs
   Table : Paid Claims Error Rates by Service Type: FIs
   Table : Paid Claims Error Rates by Service Type: QIOs

The Social Security Act established the Medicare program in 1965. Medicare currently covers health care needs of people aged 65 and over, the disabled, people with End Stage Renal Disease (ESRD), and certain others that elect to purchase Medicare coverage. Both Medicare costs and the number of Medicare beneficiaries has increased dramatically since 1965. In fiscal year (FY) 2004, more than 42 million beneficiaries were enrolled in the Medicare program, and the total Medicare benefit outlays (both Medicare Fee-for-Service (FFS) and managed care payments) was estimated at about 301.1 B ^{(1)2005 CMS Statistics: U.S. Department of Health and Human Services, CMS pub. No 03455, September 2005}. The Medicare budget represents over 11% of the total federal budget.

CMS uses several types of contractors to prevent improper payments from being made for Medicare claims and admissions including Carriers, Durable Medical Equipment Regional Carriers (DMERCs), Fiscal Intermediaries (FIs), and Quality Improvement Organizations (QIOs).

The primary goal of each Carrier/DMERC/FI is to “Pay it Right” – that is, to pay the right amount to the right provider for covered and correctly coded services.  Budget constraints limit the number of claim reviews these contractors can conduct; thus, they must choose carefully which claims to review.  To improve provider compliance, Carriers/DMERCs/FIs must also determine how best to educate providers about Medicare rules and implement the most effective methods for accurately answering coverage and coding questions.  As part of its Improper Payments Information Act (IPIA) compliance efforts, and to help all Medicare FFS contractors better focus review and education, CMS has established the Comprehensive Error Rate Testing (CERT) program and Hospital Payment Monitoring Program (HPMP) to randomly sample and review claims submitted to Medicare.

The Department of Health and Human Services (DHHS), Office of Inspector General (OIG) estimated the Medicare FFS error rate from 1996 through 2002. The OIG designed their sampling method to estimate a national Medicare FFS paid claims error rate. Due to the sample size – approximately 6,000 claims – the OIG was unable to produce error rates by contractor type, specific contractor, service type, or provider type. The confidence interval for the national paid claims error rates during these years was +/- 2.5%. Following recommendations from the OIG, CMS increased the sample size for the CERT program when production began on the Medicare FFS error rate for the November 2003 Report. The sample size for error rates concerning Carriers/DMERCs/FIs in this reporting period was 116,417 paid and denied claims.  The sample size for error rates concerning QIOs for the reporting period was 39,821 discharges.

To better measure the performance of the Carriers/DMERCs/FI and to gain insight into the causes of errors, CMS decided to calculate not only a national Medicare FFS paid claims error rate but also a provider compliance error rate.

Paid Claims Error Rate

This rate is based on dollars paid after the Medicare contractor made its payment decision on the claim. This rate includes fully denied claims for Carriers/DMERCs/FIs/QIOs. The paid claims error rate is the percentage of total dollars that all Medicare FFS contractors erroneously paid or denied and is a good indicator of how claim errors in the Medicare FFS Program impact the trust fund. CMS calculated the gross rate by adding underpayments to overpayments and dividing that sum by total dollars paid. This error rate is quantified in dollars.

Provider Compliance Error Rate

This rate is based on how the claims looked when they first arrived at the Carrier/DMERC – before the Carrier/DMERC applied any edits or conducted any reviews. The provider compliance error rate is a good indicator of how well the Carrier/DMERC is educating the provider community since it measures how well providers prepared claims for submission. This error rate is quantified in dollars. CMS does not collect covered charge data from FIs; therefore, current FI data is insufficient for calculating a provider compliance error rate. This rate is not generated for QIOs.

CMS established two programs to monitor the accuracy of the Medicare FFS Program: the CERT program and HPMP. The main objective of these programs is to measure the degree to which CMS and its contractors are meeting the goal of Paying it Right. The HPMP monitors PPS short-term acute care inpatient hospital admissions. Beginning with the November 2005 reporting period, HPMP also monitors PPS long-term acute care inpatient hospital admissions. The CERT program monitors all other claims. The following figure (Figure 1) depicts the types of claims/admissions involved in each monitoring program.

Figure 1: Types of Claims/Admissions Reviewed By CERT and HPMP

The following table (Table 1) summarizes the data that is presented in this report.

Table 1: Error Rates Available in this Report

Monitoring Program	Type of Error Rate(s) Produced	Paid Claims Error Rate	Provider Compliance Error Rate
CERT+HPMP	Medicare FFS		Not Produced
CERT	Carrier/DMERC/FI
	Carrier-Specific
	DMERC-Specific
	FI-Specific		Not Produced
	Type of Service
	Type of Provider
HPMP	QIO Specific		Not Produced
	Type of Service		Not Produced
	Type of Provider		Not Produced

CMS established the CERT program to monitor the accuracy of Medicare FFS payments made by Carriers/DMERCs/FIs. The main objective of the CERT program is to measure the degree to which CMS and Carriers/DMERCs/FIs are meeting the goal of “Paying it Right”. See Appendix H for additional details about the sample used for this report.

Sampling and Medical Record Requests

For this report, the CERT Contractor randomly sampled 116,417 claims from Carriers/DMERCs/FIs. The CERT Contractor randomly selected about 208 claims each month from each Carrier/DMERC/FI. CERT designed this process to pull a blind, electronic sample of claims each day from all of the claims providers submitted that day. 

The CERT Contractor requested the medical record associated with the sampled claim from the provider that submitted the claim. The CERT Contractor sent the initial request for medical records via letter. If the provider failed to respond to the initial request after 30 days, the CERT Contractor sent three subsequent letters and made up to three phone calls to the provider.

In cases where the CERT Contractor received no documentation from the provider once 90 days had passed since the initial request, the CERT Contractor considered the case to be a no documentation claim and counted it as an error. The CERT Contractor considered any documentation received after the 90th day “late documentation.” If the CERT Contractor received late documentation prior to the documentation cut-off date for this report, they reviewed the records and, if justified, revised the error in each rate throughout the report. If the CERT Contractor received late documentation after the cut-off date for this report, they continued to count the case as a no documentation error.

Review of Claims

Upon receipt of medical records, the CERT Contractor's clinicians conducted a review of the claims and submitted documentation to identify any improper payments. They checked the Common Working File to see if the person receiving the services was an eligible Medicare beneficiary, to see if the claim was a duplicate and to make sure that no other insurer was responsible for paying the claim. When performing these reviews, the CERT contractor followed Medicare regulations, billing instructions, National Coverage Determinations (NCDs), coverage provisions in interpretive manuals, and the respective Carrier/DMERC/FI Local Coverage Determinations (LCDs), and articles.

Appeal of Claims

In the November 2003 reporting period, the CERT Contractor did not remove an error from the error rate if a provider appeal (using the normal appeals process) of a CERT initiated denial resulted in a reverse decision. In the November 2004 Report, the CERT Contractor implemented an appeals tracking system and began to back out overturned CERT initiated denials from the error rate; however, some contractors did not enter all the appeals information into the new tracking system before the cut-off date for the report. Therefore, CERT only backed out some of the determination reversals from the error rate in the November 2004 Report. As of the November 2005 report, all Carriers/DMERCs/FIs have the opportunity to ensure that all overturned appeals are entered into the appeals tracking system in sufficient time for production of the error rates.

Variation from the General Methodology

Due to a change in the FI shared system, the system did not correctly identify payments for non-PPS Hospital Inpatient claims for April through December of 2004. To correct for this problem, CMS used data for April 2005 through September 2005 as a basis to extrapolate improper payment and error rate estimates for non-PPS Hospital Inpatient claims in this report.

The CERT program skipped ahead three months as part of its acceleration plan; January, February, and March of 2005 were not reviewed. Although this will not effect the November 2006 report, it does overlap the time period for this report. In order to compensate, the CERT program used the data collected during October, November, and December of 2004 as a proxy for the data that would normally have been collected during the first quarter of 2005.

Naming Conventions

From time to time, a Carrier/DMERC/FI will choose to leave the Medicare program. When this occurs, CMS selects a replacement contractor to take over claims processing, error rate reduction efforts, etc. The cutover date is the term used to describe the date that the incoming contractor begins to receive and process claims while the outgoing contractor ceases operations. When preparing these improper payment reports, CMS has adopted a policy of listing the name of the contractor who processed claims from that jurisdiction for more than 6 months of the reporting period. 

The CMS established the HPMP to measure, monitor, and reduce the incidence of improper PPS acute care inpatient Medicare payments.  FIs process these payments; QIOs are responsible for ensuring accurate coding, admission necessity, and coverage. HPMP operates through the QIO program as QIOs have responsibility for ascertaining the accuracy of these payments through the physician peer review process. QIOs work with acute care hospitals to identify and prevent payment errors.

Sampling
Each month a CMS contractor selected a random sample of paid short-term acute care inpatient claims for each state from a clinical data warehouse that mirrors the National Claims History (NCH) database. To allow time for hospital claims submission, HPMP sampled claims after the completion of three months from the month of discharge; claims are 97.5% complete at this time. Beginning with the November 2005 report, HPMP also sampled paid long-term acute care and FI-denied claims (both short-term and long-term). For long term acute care claims, a national random sample not stratified by state was selected monthly. Claims that had been denied at the FI were selected as a single, national random sample. The HPMP sampled a total of 39,821 claims from 52 states and jurisdictions (all 50 states plus Puerto Rico and Washington, D.C.).

Review of Claims
The CMS contractor that performed the sampling of PPS short-term acute care sample claims provided the sampled claims to the Clinical Data Abstraction Centers (CDACs) for screening. The CDACs validated Diagnosis Related Groups (DRGs), performing independent recoding and admission necessity screening based upon the information provided in the submitted record. Qualified coding specialists performed DRG coding validation. CDAC nurse reviewers performed admission necessity screening. Admission screening involved a detailed examination of each medical record using specific modules of the InterQual admission appropriateness criteria set. In addition, Maryland records were screened for length of stay (Maryland is the only waivered non-PPS state); Maryland length of stay errors are included under medically unnecessary services.

The CDACs did not follow-up with providers; the CDAC referred records that failed screening as well as those that were not received in a timely manner to the responsible QIO for case review. Under the case review process, records are again validated for coding and screened for admission necessity. Those records failing admission necessity screening are sent to peer physician review under which hospitals have further opportunity to supply documentation.

The long-term acute care sample was sent directly to QIOs and was not screened by the CDAC. Denied claims were handled only by the CDAC and were not sent to the QIOs.

The error rates were weighted so that each Carrier/DMERC/FI/QIO contribution to the error rate was in proportion to its size (as measured by the percent of allowed charges for which they were responsible). The confidence interval is an expression of the numeric range of values for which CMS is 95% certain that the mean values for the improper payment estimates will fall. As required by the IPIA, the CERT program has included an additional calculation of the 90% confidence interval for the national error rate calculation.

All national improper payment estimates from 1996 to present EXCLUDE coinsurance, deductibles and reductions to recover previous overpayments. When CMS began calculating the additional error rates for contractor-specific, service-type and provider-type in the November 2003 and November 2004 reports, these types INCLUDED coinsurance, deductibles and reductions. The CERT program was unable to exclude them from the improper payment amounts due to system limitations. CMS has since implemented new systems and revised methodology that has allowed for the EXCLUSION of coinsurance, deductibles and reductions from all improper payment amounts beginning with the November 2005 reporting period. As a result, the improper payment estimates from the November 2005 report and forward can not be compared to previously published estimates for contractor-specific, service-type, or provider-type calculations. However, since error rate estimates are unaffected, they can be compared across all reports.

Since error rates are calculated as the sum of overpayments and underpayments divided by the original dollars paid, estimated error rates >100% are possible. In particular, this situation can occur when very large underpayments are found among sampled records. The size of the associated confidence interval which represents the extent of variability should always be considered when evaluating estimated payment error rates.

In the CERT program, Carriers/DMERCs/FIs are notified of detected overpayments so that they can implement the necessary adjustments. Carriers/DMERCs/FIs are also notified of underpayments but they are not currently required to make payments to providers for underpayments identified in the CERT program. CMS plans to instruct the Carriers/DMERCs/FIs to make payments to providers in underpayment cases identified for the November 2006 and later reports. For more information about overpayments see Appendix F, for underpayments, see Appendix G. Sampled claims for which providers failed to submit documentation were considered overpayments.

QIOs in the HPMP notified FIs of adjustments necessary due to overpayment and underpayment errors identified by the program. When a QIO determined that a DRG coding change was required, the FI was also informed of the appropriate DRG. In addition, the FI was informed when: a stay was found to be inappropriate, the requested medical records were not supplied, or insufficient documentation was provided. In each case, the stay was denied and was considered an overpayment. FIs were responsible for determining payment adjustments for claims found to be in error. The QIOs did not determine adjustment amounts nor did they implement payment adjustments.

Providers can appeal denials (including no documentation denials) following the normal appeal processes by submitting documentation supporting their claims. For the November 2003 Report, the CERT program did not consider the outcome of appeal determinations. However, beginning with the claims in the November 2004 Report, the CERT program considered the outcome of any appeal determinations that reversed the CERT program’s decision when computing the error rates.  The CERT program deducted $186.5 M in appeals reversals from the error rates contained in the this report. Under the QIO case review process, hospitals have multiple opportunities to appeal a QIO decision. Cases are not included as payment errors for all HPMP calculations until all hospital case review appeals are complete.

The CERT program identified $1.1 M in actual overpayments and, as of the final cut-off date for this report, Carriers/DMERCs/FIs had collected $543,862 of those overpayments.  The HPMP identified $12.3 M in overpayments and, as of the final cutoff date for this report, the FIs had processed $10.4 M in HPMP adjustments. CMS and its contractors will never collect a small proportion of the identified overpayments because:

• The responsible provider appealed the overpayment and the outcome of the appeal overturned the CERT decision.
• The provider has gone out of business.

However, for all other situations, the Carrier/DMERC/FI will continue their attempts to collect the overpayments.

CMS aims to accomplish three error rate goals under the Government Performance and Results Act (GPRA).

1. Reduce the National Medicare FFS Paid Claims Error Rate.

• By November 2006, reduce the percent of improper payments under Medicare FFS to 5.1%.
• By November 2007, reduce the percent of improper payments under Medicare FFS to 4.9%.
• By November 2008, reduce the percent of improper payments under Medicare FFS to 4.7%.

2. Reduce the Contractor-Specific Paid Claim Error Rate

• By November 2006, 50% of Medicare claims will be processed by contractors with an error rate less than or equal to the national error rate for November 2005.
• By November 2007, 75% of Medicare claims will be processed by contractors with an error rate less than or equal to the national error rate for November 2006.
• By November 2008, every Medicare claim will be processed by contractors with an error rate less than or equal to the national error rate for November 2007.

3. Decrease the Provider Compliance Error Rate

• In November 2006, decrease the Provider Compliance Error Rate 20% over the November 2005 level.
• In November 2007, decrease the Provider Compliance Error Rate 20% over the November 2006 level.
• In November 2008, decrease the Provider Compliance Error Rate 20% over the November 2007 level.

CMS will use the error rate findings described in this report to determine underlying reasons for claim errors and to adjust its action plans to improve compliance in payment, documentation, and provider billing practices. The tracking and reporting of error rates also helps CMS identify emerging trends and implement corrective actions designed to accurately manage all Medicare FFS contractors’ performance. In addition, the error rates will provide all Medicare FFS contractors with the guidance necessary to direct claim review activities, provider education efforts, and data analysis. Carriers/DMERCs/FIs also use the error rate findings to adjust their Error Rate Reduction Plans. CMS evaluates QIOs under their contract on payment error rates.

The national paid claims error rate in the Medicare FFS program for this reporting period is 5.1% (which equates to $13.1 B). The 95% confidence interval for Medicare FFS program paid claims error rate was 4.7% - 5.5%.  The 90% confidence interval (required to be reported by IPIA) was 4.7% - 5.4%.

Table 3a summarizes the overpayments, underpayments, improper payments, and error rates by contractor type.

Table 3a: Error Rates and Projected Improper Payments by Contractor Type

Table 3b summarizes the overpayments and underpayments, improper payments and error rates by year.

Table 3b: National Error Rates by Year

Table 3c summarizes the percent of the total dollars improperly allowed by error category for this and previous reports.

Table 3d summarizes the percent of total dollars improperly allowed by error category and contractor type.

No documentation means the provider did not submit any documentation to support the services provided.^{(4)Due to the extremely low insufficient documentation error rate for QIOs, any insufficient documentation errors have been added to the no documenation rate rather than the insufficient documentation category.} No documentation errors accounted for 0.6% of the total dollars all Medicare FFS contractors allowed during the reporting period. QIO data is categorized in a different manner than the data for Carriers/DMERCs/FIs; therefore, the QIO no documentation estimates include claims that are categorized as insufficient documentation for Carriers/DMERCs/FIs. This data breaks down by contractor type as follows:

Table 4a is a combined list of the services with the highest projected improper payments due to no documentation errors for all contractor types. All series 4 tables are sorted in descending order by projected improper payments.

Table 4a: Top 20 Services with No Documentation Errors: Carriers/DMERCs/FIs/QIOs

The following are examples of No Documentation errors:

• A Carrier paid $91.89 for an office visit and services.  After repeated attempts from the CERT Contractor to obtain the supporting medical records from the provider, the provider indicated that they could not locate the records.  As a result, the CERT Contractor counted the entire payment as an error.  See Appendix E for more information about no documentation errors.
• A hospital submitted and was paid for a short-term acute care inpatient claim totaling $7,188.63.  However, when the substantiating medical record was requested, the hospital failed to provide the record.  Thus, the entire amount was considered an error.

Insufficient documentation means that the provider did not include pertinent patient facts (e.g., the patient’s overall condition, diagnosis, and extent of services performed) in the medical record documentation submitted.^{(5)Due to the extremely low insufficient documentation error rate for QIOs, any insufficient documentation errors have been added to the no documenation rate rather than the insufficient documentation category.}  

Insufficient documentation errors accounted for 1.0% of the total dollars allowed during the reporting period. This data breaks down as follows:

In several cases of insufficient documentation, it was clear that Medicare beneficiaries received services, but the physician’s orders or documentation supporting the beneficiary’s medical condition were incomplete. While these errant claims did not meet Medicare reimbursement rules regarding documentation, CMS could not conclude that the services were not provided.

In some instances, components of the medical documentation were located and maintained at a third party facility.  For instance, although a lab may have billed for a blood test, the physician who ordered the lab test maintained the medical record. If the billing provider failed to contact the third party or the third party failed to submit the documentation to the CERT Contractor, CMS counted the claim as a full or partial insufficient documentation error.

Table 4b is a combined list of the services with the highest insufficient documentation paid claims error rates for Carriers/DMERCs/FIs. This table does not include QIOs.

Table 4b: Top 20 Services with Insufficient Documentation: Carriers/DMERCs/FIs

The following is an example of an insufficient documentation error:

An FI paid an outpatient hospital $96.00 for a clinic visit. The documentation did not include a doctor’s order, a medical history, or notes to support the diagnosis listed on the claim form. As a result, the CERT Contractor counted the entire payment as an error.

Medically Unnecessary Services includes situations where the CERT or HPMP claim review staff identifies enough documentation in the medical record to make an informed decision that the services billed to Medicare were not medically necessary. In the case of inpatient claims, determinations are also made with regard to the level of care; for example, in some instances another setting besides inpatient care may have been more appropriate. If a QIO determines that a hospital admission was unnecessary due to not meeting an acute level of care, the entire payment for the admission is denied.

Medically Unnecessary Service errors accounted for 1.4% of the total dollars allowed during the reporting period. This data breaks down as follows:

Table 4c lists the top twenty medically unnecessary services for Carriers/DMERCs/FIs.

Table 4c: Top 20 Medically Unnecessary Service: Carriers/DMERCs/FIs/QIOs

The following are examples of medically unnecessary services:

• An FI paid a Skilled Nursing Facility (SNF) $49.22 for 30 minutes of therapeutic procedures; however, the physician certification for the services did not cover the dates for which the services were billed. As a result, the reviewer determined that the services were not medically necessary and counted the claim as an error.
• A short-term acute care inpatient claim for $13,412.00 was submitted and paid. The patient was admitted for an elective percutaneous transluminal coronary angioplasty (PTCA). Upon review, it was discovered that the patient was stable and underwent the procedure without preoperative or postoperative complications; therefore, the patient should have been treated in an outpatient hospital observation setting. The entire claim amount was considered an error.

Providers use standard coding systems to bill Medicare. For most of the coding errors, the medical reviewers determined that providers submitted documentation that supported a lower code than the code submitted (in these cases, providers are said to have overcoded claims). However, for some of the coding errors, the medical reviewers determined that the documentation supported a higher code than the code the provider submitted (in these cases, the providers are said to have undercoded claims).

Incorrect Coding errors accounted for 1.8% percentage of the total dollars allowed during the reporting period. This data breaks down as follows:

A common error involved overcoding or undercoding by one level on a scale of five code levels. Published studies suggest that under certain circumstances, experienced reviewers may disagree on the most appropriate code to describe a particular service. This may explain some of the incorrect coding errors in this report. CMS is investigating procedures to minimize the occurrence of this type of error in the future.

Table 4d lists the services with paid claims error rates that include undercoded coded claims for Carriers/DMERCs/FIs.

Table 4d: Top 20 Services with Incorrect Coding Underpayment Errors: Carriers/DMERCs/FIs

The following are examples of coding errors:

• A Carrier paid a physician $135.42 for the evaluation and management of an established patient. This procedure requires at least two of three key components: a detailed history, a detailed examination, and/or medical decision-making of moderate complexity. The medical reviewer determined that the services did not meet the minimum criteria for the key components since a licensed nurse rendered the services rather than physician. Instead, the medical record met the criteria for a lower level service that would have paid $91.59. The CERT reviewer determined that the service should have been billed at a lower E&M code and counted $43.83 as paid in error.
• A hospital billed a procedure without documenting that the procedure was performed. The removal of the procedure code from the billing resulted in an adjustment to a lower DRG, which, in turn, resulted in a reduced payment. In this case, $9,863.77 was determined to be in error.

The OIG and CMS have noted problems with certain procedure codes for the past several years. These problematic codes include Current Procedural Terminology (CPT) codes 99214 (office or other outpatient visit), 99232 (subsequent hospital care level 2) and 99233 (subsequent hospital care level 3). See Appendix F for more information on problematic codes.

Table 4e provides information on the impact of 1 level disagreement between Carriers and providers when coding evaluation and management codes.

Table 4e: Impact of One Level E&M (Top 20)

For more data pertaining to incorrect coding errors, see Appendix G.

Under CERT, other errors include instances when provider claims did not meet benefit category requirements or other billing requirements. Errors for services that did not meet the benefit category requirements were more common among claims submitted to DMERCs than among claims submitted to Carriers and FIs. The absence of a valid physician’s order made some DME items non-covered because an order or Certificate of Medical Necessity (CMN) was required to meet the benefit category requirements for the DME item.

Under HPMP, other errors include quality of care and billing errors. Billing errors include payments for claims where the stay was billed as non-exempt unit but was exempt, outpatient billed as inpatient, and HMO bills paid under FFS. Most other errors occur on claims for which QIOs are responsible.

Other errors accounted for 0.3% of the total dollars allowed during the reporting period. This data breaks down as follows:

Table 4f lists the services with other errors and the associated paid claims error rate.

Table 4f: Top 20 Other Errors: Carriers/DMERCs/FIs/QIOs

The following are examples of other errors:

• A Carrier paid $76.64 for routine foot care. Routine foot care is statutorily excluded. Therefore, the CERT Contractor counted the full payment as an error.  
• A hospital billed for a short-term acute care inpatient stay when the orders indicated that it should have been billed as an observation stay. The dollars paid in error were $5,778.71.

Figures 3 and 4 summarize the paid claims error rate and projected improper payments during the reporting period for each type of contractor. This data breaks down by contractor type as follows:

The following figures (Figures 3 and 4) detail the paid claim error rates and projected improper payments by contractor type.

Figure 3: Paid Claims Error Rates by Contractor Type

d

Figure 4: Projected Improper Payments by Contractor Type

d

Table 5 contains error rates and improper payment amounts for Carriers. It is sorted in descending order by error rate.

Table 5: Error Rates and Improper Payments: Carriers

For paid claim error rates, provider compliance error rates and no resolution rates by contractor and provider type, see Appendix C.

Table 6 contains DMERC specific error rates and improper payment amounts. It is sorted in descending order by error rate.

Table 7 contains error rates and improper payment amounts for FIs. It is sorted in descending order by error rate.

Table 7: Error Rates and Improper Payments: FIs

For error rates and improper payment amounts for individual contractors, paid claims error rates by cluster and type of error, and improper payment amounts for clusters, see Appendix C.

Table 8 contains QIO specific short-term PPS acute care hospital error rates and improper payment amounts, total short-term PPS acute care hospital error rates and improper payment amounts, total PPS long term acute care hospital error rates and improper payment amounts, and total error rates and improper payment amounts for all types of facilities for which QIOs are responsible. It is sorted alphabetically by state.

Table 8: Error Rates and Improper Payments: QIOs

For paid claims error rates by contractor and type of error and improper payment amounts for contractors, see Appendix C.

Table 9 displays the paid claims error rates for each type of service by type of error. This series of tables is sorted in descending order by projected improper payments. All estimates in this table are based on a minimum of 30 lines in the sample.

Table 9a: Top 20 Service Types with Highest Improper Payments: Carriers

Table 9b: Top 20 Service Types with Highest Improper Payments: DMERCs

Table 9c: Top 20 Service Types with Highest Improper Payments: FIs

Table 9d: Top 20 Service Types with Highest Improper Payments: QIOs

The table 10 series presents error rates by provider type. The tables include the top provider types based on improper payments for providers that bill each type of contractor. All estimates are based on a minimum of 30 lines in the sample. This series of tables is sorted in descending order by projected improper payments.

The CERT program is unable to calculate provider compliance error rates for FIs due to systems limitations.

Table 10a: Error Rates and Improper Payments by Provider Type: Carriers

Table 10b: Error Rates and Improper Payments by Provider Type: DMERCs

Table 10c: Error Rates and Improper Payments by Provider Type: FIs

Table 10d: Error Rates and Improper Payments by Provider Type: QIOs

CMS continues to make progress lowering the no documentation rate. Historically, the no documentation issue was more pronounced in the CERT program than in HPMP. As of this report, the DMERC and FI no documentation rates match the QIO rate of 0.1% while the Carriers remain unchanged at 0.4%. This difference is due to several factors: first, providers are more likely to respond to HPMP requests since the average claim value is much higher; second, the providers included in the HPMP were more familiar with that program; and third, HPMP pays PPS inpatient hospital providers separately for the cost of supplying medical records while CERT does not. The cost of supplying such medical records by non-PPS inpatient hospital providers is included in the fees they are paid for each service, and thus CERT is prohibited from paying the providers' cost of supplying medical records.

Reasons for no documentation errors include:

• The provider indicated that the beneficiary does not exist,
• The provider indicated that they submitted the claim for the wrong date of service,
• The provider responded but did not provide the medical record for some reason (such as fear of violating HIPAA or refusing to submit without separate payment for copying/mailing charges),
• The provider commented that they had gone out of business,
• The provider indicated that a third party is in possession of the needed medical record, or
• The provider did not respond at all.

CMS began the following corrective actions in 2005 to address the no documentation problem:

1. CMS and the Carriers/DMERCs/FIs have been educating providers about the CERT program so that providers are not hesitant about supplying medical records.
2. The CERT contractor developed a Web-based mechanism to allow Carriers/DMERCs/FIs to see which providers respond to CERT documentation requests. CMS then encouraged Carriers/DMERCs/FIs to contact non-responding providers.
3. CMS revised the medical record request letters to emphasize that faxing is the most effective way to submit medical records.
4. CMS required the CERT Review Contractor to implement an appeals tracking system. The CERT Review Contractor used the appeals information to adjust the errors when the provider appealed a CERT decision and the appeals review concluded that the claim should have been paid. Since providers that initially failed to respond to CERT requests for medical records frequently appealed the denial, this change (adjusting the error rate to account for appeals decisions) lowered the percent of the error rate due to no documentation.
5. CMS published a monthly CERT provider newsletter that contained CERT news, helpful hints, and documentation submission reminders.
6. CMS is exploring the possibility of using a secure online system to submit electronic medical records (EMR). The pilots will help CMS test whether:
   • A Medicare Carrier/DMERC/FI can realize efficiencies in their medical review program and lower their error rate by accepting computerized and imaged medical records, and
   • It would be feasible for the CERT program to accept computerized or imaged medical records from providers via a secure Web system.

CMS initiated several new corrective actions that will have an impact on the November 2006 report.  A new contractor was hired to specialize in requesting and receiving medical records reviewed by the CERT program. This new contractor, known as the CERT Documentation Contractor, has implemented new policies such as:

1. Calling providers before sending correspondence in order to verify contact information,
2. Offering to fax request letters to providers who can receive faxes,
3. Developing a website that allows providers to customize the delivery address for CERT medical record request letters, and
4. Developing a clear policy and documentation process to deal with medical records that are lost or damaged due to disaster.

The insufficient documentation problem was caused by multiple factors, including:

• Some providers remain confused about exactly what they needed to submit to the CERT contractor.
• Portions of the medical record were at a location within the billing provider organization other than the location to which the CERT contractor sent the request and the provider did not forward the request to the appropriate location (e.g., the request was sent to the home office but the record was located in a field office).
• Portions of the medical record were located at a third party and the provider did not contact the third party (e.g., the request was sent to the billing physician but the record was located at the hospital).
• Providers failed to properly document the billed service in the medical record (e.g., the plan of care lacked the required physician signature).
• Providers misplaced portions of the medical record.

CMS has already undertaken the following corrective actions aimed at reducing the insufficient documentation rate:

1. The CERT program now solicits improved addresses from Carriers/DMERC/FIs and providers themselves.
2. CMS modified the medical record request letters to clarify the components of the record needed for CERT review. The new letters also encourage the billing provider to forward the request to the appropriate location if the medical record is not on-site.
3. A new provider address customization website allows providers to supply the CERT program with alternate, third party addresses.
4. CMS now customizes the second chance letters to list the parts of the medical record that are needed to complete the review.
5. CMS encouraged Carriers/DMERCs/FIs to educate providers about the importance of submitting thorough and complete documentation.

The QIOs were responsible for the largest portion of the improper payments due to medically unnecessary services.

CMS has already undertaken actions to correct this problem:

1. CMS has developed a tool that generates state-specific hospital billing reports to help QIOs analyze administrative claims data.
2. CMS has developed projects with the QIOs that address problems identified in state-specific hospital billing reports.
3. CMS will provide hospitals with training on using comparative data reports to help them prioritize auditing and monitoring efforts with the goal of preventing payment errors.
4. CMS conducts an annual payment error cause analysis to discern sources of payment error. CMS will be developing and distributing QIO specific payment error cause analyses.
5. CMS is working to address possible issues with observation versus inpatient admission that could be contributing to inappropriate inpatient admissions.
6. CMS has completed and distributed an extensive workbook designed to be a resource for hospitals in their compliance efforts and activities.
7. CMS has tasked each Carrier/DMERC/FI with developing an Error Rate Reduction Plan (ERRP) that targets medical necessity errors in their jurisdiction.

CMS will continue the following corrective actions:

1. QIOs will continue to work with hospitals to reduce coding errors through educational efforts and the use of statewide and hospital specific reports from First Look Analysis Tool for Hospital Outlier Monitoring (FATHOM). FATHOM is designed to identify emerging problem areas through data analysis. FATHOM includes reports on DRG-based target areas such as the ratio of the count of discharges with DRG 0079 (respiratory infections and inflammations age >17 with complications or comorbidity) to the count of discharges with DRGs 079, 080, 089, or 090 (lower paying pneumonia DRGs).
2. CMS is considering a resolution passed by the American Medical Association (AMA), the owner of the physician coding system, that recommends CMS defer to the billing physician's judgment in evaluation and management cases where a reviewer and the billing physician disagree by only one coding level.

The delays in the production of the error rate reports are inherent in the current structure of the CERT and HPMP processes. For example, prior to March 2005, the CERT program requested sampled claim information from the Carriers/DMERCs/FIs daily but requested re-pricing data from them only once a month. In HPMP, claims are sampled four months after discharge in order to allow for hospital clams submission times and the records that undergo QIO case review can go through multiple levels of physician review and appeals. The majority of the impact from CMS and contractor error rate reduction efforts after November 2004 will not be evident until the November 2006 Report.

CMS has taken the following actions:

1. The CERT program now requests sampled claim information from the Carriers/DMERCs/FIs on a daily basis.
2. The CERT Documentation Contractor's medical record request letter asks the providers to respond in 30 days. However, claims are not marked as an error until day 90.
3. The CERT program now requests re-pricing data from the Carriers/DMERCs/FIs twice a month.
4. The CERT program plans to conduct a complete analysis of its procedures to identify any other areas where the program can be more time efficient. The CERT program will also advance the time period covered by each November report by three months to decrease the time lag between claim sampling and error reporting.
5. Due to issues related to claim submission and time to complete case review, it is difficult to decrease the lag time for HPMP without adversely affecting the accuracy of the estimate. However, by affecting when data is reported internally, HPMP will be able to decrease the lag time by two months to four months. Under their current contract, QIOs are investigating where efficiency in the case review process can be improved and this potentially will eliminate unnecessary time lags in the case review process and further reduce the lag time. It should be noted that for HPMP, short-term acute care claims were sampled by discharge date.

CMS continues to take the following general corrective actions:

1. CMS has directed Medicare contractors to develop local efforts to lower the error rate by submitting Error Rate Reduction Plans that address the cause of the errors, identify the steps they are taking to fix the problems, and provide recommendations to CMS. CMS closely monitors and evaluates the development and implementation of the Contractor Error Rate Reduction Plan for each each Carrier/DMERC/FI.
2. Contractors have implemented educational programs that entail both broad-based efforts and more focused communication with specific providers or provider groups concerning specific billing problems. The broad-based efforts include Websites that provide detailed information on Medicare payment policies, provider training sessions, open door forums, and written materials that explain payment policies in detail.
3. CMS has required its Carriers/DMERCs/FIs to develop annual medical review strategies to reduce the error rates. CMS ties contractor budgets to medical review strategies, evaluates contractor performance based on how well each contractor accomplishes the goals, and conforms to the procedures included in their strategies. CMS required its contractors to intensify their one-on-one educational programs to target known problems that contribute to error rates.
4. CMS will develop and install new Correct Coding Initiative edits to target identified problem areas.
5. CMS will use the Carrier/DMERC/FI-specific error rates in the contractor performance evaluation program.
6. CMS will encourage contractors to address provider billing/payment questions more consistently.
7. CMS is implementing a major initiative to determine if Recovery Audit Contractors (RACs) can lower the error rate by identifying and recovering Medicare overpayments. CMS has begun a three-year demonstration in the states of California, New York, and Florida as required by Section 306 of the Medicare Modernization Act. For more information about this demonstration, see www.cms.hhs.gov/researchers/demos/MMAdemolist.asp. CMS will closely monitor provider compliance error rates and paid claim error rates in these three states to see if providers in RAC states improve their provider compliance error rate faster than those in non-RAC states. During CY 2006, CMS will be looking to see if the Carriers/DMERCs/FIs in these states are able to lower their paid claim error rates more rapidly than other states by reducing post payment medical review and increasing provider education and prepayment medical review.
8. The CERT program completes a small area variation analysis of the Carrier/DMERC/FI error rates using data from the Improper Medicare Fee-for-Service report. This annual special study produces maps that depict local error rate problem areas. This study facilitates a better understanding of how error rates vary geographically and where CMS and the Carriers/DMERCs/FIs should focus corrective actions.
9. The Medicare Modernization Act requires that CMS publish a list of over-utilized codes. The list provides service type error rates for each CERT cluster group. The CERT program develops and distributes the list annually via the CERT public website (www.cms.hhs.gov/CERT).
10. The HPMP is developing national and state-specific models for predicting payment errors. This study facilitates a better understanding of areas prone to payment error and where QIOs should focus corrective actions.
11. CMS will form a workgroup to address the high provider compliance error rate. This workgroup will examine causes of the errors and develop recommendations for corrective actions.

The full copy of The Supplementary Appendices for the Improper Medicare Fee-for-Service Payments Report may be downloaded here. The full file is an Adobe PDF file of approximately 1.5 MB.

Error Rates by Cluster by Type of Error

Appendix C provides error rates for each cluster by type of error.

Alternate No Documentation Rate

Appendix E provides an alternate no documentation rate based on the ratio of medical records received to medical records requested. This additional information is provided in order to assist contractors with their efforts to lower the no documentation rate. The appendix provides three no documentation rates for the following categories:

• All no documentation,
• No documentation rates that have a value of less than $100 in overpayments, and
• No documentation rates with a value of $100 or more in overpayments.

The alternate no documentation rates are different from the earlier error due to no documentation rates because the alternate rates are based on the number of records requested but not received rather than the dollars in error due to no documentation.

Under CERT, no documentation errors fall into six sub-categories:

1) Beneficiary Issue - This category includes the following:

• the provider indicated that no such patient exists or,
• the provider indicated that although this patient exists, no such service was provided to the patient.

2) Wrong Date of Service – For this category, the provider indicated that they do not have a medical record for the date of service in the CERT request, but they do have a medical record for that service just a few days before or after the service in question.  The claim in question may be a duplicate claim.

3) Medical Record Issue – This category includes instances where the provider responded to a CERT documentation request but did not provide a medical record to support payment of a claim.  The category includes the following:

• The provider indicated that another department within the provider organization is responsible for fulfilling documentation requests,
• The provider indicated they have the medical record but refused to provide it without payment for copying/mailing charges,
• The provider indicated that it is a HIPAA violation to supply the record,
• The provider submitted a statement that the record was destroyed as a result of extenuating circumstances (e.g., fire, flood, explosion),
• The provider indicated in writing that they did not provide a service to the beneficiary on the date indicated on the claim, or
• The provider indicated they have the medical record but refuse to provide it for some other reason.

4) Billing Provider Issue - This category contains the following reasons for no documentation:

• The provider number has been deactivated,
• The provider has gone out of business, or
• The provider commented but failed to produce a record.

5) Third Party Record - This category contains the situations in which the provider indicated that a different provider - a third party - has the relevant medical record.

6) Did Not Respond – No response to any CERT documentation request.

Table 11a: Top 20 Service Type Error Rates: Carriers

Table 11b: Top 20 Service Type Error Rates: DMERCs

Table 11c: Top 20 Service Type Error Rates: FIs

Table 11d: Top 20 Service Type Error Rates: QIOs

Table 12a: Error Rates for Each Cluster by Type of Error: Carriers

Table 12b: Error Rates for Each Cluster by Type of Error: DMERC

Table 12c: Error Rates for Each Cluster by Type of Error: FI

Table 12d: Error Rates for Each Cluster by Type of Error: QIO

Table series 13 displays the paid claims error rate by service type for each contractor type. Each table is sorted by projected improper payments from highest to lowest. All estimates are based on a minimum of 30 claims in the sample. 

Table 13a: Paid Claims Error Rates by Service Type: Carriers

Table 13b: Paid Claims Error Rates by Service Type: DMERCs