- Email: [email protected]
Utilization Management: A Literature Review for Canadian Health Care Administrators
FORUM
Spring/Printemps 1991
Original Article
Article original
Utilization management ( U M ) , the attempt to measure, understand and reduce inappropriate hospital use, has been in development for over 20 years. It is an outgrowth of two related phenomena: (1)the increasing responsibility of large institutional third party payers for health care costs and the increasing demand of those payers for accountability; and (2) in Canada, particularly, the debate surrounding the adequacy of hospital funding and the perceived inadequacy of cost control using global budgeting, Given the interest in U M , hospital administrators, provincial and federal associations representing hospitals, hospital employees andphysicians wouldfind a review of UMprograms useful in terms of what is known about their effectiveness,and the specific initiatives in Canada. The authors underscore the critical need for formal evaluation of UM programs; to date there has been little systematic research into issues related to its implementation and impact. This issue is particularly pertinent because UM programs have not been widely implemented in Canada. La gestion de l’utilisation (GU ou UM), c’est-d-dire la tentative de mesurer, de comprendre et de rtduire les utilisations hospitalidrespeu approprites, se dtveloppe depuis plus de 20 ans. Elle est issue de deux phtnomdnes lits :(1)responsabilitt croissante de tiers comme bailleurs defonds institutionnels en services de santt et volontt grandissante de ces intervenants de sefaire rendre des comptes; ( 2 ) au Canada en particulier, dtbat sur le caractdre suflsant du financement des hdpitaux et l’insuffisance percue du contrdle des coats reposant sur des budgets globaux. fitant donnt l’intkr2t manifestt pour la GU, les administrateurs hospitaliers, les associations provinciales et ftdtrales reprtsentant des hdpitaux et le personnel et les mkdecins en milieu hospitalier jugeraient utile un examen desprogrammes GU dupoint de vue de ce que l’on connait de leur efficacittet des initiatives prtcises prises au Canada. Les auteurs insistent sur le besoin critique dune tvaluation structurte de ces programmes. Jusqu’d prksent, il y a eu peu de recherche systkmatique sur les questions d’application et d‘incidence qui s’y rapportent. Cet examen est particulidrement utile parce que les probldmes GU ne sontpas rkpandus au Canada.
Utilization Management: A Literature Review for Canadian Health Care Administrators by Sam B. Sheps, Geoff Anderson and Karen Cardiff
T
he cost of general and allied special hospitals as a percentage of the Gross National Product has increased significantly since 1978.’ At the same time, public expectations of the level and quality of care provided by hospitals have also increased? Hospitals are faced with the difficult task of ensuring quality of care while attempting to contain cost increases. Moreover, it has long been believed that a proportion of the increase in hospital costs is the result of inappropriate or unnecessary uti1izationPS In this environment, it is not surprising that there has been increased interest in and discussion of hospital-based utilization management (UM) in Canada. Payne6defined UM as “the deliberate action by hospital administrators or payers M influence providers of hospital services to increase the efficiencyand effectivenessof services provided”. @. 712) UM consists of two components: (1) comparing performance to explicit standards of appropriate care; and (2) developing initiatives to rectify identified problems? The component of UM that involves measurement of performance and
FORUM
Spring/Printemps 1991
comparison to standards is referred to as utilization review (UR): the application of defined criteria or expert opinion, or both, to the observed process of care in order to assess the efficiency of that process, and to determine the appropriateness of decision making related to the site, duration and frequency of care! The second component of UM, the response to identified problems, consists of techniques for modifying behaviour of providers, patients or administrators, either directly or indirectly, and relates not only to services provided in the hospital setting, but to the co-ordination of hospital services with community services. U R may be conducted before (prospectivereview), during (concurrent review) or after (retrospectivereview) the patient's stay. It is important to note the dual objectives of UM of achieving cost containment and quality of care; thus, UM is related to hospital risk management and, particularly, quality assurance (QA) programs. Although there is rhetorical overlap between the activities of UM and QA programs, UM usually focuses on the process of medical care while QA includes the review of the structure and process as they relate to outcome of care: the sequelae of treatment or patient satisfaction, or both. The objective of this article is to provide Canadian hospital administratorswith a brief review of the literature on hospitalbased UM to help them to understand the basic concepts involved, the types of programs that have been developed and their impact. Although the focus is on applications for Canadian hospitals, much of the literature is from the United States. The review is therefore divided into two sections. The first deals with U.S. initiatives and the second with UM programs in Canada.
Utilization review in the U.S. For three decades in the U.S., federal and state authorities have implemented many regulatory mechanisms to reduce over-utilization,to control costs and to ensure quality of care.8 In 1966, legislation requiring retrospective review for hospitals participating in Medicaid and Medicare was introduced and U R activities proliferated? It is noteworthy that the various review activities that grew out of this legislation developed in spite of there being little evidence that the activities actually worked." Hospitals across the nation used retrospectively defined length of stay as the major endpoint because (1) it was a measure objectively and easily attained;' (2) no clinical standard existed that defined inappropriate use as opposed to appropriate use;5 (3) it was thought to be an important measure of hospital performance and resource consumption; and (4) it was thought to be one of the most important variables contributing to hospital costs.12
'
However, by the mid-l970s, it was generally acknowledged that this approach to retrospective review had not been effective3 because utilization problems were identified after the patient had left the hospital and appropriate norms for evaluating medical necessity and appropriateness of care were lacking.13 In recent years, the focus in retrospective review measurement has been on the development of tools that explicitly incorporate clinical characteristics of patients in the process used to identify prolonged lengths of stay. One of the first such techniques was based on Diagnostic Related Groups (DRGS).'~In essence, the DRG system was designed to classify patients in acute care hospitals into a manageable number of groups that, on the basis of clinical characteristics, should have similar lengths of stay.'4"5 Although the use of DRGs gained some acceptance, several problems were noted: (a) they only explained a fraction of the variance in length of stay;16 (b) there were methodological problems in their construction and the classification of disease~;'~-''and c studies illustrated significant human error in coding DRGs.26. Because it was felt that clinical severity might conmbute to the large variation observed with DRG categories, tools that measured the clinical severity were developed. Studies have indicated these severity adjustment tools, such as Severity of 111ness Index (SII)22or MedisGro~ps?~ can explain some of the variance in length of stay within DRG categories.24-27 U R based on length of stay may provide a reasonable screening tool for identifying resource use that varies from standard or expected patterns of care but it is not an accurate way to identify inappropriate care!> 28 Indeed, review methods based on length of stay analysis may encourage inappropriate short stays and penalize patients who truly require longer stays?' In response to the apparent failure of retrospective review either to enhance quality of care or control costs, Professional Services Review Organizations (PSROs) expanded the use of admission and concurrent review processes in 1972. Under the PSRO mandate, all Medicaire and Medicaid admissions were reviewed to ensure that medical care in a hospital setting was necessary. In general, the criteria used to determine the appropriateness of admissions could be either criteria specific to a particular problem, diagnosis or procedure and criteria that specify the types of services that should be provided at a hospital level of care. A formal evaluation of these two components of the review process was done in 1977 and showed that the impact on utilization and health care expenditures was minimal?' By the early 1980s it had become quite evident that utilization review was not very effective at containing costs and, in 1983, the federal government enacted legislation under which reimbursement from Medicare was determined prospectively
El
a,
'
FORUM
based on the patient’s diagnosis and characteristics, instead of on the actual costs incurred in caring for the patient. This change in reimbursement made UM a vital tool for maintaining the financial viability of the hospital. Studies have indicated that the average length of stay declined by 5.4% in 1984 and that total hospital costs were approximately $2 billion less than Medicare officials had expected to pay in 1984?l More recent methods for U R use a detailed and standardized concurrent review process with predetermined, specific, objective,consistent and explicit criteria!. 32 The explicit criteria exhibit two properties that may make them acceptable to physicians: (1) they are based on clinical data and (2) each patient is assessed individually against the criteria. In contrast, length of stay from the physician’s perspective is an arbitrary measure of resource consumption,particularly because individual patients are compared to a group average.33 The three concurrent review tools most commonly discussed in the literature are: (1) AppropriatenessEvaluation Protocol (AEP) developed by a group of researchers at the Boston University School of Medicine during the 1 9 7 0 ~(2) ; ~Inten~ sity of Service, Severity of Illness and Discharge Screens (ISDA) developed by InterQual;22and (3) Standardized Medreview Instrument (SMI) developed by Sy~teMetrics.3~ Having identified an inappropriate admission or hospital day by the explicit clinical criteria, the cause may be assigned either to the hospital,the physician, the family or the environment.‘ This is particularly important because it not only promotes a broader understandingof how the various components of the health care system affect utilization but, at the same time, eliminates the tendency to focus solely on physician behaviour, fostering a more relevant set of approaches to the problems of health care, thus utilization. One study36showed that the AEP and ISD were found to be moderately valid compared to physician opinion, but that the SMI had low validity. Despite the long history of utilization analysis in the U.S., there has been astonishingly little solid e~aluation.3~ Research by Wickizer, Wheeler and Feldstein3*suggested that U R in the U.S. has contributed to cost containment and the efficient use of health care resources. However, they noted that the success of UR is linked to the incentives or disincentives associated with the review activities: “the stringency with which UR guidelines are applied and the level of penalties invoked for noncompliance,as well as other factors specific to the UR program, are likely to influence outcomes”. @. 646) In particular, as noted previously, the demonstratedeffectiveness of U R was essentially nonexistent until payment was directly and prospectively linked to UR.
Spring/Printemps 1991
Utilization review in Canada For 20 years the legislative and financial environment faced by Canadian hospital administrators has been much different from that in the U.S. However, there has been growing pressure on hospitals and increasing interest in UM. Although there is very little published material that provides descriptions or evaluations of UM programs in Canada, some specific initiatives have been undertaken.
The Value Improvement Process (VIP)The Value ImprovementProcess is a retrospective review program that was developed by the Baxter Corporation in 1984.3’ The process involves the selection of 40 to 50 patient records that fit preselected diagnostic criteria and the collection of information about the resources used for these patients. A profile of the costs associated with treating the particular type of patient is determined and compared to cost profiles of similar patients collected by Baxter from other hospitals. The review process then shifts from a determination of costs to an evaluation of practices followed by the development of recommendations. The process has been documented as improving quality of care4” 41 while reducing length of stay and saving money.42 However, the studies that evaluate VIP have been “before and after” studies with no use of control hospitals. The use of these designs raises the possibility that the observed changes in utilization were the result of factors other than the VIP process. Although VIP may decrease length of stay and costs, it is associated with a relatively limited range of procedures, in particular, total hip replacements, cataract surgery and treatment of acute myocardial infarction. This is notable because hospital admission and discharge data indicate that these procedures account for a very small proportion of all hospital admissions and di~charges.4~ Thus, the impact on overall hospital utilization, at least regarding what we knbw about VIP, might be small. This is, of course, not to say that VIP is not valid for the purposes for which it was intended; namely, to reduce costs as much as possible for specific procedures.
Resource Enhancement Program (REP) REP is a retrospective review program developed by the Health Management Resource Gr0up.4~Its goals are simultaneously to improve utilization of hospital resources and maintain or improve quality of care for patients. REP provides health care managers with a profile of the costs associated with treating patients within specific diagnostic groups (e.g. acute
Spring/Printemps 1991
FORUM
control hospitals. Similarly,small-scale evaluations of VIP have been done but they have been based on demonstrating efficacy for specific procedures instead of assessing the effectiveness at a population level (i.e. does the implementationof VIP significantly reduce length of stay or some other variable when applied to the normal operations of the hospital?). The relative lack of UM development and research in Canada compared to the U.S. can be explained by two major SWITCH is an acronym given to a concurrent review profactors. First, hospital costs have not risen as rapidly in Canada gram developed by the Peace Arch District Hospital in 1984, as in the U.S. due to the central control of expenditures and the and stands for Signs, Winds, Intramuscular Injections, Tubes, use of global budgets;” therefore, there is less pressure to use Consultant (Active) and Hospice. These criteria have since UM. Second, UM, as must be acknowledgedby administrators, been patented by MEDICUS!’ The SWITCH Index System involves modifying provider behaviour and, in the current uses six categories of explicit criteria to identify patients who Canadian context, this would entail a major change in the require acute care hospital services, either for diagnostic acrelationship between administrators and medical staff. Altivity or for treatment. Patients are monitored daily and are though managing change is never easy, it can be doubly difcategorized as “on-index”or “off-index”for that day, dependficult if the need for change emerges from outside the hospital ing on whether they meet at least one of the SWITCH criteria. environment (e.g. government pressure) and involves altering The system also identifies causes for delays in care in terms of fundamental relationships. Moreover, after having accepted the external services, internal services and physicians. Wiggins” need for change, the change process is often invested with an evaluated the SWITCH system in 1988; although there was aura of known effectiveness and implied success that make forevidence supporting a reduction in length of stay in a hospital mal evaluation a potentially hazardous undertaking because the using the SWITCH Index System, the reduction was not sigempirical evidence may not support the aura. Thus, issues of nificantly different from that observed in control hospitals over control over the evaluation become central in managethe same time period. ment/professionalinteractions,adding to the likelihood of conflict. The lack of evidence on the impact of UM has serious consequences for institutions,those who work in them and those Although HMRI has not developed any specific UM who manage them. The consequences are amply demonstrated programs, it is included in this section as an approach to colin our brief review of the early American experience with UM: lecting data on utilization. The goals of HMRI are to “collect, repeated wide-scale application of interventions for controlling analyze and report standardized comparative information on length of stay and expenditures that were never evaluated patient care provided by Canadian health care instit~tions”.~~ properly before implementationand that consistently failed to (p. 13) To accomplish this, HMRI has developed case mix achieve the desired results. Such failure not only casts doubt on groups (CMGs) that are similar to DRGs and provide CMG the whole approach but makes key groups, particularly length of stay comparison to hospitals. As an extension physicians, cynical and wary. The PPS was successful only beI-IMRI~~has recently developed Resource Intensity Weights cause as an intervention it directly influenced the key outcome, (RIWs) that compare length of stay and charge or cost-based not because it was evaluated and found to work on a small weights across CMGs. Similar to the DRG-based Prospective scale and then applied on a wide scale. Payment System (PPS) developed in the U.S., researchers at The importance of UM and subsequent interest in its use has HMRI contend that RIWs could assist in analyzing hospital not been accompaniedby systematic research into issues recase mix and perhaps in determining global budget adjustments lated to its implementation and impact. There are few sound in Canadian h0spitals.4~ studies available from the U.S. and it is difficult to generalize their findings to the Canadian setting. Thus, the message of this review is that, despite the inherent difficulty of evaluating UM strategies on a small scale, it is critical that it be done to avoid Aside from a few general reviews and 49 virtually the inevitable problems that will occur in a wider implementano empirical data exist to demonstrate the effectiveness (or tion. UM is not simply a tactic for providing technical solunot) of UM in Canada. As noted before, Wiggins” found that tions to management problems; it represents a significant the SWITCH Index System as implemented in a community change in organizational relations, which provides both opporhospital did not significantly reduce length of stay compared to myocardial infarction). Similar to VIP, participating hospital managements can compare the costs associated with treating patients in their hospital to a database of cost profiles provided by REP, helping them to identify where costs can be reduced. There has been no formal evaluation of REP.
The SWITCH Index System
The Hospital Medical Records Institute (HMRI)
Discussion
FORUM
tunities and risk. If viewed positively, with the clear objective of improving patient access to care through greater efficiency in bed resource use, and improving patient care through assessing quality of care, UM can be a powerful tool for forging better administration and professional relations. Given the growing interest in UM, we believe it is imperative that evaluation of UM programs begins to occupy a more prominent position in health services research.
References and notes 1. Evans, R.G. 1984. Strained Mercy: The Economics of Canadian Health Care, Toronto: Butterworth and Co. (Canada) Ltd. 2. Chenoy, N.C. 1984. Strategic planning-understanding and responding to a rapidly changing world. Health Management Forum 5(2): 3-19. 3. Dans, P.E., Weiner, J.P. and Otter, S.E. 1985. Peer review organizations: promises and potential pitfalls. New England Journal of Medicine 3 13(18): 1131-1137. 4. Hughes, R.A., Gertman, P.M., Anderson, J.J. et al. 1984. The ancillary services review program in Massachusetts. Journal of the American Medical Association 252(13): 1727-1732. 5. Goran, M.J. 1979. The evolution of the PSRO hospital review system. Medical Care (suppl) 17(5): 1-46. 6. Payne, S.M.C. 1987. Identifying and managing inappropriate hospital utilization: a policy synthesis. Health Services Research 22(5): 709-769. 7. Anderson, G., and Lomas, J. 1988.Development of UtilizationAnalysis: How, Why and Where It’s Going, Vancouver, B.C.: Health Policy Research Unit, University of British Columbia, 88(7), with comment by Barer, M.L., You Can Take the System to Its Analyst, But Should You Let the Analyst Loose on the System. 8. Brown, L.D. 1986. Introduction to a decade of transition. Journal of Health Politics and Policy Law 11 (4): 569-583. 9. McDonough, Catherine and Vaz, Andrew P. 1987. Hospital utilization management. Health Management Forum 8(1): 42-52. 10. Wiggins, Sandra. 1988. Utilization Management of Acute Care Services: Evaluation of the SWITCH Index System. M.Sc. diss., University of British Columbia, Vancouver, B.C. 11. Goldberg, G.A. and Holloway, D.C. 1975. Emphasizing “level of care” over “length of stay” in hospital utilization review. Medical Care 13(6): 474-485.
Spring/Printemps 1991
12. Lave, Judith, R. and Leinhardt, Samuel. 1976. An evaluation of a hospital stay regulatory mechanism. American Journal of Public Health 66(10): 957-967. 13. Flashner, B.A., Reed, S . , Coburn, R.W. et al. 1973. Professional Standards Review Organizations: analysis of their development and implementation based on a preliminary review of the hospital admission and surveillanceprogram in Illinois. Journal of the American Medical Association 223 (13): 1473-1484. 14.Fetter, R.B., Shin, Y.L., Freeman, J.L. et al. 1980. Case mix definition by diagnosis-related groups. Medical Care (Suppl) 18(2): 1-53. 15. Grimaldi, P.L. 1983. New medicare DRG payment calculation issues. Nursing Management 14(11): 19-23. 16. MacKenzie, T.A., Markle, F. and Croke, M. 1987. CMG’s: variations on a theme. Health Management Forum 8( 1): 21-24. 17. Mullin, R.L. 1985. Diagnosis-related groups and severity: ICD-9-CM, the real problem. Journal of the American Medical Association 254(9): 1208-1210. 18. Smits, H.L., Fetter, R.B. and McMahon, L.F. November 1984. Variation in resource use within diagnosis-related groups: the severity issue. Health Care Financing Review (annual suppl): 71-78. 19. Mendenhall, S . 1984. DRGs must be changed to take patient’s illness severity into account. Modern Healthcare 14(15): 86,88. 20. Gertman, P.M. and Lowenstein, S . November 1984. A research paradigm for severity of illness: issues for diagnosis-relatedgroups system. Health Care Financing Review (annual suppl): 79-80. 21. Johnson, A.N. and Appel, G.L. 1984. DRGs and hospital care records: implications for medicare case mix accuracy. Inquiry 21: 128-134. 22. InterQual. 1978. The I S k A R e v i e w System with Adult Criteria. InterQual Incorporated, Westborough, Ma. 23. MediQual, MedisGroups. MediQual Systems Inc., 1900 West Park Drive, Westborough, Mass. 24. Horn, S.D., Horn, R.A. and Sharkey, P.D. November 1984. The severity of illness index as a severity adjustment to diagnosis-relatedgroups. Health Care Financing Review (annual suppl): 33-45. 25. Horn, S.U., Sharkey, P.D., Chambers, A.F. et al. 1985. Severity of illness within DRGs: impact on prospective payment. American Journal of Public Health 75( 10): 1195-1199.
Spring/Printemps 1991
26. Jencks, S.F. and Dobson, A. 1987. Refining case-mix adjustment: the research evidence. New England Journal of Medicine 3 17(11): 679-686. 27. Iezzoni, L.I., Ash, A.A., Cobb, J.L. et al. 1988. Admission MedisGroups score and the cost of hospitalizations. Medical Care 26(11): 1068-1080. 28. Owens, R.R. and Averill, R.F. October 1984. The role of utilization management under PPS. Healthcare Financial Management: 60-62,64,66. 29. Kemper, K.J. 1988. Medically inappropriate hospital use in a pediatric population. New England Journal of Medicine 388(16): 1033-1037. 30. Dobson, A., Greer, J.G., Carlson, R.H. et al. 1978. PSROs: their current status and their impact to date. Inquiry 15: 113-128. 31. Kelly, J. and Bankhead, C.D. 1985. DRGs: how are they stacking up? Medical World News 26(5): 80-103. 32. Clemenhagen, C., Champagne, F., Contandriopoulos, A. et al. 1985. Hospital-Based Quality Assurance: A Starting Point for Physicians and Managers, Ottawa, Ont.: Canadian Hospital Association. 33. McIlrath, S. January 11, 1985. Review system based on illness severity. American Medical News. 34. Boston University School of Medicine. Appropriateness Evaluation Protocol (AEP), Health Care Research Unit, Boston University School of Medicine, Suite 1102,720 Harrison Avenue, Boston, Mass. 02118. 35. SysteMetrics. 1983. Standardized Medreview Instrument (SMI). SysteMetrics Incorporated, 104 West Anapamu Street, Santa Barbara, Calif. 93101. 36. Strumwasser, I., Paranjpe, N.V., Ronis, D.L. et al. 1990. Reliability and validity of utilization review criteria: appropriateness evaluation protocol, standardized . medreview instrument, and intensity-severity-discharge criteria. Medical Care 28(2): 95- 111. 37.Feldstein, P.J., Wickizer, R.M. and Wheeler, J.R.C. 1988. Private cost containment: the effects of utilization review programs on health care use and expenditures.New EngIand Journal of Medicine 318 (20): 1310-1314. 38. Wickizer, T.M., Wheeler, John, R.C. and Feldstein, Paul, J. 1989. Does utilization review reduce unncessary hospital care and contain costs? Medical Care 27(6): 623-647. 39. Baxter Corporation, Value Improvement Process. Baxter Corporation, 2930 Argentia Road, Missisauga, Ont.
FORUM
40. Winchell, J.R. October 1985. Reduced costs can improve care. Health Care. 41. Coombs, R. 1988. Implementing a value improvement process. Healthcare Management Forum 1 (2): 14-16. 42. Eliasoph, H. and Hassen, P. 1986. VIP streamlines care and reduces length of stay. Dimensions 63 (7): 42-53. 43. Hospital admission and discharge data (1987/88) for British Columbia indicate that cholecystectomiesaccounted for only 1.1% of all hospital admissions, total hip replacements accounted for .15% and acute myocardial infarction accounted for 1.2%. 44. Ryan, Patricia and De Paoli, Lisa. 1989. Comparing the Costs of Managing Patients with Acute Myocardial Infarction in Four B.C. Hospitals. Vancouver, B.C.: Health Management Resource Group. 45. MEDICUS, The MEDICUS UM Program “SWITCH’. 1989. MEDICUS Canada Incorporated, Toronto, Ont. 46. Ontario Hospital Association. 1988. Guidefor Hospital Utilization Review and Management in Ontario. Don Mills, Ontario: Ontario Hospital Association. 47. Hospital Medical Records Institute. 1988. Case Weighting Methods: A Review by the HMRI Database Committee. Ottawa, Ontario: Don Carmichael and Company. 48. Rachlis, M., and Fooks, C. 1988. Utilization Analysis: Current Initiatives Across Canada. Conference Paper. Conference Proceedings, First Annual Health Policy Conference, McMaster University, Hamilton, Ont. 49. Suttie, B., Helliwell, B., Villenure, D. 1988. Some Aspects of Utilization Management in Twenty Ontario Hospitals, Waterloo, Ontario: Centre for Applied Health Research, University of Waterloo. 50. Evans, R.G., Lomas, J., Barer, M.L. et al. 1989. Controlling health expenditures-the Canadian reality. New England Journal of Medicine 320 (9): 571-577. Sam B. Sheps, MD, MSc, FRCPC, is Associate Professor and Head, Department of Health Care and Epidemiology, University of British Columbia. Geoff Anderson, MD, PhD, is Associate Director, Health Policy Research Unit, and Assistant Professor, Department of Health Care and Epidemiology, University of British Columbia. Karen Cardiff,BScN, BSc, MHSc, is Research Associate, Division of Health Services Research and Development, University of British Columbia.
Spring/Printemps 1991
Original Article
Article original
Utilization management ( U M ) , the attempt to measure, understand and reduce inappropriate hospital use, has been in development for over 20 years. It is an outgrowth of two related phenomena: (1)the increasing responsibility of large institutional third party payers for health care costs and the increasing demand of those payers for accountability; and (2) in Canada, particularly, the debate surrounding the adequacy of hospital funding and the perceived inadequacy of cost control using global budgeting, Given the interest in U M , hospital administrators, provincial and federal associations representing hospitals, hospital employees andphysicians wouldfind a review of UMprograms useful in terms of what is known about their effectiveness,and the specific initiatives in Canada. The authors underscore the critical need for formal evaluation of UM programs; to date there has been little systematic research into issues related to its implementation and impact. This issue is particularly pertinent because UM programs have not been widely implemented in Canada. La gestion de l’utilisation (GU ou UM), c’est-d-dire la tentative de mesurer, de comprendre et de rtduire les utilisations hospitalidrespeu approprites, se dtveloppe depuis plus de 20 ans. Elle est issue de deux phtnomdnes lits :(1)responsabilitt croissante de tiers comme bailleurs defonds institutionnels en services de santt et volontt grandissante de ces intervenants de sefaire rendre des comptes; ( 2 ) au Canada en particulier, dtbat sur le caractdre suflsant du financement des hdpitaux et l’insuffisance percue du contrdle des coats reposant sur des budgets globaux. fitant donnt l’intkr2t manifestt pour la GU, les administrateurs hospitaliers, les associations provinciales et ftdtrales reprtsentant des hdpitaux et le personnel et les mkdecins en milieu hospitalier jugeraient utile un examen desprogrammes GU dupoint de vue de ce que l’on connait de leur efficacittet des initiatives prtcises prises au Canada. Les auteurs insistent sur le besoin critique dune tvaluation structurte de ces programmes. Jusqu’d prksent, il y a eu peu de recherche systkmatique sur les questions d’application et d‘incidence qui s’y rapportent. Cet examen est particulidrement utile parce que les probldmes GU ne sontpas rkpandus au Canada.
Utilization Management: A Literature Review for Canadian Health Care Administrators by Sam B. Sheps, Geoff Anderson and Karen Cardiff
T
he cost of general and allied special hospitals as a percentage of the Gross National Product has increased significantly since 1978.’ At the same time, public expectations of the level and quality of care provided by hospitals have also increased? Hospitals are faced with the difficult task of ensuring quality of care while attempting to contain cost increases. Moreover, it has long been believed that a proportion of the increase in hospital costs is the result of inappropriate or unnecessary uti1izationPS In this environment, it is not surprising that there has been increased interest in and discussion of hospital-based utilization management (UM) in Canada. Payne6defined UM as “the deliberate action by hospital administrators or payers M influence providers of hospital services to increase the efficiencyand effectivenessof services provided”. @. 712) UM consists of two components: (1) comparing performance to explicit standards of appropriate care; and (2) developing initiatives to rectify identified problems? The component of UM that involves measurement of performance and
FORUM
Spring/Printemps 1991
comparison to standards is referred to as utilization review (UR): the application of defined criteria or expert opinion, or both, to the observed process of care in order to assess the efficiency of that process, and to determine the appropriateness of decision making related to the site, duration and frequency of care! The second component of UM, the response to identified problems, consists of techniques for modifying behaviour of providers, patients or administrators, either directly or indirectly, and relates not only to services provided in the hospital setting, but to the co-ordination of hospital services with community services. U R may be conducted before (prospectivereview), during (concurrent review) or after (retrospectivereview) the patient's stay. It is important to note the dual objectives of UM of achieving cost containment and quality of care; thus, UM is related to hospital risk management and, particularly, quality assurance (QA) programs. Although there is rhetorical overlap between the activities of UM and QA programs, UM usually focuses on the process of medical care while QA includes the review of the structure and process as they relate to outcome of care: the sequelae of treatment or patient satisfaction, or both. The objective of this article is to provide Canadian hospital administratorswith a brief review of the literature on hospitalbased UM to help them to understand the basic concepts involved, the types of programs that have been developed and their impact. Although the focus is on applications for Canadian hospitals, much of the literature is from the United States. The review is therefore divided into two sections. The first deals with U.S. initiatives and the second with UM programs in Canada.
Utilization review in the U.S. For three decades in the U.S., federal and state authorities have implemented many regulatory mechanisms to reduce over-utilization,to control costs and to ensure quality of care.8 In 1966, legislation requiring retrospective review for hospitals participating in Medicaid and Medicare was introduced and U R activities proliferated? It is noteworthy that the various review activities that grew out of this legislation developed in spite of there being little evidence that the activities actually worked." Hospitals across the nation used retrospectively defined length of stay as the major endpoint because (1) it was a measure objectively and easily attained;' (2) no clinical standard existed that defined inappropriate use as opposed to appropriate use;5 (3) it was thought to be an important measure of hospital performance and resource consumption; and (4) it was thought to be one of the most important variables contributing to hospital costs.12
'
However, by the mid-l970s, it was generally acknowledged that this approach to retrospective review had not been effective3 because utilization problems were identified after the patient had left the hospital and appropriate norms for evaluating medical necessity and appropriateness of care were lacking.13 In recent years, the focus in retrospective review measurement has been on the development of tools that explicitly incorporate clinical characteristics of patients in the process used to identify prolonged lengths of stay. One of the first such techniques was based on Diagnostic Related Groups (DRGS).'~In essence, the DRG system was designed to classify patients in acute care hospitals into a manageable number of groups that, on the basis of clinical characteristics, should have similar lengths of stay.'4"5 Although the use of DRGs gained some acceptance, several problems were noted: (a) they only explained a fraction of the variance in length of stay;16 (b) there were methodological problems in their construction and the classification of disease~;'~-''and c studies illustrated significant human error in coding DRGs.26. Because it was felt that clinical severity might conmbute to the large variation observed with DRG categories, tools that measured the clinical severity were developed. Studies have indicated these severity adjustment tools, such as Severity of 111ness Index (SII)22or MedisGro~ps?~ can explain some of the variance in length of stay within DRG categories.24-27 U R based on length of stay may provide a reasonable screening tool for identifying resource use that varies from standard or expected patterns of care but it is not an accurate way to identify inappropriate care!> 28 Indeed, review methods based on length of stay analysis may encourage inappropriate short stays and penalize patients who truly require longer stays?' In response to the apparent failure of retrospective review either to enhance quality of care or control costs, Professional Services Review Organizations (PSROs) expanded the use of admission and concurrent review processes in 1972. Under the PSRO mandate, all Medicaire and Medicaid admissions were reviewed to ensure that medical care in a hospital setting was necessary. In general, the criteria used to determine the appropriateness of admissions could be either criteria specific to a particular problem, diagnosis or procedure and criteria that specify the types of services that should be provided at a hospital level of care. A formal evaluation of these two components of the review process was done in 1977 and showed that the impact on utilization and health care expenditures was minimal?' By the early 1980s it had become quite evident that utilization review was not very effective at containing costs and, in 1983, the federal government enacted legislation under which reimbursement from Medicare was determined prospectively
El
a,
'
FORUM
based on the patient’s diagnosis and characteristics, instead of on the actual costs incurred in caring for the patient. This change in reimbursement made UM a vital tool for maintaining the financial viability of the hospital. Studies have indicated that the average length of stay declined by 5.4% in 1984 and that total hospital costs were approximately $2 billion less than Medicare officials had expected to pay in 1984?l More recent methods for U R use a detailed and standardized concurrent review process with predetermined, specific, objective,consistent and explicit criteria!. 32 The explicit criteria exhibit two properties that may make them acceptable to physicians: (1) they are based on clinical data and (2) each patient is assessed individually against the criteria. In contrast, length of stay from the physician’s perspective is an arbitrary measure of resource consumption,particularly because individual patients are compared to a group average.33 The three concurrent review tools most commonly discussed in the literature are: (1) AppropriatenessEvaluation Protocol (AEP) developed by a group of researchers at the Boston University School of Medicine during the 1 9 7 0 ~(2) ; ~Inten~ sity of Service, Severity of Illness and Discharge Screens (ISDA) developed by InterQual;22and (3) Standardized Medreview Instrument (SMI) developed by Sy~teMetrics.3~ Having identified an inappropriate admission or hospital day by the explicit clinical criteria, the cause may be assigned either to the hospital,the physician, the family or the environment.‘ This is particularly important because it not only promotes a broader understandingof how the various components of the health care system affect utilization but, at the same time, eliminates the tendency to focus solely on physician behaviour, fostering a more relevant set of approaches to the problems of health care, thus utilization. One study36showed that the AEP and ISD were found to be moderately valid compared to physician opinion, but that the SMI had low validity. Despite the long history of utilization analysis in the U.S., there has been astonishingly little solid e~aluation.3~ Research by Wickizer, Wheeler and Feldstein3*suggested that U R in the U.S. has contributed to cost containment and the efficient use of health care resources. However, they noted that the success of UR is linked to the incentives or disincentives associated with the review activities: “the stringency with which UR guidelines are applied and the level of penalties invoked for noncompliance,as well as other factors specific to the UR program, are likely to influence outcomes”. @. 646) In particular, as noted previously, the demonstratedeffectiveness of U R was essentially nonexistent until payment was directly and prospectively linked to UR.
Spring/Printemps 1991
Utilization review in Canada For 20 years the legislative and financial environment faced by Canadian hospital administrators has been much different from that in the U.S. However, there has been growing pressure on hospitals and increasing interest in UM. Although there is very little published material that provides descriptions or evaluations of UM programs in Canada, some specific initiatives have been undertaken.
The Value Improvement Process (VIP)The Value ImprovementProcess is a retrospective review program that was developed by the Baxter Corporation in 1984.3’ The process involves the selection of 40 to 50 patient records that fit preselected diagnostic criteria and the collection of information about the resources used for these patients. A profile of the costs associated with treating the particular type of patient is determined and compared to cost profiles of similar patients collected by Baxter from other hospitals. The review process then shifts from a determination of costs to an evaluation of practices followed by the development of recommendations. The process has been documented as improving quality of care4” 41 while reducing length of stay and saving money.42 However, the studies that evaluate VIP have been “before and after” studies with no use of control hospitals. The use of these designs raises the possibility that the observed changes in utilization were the result of factors other than the VIP process. Although VIP may decrease length of stay and costs, it is associated with a relatively limited range of procedures, in particular, total hip replacements, cataract surgery and treatment of acute myocardial infarction. This is notable because hospital admission and discharge data indicate that these procedures account for a very small proportion of all hospital admissions and di~charges.4~ Thus, the impact on overall hospital utilization, at least regarding what we knbw about VIP, might be small. This is, of course, not to say that VIP is not valid for the purposes for which it was intended; namely, to reduce costs as much as possible for specific procedures.
Resource Enhancement Program (REP) REP is a retrospective review program developed by the Health Management Resource Gr0up.4~Its goals are simultaneously to improve utilization of hospital resources and maintain or improve quality of care for patients. REP provides health care managers with a profile of the costs associated with treating patients within specific diagnostic groups (e.g. acute
Spring/Printemps 1991
FORUM
control hospitals. Similarly,small-scale evaluations of VIP have been done but they have been based on demonstrating efficacy for specific procedures instead of assessing the effectiveness at a population level (i.e. does the implementationof VIP significantly reduce length of stay or some other variable when applied to the normal operations of the hospital?). The relative lack of UM development and research in Canada compared to the U.S. can be explained by two major SWITCH is an acronym given to a concurrent review profactors. First, hospital costs have not risen as rapidly in Canada gram developed by the Peace Arch District Hospital in 1984, as in the U.S. due to the central control of expenditures and the and stands for Signs, Winds, Intramuscular Injections, Tubes, use of global budgets;” therefore, there is less pressure to use Consultant (Active) and Hospice. These criteria have since UM. Second, UM, as must be acknowledgedby administrators, been patented by MEDICUS!’ The SWITCH Index System involves modifying provider behaviour and, in the current uses six categories of explicit criteria to identify patients who Canadian context, this would entail a major change in the require acute care hospital services, either for diagnostic acrelationship between administrators and medical staff. Altivity or for treatment. Patients are monitored daily and are though managing change is never easy, it can be doubly difcategorized as “on-index”or “off-index”for that day, dependficult if the need for change emerges from outside the hospital ing on whether they meet at least one of the SWITCH criteria. environment (e.g. government pressure) and involves altering The system also identifies causes for delays in care in terms of fundamental relationships. Moreover, after having accepted the external services, internal services and physicians. Wiggins” need for change, the change process is often invested with an evaluated the SWITCH system in 1988; although there was aura of known effectiveness and implied success that make forevidence supporting a reduction in length of stay in a hospital mal evaluation a potentially hazardous undertaking because the using the SWITCH Index System, the reduction was not sigempirical evidence may not support the aura. Thus, issues of nificantly different from that observed in control hospitals over control over the evaluation become central in managethe same time period. ment/professionalinteractions,adding to the likelihood of conflict. The lack of evidence on the impact of UM has serious consequences for institutions,those who work in them and those Although HMRI has not developed any specific UM who manage them. The consequences are amply demonstrated programs, it is included in this section as an approach to colin our brief review of the early American experience with UM: lecting data on utilization. The goals of HMRI are to “collect, repeated wide-scale application of interventions for controlling analyze and report standardized comparative information on length of stay and expenditures that were never evaluated patient care provided by Canadian health care instit~tions”.~~ properly before implementationand that consistently failed to (p. 13) To accomplish this, HMRI has developed case mix achieve the desired results. Such failure not only casts doubt on groups (CMGs) that are similar to DRGs and provide CMG the whole approach but makes key groups, particularly length of stay comparison to hospitals. As an extension physicians, cynical and wary. The PPS was successful only beI-IMRI~~has recently developed Resource Intensity Weights cause as an intervention it directly influenced the key outcome, (RIWs) that compare length of stay and charge or cost-based not because it was evaluated and found to work on a small weights across CMGs. Similar to the DRG-based Prospective scale and then applied on a wide scale. Payment System (PPS) developed in the U.S., researchers at The importance of UM and subsequent interest in its use has HMRI contend that RIWs could assist in analyzing hospital not been accompaniedby systematic research into issues recase mix and perhaps in determining global budget adjustments lated to its implementation and impact. There are few sound in Canadian h0spitals.4~ studies available from the U.S. and it is difficult to generalize their findings to the Canadian setting. Thus, the message of this review is that, despite the inherent difficulty of evaluating UM strategies on a small scale, it is critical that it be done to avoid Aside from a few general reviews and 49 virtually the inevitable problems that will occur in a wider implementano empirical data exist to demonstrate the effectiveness (or tion. UM is not simply a tactic for providing technical solunot) of UM in Canada. As noted before, Wiggins” found that tions to management problems; it represents a significant the SWITCH Index System as implemented in a community change in organizational relations, which provides both opporhospital did not significantly reduce length of stay compared to myocardial infarction). Similar to VIP, participating hospital managements can compare the costs associated with treating patients in their hospital to a database of cost profiles provided by REP, helping them to identify where costs can be reduced. There has been no formal evaluation of REP.
The SWITCH Index System
The Hospital Medical Records Institute (HMRI)
Discussion
FORUM
tunities and risk. If viewed positively, with the clear objective of improving patient access to care through greater efficiency in bed resource use, and improving patient care through assessing quality of care, UM can be a powerful tool for forging better administration and professional relations. Given the growing interest in UM, we believe it is imperative that evaluation of UM programs begins to occupy a more prominent position in health services research.
References and notes 1. Evans, R.G. 1984. Strained Mercy: The Economics of Canadian Health Care, Toronto: Butterworth and Co. (Canada) Ltd. 2. Chenoy, N.C. 1984. Strategic planning-understanding and responding to a rapidly changing world. Health Management Forum 5(2): 3-19. 3. Dans, P.E., Weiner, J.P. and Otter, S.E. 1985. Peer review organizations: promises and potential pitfalls. New England Journal of Medicine 3 13(18): 1131-1137. 4. Hughes, R.A., Gertman, P.M., Anderson, J.J. et al. 1984. The ancillary services review program in Massachusetts. Journal of the American Medical Association 252(13): 1727-1732. 5. Goran, M.J. 1979. The evolution of the PSRO hospital review system. Medical Care (suppl) 17(5): 1-46. 6. Payne, S.M.C. 1987. Identifying and managing inappropriate hospital utilization: a policy synthesis. Health Services Research 22(5): 709-769. 7. Anderson, G., and Lomas, J. 1988.Development of UtilizationAnalysis: How, Why and Where It’s Going, Vancouver, B.C.: Health Policy Research Unit, University of British Columbia, 88(7), with comment by Barer, M.L., You Can Take the System to Its Analyst, But Should You Let the Analyst Loose on the System. 8. Brown, L.D. 1986. Introduction to a decade of transition. Journal of Health Politics and Policy Law 11 (4): 569-583. 9. McDonough, Catherine and Vaz, Andrew P. 1987. Hospital utilization management. Health Management Forum 8(1): 42-52. 10. Wiggins, Sandra. 1988. Utilization Management of Acute Care Services: Evaluation of the SWITCH Index System. M.Sc. diss., University of British Columbia, Vancouver, B.C. 11. Goldberg, G.A. and Holloway, D.C. 1975. Emphasizing “level of care” over “length of stay” in hospital utilization review. Medical Care 13(6): 474-485.
Spring/Printemps 1991
12. Lave, Judith, R. and Leinhardt, Samuel. 1976. An evaluation of a hospital stay regulatory mechanism. American Journal of Public Health 66(10): 957-967. 13. Flashner, B.A., Reed, S . , Coburn, R.W. et al. 1973. Professional Standards Review Organizations: analysis of their development and implementation based on a preliminary review of the hospital admission and surveillanceprogram in Illinois. Journal of the American Medical Association 223 (13): 1473-1484. 14.Fetter, R.B., Shin, Y.L., Freeman, J.L. et al. 1980. Case mix definition by diagnosis-related groups. Medical Care (Suppl) 18(2): 1-53. 15. Grimaldi, P.L. 1983. New medicare DRG payment calculation issues. Nursing Management 14(11): 19-23. 16. MacKenzie, T.A., Markle, F. and Croke, M. 1987. CMG’s: variations on a theme. Health Management Forum 8( 1): 21-24. 17. Mullin, R.L. 1985. Diagnosis-related groups and severity: ICD-9-CM, the real problem. Journal of the American Medical Association 254(9): 1208-1210. 18. Smits, H.L., Fetter, R.B. and McMahon, L.F. November 1984. Variation in resource use within diagnosis-related groups: the severity issue. Health Care Financing Review (annual suppl): 71-78. 19. Mendenhall, S . 1984. DRGs must be changed to take patient’s illness severity into account. Modern Healthcare 14(15): 86,88. 20. Gertman, P.M. and Lowenstein, S . November 1984. A research paradigm for severity of illness: issues for diagnosis-relatedgroups system. Health Care Financing Review (annual suppl): 79-80. 21. Johnson, A.N. and Appel, G.L. 1984. DRGs and hospital care records: implications for medicare case mix accuracy. Inquiry 21: 128-134. 22. InterQual. 1978. The I S k A R e v i e w System with Adult Criteria. InterQual Incorporated, Westborough, Ma. 23. MediQual, MedisGroups. MediQual Systems Inc., 1900 West Park Drive, Westborough, Mass. 24. Horn, S.D., Horn, R.A. and Sharkey, P.D. November 1984. The severity of illness index as a severity adjustment to diagnosis-relatedgroups. Health Care Financing Review (annual suppl): 33-45. 25. Horn, S.U., Sharkey, P.D., Chambers, A.F. et al. 1985. Severity of illness within DRGs: impact on prospective payment. American Journal of Public Health 75( 10): 1195-1199.
Spring/Printemps 1991
26. Jencks, S.F. and Dobson, A. 1987. Refining case-mix adjustment: the research evidence. New England Journal of Medicine 3 17(11): 679-686. 27. Iezzoni, L.I., Ash, A.A., Cobb, J.L. et al. 1988. Admission MedisGroups score and the cost of hospitalizations. Medical Care 26(11): 1068-1080. 28. Owens, R.R. and Averill, R.F. October 1984. The role of utilization management under PPS. Healthcare Financial Management: 60-62,64,66. 29. Kemper, K.J. 1988. Medically inappropriate hospital use in a pediatric population. New England Journal of Medicine 388(16): 1033-1037. 30. Dobson, A., Greer, J.G., Carlson, R.H. et al. 1978. PSROs: their current status and their impact to date. Inquiry 15: 113-128. 31. Kelly, J. and Bankhead, C.D. 1985. DRGs: how are they stacking up? Medical World News 26(5): 80-103. 32. Clemenhagen, C., Champagne, F., Contandriopoulos, A. et al. 1985. Hospital-Based Quality Assurance: A Starting Point for Physicians and Managers, Ottawa, Ont.: Canadian Hospital Association. 33. McIlrath, S. January 11, 1985. Review system based on illness severity. American Medical News. 34. Boston University School of Medicine. Appropriateness Evaluation Protocol (AEP), Health Care Research Unit, Boston University School of Medicine, Suite 1102,720 Harrison Avenue, Boston, Mass. 02118. 35. SysteMetrics. 1983. Standardized Medreview Instrument (SMI). SysteMetrics Incorporated, 104 West Anapamu Street, Santa Barbara, Calif. 93101. 36. Strumwasser, I., Paranjpe, N.V., Ronis, D.L. et al. 1990. Reliability and validity of utilization review criteria: appropriateness evaluation protocol, standardized . medreview instrument, and intensity-severity-discharge criteria. Medical Care 28(2): 95- 111. 37.Feldstein, P.J., Wickizer, R.M. and Wheeler, J.R.C. 1988. Private cost containment: the effects of utilization review programs on health care use and expenditures.New EngIand Journal of Medicine 318 (20): 1310-1314. 38. Wickizer, T.M., Wheeler, John, R.C. and Feldstein, Paul, J. 1989. Does utilization review reduce unncessary hospital care and contain costs? Medical Care 27(6): 623-647. 39. Baxter Corporation, Value Improvement Process. Baxter Corporation, 2930 Argentia Road, Missisauga, Ont.
FORUM
40. Winchell, J.R. October 1985. Reduced costs can improve care. Health Care. 41. Coombs, R. 1988. Implementing a value improvement process. Healthcare Management Forum 1 (2): 14-16. 42. Eliasoph, H. and Hassen, P. 1986. VIP streamlines care and reduces length of stay. Dimensions 63 (7): 42-53. 43. Hospital admission and discharge data (1987/88) for British Columbia indicate that cholecystectomiesaccounted for only 1.1% of all hospital admissions, total hip replacements accounted for .15% and acute myocardial infarction accounted for 1.2%. 44. Ryan, Patricia and De Paoli, Lisa. 1989. Comparing the Costs of Managing Patients with Acute Myocardial Infarction in Four B.C. Hospitals. Vancouver, B.C.: Health Management Resource Group. 45. MEDICUS, The MEDICUS UM Program “SWITCH’. 1989. MEDICUS Canada Incorporated, Toronto, Ont. 46. Ontario Hospital Association. 1988. Guidefor Hospital Utilization Review and Management in Ontario. Don Mills, Ontario: Ontario Hospital Association. 47. Hospital Medical Records Institute. 1988. Case Weighting Methods: A Review by the HMRI Database Committee. Ottawa, Ontario: Don Carmichael and Company. 48. Rachlis, M., and Fooks, C. 1988. Utilization Analysis: Current Initiatives Across Canada. Conference Paper. Conference Proceedings, First Annual Health Policy Conference, McMaster University, Hamilton, Ont. 49. Suttie, B., Helliwell, B., Villenure, D. 1988. Some Aspects of Utilization Management in Twenty Ontario Hospitals, Waterloo, Ontario: Centre for Applied Health Research, University of Waterloo. 50. Evans, R.G., Lomas, J., Barer, M.L. et al. 1989. Controlling health expenditures-the Canadian reality. New England Journal of Medicine 320 (9): 571-577. Sam B. Sheps, MD, MSc, FRCPC, is Associate Professor and Head, Department of Health Care and Epidemiology, University of British Columbia. Geoff Anderson, MD, PhD, is Associate Director, Health Policy Research Unit, and Assistant Professor, Department of Health Care and Epidemiology, University of British Columbia. Karen Cardiff,BScN, BSc, MHSc, is Research Associate, Division of Health Services Research and Development, University of British Columbia.