4
Excess Administrative Costs

INTRODUCTION

Administrative costs in the United States consumed an estimated $156 billion in 2007, with projections to reach $315 billion by 2018 (Collins et al., 2009). With the time, costs, and personnel necessary to process billing and insurance-related (BIR) activities from contracting to payment validation on the provider side and the needs of payers to process claims and credential providers, significant redundancy and inefficiency arises from healthcare administration. Adding to concerns is emerging evidence of an inverse relationship between administrative complexity and quality of care (Himmelstein and Woolhandler, 2002). The presenters in this session approach estimating excess administrative costs from a variety of macro-and microeconomic levels, all with the goal of identifying the portion of expenditures spent on administration that could be reduced by increasing the efficiency of the delivery system, which highlights the need for administrative simplification and harmonization.

James G. Kahn of the University of California-San Francisco discusses BIR costs at the provider level. He puts these costs in the context of a complex payment system, describing three main drivers of BIR costs: complexity, variability, and friction. Using available evidence, he estimates that up to $183 billion of expenditures on BIR activities in the United States may be due to inefficiency. However, he also encourages caution in interpreting the results, given the lack of adequate data on the BIR costs in several settings, such as in pharmacies and nursing homes. Lawrence P. Casalino of Weill Cornell Medical College builds on this presentation, citing evidence that



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 141
4 Excess Administrative Costs INTRODUCTION Administrative costs in the United States consumed an estimated $156 billion in 2007, with projections to reach $315 billion by 2018 (Col- lins et al., 2009). With the time, costs, and personnel necessary to process billing and insurance-related (BIR) activities from contracting to payment validation on the provider side and the needs of payers to process claims and credential providers, significant redundancy and inefficiency arises from healthcare administration. Adding to concerns is emerging evidence of an inverse relationship between administrative complexity and quality of care (Himmelstein and Woolhandler, 2002). The presenters in this session approach estimating excess administrative costs from a variety of macro- and microeconomic levels, all with the goal of identifying the portion of expenditures spent on administration that could be reduced by increasing the efficiency of the delivery system, which highlights the need for admin- istrative simplification and harmonization. James G. Kahn of the University of California-San Francisco discusses BIR costs at the provider level. He puts these costs in the context of a com- plex payment system, describing three main drivers of BIR costs: complex- ity, variability, and friction. Using available evidence, he estimates that up to $183 billion of expenditures on BIR activities in the United States may be due to inefficiency. However, he also encourages caution in interpreting the results, given the lack of adequate data on the BIR costs in several settings, such as in pharmacies and nursing homes. Lawrence P. Casalino of Weill Cornell Medical College builds on this presentation, citing evidence that 

OCR for page 141
2 THE HEALTHCARE IMPERATIVE the average American physician spends 3.8 hours a week—the equivalent of more than 3 workweeks a year—on interactions with payers (Casalino et al., 2009). He estimates that Canada spends between $15 and $32 bil- lion less on BIR activities compared to the United States because of greater administrative standardization. James L. Heffernan from the Massachusetts General Physicians Orga- nization highlights how other economic sectors, such as industrial manu- facturers, commit significantly less resources to administration compared to the healthcare sector. Comparing the administrative costs of a single professional billing office to that of Medicare, he surmises that standard- izing administrative complexity could save $26 billion for physician and clinical services’ billing operations along with 4 hours of professional time per physician per week and 5 hours of practice support staff time per week. He therefore concludes that a single, transparent set of payment rules in a multipayer healthcare system would potentially reduce the burden common in a provider’s billing office. Concluding this session, Andrew L. Naugle of Milliman explores ad- ministrative costs for payers, focusing on the commercial market. He esti- mates that, if commercial insurers could all adopt the best-practice level of administrative expenses being no more than approximately 7.6 percent of fully insured commercial premiums, up to a $23 billion savings opportunity exists for the commercial market in total administrative expense reduction. As these estimates applied data across the entire commercial marketplace, Naugle cautions that variation in savings could occur across specific indi- vidual payers as they each will be variously impacted by their respective marketplace and organizational characteristics. Outlining opportunities to capitalize on these savings, he discusses possible policy options, including the elimination of manual transactions between payers and providers; sim- plifying the sales process; maximizing self-service capabilities and adoption; and standardizing payer and provider interaction processes and rules. EXCESS BILLING AND INSURANCE- RELATED ADMINISTRATIVE COSTS James G. Kahn, M.D., M.P.H. University of California-San Francisco There are very substantial administrative costs in the U.S. healthcare system, making up, by one estimate, nearly one-third of all spending (Wool- handler et al., 2003). A major portion of these costs are BIR activities undertaken to fulfill the requirements of getting paid, from contracting through collections. Most BIR activities occur at the provider level, with a smaller amount at the insurer level. BIR activities thus reflect the trans-

OCR for page 141
 EXCESS ADMINISTRATIVE COSTS action cost of a complex payment system. As such, it largely constitutes inefficiency rather than added value. This economic loss can be mitigated if there are secondary benefits (e.g., enhanced quality or use management). Three features of managing health insurance drive BIR costs at the provider level. The first is complexity. The insurance process has multiple steps, often demanding precise accuracy and attention to detail. BIR steps include contracting with insurers and subcontracted providers; maintain- ing benefits databases; determining patient insurance and cost sharing; collecting copayments, formulary, and prior authorization; coding of ser- vices delivered; checking and submitting claims; receiving and depositing payments; appealing denials and underpayments; collecting from patients; negotiating end-of-year resolution of unsettled claims; and paying subcon- tracted providers. The second burdensome feature of managing insurance is variation. Due to consolidation of insurers in recent decades, a provider practice likely has fewer payers to deal with. However, each payer offers multiple products and often further customizes products to individual purchasers (such as a large employer). Each provider may have to deal with dozens to hundreds of different plans. Providers must track plan-specific benefits and reimbursement rules, maintain special databases and benefit experts, and conduct time-consuming checks of plan details prospectively and in response to claims denials. This situation is in stark contrast to privately administered plans in other developed countries, where there is typically a single primary benefits package. The third feature is friction. Many BIR steps slow and complicate the process of getting paid. These include priority authorizations and formu- lary restrictions, high rates of nonpayment for initial submissions (10 to 15 percent), underpayments, and ultimate non- and underpayment (5 to 10 percent) (Gans, 2009). Providers express frustration and occasionally a suspicion that the process is kept complicated to lower ultimate payment levels. This report estimates total U.S. BIR costs for providers, using avail- able evidence on BIR rates applied to National Health Expenditures. For completeness, it includes a similar estimate of BIR costs for private payers. Finally, to facilitate synthesis, the report includes a tabular summary of the five administrative cost estimates presented at the IOM Roundtable on Value & Science-Driven Health Care’s workshop titled The Healthcare Imperative in May 2009. Billing and Insurance-Related Administrative Costs Existing estimates of the BIR component of administrative costs at providers fall into two broad categories: macroanalyses and microanalyses.

OCR for page 141
 THE HEALTHCARE IMPERATIVE Using a macro approach, some studies have compared total administrative costs in U.S. settings to similar settings in Canada and mainly attributed the observed differences to demands of a multipayer system. More promising for this analysis, subsequent micro studies have described and inventoried BIR tasks and associated costs, in increasing detail over time. The first study (Woolhandler et al., 2003) found a roughly threefold difference in total administrative spending per capita between the two countries in 1999. For physician practices, administrative spending was $107 (16 percent of revenue) per capita in Canada, and $324 (27 percent) in the United States. For hospitals, the amounts were $103 (13 percent) and $315 (24 percent), respectively. Thus, BIR costs were estimated indirectly (if imprecisely) as $217 and $212 per capita in 1999, in the two provider settings. Later studies increasingly honed in on BIR tasks. Kahn and colleagues (2005) studied BIR costs in 2001 in California, finding that BIR administra- tion represents 14 percent of physician revenue and 6.6 to 10.8 percent of hospital revenue. Sakowski and colleagues (2009) studied a large multispe- cialty group practice in California, finding that BIR activities represents 10 to 12 percent of revenue, with higher percentages related to the clinician time needed to code services for billing purposes. Casalino and colleagues (2009) found that the portion of BIR activities related to staff compensation in dealing with private payers is 6.9 percent of revenue. When adjusted to include public payers, overhead costs, and a portion of clinician coding of services, this translates to a total BIR cost equal to 13 percent of revenue. Estimates of Billing and Insurance-Related Costs This report derives estimates of total BIR costs in the United States, drawing on existing research and reference data to create as comprehensive a picture as possible. In addition, to foster a clear overview and synthesis of all administrative cost estimates presented in these proceedings, this report includes a table that systematically summarizes the estimates, identifies and reconciles differences in scope and method, and presents a best estimate for each component of BIR costs. Total BIR costs in each healthcare system setting (e.g., physician prac- tices) was calculated as the product of two factors: BIR cost as a percent of revenue (from published studies) and National Health Expenditures (CMS, 2007). This analysis defined BIR-specific activities as potential excess, as com- pared with a system that greatly simplifies BIR requirements. An attempt to formally benchmark using the Canada vs. United States macrocomparison described above failed—observed differences in administrative costs slightly exceed the BIR costs calculated directly at U.S. providers. That is, BIR costs

OCR for page 141
 EXCESS ADMINISTRATIVE COSTS are greater when calculated with the macro approach than with the micro approach. This suggests that either the cross-national macrocomparison overestimates BIR costs or the U.S. ground-up (micro) measurement un- derestimates BIR costs. Thus the estimates of excess BIR costs presented here are upper bounds; we attempt to correct for benchmarking in the summary table. We also considered comparing different U.S. providers to identify best practices (lowest BIR cost) and thus the excess BIR expenditure attribut- able to less efficient providers would be highlighted by examining statistical distributions across practices. This is analogous to the approach used for differences in clinical services use and cost presented in other sections of these proceedings. In this method, we would examine statistical distribu- tions across practices. However, there are serious technical impediments. Crude statistical distributions may obscure real explanatory differences (e.g., greater administrative burden or lower practice income). For example, an HMO with high-market power may negotiate high-payment rates, with no added administrative burden, leading to an artificially deflated BIR burden. Or, a practice may operate in an especially complex payer environ- ment, driving up BIR costs. Further, this approach might be taken to imply that there are only minimal savings to be obtained at all providers from sim- plified BIR demands (e.g., standard billing forms or benefit plans). In this way, reducing BIR costs differs from efforts to reduce the well-understood variation in clinical practices. That is, the largest savings in administrative costs may derive from lowering everyone’s costs by simplifying the system, whereas the largest savings in clinical practice may stem from emulating current best practices. Savings Opportunities For physician care, annual expenditures from National Health Expen- ditures (NHE) projections for 2009 are $539 billion (CMS, 2007). The BIR portion of physician revenue is estimated at 13 percent, based on data from three studies (Casalino et al., 2009; Kahn et al., 2005; Sakowski et al., 2009). The result is an estimated $70 billion per year in BIR expenditures, representing an upper bound on “excess” due to the lack of an adequate benchmark (Table 4-1). For hospital care, the $789 billion in annual spend- ing is multiplied by the midpoint of 8.5 percent from one study (Kahn et al., 2005), yielding an estimate of $67 billion. The total for physicians and hospitals is $137 billion per year. If a similar rate applies to other pro- viders (e.g., pharmacies and nursing homes), the total for BIR costs at all providers is $254 billion. We also present an estimate of BIR costs at private insurers. Private insurers have an estimated $854 billion in annual revenue in 2009 (CMS,

OCR for page 141
 THE HEALTHCARE IMPERATIVE TABLE 4-1 Estimate of Billing and Insurance-Related (BIR) Costs in the U.S. Healthcare System in 2009 Annual NHE Percentage Annual BIR Costs (in billions) for BIR Costs (in billions) Physician care $539 13 $70 Hospital $789 8.5 $67 Subtotal $137 Other providers $771 10 $77 Cumulative subtotal $214 Private insurers $854 12.3 $105 Public programs $1,191 3.5 $42 Cumulative total $361 NOTE: NHE = national health expenditure. SOURCE: CMS, 2007. 2007). They have an overhead (administration and profits) of 12.3 percent (CMS, 2007). Thus, by multiplying total annual revenues by the percentage consumed by overhead, the total BIR cost is estimated at $105 billion. This is consistent with reporting for 2007 of $94.6 billion for the administration and net cost of private health insurance (CMS, 2007). An appropriate benchmark is the overhead for public programs, which is 2 percent and 4 to 5 percent for Medicare and Medicaid, respectively (CMS, 2007); this is incorporated in the summary table for all the admin- istrative cost estimates presented in this session. Adding each of the individual BIR estimates together, the total up- per bound for BIR costs is estimated at $361 billion in 2009. Adjust- ment for estimated benchmarks decreases this amount by about one-third (Table 4-1). Primary Assumptions and Caveats This analysis assumes that it is possible to distinguish BIR costs from other administrative functions. The triangulated, mutually consistent data using varied methods suggests that this is true, to reasonable precision. Consistency with qualitative data (e.g., physician description of a major BIR burden) is also encouraging (Casalino et al., 2009; Sakowski et al., 2009). Second, we assume that BIR costs are not dropping since these data were collected. The recent findings of Casalino and colleagues (2009) are comparable to Kahn and colleagues (2005), with data that are 5 years

OCR for page 141
 EXCESS ADMINISTRATIVE COSTS apart. Further, the majority of physician groups believe that the effort needed to deal with plans is increasing (Casalino et al., 2009). Third, there is uncertainty in the BIR estimates. Our knowledge of BIR costs is perhaps best for physician offices, with much more limited data for hospitals, and almost none for other providers (e.g., nursing homes, labs, and pharmacies). Even in physician practices, there is uncertainty: clinician coding of services provided (about 2 percent of revenue) is necessary for billing but may also provide useful information for outcomes measure- ment and quality improvement. Hospital administrative reporting includes a “general administration” category that comprises 4 percent of revenue, with insufficient information to apportion to BIR activities and other ad- ministration. Notably, the BIR cost at other providers (e.g., pharmacies) is a major unknown (the analysis assumed 10 percent, based on informal observations and discussions about pharmacy operations and anecdotal reports on pharmacy benefit management procedures and fees). For public programs, there are multiple and potentially inconsistent sources of BIR burden; we rely on estimates from the Centers for Medicare & Medicaid Services (CMS), which are lower than aggregate values in NHE. This analysis could identify no definitive benchmarks. In the summary table (Table 4-2), we incorporate benchmarks used by others, leading to a drop in the estimate of “excess” BIR costs. Finally, the BIR costs reported here may overlap with excess clinical services. That is, if expensive clinical services are reduced 5 percent through more proactive, patient-centered care, there will also be a drop in BIR costs. However, this drop may be less than proportional, for two reasons. First, some BIR costs are a function of the number of patients or plans, not the number of services. Second, the services targeted for reduction are more expensive (e.g., MRIs cost more than office visits), so that relatively fixed BIR costs per service will represent a smaller portion of the reduced rather than of the retained services. Next Steps There would be value in broadened and improved BIR research. Better BIR studies are very much needed for hospitals, due to the presence of only one study, and hampered by a large undefined administrative cost category. The need for BIR research for other providers is even more acute. For phy- sician groups, it could be useful to conduct more in-depth studies (i.e., the depth of Sakowski et al., 2009, with the sample of Casalino et al., 2009). Intervention studies are a critical next phase. Policy makers will need to know how much BIR cost will be reduced with changes in procedures that are small (e.g., single billing form), medium (e.g., standard core benefits

OCR for page 141
 THE HEALTHCARE IMPERATIVE TABLE 4-2 Synthesis of Estimates from Presentations on Excessa Administrative Costs Billing and Insurance- Related Administrative Costs Method Basis for Roundtable Data Types of Costs Estimating Setting Presenter Total Excess* Source(s) Included Excess Private Jensen n/a $63 OECD All Comparison Insurers billion administration U.S vs. other & profits OECD, adjusted for wealth Kahn $105 $75 U.S. national All Difference in billion billion health administration overhead for expenditures & profits private vs. public payers Synthesis $105 $63-75 See above All Range from billion billion administration above & profits Physicians Casalino $65 $32 U.S. 6 major Ratio based billion billion representative activities. No on Canadian survey, service coding. survey applied to (preliminary, NHE potentially conservative) Kahn $70 n.s. Two All BIR tasks None billion California (with half available studies, of service applied to coding), all NHE payers & cost Heffernan n.s. $26 Mass. All BIR tasks, Micro-costing billion General Phys. for private of current Org, applied payers only, private payers to NHE for 2009 vs. Medicare Synthesis $65- $32-35 As above Similar to Use of 70 billion Kahn: all Casalino billion payers and preliminary BIR tasks ratio for physician practices

OCR for page 141
9 EXCESS ADMINISTRATIVE COSTS TABLE 4-2 Continued Billing and Insurance- Related Administrative Costs Method Basis for Roundtable Data Types of Costs Estimating Setting Presenter Total Excess* Source(s) Included Excess Hospitals Kahn $67 n.s. One All BIR None billion California activities available study, applied to NHE Synthesis $67 $34 As above As above Use of billion billion Casalino preliminary ratio for physician practices Other Kahn $77 n.s. NHE, with Assumed 10% None providers billion assumed BIR BIR, based available on physicians and hospital data Synthesis $77 $39 As above As above Use of billion billion Casalino preliminary ratio for physician practices TOTALb $168- 183 billion NOTE: BIR = billing-and-insurance related; n/a = not applicable; NHE = national health expenditures; n.s. = not significant; OECD = Organisation for Economic Co-operation and Development. a By “excess” we mean spending above the indicated benchmark comparison. We make no judgment on whether that excess spending brings value. b Estimates of provider BIR excess rely on the preliminary U.S.:Canada ratio used by Casa- lino for physicians. As this ratio is finalized, the estimates will evolve.

OCR for page 141
0 THE HEALTHCARE IMPERATIVE package), or pervasive (e.g., single payer) in order to effectively minimize the waste associated with these costs in the U.S. healthcare system. Synthesis of Presentations on Excess Administrative Costs Two analyses of BIR costs among private insurers found very similar results. Eric Jensen estimated $63 billion in excess BIR costs at private insurers, as compared with OECD countries (which have a much lower private payer role). Kahn and colleagues estimated $105 billion in total BIR costs at private insurers, based on U.S. NHE data, and $75 billion in excess as compared with U.S. public payers (CMS, 2007). The synthesis range is $- billion. Three assessments of administrative costs in physician practices applied data from practices to the physician category in the National Health Expen- ditures. The analysis presented by Casalino and colleagues (2009) focused on staff costs in a national sample, for six major tasks required to deal with private health plans. For these written proceedings, Casalino and col- leagues adjusted their analysis to include private and public payers, as well as nonstaff overhead, estimating $64.7 billion in BIR costs. As compared with costs for similar activities in Canada, the estimated annual excess in the United States is $32 billion. The analysis presented by Heffernan used data from the Massachusetts General Physician Organization, which has a relatively efficient billing operation as compared with national means. A microcosting of tasks to obtain private insurer payment versus Medicare (admittedly an imperfect payer itself) found a 10 percent excess burden. The analysis presented by Kahn and colleagues used data from two California studies that included a broader range of billing and insurance-related activi- ties, including a portion of clinician coding of services provided. All payers and costs (e.g., overhead) are included, yielding an estimated $70 billion per year in total BIR costs. The differences between these estimates can be explained as follows. The revised estimate by Casalino and colleagues is $5 billion less than the estimate from Kahn and colleagues, and probably reflects the different treatment of clinician service coding and imprecision.1 Heffernan used a method that includes all BIR tasks but estimates national burden based only on private payers (i.e., assuming no excess burden to providers from dealing with Medicaid or Medicare). 1 The $39 billion difference between the Kahn et al. estimate and the Casalino et al. estimate, presented earlier, reflects the public payer portion ($14 billion); additional BIR activities, such as health information technology and answering patient billing questions ($6 billion); overhead at 12 percent ($7 billion); a correction in the denominator derived from NHE data ($5 billion); and half of clinician service coding ($7 billion).

OCR for page 141
 EXCESS ADMINISTRATIVE COSTS The synthesis estimate for excess BIR costs for physician practices is $2- billion. This includes all BIR activities, all payers, and all costs, with Canada preliminary analysis on staff time (Casalino et al., 2009) used to generate a benchmark ratio (i.e., excess = 50 percent total BIR). For hospitals, Kahn presented the only estimate of $67 billion total BIR cost based on study of BIR percentages and NHE expenditures. Applying the United States vs. Canada preliminary benchmark ratio used by Casalino and colleagues for physician practices yields a net excess BIR estimate of $ billion. For other providers (e.g., labs, pharmacies, and nursing homes), an as- sumption of 10 percent BIR (between physician and hospital levels) yields an estimated $77 billion in total BIR costs. Again, applying the preliminary benchmark ratio used by Casalino and colleagues for physicians yields an estimated $9 billion in excess BIR costs for these providers. The total excess BIR cost is estimated at $-$ billion per year, in 2009 dollars. This is the sum of values presented above. It relies heavily on the preliminary benchmark ratio used by Casalino and colleagues for physician practices. As this ratio is finalized, the BIR estimate using this method will evolve. This estimate is conservative in assuming no BIR excess for public programs, though some of these are more costly to administer than others, and no BIR outside of the health sector (e.g., at employers). As well, further research on BIR costs in specific sectors in the United States and Canada will lead to evolving estimates of BIR costs, total and excess. Nonetheless, we believe that this synthesis represents the best integration of existing data on BIR costs and the most accurate comprehensive esti- mate of current excess BIR costs—about 7 to 8 percent of U.S. spending on health care. WHAT DOES IT COST PHYSICIAN PRACTICES TO INTERACT WITH PAYERS? Lawrence P. Casalino, M.D., Ph.D., Weill Cornell Medical College; Sean Nicholson, Ph.D., Cornell University; David N. Gans, MSHA, and Terry Hammons, M.D., M.S., Medical Group Management Association; Dante Morra, M.D., M.B.A., and Wendy Levinson, M.D., University of Toronto Physicians in the United States have multiple forms of interaction with different payers. Interactions with payers, such as private health insurance plans, Medicare, and Medicaid, include obtaining prior authorization, dealing with formularies, submitting claims and verifying the accuracy with which they are paid, submitting quality data and reviewing payer-generated quality performance reports, negotiating contracts, and having physicians credentialed by the payers.

OCR for page 141
 THE HEALTHCARE IMPERATIVE TABLE 4-7 How Could Administrative Complexity Burden Be Reduced if a Single Set of Rules Were Used? Group Practice Management • Time to research and understand payer rules/guidelines would be reduced • Time reviewing and analyzing rejections would decrease • Work flow could be more streamlined and efficient • Time saved working transaction edits Third Party • Reduction in rejection claim follow-up time • Reduction in effort to maintain different formats for scrubbed claims • Reduction in overall billing effort due to easily accessible online EOB information Coding • Reduction in time working payer-specific TES edits and PCS work files • Reduction in time dedicated to Radiology bundling edits and Radiology local policy review edits Production • Elimination of all manual processing of paper checks and EOBs • Elimination of resources required to scan paper EOBs Management Information Systems • Elimination of referral manager queues • Elimination of open referral module maintenance • Elimination of payer-specific dictionary fields • Reduction in time to implement 835 receipt files • Reduction in PCS work file compile routines Customer Service • Reduction in volume of insurance-related questions Payer Relations • Reduction in time related to the research of payer policies NOTE: EOB = explanation of benefits; PCS = paperless collection system; TES = transaction- editing system. SOURCE: Prepublication data prepared by the authors for research funded by the Robert Wood Johnson Foundation. list of opportunities to pursue common policies and procedures leading to lower cost. Efforts continue to realize the savings, including working with the Council on Affordable Quality Healthcare to develop a more efficient process for large groups, in addition to the progress they have made for individual practitioners and small groups. A new effort was launched in the last 3 months that brings together payers and providers with employers in an expanded effort. The program is called EACH, or Employer Action Coalition on Healthcare. It is a three-

OCR for page 141
 EXCESS ADMINISTRATIVE COSTS pronged effort, with administrative simplification being one of the major goals. Providers and payers continue to take these voluntary actions in many communities. But, both payers and providers find it difficult to make the investment to change to a more efficient collaborative system when their own systems are working well for their purpose. Effectively these fac- tors prevent the achievement of the important administrative simplification goals. Some commentators have concluded that this is evidence that we need a single national payer. Consider as an alternative, that a common administrator might be a more effective solution. The analysis described in this paper used Medicare rules as a possible benchmark. In addition, CMS is moving toward us- ing common Medicare contractors in each region. The contractors have launched the major effort to coordinate hospital and physician payments. These organizations will be in a position to compete to administer services consistently across payers in a region. The Medicare contractors are already required to meet performance requirements, maintain an extensive system of reporting, deliver provider education, and a track record of low cost. The processes involved in the revenue cycle and therefore in the admin- istrative simplification effort are not static. They change with the addition of new technology for both payers and providers. They are impacted by new developments in benefit management. They are potentially significantly impacted by the new payment models that may be anticipated with health reform. Administrative simplification requires a dynamic measurement pro- cess to allow for continuous improvement and adjustment as the perfor- mance changes. A few areas of the country have experience with provider and payer–provider performance reporting tools. These have the potential to transparently monitor the effectiveness of both payer and provider sys- tems to improve both the effectiveness and efficiency of the revenue cycle. Primary Caveats and Assumptions We recognize that there are other important limitations to our study. First, our study was limited to just one PBO. Although there is some evidence to suggest that the operations of that PBO are relatively efficient (thus biasing our savings estimates toward being conservative) there may be significant variation in administrative costs between PBOs, which could be similar to that found among hospitals (McKay et al., 2008). For example, PBOs in less competitive markets with a single dominant commercial payer could have lower administrative costs due to a smaller number of bill- ing rules and processes. Second, we focused on excessive administrative complexity in fee-for-service payments. We recognize that with underman- aged care the payment rules may be purposefully more restrictive so our results may not be directly generalized to capitated and other managed care arrangements.

OCR for page 141
 THE HEALTHCARE IMPERATIVE We focused on a PBO perspective, recognizing that additional savings would also accrue to payers (who have corresponding costs associated with adjudicating claims and appeals) under our single payment rule scenario, but they could not be quantified here. In addition, there are emerging costs related to quality reporting, improvement, and pay-for-performance administration that are not included here because they could not be char- acterized as “non-value-added” activities. The incremental costs incurred by the PBO to meet different performance management measures have not been identified, yet contribute to the dynamic nature of the administrative simplification in light of the health reform debate. Conclusion An incremental move to one set of payment rules would yield sig- nificant dollar savings and work-life and productivity opportunities for physicians would be created. The savings from reducing administrative complexity would be translated into decreased costs in general and provide resources that could be passed on as savings to purchasers and patients or provide additional needed health services. Achieving these savings would not require restructuring the basic market system of our complex healthcare system through mandating a single payer. Rather, mandating a single set of rules, a single claim form, standard rules of submission, and transparent payment adjudication—with corresponding savings to both providers and payers—could provide systemwide savings that could translate into better care for Americans. EXCESS HEALTH INSURANCE ADMINISTRATIVE EXPENSES Andrew L. Naugle, M.B.A. Milliman, Inc. Within the context of the 2009 U.S. healthcare reform discussion, significant attention has been paid to identifying opportunities to reduce administrative expenses. Every stakeholder in the health insurance system incurs some administrative expense—payers, providers, purchasers, and even patients. Efforts to reduce these costs, especially those of payers and providers, have the potential to produce substantial financial savings, which could be used to fund additional care or be redirected for other purposes. Our experience working with both payers and providers convinces us that there is widespread agreement that administrative expense reduction is both worthwhile and possible. In many cases, we believe that there is also agreement regarding viable high-level tactics for reducing administrative expenses. The points of contention and disagreement, which have precluded

OCR for page 141
 EXCESS ADMINISTRATIVE COSTS significant administrative expense reduction in recent years, tend to involve funding of cost-reduction initiatives (who will pay for them?), avoidance of risk associated with changes to the status quo, and the potential for loss of payer competitiveness through product commoditization. This paper quantifies the industry-wide administrative expense-reduc- tion opportunity that the commercial payer community could achieve by transitioning from today’s average administrative expense level to a best- practice administrative expense level. In addition, the paper identifies some tactics that could be employed by the industry to achieve these potential cost reductions. For the purposes of this analysis, we have defined administrative ex- pense as all expenses incurred by payers for common administrative func- tions such as claim processing, customer service, underwriting, medical management, and sales and marketing, as well corporate overhead and external broker commissions. We have excluded premium taxes from the analysis. It is important to note that this paper only focuses on the commercial market (e.g., not Medicare, Medicaid, or TRICARE) and excludes supple- mental products (e.g., vision, dental, and hospital indemnity plans). It also ignores potential savings that could be realized by other stakeholders (specifically, providers and purchasers) through implementation of cost- reduction strategies by payers. Quantifying the Expense Reduction Opportunity We used the following methodological approach to quantify the mag- nitude of the administrative expense-reduction opportunity: • Estimated the total dollar value of commercial premiums for the entire U.S. health insurance marketplace; • Estimated the distribution of commercial premiums between self- insured and fully insured products; • Estimated total administrative expense associated with fully in- sured commercial products; • Estimated total administrative expense for fully insured commercial products assuming a shift from current expense levels to a best- practice level; • Calculated the savings opportunity for fully insured commercial products as the difference between the current administrative ex- pense level and the estimated best-practice expense level; • Estimated the marginal expense reduction opportunity for self- insured business as a percentage of the marginal expense-reduction opportunity for fully insured business; and

OCR for page 141
 THE HEALTHCARE IMPERATIVE • Calculated the range of total possible savings as the sum of the value for fully insured commercial business and the range of pos- sible savings for self-insured commercial business. The methodology and data sources we used to develop these estimates are described below. Value of Current Total Commercial Health Insurance Premiums Our estimate of the total value of health insurance premiums for the commercial health insurance market is based on the Milliman Healthcare Reform Database. The Milliman database contains cost details for U.S. sub- populations (market segments), with the total reconciling National Health Expenditures data for 2008. According to this data source, 2008 U.S. health insurance commercial premiums, including premium equivalents for self- funded products, totaled approximately $700 billion. Distribution of Commercial Health Insurance Premiums There are two primary types of risk arrangements in the health insur- ance market: fully insured and self-insured (also known as “self-funded”). For fully insured products, the insurance company (the payer) takes the financial risk on the claims cost. For self-insured products, the purchaser (typically the employer) takes that financial risk. The self-insured market has grown substantially since implementation of the Employee Retirement Income Security Act of 1974 (ERISA), which recognized self-funded plans as a viable option and exempted them from most state-mandated benefits. Self-funded products are most prevalent for group sizes greater than 500 covered lives, but are a viable option for much smaller groups. This approach is typically unadvisable for groups of less than 100 covered lives because of the risk exposure. Self-funding offers several characteristics that are desirable to purchasers, including benefit design flexibility, and lower cost owing to exemption from state premium taxes (which can add 2 per- cent to the cost of a fully insured product) and the insurer’s risk margin on the claims cost. In our experience, fully insured products tend to generate a greater amount of administrative expense than self-insured products. This situation exists because of a variety of factors such as unbundling of administra- tive services, shifting of administrative responsibilities from the payer to the employer’s human resources department, and price pressure. For that reason, it was necessary to estimate the distribution of total commercial premiums between these two risk arrangements. The data sources we used to make this distribution were the Medical Expenditure Panel Survey from

OCR for page 141
9 EXCESS ADMINISTRATIVE COSTS the Agency for Healthcare Research and Quality (AHRQ), and proprietary Milliman data. Through the combination of these two data sources, we estimated that approximately $375 billion of premiums are associated with fully insured products and $325 billion of premium-equivalents with self- insured products. Administrative Expense for Fully Insured Commercial Products We estimated 2008 total administrative expense for fully insured com- mercial products using benchmarks developed from administrative expense data collected from more than 100 payers. According to these proprietary benchmarks, median payer administrative expense for fully insured com- mercial products, expressed as a percentage of fully insured commercial premiums, was 11.3 percent. Note that this definition of administrative expense is inclusive of external broker commissions, but excludes premium taxes. Using the combination of the total fully insured premiums in the com- mercial market and the median administrative expense level (using the median to approximate the mean) we calculated an estimate of $42.4 bil- lion ($375 billion × 11.3 percent) to represent total payer administrative expense for fully insured commercial products. Administrative Expense for Fully Insured Commercial Products at Best Practice Next, we developed an estimate of what total payer administrative expense for fully insured commercial products would have been in 2008 if administrative expense as a percentage of premiums was shifted from 11.3 percent to a level equivalent to that exhibited by best-practice orga- nizations. Best-practice payers tend to exhibit certain characteristics that allow them to offer high-quality service in a very efficient manner. For ex- ample, they maximize use of electronic transactions, leverage information systems to achieve high levels of automation, minimize low-value adminis- trative activities, and generally avoid unnecessary complexity. In terms of administrative expense, we defined the best-practice level, based on our experience, to be approximately 7.6 percent of fully insured commercial premiums. Although it is possible for organizations to operate effectively at lower administrative expense ratios, we find it is more com- mon for organizations with administrative costs below this level to exhibit characteristics of poor performance (e.g., high claims turnaround times, long customer service call hold times, inadequate or ineffective medical management programs) that are due to insufficient staffing. Furthermore, it is important to consider that certain administrative

OCR for page 141
0 THE HEALTHCARE IMPERATIVE costs can have an offsetting impact on benefit cost. For example, some medical management programs can help to avoid unnecessary use. Adminis- trative spending on these programs can be considered an investment, which can result in lower expenditures for healthcare services and therefore a lower total cost. Elimination of such “good” administrative expenses must be carefully considered to ensure that any administrative expense savings are not offset by increases in benefit costs. Using the best-practice administrative expense level defined above and our estimate of total fully insured commercial premiums, we estimated that total payer administrative expense would be approximately $28.5 billion ($375 billion × 7.6 percent). Administrative Expense Reduction Opportunity Fully insured commercial business Using the administrative estimates de- veloped in the two prior sections, we calculated the total administrative expense reduction opportunity for fully insured commercial products as the difference between the 2008 median and the best practice: $13.9 billion ($42.4 billion-$28.5 billion). This amount represents an estimate of the sav- ings that could be achieved by shifting the industry median administrative cost level to a level representing current best practice. Self-insured commercial business As previously stated, in our experience, self-insured products incur lower levels of administrative expense than do fully insured products. Therefore, we estimated the administrative expense reduction opportunity for these products by assuming the effect would be in the range between 50 and 75 percent of the marginal reduction for fully insured products. Given that, we estimate that additional administrative expense savings for self-insured businesses could be in the range between $6.2 billion and $9.1 billion. We calculated these estimates as shown in Table 4-8. Commercial Administrative Expense Reduction Opportunity In summary, we estimate the total administrative expense-reduction opportunity for the commercial market as the sum of the estimate for the fully insured market ($13.9 billion) and the range of estimates for the self-insured market ($6.2 billion to $9.1 billion). The resulting range is $20.1 billion to $23.0 billion, or approximately 3 percent of total com- mercial premiums. Within the context of healthcare reform, this may be a relatively con- servative estimate. It assumes that the entire payer community achieves an administrative expense level consistent with current best practices. If the

OCR for page 141
 EXCESS ADMINISTRATIVE COSTS TABLE 4-8 Estimates of Payer Administrative Expense-Reduction Opportunity for Self-Insured Business Scenario 1 Scenario 2 Percentage of marginal FI savings that can be applied to SI 50% 75% business 2008 administrative expense ratio for FI business 11.3% 11.3% 2008 best-practice administrative expense ratio for FI 7.6% 7.6% business Marginal improvement opportunity for FI business 3.7% 3.7% Marginal improvement opportunity for SI business based on 1.9% 2.8% percentage of marginal FI business reduction Estimate of total SI commercial premium equivalents $325 billion $325 billion Estimate of administrative expense-reduction opportunity $6.2 billion $9.1 billion NOTE: FI = fully insured; SI = self-insured. definition of best practices changes due to significant changes to the ad- ministrative paradigm, then even greater administrative expense reductions may be possible. Furthermore, we caution users of this report to consider the caveats and assumptions described in the next section. Caveats and Assumptions Reviewers of this document should consider the following caveats and assumptions when evaluating the results: • The savings estimates provided herein are only for payers. Second- ary savings would likely accrue to providers, purchasers, and po- tentially patients. Those savings are not estimated in this paper. • The savings estimates provided herein are only for commercial products. Additional savings may be achieved in noncommercial products (e.g., Medicare, Medicaid, TRICARE). Those savings are not estimated in this paper. • The calculation methodology applies data in a general manner across the entire marketplace. These estimates are not intended to represent what is possible for a specific plan or group of plans. It may not be possible for all payers, especially small payers, to achieve the best-practice benchmark because of a variety of circum- stances, most notably the effects of economies of scale. • We do not guarantee an organization’s or the industry’s ability to achieve the savings estimates described herein, and Milliman dis- claims any and all liability that may result from a third party’s use of this white paper.

OCR for page 141
2 THE HEALTHCARE IMPERATIVE • The opinions expressed in this white paper represent those of the author and not the opinions of Milliman, Inc. Next Steps While the opportunity to reduce payer administrative expenses in the U.S. health insurance system is great, the realization of those savings pres- ents many challenges. If the historical context is an indicator of the future, then the achievement of material administrative cost reductions will require concerted, collaborative expense-reduction efforts coordinated among all stakeholders. We believe there are opportunities to reduce the complexity that drives inefficiency in the system. To that end, we have identified a few tactics targeting those functions that drive the majority of administrative expense, and therefore represent, in our opinion, high-priority areas of focus for administrative expense reduction efforts. Eliminate Manual Transactions Between Payers and Providers According to the U.S. Healthcare Efficiency Index (www.ushealth careindex.com), the majority of common transactions between payers and providers are performed using labor-intensive, manual means. This is de- spite the fact that, in accordance with HIPAA, nearly every payer in the nation has the capability to accept electronic transactions, and significant financial benefits accrue to payers through their use. Eliminating manual transactions for claim submission, claim status inquiries, eligibility verifica- tion, claim payment, and remittance advices will substantially reduce both payer and provider administrative expenses. Simplify the Sales Process Today approximately 30 percent of payer administrative cost is driven by sales and marketing activities. Approximately one-half of that amount is driven by external broker fees. The process of purchasing group health insurance, and soliciting and evaluating proposals from multiple payers, is complicated and time consuming. Furthermore, most group insurance purchasers are not health insurance experts. These realities drive many purchasers to employ the services of a broker. Although the broker provides some valuable administrative services for less sophisticated purchasers, substantial reductions in sales and marketing expense could be achieved if purchasers could more easily compare products and prices, thus minimizing the broker’s role, and associated costs, in the sales process.

OCR for page 141
 EXCESS ADMINISTRATIVE COSTS Maximize Self-Service Capabilities and Adoption Although health insurers have made significant investments in self- service capabilities (online and telephonic), adoption rates for these services could improve significantly. The administrative expense associated with a self-service transaction is negligible when compared to the cost of handling a telephone call or processing written correspondence. Standardize Payer–Provider Interaction Processes and Rules A typical provider may have contracts with 10 or more insurers and interact with others as a nonnetwork provider. Every payer has different processes, policies, and rules. Standardization of processes for common types of interactions could reduce both provider and payer administrative expense. Scrutinize Medical Management Programs for Effectiveness Since the advent of managed care, payers have implemented many pro- grams intended to manage use of healthcare services. The clinical personnel (e.g., physicians, nurses, and other clinicians) responsible for these pro- grams are often among the most expensive administrative staff. Although some of these programs are effective in managing use and cost, others have dubious value, especially when compared to the administrative burden they impose on payers, providers, and patients. The elimination of medical management programs that do not demonstrate value could significantly reduce administrative cost. Of course, this is not a complete list, and successful implementation of all of these tactics does not guarantee realization of the full savings op- portunity. However, we believe it is possible to substantially reduce payer administrative expense to the benefit of the U.S. healthcare system. We also believe that material administrative expense reduction can be achieved without harming competition among insurers, and without reducing pro- vider reimbursement levels or diminishing quality and service to purchasers and patients. Such initiatives will, however, require coordination among all stakeholders, and implementation of carefully considered strategies adopted by all payers, to reduce complexity and eliminate administrative variation. REFERENCES American Academy of Nurse Practitioners. Number of nurse practitioners. http://www.aanp. org/AANPCMS2 (accessed July 1, 2009).

OCR for page 141
 THE HEALTHCARE IMPERATIVE American Academy of Physician Assistants. http://www.aapa.org/about-pas/faq-about-pas (accessed July 1, 2009). Casalino, L. P., S. Nicholson, D. N. Gans, T. Hammons, D. Morra, T. Karrison, and W. Levin- son. 2009. What does it cost physician practices to interact with health insurance plans? Health Affairs (Millwood) 28(4):w533-w543. CMS (Centers for Medicare & Medicaid Services). 2007. National Health expenditure pro- jections 200-20. http://www.cms.hhs.gov/NationalHealthExpendData/Downloads/ proj2007.pdf (accessed July 1, 2009). Collins, S. R., R. Nuzum, S. D. Rustgi, S. Mika, C. Schoen, and K. Davis. 2009. How health care reform can lower the costs of insurance administration. Issue Brief (Commonw Fund) 61:1-19. Credit Today. 2006. MGPO Analysis Comparing Data from Credit Today 200 Credit, Col- lection & A/R Staff Benchmarking Survey. Roanoke: Credit Today. Gans, D. 2009. Unpublished Data, June 21. The Hackett Group. 2009. Hackett Certified Best Practices. http://www.thehackettgroup.com (accessed November 6, 2009). Healy, B. 2003. Paperwork pileup. Boston Globe, April 21. Hellander, I., D. U. Himmelstein, S. Woolhandler, and S. Wolfe. 1994. Health care paper chase, 1993: The cost to the nation, the states, and the District of Columbia. International Journal of Health Services 24(1):1-9. Himmelstein, D. U., and S. Woolhandler. 2002. Taking care of business: HMOs that spend more on administration deliver lower-quality care. International Journal of Health Ser- vices 32(4):657-667. Kahn, J. G., R. Kronick, M. Kreger, and D. N. Gans. 2005. The cost of health insurance administration in California: Estimates for insurers, physicians, and hospitals. Health Affairs (Millwood) 24(6):1629-1639. Keehan, S., A. Sisko, C. Truffer, S. Smith, C. Cowan, J. Poisal, and M. K. Clemens. 2008. Health spending projections through 2017: The baby-boom generation is coming to Medicare. Health Affairs (Millwood) 27(2):w145-w155. Martin, J. 1999. Two simple steps that would cut costs and broaden access to health care. Boston Globe, January 20. Massachusetts Medical Society. 2007. MMS Index Report. Waltham: Massachusetts Medical Society. McKay, N. L., C. H. Lemak, and A. Lovett. 2008. Variations in hospital administrative costs. Journal of Healthcare Management 53(3):153-166; discussion 166-157. Pope, C. 2004. The cost of administrative complexity. Administrative intricacies add no value to health care—but the costs keep stacking up. MGMA Connexion 4(10):36-41. Sakowski, J. A., J. G. Kahn, R. G. Kronick, J. M. Newman, and H. S. Luft. 2009. Peering into the black box: Billing and insurance activities in a medical group. Health Affairs (Millwood) 28(4):w544-w554. Woolhandler, S., and D. U. Himmelstein. 1991. The deteriorating administrative efficiency of the U.S. health care system. New England Journal of Medicine 324(18):1253-1258. Woolhandler, S., T. Campbell, and D. U. Himmelstein. 2003. Costs of health care administration in the United States and Canada. New England Journal of Medicine 349(8):768-775. ———. 2004. Health care administration in the United States and Canada: Micromanage- ment, macro costs. International Journal of Health Services 34(1):65-78.