St. Francis Home in the Park v. Department of Health and Family Services

No. 98-0986
STATE OF WISCONSIN	IN COURT OF APPEALS DISTRICT III
St. Francis Home in the Park, Petitioner-Respondent-Cross- Appellant, v. Department of Health and Family Services, Respondent-Appellant-Cross- Respondent.

APPEAL and CROSS-APPEAL from an order of the circuit court for Douglas County: joseph a. McDONALD, Judge. Affirmed in part; reversed in part and cause remanded.

Before Cane, C.J., Myse, P.J., and Hoover, J.

PER CURIAM. The Wisconsin Department of Health and Family Services (DHFS) appeals a trial court order that reversed part of DHFS’s administrative ruling. DHFS disallowed sundry operating costs that St. Francis Home in the Park sought to make part of its rate base for setting its future reimbursement rate for Medical Assistance (Medicaid). St. Francis cross-appeals part of the same order that upheld other aspects of DHFS’s ruling. Both the appeal and cross-appeal concern St. Francis’ Medicaid cost reports for its fiscal 1992-93 cost year. That was the first cost year for which DHFS conducted an extensive review and examination of St. Francis’ cost-allocation methods. The disallowed costs involve food costs St. Francis incurred to prepare meals for its residents, as well as labor and utility costs St. Francis incurred to prepare those meals. DHFS ruled that St. Francis had overstated those costs. We will describe DHFS’s rulings and the parties’ arguments on these matters in the opinion. DHFS also claims that the trial court lacked jurisdiction over St. Francis’ ch. 227, Stats. review proceeding, St. Francis having failed to serve the Bureau of Health Care Financing with its ch. 227 petition. We affirm the trial court’s order in part, reverse it in part, and remand the matter directly to DHFS for proceedings consistent with this opinion.

St. Francis is a not-for-profit nursing home that receives Medicaid reimbursement from DHFS. Royalton Manor, an apartment complex for retirees, lies adjacent to the nursing home. Catholic Charities is the parent organization of both institutions. The Royalton kitchen prepares all food for both Royalton and St. Francis residents, and Royalton purchases all food for both institutions. DFHS pays St. Francis for St. Francis’ share of those costs. St. Francis residents eat their meals in their rooms, while Royalton residents eat theirs in the Royalton cafeteria. Royalton residents also have apartment kitchenettes usable for food preparation. St. Francis specifically assigns some of these food costs to St. Francis and some to Royalton. For those food costs it can trace to St. Francis or Royalton, St. Francis specifically assigns those costs to St. Francis and Royalton respectively. St. Francis uses a 60%-40% allocation for all food costs not traceable directly to St. Francis or Royalton. This allocation stands on the ratio of meals the kitchen serves to St. Francis residents (60%) and Royalton residents (40%). In other words, St. Francis estimates that St. Francis residents benefit 60% and Royalton 40% from these costs. DHFS conducted a review of St. Francis’ records for fiscal cost year 1992-93 for food costs, including food purchases, labor costs for meal preparation, and utility costs for meal preparation. After this review, DHFS auditors disallowed certain costs that St. Francis had claimed, either through direct assignment or allocation by the meals ratio.

DHFS determines nursing homes’ future Medicaid reimbursement rates under a prospective payment system (PPS) that DHFS updates annually. See § 49.45(6m)(ag), Stats. For the annual update, DHFS relies on cost information it gathers from each nursing home in the prior year. As part of that process, nursing homes must maintain records sufficient for DHFS to verify the providers’ cost data. See § 49.45(3)(f), Stats. DHFS may disallow a nursing home’s asserted costs if DHFS cannot verify them from the home’s financial records. See § 49.45(3)(f), Stats. Under DHFS Medicaid rules, nursing homes must keep truthful, accurate, complete, and concise documentation, as well as medical and financial records. See Wis. Adm. Code § HFS 106.02(9)(a). Nursing homes, not DHFS, are responsible for the truthfulness, accuracy, timeliness, and completeness of all claims, cost reports, and any supplementary information relating to the provider’s Medicaid certification and reimbursement. See Wis. Adm. Code § HFS 106.02(9)(e). This duty includes, but is not limited to, the truthfulness, accuracy, timeliness, and completeness of all documentation necessary to support each claim. See Wis. Adm. Code § HFS 106.02(9)(e). DHFS may refuse to pay claims of any nursing home that fails or refuses to keep records supporting its claims. See Wis. Adm. Code § HFS 106.02(9)(g). DHFS may also disallow costs that are not necessary and proper to patient care. See § 49.45(6m)(ag), Stats. DHFS may further disallow and make adjustments to claims for mathematical errors in computations. See § 49.45(3)(f)2, Stats.

DHFS’s § 49.45(6m)(ag) prospective payment system (PPS) appears in its Methods of Implementation of Nursing Home Payments (“Methods”). The Methods give more instructions on nursing homes’ financial reporting duties. Under Methods § 1.254, DHFS requires nursing homes to keep adequate documentation to support their cost calculations, including the basis for allocating direct and indirect costs between healthcare providers like St. Francis and related organizations like Royalton. The Methods do not expressly cover all cost issues that may arise under the Medicaid program. Under Methods § 1.255, however, DHFS puts nursing homes on notice that it will generally refer to Medicare guidelines and interpretations when examining payment issues arising out of costs of related organizations. Implicit in this statement is the principle that DHFS will apply only those Medicare guidelines that make sense in a particular instance. In addition, the Methods provide that DHFS will develop necessary and proper administrative policies and procedures to cover the usual and customary situations. See Methods § 1.110. DHFS acknowledges that such policies and procedures will not necessarily apply to special situations and circumstances. See Methods § 1.110. The Methods require nursing homes to file cost reports and to disclose financial and other information necessary for verification of the costs incurred. See Methods § 1.160. Provider financial records must have sufficient accuracy and give sufficient detail to substantiate the reported cost data. See Methods § 1.174.

The Methods themselves do not answer the question of how nursing homes should allocate costs to products like processed meals. That answer lies in the discipline of cost accounting. The discipline uses various means (1) to analyze cost behavior, product flows, and cost flows through an enterprise, and then (2) to derive cost allocation methods from such analysis. First, the discipline specifically assigns to a specific product all costs it can physically trace directly to that product. See Horngren, Cost Accounting 404-05, 810-11 (3d ed. 1972). Second, for those costs not physically traceable to a specific product, the discipline relies on cost allocations derived from various techniques of cost flow analysis. One of these is the industrial-engineering method. It relies heavily on time studies and engineering measurements to analyze cost flows. See Horngren at 809-10. The method is expensive, and enterprises use it only (1) if the costs measured are a major part of total costs or (2) if the method’s benefits exceed its expense. See Horngren at 810; Dopuch, Birnberg & Demski, Cost Accounting 51-52 (2d ed. 1974). For these reasons, enterprises usually rely on a third method. Known as historical account inspection, it uses historical cost data from accounting records, rather than industrial-engineering studies, to ascertain cost flows. See Horngren at 810; Dopuch at 50. This method uses several techniques, in order of increasing complexity and expense: (1) linear analysis of two account readings; (2) high-low analysis of multiple account readings; (3) visual curve fitting to multiple account readings; and (4) statistical account analysis, including sampling, simple regression, and multiple regression techniques. See Horngren at 810-20; Dopuch at 50-80.[1]

On appeal, we review DHFS’s decision on St. Francis’ cost reports, not the trial court’s decision. See Nu-Roc Nursing Home, Inc. v. DHSS, 200 Wis.2d 405, 422, 546 N.W.2d 562, 568 (Ct. App. 1996). We review de novo DHFS’s interpretations of statutes or administrative regulations. See id. We apply, however, one of three levels of deference to DHFS’s conclusions of law if they involve more than pure interpretations of statutes and regulations. See id. We give the highest deference, great weight, to conclusions that depend on agency experience and specialized knowledge, that rest on a long-standing agency interpretation, or that deal with intertwined factual, value, and policy determinations. See id. at 422-23, 546 N.W.2d at 568-69. We apply “due weight” to interpretations of very nearly first impression, and no weight to determinations of first impression. See id. On other factual issues, we uphold DHFS’s ruling if it had relevant, credible and probative evidence on which reasonable persons could rely to reach a conclusion. See Princess House, Inc. v. DILHR, 111 Wis.2d 46,54, 330 N.W.2d 169, 173 (1983). We will not substitute our judgment for DHFS on the weight of the evidence on disputed findings as long as substantial evidence supports its finding. See CNW v. Comm’r of Railroads, 204 Wis.2d 1, 7, 553 N.W.2d 845, 848 (Ct. App. 1996). We owe no such deference, however, to questions of undisputed facts, and we have the power to view such facts independently of the agency’s determination. See Dept. of Revenue v. Milwaukee Refining Corp., 80 Wis.2d 44, 48, 257 N.W.2d 855, 858 (1977).

DHFS first claims that the trial court lacked jurisdiction over the ch. 227 review proceedings, citing the fact that St. Francis failed to serve the Bureau of Health Care Financing with its petition for review. DHFS points out that its own administrative law judge listed the bureau as a party in his decision. DHFS claims that St. Francis had a duty to serve all parties in the administrative proceedings, including all those listed in the ALJ’s decision. See § 227.53(1)(c), Stats. The bureau is a subunit of DHFS, and St. Francis claims that service on the parent agency met the terms of the statute, thereby sufficing to give the trial court jurisdiction. We agree with St. Francis. Wisconsin courts have taken a pragmatic, common-sense view of such matters, ruling that service on the parent agency suffices for service on the subunit and that service on the subunit suffices for service on the parent. See Dremel v. Nursing Home Review Bd., 119 Wis.2d 75, 79 n.5, 349 N.W.2d 725, 727 n.5 (Ct. App. 1984). Although § 227.53(1(c) requires ch. 227 petitioners to serve all parties to the administrative proceedings, no published decision has applied this rule to require separate service on both the parent and the subunit, especially subunits like the bureau that nominally appear as parties. The bureau had no independent standing in the administrative proceeding. DHFS may have the bureau act as a party under its own internal conventions, but the bureau had no rights itself. For example, had DHFS ruled in St. Francis’s favor, the bureau could not secure ch. 227 judicial review. Rather, DHFS’s decision would bind the bureau. In short, the fact that DHFS denominated the bureau a titular party has no jurisdictional import.

DHFS next challenges St. Francis’ specific identification and assignment of certain milk, condiments, and paper product costs to the nursing home. The assignment rested on physical tracing. St. Francis claimed that Royalton purchased packaged milk, packaged condiments, and certain paper products exclusively for use in the nursing home. St. Francis also claimed that Royalton purchased bulk milk and bulk condiments exclusively for Royalton residents and that St. Francis specifically assigned those costs to Royalton. These specific assignments brought about an allocation that deviates from the 60%-40% cost allocation St. Francis used for food costs not specifically identified and assigned. While DHFS does not oppose the principle of specific assignment of those costs physically traceable to meals eaten by nursing home or Royalton residents, DHFS claims that St. Francis gave equivocal evidence and documentation on these tracings and assignments. On that basis, DHFS wants St. Francis to revert to the 60%-40% allocation for such costs. The trial court rejected this argument, and we agree with trial court. St. Francis management gave direct, straightforward testimony that the nursing home was the exclusive user of the packaged milk, packaged condiments, and specified paper products. According to that testimony, the nursing home used these products to help stop the spread of disease, a greater risk to nursing home residents than apartment residents. DHFS’s attack on this testimony is little more than selective parsing of statements by various witnesses. We are satisfied that their testimony, when viewed as a whole, supports the specific identification and assignment of these costs to the nursing home.

In the same vein, DHFS disputes St. Francis’ corresponding claim that St. Francis specifically assigns to Royalton the costs of all bulk milk, bulk condiments, and bulk paper used by Royalton. DHFS suspects that St. Francis is assigning these bulk costs to Royalton in name only, furtively burying them in the other allocable costs. DHFS believes that St. Francis underreports these bulk costs as costs specifically assigned to Royalton, instead routing them back through its cost allocation system to the nursing home as part of the 60%-40% allocable costs. In other words, DHFS suspects that St. Francis actually runs 60% of these bulk costs back to St. Francis on the basis of the meals ratio. DHFS points out that the nursing home and Royalton specific cost assignments depart radically from the 60%-40% meals ratio; St. Francis specifically assigned $100,008.03 in food costs to the nursing home and only $9,460.26 in such costs to Royalton, a 91%-9% split. DHFS apparently believes that specifically assigned food costs should track the meals ratio more or less, except for the extra costs to the nursing home from the packaging itself. The administrative law judge seems to have adopted this view. DHFS has pointed to no substantial evidence for this analysis. St. Francis management maintained unequivocally that St. Francis directly assigned 100% of the above-cited bulk costs to Royalton. DHFS cites inconsistencies in management’s testimony and the lack of hard documentation to support the costs specifically assigned. DHFS has not shown, however, that it asked St. Francis for such documentation, and the record has nothing undermining the basic persuasiveness of management’s testimony.[2]

DHFS next challenges St. Francis’ specific identification and assignment of so-called “nourishment” costs to the nursing home.
This assignment rested on physical tracing. “Nourishments” are special foods nursing home residents consume at bedtime, something unique to the particular health needs of nursing home residents. These include liquid food supplements, some puree foods, some “sculptured foods,” and cookies. According to DHFS, these “nourishments” act merely as deferred meals, food that differs in timing, not substance, eaten at bedtime rather than mealtime. Nursing home residents consume the “nourishments” in part to counteract the fact that their frail health sometimes impedes them from eating adequately during their meals. DHFS views the “nourishments” as something akin to “seconds” at mealtime and therefore disallowed their specific assignment as food costs, reverting to the 60%-40% allocation used for other food costs. Implicit in this argument is the premise that the Royalton kitchen prepares smaller meals for St. Francis residents than it does for Royalton residents. In other words, the Royalton kitchen (1) adjusts meal size for the undereating by St. Francis residents by serving those residents smaller meals by unit measure than it serves Royalton residents, and then (2) makes up the difference to the underserved St. Francis residents with “nourishments” at bedtime. Under DHFS’ theory, the “nourishments” are just part of the common food costs and should follow the same 60%-40% allocation as other food costs.

We agree with the trial court that DHFS had no substantial basis for disallowing the “nourishments” specific assignment. St. Francis officials testified unequivocally that nursing home residents were the exclusive consumers of the “nourishments” and that the “nourishments” had no analog in Royalton. According to this testimony, nursing home residents consumed the “nourishments” as a distinct supplement to their meals to meet their special needs. Contrary to DHFS’s assertions, the record has no substantial evidence that the “nourishments” acted as a deferred part of the meals, a substitute for the “seconds” available to Royalton residents in the cafeteria lines. Likewise, the record has no substantial evidence that the Royalton kitchen serves smaller amounts of food to St. Francis residents at mealtime than it serves Royalton residents. Rather, the record suggests that the kitchen generally serves the same meal size by unit measure to both St. Francis and Royalton residents and that St. Francis residents simply leave more food uneaten at mealtime than Royalton residents. Viewed in this light, the “nourishments” represents food costs above and beyond the ordinary meal costs, and St. Francis rightly assigned them to the nursing home above and beyond the meals-ratio food cost allocation. In short, DHFS had no substantial basis for disallowing the specific assignment of the “nourishment” food costs.

DHFS next challenges St. Francis’ method for allocating labor costs of three Royalton kitchen employees. Specifically, DHFS attacks the means St. Francis used to measure how these three employees spent their time. Once each year in November, over a one- or two-day time frame, the three employees kept track of how much time they worked in various aspects of meal processing (1) for the nursing home residents and (2) for the Royalton residents. They then reported that to management. On the basis of this self-reported time worked, St. Francis allocated the three employees’ yearly labor costs between the nursing home and Royalton. DHFS disallowed the cost allocation and reverted to a meals-ratio allocation. DHFS claims that one- or two-day self-reporting supplies an untrustworthy measure of time worked throughout the year with its seasonal variations. According to DHFS, St. Francis should have measured work times on a periodic basis, such as monthly, and then averaged the results in a statistically valid way. DHFS notes that Medicare rules require more extensive time studies. Although DHFS has shown that the St. Francis method was imperfect—St. Francis did not conduct a rigorous industrial-engineering time study—DHFS has not shown that the method misstated actual time worked in a material respect. This was a practical, cost-effective method for the small number of employees involved, and DHFS has not shown that its 60%-40% meals-ratio allocation more accurately represented time worked.[3] In short, DHFS had no substantial basis for disallowing the labor cost allocations in full.

DHFS next challenges St. Francis’ method for allocating electricity costs. St. Francis first allocated the electricity between the Royalton kitchen and Royalton apartments. St. Francis then allocated 60% of the kitchen’s electricity use to the nursing home on the basis of the 60%-40% meals ratio. St. Francis did not have separate meters for the kitchen and apartments. Instead, it used two power company studies as the basis for its allocation. The power company conducted the first study in 1979 and the second in June 1995, after the 1992-93 cost year. DHFS claims that these industrial-engineering studies are unreliable for several reasons. First, St. Francis no longer has any documentation for the 1979 study and could not explain its methodology. In addition, according to DHFS, the 1979 measurement was too remote in time to the 1992-93 cost year to reliably support the allocation, in light of likely and unknown changes in occupancy, consumption, climate, and kitchenette use in individual apartments since 1979. Second, the power company conducted the 1995 study over no more than five days, possibly over only one day, in June 1995. DHFS states that this is not a valid measurement time frame, claiming that usage over one or five days in June 1995 does not accurately represent typical usage over a twelve-month period with seasonal variations. DHFS believes that St. Francis should have gathered more data over a longer-term time frame and then averaged them in a statistically valid way. Last, DHFS points out that St. Francis installed a “heater line” after the 1992-93 cost year but before the 1995 power company study, and that this could skew the 1995 measurement.

The trial court rejected these arguments, and we agree with the trial court’s decision. DHFS seeks to impose a square-footage allocation, in lieu of the allocation based on the 1979 and 1995 power company studies. Under the square-footage method, DHFS would allocate Royalton’s electricity usage between the kitchen and the apartments on the basis of those areas’ respective square footage. DHFS would then allocate 60% of the kitchen’s share of Royalton’s electricity to St. Francis, on the basis of the 60%-40% meals ratio. Both parties agree that a square-footage allocation does not represent actual electricity usage in Royalton. No one argues that the apartments use the same amount of electricity per square foot as the kitchen, whose equipment consumes large amounts of power. In other words, the square-footage allocation wrongly assumes that electricity is a fixed overhead cost that does not vary with the level of meal production.[4] Nonetheless, DHFS imposed that allocation as a default method, on the ground that the St. Francis industrial-engineering allocation rests on too many assumptions and too many unknown variables. We agree with the trial court that this did not merit DHFS’s square-footage method. Although DHFS has shown that the St. Francis method was imperfect, DHFS has not shown that the St. Francis method materially misallocated electricity between the kitchen and the apartments. On the contrary, the St. Francis allocation treated electricity as a variable overhead cost, and this more closely tracked actual variable electricity usage than DHFS’s square-footage fixed overhead allocation. Under the circumstances, DHFS had no substantial basis to disallow St. Francis’ allocation in full.

DHFS next challenges the St. Francis method for allocating gas, water, and sewer charges between the Royalton kitchen and apartments. St. Francis did not have separate meters for those areas for these utilities, and it had no evidence at the time of the DHFS review to support the allocation. After DHFS’ review, however, St. Francis obtained new evidence for the allocation. St. Francis compared these utilities’ usage for the 1976-77 time frame with the 1994-95 time frame. St. Francis remodeled the kitchen in 1978 to increase its meal-production capacity. The 1976-77 figures depict usage before the 1978 kitchen remodeling; the 1994-95 figures depict post-remodeling usage. This allocation rested on historical account inspection. St. Francis sought to correlate the increased utility usage to the increased post-remodeling meal production as a linear variable overhead cost with no fixed cost component. DHFS claims that too many variables besides production levels could skew this correlation and that the time frames used were too far apart for a systematic correlation. DHFS imposed a square-footage allocation in lieu of St. Francis’ method. Both parties agree, however, that a square-footage allocation deviates from actual usage; it wrongly treats the utility charges as fixed overhead costs that do not vary with meal production. DHFS has again shown that the St. Francis method was imperfect. For example, St. Francis used no data from other years to help support its linear correlation theory, and its method does not account for fixed aspects of the utility costs. Nonetheless, DHFS has not shown that the St. Francis method materially misallocated kitchen utility costs, and DHFS’ own square-footage allocation had major shortcomings. Under the circumstances, DHFS did not have a substantial basis to disallow the St. Francis allocation in full.

We have rejected DHFS’s blanket disallowance of St. Francis’ labor and utility cost allocations. We believe, however, the DHFS has the power to work partial cost disallowances under the unique circumstances of this case. As noted above, DHFS may correct errors in calculations without disallowing the entire cost item. See § 49.45(3)(f)2, Stats. In addition, other courts have recognized the power of state agencies to work partial disallowances of reported Medicaid costs. See Indiana Dept. of Public Welfare v. Crescent Manor, Inc., 416 N.E.2d 470, 474 (Ind. App. 1981); Maryland Dept. of Health v. Riverview Nursing Centre, Inc., 657 A.2d 372, 377-79 (Md. App. 1995). Here, the record has substantial evidence that St. Francis’ allocations of labor and utility costs carry a significant level of estimate and a discernible risk of error. For example, the labor allocation stood on a two-day, self-reported time study, not more extensive time studies by an internal auditor or industrial engineer. Likewise, the utility allocations used methods affected by unknown variables and failed to account for some fixed components of utility costs. Although DHFS’s meals-ratio and square-footage allocations fail to fairly allocate these costs, DHFS may be able to work other downward adjustments to offset the risk of error imbedded in the unknown variables. In other words, DHFS may write down St. Francis’ labor and utility cost allocations by an “uncertainty discount” to protect taxpayers from the nursing home’s inability to supply cost data with a high level of confidence.[5] As a result, we remand the cause to DHFS for it to make objective and reasonable partial disallowances of St. Francis’ labor and utility cost allocations, short of straight meals-ratio and square-footage allocations, to offset the risk of error in St. Francis’ methods.

By the Court.—Order affirmed in part and reversed in part; cause remanded to DHFS for proceedings consistent with this opinion; no costs to either party.

This opinion will not be published. See Rule 809.23(1)(b)5, Stats.

[1] Except for multiple regression, the historical account inspection methods make two assumptions: (1) cost behavior is linear; and (2) one independent variable explains the behavior. See Horngren at 807. Multiple regression uses more than one independent variable. See id. at 831. Most often, some measure of production volume is the independent variable or activity base. Except for those costs that can be physically traced to a particular product, however, all cost allocations involve a degree of estimate and approximation. See Dopuch at 32-33. This applies not only to the simpler methods, but also to statistical sampling and regression analysis, which are subject to misuse if applied indiscriminately. See Horngren at 807, 814-21; Dopuch at 62-81. For that reason, some adjustment is sometimes needed, and only a case-by-case analysis can answer whether such simplifications give sufficiently accurate approximations. See Horngren at 252, 807. In the end, all cost estimates, even highly precise statistical regression techniques and industrial-engineering studies, need adjustments and allowances on the basis of personal judgment. See Davidson & Weil, Handbook of Modern Accounting 39:22 (2d ed. 1977). This appeal involves each of the above-cited approaches: (1) physical tracing cost assignments; (2) industrial-engineering cost allocations; and (3) historical account inspection cost allocations.

[2] Other courts have taken this view in parallel cost-reporting contexts. In Merck, Sharp & Dohme Int’l v. United States, 915 F. Supp. 405 (Ct. Int’l Trade 1996), the United States Court of International Trade faced similar circumstances. In examining a customs question, the court ruled that Merck did not need to produce source documentation to prove the accuracy of its cost-of-products figures. Id. at 408. The court noted that Merck’s proof need not be perfect and that Merck need not prove its cost of production in a rigid way. Id. The court also looked at whether the government had given a workable alternative to Merck’s cost methods. Id. at 411. The court refused to permit the government to use an unworkable default method in lieu of Merck’s cost data. Id. at 412.

[3] We doubt that the Methods’ reference to Medicare rules requires nursing homes to use the more extravagant kinds of Medicare-compliant time studies when the studies’ costs exceed their benefits. Here, the small number of employees weighs against costly, repetitive and more rigorous kinds of Medicare-compliant time studies.

[4] Variable overhead costs vary with the level of production; fixed overhead costs do not. See Horngren at 23-24.

[5] Security analysts recognize that cost reporting has a large element of estimate and assumption. For that reason, they often adjust such figures to compensate for their imperfections. See Cottle, Murray & Block, Security Analysis 145 (5th ed. 1988). Accounting has a variety of imperfections; it is full of judgmental decisions and is sometimes highly subjective. See id. Someone must determine allocations of costs, and “it is difficult to prove whether decisions are right or wrong.” See id. As a result, security analysts often reduce reported amounts that management may have unconsciously given an upward bias. See id.; see also Altman, ed., Financial Handbook 25:13-25:14 (5th ed. 1981); Handbook of Modern Accounting at 2:4-2:5.