« 이전계속 »
One apparent reason for the high turnover of Medicare Part B claims examiners
has been the fact that they have been paid at a lower rate than have claims rideles examiners in other parts of the carirer's operations. In May 1975, the carrier
approved raising the pay of Medicare Part B claims examiners to that of other f the dire claims examiners in the organization, but this raise has not yet been put into
One result of the high turnover rate is an excessively high error rate being experienced by Blue Shield of Florida in its processing of Medicare Part B claims.
As the direction of SSA, Medicare Part B carriers each week perform an "end.
of-line review" of a sample of claims processed that week to identify errors which able contre remain uncorrected at completion of initial processing. The results of the end-of
line review provide an indication of the quality of work being done during the inforzi "routine manual processing" portion of claims processing.
The end-of-line review was initiated on January 1, 1974. Results are expressed Ise tenis as the ratio between the number of errors detected and the number of line items
examined. For the period February 1974 (first reporting month) through Decemshed ber 1974, the error ratio reported by Blue Shield of Florida ranged from a low of
20 in February to a high of .42 in July. In other words, for every 100 line items processed during July, 42 errors were made and remained undetected throughout
the claims processing cycle. rable
The most recent report showed an error rate of .33 for the month of April, effice during which the average claim reviewed contained about 1.5 errors. Blue Shield
of Florida's error rate was among the highest of all the carriers.
This error rate affects the program in three ways. First, errors cause significant processing delays. Second, errors which slip through the processing cycle undetected may result in underpayments to claimants who must write in to request
a review of their claim and who will experience another long delay before resolure ata tion of their complaint and ultimate payment of amounts to which they are
entitled. Finally, they may result in overpayments which in all probability will remain undetected.
Mr. Ahart informed the subcommittee that the Blue Shield Corp. did not have an unduly complicated system of claim processing:
l'pon receipt in the mallroom, claims are sorted and batched according to the nature of the claim—such as routine claims by physicians or other suppliers of health services (called assigned claims), routine claims by beneficiaries (called unassigned claims), claims for the cost of purchase or rental of durable medical equipment, or claims submitted on behalf of deceased beneficiaries. At the present time there are 18 categories being used to group claims into batches.
After the claims are sorted and batched, each claim is stamped with a control
number which includes the year, Julian date, batch number and the number of led in the claim within that batch.
Next, certain information from each claim is entered into the computer and compared with information relating to the particular beneficiary which is already In the computer's address file.
The claims are then delivered to claims examiners who have responsibility for the particular type of claims included in each batch. About 80 percent of the claims go to the routine claims examining section with the remainder going to the special claims examining section.
Information from the simpler claims reviewed is entered directly into the computer by the claims examiners. For the more complicated claims, the examiners prepare work sheets from which information is entered into the computer,
If the claims examiner finds that all information necessary to process claim has not been submitted, he notes on the claim that additional information is required. During processing by the computer, each claim is subjected to five "screens." Fallure to pass any one of the screens will cause an error suspense sheet to be generated and the claim to be placed in suspense until the question is satisfactorily resolved. The first screen, which is really two screens in one, tests for "edit errors" and "reasonable charges.” In addition, any claims preivously identified by the claims
examiners as being in need of additional information will kick-out during the B
Brst screen and be held in suspense until the additional information is obtained either through telephone calls or correspondence with the beneficiary or provider. Other computer screens test the claim for correctness of basic data (claims failing
57-399 0 - 75 - 2
to pass this series are called transaction rejects), possible duplicate payments and cases of suspected over-utilization (prepayment screens). We will discuss these screens in greater detail as we go on.
When these computer screens have been successfully passed, SSA records in Baltimore must be queried, for certain claims, to ascertain the eligibility of the beneficiary and/or the status of his deductible. This is needed because beneficiaries might be filing claims with two or more carriers, information concerning charges incurred and applied against the $60 annual deductible must be accumulated at a central location.
After the claim has been fully developed, all computer screens have been satisfied, and the necessary information obtained from Baltimore, an Explanation of Medicare Benefits (EOMB) and a payment check (if appropriate) are generated to be mailed to the claimant.
After the initial rejection or payment of his claim, a beneficiary dissatisfied with the determination may request that Blue Shield of Florida review the claim. If the beneficiary is still dissatisfied and the amount in dispute is $100 or more he may request a hearing.
Findings-Reported processing time:
Reports prepared by Blue Shield of Florida show that in calendar year 1974 the carrier processed 3,858,535 Medicare Part B claims; and that 80 percent of these were processed in 30 days or less; 14 percent in 31 days to 60 days; three percent in 61 to 90 days, and three percent in more than 90 days. Thus, although serious delays occurred in the processing of a small percentage of total claims 94 percent were processed within 60 days—the number of claims encountering long delays involves thousands of people-236,613 claims took over 60 days to process, and 103,400 of these took over 90 days.
Fifty-nine percent of the claims were unassigned—that is, the claim was submitted by the Medicare beneficiary. Thirty-five percent were assigned claims submitted by providers, and six percent were for the services of hospital-based physicians. Using information in monthly reports prepared by the carrier, we computed the average processing time for each of these types of claims.
Our analysis showed that claims for hospital-based physicians were processed in an average of 12.5 days, more quickly than were the other two types. The average processing time fluctuated from a low of nine days during August when 19,814 claims were processed to a high of 16 days during February when 20,403 claims were processed.
Assigned claims took a little longer, averaging 17.8 days. Again there was a fluctuation in the average processing time from month to month ranging from a low of 11 days during August when 115,760 assigned claims were processed to a high of 26 days in December when 118,695 claims were processed.
Later in his testimony, Mr. Ahart commented on the delays within the carrier's processing system:
We noted, however, that claims normally are in the carrier's office for some period before control numbers are assigned and that some period elapses between preparation of checks and EOMB forms and their mailing to claimants. Based on our tests and information available at the carrier, we estimate that a total of about seven (7) days elapse at these two stages.
Further testimony follows:
We also found that two types of claims processed through the payment cycle distorted the computation of processing time to some degree. These are referred to as "set-ups" and "deletions."
A set-up occurs when a single claim is divided into two or more claims for processing purposes. The following situations necessitate set-ups:
The claim includes services performed in more than one calendar year; The claim includes charges incurred by both husband and wife; The claim contains more than 32 line items, or The claim contains both routine items and complicated procedures such as multiple surgery.
Set-ups are not prepared until the original claim reaches a claims examiner. They are then sent back through the front-end control procedures and routed to
the proper claims examiner for processing. The distortion occurs because a new
As in the case of set-ups, these claims will have been on hand for some period 1080:
before they are routed to the correct claims examining area.
The carrier could provide us with records showing number of deletions only
for a la month period from December 23, 1974, to February 8, 1975. During that started period, there were 10,094 deletions. It should be noted, however, that the volume Priet?
of deletions may not have been as high before October 1974, when current procedures for sorting and batching claims were put into effect.
Our verification of the processing time reported by Blue Shield of Florida was based upon analysis of a random sample of 1,961 Medicare Part B claims processed during calendar year 1974 which was extracted for us by the carrier's Electronic Data Processing Department.
MEDICARE HEARINGS AND FINDINGS
The GAO team was completely satisfied with the process used for extracting samples to study.
A detailed analysis of the 1,961 claims revealed the following: 377 required over 30 days to process. A further analysis of one half of those--189-showed that:
3 of the cases did not present sufficient data to make a determination as to the disposition of the cases;
3 other cases were delayed for an extended period by the Social Ben Security Administration;
183 remaining cases required a total of 9,925 days to process an average of about 54 days. 76 percent of the 183 cases were processed in 31 to 90 days; An additional 16 percent were processed in 61 to 90 days. Actual time and location of claims were as follows: 41 percent in routine manual processing which includes receiving, sorting, stamping, microfilming, and examining and entering the claim into the computer for further processing;
16 percent in “additional development" which involves obtaining from the beneficiary or the provider additional information required to properly process the claim;
28 percent in "edit kickouts” resulting from computer screens of such items as reasonable charges, possible duplicate charges, and the accuracy of the data entered; and
15 percent in “queries" of SSA central files to determine the benefciary's eligibility and the status of the deductible. Additional work in selected areas where delays occur: Since the procedures followed during "routine manual processing” have been altered recently as a result of the installation of direct data entry equipment, GAO did no additional review work in that area. Also
, since queries to the central files are necessary and required, and delays in this area generally are outside the control of Blue Shield of Florida, we did no additional work in that area.
Reports prepared by BHI show that in calendar year 1974 about 19.7 percent of all Medicare Part B claims handled by Blue Shield or Florida required additional information before they could be processed. This is almost double the national average of 10.2 percent.
SSA guidelines prohibit a carrier from returning claims to the claimants, in most instances, requiring instead that the carrier develop the needed information. The carrier attempts to develop the required information by telephone whenever feasible. Otherwise, it is requested by mail.
Within our sample of 183 claims which took over 30 days to process, 69 claims, or 38 percent, required additional information-obtaining the information took an average of about 23 days.
For these 69 claims, an average of nearly 15 days elapsed between the date the claims examiner requested the additional information and the date the telephone call was made or a request letter was sent. Based on observations made during May 1975, we believe that a similar delay is still being experienced at this point in the claims processing cycle.
To gain a better understanding of the type of information being requested when claims are forwarded to the additional development section, we sampled an additional 250 claims which were being processed during our review and which needed additional information. Our analysis of the 250 claims and the 69 claims discussed above showed that the additional information requested in about 60 percent of these 319 cases was either an itemized breakdown of the services performed and the charges for the services, or a statement of the diagnosis of the medical condition which necessitated the service.
Further testimony by Mr. Ahart also revealed that:
Fifty-seven percent of all edit error kick-outs are attributed to errors made by claim examiners or the person who processes the claim before it went to the examiners;
GAO believed that a substantial reduction in edit error kick-out could be achieved by a stable work force;
A preliminary finding that the lack of adequate management attention at all levels, is the most significant factor in causing the time delay problem;
An appearance that management has not acted to alleviate back-logs until they have reached crisis proportion.
Under questioning by the subcommittee Chairman, Senator Lawton Chiles, Mr. Ahart stated in response to a query concerning cost on case processing, thusly:
Our source is a statistical report comparing the average cost as well as the productivity of different carriers which BHI pulls together
Senator CHILES. So it is because they are producing fewer cases per man-hour that the average cost per case is higher ?
Mr. AHART. I think that we can relate this to the fact their personnel turnover is so high, 77 percent in the claims examining function and with each new examiner coming in, it will be less efficient than the typical operation.
Senator CHILES. Would you concur, then, judging from the error rate in the processing time for claims that Blue Shield of Florida because of its inefficiency is the reason for its costs being 30 percent higher than average?
Mr. AHART. Yes, it would be a matter of inefficiency in its operations.
Senator CHILES. The 10 largest Medicare B carriers in Florida is eight from
Mr. Ahart. There has been a lot of growth and the seasonal fluctuation would
Senator CHILES. Did you determine anything about the high turn-over rate, this 77 percent; do you think that was primarily because of wages paid or were there other factors ?
Mr. Ahart affirmed that the low salary paid to workers is probably an inhibiting factor in worker performance. Further, the General Accounting Office indicated that a full review would be forthcoming which would present a more detailed study of the situation in Florida.
Government Participation and Supervision/Bureau of liealth Insurance.
Mr. Thomas M. Tierney gave testimony for the BHI supervision on the Florida situation.
Mr. Tierney stated the overall situation thusly: Let me start out by saying, Mr. Chairman, from what you have already heard and I know from your own interest and background in the past Medicare is a very complicated program. I think it has been a tremendously successful program in many ways, but it was probably the first massive effort to enter into whole new areas of health care and the mechanisms of paying for health care that had ever been undertaken in this country, and certainly in the beginning there were a lot of headaches. Things are improving and things I would dare say even in Florida are improving. Senator CHILES. No, sir, don't dare say that. Don't dare say that because I don't believe you can dare say that in Florida yet.
Mr. TIERNEY. I would only add I think things in Florida, and I think Mr. Ahart
Let me make it clear, Mr. Chairman, we have been probably the greatest critics
I think there are one or two things you would want to know about their performance and costs and particular problems as we have viewed them.
You raised a question for example about the hospital claim. I think the one element left out of the GAO response to that was this assignment situation. Assigned claims are very much easier as the figures indicate, the processed claims. They are prepared by doctors' secretaries and they are well done with few errors and they go through fairly well. Those prepared by hospital bookkeeping departments are perhaps the best and they go through very well. An unassigned claim is usually one prepared by himself or herself, usually it contains a large number of errors so there is a correlation between the overall figure of claims processing and the assignment ratio.
Mr. Tierney stated that the Government paid private carriers on a prenegotiated rate and that audit power over the carriers is exercised by both GAO and the Department of Health, Education, and Welfare.