Sources of Error in Delayed Payment of Physician Claims

Transcription

1 Vol. 35, No Practice Management Sources of Error in Delayed Payment of Physician Claims Jessica M. Lundeen; Wiley W. Souba, MD, ScD, MBA; Christopher S. Hollenbeak, PhD Background and Objectives: Our objectives were to determine the distribution of errors and estimate the magnitude of the burden of delayed payments in a large physician group practice. Methods: A 25% random sample (n=775) was taken from all billed records of a physician group practice in the Pacific Northwest that were delayed 6 months or more as of June 30, The source and specific reasons for payment delays, as well as the amount of each unpaid invoice, were determined by electronic documentation or telephone calls to the payor. Analysis of variance was used to determine whether the amount of the invoice was associated with the source and reason of error. Results: The source of delayed payments due to provider, payor, patient, and technical error were 36.1%, 28.1%, 14.5%, and 21.3%, respectively. The most-frequent reasons for delayed payment were that the provider incorrectly set up the account (15.2%), the provider did not follow up on denial (12.9%), and the payor incorrectly processed the invoice (11.6%). Analysis of variance suggested that the invoice amount was not significantly associated with the source but was significantly different across reasons for delayed payment. The potential financial impact of earlier recovery of payment was $262,270. Conclusions: In these data, provider and payor errors accounted for almost two thirds of delayed payments. The most promising avenue for providers to reduce delayed payments is by reducing their own errors. Eliminating the two most common errors would result in a more timely recovery of nearly $70,000 in revenues. (Fam Med 2003;35(5):355-9.) Late payment of medical claims by insurers is recognized as a growing problem for health care providers. 1-4 A considerable portion of medical practice business office expenses is devoted to resolving and collecting late payments, and many states have enacted legislation that regulates acceptable payment schedules for medical insurance claims. 3 Contractually, payors are obligated to pay clean claims in a timely manner; clean claims are defined as claims that have all required information contained in the initial billing on the HCFA-1500 form. 3 The medical practice obligation includes submitting a timely claim to the payor and providing additional information necessary for the payor to process the claim. A common payment standard for clean claims is 14 days for electronic claims and 30 days for paper claims. 5 In spite of legislation and contracts, however, payments are often delayed. Delays may result from errors made by the payor, provider, or patient. Examples of these errors include (1) payors may incorrectly reject the claim, (2) providers may submit a claim with incorrect or inaccurate insurance information, causing the From the Departments of Surgery and Health Evaluation Sciences, Pennsylvania State College of Medicine, Hershey, Pa. payor to be unable to identify the patient, and (3) patients may provide inaccurate information to the provider. Payments may also be delayed due to technical errors, such as when the provider s record indicates the claim has been sent, but the payor has no record of receiving the claim, or when the payor sends an explanation of denial, but the provider has no record of receiving it. Expenditures on efforts to recover delayed payments could be allocated more appropriately, and some errors may be prevented altogether if patterns are identified in the source of delayed medical claims. This study (1) determined the distribution of delayed payments attributable to payor, provider, patient, and technical error, (2) determined whether the amount of expected payment is associated with the cause of delayed payment, and (3) estimated the potential financial effect of delayed payments on a large primary care medical practice. Methods Setting Data were obtained from a group of 63 physicians in the Pacific Northwest, including 43 physicians who went through a family practice residency, 13 physicians

2 356 May 2003 Family Medicine who went through an internal medicine residency, five physicians who went through a pediatric residency, and two nurse practitioners. The practice includes a family practice residency program. At this practice, the physicians code their own bills, and two coding and reimbursement specialists assist them in optimizing coding. In addition, the billing office employs approximately 60 full-time employees for data entry, insurance follow-up, and claims collection. Approximately 5.2% of claims are denied on the initial submission,.3% of claims cannot be transmitted to payors due to data entry errors, and approximately 7% of outstanding claims are more than 180 days old. Claims Reviewed We extracted all billing records of claims with a date of service prior to January 1, 2001, and that were older than 180 days (n=3,124). We then took a 25% sample of these claims (n=781), the largest random sample that our resource constraints would allow, using a computerized random number generating process. Six invoices from the sample for obstetrical services performed by family physicians were excluded; all six invoices were extreme outliers, at least 30 standard deviations above the mean. Further, like many physician groups across the country, this group practice more aggressively collects on larger invoices. As a result, these six invoices had been resubmitted several times and had numerous reasons for being denied. The remaining invoices (n=775) were reviewed in the patient s computer record to verify the age of the claim, verify that it was billed to a payor, record the amount of the invoice, and record the reason why the claim was still outstanding, if evident. If the reason was not evident, phone calls were made to the payor to determine why the claim had not been paid. Capitated managed care claims were excluded since this practice did not collect individual claims from these payors but instead received a lump sum payment per patient per month. Data Analysis The source of the delay and the specific error that resulted in the delay were recorded for each claim. Errors were attributed to four different sources: providers, payors, patients, and technical errors. Provider errors were those errors made by the physician or the administrative support staff. For example, the billing staff setting up the insurance information incorrectly would qualify as a provider error. Payor errors were delays or denials that resulted from a mistake by the payor. Examples include payments sent to the wrong provider. Patient errors were errors made by the patient, in most cases by providing incorrect information to the provider. Finally, there were some cases in which the provider submitted a clean claim, but the payor had no record of receiving it. Other claims were received by the payor, and a denial or payment was sent, but the provider had no record of receiving it. Such errors with ambiguous culpability were considered technical errors. The frequency of errors was tabulated, and analysis of variance (ANOVA) was used to test whether the source of error and the specific error that led to the delay were associated with the invoice amount. The financial effect of recovery of delayed payment for the practice was extrapolated by summing across error categories and multiplying by four (because we only sampled 25% of late claims). Results Table 1 reports percentages of the total error attributable to each source, the mean charge per invoice for each source of error, and the projected sum for each source of error if the claim had been paid in a timely fashion. As shown in Table 1, 36.1% of delayed payments can be attributed to provider error, 28.1% to payor error, 21.3% to technical error, and the remaining 14.5% to patient error. The most costly errors occurred with technical errors, which had a mean charge of $91.20 per invoice, followed by payor errors with $83.76, patient error with $83.70, and provider error with $ ANOVA suggested that differences in the average invoice amount were not significantly different across the sources of delayed payments (P=.48). Table 2 displays the specific errors made by payors, providers, and patients, including the number, percentage, and mean invoice amount for each error. The mostfrequent provider error occurred when the provider set up the patient s insurance record incorrectly by entering an incorrect identification number or group number (42.1%). Without this information, the payor cannot identify the patient, and the claim is denied. The second most-frequent provider error occurred when a denial was received, but the provider did not follow up (35.7%). For example, the payor may request medical records from the provider to pay the claim. Under most contracts signed by the provider and payor, Table 1 Description of Sources of Error for Delayed Payment of Physician Claims Error Standard Projected Source # % Mean Deviation Sum Payor error $83.76 $65.91 $73,041 Provider error $81.73 $58.43 $91,536 Patient error $83.70 $62.94 $37,499 Technical error $91.20 $63.87 $60,195 Total $84.60 $62.40 $262,270

3 Practice Management Vol. 35, No the provider is obligated to obtain this information, and the payor is not obligated to pay the claim until this information is received, so claim payment is delayed until the requested information is received. All other provider reasons for delayed payments accounted for only 7.9% of total outstanding invoices. These errors included charges entered incorrectly, a patient encounter without a referral for service, or failure to bill the primary or secondary payor for the services. The largest payor error occurred when the payor withheld payment (also known as pending ) on the claim for no identifiable reason or processed the claim incorrectly (41.3%). Almost all of the patient errors occurred when the patient provided inaccurate insurance information to the provider (88.4%). From the provider s perspective, the insurance information was entered correctly and billed to the insurance company, and the error is only realized after receiving a denial from the payor. Technical errors made up 21.3% of the total errors. Most of these errors occurred when the provider s records indicated that a claim has been sent, but the payor had no record of the claim (71.5%). The rest occurred when the payor records showed that they sent an explanation of benefits to the provider, but the provider did not receive it. Across all specific errors, there were significant differences in average invoice amounts (P=.026). The specific errors that carried the highest average invoice amount were providers mailing the bill to the wrong address ($188.50) and improper handling of denials ($159), but there were relatively few of these errors. Post hoc comparisons revealed that the average invoice for these two reasons were not significantly different (P>.05). Further, average invoice amounts for improper handling of denial, incorrect information on payor system, and incorrect charge entry were not significantly different (P<.05) but were significantly larger than average invoice amounts for all other reasons. Total Lost Income Extrapolating from our 25% sample suggests that recovery of all delayed payments would result in projected payments of $262, Assuming that none of these claims would be recovered without additional action by the provider, this amount represents the upper bound on the amount the physician group could Table 2 Specific Errors That Led to Delayed Payment of Physician Claims Stratified by Source of Error % of % of Average Total Extrapolated Specific Error # Error Source Total Errors Claim Claims Claims Provider Error Set up insurance incorrectly $85.10 $10,046 $40,185 No follow-up on denial received $74.00 $7,405 $29,618 Incorrect charge entry $ $2,135 $8,539 Kept as pending in error $55.40 $942 $3,766 No referral and incorrect follow-up or processing $84.90 $764 $3,057 Denial received but not posted $71.30 $357 $1,426 Failed to bill primary insurance $58.00 $232 $928 Improper handling of denial $ $477 $1,908 Mailed to incorrect address $ $377 $1,508 Failed to bill to secondary insurance $75.00 $150 $600 Payor Error Processed incorrectly or pending for no reason $76.00 $6,837 $27,348 Rejected in error $67.10 $2,148 $8,592 Problem with pricing network $95.20 $2,857 $11,427 Paid to wrong provider $91.20 $2,553 $10,210 Incorrect information on payor system $ $2,426 $9,705 Did not forward claim to correct payor $82.10 $1,068 $4,270 Problem with participating providers $93.00 $372 $1,488 Patient Error Patient gave inaccurate information $83.00 $8,216 $32,863 Worker s compensation form not complete or denied $89.20 $1,159 $4,636 Technical Error Payor sent denial; provider never received $89.40 $4,204 $16,816 Provider sent bill; payor never received $91.90 $10,845 $43,379

4 358 May 2003 Family Medicine rationally expend in recovering and avoiding delayed payments. Extrapolating from our 25% sample, we computed the relative financial importance of each type of error. The most expensive errors were (1) patient errors in which the patient provides the incorrect insurance information; these carry a projected total of $48,308 or $767 per physician per year and (2) technical errors in which the provider shows the claim was mailed, but the payor states it wasn t received; these errors cost $43,377 or $689 per physician per year. Discussion Delayed payment of claims is a growing concern among health care providers. In our study of delayed claims at a primary care practice in the Pacific Northwest, we found the most-common source of delayed payments to be errors made by providers, which accounted for approximately 36% of delayed payments. Errors made by payors accounted for approximately 28% of delayed payments, technical errors not directly attributable to any party accounted for about 21% of delays, and patient errors accounted for about 14% of delays. The average claim amount was not significantly different for each of these sources of error (P=.48). However, the average claim amount for specific errors was significantly different (P=.026), with the provider mailing the claim to the wrong address and providers posting the invoice incorrectly being the errors associated with the highest average claim amount. If all errors that led to delayed payments could be eliminated, our results suggest that a theoretical total of $262,270, or $6,099 per physician per year, could be recovered by the practice. Admittedly, not all of this amount would be recovered because of contractual differences between charged amounts and reimbursed amounts. The nature of the distribution of errors, too, makes it unlikely that all of the potential reimbursements could be recovered. Further, some errors may not be eliminated. For example, technical errors presumably arose from procedural problems, computer errors, and mail delivery errors. Other than automating the billing process so that outstanding invoices are billed again after a certain delay, there is little that a provider could do to remedy these errors. Likewise, the ability of providers to reduce errors arising from patients who provide inaccurate information is limited. One possible remedy is an on-line system of eligibility that would allow the billing staff to access patient account information on-line. This would reduce the need to rely on the patient s memory and potentially outdated insurance card information. Such systems are becoming available but are not presently in widespread use. Errors on the part of the payor are similarly problematic. Economic theory would suggest that payors have an incentive to delay payment until the marginal cost of doing so equals the additional interest that accrues during the delay. The most-frequent payor error was pending (withholding payment for no cause), providing some suggestion that payors are responding to the aforementioned incentive, and little can be done by the provider to change this practice besides pursuing a legal remedy. As mentioned previously, some states are using legislation to require prompt payment of clean claims. These legislative actions can be expected to be effective to the extent that the marginal cost of penalties for delayed payments outweigh the marginal, in terms of gains from recovered claims, revenue to the payor. The most promising avenue for providers to reduce delayed payments is by reducing their own errors. Most provider errors occurred when the provider set up the patient account incorrectly (42%) or failed to follow up on denials from the payor (36%). Eliminating these two errors alone would result in a more timely recovery of nearly $70,000, which is more than $1,000 per physician per year in revenues, realizing that there will be contractual adjustments in actual reimbursement. There are methods for providers to reduce their errors, such as employing an electronic claims administrator, using an electronic claims clearinghouse, and reallocation of resources. These methods would need to be tailored to the unique circumstances of each individual practice. Limitations We recognize several limitations to this study. First, in our discussion of recovery of payments, we assume that the claim will be paid at some point by either the payor or patient. In many cases, it will be paid by the payor, especially in cases where the provider set up the insurance incorrectly, or the payor rejected the claim in error. In other cases, the patient will be billed, especially when the payor did not send an explanation of denial or when the patient has given incorrect insurance information. As the claim ages, however, the possibility of reimbursement for the service decreases. Many contracts between providers and payors stipulate that the claim must be submitted within 1 year. If the provider does not identify the error and send a timely bill to the payor, the claim is not reimbursed and is considered a contractual write-off. In contrast to commercial payors, Medicare only withholds 10% of the payment if they receive the bill more than 1 year since the date of service. The unpaid amount cannot legally be billed to the patient but must be taken as a contractual write-off. Another limitation of the current study is that data on payment timing and individual payor patterns were not available. Therefore, we were not able to consider such other important issues as whether there was a relationship between the amount of the claim and the

5 Practice Management Vol. 35, No length of the delay. Such a relationship would be additional evidence of payors responding to their economic incentives to withhold payments. In addition, with data on individual payors, we may have been able to comment on whether certain payors or classes of payors were more likely than others to act on their incentives to delay payment. Finally, all data from this study were obtained from a single medical practice in the state of Washington, and some of the payors were located in the same state. Washington enacted legislation for payment of clean claims at the end of 1999 (WAC ). The change in legislation may have diminished the incentive to delay payment in the latter part of our data. Further, because data came from a single state, our results may not be representative of the country generally. Group practice variations may also reflect different billing techniques and different methods of collecting on claims. Therefore, other group practice types may have errors that were not identified in our study. Conclusions Delayed payments are growing in importance to medical care providers, and there is a growing need to teach medical students and residents the nature and pitfalls of medical billing, at least to the extent that delayed payments can be minimized. This research is one of the first empirical analyses of the source of delayed payment of physician claims and has important implications for both current and future health care providers. Providers who are concerned about delayed payment of claims should consider their own processes for setting up insurance accounts, following up on denials, and facilitating charge entries. Acknowledgment: Funding for this study was salary support from the authors institutions. This study was presented at the Academy for Health Services Research and Health Policy June 25, 2002, in Washington, DC. Corresponding Author: Address correspondence to Dr Hollenbeak, Penn State College of Medicine, PO Box 850, MC H113, Hershey, PA Fax: chollenbeak@psu.edu. REFERENCES 1. Coleman R. Monitoring fiscal intermediary payment performance. Patient Accounting 2000;23(9): Collect data, monitor claims receipts to ensure prompt Medicare, Medicaid payment. Public Sector Contracting Report 1999;5(10): Hoffer K. Resolving payment disputes in small claims court. Patient Accounting 1998;21(1): Pallarito K. Falling behind in the payment game. Mod Healthc 1999; 29(25): Silversmith J. Late payment of physician claims: results of an MMA survey. Minn Med 2000;83(2): Traska M, Gentile B. Late receivables squeeze cash flow. Hospitals 1985;59(14): Lewers DT. Insurers must pay on time or pay the consequences. Am Med News 2001;144(5):Feb 5.