Method of optimizing healthcare services consumption

[0001] The present invention generally relates to a method of optimizing healthcare services consumed by patients including employees and their family members by improving the overall quality of care and reducing the overall cost incurred by the employer, and more particularly to a method for application by a healthcare quality management firm (HOM) of characterizing the healthcare situation of an employer who pays for healthcare, comparing that healthcare situation to that of a geographic area in which the employer resides, identifying factors affecting the quality and cost of the healthcare, and recommending action for addressing the factors by applying resources at levels corresponding to the relative affect of the factors on the quality and cost of the healthcare.

[0002] Employer sponsored healthcare benefits are of tremendous value to employees and their families. Such benefits, on the other hand, typically constitute a significant portion of an employer's total operating costs. Unfortunately, as medical costs continue to increase, the cost of providing employer sponsored healthcare benefits will continue to increase.

[0003] Currently, many employers attempt to offset the rising costs of providing healthcare benefits by shifting the cost to employees. Of course, only so much of the expense can be shifted to employees. At some point, the cost incurred by the employees will become prohibitive, and employer sponsored healthcare will no longer be seen as a benefit. Some employers attempt to monitor the price of certain healthcare services, but without information relating to the quality of the services, cost information is of limited value. Other employers have attempted to reduce their healthcare expenses by sponsoring health fairs or wellness screenings. This approach, while somewhat effective in prompting preventative healthcare, is not a focused expenditure of resources. For the majority of employees who are healthy, the money spent on wellness screenings is essentially wasted. Finally, employers sometimes attempt to negotiate the fixed costs associated with administering healthcare benefits. Again, since these costs typically make up only a small portion of the total cost, even successful negotiation attempts will have a limited impact on the employer's bottom line.

[0004] In short, employers have been largely unsuccessful in their attempts to control healthcare costs while ensuring a high level of care. Employers simply lack the information necessary to identify the most significant factors affecting their healthcare costs, to quantify and compare the performance of healthcare providers, and to apply their resources in a way that most effectively reduces both the overall consumption of healthcare and the costs of the services consumed while maintaining or improving the quality of the healthcare benefits they provide.

[0005] The present invention provides a method of optimizing healthcare services consumption through analysis of the demographic and wellness characteristics of an employee population (including employees and employee family members, hereinafter, patients), analysis of the quality and cost efficiency of the practices of providers used by the patients, and intervention with patients and providers to improve the overall health of the patients, the practices of the providers, and the cost efficiency of the employer provided healthcare plan. The method, in one embodiment thereof, includes the steps of assessing the healthcare situation of the employer as it relates to normative characteristics of a health economic zone including the patients, identifying patients from the covered population likely to generate expensive healthcare claims relative to the other patients based on data representing past healthcare claims generated by the patients, periodically determining whether these patients have obtained healthcare services that satisfy predetermined requirements, identifying qualified providers in the health economic zone who provide high quality, cost efficient healthcare services relative to other providers in the health economic zone based on data representing past practice patterns of the providers, prompting patients who have not obtained healthcare services that satisfy the predetermined requirements to obtain additional healthcare services from the qualified providers, and responding to healthcare requests from patients by determining whether the requesting patient is seeking to obtain healthcare services from a qualified provider, and, if not, urging the patient to obtain services from a qualified provider.

[0006] The features and advantages of the present invention described above, as well as additional features and advantages, will be readily apparent to those skilled in the art upon reference to the following description and the accompanying drawings.

[0007] FIG. 1 is a conceptual diagram of participants in a healthcare consumption situation that may be optimized using a method according to the present invention.

[0008] FIG. 2 is a conceptual diagram of an interrelationship between a central database and the participants shown in FIG. 1.

[0009] FIG. 3 is a flow diagram depicting steps included in one embodiment of the present invention.

[0010] FIG. 4 is a conceptual diagram of a health economic zone.

[0011] FIGS. 5-15 are illustrations of reports generated according to an embodiment of the present invention.

[0012] FIG. 16 is a flow diagram of a process for evaluating the practice characteristics of healthcare providers.

[0013] FIG. 17 is a flow diagram depicting steps included in one embodiment of the present invention.

[0014] The embodiments described below are merely exemplary and are not intended to limit the invention to the precise forms disclosed. Instead, the embodiments were selected for description to enable one of ordinary skill in the art to practice the invention.

[0015] FIG. 1 depicts a relationship among participants in a typical employer provided healthcare situation. In this example, the employer 10 is self-insured and provides funds, based on predicted healthcare costs, to a third party administrator (TPA 12) of healthcare benefits for paying employee healthcare claims. Of course, also involved in this relationship are the healthcare consumer, patient 14, the healthcare provider 16 (e.g., a physician or a facility such as a hospital, laboratory, etc.), a pharmacy 18, a pharmacy benefit manager (PBM 19), a PPO 20, and a healthcare quality management firm (HQM 13). As should become apparent from the following description, HQM 13 could perform the functions of TPA 12. Thus, except where expressly indicated otherwise or mandated by the context of this description, references to HQM 13 may include HQM 13 and TPA 12.

[0016] In a typical transaction associated with a healthcare claim, patient 14 visits provider 16 to obtain healthcare services and/or products such as drugs. For simplicity, this description collectively refers to services and products as healthcare services. Provider 16 submits a claim to PPO 20 (or alternatively directly to TPA 12) in an amount corresponding to the cost of the services. Provider 16 may also write a prescription that is received by a pharmacy 18. In that event, pharmacy 18 submits a claim to PBM 19, which in turn submits a claim to TPA 12. As is well known in the art, PPO 20 (or alternatively TPA 12) typically discounts or reprices the claimed charges based on an agreement between provider 16, pharmacy 18, and PPO 20. The repriced claim is submitted to TPA 12 for payment. TPA 12 accesses funds in the healthcare account of employer 10 to pay provider 16 and PBM 19 the repriced claim amounts. PBM 19 then forwards a payment to pharmacy 18. TPA 12 then also informs patient 14 of the patient's payment responsibility that arises as a part of the application of the terms of the underlying benefit plan when it does not pay 100% of eligible charges. Patient 14 then sends a payment to provider 16. The above-described example assumes that TPA 12 is separate from HQM 13. If HQM 13 functions as a combination of HQM 13 and TPA 12, then HQM 13 interacts directly with employer 10, patient 14, provider 16, PBM 19, and PPO 20 in the manner described with reference to TPA 12 above.

[0017] As should be apparent from the foregoing, throughout each such transaction, TPA 12 has access to all of the material claim information. TPA 12 shares this information with HQM 13, which may contact employer 10, patient 14, and/or provider 16. Accordingly, as will be described in detail below, HQM 13 is in a position to facilitate change in and/or directly influence the healthcare situation to control the cost incurred by employer 10 and to encourage consumption of healthcare from high quality providers 16. Thus, HQM 13 is described below as practicing the present invention as a service for the benefit of its clients, employers 10, and patients 14 including the clients' employees and their family members.

[0018] According to one embodiment of the present invention, TPA 12 maintains a database 22 including a variety of different types of information from employer 10, provider 16, PBM 19, and PPO 20 as depicted in FIG. 2. As is further described below, TPA 12 also updates information included in database 22 as a result of its interaction with HQM 13. Database 22 may be maintained on any of a variety of suitable computer-readable media such as a hard drive of a computer. While FIG. 2 suggests contributions of information to database 22 by each of employer 10, TPA 12, provider 16, PBM 19, and PPO 20, it should be understood that such information may not be provided directly to database 22. Instead, TPA 12 may receive information from the other participants and enter and/or otherwise process the information for storage in database 22. For example, information may be transferred electronically from employer 10, provider 16, PBM 19, PPO 20, and HQM 13 to TPA 12 via a network or multiple networks. Moreover, TPA 12 may physically reside at multiple locations, each of which receives information from the other participants. Such multiple locations may be connected together via a network configured to permit simultaneous access to database 22 through a server. Any suitable method of transferring information and storing such information in either a centralized or distributed database 22 is within the scope of the invention. For simplicity, the transfer of information is described herein as occurring electronically over a network, and database 22 is described as a centralized database accessible by a single TPA 12 location.

[0019] As is further described below, the information stored in database 22 permits HQM 13 to evaluate the healthcare situation of employer 10, including the cost information, the healthcare characteristics of patients 14, and the performance of providers 16 used by patients 14 covered under the healthcare plan provided by employer 10. Accordingly, the information in database 22 includes employer information, patient information, provider information, pharmacy information, and claims information that may relate to some or all of the other types of information. The employer information includes information identifying employer 10, patients 14 covered under the employer provided healthcare plan, PPO 20 associated with employer 10, as well as historical data that characterizes changes in the healthcare situation of employer 10 over time. The patient information includes the name, address, social security number, age, and sex of each patient 14 covered under the healthcare plan provided by employer 10. The provider information includes the name, tax identification number, address, and specialty of a plurality of healthcare providers across a large geographic region, such as the entire United States. As is further described below, portions of the provider information are associated with employer 10. These portions correspond to the providers 16 that provide services to patients 14. The pharmacy data includes information identifying the type, quantity, and dosage of drugs associated with a particular prescription for a particular patient 14 as well as the social security number of the patient 14. This information permits association of prescription drug claims with patients 14. These claims can be further associated with the provider 16 that wrote the prescription by accessing the claims data (described below) associated with the patient 14 who filled the prescription to determine which provider 16 patient 14 saw prior to obtaining the prescription. Alternatively, an identifier may be included in the pharmacy data with each prescription entry that identifies provider 16.

[0020] The claims data stored in database 22 include portions of the above-described data, but may be organized or associated with a particular claim. More specifically, a claim may include information identifying and/or describing employer 10, patient 14, provider 16, pharmacy 18, PBM 19, and PPO 20. The claim may further include information describing the condition or symptoms of patient 14 that generated the claim, the diagnosis of provider 16, the procedures ordered by provider 16 to treat the diagnosed condition as identified by commonly used procedure codes, and the costs (both original charges and repriced amounts) of the healthcare services associated with the claim.

[0021] As indicated above, the information stored in database 22 comes from a variety of sources. For example, when an employer 10 becomes a new client of HQM 13, PPO 20 servicing employer 10 may provide HQM 13 with enrollment data including employer information, employee information, and associated past claims information. HOM 13 may then process that information for addition to database 22. Periodically, PPOs 20 of employers 10 transfer claims information to TPA 12 (i.e., as the claims information is processed by PPOs 20). As indicated above, in addition to information relating to associated healthcare services, this claims information may include employee information, provider information, and pharmacy information. Additionally, PBMs 19 (or data transfer services working with PBMs 19) periodically transfer pharmacy information to TPA 12. As further described below, each time new information is provided to TPA 12, TPA 12 and/or HOM 13 may process the information such that it is associated with a particular employer 10, a particular patient 14, or a particular provider 16.

[0022] Referring now to FIGS. 3 and 4, one embodiment of the method according to the present invention may be generally described as involving three basic steps: analyzing the healthcare situation of employer 10, improving the healthcare consumption characteristics of patients 14, and improving the overall performance characteristics of providers 16 used by patients 14. One process for analyzing a healthcare situation of an employer 10 is depicted in FIG. 3. In general, after all of the relevant information regarding employer 10, patients 14 associated with employer 10, and providers 16 used by patients 14 resides in database 22, HQM 13 executes software (as further described below) to access database 22 and identify a Healthcare Economic Zone (HEZ 24, FIG. 4) corresponding to employer 10 (step 26). As shown in FIG. 4, HEZ 24 corresponds to a geographic area that includes all patients 14 associated with all employers 10 and providers 16 used by patients 14 (including physicians 28 and facilities 30, such as hospitals). HEZ 24 may be defined to correspond to Hospital Service Areas set forth by the Dartmouth Atlas project, a funded research effort of the faculty of the Center for the Evaluative Clinical Sciences at Dartmouth Medical School. Essentially, HEZs are based on the zip codes of the residential addresses of patients 14 stored in database 22 and the locations of providers 16 servicing those zip codes. In other words, an HEZ 24 includes a geographic region in which patients 14 tend to obtain their primary healthcare. For example, assuming patients 14 associated with employer 10 all reside in three adjacent zip codes that are serviced by one facility 30 (also within one of the three zip codes), then those three zip codes are included in HEZ 24. However, if facility 30 also refers patients 14 to, for example, specialist providers 16 in a fourth zip code, then the fourth zip code is also included in HEZ 24. FIG. 4 shows HEZ 24 fully contained within a larger geographic area such as a state 32. It should be understood, however, that HEZs 24 (or the equivalent of HEZs 24) may extend across state lines.

[0023] Referring again to FIG. 3, step 34 indicates that information in database 22 corresponding to employer 10 (i.e., employer information, patient information, claims information corresponding to patients 14 associated with employer 10, and provider information) is analyzed to evaluate the healthcare situation of employer 10. In step 36, the employer specific data is compared to generalized data relating to HEZ 24 as is further described below. As indicated in FIG. 3, the results of the analyses performed in steps 26, 34, and 36 may be processed in the form of provider reports 38, employer reports 40, and patient reports 42, some or all of which may be provided to employer 10 as shown in FIG. 1 as part of the process of analyzing the healthcare situation of employer 10. Step 44 depicts the process of updating database 22 as HOM 13 and/or TPA 12 receive claims information and/or changes in the population of patients 14 associated with employer 10 as a result of employees being hired by or departing from employer 10, or changes in the family situation of the employees. As should be apparent from the figure, the process of analyzing the healthcare situation of employer 10 is therefore continuously updated and may result in generation of periodic reports for employer 10 and HQM 13 to track changes in the healthcare situation over time.

[0024] FIG. 5 depicts an example of an employer report 40. Although chart 46 of FIG. 5 does not compare employer 10 information to HEZ 24 information, it is an employer report 40 because it provides employer 10 information regarding the costs of healthcare services in the HEZ 24 in which employer 10 (more accurately, patients 14 associated with employer 10) resides. Chart 46 includes a specialty column 48, a total allowed charges column 50, a total allowed charges at normative costs column 52, a percent of excess charges column 54, and an excess charge per life per year column 56. Chart 46 provides employer 10 information regarding the relative costs of healthcare services (by specialty) in the employer's HEZ 24 as compared to the costs in a larger geographic area that includes HEZ 24 (e.g., state 32, the Midwest, the southeast, etc.). In this example, providers 16 in HEZ 24 charged $4,251,526 (column 50) for cardiology services over the course of a predetermined time period, such as two years. Column 52 shows that the normative costs for such services is $4,488,559 for the same number of healthcare consumers (i.e., patients 14) over the same predetermined time period. More specifically, the dollar amounts in column 52 are derived by first adding all of the charges for cardiology services in the larger geographic area for the predetermined time period and dividing the total by the number of healthcare consumers in the larger geographic area. Then, this average cardiology charge per healthcare consumer is multiplied by the number of healthcare consumers in HEZ 24. As shown in column 54, HEZ 24 experienced cardiology costs that were 5.3 percent below the normative cardiology charges. Finally, column 56 simply converts the percentage deviation from the normative charge into a dollar value divided by the number of healthcare consumers in HEZ 24 and the number of years in the predetermined time period.

[0025] Line 58 shows the totals for all specialties or Major Practice Categories (MPCs). Lines 60 and 62 illustrate a situation wherein HEZ 24 is serviced by more than one PPO 20. Since all of the claims information in database 22 is associated with a particular PPO 20, the charges associated with all claims of healthcare consumers in HEZ 24 corresponding to PPO network A and PPO network B may be separated based on the PPO that handled the claim. Thus, lines 60 and 62 depict the relative usage of the PPOs by healthcare consumers in HEZ 24 (column 50), the normative usage values for each PPO in a larger geographic area (e.g., state 32) (column 52), the cost performance of the PPOs for HEZ 24 relative to the cost performance of the PPOs across state 32 (column 54), and the meaning of that relative performance on a dollars per patient 14 per year basis (column 56). Lines 64 and 66 provide similar information for two hospitals used by healthcare consumers in HEZ 24.

[0026] As should be apparent from the foregoing, employer 10 may readily scan down total allowed charges column 50 to determine the specialties most likely to contribute significantly to the employer's overall healthcare costs. Columns 54 and 56 permit employer 10 to readily identify those practice categories having charges that deviate most from the average or normative charges. In this manner, employer 10 (and HOM 13) can isolate the practice categories that have the most potential for providing the most significant reduction in the overall healthcare costs of employer 10.

[0027] Another employer report 40 (chart 68 of FIG. 6) follows the same format as chart 46, but compares the actual healthcare costs of employer 10 to the typical costs in HEZ 24. Chart 68 includes a specialty column 70, a total costs column 72, a normalized costs in HEZ 24 column 74, a percent excess column 76, and an excess cost per life per year column 78. Column 72 represents the total costs employer 10 incurred for the various specialties listed in column 70 during a predetermined time period. The normalized amounts in column 74 represent the expected cost in HEZ 24 for an employer having the same number of patients 14 as are associated with employer 10. For example, assuming a total cost for cardiology in HEZ 24 of $17,122,789 for 35,623 healthcare consumers in HEZ 24, the average cardiology cost per healthcare consumer is $480.67. Assuming that employer 10 has 150 patients 14, then the expected total cost for cardiology services (i.e., the normalized costs in HEZ 24, column 74) is $72,100. Accordingly, employer 10 has incurred costs for cardiology services that are 23.7% below the anticipated amount for an employer the size of employer 10 located in HEZ 24 as shown by column 76. Column 78 reflects this percentage in a per patient 14 per year dollar value.

[0028] As should be apparent from the foregoing, chart 68 could readily be revised to reflect similar information for actual consumers of the particular specialties as opposed to patients 14 and healthcare consumers generally. In other words, if only nine patients 14 used cardiology services over the predetermined time period (resulting in a total cost of $55,000), column 74 could be modified to reflect the expected amount for nine of the average consumers of cardiology services in HEZ 24 over the predetermined time period. Of course, columns 76 and 78 would then reflect the difference between these values on a percentage and per life per year basis, respectively.

[0029] FIG. 7 is another employer report 40 that summarizes the illness burden and demographics of HEZ 24 associated with employer 10. Chart 80 includes a description column 82, an HEZ 24 data column 84, a normative value a larger geographic area including column 86 for HEZ 24, a percent excess column 88, and an excess per life per year column 90. It is well known that healthcare consumption is greater for adults verses children (other than newborn children), for females verses males, and for older adults verses younger adults. Obviously, healthcare consumption is also greater for individuals having certain types of pre-existing illnesses as compared to healthy individuals. The method of the present invention uses these factors to compute a healthcare index (line 92 in FIG. 7) for HEZ 24 in which patients 14 associated with employer 10 reside. The method of the present invention calculates the healthcare index for an HEZ 24 using Episode Risk Group (ERG) scores inherent in the health risk assessment process provided by Symmetry Health Data Systems, Inc. and described in A New Approach to Health Risk Assessment, a white paper available from Symmetry Health Data Systems, Inc., the disclosure of which is hereby incorporated herein by reference. A healthcare index for each patient 14 in HEZ 24 is computed using a retrospective analysis, and the index for HEZ 24 is derived by calculating an average index for all patients 14 in HEZ 24. As shown in column 84 of FIG. 7, HEZ 24 has 9,808 patients 14 having an average age of 42, and comprising 74.8% adults, 38.6% of whom are female. These factors result in a healthcare index for HEZ 24 of 1.506. As shown in column 88, this healthcare index is 50.6% above the normative healthcare index of 1.0 for the larger geographic area. This high healthcare index results from a higher than typical percentage of females and adults in HEZ 24 and a higher than typical percentage of individuals with health risk factors. More specifically, as shown in column 88 of FIG. 7, 10.8% of the overage is due to atypical demographics (i.e., an older and more heavily female population). 39.8% of the overage is due to the atypical illness burden of the population (i.e., a population with health conditions corresponding to higher than typical health risk factors). Accordingly, an employer 10 in HEZ 24 should expect to have healthcare costs that are greater than the typical costs of the larger geographic region. It should be understood that a similar report could readily be generated comparing the illness burden and demographic information of a particular employer 10 to information describing the HEZ 24 in which patients 14 associated with employer 10 reside.

[0030] Referring now to FIG. 8, a patient report 42 is shown summarizing the chronic illnesses of patients 14 associated with employer 10. It is well known that typically 80% of an employer's healthcare costs are generated by approximately 20% of the covered population of patients 14. That 20% of the population generally has a high incidence of chronic illness. Accordingly, chart 94 of FIG. 8 is generated to provide employer 10 a summary of its chronically ill patients 14.

[0031] As shown, column 96 lists various chronic illnesses. While the method of the present invention may track any number of chronic illnesses, only six are shown in FIG. 8. Column 98 shows the number of patients 14 having each of the listed illnesses. Column 100 shows the number of those patients 14 listed in column 98 that have at least one year of claims history (i.e., have submitted claims that were added to database 22). Column 102 shows the number of patients 14 that have satisfied the minimum annual care requirements (MACRs) recommended for treating the chronic illness or illnesses from which they suffer. Column 104 simply expresses the number in column 102 in the form of a percentage of the total patients 14 suffering from the listed illness. The MACRs for each chronic illness of chart 94 are listed in column 106 and obtained using software available from McKesson Corp., a supplier of information and managed care products and services for the health care industry. In particular, McKesson's CareEnhance Resource Management Software (CRMS) provides such information. As the method of optimizing healthcare services consumption described below is practiced, periodic reports such as chart 94 of FIG. 8 will show improvements in the number of patients 14 that satisfy the MACRs associated with their particular illness(es).

[0032] Chart 110 of FIG. 9 shows the chronic illness status of patients 14 associated with employer 10 in terms of co-morbidities. Chart 110 includes a description column 112, a current patient column 114, a percent of current covered patients column 116, a previous patient column 118, a percent of previous covered patients column 120, and a percent of database driven norms column 122. As shown in column 114, of the 993 total patients 14 covered under a healthcare plan provided by employer 10, a total of 619 have a single chronic illness, 236 have two chronic illnesses, 86 have three chronic illnesses, etc. Column 116 expresses the number of patients 14 listed in column 114 in terms of the percentage of the total patient 14 population. Columns 118 and 120 include similar information representing the status of the chronically ill at a previous date. Employer 10 can monitor changes in the chronic illness status of its patients 14 by comparing these two sets of columns. Finally, column 122 shows the typical percentage of individuals (based on all individuals reflected in the database) with the particular number of chronic illnesses.

[0033] In addition to summarizing patients 14 having chronic illnesses, the method of the present invention also includes the step of performing a risk stratification of all patients 14 covered by employer 10. The results of this risk stratification step are provided to employer 10 as an patient report 40. Chart 124 of FIG. 10 is an example of such an patient report 42. As shown, chart 124 includes a family identification number column 126, a patient identification column 128, an age column 130, a gender column 132, a healthcare index column 134, and a predicted cost column 136. The primary purpose of chart 124 is to display patients 14 in order of their associated healthcare index listed in column 134. The healthcare index is derived using the McKesson CRMS software as described above, which takes into account the age, gender, chronic illnesses, and co-morbidities of each patient 14. Also, by analyzing claims data describing prescriptions, the CRMS software imputes illnesses of patients 14 based on the number and types of medications prescribed for patients 14. Thus, the healthcare index is used to rank patients 14 in terms of their likelihood of generating large medical expenses in the near future. It should be noted that not only the chronically ill are identified by the healthcare index. Other patients 14 having conditions that are not considered chronic may have high healthcare indices. Column 136 provides a predicted cost associated with each patient 14 based on their healthcare index. More specifically, column 136 is derived by calculating the total expense associated with the normative population, and dividing that amount by the total number of ERG risk points of the normative population to get dollars per risk point (prospectively). Then, using the method of the present invention (and not the CRMS software), the healthcare index of column 134 is multiplied by the dollars per risk point value.

[0034] The above-described employer reports 40 and patient reports 42 are illustrative of the way in which the method of the present invention determines which patients 14 covered by employer 10 should receive intervention or proactive coaching (as further described below and depicted in FIG. 1), and at what level of intensity. In other words, since chronically ill patients 14 generally generate large healthcare costs, chronically ill patients 14 should be monitored and coached most actively and at levels corresponding to the number of chronic illnesses from which they suffer. Likewise, patients 14 having high healthcare indices because of their age, gender, illnesses, etc. should be monitored and coached most actively and at levels corresponding to their healthcare index. Using the above-described approach, patients 14 that require proactive coaching typically constitute approximately 25% of the total patient 14 population. It has been shown that this 25% portion of the patient 14 population typically generates 90% of the total healthcare costs incurred by employers 10.

[0035] As indicated above, the method of the present invention also generates physician reports 38 such as chart 138 shown in FIG. 11. Chart 138 is an example of a comparison of various characteristics of the practice of a particular provider 16 to the practices of other providers 16 in the same specialty. In order to make such comparisons, the claims information in database 22 may be analyzed on the basis of episodes of healthcare. This analysis is performed using software applications available from McKesson Corp., which analyze the services and costs associated with claims originated by a particular provider 16. An episode is defined as a healthcare consumption sequence including all healthcare services consumed by a patient 14 for a particular healthcare problem. Episodes may include healthcare services ordered by a physician as a result of an initial office visit (e.g., tests, X-rays, etc.), healthcare services associated with a subsequent hospital visit (e.g., for surgery), and healthcare services associated with aftercare or follow-up visits to the physician.

[0036] The analysis of claims information grouped by specialty episodes permits identification of providers 16 having practice patterns that result in low total costs for the types of healthcare problems they treat as compared to other providers 16 in the specialty. Additionally, providers 16 who deliver high levels of post-primary preventative care services for chronically ill patients 14 can be identified. Finally, specific undesirable characteristics of a provider's 16 practice patterns can be identified such as up-coding, ordering inappropriate services, vague or invalid diagnostic codes, and services that are performed too frequently. All of this information is available from the claims information stored in database 22.

[0037] Referring back to FIG. 11, Bar 140 of chart 138 represents the percentage of procedures ordered by a particular provider 16 (physician ID 223776) that were determined to be inappropriate for the diagnosis reflected in the claims information associated with the evaluated episodes. Bar 142 represents similar data for the entire specialty. Comparing bar 140 to bar 142 shows that this particular anesthesiologist ordered inappropriate procedures at nearly double the rate of others in the specialty. The remaining bar groups 144, 146, 148, 150, and 152 permit similar comparisons for the practice pattern characteristics indicated on chart 138.

[0038] As further described below, one of the steps of a method according to the present invention involves determining whether providers 16 used by employees 14 of employer 10 provide healthcare in a manner that satisfies certain criteria. If so, these providers 16 are identified as Quality Service Providers or QSPS. To achieve a QSP designation or rating, providers 16 must, based on claims information stored in database 22, pass three screens or quantitative tests of the providers' 16 performance or practice characteristics. Any provider 16 who fails one or more of these tests is identified for purposes of practicing the present invention as a non-certified QSP (NCQSP).

[0039] The first test (the CEI test) is primarily economic. Using claims information in database 22, the software of the present invention generates a Cost Efficiency Index (CEI) for each provider 16. The CEI represents the actual total cost of care provided and/or ordered by provider 16 for completed episodes, divided by the total average cost of such care for similar episodes treated by other providers 16 in the specialty. In other words, the cost to employer 10 for the healthcare delivered and/or ordered by provider 16 for all completed episodes for all patients 14 is first extracted from the claims information in database 22. Then, the total cost for all similar episodes handled by all providers 16 tracked in database 22 is determined, and divided by the total number of episodes to arrive at an average cost per episode in the specialty. Finally, the average cost per episode for provider 16 is divided by the average cost per episode in the specialty to arrive at the CEI for provider 16. If provider 16 has a CEI that exceeds a predetermined threshold (e.g., 125% or more above that of others in the specialty of provider 16) and is statistically higher that the average for the specialty (i.e., sufficient claims information is contained in database 22 to calculate the CEI of provider 16 with a statistically acceptable confidence level such as at the p 0.1 level), then provider 16 failed the CEI test and will be designated a NCQSP. A sample report of the data used to complete a CEI analysis is shown in FIG. 12.

[0040] The second test in the QSP rating process (the service rate test) evaluates the preventative care practices of providers 16. As is well known in the field of medical care, preventative care services may significantly affect the overall cost of healthcare, particularly those services provided to treat chronic illnesses to prevent those illnesses from progressing or resulting in other health complications. The McKesson software permits extraction of data representing the number and types of preventative care services ordered by providers 16 for treatment of chronic conditions. In one embodiment of the invention, nineteen chronic conditions are tracked. To evaluate a particular provider 16, the data representing the preventative care services for provider 16 is extracted and compared (according to the method of the present invention) to a minimum number and particular types of services considered acceptable in treatment of the particular chronic conditions treated by provider 16. This analysis results in a service rate for provider 16. More specifically, the total number of services ordered for chronically ill patients treated by provider 16 is determined, and then divided by the number of services required for such patients to achieve compliance with the associated MACRs. This service rate, or fraction of recommended MACRs, is then compared to the typical service rate in the appropriate specialty. If provider 16 has a service rate that is both less than a certain percentage of the typical service rate (e.g., has ordered 75% or less of the services required to achieve compliance with the associated MACRs) and statistically significantly lower than the average for the specialty (i.e., a statistically significant sample size is available in database 22 to obtain confidence at the p 0.1 level), then provider 16 failed the service rate test and is designated a NCQSP. A sample report representing the results of a service rate analysis is shown in FIG. 13.

[0041] The third test (the practice patterns test) involves an evaluation of the overall practice patterns of providers 16. More specifically, the McKesson clinical software is used to extract the number of occurrences of up-coding, ordering inappropriate services, vague or invalid diagnostic codes, and services that are performed too frequently, both for the particular provider 16 being evaluated, and for the specialty as a whole. Each practice pattern category is evaluated according to the method of the present invention to determine whether provider 16 practices in a manner that results in a practice patterns challenge rate that exceeds a predetermined multiple of the typical practice pattern percentages (e.g., 200% or more than the typical practice patterns) and is statistically significantly higher than the average percentages (e.g., at the p 0.01 level). If so, provider 16 failed the practice patterns test and is designated a NCQSP. A sample report representing the results of a practice patterns analysis is shown in FIG. 14.

[0042] According to the present invention, providers 16 that pass each of the three tests are assigned a QSP designation, indicating that providers 16 practice high quality medicine in a cost effective manner. As will be further described below, these QSP providers 16 are targeted by the present method for providing a maximum percentage of the overall healthcare consumed by patients 14 of employer 10. In addition to the basic QSP/NCQSP distinction resulting from the above-described process, providers 16 may be further ranked based on the results of the above-described tests. For example, the QSP category of providers 16 may be divided into A level QSP providers 16 and B level QSP providers 16. A level QSP providers 16 may be defined as providers 16 who have historical claims data in database 22 representing at least five episodes of the relevant type (sufficient episodic data), pass the CEI test with a CEI of less than 100% of the typical CEI in the specialty, and pass both the service rate test and the practice patterns test. B level QSP providers 16 may include providers 16 who (1) do not have sufficient episodic data, or (2) have sufficient episodic data and pass all three tests, but with a CEI of greater than or equal to 100% of the typical CEI in the specialty.

[0043] Similarly, providers 16 falling into the NCQSP category may be further ranked relative to one another to provide an ordered listing of NCQSPs. For example, C level NCQSP providers 16 may be defined as providers 16 who have sufficient episodic data, pass the CEI test, but fail one of the service rate or practice patterns tests (not both). D level NCQSP providers 16 may be defined as providers 16 who have sufficient episodic data and (1) fail the CEI test with a CEI of less than 150% of the typical CEI in the specialty or (2) fail both the service rate and practice patterns tests. Finally, an E level NCQSP provider 16 may be defined as a provider 16 with sufficient episodic data who fails the CEI test with a CEI that is at least 150% greater than the typical CEl in the specialty. Thus, providers 16 may be categorized in levels A through E. This ranking permits targeting not only QSPs, but A level and B level QSPs, or NCQSPs that at least have the best relative rankings on the list of NCQSPs.

[0044] Another example provider report 38 is shown in FIG. 15. Chart 154 of FIG. 15 is a listing of NCQSPs in descending order. The group of columns collectively assigned reference designation 156 identifies the providers 16 by ID, name, and location. Column 158 lists the number of episodes in database 22 associated with each provider 16. Column 160 lists the above-described CEI for each listed provider 16. The greater the CEI listed in column 160, the more significant the provider's deviation from the practice patterns of other providers 16 in the specialty. Consequently, those providers 16 listed near the top of chart 154 will provide healthcare resulting in a greater cost to employer 10. In one embodiment of the invention, listings of NCQSPs such as chart 154 are divided into thirds for purposes of practicing the invention as further described below.

[0045] Referring now to FIG. 16, a flow diagram of the above-described process for assigning QSP or NCQSP designations to providers 16 is shown. At step 162, claims information corresponding to a particular provider 16 is extracted from database 22 to determine whether provider 16 has sufficient episodic data (e.g., at least five episodes of the relevant type). If not, then provider 16 is designated an unknown, B level QSP. If provider 16 has sufficient episodic data stored in database 22, then each of the three above-described tests are performed as indicated at step 163. At step 164, the results of the CEI test are analyzed. If the CEI is 125% or more greater than the typical CEI in the specialty and satisfies the above-described statistical significance criteria, then provider 16 is marked as failing the CEI test (step 165). Otherwise, provider 16 is marked as passing the CEI test (step 166). Similarly, the results of the practice patterns test are analyzed at step 167. If provider 16 has a service challenge rate of 200% or more than the typical rate in the specialty and satisfies the above-described statistical significance criteria, then provider 16 is marked as failing the practice patterns test (step 168). Otherwise, provider 16 is marked as passing the practice patterns test (step 169). Finally, the results of the service rate test are analyzed in a similar manner at step 170, and provider 16 is marked as failing (step 171) or passing (step 172) the service rate test as a result of the analysis.

[0046] As shown at step 173, A level QSPs are identified as providers 16 who are marked as passing all three tests and achieved a CEI of less than 1. If a provider 16 is marked as passing all three tests, but has a CEI that is greater than or equal to 1, then provider 16 is designated a B level QSP as indicated by step 174. The remaining providers 16 are NCQSP providers 16. At step 175, the method of the present invention identifies C level NCOSPs at step 176 as providers 16 who are marked as passing the CEI test, but failing one of the other two tests (but not both). At step 177, the lowest level providers 16 (E level NCQSPs) are identified as providers 16 who are marked as failing the CEI test with a CEI of at least 1.5. Any remaining providers 16 are designated D level NCQSPs as indicated at step 161. D level NCQSPs include providers 16 who are marked as failing the CEI test, but with a CEI of less than 1.5, and providers 16 who are marked as failing both the service rate and practice patterns tests. This process of evaluating providers 16 for purposes of determining QSP/NCQSP status and levels within each category is repeated periodically to maintain an updated listing in database 22. It should be further understood that the particular numeric threshold values used in each of the three tests may readily be changed to affect the number of providers 16 falling into each of the five levels without departing from the principles of the invention. The designations for providers 16 resulting from the above-described process are used to improve the quality and cost-efficiency of the healthcare services consumed by employees 14 of employer 10 in the manner described below.

[0047] Referring now to FIG. 17, a flow diagram representing a portion of a method for optimizing healthcare services consumption is provided. At step 178, the claims information corresponding to patients 14 is extracted from database 22. Step 178 results in the data necessary to identify patients 14 having chronic illnesses (step 180) and to rank patients 14 according to the above-described risk stratification process (step 182). As indicated above and shown in FIG. 1, the method of the present invention, in one form thereof, involves intervention with patients 14 by registered nurses and other staff of HQM 13. This intervention or proactive coaching follows one or both of the two paths depicted in FIG. 17. First, for patients 14 identified as having one or more chronic illness, the method of the invention determines at step 184, based on claims information associated with such patients 14, whether the MACRs associated with the illness(es) have been satisfied. If the MACRs for a particular patient 14 have not been satisfied, then a representative of HQM 13 (e.g., a registered nurse or other staff member) contacts patient 14 to remind patient 14 of the need to schedule the healthcare necessary to satisfy the MACRs. This contact may be accomplished by any mode of communication including by phone, email, fax, mail, or any combination thereof. Preferably, the representative of HQM 13 has a live conversation with patient 14 to impress upon patient 14 the importance of satisfying the MACRs associated with the patient's chronic illness.

[0048] At step 188, the representative of HQM 13 may also contact provider 16 of healthcare services associated with the chronic illness(es) of patient 14. As a result of this contact, the representative enlists the cooperation of provider 16 in the effort to persuade patient 14 to satisfy the MACRs. As should be apparent from the foregoing, a goal of this intervention is to improve the health of patient 14 and minimize the cost to employer 10 by avoiding the increased healthcare expenses typically accompanying untreated chronic illnesses.

[0049] Steps 186 and 188 may result in the generation of a healthcare request. Specifically, patient 14 may respond to contact by the representative of HQM 13 by scheduling an evaluation by provider 16 or other action toward satisfying the MACRs associated with the chronic illness(es) of employee 14. Step 190 represents the possibility that a healthcare request is generated. If so, the healthcare request is processed as described below with reference to the second path depicted in FIG. 17. Otherwise, a predetermined time period is allowed to pass before repeating the process of checking the compliance of patient 14 with the MACRs associated with the chronic illness(es) of patient 14 and contacting patient 14 and provider 16. Step 192 indicates this delay period.

[0050] When healthcare requests are generated, either as a result of the first path of FIG. 17 described above, or simply during the ordinary course of employee healthcare consumption, HQM 13 receives the healthcare request at step 194. The healthcare request is associated with a particular patient 14 based on the risk stratification process represented by step 182. By determining the risk ranking of the requesting patient 14, HQM 13 can perform intervention actions (as described herein) in the order of ranking of patients 14. In other words, since it is not possible to contact every patient 14 submitting a healthcare request, the ranking of patients 14 permits HQM 13 to focus first on patients 14 having a highest risk ranking, and then (time and resources permitting) patients 14 have a smaller likelihood of generating high cost healthcare claims. The method of the present invention next accesses database 22 to determine whether provider 16 associated with the healthcare request is currently designated a QSP according to the process described above. If patient 14 is requesting to obtain healthcare services from a QSP, then the healthcare request may be processed according to conventional procedures without intervention by representatives of HQM 13 as indicated by step 196. Alternatively, the QSPs resulting from the above-described evaluation process may be ranked relative to one another and categorized into, for example, the A and B level QSP classifications described above. In such an alternative embodiment, an additional step (not shown) between step 196 and step 194 of contacting an patient 14 requesting healthcare from a B level QSP may be provided. At that step, a representative of HQM 13 may attempt to influence patient 14 to obtain such services from a A level QSP.

[0051] If, on the other hand, the healthcare request seeks services from a NCQSP, then the ranking of the NCQSP (derived as explained above with reference to FIG. 16) is determined at step 198. At step 200, a representative of HQM 13 contacts patient 14 who generated the healthcare request to urge patient 14 to obtain the requested services from a QSP. The representative may explain to patient 14 that various other providers 16 within geographic proximity to patient 14 (determined in the manner described below) have achieved the OSP designation for high quality, cost efficient healthcare, while provider 16 selected by patient 14 has not achieved that designation. The representative may further explain the implications of obtaining healthcare services from NCQSPs, and attempt to assist patient 14 in rescheduling the requested healthcare services with a QSP. Additionally, if patient 14 refuses to switch to a QSP, the representative may attempt to persuade patient 14 to at least switch to a NCQSP that is ranked at a higher level than the currently selected NCQSP.

[0052] As described above, at step 200 of FIG. 14, the representative of HQM 13 may list for patient 14 the variety of other providers 16 (specifically, QSPs) within a specific geographic proximity to patient 14. Such a list is generated by accessing database 22 using a software interface configured to permit the HQM 13 representative to input a desired radius extending from the location of patient 14, thereby defining an area of geographic proximity surrounding patient 14. The software accesses database 22, identifies the QSPs located within the selected geographic area, and provides a listing to the HOM 13 representative. Using this software and method, the representative may access listings of QSPs within, for example, a five mile, ten mile, and/or fifteen mile radius of patient 14.

[0053] In the event patient 14 refuses to obtain healthcare services from a provider 16 other than the currently selected NCQSP, the method of the present invention determines (at step 202) the level of intervention required to minimize the costs of such services while maintaining high quality healthcare and the specific actions associated with that intervention level. A plurality of actions may be taken by the representative of TPA 12, depending upon the level of intervention required. As described above, the NCQSP listings generated by the present invention may, for example, be divided into thirds (C, D, and E level NCQSPs). E level NCQSPs require the greatest level of intervention because the healthcare provided by such NCQSPs, as evaluated by the three QSP tests described herein, most significantly deviates from characteristics associated with desirable healthcare services. D level NCQSPs require less intervention. Finally, providers 16 designated C level NCQSPs require the least intervention. This stepped-down approach to intervention permits efficient usage of the resources available to HQM 13 in managing the healthcare expenses of employer 10.

[0054] As indicated above, providers 16 at the top third of a NCQSP listing (E level NCQSPs) receive the highest level of monitoring and individual contact by representatives of HQM 13. If an E level NCQSP is identified at step 198 of FIG. 17, then step 202 obtains a listing of intervention actions associated with E level NCQSPs. These actions may include the following:

[0055] (1) Obtain criteria for any admission associated with the healthcare request, including medical history, tests, and lab work;

[0056] (2) Delay any admission for employee 14 until all days of admission are approved by an appropriate representative of HQM 13;

[0057] (3) Complete a telephone evaluation with the NCQSP provider 16, conducted by an appropriate HQM 13 representative, to evaluate and discuss the need for any admission;

[0058] (4) Review the need to continue an admission after each day of the admission;

[0059] (5) Delay any additional days of admission beyond the initial length of stay until such additional days are approved by an appropriate representative of HQM 13;

[0060] (6) Assign a representative of HQM 13 to provide assistance to provider 16 in determining appropriate services to address the healthcare problem and to report treatments proposed by provider 16 to an appropriate representative of HOM 13; and

[0061] (7) Contact provider 16 directly to discuss any questionable proposed treatments as determined by an appropriate representative of HQM 13.

[0062] If a D level NCQSP is identified at step 198 of FIG. 17, then step 202 obtains a listing of intervention actions associated with D level intervention. These actions may include the following:

[0063] (1) Obtain criteria for any admission associated with the healthcare request, including medical history, tests, and lab work;

[0064] (2) Assign a one-day length of stay and perform daily concurrent review of additional days, requiring approval by an appropriate representative of HQM 13 as needed;

[0065] (3) Require provider 16 to send notifications of admissions to an appropriate representative of HQM 13;

[0066] (4) Complete a telephone consultation with provider 16, conducted by an appropriate representative of HQM 13, if deemed necessary by the representative of HQM 13; and

[0067] (5) Assign a representative of HQM 13 to provide assistance to provider 16 in determining appropriate services to address the healthcare problem and to report treatments proposed by provider 16 to an appropriate representative of HQM 13.

[0068] Finally, if a C level NCQSP is identified at step 198 of FIG. 17, then step 202 obtains a listing of intervention actions associated with a C level intervention. These actions may include the following:

[0069] (1) Obtain criteria for any admission associated with the healthcare request, including medical history, tests, and lab work;

[0070] (2) Assign a maximum two-day length of stay or less based on conventional length of stay guidelines, and perform daily concurrent review of additional days, requiring approval by an appropriate representative of HOM 13 as needed; and

[0071] (3) Assign a representative of HQM 13 to provide assistance to provider 16 in determining appropriate services to address the healthcare problem and to report treatments proposed by provider 16 to an appropriate representative of HQM 13.

[0072] All of the various intervention actions listed above are represented by steps 204 and 206 of FIG. 17. After all of the appropriate intervention actions have been completed, the healthcare request is fully processed. Additional healthcare requests may be received at step 194 and simultaneously processed.

[0073] By applying the resources of HQM 13 to intervene with those patients 14 presenting the greatest risk of generating high healthcare costs and providers 16 most likely to provide the least desirable healthcare, the method of the present invention may result in improvements to the healthcare consumption habits of patients 14 and to the practice patterns of providers 16, thereby resulting in an overall improvement of healthcare services consumed by patients 14 and cost efficiency realized by employer 10.

[0074] The foregoing description of the invention is illustrative only, and is not intended to limit the scope of the invention to the precise terms set forth. Although the invention has been described in detail with reference to certain illustrative embodiments, variations and modifications exist within the scope and spirit of the invention as described and defined in the following claims.