State Employees Workers' Compensation Study Appendicies
Table of State Employees Workers' Compensation Appendices
Appendix D.1: State Employees Workers' Compensation Study Participants
Organization Description of Role Comptroller of Public Accounts Provided overall project management. Solicited bids and selected a contractor for the study. Evaluated and enhanced study design from the first study. Coordinated project meetings, interviews and information requests from all agencies involved with the Medicaid, State Employees Workers' Compensation and Medicaid Managed Care studies. Compiled and validated all study information received. Worked with consultants on aspects of the study, including completing information sheets, assisting with medical record reviews and resolving issues as they arose. J. Shank, R.N. Contracted consultant to perform the online injury claim and medical record reviews. Participated in SORM meetings and on-site Argus tour. Provided program knowledge to the CPA project team and assisted with the compilation of study findings. State Auditors Office (SAO) Provided assistance with evaluating and enhancing the Medicaid study methodology. Validated the Medicaid sample universe data. Performed the overpayment rate computation. State Office of Risk Management (SORM) Provided on-site access to the study claimant's online injury claim and medical bill information. Provided copies of all medical record documentation received with medical bills and coordinated receipt and delivery of requested medical records. Assisted with program information. Participated in the findings review.
Appendix D.2: State Employees Workers' Compensation Organization Chart
Source: State Office of Risk Management.
Appendix D.3: State Employees Workers' Compensation Comparison Table: 1998, 2001 and 2003
The table below compares the criteria of the three studies published in December 1998, January 2001 and January 2003.
* Claims refer to state employee workers' compensation injury or illness claims made to SORM; bills refer to medical bills submitted by providers to SORM for reimbursement.
Criteria December 1998 Study January 2001 Study March 2003 Study March 2005 Study Sample Size 200 200 200 200 Number of
278 bills 284 bills 256 bills 732 bills Sample
Medical bills with services rendered in November 1997 and paid through date of profiling/sampling. Medical bills with dates of service in September, October and November 1999, and paid through March 1, 2000. Medical bills with dates of service in September, October and November 2001, and paid through January 4, 2002. Medical bills with dates of service in September, October and November 2003, and paid through January 31, 2004. Sample Service Exclusions None None Transportation bills paid directly to claimants Transportation bills paid directly to claimants Sampling Selection Program 200 SORM claims* were randomly selected, pulled and grouped into four categories: hospitals, medical, miscellaneous and pharmacy. The sampling methodology was the same as that used in the previous study. The sampling methodology was the same as that used in the previous studies, except transportation bills were excluded. The sampling methodology was the same as that used in the 2003 study. Patient Telephone Interviews Telephone interviews were attempted for all patients with a telephone number. A professional telephone survey firm conducted these interviews. Face-to-face interviews were attempted with patients who could not be interviewed by phone. As in the 1998 study, telephone interviews were attempted with all claimants who had a telephone number. Face-to-face interviews were not attempted with patients who could not be interviewed by phone. Telephone interviews were attempted with all claimants who had a telephone number. Face-to-face interviews were not attempted with claimants who could not be interviewed by phone. Eliminated due to the unreliability of results from the 2001 and 2003 studies. Online Claim Review (Contextual Data Analysis) Analysts reviewed a three-month period of claims surrounding the sample day using its SORM's Claims Management System. Analysts reviewed a four-month period of claims surrounding the sample day using its SORM's Claims Management System. In addition, adjuster notes and references to medical information found in the patient's file in the Claims Management System were reviewed. Same as the 2001 study. Same as the 2001 and 2003 studies. Medical Records Review Medical records were requested if there was a problem noted in either the patient interview or the contextual data analysis. Analysts conducted an initial review of the medical records; no records were sent for a third-party review. Medical records were not requested for the SORM sample due to the inadequate response from providers in the previous study and the lack of SORM authority to obtain medical records. The only hard-copy documentation provided was remittance advice for the services provided. A professional nurse utilization reviewer with expertise in the Texas Workers' Compensation program reviewed medical records submitted with the medical bills along with medical bills. A professional nurse utilization reviewer with expertise in the Texas Workers' Compensation program reviewed medical records submitted with the medical bills along with medical bills.
Appendix D.4: State Employees Workers' Compensation Medical Record Review Worksheet
Appendix D.5: State Employees Workers' Compensation Discrepancy Error Code Table
Medical Record Review Error Code Error Code
Review Error Description 1 Processing The service billed exceeds the National or Local Coverage Determinations for a physician, MD/DO 2 Processing The service billed exceeds the National or Local Coverage Determinations for a chiropractor, DC 3 Processing The service billed is incorrect per the Medicare Correct Coding Initiative (CCI) Edits 4 Processing Procedures billed separately that should be billed as a panel (laboratory tests) or global (surgical procedures) under the Texas Workers' Compensation Program 5 Processing Not according to Medical Fee Guidelines (i.e. incorrect procedure coding, including descriptors) 6 Documentation The medical record has no documentation of the sample service 7 Duplication Duplicate payment 8 Clerical/Billing Clerical/billing error 9 Upcoding Global charge to procedure 10 Non-covered Service The procedure billed is not payable according to administrative rule 11 Service Not Related to Injury The service billed is unrelated to the claimant's documented injury claim. 12 Medical Necessity Not medically necessary 13 Payment Exceeds Pharmacy Fee Guidelines 14 Not Authorized Preauthorization issue
Appendix D.6: State Employees Workers' Compensation Sample Statistical Data Tables
Exhibit 1: State Employees Workers' Compensation Study Overpayment Calculation Table
Total Number of Sample Services (Medical Bills) Amount Paid for Sample Services (Medical Bills) Total Number of Services (Medical Bills) Overpaid Total of Dollars Overpaid
Proportion of Dollars in Sample Weighted
Overpayment Error Rate
Amount Paid for Universe of Paid Services (Medical Bills) Weighted Lower 95% Confidence Interval Weighted Upper 95% Confidence Interval Margin of Error on Proportion of Sample Dollars Overpaid Hospital 32 $13,452 4 $757.02 20.08% 7.80% $2,161,994 7.35% 8.25% 0.45% Medical 534 $36,224 115 $6,969.90 54.07% 18.95% $4,136,580 18.54% 19.35% 0.40% Pharmacy 166 $17,320 18 $4,464.05 25.85% 6.19% $1,470,531 5.83% 6.55% 0.36% Totals 732 $66,996 137 $9,190.97 100.00% 13.72% $7,769,105 13.46% 13.98% 0.265
Exhibit 2: State Employees Workers' Compensation Study Computation of the Payment Accuracy Rate performed by the State Auditor's Office
Step Computation Description 1 Total dollars in the sample were calculated in each sample stratum. 2 Total dollars in the universe were calculated in each sample stratum. 3 Proportion of dollars in the sample were calculated in each stratum by dividing the number of dollars in that stratum by the total dollars sampled. 4 Proportion of dollars in the universe were calculated in each stratum by dividing the number of dollars in that stratum by the total dollars in the universe. 5 Each sample stratum was weighted by the ratio of the proportion of dollars in the universe divided by the proportion of dollars in the sample. 6 The total dollars overpaid in each sample stratum was calculated. 7 An error rate for each stratum was calculated by dividing the number of dollars overpaid by the number of dollars in the Sample Stratum. 8 The above error rate was then multiplied by the weight calculated in step 5 to yield the point estimate for the error rate in the sample stratum. 9 A margin of error at 95% confidence was calculated for the above weighted error rate by taking the square root of [(1.96^2) x (weighted error rate) x (1 - weighted error rate)/dollars sampled] for each sample stratum. 10 The above margin of error was multiplied by [1-(sample dollars in stratum/universe dollars in stratum)] to adjust for the finite Universe. 11 The above adjusted and weighted margin of error was added to and subtracted from the point estimate error rate to render a 95% confidence Interval for each stratum. 12 The dollars in the universe in each sample stratum was multiplied by the lower confidence limit, point estimate, and upper confidence limit of the 95% confidence interval to extrapolate the weighted error rate to the universe of dollars. 13 A similar confidence interval was constructed for total dollars in the sample, the endpoints and midpoint of which were multiplied by the total dollars in the universe to render the overall extrapolation to the universe.