The Mystery of the Missing Collections
How well do you really know your payors? A recent analysis for one practice turned up some very interesting revelations with operational ramifications for the provider, but this was possible only because the provider had access to the entire fee schedule of the insurance company.
The larger insurance companies use variations of the RBRVS, applying their own conversion factors. For this reason, radiology providers would be wise to secure a complete schedule of their contracted fees. Only then can providers manage the complex relationships between their imaging-center costs and the revenues that they must receive to cover them.
Prudent owners should know their fixed and variable costs by modality—information useful in determining whether the fees proposed by an insurance company adequately compensate them. They also have to consider the evolving dynamics of the current marketplace, in which insurance companies promise a certain fee level and then shift increasing amounts of the obligation to the patient, in the form of copayments and deductibles.
Reason for Evaluation
A radiology practice, owner of an imaging center, did not find the payment trends for a major payor to be acceptable. It had performed some testing (on a limited sample of exams) that had provided few answers, and it was unwilling to spend a large amount on a detailed audit. The use of database-extraction techniques gave the practice the answers that it sought.
The project involved the extraction of billing-system files accounting for 2,466 exams and containing 28 fields of data. This was performed in mid-February 2011. Two additional fields were added to facilitate the evaluation. The first was the negotiated fee schedule of the primary payor being evaluated, provided in an electronic spreadsheet, in order by CPT® code.
The second added field contained one of 11 numeric flags assigned to exam families for which technical-component payment is discounted if two or more exams in the same family are performed during the same patient visit. This discounting policy was based on a CMS regulation, but other major payors also instituted it. The discounting did not apply when the multiple exams performed during a single visit were in different families. This regulation was changed (effective January 1, 2011) to merge the 11 families into one, making providers unable to avoid discounting for multiple exams, even when the exams are from different families.
Account Resolution
In Table 1, six hypothetical patients underwent the same exam; all were apparently covered by the same insurer, with different benefit programs. The radiology charge was $100; the agreed-upon fee was $49.50. Therefore, the radiologist was obligated to write off $50.50 as a contract adjustment. Patient A had a high-end plan (slowly disappearing), in which the insurer paid the entire amount due. Patient B had a 20% copayment, but also had a secondary insurance plan that covered this. Patient C was obligated to pay the 20% and called the billing office to pay the amount by credit card; the practice paid the credit-card fee. 
Table 1. Mathematics of Account Resolution Patient D, who had an annual deductible that had not been met, owed the entire bill. Only part of the fee was received, and the balance was eventually sent to a collection agency. Patient E was referred to the office by a physician who did not comply with the insurer’s rule for preauthorization. The patient was not asked to return at a later date (the front office knew that it needed an authorization number). The amount due was lost as a compliance credit. Patient F formerly had insurance coverage, but was uninsured at the time of service. He had been to the office before, and the front desk did not verify coverage. The claim was rejected; the patient made a partial payment, and the rest was written off as a courtesy. Four Universe Subsets A perfect universe is one containing only two credits: cash and contract adjustment. The marketplace is moving in a different direction, however, where other types of noncash credits apply: courtesy, credit-card fees, and adjustments for bad debt and compliance failures. This unweighted sample implies perfect resolution at 49.5% (which is not that far from reality, in this case), but other market factors drag down the final number. A review of the imaging center’s accounts receivable revealed that the 2,466 exams could be organized into 22 different payment scenarios (Table 2). These universes were organized into four groups. The first group contained single exams and multiple exams, performed on the same day, that were not in the 11 families where discounting might have applied in 2010.
Table 2. Summary of Four Universe Subsets The second group consisted of alphanumeric supply codes (such as those for contrast and disposables). No payor fee existed for these codes because they were variably priced. A fee schedule was reverse engineered by organizing the records by CPT code and looking for consistent payment patterns. The third group included single exams, by date of service, in one of the 11 discounting families, and the fourth group was composed of multiple exams in one of those 11 families that had been performed on the same day. The four universe subsets contained 22 universes. Universe 1: This was the perfect universe; resolved at the payor-fee amount, it accounted for 50.36% of the total volume as of mid-February 2011. Universes 2 through 7: These include zero-balance records where cash (plus the credit-card fee, bad debt, courtesy, or balance) equals the payor fee. The only other noncash credit is the contract adjustment. Universes 2 through 5 would have resolved at the fee-schedule amount, if not for the practice’s failure to collect on self-pay balances. Universe 6 was made up only of records with cash and contract-adjustment credits where balances were due from patients. If these had been fully collected, they would have been in universe 1. Universe 7 was unique to the supply codes in that the payor fee exactly matched the practice charge. Universes 8 through 10: Universes 8 and 9 held exams where the patient was responsible for the entire payor fee because of an annual deductible. In spite of numerous statements being sent to the patients, no cash was collected, and these amounts were written off as either bad debt or courtesy. Universe 10 was similar to universe 6, except that no cash had been collected as of mid-February 2011. These records would have been shifted into any of the prior 9 universes once their balances reached zero. Universes 11 through 16: Universe 11 was a useful byproduct of acquiring the payor’s fee schedule. It only contained records where the payor fee was higher than the practice charge. Naturally, the payor reimbursement was at the practice-charge level; there was no contract adjustment, and the balances were zero. Universe 12 held all zero balances, resulting only from cash and contract-adjustment credits. None of the receipts matched the payor fee schedule; they were either higher or lower. These could have been exams for patients who had different coverage than that provided by the payor for which records were evaluated, but information on that other coverage was not acquired at the time of the visit. Universe 13 contained records where the cash was less than the payor fee, the balance was zero, and there were multiple noncash credits, including contract adjustments. Universe 14 consisted of exams performed for patients who were classified as covered by the payor of interest; nonetheless, the charge was written off as bad debt. Universe 15 held exams with no cash and zero balances resulting from combinations of compliance credits, bad debt, and courtesy. Universe 16 consisted of 13 records that met the criteria for potential discounting because of the 11-family rule. Universes 17 through 22: Universe 17 represented the only records where no response had been received on the original claims at the time of evaluation. Universe 18 held records where cash had been credited and noncash credits had not been posted. Universe 19 differed from universe 11 because there were balances. Universes 20 and 21 showed cash and contract adjustments inconsistent with the payor fee and either debit or credit balances. The last two records (universe 22) pertained to supply codes where a payor fee could not be determined. Accounts-receivable Results Table 2 shows exam volume for all modalities combined, but without including dollars, it is only half the picture. Table 3 illustrates the financial implications of Table 2. It is organized as a conventional reconciliation, except for the inclusion of the payor-fee information. This is the amount that the practice would have received if: no copayment or deductible requirements had existed, no discounting rules on multiple exams had been applied, and no compliance with precertification and other rules had been required. Net cash consists of the actual receipts posted to these accounts by mid-February, 2011—about $90,000 less than the total payor fees. The remaining balances on all nonzero records were $44,819. With nearly $45,000 in receivables unaccounted for, it was necessary to make a speculative attempt at bringing the balance to zero using the statistical techniques illustrated in Table 4. The balance-resolution row shows hypothetical amounts of cash and credits with a total equal to the remaining balance. The residual cash estimate is based on a belief that residual balances mostly end up as charity or bad-debt write-offs, accounting for approximately 25% of the average resolution rate. The assumption has also been made that the contract adjustments had already been applied to the remaining procedures, meaning that the balance would be attributed to the other credits (using the proportions achieved through mid-February). A lower resolution rate was assumed for the unaccounted shortfall, based on the knowledge that this was the hard-to-collect balance. The bottom three rows illustrate the dollar impact of the hypothetical resolution of all records to a zero balance, measured against the perfect universe (the payor-fee level). The perfect universe would only have two credits: cash and contract adjustment. The combined cash and contract-adjustment shortfall was shifted to the other four credit categories in the form of losses attributed to the current reimbursement environment. The whole purpose of the hypothetical universe was to examine how much the practice lost because of the introduction of multiple-procedure discounting and compliance rules, in addition to the shifting of responsibility for promised fee-schedule payments to patients (in the form of higher deductibles). At this point in the investigation, it became clear that the practice’s suspicion that a payor was responsible for the missing accounts receivable was unfounded—and that patients were the probable culprits. Reconciliation by Month of Service To illustrate the implications of needing to collect increasingly large deductibles from patients, the database extraction for Table 5 was organized by month of service to show how the 2010 population was managed. The information on each row is interrelated; the cash and noncash credits pertain to the charges generated in that month of service. The accounts column lists the patient accounts that were either created or updated in each month of service. The billing system maintains a permanent account number for each patient; this is why the number of accounts in the monthly history changes according to whether the data cover patients who had only a single exam or those who had either multiple exams during one visit or single exams during more than one visit in 2010. The lower of the two numbers is the more accurate. Modeling the database in this fashion helps to explain payor dynamics that affect collections. the last three columns are resolution ratios that illustrate this. Since the imaging center maintained a constant payor relationship for all of 2010, the degree of variation in the resolution rates seems puzzling. The resolution percentage was based only on cash and noncash credits; charges were not considered. Payor fee divided by charge is a more relevant measurement; it assumes perfect resolution at the negotiated-fee level. Note how uniform its ratios are for all months except July. January and March have low resolution rates. The reasons are easily understood because of the extent of compliance and bad-debt losses attributed to cases in those month of service. Compliance write-offs are random, but the large bad-debt credits in the early months of the year are triggered because patients must meet annual deductibles, making them responsible for the entire payor fee. The bad-debt and courtesy trends could reflect both timing issues and policy decisions. The absence of bad debt in the last three months was probably due to the timing of the data extraction, since fewer patients will have deductibles to meet that late in the year. The courtesy adjustments seemed to increase to a new level suddenly in August. The receivables balances for the last four months of the year dropped more from bad debt and courtesy credits than from cash. The success rate of the practice’s collection effort was expressed in the success-rate column, which was the achieved resolution percentage rate divided by the payor-fee resolution percentage rate. The most noteworthy month was July, measuring 98.85% of the maximum attainable amount. There was an insignificant balance remaining. The main reason for this is case mix; note the maximum achievable level (42.44%). In other words, the mix of exams in July was so favorable that actual collections almost matched the perfect universe. Perhaps there was a low incidence of multiple-procedure discounts or a higher rate of better-paying procedures. Analysis Findings Slightly more than 50% of the exams were resolved exactly at the payor-fee level, at rates approximately twice those of the Medicare schedule for this region. Exactly 31% of the exams could have been resolved at the payor-fee level, but revenue was lost—either from failure to collect self-pay balances or due to amounts that remained to be collected at the time of data extraction. This percentage could have declined to 20% thereafter, as more outstanding balances were paid. The bad-debt and courtesy losses attributed to just this one payor are significant enough to warrant exploration, by the imaging center, of making copayments mandatory at the time of service. Any subsequent need to refund overpayments would still be a better alternative than the loss of copayments that are never collected. The remaining 19% of accounts failed to resolve at the payor-fee level and will not do so without corrective action (assuming that some action is appropriate). Within this number is a population of exams that might not belong in this payor universe, suggesting that the imaging center’s staff either did not acquire information about changes to an existing patient’s account or that a new patient provided incorrect information. The billing system purposely keeps the original payor as the primary one, even if it subsequently acquires different information. The value per exam at the payor-fee level is $323; this amount, however, is only relevant to this provider and this payor, and is not applicable elsewhere. Actual cash per exam has reached $286. The cash value of the remaining receivables ($45,000) contains a large proportion of bad debt and courtesy, and it might only increase the final number to $288. This implies a shortfall of $86,000. The average number of procedure codes per account (considering only those with multiple transactions) was 3.84. Supply (disposables) codes are included in this average. While this inquiry was triggered by the practice’s belief that one payor was delinquent, it turned out that many payors were in default—a considerable number of patients. As insurance companies and employers shift a greater share of health-care costs to the patient, self-pay balances are a growing problem for radiology groups. Table 3. Accounts-receivable Results for Four Universes (click here to view table) Table 4. Hypothetical Final Resolution Versus Perfect Universe (click here to view table) Table 5. Receivables Reconciliation (by Month of Service) (click here to view table)James A. Kieffer, MBA, is president of Proforma Financial Group, Inc, Nashua, New Hampshire; James.Kieffer@comcast.net.
Table 1. Mathematics of Account Resolution Patient D, who had an annual deductible that had not been met, owed the entire bill. Only part of the fee was received, and the balance was eventually sent to a collection agency. Patient E was referred to the office by a physician who did not comply with the insurer’s rule for preauthorization. The patient was not asked to return at a later date (the front office knew that it needed an authorization number). The amount due was lost as a compliance credit. Patient F formerly had insurance coverage, but was uninsured at the time of service. He had been to the office before, and the front desk did not verify coverage. The claim was rejected; the patient made a partial payment, and the rest was written off as a courtesy. Four Universe Subsets A perfect universe is one containing only two credits: cash and contract adjustment. The marketplace is moving in a different direction, however, where other types of noncash credits apply: courtesy, credit-card fees, and adjustments for bad debt and compliance failures. This unweighted sample implies perfect resolution at 49.5% (which is not that far from reality, in this case), but other market factors drag down the final number. A review of the imaging center’s accounts receivable revealed that the 2,466 exams could be organized into 22 different payment scenarios (Table 2). These universes were organized into four groups. The first group contained single exams and multiple exams, performed on the same day, that were not in the 11 families where discounting might have applied in 2010.
Table 2. Summary of Four Universe Subsets The second group consisted of alphanumeric supply codes (such as those for contrast and disposables). No payor fee existed for these codes because they were variably priced. A fee schedule was reverse engineered by organizing the records by CPT code and looking for consistent payment patterns. The third group included single exams, by date of service, in one of the 11 discounting families, and the fourth group was composed of multiple exams in one of those 11 families that had been performed on the same day. The four universe subsets contained 22 universes. Universe 1: This was the perfect universe; resolved at the payor-fee amount, it accounted for 50.36% of the total volume as of mid-February 2011. Universes 2 through 7: These include zero-balance records where cash (plus the credit-card fee, bad debt, courtesy, or balance) equals the payor fee. The only other noncash credit is the contract adjustment. Universes 2 through 5 would have resolved at the fee-schedule amount, if not for the practice’s failure to collect on self-pay balances. Universe 6 was made up only of records with cash and contract-adjustment credits where balances were due from patients. If these had been fully collected, they would have been in universe 1. Universe 7 was unique to the supply codes in that the payor fee exactly matched the practice charge. Universes 8 through 10: Universes 8 and 9 held exams where the patient was responsible for the entire payor fee because of an annual deductible. In spite of numerous statements being sent to the patients, no cash was collected, and these amounts were written off as either bad debt or courtesy. Universe 10 was similar to universe 6, except that no cash had been collected as of mid-February 2011. These records would have been shifted into any of the prior 9 universes once their balances reached zero. Universes 11 through 16: Universe 11 was a useful byproduct of acquiring the payor’s fee schedule. It only contained records where the payor fee was higher than the practice charge. Naturally, the payor reimbursement was at the practice-charge level; there was no contract adjustment, and the balances were zero. Universe 12 held all zero balances, resulting only from cash and contract-adjustment credits. None of the receipts matched the payor fee schedule; they were either higher or lower. These could have been exams for patients who had different coverage than that provided by the payor for which records were evaluated, but information on that other coverage was not acquired at the time of the visit. Universe 13 contained records where the cash was less than the payor fee, the balance was zero, and there were multiple noncash credits, including contract adjustments. Universe 14 consisted of exams performed for patients who were classified as covered by the payor of interest; nonetheless, the charge was written off as bad debt. Universe 15 held exams with no cash and zero balances resulting from combinations of compliance credits, bad debt, and courtesy. Universe 16 consisted of 13 records that met the criteria for potential discounting because of the 11-family rule. Universes 17 through 22: Universe 17 represented the only records where no response had been received on the original claims at the time of evaluation. Universe 18 held records where cash had been credited and noncash credits had not been posted. Universe 19 differed from universe 11 because there were balances. Universes 20 and 21 showed cash and contract adjustments inconsistent with the payor fee and either debit or credit balances. The last two records (universe 22) pertained to supply codes where a payor fee could not be determined. Accounts-receivable Results Table 2 shows exam volume for all modalities combined, but without including dollars, it is only half the picture. Table 3 illustrates the financial implications of Table 2. It is organized as a conventional reconciliation, except for the inclusion of the payor-fee information. This is the amount that the practice would have received if: no copayment or deductible requirements had existed, no discounting rules on multiple exams had been applied, and no compliance with precertification and other rules had been required. Net cash consists of the actual receipts posted to these accounts by mid-February, 2011—about $90,000 less than the total payor fees. The remaining balances on all nonzero records were $44,819. With nearly $45,000 in receivables unaccounted for, it was necessary to make a speculative attempt at bringing the balance to zero using the statistical techniques illustrated in Table 4. The balance-resolution row shows hypothetical amounts of cash and credits with a total equal to the remaining balance. The residual cash estimate is based on a belief that residual balances mostly end up as charity or bad-debt write-offs, accounting for approximately 25% of the average resolution rate. The assumption has also been made that the contract adjustments had already been applied to the remaining procedures, meaning that the balance would be attributed to the other credits (using the proportions achieved through mid-February). A lower resolution rate was assumed for the unaccounted shortfall, based on the knowledge that this was the hard-to-collect balance. The bottom three rows illustrate the dollar impact of the hypothetical resolution of all records to a zero balance, measured against the perfect universe (the payor-fee level). The perfect universe would only have two credits: cash and contract adjustment. The combined cash and contract-adjustment shortfall was shifted to the other four credit categories in the form of losses attributed to the current reimbursement environment. The whole purpose of the hypothetical universe was to examine how much the practice lost because of the introduction of multiple-procedure discounting and compliance rules, in addition to the shifting of responsibility for promised fee-schedule payments to patients (in the form of higher deductibles). At this point in the investigation, it became clear that the practice’s suspicion that a payor was responsible for the missing accounts receivable was unfounded—and that patients were the probable culprits. Reconciliation by Month of Service To illustrate the implications of needing to collect increasingly large deductibles from patients, the database extraction for Table 5 was organized by month of service to show how the 2010 population was managed. The information on each row is interrelated; the cash and noncash credits pertain to the charges generated in that month of service. The accounts column lists the patient accounts that were either created or updated in each month of service. The billing system maintains a permanent account number for each patient; this is why the number of accounts in the monthly history changes according to whether the data cover patients who had only a single exam or those who had either multiple exams during one visit or single exams during more than one visit in 2010. The lower of the two numbers is the more accurate. Modeling the database in this fashion helps to explain payor dynamics that affect collections. the last three columns are resolution ratios that illustrate this. Since the imaging center maintained a constant payor relationship for all of 2010, the degree of variation in the resolution rates seems puzzling. The resolution percentage was based only on cash and noncash credits; charges were not considered. Payor fee divided by charge is a more relevant measurement; it assumes perfect resolution at the negotiated-fee level. Note how uniform its ratios are for all months except July. January and March have low resolution rates. The reasons are easily understood because of the extent of compliance and bad-debt losses attributed to cases in those month of service. Compliance write-offs are random, but the large bad-debt credits in the early months of the year are triggered because patients must meet annual deductibles, making them responsible for the entire payor fee. The bad-debt and courtesy trends could reflect both timing issues and policy decisions. The absence of bad debt in the last three months was probably due to the timing of the data extraction, since fewer patients will have deductibles to meet that late in the year. The courtesy adjustments seemed to increase to a new level suddenly in August. The receivables balances for the last four months of the year dropped more from bad debt and courtesy credits than from cash. The success rate of the practice’s collection effort was expressed in the success-rate column, which was the achieved resolution percentage rate divided by the payor-fee resolution percentage rate. The most noteworthy month was July, measuring 98.85% of the maximum attainable amount. There was an insignificant balance remaining. The main reason for this is case mix; note the maximum achievable level (42.44%). In other words, the mix of exams in July was so favorable that actual collections almost matched the perfect universe. Perhaps there was a low incidence of multiple-procedure discounts or a higher rate of better-paying procedures. Analysis Findings Slightly more than 50% of the exams were resolved exactly at the payor-fee level, at rates approximately twice those of the Medicare schedule for this region. Exactly 31% of the exams could have been resolved at the payor-fee level, but revenue was lost—either from failure to collect self-pay balances or due to amounts that remained to be collected at the time of data extraction. This percentage could have declined to 20% thereafter, as more outstanding balances were paid. The bad-debt and courtesy losses attributed to just this one payor are significant enough to warrant exploration, by the imaging center, of making copayments mandatory at the time of service. Any subsequent need to refund overpayments would still be a better alternative than the loss of copayments that are never collected. The remaining 19% of accounts failed to resolve at the payor-fee level and will not do so without corrective action (assuming that some action is appropriate). Within this number is a population of exams that might not belong in this payor universe, suggesting that the imaging center’s staff either did not acquire information about changes to an existing patient’s account or that a new patient provided incorrect information. The billing system purposely keeps the original payor as the primary one, even if it subsequently acquires different information. The value per exam at the payor-fee level is $323; this amount, however, is only relevant to this provider and this payor, and is not applicable elsewhere. Actual cash per exam has reached $286. The cash value of the remaining receivables ($45,000) contains a large proportion of bad debt and courtesy, and it might only increase the final number to $288. This implies a shortfall of $86,000. The average number of procedure codes per account (considering only those with multiple transactions) was 3.84. Supply (disposables) codes are included in this average. While this inquiry was triggered by the practice’s belief that one payor was delinquent, it turned out that many payors were in default—a considerable number of patients. As insurance companies and employers shift a greater share of health-care costs to the patient, self-pay balances are a growing problem for radiology groups. Table 3. Accounts-receivable Results for Four Universes (click here to view table) Table 4. Hypothetical Final Resolution Versus Perfect Universe (click here to view table) Table 5. Receivables Reconciliation (by Month of Service) (click here to view table)James A. Kieffer, MBA, is president of Proforma Financial Group, Inc, Nashua, New Hampshire; James.Kieffer@comcast.net.
