STATE OF WEST VIRGINIA
FULL PERFORMANCE EVALUATION
OF THE

DEPARTMENT OF HEALTH AND HUMAN RESOURCES
MEDICAID


Goods and Services Purchased for Deceased Medicaid Recipients


OFFICE OF LEGISLATIVE AUDITOR
Performance Evaluation and Research Division
Building 1, Room 314W
State Capitol Complex
CHARLESTON, WEST VIRGINIA 25305
(304) 347-4890




Medicaid was created in 1965 as Title XIX of the Social Security Act. It is a federal/state program, administered by states and funded from federal, state and, in some states, local revenues. Federal funds are made available contingent on a state match that varies among states from year to year. West Virginia has the second most favorable match rate in the nation, at an approximate 74 percent federal and 26 percent state match. The match rate is determined by a formula that takes into account the State's per capita income compared to the national average. After rising steadily between 1980 and 1992, the match rate began to decline in 1993. The decrease in the match rate alone required State funding increases of $11 million in State Fiscal Year (SFY) 1995 and $16 million in SFY 1996. Today, one in five West Virginians receive health care benefits through the Medicaid program, a $1.4 billion program, with an approximate $322 million state match, serving approximately 431,000 citizens.

Health care services for the aged and disabled account for nearly two-thirds of all Medicaid expenditures. Medicaid and Medicare together have an enormous economic impact in West Virginia. Almost 70 percent of nursing home revenue is attributable to Medicaid and more than 60 percent of all hospital revenue in West Virginia is attributable to these two programs.

Nationally, the Medicaid program consumed nearly 20 percent of State expenditures in 1995, up from an average of 10 percent in 1987. In West Virginia, Medicaid consumed 15.2 percent of the State budget for Fiscal Year 1995. Medicaid is essentially three programs in one: first, it is a health insurance program for low income parents and children; secondly, it is a long-term care program for the elderly; and lastly, it is a funding source for services to people with disabilities. Medicaid is one of the largest expenditures within the Department of Health and Human Resources.

Programs within the Department are categorized by similarity of service and function into four bureaus; two of these bureaus involve Medicaid.

Bureau for Children and Families. This Bureau manages public assistance programs, social service programs, and child support enforcement and collections. A citizen might have contact with this Bureau's programs in such matters as eligibility determination for Temporary Assistance to Needy Families, Food Stamps and Medicaid, subsidized work programs, child and adult abuse concerns, and child day care, foster care and adoptions.

Bureau for Medical Services. This Bureau is the single state agency charged with administering the State's Medicaid program, which provides medical coverage for eligible clients. The Bureau provides administration and reimbursement for medical services to eligible individuals, such as inpatient and outpatient hospital care, physician services, laboratory, x-ray, behavioral health services, prescription drugs, nursing home care and several in-home services which keep individuals out of institutional services.

As related to the single issue addressed in this report, a stratified random sample was taken from 1995 death records maintained by the Office of Vital Statistics of the Department of Health and Human Resources. The sample was stratified by place of death, including the following strata: hospital, en route to hospital, clinic or center, licensed nursing home, unlicensed personal care facilities and home (see APPENDIX A). Social Security numbers, names, places and dates of death were provided by the Office of Vital Statistics. In these six places, a total of 18,250 West Virginians died (in West Virginia) in calendar year 1995. Total deaths and sample size are represented in TABLE 1 on page 8. Social Security numbers from the random sample were cross checked against Social Security numbers in Medicaid Management Information System (MMIS) Archives by the Bureau for Medical Services and Consultec. By matching Social Security numbers from Vital Statistics to those in the Medicaid billing system, a consumption history for each deceased Medicaid beneficiary was obtained. A computer program then identified charges in which the date of service followed the date of death.

Decedent names were included from both data sources and used to reconcile the identity of all individuals for which hits were obtained. There were some instances in which the individual's name, as provided by Vital Statistics, did not match the individual's name provided by the Bureau for Medical Services. Each of these cases of uncertain identity was fully investigated to determine whether or not the Medicaid consumption was by the deceased individual sampled from Vital Statistics. This often involved communication with county officials and obtaining copies of death certificates, obituaries and information from Medicaid eligibility files. Cases of mistaken identity were purged from the database.

The execution of the cross match by Social Security numbers leads to a situation where there is some likelihood that postmortem consumption could go undetected because one or both Social Security numbers could be incorrect. It has been reported that the eligibility information system, RAPIDS, has had some difficulty managing Social Security numbers, and that for a time case workers entered nine zeros into the data field. Any case in which Medicaid paid for such bogus claims but the recipient was recorded as having 000-00-0000 as his/her Social Security number would be invisible to this study. Fourteen of the randomly sampled recipients had no Social Security number in the Office of Vital Statistics' database. Cases lacking Social Security numbers could not be cross matched with consumption data and were therefore excluded from statistical calculations.

After removing all cases of mistaken identity, the cross match identified 117 postmortem transactions. Because the Medicaid Management Information System retains only two years of on-line claims history, the Department was initially unable to provide documentation concerning the reasons for date of service to fall after date of death on a case-for-case basis. Because of the high number of transactions that occurred within thirty or fewer days after the date of death, a sub-sample of transactions was drawn from the thirty or less sub-population, to be targeted for more aggressive investigation (usually this involved contacting providers for further documentation). Within the sub-sample, several transactions were found to be legitimate and were removed from the analysis. These findings were then projected to the transactions in the sub-population that were not part of the sub-sample. For these transactions dollar amounts were decreased as indicated by the sub-sample, and transactions were counted as fractional transactions to derive the expected aggregate effect. Some information from the sample pertaining to the Department's function and management controls, remain a part of this evaluation, though the postmortem consumption was proven to be legitimate and was not included in other parts of the analysis.

All Durable Medical Equipment and Laboratory charges occurring thirty days or less after the date of death were deemed to be legitimate because of standard policy governing contract duration and billing procedure. Five pharmacy charges were also deemed to be legitimate because the Department was able to provide documentation from the provider indicating that the pharmaceutical was dispensed on or before the date of death, even though the reported service date fell after. Two cases were determined to be legitimate because the charge was for a drug taken from the emergency supply box in the hospital. The recipient's consumption required the hospital to replace its emergency supply, which occurred after the date of death. Several Pharmacy Providers were unable to provide documentation pertaining to the legitimacy of claims in the sub-sample. Medicaid Regulations for Pharmacy Services state:

To meet federal requirements of utilization review and quality control, the provider of medical supplies and durable medical equipment must keep a file on each Medicaid recipient for which the Department of Health and Human Resources is billed for medical supplies and/or durable medical equipment. This file must contain documentation of the medical necessity of the supplies and/or durable medical equipment provided to the patient including the prescription signed and dated by the physician.
Therefore, pharmacy claims in the sub-sample in which documentation was not provided are considered to be invalid claims. The Legislative Auditor considers documentation necessary to make a valid determination of pharmacy claims.

Durable Medical Equipment is viewed somewhat differently because the equipment is provided on a thirty day contractual basis, with the last day in the contract period representing the "date of service". Durable Medical Equipment Claims with service dates occurring less than thirty days after the date of death were considered to be valid, by virtue of this policy. Pharmacy claims, though often billed for on a monthly basis, are not contractual rental agreements.

Fourteen of the randomly sampled recipients had no Social Security number in the Office of Vital Statistics' database. Cases lacking Social Security numbers could not be cross matched with consumption data and were therefore excluded from statistical calculations. See Table 1 and Appendix A.


TABLE 1

1995 Deaths and Sample Size
PLACE OF DEATHTOTAL DEATHSSAMPLE SIZE
Hospital10,219370
En Route to Hospital1,005281
Clinic or Center87
Licensed Nursing Home2,397334
Unlicensed Personal Care144104
Home4,477352
TOTALS18,2501,448
Source: Legislative Auditor's Study on Postmortem Medicaid Consumption, 1997

ISSUE AREA: A Projected $63,275 of $1.3 billion service dollars Was Paid to Providers for Invoices in which a Recipient's Death Preceded the Invoiced Service Date.

This performance evaluation found that approximately $63,275 of greater than $1.273 billion service dollars was paid for Medicaid services for deceased recipients. The Legislative Auditor's sample of cases found that 7.8 percent of Medicaid recipients who died in 1995, or approximately 315 deceased persons, were provided goods and services after death. These 1995 deaths predated managed care. The implications of this finding could be very costly in the future, given the greater proportion of health care costs that will be incurred prospectively rather than retrospectively.

The postmortem payments identified in this study occurred because of the absence of a simple management control involving cross matching information from the DHHR's Office of Vital Statistics with Medicaid records. Federal regulations require the DHHR to recover funds improperly paid resulting from overpayments, false claims, and/or misrepresentation or concealment of facts related to a provider's qualifications or his/her costs as filed with the Bureau for Medical Services.

Incidents of Medicaid and Medicare fraud have been widely reported. National findings indicate that the long term care and home health industries are especially susceptible to fraud and abuse. Given the 297 percent increase in Medicaid costs that occurred between 1988 and 1995, and long-term care expenditures totaling $371 million in 1994, the need for management controls to prevent and detect fraud cannot be understated. This study was designed to detect this specific type of fraud and abuse across all types of providers, especially long-term care and home health, and to test the Department's existing management controls.

A stratified random sample was taken from 1995 death records maintained by the Office of Vital Statistics of the Department of Health and Human Resources and cross matched to Medicaid consumption records (See Background, Objective, Scope and Methodology for details). In the six places of death examined in the study, a total of 18,250 West Virginians died (in West Virginia) in calendar year 1995. By matching Social Security numbers from Vital Statistics to those in the Medicaid billing system, a consumption history for each deceased Medicaid beneficiary was obtained. A computer program then identified paid claims in which the date of service followed the date of death.

The recipient, in Medicaid terms, is an individual who has been declared eligible to receive State Medicaid Program Benefits, based upon certain eligibility criteria, one of which is to live in West Virginia. Therefore, in this report, "recipient" shall be used to refer to the Medicaid Recipient number, assigned to a now deceased individual, whose Medicaid benefit consumption record became part of this evaluation due to the individual's selection in a random sample taken from the West Virginia Department of Health and Human Resources Office of Vital Statistics.

RESULTS

Of the 1,448 sampled, 33 recipients were found to have consumed benefits after date of death, at a total cost of $3,970. An analysis of findings by place of death is found in TABLE 2.

TABLE 2
Place of Death by Provider Type, Number of Transactions and Cost*
PLACE OF DEATH (recipients) TRANSACTIONSTOTAL
HOSPITAL (5 recipients)131,414.72
Prenatal Case Management (RFS)373.02
Case Management Agencies163.90
Outpatient Hospital1580.30
General Practice298.69
Pharmacy18598.81
EN ROUTE to Hospital (4 recipients)4111.19
Case Management Agencies149.70
Mental Health Clinic17.46
Pharmacy254.03
LIC. NURSING HOME (20 recipients)40899.85
General Practice146.54
Radiology18.75
Pharmacy38844.56
UNLIC. PERSONAL CARE (3 recipients)9203.24
Laboratory117.00
Pharmacy8186.24
HOME (2 recipients)21,341.71
Homemaker Agencies & Com. Care21,341.71
TOTAL (33 recipients)793,970.71
* The numbers of recipients and transactions have been rounded for presentation. The totals were computed using mean values from a sub-sample and therefore may not be equal to the sum of rounded values.
The best perspective on how the money was spent can be gained from an examination of the provider types that invoiced for these goods and services. TABLE 3 summarizes all charges by provider type. Of approximately 132 provider types recognized by the Bureau for Medical Services, all postmortem consumption identified by the sample occurred within 9 provider types (for analytical purposes, the provider types "pharmacy" and "hospital-based pharmacy" have been combined in TABLE 3).


TABLE 3
Charges by Provider Type*
PROVIDER TYPE TRANSACTIONSCOST% of TOTAL COST
Case Management Agencies2$113.602.8
Homemaker Agencies & Community Care6$1341.7133.7
Right From the Start3$73.02 1.8
Mental Health Clinic1$7.46.18
Outpatient Hospital1$580.3014.6
Radiology1$8.75.22
Laboratory 1$17.00 .42
General Practitioners3$145.23 3.65
Pharmacy67$1,683.6442.4
TOTAL79$3,970.71100
Source: Legislative Auditor's Study, 1997.
* The numbers of recipients and transactions have been rounded for presentation. The totals were computed using mean values from a sub-sample and therefore may not be equal to the sum of rounded values.

Homemaker Agencies and Community Care account for 2 out of 79 transactions (see Background, Objective, Scope, and Methodology for explanation of fractioned transactions). Two recipients were charged for services from two providers at a cost to the State of 1/3 of the total cost. According to the Bureau for Medical Services, "billing for this type of service is generally done on a monthly cycle." Because goods and services are provided in monthly increments, the service date could be any given day of the month. If a recipient dies in the middle of a billing cycle, costs will be incurred for that cycle. For these three transactions, all service dates did occur within one month of the date of death.

Pharmacy charges comprised $1,683.64 or 42.4 percent of all postmortem charges. However, unlike homemaker services and community care charges, pharmacy charges were spread over many more transactions. A total of 66.84 pharmacy transactions were identified. The largest pharmacy charge was $193.64. Through examination of charges appearing in the sub-sample, the Department of Health and Human Resources identified five pharmacy charges which documentation proved to be legitimate charges.

An employee of the Department of Health and Human Resources stated that "In the case of pharmacies doing the billing, they generally will have a recipient on file and do not need to see a card every month." However, State Medicaid Regulations for pharmacy services clearly state:

It is important to examine the medical identification card to determine that the patient is eligible for the current month of service. Carefully check the recipient identification number and provider number for accuracy.

According to the Secretary of the Department of Health and Human Resources, this provider type has the highest potential for provider and recipient fraud.

Of the 79 transactions indicated in the study, two charges exceeded $500. One charge was for $896.40 by a homemaker services provider. The other charge was for $580.30 from an outpatient hospital. Another large charge, $450, was also billed for by a homemaker service provider.

An analysis of the time between date of death and date of service is provided in TABLE 4. For the sample, 94 percent of all costs were incurred 30 days or less after the date of death, eight percent were incurred between 31 and 90 days after the date of death, leaving 6 percent that occurred between 91 and 365 days.


TABLE 4
Summary of Postmortem Service Time Periods
Days After Date of DeathTransactionsCost Percentage of Total
Between 1 and 30 603415.1286%
Between 31 and 90 10325.398%
Between 91 and 3659230.206%
TOTAL79$3970.71100%
Source: Legislative Auditor's Study on postmortem Medicaid Consumption, 1997

No transactions occurred more than a year after date of death. However, at least one recipient remained eligible 21 months after date of death.

Each transaction, with the exception of third-party insurance claims, has a Procedure Code. The West Virginia Medicaid Program uses the Health Care Financing Administration Common Procedure Coding System, including: American Medical Association Common Procedure Terminology codes, alpha-numeric codes, and local codes. Procedure Codes indicate the service or good for which the provider is billing. It allows the third-party provider reimbursement contractor and State program staff to determine that the charge is for a covered service. By definition, charges for any goods or services represented as having been provided after death are suspect. Items involving complex procedures occurring weeks after a patient's death are even more highly suspect. Other procedures, such as surgical pathology, may have a legitimate explanation, especially when it occurs only three days after the date of death.

In one case a recipient remained eligible to consume benefits, according to state information systems, well after case closure would have been appropriate. The recipient, who died in November 1995, remained eligible to consume benefits until August 1997. The Bureau for Medical Services reported that the RAPIDS case had been closed in early August. A computer printout from the RAPIDS eligibility system indicated that the case was closed as of August 15, 1997, with code number 054, indicating that the client refused continued assistance. However, MMIS, the information system used for provider reimbursement, showed client eligibility until August 31, 1997. The correct date of death was reported to the Social Security Administration. The next month's check was returned to the Social Security Administration and no subsequent checks were issued to the recipient. The Social Security Administration was notified in a timely manner of the recipient's eligibility change, yet failed to notify the state Medicaid program of the recipient's eligibility change. Social Security related Medicaid cases (such as this recipient's case) are to be turned on and off only at the instruction of the Social Security Administration, except in cases of death or when a client moves out of state. The local Medicaid office may close a case when the client is obviously ineligible for benefits.

The Department of Health and Human Resources receives information about Social Security recipients from the Social Security Administration through the State Data Exchange (SDX) tape. The SDX contains information about the amount of benefits, beginning date of eligibility, ending date and changes which occur. Using this data, the Need to Open and the Need to Evaluate listings are produced. Cases meeting certain selection criteria are printed to hard copy. The Need to Evaluate Printout is used to identify recipients who are no longer eligible for SSI so that caseworkers may evaluate the individuals for other types of Medicaid eligibility and close appropriate SSI Medicaid cases. Without accurate information from the Social Security Administration, states are unable to make reliable eligibility verifications. The caseworker must then rely on information provided by family members, obituaries and the possibility that the recipient's Medicaid card will be returned to the county office. In the case of the recipient who remained eligible for almost two years past the date of death, the Social Security Administration failed to notify the West Virginia Department of Health and Human Resources of the recipient's change in eligibility. Regular exchange of death statistics between the Office of Vital Statistics and DHHR would keep this kind of situation from becoming an expensive problem.


PRIMARY CAUSE: Failure to Use Death Information from DHHR Office of Vital Statistics

The only reasonably foolproof management control for preventing payments to providers for goods and services charged to deceased recipients would be one in which death records from the Office of Vital Statistics, also within DHHR, would be interfaced or regularly cross matched with Medicaid eligibility and consumption records. Such information processing would allow Department of Health and Human Resources to either deny claims for reimbursement or charge back providers for such payments which have already been made. Department of Health and Human Resources currently has no such control.

The Secretary of the Department of Health and Human Resources has general supervision over the System of Vital Statistics, which is under the immediate supervision of the State Registrar of Vital Statistics, pursuant to Chapter 16, Article 5, of the West Virginia State Code, as amended. The State Registrar prepares and publishes annual reports of vital statistics and supervises the activities of local registrars, whose duties include the transmittal, on the first and fifteenth day of each month, of certificates and reports to the State Registrar. A death certificate for each death which occurs in the State is filed with the local registrar of the district of which the death occurred, within three days after the death. Given that the Office of Vital Statistics is within the Department of Health and Human Resources, the Department has the ability to access the necessary information to ensure that State Medicaid monies are not paid for goods and services rendered after the date of death.

SECONDARY CAUSES

These findings can undoubtedly be attributed to many other causes. At the request of the Legislative Auditor's Office, the Department of Health and Human Resources identified several ways by which some postmortem charges occurred and provided available information. However, the Department was unable to provide the Legislative Auditor with information concerning the specific causes for many postmortem transactions. It is therefore not possible to ascertain whether some charges are (1) legitimate (i.e.,pharmaceuticals dispensed while a patient was living but with a cyclical service date falling after death); (2) illegitimate (i.e., provider or recipient fraud, billing errors, etc.); or, (3) partially legitimate/partially illegitimate (i.e., services billed on a monthly cycle which should have been prorated for the month of death).

Among the possibilities for secondary causes are provider and recipient fraud. Because of the potential for criminal activity, few providers were contacted. Families of recipients were not contacted to avoid jeopardizing any future fraud investigations. Because of this and the Department's lack of information concerning the validity of specific transactions, and especially providers inability to present documentation required by federal regulations, discussions of causal elements are necessarily speculative in nature.


SECONDARY CAUSE 1: Possible Provider Fraud

Medicaid regulations specifically state that "the medical services provider is completely and solely responsible for the content of his/her claim", and that, "...the provider carefully check the medical identification card each time a service is rendered in order to make certain the patient is currently eligible on the date service is rendered, including those services which have been prior authorized". Activities which constitute fraudulent practices or abuse of the program, by providers, which are pertinent to possible causes in this study are:

A. Billing for services, supplies or equipment which were not rendered to or used for Medicaid patients.
B. Misrepresentations of dates and descriptions of services rendered or of the identity of the recipient or the individual who rendered the services.
C. Charging the Medicaid Program, by subterfuge, costs not incurred or which were attributable to non-program activities, other enterprises, or personal expenses.

In order to be eligible to provide services in the State Medicaid program, a provider must be licensed. The Bureau for Medical Services does not conduct automatic reviews with licensing boards to verify licensing of providers, because no state or federal regulation requires that such a review be performed. Instead, the Bureau relies upon notification from State licensing boards of actions they have taken with regard to licensing and/or sanctioning of a provider. There are four staff persons at the Bureau for Medical Services who are assigned to provider enrollment functions. Their responsibilities include: enrolling new providers; purging the system of outdated provider information; and updating information in provider files.

SECONDARY CAUSE 2: Process Errors

Medical billing is a complicated process. Federal regulations mandate that the Medicaid agency must require providers to submit claims no later than 12 months after the date of service; however, there is no regulation that specifically states that the date of service must fall on or before the date of death. Nor is there any regulation stating that death is grounds for an immediate change in eligibility.

Though the Bureau for Medical Services provides training seminars each year, the reliability of provider generated data is still subject to mistakes. It is the provider's responsibility to ensure that the information on the invoice is correct. The Bureau for Medical Services provides training seminars, manuals, including supplemental information to providers and a help line for questions concerning billing procedures to providers. They answer between 200 and 300 provider inquiries a day. High turnover rates among Provider clerical staff may be a causal factor to a portion of the cases in the study. In one case on which the Department was able to provide documentation, the deceased recipient's number differed by only one digit to that of the still living spouse. The pharmacy charge appearing on the deceased's claims history was actually intended for the spouse.

Medicaid is automatically billed for some goods and services, such as durable medical equipment. Medicaid regulations state that Medicaid claims are processed through an automated Electronic Data Processing (EDP) system at the Office of Administration and Claims Processing, Bureau for Medical Services. This office processes 15 million claims annually through Consultec. Consultec is the contractor to which the Medical Services has out-sourced the responsibility for claims processing. Paper and electronic claims go directly to Consultec which has separate Post Office boxes for the different types of claims. Electronic claims arrive via modem or on diskette. The tenth digit of the Claim Reference Number represents the "input media," or the form the claim is filed in. Providers may submit electronic claims directly, or through one of several firms that offer electronic submission services.

Claims are sent to Consultec's Atlanta processing site. There, claims pass through three system edits which include: (1) Is the recipient eligible; (2) Is the service covered; (3) Is the provider eligible to provide the service. After claims pass through the system edits, the claim goes on status. There are three status types: (1) Paid; (2) Pending; (3) Denied. Once a claim goes on status, the system reports back to the Bureau for Medical Services and then the State Central Data Processing prepares the checks. Technicians at Consultec are available to provide assistance to the provider's office for electronic claims submission.

One case taken in the sub-sample was found to be a duplicate claim. The payment was recovered in an adjustment process but the claims history was not adjusted because it was done as a mass recovery, not as a single void. This occurred because the claims processing staff at Consultec mistakenly processed a tape twice, causing many providers to receive duplicate payments.

SECONDARY CAUSE 3: Possible Recipient Fraud

Recipient Fraud is another possible cause. Someone close to the now-deceased individual may have taken over the identity of the deceased for the purpose of receiving health care. Prescription drug charges are especially suspect and vulnerable, due to the casual regard of policy and the demand for certain drugs having black market value.

SECONDARY CAUSE 4: Billing Cycles and Time Lags

Of the nine provider types indicated in this study, pharmacies and homemaker service providers bill on a monthly cycle. Medicaid regulations do not allow payment for the same prescription twice in a monthly billing cycle. Homemaker Services are often billed on a monthly cycle, since services occur regularly and frequently.

There are two areas in which time lags are a factor. The time which elapses between the determination of a recipient's eligibility status and the time the information is communicated to the Medicaid Claims Processing System is a possible cause for error. Eligibility begin and end dates are passed from the C and M219 systems to Consultec in Atlanta. There, they are downloaded to the Medicaid Management Information System (MMIS) every night. A pertinent System Requirements of MMIS is to Receive appropriate Medicaid recipient eligibility data from all sources of eligibility determination. A basic function and objective of the MMIS claims processing system is to Ensure that all recipients for whom input is submitted were eligible for the type of service at the time the service was rendered. The system is also required to enter the date of death for a recipient as indicated in the social services or SSI file after an official notice of death has been received. (The SDX Need to Evaluate Tape) Another code which must be entered into a recipients data file indicates a recipient's destination upon discharge from a medical institution. "Died" is one of the codes available to system operators.

The time between a recipient's death and the communication of that information to the DHHR county office, which is responsible for initiating eligibility transactions, is another possible cause for error. Since eligibility cards are issued every month, the earliest eligibility end date the Bureau for Medical Services could have for a recipient would be the last day of the month in which the recipient died. Because eligibility cut-off for issuing cards for the following month is four to five working days from the last day of the month, recipients who die during the last week in the month do not receive an eligibility end date until the second month following the month in which they died.
Providers are notified of a recipient's eligibility through two mechanisms. The recipient's Medicaid card, issued monthly, is to be presented by the recipient to the provider at the time of service. The second mechanism is the Medicaid Management Information System (MMIS) Voice Response System, which is available to providers. The information available through the Voice Response System reflects the time lag between the eligibility determination and the transfer of that information to the Medicaid Claims Processing System. Therefore, days immediately after a recipients death are a gray area for eligibility determination. The system is not stopping charges which pass through with service dates after the date of death. Granted some of these charges may be legitimate, though the accounting is sloppy. Others are not legitimate. These time lapses in accurate information provide a window for fraud, error and abuse to occur.

EFFECTS

It is projected that the State of West Virginia's Medicaid Program paid $63,275 for medical goods and services provided to 315 recipients after their 1995 deaths (see TABLE 5). Attempts to identify the number of Medicaid recipients that died in 1995 were unsuccessful. However by estimating Medicaid deaths by dividing the number of Medicaid recipients by the State population, then multiplying by the number of West Virginia Deaths, yields an approximate error rate of 7.8%. This means that unexplained postmortem charges were incurred for approximately 7.8% of the Medicaid recipients that died in 1995.

As discussed in the Results section under TABLE 5, most of the charges occurred within 30 days of date of death. Transactions for which service dates occurred more than 30 days after the date of death, represent 14 % of total charges.

TABLE 5
Projected Costs by Place of Death
PLACE OF DEATHTOTAL
EXPENDITURES
SAMPLE
AVERAGE
TOTAL FOR
POPULATION
Hospital1,414.723.8239,073.04
En route to Hospital 111.19 0.39397.67
Clinic or Center0.000.000.00
Licensed Nursing Home899.852.696,457.91
Unlicensed Personal Care 203.241.95281.41
Home1,341.71 3.817,064.87
TOTAL3,970.71 -----------63,274.90
Source: Legislative Auditor's Study on postmortem Medicaid Consumption, 1997
Medicaid regulations state: "The Bureau for Medical Services is responsible for establishing and maintaining communication with providers participating in the program," and that "appropriate staff is [ to be ] available to respond to inquiries regarding program policy or reimbursement, receive and discuss suggestions for program coverage, assist with filing claims, or other problems." The Bureau is also responsible for recovery of State and Federal Medicaid funds improperly paid resulting from overpayments, false claims, and/or misrepresentation or concealment of facts related to a provider's qualifications or his/her costs as filed with the Bureau. The Bureau's procedure for auditing providers may include the use of sampling and extrapolation. If sampling reveals a consistent pattern of overcharging or other fiscal abuse of the medical assistance program by the provider, the provider will be required to reimburse the Bureau for the projected total overpayment. The Office of Surveillance and Review is responsible for detecting and examining any unusual patterns of payments and unnecessary or inappropriate utilization of care and services covered under the Medicaid Program. When inappropriate practices are identified, the Bureau may pursue one or more of the following:

1. Requirement of a satisfactory written plan of correction;
2. Limited participation in the plan;
3. Disenrollment, or exclusion from participation.


IMPLICATIONS FOR THE MANAGED CARE ENVIRONMENT

These 1995 deaths predate managed care. The implications of this finding could be very costly in the future, given the greater proportion of health care costs that will be incurred prior to treatment. At a recent Welfare Reform Summit, a representative of one of the participating managed care providers estimated that the average premium cost, per member, per month, was $100. In a case identified in this study, an individual's eligibility status was maintained for 21 months after her death. If she had been enrolled in a managed care plan, the State would have paid an estimated $2,100 in premiums to an HMO to insure just one deceased recipient.

To prevent payments to providers for goods and services charged to deceased recipients, DHHR should interface or regularly cross match death records from the Office of Vital Statistics with Medicaid eligibility and consumption records. Because death statistics are collected and entered by the DHHR's Office of Vital Statistics as a matter of law, maintaining this control should be as inexpensive as sharing data extracts or interfacing the two systems.


CONCLUSION

A $63,275 finding in an evaluation of the State Medicaid Program, which had almost $1.273 billion in expenditures in FY 95, seems trivial in comparison. However, postmortem consumption for 7.8 % of the recipients that died in 1995 reveals weaknesses in the Division's system of controls. Eligibility verification and determination of recipients and providers is essential to maintaining the fiscal integrity of the program. If fraudulent providers are identified by this study, the extent of their improprieties may go beyond this type of fraud. Therefore, the Legislative Auditor's Performance Evaluation and Research Division will turn this study over to the Office of Inspector General Investigations and Fraud Management Section, with the recommendation that their findings be reported back to the Legislative Auditor.

RECOMMENDATION 1

The Department of Health and Human Resources should perform annual (or more frequent) reviews of postmortem consumption by cross matching death statistics from the Office of Vital Statistics with Medicaid consumption/eligibility data as it has been done in this study. Because the process is highly automated, the Department should perform the cross match for all deaths, rather than a sample. The Bureau can then focus its efforts on larger, more suspicious charges.


RECOMMENDATION 2


The Department of Health and Human Resources should direct the Office of Inspector General, Investigations and Fraud Management Section to investigate significant postmortem charges identified by this study and recover any State and Federal funds spent inappropriately.