Economic Externalities of Healthcare Information Technology

A presidential executive order in 2004 called for widespread adoption of Electronic Health Records (EHRs) within 10 years. Proponents have shown this will lead to safe, affordable, and consumer oriented healthcare. Current EHR adoption has not met pace, with estimates indicating widespread EHR adoption to occur over a significantly longer period, with implementation costs and Return on Investment (ROI) listed among others, as the predominant limiting factors. A widespread EHR adoption plateau is expected, with entities unable or unwilling to adopt EHR. This will lead to incentive-based requirements to achieve widespread adoption, and the full potential of EHRs. This paper looks at externalities of health information technology between the major entities: payers, providers, and consumers. These externalities necessitate implementation of incentive based programs to achieve benefit equilibrium. Game theory is employed to model the behavior of these entities to capture the most equitable outcome. Prescriptive analysis is utilized to interpret and suggest optimal adoption behavior.

INTRODUCTION

During the late 1990's, American industry information technology investment averaged eight times higher per worker than within the healthcare industry. Despite the large amount of spending on health information technology, there are still issues relating to costs, errors, efficiency, and coordination of care, which reflect the limited saturation of health information technology within healthcare systems. For example, inefficient paper-based systems, inaccessible medical information during care, limited patient access to health information, misinterpreted hand-writing, and unavailable best treatment options, affect the current healthcare systems [1].

To counter these issues, a 2004 executive order was established, calling for widespread and nationwide adoption of Electronic Health Records (EHRs) by 2014. Estimates are less than 1 in 5 physicians have adopted EHRs, and in another survey regarding patient access of EHR, only a small number (1 in 25) had accessed their information. Small practices, or those with less than 10 physicians are seen to be the final adopters of EHRs. The tipping point, or time when the greatest number of adoptions will occur is estimated between 2009 and 2012. Mid-range estimates are 9 in 10 practices will adopt EHRs by 2024, a decade beyond the expected goal, with complete saturation unattained even then. Barriers to implementation include upfront costs, and savings that benefit third-party payers. In other parts of the world, EHR adoption has been spurred by governmental subsidizes to provider technology [2,3].

A widespread EHR adoption plateau is anticipated, with entities unable or unwilling to adopt EHRs. The management dilemma is to determine methods that will encourage health information technology adoption, specifically EHRs, and lead to healthcare industry cost and quality of care improvements. This process will lead to incentive-based requirements, which do not necessitate governmental subsidies for EHR adoption. This paper looks at the externalities involved in the adoption of EHR, and utilizes game theory to model outcomes.

LITERATURE REVIEW

EHRs

Healthcare stakeholders promote the use of HIT in terms of safe, affordable, and consumer oriented healthcare. This includes the avoidance of medical errors, improved use of resources, accelerated diffusion of knowledge, reduction in access variability, consumer role advancement, privacy and data protection, and public health and preparedness. EHRs have been shown to decrease billing issues, medical/drug errors, and improve patient health, use of medical evidence, cash flow/collections, paper cost, quality, safety, research, compliance, and preventative care. Personal Health Records (PHRs) are consumer-oriented information, which allows the consumer to better manage their healthcare. This includes data on medical history, medication, immunizations, allergies, etc. Typically these are seen as electronic systems integrated with or a component of EHRs. Other technologies that are integrated with or a component of EHRs include decision support and computerized physician order entry (CPOE) [4,5,6,7].

The need for quality measures is another reason for adoption of EHRs. Increasingly, providers are being measured on quality, and that quality is tied to pay for performance or other programs that directly affect the payment to the provider. EHR systems make the quality measures available for reporting and compliance requirements. In perhaps one of the largest EHR implementation case studies, Veterans Health Administration (VHA), showed improvements in employee-patient ration, and cost-per-patient decreases as compared with the US consumer price index (CPI) increase. Quality also improved through preventative screening measures, and disease management [8,9].

The president's plan for EHRs coincides with these benefits, with the intention to empower the patient and solve issues with regard to errors, quality, and increasing costs. By 2014, healthcare information is expected to be accessible for most Americans at the time and location of service, and patient participation will be voluntary. To meet the target, the following activities are taking place to speed the adoption rate: creation of standards, funding for demonstration projects, governmental adoption of HIT, and national leadership position [10].

Despite the political pressures and potential benefits, EHR adoption has been lacking. One recent study showed the relationship between EHR adoption and provider size, in which the adoption rate decreased as provider size in terms of full-time physicians decreased. The total cost of an EHR, which includes purchase of hardware, software, implementation, maintenance, training, customization and support for operational changes, remains high [4,6,11,12].

Resources, such as insufficient capital, return on investment (ROI), data use concerns, consumer education, and costs were listed as the top limiting factors and barriers to EHR adoption. While pressure increases to adopt EHRs, issues remain regarding costs, since a typical provider ROI cycle for an EHR application is three years. While benefits of EHR adoption extend to others, the provider must make the investment. Key stakeholders such as payers, consumers or patients, and physicians and practices should collaborate to ensure sufficient capital, and share in the risk/reward. Collaborative purchases and government incentives are expected to improve adoption and lower price points [4,5,7,11,12].

Incentive

While the healthcare industry spends less than others in comparison, electronic systems are still expensive. To encourage information technology adoption, some advocate direct or indirect incentives. Examples of direct rewards include regional grants and contracts, low-rate loans, pay for use, and an EHR system in exchange for data utilized for third-parties. Examples of indirect rewards are pay-for-performance. Regional grants and contracts would be used to promote EHR at local levels, with the hope of creating local and regional data exchanges. Some solutions intended to encourage adoption, such as pay for performance do not apply universally, and require localized incentives. Federal low-rate loans could also reduce the entry barrier for EHRs. Pay for use explores ways to provide reimbursement based on new codes/modifiers, or through direct incentives. Pay for performance would provide incentives for those practices with the highest quality not quantity, which is expected to be enhanced through EHR adoption [4,12,13,14].

Incentives are intended to provide adoption momentum in the market. Providers of care are not adopting at a high rate as a result of the cost-benefits, business process re-engineering, and potential legal barriers. It has been shown however that quality and efficiency can improve which leads to reduced medical errors and utilization. This creates the issue whether the market supports technology in terms of societal benefit. In terms of interoperability, support, and IT infrastructure, smaller practices have a larger need. Some suggest targeting larger providers, which require lower incentives, since they typically have a well established IT system and support structure. Once large providers adopt IT, smaller providers may follow. One study looked at targeting imminent adopters of EHRs. This includes those providers ready to implement within one year. The idea is to move the adoption curve forward, until the market is saturated, and requires others to implement. Larger practices or those with larger population centers had a higher percentage of adoption [12,14].

The government could also implement policy that adds to the business case of adoption. Various payers have tested different incentive approaches, but there is still a learning curve. Stakeholders are wary to commit incentives, until the case is proven in support of IT adoption. In addition to continued incentives, education such as policies, procedures, methodologies, analysis, etc. will be needed to ensure continued success. This information can be shared amongst the healthcare community to improve healthcare and IT adoption [12].

Providers currently assume the majority of the cost and risk of investment, and do not receive the full return on benefits. The costs and benefits factors may include practice size, specialty, geographic location, operational efficiency, affiliations, IT support, and market incentives. An agreed upon level of incentives, is one that compensates the additional cost of obtaining data, and is fair, equitable, attainable, and reviewed with increases. An incentive by only limited payers, generates a first adopter disadvantage, and lacks sufficient motivation. Payer adoption of incentives will necessitate standards to be created, in order to measure the EHR utilization. Past implementations have shown that physicians are often resistant to practice changes required by EHR. Given this, incremental approaches to adoption have been suggested. These include e-prescribing and online tools [12].

In a state of rising costs, Health Information Technology is commonly asked to show ROI, and may be challenging where indirect benefits are realized. In other cases, while providers typically make the investment, the benefits are realized by non-investing entities such as payers and consumers. This disincentive may influence safety and quality of care by providers [6].

Economics

Network externalities, specifically indirect network effects, are commonly referred to as chicken and egg problems. Current examples include broadband and 3G wireless. The demand is dependent on the infrastructure and the infrastructure dependent on the demand. In supply-side economics, the costs decrease while demands increase, with regard to increases in scale. These network effects produce multiple equilibriums, with adoption perpetuating network effects. The demand curve increases as the adopters increase, but then decreases as unwilling adopters are introduced. In perfect form, the two end equilibriums of supply and demand are stable, whereas the center equilibrium is the critical mass. Once this point is reached, the perpetuating positive loop occurs. In cases of network effects, it is advantageous for early adopters to receive a lower cost than later adopters due to the perceived value and to achieve critical mass. These network effects also perpetuate lock-in as well, with the switching costs involving the cooperation of others [15].

An externality is a cost or benefit that affects the entity external to the immediate production or consumption. For example, if a provider produces information in non-computerized format, this creates an additional cost for the payer in terms of processing. Likewise if payer processes and produces return information in a non-computerized format, this creates an additional cost for the provider. These are examples of negative production externalities. Additionally this creates a negative consumption externality for consumers of healthcare services, including patients of providers and subscribers of payers. Now consider if the opposite were true. If a provider produces information in computerized format, this creates a cost reduction for the payer. Likewise if payer processes and produces return information in a computerized format, this creates a cost reduction for the provider. This in turn creates a positive consumption externality for consumers through reduced costs in addition to improved quality [16].

The marginal private cost is the additional per unit amount incurred by the producer. In this case, the marginal private cost would be the additional cost of a healthcare service. In other words any support or costs associated with providing an episode of care. This can include but is not limited to the patient visit, claim processing, documentation, payment processing, etc. These functional units can be further divided into categories for the provider of services, payer of services, and the consumer of services. The marginal external cost is the additional per unit amount incurred by others than producer. These include the costs incurred by the payer or consumer as a result of the provider producing a service. This may also apply to the provider through output from the payer or consumer. The marginal social cost is the additional per unit amount incurred by all others and is the combination of the marginal private cost and marginal external cost. This may be generalized as the cost of healthcare. As healthcare costs continue to increase, these are a combination of the cost functions from the producers and consumers. The marginal private benefit is the additional per unit benefit realized by the consumer. The marginal external benefit is the additional per unit benefit realized by others than the consumer. The marginal social benefit is the additional per unit benefit realized by all others and is the combination of the marginal private benefit and marginal external benefit. Given this, we seek to reach the equilibrium between supply and demand. If the supply is equal to the marginal cost less any subsidy, and demand is equal to the marginal benefit, supply is equal to demand when the marginal benefit equals the marginal cost less any subsidy. In each of these cases as before, we consider EHRs capability to improve benefits to the producers, consumers, and society as a result of improved quality and cost of healthcare [16].

The productivity growth period of the 1990s was largely attributed to the prior IT investment. Innovation is often cyclical, and may arise from complementary products. When comparing past technological innovations to information innovations, the adoption time is much less for information; consider for example the gasoline engine compared with Internet adoption timelines. Often times, the benefits of the innovation may also take many years to reach their potential. With information, there is an increasing dependence on complements, or the value of combination. Total cost and benefit sharing is important, since the individual costs or profits may not affect one another. It is vital for the stakeholders to form a type of integration or collaboration, to maximize the benefits for all. With information technology, data can be utilized for individuals to contract on attributes of transactions that were previously unknown or unrealized. In one example in the movie rental industry, a model to charge a high fixed cost was phased out, with a low fixed cost introduced, and a share in the variable revenue. This reduced prices for consumers and increased revenues for producers. The information systems facilitated this shift, as it allowed all parties to accurately track and report on the transactional data. With diffusion of computerized methods, monitoring costs decrease and allow for efficient arrangements to be reached amongst parties [15].

These topics are used in conjunction with HIT and EHR adoption. Potential productivity growth realized by other industries has been largely unrealized in the healthcare industry due to the investment. EHR adoption is plagued by the same network externalities impacting other industries, where the demand for EHR has not yet reached the demand curve apex. The switching costs involve the cooperation of the existing market participants, namely the payers, providers, and consumers in EHR adoption. As EHR adoption has externality implications, it is vital for the participants to maximize the benefits involved. Transactional data standards exist, and can be easily tracked between participants, permitting successful agreements to be reached.

Game Theory

Game theory is utilized for its methodology in the construction and evaluation of decision problems. The game modeling process involves each player, and their decisions to be specified, taking into account each player's inclinations. The purpose of creating a game model is to allow an enhanced outlook on the problem at hand. This prescriptive application of game theory seeks to improve decision making strategy, and to provide direction on the best decision choices. The Nash equilibrium provides a strategy for each rational player to expect, so that no individual player can modify their strategy to achieve a higher payoff. This allows for rational players to adhere to the recommended strategic guidelines and for the same expectation from other players in the game [17].

Zang, Niv introduce a "costless" regulation, in which the entry cost is subsidized by the incumbent, thereby limiting governmental intrusion to a regulator rather than the cost bearer. This regulation model may be applied to the healthcare industry, in terms of the incumbent players subsidizing the adoption of HIT. This eliminates governmental interaction in terms of monetary support, and allows the players to work towards the most beneficial outcome [18].

MODEL AND RESEARCH DESIGN

Game Setup

Backward induction is explored to solve the game. First we consider the last move in the expected game outcome, which is EHR adoption. Taking this final move as a future activity, we make our way backwards in time to the start of the game, in order to determine the best set of moves the player(s) can make given the final move and game outcome. This game contains three players, or those agents who make a decision. Player I is the payer, player II is the provider, and player III is the consumer. All players are expected to act rationally, that is play the game in a manner which maximizes their payoff. The costs involved are referred to generally and may include direct costs such as computerization, or indirect costs such as quality. For EHR adoption, we initially consider a zero-sum game upon adoption; that is each player's outcome equalizes with one another. Otherwise, if the sum of payoffs were less than zero, there would be no incentive to adopt EHR, and if the payoffs were greater than zero, a player would not be maximizing their potential payoff in the game.

During the first stage of the game, the payer establishes a given subsidy to any provider willing to adopt an EHR. This subsidy is variable based on the number of covered consumers by the third-party payer and the estimated number of services and related savings through use of an EHR. This stage requires the third-party payer and provider establish a set number of covered services and expected savings over a set period of time, otherwise the payer could oversubsidize the provider, or the provider could be undersubsidized. This would be established through a contract or other such agreement. If there is a case such that the subsidy is greater than the expected marginal private benefit, the payer may choose not to subsidize.

The second stage determines whether the provider accepts the subsidy and adopts EHR. If there is a case where the entry cost is greater than the third-party subsidy plus the marginal private benefit, the provider may choose not to adopt. This stage also involves the EHR adopted provider's desire to move towards those payers willing to subsidize the EHR, as this will maximize the provider payoff. This will force those payers unwilling to subsidize into market entry, as it may result in a potential loss of provider coverage. The second stage endpoint will involve completion of set payer subsidies to providers until a "tipping point" is reached; this point is desired to be less than or equal to the maximum subsidy benefit level realized by the payer.

The third stage involves creation of a payer-based incentive to the consumer willing to participate with providers that have adopted an EHR. This is variable based on the number of consumers who have yet to adopt, and the savings available by switching that consumer. This would require some arrangement in terms of which consumer service locations were covered and the particular rate invoked. In a case where the consumer switching cost is greater than the benefit, the payer may choose not to subsidize.

The fourth stage determines whether the consumer receives services from an EHR adopted provider and accepts the subsidy or lowered payment rate. In order for the consumer to maximize their gains they would accept the subsidy with an EHR adopted provider; as this would provide the highest quality and lowest cost of care. Each player seeks to maximize their gains; the payer seeks to enable as many providers as possible, with the least subsidy. The provider seeks to achieve the highest amount of subsidy through covered payers. The consumer seeks to achieve the lowest cost, highest quality of care.

Two equilibriums are reached during the game. The first occurs when the payer subsidizes the provider to the extent that it is profitable. The second occurs when the payer subsidizes the provider indirectly through the consumer to the extent that it is profitable. The consumer subsidizes by paying a higher price to non-EHR provider services.

Following the possible programmatic modeling paths, each player would choose the move that would maximize their payoff. Initially, in the first decision, the payer (I) would always prefer not to subsidize the provider (II), in the second decision, the payer (I), would again always prefer not to subsidize the consumer (III). However in each case, either the provider (II), or the consumer (III) may only choose to adopt if a subsidy is offered, thereby reducing the payer’s (I) maximum payoff. The provider’s (II) decision will be whether or not to adopt. The provider (II) may choose to adopt without a subsidy, but would decrease their payoff. Additionally if the provider (II) chose not to adopt, efficiencies may be unrealized and the payoff would decrease. If the consumer (III) chooses the subsidy and the provider (II) does not adopt, the provider (II) may decrease their payoff through consumer services. The provider (II) would always prefer a subsidy to adopt. The consumer (III) may choose whether to accept a subsidy from a payer (I) for services rendered by the provider (II). In this case, the consumer (III) would always prefer a lower-cost, higher quality service. Using backward induction, based on rational players maximizing their payoff, the final move is for the consumer (III) to accept a subsidy provided by the payer (I) for the provider (II) services. The payer (I) would move to subsidize the consumer (III), the provider (II) would adopt, and the payer (I) would subsidize.

CONCLUSION

Given the preceding, a game theoretic model is developed to identify the optimal method for the game outcome, in this case EHR adoption, and may be utilized for improving adoption rates. A programmatic model, as with the example given can be employed universally or locally to systematically calculate cost, benefit, and subsidy, with the goal of maximizing the system benefit. The computerized systems permit for real-time tracking of these items for reporting and refinement purposes. Preference of outcome value is given to address current limiting factors, and based on these factors displays the prescriptive analysis for each player to maximize their benefit. This occurs until the end stage of the game and maximum EHR adoption has been achieved to the point where it is beneficial in terms of financial and qualitative costs. Future directions include examining choice behavior, and the factors that influence the player's decisions. In other words, to determine factors that may cause the player(s) to act irrationally, thereby minimizing their payoff. In addition to uncovering methods that mitigate, modify, or explain player's decision making behavior.

REFERENCES

1. N/A (2005). Fact Sheet: Improving Care and Saving Lives Through Health IT.

2. Ford, E.W., Menachemi, N., Phillips, T. M. (2006). Predicting the Adoption of Electronic Health Records by Physicians: When Will Healthcare be Paperless? Journal of the American Medical Informatics Association, 13.

3. Pyper, C., Amery, J., Watson, M., Crook, C., Thomas, B. (2002). Patients access to their online electronic health records. Journal of Telemedicine and Telecare, S2, 103-105.

4. Thompson, T.G., Brailer, D. J. (2004). The Decade of Health Information Technology:Delivering Consumer-centric and Information-rich Healthcare.

5. N/A (2006). Roundtable Summary Report: Personal Health Records and Electronic Health Records: Navigating the Intersections. The Agency for Healthcare Research and Quality.

6. Menachemi, N., Brooks, R.G. (2006). Reviewing the Benefits and Costs of Electronic Health Records and Associated Patient Safety Technologies. Journal of Medical Systems, 30, 156-168.

7. Ash, J.S., Bates, D.W. (2005). Factors and Forces Affecting EHR System Adoption: Report of a 2004 ACMI Discussion. Journal of American Medical Informatics Association, 12, 8-12.

8. N/A (2006). Importance of Quality Measures in Relationship to the EHR. Medical Strategic Planning.

9. Evans, D.C., Nichol, W.P., Perlin, J.B. (2006). Perspective Effect of the implementation of an enterprise-wide Electronic Health Record on productivity in the Veterans Health Administration. Health Economics, Policy and Law, 1, 163-169.

10. N/A (2004). Transforming Healthcare: The President’s Health Information Technology Plan.

11. Medical Group Management Association (MGMA) Center for Research, Medical groups’ adoption of electronic health records and information systems.

12. Connecting for Health (2004). Achieving Electronic Connectivity in Healthcare.

13. Manos, D. (2006). Vendor to Offer Web-based EHRs to Docs in Exchange for Data, in Healthcare IT News.

14. Ackerman, K. (2006). Changing the Target of EHR Incentive Programs. iHealthBeat.

15. Varian, H.R. (2003). Economics of Information Technology. NSF grant SES-9979852.

16. Bade, R., Parkin, M. (2002). Foundations of Microeconomics. CITY, OH: Pearson Education, Inc.

17. Turocy, T.L., Stengel, B. V. (2001). Game Theory. CDAM Research Report LSE-CDAM-2001-09.

18. Niv, M.B., Zang, I. (1999). Costless regulation of monopolies with large entry cost: A game theoretic approach. International Journal of Game Theory, 28, 35-52.

Definitive Source and Citation:

Woodside, Joseph M. (2007). Economic Externalities of Healthcare Information Technology. Journal of Health Information Management.

Available at: http://s3.amazonaws.com/rdcms-himss/files/production/public/HIMSSorg/Content/files/jhim/21_4/08_focus_ehr.pdf.