IT Capability for Health Care Delivery

Abstract

In recent years, there has been a proliferation of health information technologies (HITs) that promise to improve the delivery of care. Health care service providers are faced with an increasing push to develop electronic medical record (EMR) capability, which is the ability to leverage health IT to enable and link the clinical processes for an effective and efficient delivery of care. However, there is little guidance in the literature on the performance implications of EMR capability and whether providers should commit to higher stages of EMR capability. Based on data from 1,011 acute care providers in the United States, the findings of this study suggest that the operational performance of providers, measured as case mix index-adjusted discharges per licensed bed, is positively related to the stage of EMR capability. However, the findings also point to a cautionary insight—committing to higher stages of EMR capability may not be uniformly beneficial to all providers. The findings suggest that the choice of the stage of EMR capability is self-selected. Interestingly, if health care service providers were to be assigned to a higher stage of EMR capability (i.e., incommensurate with their idiosyncratic technological, organizational, and environmental characteristics), the potential operational performance benefit of that stage of EMR capability may remain unrealized.

Keywords

IT capability electronic medical record (EMR) capability operational performance self-selection health care delivery

Introduction

Health care delivery is one of the most critical and complex services, which has traditionally been characterized by disparate processes that often lack rudimentary information technologies (ITs; Garets and Davis 2006; Institute of Medicine [IOM] 2002). The inability to share relevant medical information across the various clinical processes in a timely manner represents a major impediment to progress toward improving the delivery of care (Grimson, Grimson, and Hasselbring 2000). ITs available now to health care service providers promise significant improvement in this regard (Agarwal et al. 2010; Ayal and Seidmann 2009; McCullough et al. 2010) and form the focus of this study.

Electronic Medical Record (EMR) capability of health care service providers is the ability to leverage ITs to enable and link the clinical processes for an effective and efficient delivery of care (Barua et al. 2004; Bharadwaj 2000; Pavlou and El Sawy 2006). EMR capability is not achieved all at once but is developed gradually in multiple stages (see Electronic Medical Record (EMR) Capability: The Stages section). The increasing levels of EMR capability suggest an increasing capability to capture, store, retrieve, share, monitor, and analyze patient-specific health care related data in a timely manner from disparate sources in the delivery of care (Brailer and Terasawa 2003) with the promise of improved operational performance relevant to health care delivery (Angst and Agarwal 2009; Angst, Agarwal, and Sambamurthy 2010; Hillstead et al. 2005; Palacio, Harrison, and Garets 2010). The growing significance of EMR capability in the delivery of care was acknowledged by the U.S. President in his State of the Union Address in 2004 and forms the basis for the federal government initiatives that followed as part of the Health Information Technology for Economic and Clinical Health Act of 2009, mandating health care service providers to develop EMR capability.

The development of EMR capability, however, has been fraught with a myriad of challenges and remains an underachieved goal (e.g., American Hospital Association [AHA] 2010; Blumenthal 2009; IOM 2002; Jha et al. 2009; Palacio, Harrison, and Garets 2010). While providers are faced with an increasing push to develop EMR capability (the federal government incentives attest to this drive), there is little guidance on whether providers should commit to higher stages of EMR capability (Agarwal et al. 2010; Chaudhury et al. 2006). There remains considerable ambiguity about the operational performance benefits of increasing levels of EMR capability (Jha et al. 2009; Tang et al. 2006), contributing to the slow pace of the development of highest stages of EMR capability among providers in the United States (e.g., AHA 2010; Blumenthal 2009; IOM 2002; Jha et al. 2009; Palacio, Harrison, and Garets 2010). Highlighting the paucity of research, specifically, “empirical research” related to these issues, Ayal and Seidmann (2009, p. 47) observe that “the literature lacks a quantification of the effects of the implementation of large health care IT systems.” In the same vein, Agarwal et al. (2010) highlight the need for research on the selection and evaluation of the operational performance implications of HITs. Addressing this need is central to the research agenda of this study. Specifically, the article is motivated by the questions:

What are the operational performance implications of increasing stages of EMR capability?

Should health care service providers commit to higher stages of EMR capability?

The remainder of the article is organized as follows. The second section provides the conceptual foundation of the study. This section describes the stages of EMR capability and the normative performance implications of increasing stages of EMR capability. The third section develops the theoretical foundation for self-selection into a stage EMR capability by health care service providers and the selection model used in the study. The fourth section discusses the variables, model specification, and data collection for the study. The fifth section presents the study results. The sixth section contains our concluding remarks.

Conceptual Development

Electronic Medical Record (EMR) Capability: The Stages

As is evident from the recent literature and the press (e.g., Angst et al. 2011; New York Times 2010; Palacio, Harrison, and Garets 2010), EMR capability is not achieved all at once. Instead, such IT capability is developed gradually. We build on the recent work in this area (e.g., Healthcare Information and Management Systems Society [HIMSS] 2009; Palacio, Harrison, and Garets 2010) to identify an ordered set of five EMR capability stages (i.e., Stages 0–IV). Each stage of EMR capability is identified based on the composition of HITs (see Table 1) implemented by a provider (HIMSS 2009; Palacio, Harrison, and Garets 2010). The ordered set of five EMR capability stages is reflective of an increasing capability to capture, store, retrieve, share, monitor, and analyze electronic medical information in a timely manner from disparate sources in the delivery of care. Each stage progressively provides the health care service provider timely information capture, storage, access, and analysis capabilities that serve to improve the delivery of care end to end (viz., diagnosis, treatment, prescription, evaluation, discharge, and follow-up).

Table 1.

Stages 0 to IV of EMR Capability.

EMR Capability	Health information technologies implemented
Stage 0	Starting stage—none or a subset of EMR systems implemented
Stage I	All ancillary information systems (laboratory, pharmacy, and radiology information systems)
Stage II	All of Stage I and clinical data repository, clinical documentation, and clinical decision support systems
Stage III	All of Stage II and picture archiving communication systems and medical records imaging
Stage IV	All of Stage III and computerized physician order entry

Note. EMR = electronic medical record.

Specifically, Stage I EMR capability is developed from the ancillary IT applications (related to laboratory, radiology, and pharmacy). These are the foundational technologies (Angst et al. 2011) with which the development of EMR capability can be initiated. Providers that have all three HITs implemented are deemed to have Stage I EMR capability, and those that do not belong to Stage 0. Stage I EMR capability is developed from technology applications that enable the diagnostic processes in the delivery of care. This stage provides physicians with easy access to results when needed, reduces the decision-making time by eliminating redundancy in medical tests, improves patient care owing to the timely diagnosis of medical problems, and improves task efficiency due to faster turnarounds (Bates et al. 2003; IOM 2002; Overhage, Suico, and McDonald 2001; Schiff et al. 2003; Tierney et al. 1987).

Stage II EMR capability is developed from the clinical data repository, clinical documentation (e.g., vital signs, flow sheets), and the clinical decision support systems in addition to those from Stage I capability. Specifically, Stage II capability enables the storage of patients’ medical information from the ancillary systems and provides decision support based on the information stored. This stage makes it possible to play a key role in the treatment processes including evaluating alternative treatment options in the delivery of care. For example, Stage II EMR capability enables effective physician decision making in drug prescription (selection, dosage, and interactions) based on patient medical information (Abookire et al. 2000; Bates and Gawande 2003; IOM 2002).

Stage III EMR capability is developed from the picture archiving communication systems (PACS) and the medical records imaging technologies in addition to the Stage II capability. Stage III capability provides the ability to store and retrieve medical imaging, a key element in the delivery of health care. This stage makes it possible to decrease the waiting times for radiology reports with the potential to improve accuracy of patient care, reduce the length of stays, and increase clinician’s efficiency (Ayal and Seidmann 2009; Chesson 2006; Ondategui-Parra et al. 2005).

Stage IV EMR capability is developed from a computerized physician order entry (CPOE) system in addition to the Stage III capability. With a CPOE system, physicians can use all of the relevant information available to prescribe medications electronically (Kuperman and Gibson 2003). Stage IV capability includes the provision for validating medication dosage and frequency (Bates and Gawande 2003), displaying relevant laboratories, and checking for drug allergy and drug-drug interactions (IOM 2002, p. 8; Queenan, Angst, and Devaraj 2011). This capability serves to enhance the treatment and prescription process by ensuring increased compliance to clinical standards, reducing medication errors from illegible medication orders, and improving clinical administration (Davidson and Chismar 2007; Mekhjian et al. 2002).

Thus, the Stages 0 to IV of EMR capability exemplify an increasing ability to capture, store, retrieve, share, monitor, and analyze patient-specific, care-related data from various clinical processes for effective delivery of care (Brailer and Terasawa 2003; Hillstead et al. 2005; Raghupathi 1997).¹ Next, we discuss the performance implications of developing the different stages of EMR capability.

EMR Capability Stages and Operational Performance

Health care service providers realize the potential performance benefits of a stage of EMR capability by leveraging the technologies relevant to that stage in their activities, routines, and processes (Barua et al. 2004; Goh, Gao, and Agarwal 2011; Ray, Barney, and Muhanna 2004; Zhu, Kraemer, and Xu 2006). The potential benefits of EMR capability may not be realized if the HITs do not enable the operational processes for health care delivery (Ayal and Seidmann 2009; Palacio, Harrison, and Garets 2010). Further, as Barua, Kriebel, and Mukhopadhyay (1995) highlight, the association between the IT and performance attenuates with distance between the cause and effect, and process-related operational performance metrics are likely to be more directly affected by EMR capability.

In this study, the operational performance of providers is assessed by way of case mix index (CMI)-adjusted Discharges per Licensed Bed, which measures the number of patients using a health care service provider’s beds during a given period, adjusted for the CMI to account for the severity of the patient mix and the heterogeneity in the patient population across providers (Angst et al. 2011; Grosskopf and Valdmanis 1993). While technologies relevant to each stage of EMR capability add unique information capabilities, increasing levels of EMR capabilities contribute to an increase in the operational performance by improving the workflow in the delivery of care (Agarwal et al. 2010; Samaan et al. 2009) in three key ways.

First, higher EMR capability levels enable the streamlining and automation of the individual processes at various steps of health care delivery. For example, the laboratory system in Stage I EMR capability enables improvements in an otherwise paper-based manual process of organizing, testing, recording, and communicating results by streamlining and automation the repetitive and routine departmental processes. The digitization of patient information in standard electronic format not only reduces the time needed to reenter patient information multiple times but also reduces clinical mishaps and medication errors due to illegible handwritten orders (Holstein 2009; Palacio, Harrison, and Garets 2010).

Second, higher EMR capability levels enable the timely exchange of patient information within and between the various technology-enabled processes in the delivery of care. As highlighted by the National Institutes of Health (NIH) report (2006, p. 7), if professionals at each step in the delivery of care “work from a data silo, each will have an incomplete picture of the patient’s condition.” Higher levels of EMR capability enable the linkage of the hitherto fragmented enabled processes central to the effective delivery of care (Angst et al. 2011; Kohli and Grover 2008). For example, when clinical ITs (such as decision support systems) are enabled to exchange information with storage technologies (such as medical imaging) and ancillary technologies (such as radiology or pharmacy), not only are the repetitive and routine processes automated, but there is also enhanced physician decision support for making rapid and informed decisions pertaining to the delivery of care. Such linkage of the workflow is central to the operational performance improvement of a health care provider (Ayal and Seidmann 2009).

Third, higher EMR capability levels enhance decision making by providing relevant and consolidated information at the point of care. For example, the decision support technologies enable physicians to identify potential drug interactions and the spread of illnesses and receive reminders of abnormal test results and real-time decision support including alternative treatment options, thereby facilitating effective delivery of care.

In sum, the above arguments suggest that increasing levels of EMR capability will enable increasingly efficient access to timely and relevant medical information from various processes in the delivery of care. Such improved workflow relevant to the delivery of care is associated with a generic improvement in the operational performance of health care service providers. This motivates the central hypothesis of the study:

Hypothesis 1: Ceteris paribus, the stage of EMR capability is positively related to the CMI-adjusted discharges per licensed bed of health care service providers.

Modeling the Selection of EMR Capability Stages

Theoretical Background

The previous section highlighted the operational performance benefits that may be realized from increasing levels of EMR capability. Note, however, that developing increasing levels of EMR capability requires allocation of organizational resources toward hardware and software technology, process redesign, employee training, and other relevant management initiatives, and yet, the realization of the intended performance outcomes is fraught with uncertainty and risk (Blumenthal 2009; Congressional Research Service (CRS) Report 2005; Devaraj and Kohli 2003; Jha et al. 2009; Palacio, Harrison, and Garets 2010). Also, prior research (e.g., Hamilton and Nickerson 2003; Shaver 1998; Thirumalai and Sinha 2012) suggests that managerial choices (e.g., selecting a stage of EMR capability) are based on not only the expected performance gains from the choice but also whether the choice is commensurate with their idiosyncratic conditions. Hence, it can be inferred that health care service providers’ choice of their stage of EMR capability is strategic and influenced by a host of idiosyncratic factors related to their technological, organizational, and environmental (TOE) contexts (Angst, Agarwal, and Sambamurthy 2010; Jha et al. 2009; Kazley and Ozcan 2007; Thirumalai and Sinha 2012; Zhu, Kraemer, and Xu 2006).

The technological context of a health care service provider includes the existing technology applications and the supporting technology infrastructure at the provider (Angst, Agarwal, and Sambamurthy 2010; Zhu, Kraemer, and Xu 2006; Zhu and Kraemer 2005). The role of the technological context in influencing the technology decisions of organizations is well recognized in the literature (e.g., Angst, Agarwal, and Sambamurthy 2010; Forman 2005; Premkumar, Ramamurthy, and Nilakanta 1994; Thirumalai and Sinha 2012; Zhu, Kraemer, and Xu 2006). In the same vein, the influence of idiosyncratic factors related to the organizational context such as people, processes, scale, and scope of operations, and other resources in the technology decisions of organizations is well recognized in the extant literature (e.g., Chatterjee, Grewal, and Sambamurthy 2002, Jensen 1994; Mata, Fuerst, and Barney 1995; McAfee 2002; Palacio, Harrison, and Garets 2010; Thirumalai and Sinha 2012; Zhu, Kraemer, and Xu 2003). Finally, the technology decisions of a health care service provider are also influenced by mimetic, coercive, and normative pressures from its institutional environment consisting of government and regulatory agencies, network partners, competing providers, patients, employer groups, suppliers, and insurance providers (Ang and Cummings 1997; Angst, Agarwal, and Sambamurthy 2010; Teo, Wei, and Benbasat 2003; Thirumalai and Sinha 2012).

Many of the idiosyncratic TOE factors discussed above are often unobserved, tacit, or private information to the provider (Chaney, Jeter, and Shivakumar 2004; Hamilton and Nickerson 2003; Li and Prabhala 2007; Shaver 1998; Thirumalai and Sinha 2012). This affects the empirical evaluation of the performance implications of the EMR capability stages, the central objective of this research. If providers self-select themselves into EMR capability stages appropriate to their idiosyncratic TOE contexts, it is imperative that the self-selection be accounted for in the empirical model (Ghosh and John 2009; Hamilton and Nickerson 2003; Heckman 1979; Maddala 1983; Masten 1993; Shaver 1998; Thirumalai and Sinha 2012). Consequently, in this research, we utilize an econometric model that takes into account the self-selection in estimating the effect of EMR capability stages on the providers’ performance.

Selection Model

The self-selection model involves two key steps. In the first step, the choice of a health care service provider to select into a stage of EMR capability is modeled as a function of its idiosyncratic TOE characteristics. In the second step, the normative effect of the chosen stage of EMR capability on its operational performance is evaluated accounting for the selection in the first step. This two-step model also serves as the foundation for the subsequent counterfactual analyses in the study. Below, we describe the selection model, performance evaluation in the presence of selection, and the counterfactual analyses.

Model Description

Step 1. EMR capability stage selection

This step models the choice of the stage of EMR capability by a health care service provider. The five stages of EMR capability (see Table 1) form the five ordered discrete choices. The choice is modeled as a continuous latent variable ${C a p a b i l i t y}_{i}^{*}$ , given by:

{C a p a b i l i t y}_{i}^{*} = δ Z_{i} + ν_{i}

where i denotes the health care service provider, δ is the vector of parameters to be estimated, Z_i represents factors that affect the choice of the stage of EMR capability, and ν _i represents the unobserved factors influencing the stage choice. The provider’s choice of the stage of EMR capability is based on its ${C a p a b i l i t y}_{i}^{*}$ value relative to four cutoff points: Cut₁, Cut₂, Cut₃, and Cut₄. Assuming the error term to be normally distributed with zero mean and unit variance, we can use an ordered probit model for the choice where the cutoffs are obtained as the maximum likelihood estimates (Idson and Feaster 1990; Maddala 1983) from selection Equation 1. For example, provider i would select Stage I if Cut₁ < Capability*_i < Cut₂. Next, we focus on the performance implications accounting for selection at each stage of EMR capability.

Step 2. Performance evaluation

Continuing with the example above, consider estimating the operational performance benefits of Stage I EMR capability. The key idea is to compare providers at Stage I EMR capability to those that are not at Stage I. However, as discussed earlier, the choice to belong to Stage 1 of EMR capability is not arbitrary—that is, there are some health care service providers that have selected into this stage and others (in Stage 0) who have not. Following the selection model described earlier, this choice of the provider can be modeled as a continuous latent variable Capability*_i given by Capability*_i = δZ_i + ν _i . The provider will choose Stage I if Capability*_i > 0, with corresponding probability Pr(Capability*_i > 0) = Pr(δZ_i + ν _i > 0). While Capability*_i is unobservable, Capability _ij , the actual choice j of the provider to select into Stage I, is observed. Assuming the error term to be normally distributed, the specification becomes a probit model.

Now, the performance outcome Π _ij of provider i under EMR capability stage choice j is given by Π _ij = βX_i + θ Capability _ij + ∊ _ij . The error terms in the performance equation is set to capture the effect of variables that are not identified in the specified X_i . However, if there are unobserved factors affecting the choice of both the stage and performance of EMR capability, then E(∊_ij |j) ≠ 0 which leads to biased estimates of the coefficients (Hamilton and Nickerson 2003). To that end, the expected performance of providers that select into Stage I of EMR capability can be corrected for the nonzero covariance of ∊ _ij and ν _i , owing to unobserved factors common to the selection and performance equations and respecified (Greene 2002) as:

E [Π_{i 1} | j = 1] = β X_{i} - ρ σ [\frac{φ (δ Z_{i})}{Φ (δ Z_{i})}]

Similarly, the expected performance of providers not selecting into Stage I EMR capability is obtained as:

E [Π_{i 0} | j = 0] = β X_{i} + ρ σ [\frac{φ (δ Z_{i})}{1 - Φ δ (Z_{i})}]

where $Φ (δ Z_{i})$ represents the cumulative density function of the standard normal distribution. The error terms ν _i and the ∊ _ij are assumed to be joint normally distributed so that expressions for E(∊ _ij |j) are tractable. The expected values for each error term are now zero due to inclusion of the second component in each equation referred to as the inverse Mills ratio which, in turn, enables the estimation of the treatment effect (Greene 2002) of the EMR capability stage. This component captures the unobserved idiosyncratic factors or private information at each provider (Li and Prabhala 2007). A significant selection term indicates the presence of unobserved idiosyncratic factors that not only are conducive to the choice of the stage of EMR capability but also impact the operational performance of health care service providers (Hamilton and Nickerson 2003).

Counterfactual Analyses

Following the estimation of the above equations for expected performance evaluation, “it is also important to estimate the mean values of the dependent variable for the alternative choice” (Maddala 1983, pp. 259–260). A comparison of the expected performance of providers in their self-selected choice relative to their expected performance in the nonpreferred alternative serves to highlight the performance implications of strategy choices (Chaney, Jeter, and Shivakumar 2004; Maddala 1991; Shehata 1991; Thirumalai and Sinha 2012). Such counterfactual analysis addresses the questions: What if health care service providers were assigned to a stage of EMR capability different from their selected stage? How would their (counterfactual) performance compare to the performance of providers that self-selected into that stage of EMR capability? We conduct this analysis at each stage of EMR capability.

Continuing with the example discussed earlier related to Stage I of EMR capability, the counterfactual performance of providers that did not actually select into Stage I of EMR capability, but if they had been assigned to Stage 1 of EMR capability can be estimated as follows (Ghosh and John 2009; Maddala 1983):

E [Π_{i 1} | j = 0] = β X_{i} + ρ σ [\frac{φ (δ Z_{i})}{1 - Φ δ (Z_{i})}]

This estimate can be compared to the observed performance of providers self-selecting into Stage I. In a similar manner, at each stage of EMR capability, we can compare the performance of providers self-selecting into that stage relative to the counterfactual performance of providers from lower stages if they had been assigned to that stage of EMR capability, in order to understand the implications of selection.

Research Design

Variables and Model Specification

Selection Equation

As discussed earlier, the selection of the stage of EMR capability by a health care service provider is influenced by the TOE characteristics of the provider. Below, we describe the TOE variables² identified in our analyses based on our review of the literature.

Technological context. Health care service providers usually employ a large number of administrative and medical technologies that support their strategic, administrative, and medical functions (Austin and Bowerman 2003). These administrative and medical technologies are grouped into various categories of IT applications such as Business Office, Financial Management, Human Resources, Decision Support, Managed Care, Medical Charts, and Medical Applications, based on their functionality and focus (Borzekowski 2002). Specifically, the Business Office technologies provide administrative support to patient registration, scheduling, billing, claims, and collection operations of the provider. The Financial Management technologies serve to support the resource planning, accounts payable, and materials management aspects of the provider operations. The Human Resources technologies are related to the employee payroll, time and attendance, personnel, and benefits administration applications at the provider. The Decision Support technologies are related to the cost and budgeting systems, outcomes and quality management systems, and executive information system applications at the provider. The Managed Care technologies are related to patient eligibility, contract management, and premium billing applications at the provider. The Medical Chart technologies constitute applications related to chart tracking, master patient indexing, abstracting, and encoding. The Medical applications include surgery, cardiology, obstetrical, intensive care, and bedside systems. As discussed earlier, ITs such as the administrative and medical technologies highlighted above that are currently employed by providers influence their technology decisions and their selection into a stage of EMR capability (Angst, Agarwal, and Sambamurthy 2010; Zhu, Kraemer, and Xu 2006). Hence, we account for the technology characteristics above in the specification of the selection model for health care service providers’ choice of the stage of EMR capability.

Organizational context. Variables related to the organizational context in the selection model include the Number of Licensed Beds, Profit Status, presence of Formal Steering Committee, Number of Executives in Steering Committee, Number of Nonexecutives in Steering Committee, chief information officer (CIO) Reporting Status, presence of Purchasing Subcommittee, and Service Type.

We include the Number of Licensed Beds at the acute care facility, a commonly used measure of organizational size. Accounting for size is important since it has been found to play a significant role in the adoption of a variety of improvement initiatives and technologies including total quality management programs, administrative innovations, magnetic resonance imaging technology (Friedman and Goes 2000; Kimberly and Evanisko 1981; Parente and McCullough 2009), integrated service digital networks, open systems technology, and electronic data interchange (Chau and Tam 2000; Premkumar, Ramamurthy, and Crum 1997).

We also account for the Profit Status of providers, a dichotomous variable that captures the type of governance (for profit or nonprofit). It is considered to be an important factor affecting not only performance but also technology-related decisions (Chesteen et al. 2005; Jha et al. 2009; McCullough et al. 2010; Palacio, Harrison, and Garets 2010; Shukla, Pestian, and Clement 1997).

Steering committees are high-level teams of representatives from across organizational subunits, which are entrusted with the task of linking IT strategy with business strategy (Nolan 1982). We account for the presence of a Formal Steering Committee and the membership structure of such committees (Number of Key Executives in Steering Committee and Number of Nonexecutives in Steering Committee) to account for their influence on technology-related decision making. Steering committees not only communicate the benefits and implications of IT to top management and user groups but also influence strategic IT planning and its effectiveness (Doll and Torkzadeh 1987; Drury 1984; Gupta and Ragunathan 1989; Reich and Benbasat 2000).

We account for the CIO Reporting Status to the chief executive officer (CEO) and the presence of a Purchasing Subcommittee to account for the influence of the top management in the IT planning. Support from top management and physicians has been found to be a key factor in the technology-related decision making (Friedman and Goes 2000). CIO reporting directly to the CEO, a practice by leading users of IT (Preston, Leidner, and Chen 2008; Smaltz, Sambamurthy, and Agarwal 2006), sends a visible signal to other organizational actors of the value of IT, synchronizes IT considerations with the business model of the organization, and places the CIO in a position of considerable power, a predictor of innovation in organizations (Preston, Karahanna, and Rowe 2006; Sobol and Klein 2009). Also, a Purchasing Subcommittee decides on issues relating to technology selection and its presence indicates top management's willingness to focus on IT (Abramson et al. 2009).

In health care settings, the information and the technology requirements vary with the type of care delivered (Angst, Agarwal, and Sambamurthy 2010). Hence, we control for the Service Type (general medical and surgery, academic, long-term acute, and others comprising critical access, pediatric, and orthopedic hospitals) in our analysis.

Environmental context

Health care service providers are influenced by policy makers, network members, and patients to develop higher stages of EMR capability. However, the influence of the participating network providers and alliance members may vary across providers. Variables related to the environmental context in the specification for the selection model include Number of Members in integrated health care delivery systems (IHDS), Member of Purchasing Group, and Member of Voluntary Purchase Alliance, which capture the institutional influence of the network and alliance members.

Membership in IHDS, or simply networks of health care service providers, significantly affects IT choices (Burns et al. 2001; Mousin, Remmlinger, and Weil 1999). As members of learning since member organizations share their experiences, starting with technology selection, and continuing through the implementation stage (Furukawa et al. 2008; McCullough 2005). The higher the Number of Members in IHDS, the greater the network learning; this affects the choice of technologies and the stage of EMR capability. Hence, we include this variable in our selection model.

Similarly, we consider Member of Purchasing Group and Member of Voluntary Purchasing Aalliance to account for the influence of communication channels in a health care service provider’s selection of the stage of EMR capability. Communication within and across organizations arising from exposure to professional societies plays a pivotal role in the selection of technologies, particularly in health care (Schneller 2000; Tabak and Jain 2000). Participation in strategic hospital alliances increases the number of communication channels available to the participating members (Burns and Lee 2008; Clement et al. 1997).

Performance Evaluation Equation

For performance evaluation and counterfactual analyses, CMI-adjusted Discharges per Licensed Bed is the dependent variable. Explanatory variables in the specification include those related to the technological context such as Medical Applications, Medical Charts, Total Networking Applications, Total Outsourced Services, Presence of Wireless and Intranet, and Usage; the organizational context such as service type and profit status; the environmental context such as the Number of Members in IHDS; and the inverse Mills ratio terms.

Medical Applications and Medical Charts include technologies implemented by the health care service provider, such as surgery systems, cardiology systems, nurse staffing systems, intensive care and point of care systems, master patient indexing, abstracting, encoding, and chart tracking/locator. It is conceivable that health care service providers implementing these technologies would be able to perform better than those that are not. We control for Medical Applications and Medical Charts to account for such variation.

The operational performance benefits realized from the different HITs are influenced by the availability of networking technologies, the availability of intranet within the organization, and the availability of wireless technologies on the premises of a health care service provider (Angst and Agarwal 2009; Zhu, Kraemer, and Xu 2006). Each of these, in turn, affects the operational performance. Hence, we consider the variables Total Networking Applications, Presence of Intranet and Wireless, and the Total Number of Outsourced Services in our estimation of the performance model. We also account for the usage of these technologies, an often overlooked factor (Devaraj and Kohli 2003), in our performance model. We assess Usage as the ratio of the number of technologies in use to the total number of technologies at a provider.

As mentioned earlier, Service Type refers to the type of health care (general medical and surgery, academic, long-term acute, or others comprising of critical access, pediatric, and orthopedic hospitals). We control for the Service Type to account for the variability in the process and outcomes across different types of care delivery. We also control for Number of Members in IHDS, as mentioned earlier, to account for the performance benefits realized by learning from the network partners. In the next subsection, we describe the data collection process for the variables discussed above.

Data Collection

The unit of analysis for this research is an individual health care service provider in a given physical location, and the sampling frame consists of acute care providers in the United States. The data for this study combine two databases: (i) The Dorenfest Complete IHDS+ Database and (ii) The Medicare Cost Reports Database. The Dorenfest Complete IHDS+ Database, collected by Dorenfest and Associates (now part of HIMSS), contains detailed information about the health care service providers associated with 1,444 IHDS in the United States.

To construct the study sample, we started with data from the Dorenfest database. For each health care service provider, we tabulated the information on their identifier, demographics, and technology characteristics. Based on this information, we also identified the stage of EMR capability at each provider, as discussed earlier. In the next step, the data were linked with the cost reports data set using medicare identification numbers. We were able to match approximately 80% of available health care service providers from the Dorenfest database with the cost reports data set. The cost reports enabled us to compute the operating performance measure for individual acute care providers included in our data set. Overall, this information was collected for 1,011 individual acute care providers. In the next section, we discuss the results of the empirical analyses. The empirical analyses in the study were conducted using the statistical software STATA.

Results and Discussion

Estimation of the Selection Equation

Table 2 presents the results from the estimation of the ordered probit model representing the selection of the five stages of EMR capability. First, we note that the coefficients of variables related to the technological context such as Business Office, Decision Support, Medical Charts, and Medical Applications are positive and significant. This suggests that health care service providers that have implemented these technologies are more likely to select into higher stages of EMR capability. This result is consistent with the findings of the prior studies (e.g., Angst, Agarwal, and Sambamurthy 2010; Forman 2005; Zhu, Kraemer, and Xu 2006). Second, variables related to the organizational context such as Number of Licensed Beds, Profit Status, CIO Reporting Status, Number of Nonexecutives in the Steering Committee, and Service Type have a significant association with the choice of the stage of EMR capability by providers. Specifically, the positive and significant coefficient for Number of Licensed Beds suggests that the larger sized providers are more likely to select into higher stages of EMR capability. This may be attributed to not only the availability of resources at the larger sized providers but also the need for coordination to operate in a large, complex health care supply chain, as is acknowledged in the extant literature (e.g., Chau and Tam 2000; Friedman and Goes 2000; Palacio, Harrison, and Garets 2010; Parente and McCullough 2009). The coefficient of Profit Status is negative and significant. This suggests that nonprofit providers are more likely to select into higher stages of EMR capability. Further, the significant negative coefficients of the Number of Nonexecutives in the Steering Committee and the CIO Reporting Status highlight the influence of considerations from multiple organizational constituencies in the selection of the stage of EMR capability. For example, as discussed earlier, while nonexecutives in the steering committee may bring to bear user (physicians and nurses)-related issues, the involvement of the CIO may serve to highlight the IT considerations in developing higher stages of EMR capability (e.g., Doll and Torkzadeh 1987; Drury 1984; Gupta and Ragunathan 1989; Smaltz, Sambamurthy, and Agarwal 2006). Third, we note that variables related to the environmental context such as Number of Members in IHDS and Membership in a Voluntary Purchase Alliance are significantly associated with the choice of the stage of EMR capability. This highlights the influence of the knowledge spillovers and learning from the institutional environment (i.e., the partners in the health care supply chain) in the selection of the stage of EMR capability by a provider (Agarwal et al. 2010; Tabak and Jain 2000).

Table 2.

Stage of EMR Capability and the TOE Factors.

Variable^a	Coefficient	SE
Business Office	0.261**	0.091
Financial Management	0.001	0.104
Human Resources	0.067	0.075
Decision Support	0.375**	0.050
Managed Care	0.028	0.053
Medical Charts	0.123**	0.050
Medical Applications	0.406**	0.029
Number of Licensed Beds	0.340**	0.056
Profit Status	−0.688**	0.127
Formal Steering Committee	−0.218	0.156
Number of Key Executives in Steering Committee	0.021	0.025
Number of Nonexecutives in Steering Committee	−0.072**	0.018
CIO reports to CEO	−0.173*	0.100
Presence of Purchasing Subcommittee	−0.193	0.127
Number of Members in IHDS	0.004**	0.001
Member of Purchasing Group	−0.125	0.100
Member of Voluntary Purchasing Alliance	−0.287**	0.131
Wald χ²		680.72**
Log-likelihood (n)		−794.97 (1011)

Note. EMR = electronic medical record; TOE = technological, organizational, and environmental; CIO = chief information officer; CEO = chief executive officer; IHDS = integrated health care delivery systems.

^a Dependent variable: stage of EMR capability. Model additionally contains a control for service type.

*p < .10. **p < .05.

Overall, these results suggest that the selection of the stage of EMR capability by a health care service provider is significantly influenced by the TOE characteristics of the provider.

Estimation of the Performance Evaluation Equation

Table 3 reports the results from the analysis evaluating the operational performance implications of increasing levels of EMR capability, accounting for the selection into each stage of EMR capability. The focus of this analysis is on the coefficient of the EMR capability level and the tests for self-selection. As discussed earlier in the model specification, the analysis additionally includes control variables to account for extraneous sources of heterogeneity across providers that may affect performance outcomes. The coefficients of the control variables are indicative of the extent to which the control variables account for the variation in the operational performance of providers in a given stage of EMR capability relative to that of providers in the next lower level of EMR capability. The results from the analysis indicate that while a few control variables including Medical Applications, Profit Status, Number of Members in IHDS, and Usage are associated with operational performance of providers belonging to various stages of EMR capability, they do not exhibit a consistent pattern of association with the operational performance. Broadly, this result suggests that the incremental variation in the operational performance of providers between successive EMR capability levels that may be attributed to the control variables is not significant. The results related to the selection term and the coefficient of EMR capability level highlight two key insights.

Table 3.

EMR Capability Stages and Operational Performance.

Variable^a	Stage 0	Stage I	Stage II	Stage III	Stage IV
Medical Applications	4.6546**	0.8647	1.0234	0.5855	1.6372
Medical Charts	0.3069	−0.8828	1.5520	2.9889	5.8185
Networking technologies	0.5841	0.3619	0.3619	0.5233	0.1431
Outsourced IT Services	0.8491	−0.3515	0.9531*	1.0813**	1.4063
Wireless	5.0146	−2.0113	3.4947	2.9483	−0.0356
Intranet	2.5083	−2.4148	−0.0434	0.9366	14.893*
Number of Members in IHDS	−0.2125**	−0.0514	0.0030	0.0181	−0.2072
Usage	69.682*	−0.0013	9.6771	32.259	−274.8**
Profit Status	11.440	7.1580	−14.746**	−4.8991	−7.3504
EMR Capability		37.37**	11.75	38.70**	39.65**
Test for Selection (χ²)		18.91**	2.75*	50.34**	5.42**
Coefficient of Selection (ρ × σ)		−17.59**	−10.82	−21.77**	−22.27**

Note. EMR = electronic medical record; IHDS = integrated health care delivery systems; IT = information technology.

^a Dependent Variable—case mix index (CMI)-Adjusted Discharges per Licensed Bed. Specifications additionally contain a constant and control for service type.

*p < .10. **p < .05.

First, we find that the tests for independence of the selection and performance equations (chi-square statistic) are significant. This suggests that in addition to the observed TOE characteristics of a provider, there are significant unobserved idiosyncratic factors of a provider that affect not only the choice of the stage of EMR capability but also the performance benefits realized from the selected stage. This highlights the private information (Li and Prabhala 2007) relevant to providers (e.g., by way of people, processes, knowledge, and related path dependencies), which is associated with the systematic self-selection into the various stages of EMR capability. It is worth noting here that the coefficient of the selection term is negative. This suggests that the ordinary least squares (OLS) estimate of the treatment effect of EMR capability that does not account for the unobserved, idiosyncratic characteristics (that is private information to the provider) will underestimate the benefits of EMR capability.

Second, results of the analyses also highlight the treatment effect of the multiple stages of EMR capability. Specifically, we find that Stages I to IV of EMR capability have a positive coefficient with statistically significant values for Stages I, III, and IV of EMR capability. These coefficients are interpreted relative to the next lower EMR level. The coefficients indicate that providers at Stage I EMR capability realize significantly higher operational performance than Stage 0 providers, providers at Stage III EMR capability realize significantly higher operational performance than Stage II providers, and providers at Stage IV EMR capability realize significantly higher operational performance than Stage III providers. This suggests that higher stages of EMR capability are generally associated with higher operational performance by way of CMI-adjusted Discharges per Licensed Bed, supporting the study hypothesis.

Counterfactual Analyses

Results of the counterfactual analyses for Stages I to IV of EMR capability, depicted in Figure 1, reveal a notable pattern. Specifically, the counterfactual analyses suggest that if providers in Stage 0 are assigned to Stage I, their operational performance, measured as CMI-adjusted Discharges per Licensed Bed, under the counterfactual option would be 48.02, while the performance of providers who actually selected into Stage I is 57.0. In other words, if providers that did not choose Stage I are assigned to Stage I, their performance would be lower than the performance of providers who actually self-selected into Stage I (48.02 < 57.0, p < .01), indicating their inability to realize the potential operational performance benefits of Stage I. The counterfactual analyses at Stages II, III, and IV reveal a similar trend. The counterfactual estimates at Stage II suggest that if providers from lower stages that did not select Stage II are assigned to Stage II, their performance would be lower than the performance of providers who actually self-selected into Stage II (48.23 < 61.69, p < .01), indicating that the potential operational performance benefits of Stage II would remain unrealized for them. Likewise, the analyses suggest that if providers that did not select Stage III are assigned to Stage III, their performance would be lower than the performance of providers who actually self-selected into Stage III (55.57 < 75.78, p < .01), indicating that the potential operational performance benefits of Stage III would remain unrealized for them. Similarly, the counterfactual estimates at Stage IV suggest that if providers that did not select Stage IV are assigned to Stage IV, their performance would be lower than the performance of providers who actually self-selected into Stage IV (56.34 < 83.12, p < .10), indicating that the potential operational performance benefits of Stage IV would remain unrealized for them. Below we discuss why providers assigned to a stage of EMR capability that is not self-selected may fail to realize the potential benefits of that stage of EMR capability.

The counterintuitive findings above reflect the self-selection into EMR capability stages and the significant role of the idiosyncratic characteristics of providers. As noted earlier, realizing the potential operational performance benefits of EMR capability stages requires leveraging the relevant HITs in their activities, routines, and processes (Barua et al. 2004; Bharadwaj 2000; Devaraj and Kohli 2003; Fichman and Kemerer 1999; Goh, Gao, and Agarwal 2011; Ray, Barney, and Muhanna 2004). However, this ability of a health care service provider to leverage the ITs is a key component of its absorptive capacity and specific to the observed and unobserved idiosyncratic TOE characteristics of the provider. For example, organizational factors including proprietary routines, processes, and equipment; organizational reconfiguration and process reengineering efforts at the provider; and the social construction (such as the hierarchy, culture, and reward structure) at the provider affects its ability to assimilate and benefit from the HITs. Similarly, technological characteristics of the providers including the compatibility of their medical systems and ITs; network infrastructure; and technology support for users may affect the assimilation of the HITs into their operational routine. Also, environmental factors including providers’ interactions with their network partners, insurance providers, regulatory agencies, and patients affect their ability to implement and benefit from HITs. Developing a technological, organizational, and institutional environment conducive to the assimilation of HITs, developing the absorptive capacity and the path dependencies related to problem solving and learning are not easily achieved in the short term (Barua et al. 2004; Bharadwaj 2000; Cohen and Levinthal 1990; Devaraj and Kohli 2003; Ray, Barney, and Muhanna 2004; Zahra and George 2002).

The self-selection of a provider into a stage of EMR capability and the normative benefits realized from the EMR stage is reflective of these (observable and unobservable) idiosyncratic characteristics (Chaney, Jeter, and Shivakumar 2004; Ghosh and John 2009; Hamilton and Nickerson 2003; Li and Prabhala 2007; Shaver 1998). Health care service providers normatively self-select into a stage of EMR capability that is commensurate with their idiosyncratic characteristics, and by doing so realize the potential benefits of that stage of EMR capability (Ang and Cummings 1997; Chaney, Jeter, and Shivakumar 2004; Ghosh and John 2009; Hamilton and Nickerson 2003; Shaver 1998). However, when the stage of EMR capability is not self-selected, that is, when a provider is assigned to a stage of EMR capability that is incommensurate with the provider’s idiosyncratic considerations (e.g., owing to regulatory forces), the lack of appropriate processes, routines, knowledge, and employee skills, along with the related path dependencies highlighted above results in the health care service provider failing to realize the potential benefits of that stage of EMR capability (Ray, Barney, and Muhanna 2004). In other words, the idiosyncratic characteristics of a health care service provider, such as those alluded to earlier, inform not only the provider’s choice of a stage of EMR capability, but also the extent to which providers realize the performance potential of EMR capability.

The results of the empirical analyses discussed above provide three key insights related to EMR capability of health care service providers: First, the results suggest that the operational performance of health care service providers, measured as CMI-adjusted Discharges per Licensed Bed, generally improves with the stage of EMR capability. This result provides support for the central hypothesis in the study—that is, enabling the various processes in the delivery of care with HITs and linking the processes to develop EMR capability facilitates information sharing and effective decision making related to the delivery of care, resulting in higher operational performance.

Second, the choice of the stage of EMR capability by a health care service provider is influenced by idiosyncratic factors related to its TOE contexts of the provider. Further, we also find that while operational performance may be higher at higher stages of EMR capability, some providers nevertheless select into a lower stage, that is, providers systematically self-select into different stages of EMR capability.

Third, findings from the counterfactual analysis in the study also indicate that if health care service providers were to develop a higher stage of EMR capability that is incommensurate with their idiosyncratic characteristics, the potential operational performance benefits of that stage of EMR capability would remain unrealized. This, in turn, provides credence to the concerns of health care service providers (e.g., AHA 2010) echoed in the quote below from the New York Times (June 7, 2010):

Doctors and Hospitals Say Goals on Computerized Records Are Unrealistic … the president’s all-or-nothing approach could discourage efforts to adopt electronic health records because some of the proposed standards are impossibly high and the risk of failure is great. They pleaded with the administration to take a more gradual approach and reward incremental progress …

Conclusion

The findings of this study make a threefold contribution toward advancing the literature:

Our study provides a theoretically grounded empirical assessment of the operational performance benefits of increasing stages of EMR capability. The empirical analysis is based on the data from a large sample (n = 1,011) of health care service providers in the United States. The significance of this contribution is further heightened given that while the emerging literature on EMR has predominantly viewed EMR capability from a dichotomous (yes or no) standpoint, this study is among the first to explicitly acknowledge the multiple stages of EMR capability in evaluating the performance implications.

Recently, it was acknowledged that “research providing insights on HIT selection and how to optimally execute the complex set of trade-offs involved in selection would be extremely valuable” (Agarwal et al. 2010, p. 9). To that end, based on a synthesis of multiple theoretical perspectives and empirical analyses, this study highlights the systematic self-selection by health care service providers into different stages of EMR capability commensurate with their idiosyncratic TOE characteristics.

Our study underscores the implications of prematurely developing a higher stage of EMR capability for health care service providers. Specifically, the study findings suggest that health care service providers developing a higher stage of EMR capability that is incommensurate with their idiosyncratic TOE characteristics are unlikely to realize the potential benefits of that stage of EMR capability. In other words, simply incentivizing health care service providers to move up the stages of EMR capability may not lead to the realization of the potential benefits of the higher stages of EMR capability. The practical implication of this finding is that health care service providers need to assess whether their choice of a stage of EMR capability is commensurate with their idiosyncratic TOE characteristics before committing to a stage of EMR capability. Developing EMR capability and, more importantly, the ability to realize the potential operational performance benefits of the EMR capability requires an informed and tailored approach and not a one-size-fits-all approach.

From a policy standpoint, improving the delivery of health care as envisioned in the U.S. Federal Government’s plan³ requires a further understanding of the TOE enablers and barriers to the development of EMR capability and realizing the operational performance benefits of HITs. Devoid of such understanding, pure regulatory forces would be ineffective in realizing the potential performance benefits of the HITs and improving health care delivery in the United States.

Footnotes

Appendix

Acknowledgment

The authors would like to thank HIMSS (Healthcare Information and Management Systems Society) Analytics for access to the HIMSS Analytics Database for research purposes. Support by way of a grant from the Association for Healthcare Resource & Materials Management (AHRMM) of the American Hospital Association (AHA) is gratefully acknowledged.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Notes

References

Abramson

Hyman

Osorio

S. N.

Kaushal

(2009), “Implementing a Patient Safety and Quality across Two Merged Pediatric Institutions,” Joint Commission Journal of Quality and Patient Safety, 35 (1), 43–48.

Abookire

S. A.

Teich

J. M.

Sandige

Paterno

M. D.

Martin

M. T.

Kuperman

G. J.

Bates

D. W.

, (2000), “Improving Allergy Alerting in a Computerized Physician Order Entry System,” Proceedings of the AMIA Symposium, Boston, MA.

Agarwal

Gao

DesRoches

Jha

A. K.

(2010), “The Digital Transformation of Healthcare: Current Status and the Road Ahead,” Information Systems Research, 21 (4), 796–809.

American Hospital Association (2010), “Letter to the Secretary of the Department of Health and Human Services,” (accessed May 3, 2010), [available at http://www.aha.org/aha/letter/2010/100503-cl-cms-0033-p.pdf].

Ang

Cummings

L. L.

(1997), “Strategic Response to Institutional Influences on Information Systems Outsourcing,” Organization Science, 8 (3), 235–256.

Angst

C. M.

Devaraj

Queenan

Greenwood

(2011), “Performance Effects Related to the Sequence of Integration of Healthcare Technologies,” Production Operations Management, 20 (3), 319–333.

Angst

C. M.

Agarwal

Sambamurthy

(2010), “Social Contagion and Information Technology Diffusion: The Adoption of Electronic Medical Records in US Hospitals,” Management Science, 56 (8), 1219–1241.

Angst

C. M.

Agarwal

(2009), “Adoption of Electronic Health Records in the Presence of Privacy Concerns: The Elaboraiton Likelihood Model and Individual Persuasion,” MIS Quarterly, 33 (2), 339–370.

Austin

C. J.

Bowerman

S. B.

(2003), Information Systems for Health Care Management. Chicago, IL: Health Administration Press.

10.

Ayal

Seidmann

(2009), “An Empirical Investigation of the Value of Integrating Enterprise Information Systems: The Case of Medical Imaging Informatics,” Journal of Management Information Systems, 26 (2), 43–68.

11.

Barua

Konana

Whinston

A. B.

Yin

(2004), “An Empirical Investigation of Net-Enabled Business Value,” MIS Quarterly, 28 (4), 584–620.

12.

Barua

Kriebel

C. H.

Mukhopadhyay

(1995), “Information Technologies and Business Value: An Analytic and Empirical Investigation,” Information Systems Research, 6 (1), 3–23.

13.

Bates

D. W.

Gawande

A. A.

(2003), “Improving Safety With Information Technology,” The New England Journal of Medicine, 348 (25), 2526–2534.

14.

Bates

D. W.

Ebell

Gotlieb

Zapp

Mullins

H. C.

(2003), “A Proposal for Electronic Medical Records in U.S. Primary Care,” Journal of the American Medical Informatics Association, 10 (1), 1–10.

15.

Bharadwaj

A. S.

(2000), “A Resource-Based Perspective on Information Technology Capability and Firm Performance: An Empirical Investigation,” MIS Quarterly, 24 (1), 169–193.

16.

Blumenthal

(2009), “Stimulating the Adoption of Health Information Technology,” The New England Journal of Medicine, 360 (15), 1477–1479.

17.

Borzekowski

(2009), “Measuring the Cost Impact of Hospital Information Systems: 1987–1994,” Journal of Health Economics, 28 (5), 938–949

18.

Brailer

D. J.

Terasawa

E. L.

(2003), “Use and Adoption of Computer-based Patient Records in the United States: A Review and Update,” California Health Care Foundation. March 28, 2003, (accessed May 3, 2010), [available at http://www.chcf.org/*/media/MEDIA%20LIBRARY%20Files/PDF/U/PDF%20UseAdoptionComputerizedPatientRecords.pdf].

19.

Burns

L. R.

Lee

(2008), “Hospital Purchasing Alliances: Utilization, Services, and Performance,” Health Care Management Review, 33 (3), 203–215.

20.

Burns

L. W.

Walston

S. L.

Alexander

J. A.

Zuckerman

H. S.

Andersen

R. M.

Torrens

P. R.

Hilberman

(2001), “Just How Integrated are Integrated Delivery Systems? Results from a National Survey,” Health Care Management Review, 26 (1), 20–39.

21.

Chaney

P. K.

Jeter

D. C.

Shivakumar

(2004), “Self-Selection of Auditors and Audit Pricing in Private Firms,” The Accounting Review, 79 (1), 51–72.

22.

Chatterjee

Grewal

Sambamurthy

(2002), “Shaping Up for E-Commerce: Institutional Enablers of the Organizational Assimilation of Web Technologies,” MIS Quarterly, 26 (2), 65–89.

23.

Chau

P. Y. K.

Tam

K. Y.

(2000), “Organizational Adoption of Open Systems: A Technology-Push, Need-pull Perspective,” Information & Management, 37 (5), 229–239.

24.

Chesson

(2006), “Radiology is Streamlining,” Health Imaging & IT, 8 (Summer), 36–39.

25.

Chesteen

Helgheimb

Randall

Wardell

(2005), “Comparing Quality of Care in Non-Profit and For-Profit Nursing Homes: A Process Perspective,” Journal of Operations Management, 23 (2), 229–242.

26.

Chaudhury

Wang

Maglione

Mojica

Roth

Morton

S. C.

Shekelle

P. G.

(2006), “Systematic Review: Impact of Health Information Technology on Quality, Efficiency, and Costs of Medical Care,” Annals of Internal Medicine, 144 (10), 742–752.

27.

Clement

J. P.

McCue

M. J.

Luke

R. D.

Bramble

J. D.

(1997), “Strategic Hospital Alliances: Impact on Financial Performance,” Health Affairs, 16 (6), 193–203.

28.

Cohen

W. M.

Levinthal

D. A.

(1990), “Absorptive Capacity: A New Perspective on Learning and Innovation,” Administrative Science Quarterly, 35 (1), 128–152.

29.

Congressional Research Service (CRS) Report for Congress (2005), “Health Information Technology: Promoting Electronic Connectivity in Health Care,” Information Systems, 16 (4), 121–143.

30.

Davidson

E. J.

Chismar

W. G.

(2007), “The Interaction of Institutionally Triggered and Technology-Triggered Social Structure Change: An Investigation of Computerized Physician Order Entry,” MIS Quarterly, 31 (4), 739–758.

31.

Devaraj

Kohli

(2003), “Performance Impacts of Information Technology: Is Actual Usage the Missing Link?” Management Science, 49 (3), 273–289.

32.

Doll

W. J.

Torkzadeh

(1987), “The Relationship of MIS Steering Committees to Size of Firm and Formalization of MIS Planning,” Communications of the ACM, 30 (11), 972–978.

33.

Drury

D. H.

(1984), “An Evaluation of Data Processing Steering Committees,” MIS Quarterly, 8 (4), 257–265.

34.

Fichman

R. G.

Kemerer

(1999), “The Illusory Diffusion of Innovation: An Examination of Assimilation Gaps,” Information Systems Research, 10 (3), 255–275.

35.

Forman

(2005), “The Corporate Digital Divide: Determinants of Internet Adoption,” Management Science, 51 (4), 641–654.

36.

Friedman

L. H.

Goes

J. B.

(2000), “The Timing of Medical Technology Acquisition: Strategic Decision Making in Turbulent Environments,” Journal of Health Care Management, 45 (5), 317–331.

37.

Furukawa

M. S.

Raghu

T. S.

Spaulding

T. J.

Vinze

(2008), “Adoption of Health Information Technology for Medication Safety in U.S. Hospitals, 2006,” Health Affairs, 27 (3), 865–875.

38.

Garets

Davis

(2006), “Electronic Medical Records v. Electronic Health records: Yes, there is a Difference,” HIMSS Analytics White Paper, (accessed January 26, 2006), [available at http://www.himssanalytics.org/docs/wp_emr_ehr.pdf].

39.

Ghosh

John

(2009), “When Should Original Equipment Manufacturers Use Branded Component Contracts with Suppliers?” Journal of Marketing Research, 46 (October), 597–611.

40.

Goh

Gao

Agarwal

(2011, September), “Evolving Work Routines: A Model of Successful Adaptation to Information Technology in Healthcare,” Information Systems Research, , 22 (3), 565–585.

41.

Greene

W. H.

(2002), Econometric Analysis, 5th ed. Upper Saddle River, NJ, Prentice Hall.

42.

Grimson

Hasselbring

(2000), “The SI Challenge in Health Care,” Communications of the ACM, 43 (6), 46–55.

43.

Grosskopf

Valdmanis

(1993), “Evaluating Hospital Performance with Case-Mix-Adjusted Outputs,” Medical Care, 31 (6), 525–532.

44.

Gupta

Y. P.

Ragunathan

T. S.

(1989), “Impact of Information Systems (IS) Steering Committees on IS Planning,” Decision Sciences, 20 (4), 777–793.

45.

Hamilton

B. H.

Nickerson

J. A.

(2003), “Correcting for Endogeneity in Strategic Management Research,” Strategic Organization, 1 (1), 51–78.

46.

Heckman

(1979), “Sample Selection Bias as a Specification Error,” Econometrica, 47 (1), 153–161.

47.

Hillstead

Bigelow

Bower

Girosi

Meili

Scoville

Taylor

(2005), “Can Electronic Medical Record Systems Transform Health Care? Potential Health benefits, Savings and Costs,” Health Affairs, 24 (5), 1103–1117.

48.

Healthcare Information and Management Systems Society (2009), U.S. EMR Adoption Model Trends. (accessed May 3, 2010), [available at http://www.himssanalytics.org/docs/ha_emram_overview_eng.pdf].

49.

Holstein

W. J.

(2009), “Does Health Care Have an Electronic Future? Strategy+Business,” Booz, (accessed August 18, 2009), [available at http://www.strategy-business.com/article/li00138?gko=1a4d8].

50.

Idson

Feaster

(1990), “A Selectivity Model of Employer-Size Wage Differentials,” Journal of Labor Economics, 8 (1), 99–122.

51.

Institute of Medicine (IOM) (2002), “Key Capabilities of an Electronic Health Record System,” Letter Report, National Academies Press, Washington, DC.

52.

Jensen

M. C.

(1994), “Self-Interest, Altruism, Incentives, and Agency Theory,” Journal of Applied Corporate Finance, 7 (2), 40–45.

53.

Jha

A. K.

DesRoches

C. M.

Campbell

E. G.

Donelan

Rao

S. R.

Ferris

T. G.

Shields

Rosenbaum

Blumenthal

(2009), “Use of Electronic Health Records in U.S. Hospitals,” The New England Journal of Medicine, 360 (16), 1628–1638.

54.

Kazley

A. S.

Ozcan

Y. A.

(2007), “Organizational and Environmental Determinants of Hospital EMR Adoption: A National Study,” Journal of Medical Systems, 31 (5), 375–384.

55.

Kimberly

Evanisko

(1981), “Organizational Innovation: The Influence of Individual, Organizational, and Contextual Factors on Hospital Adoption of Technological and Administrative Innovations,” Academy of Management Journal, 24 (4), 689–713.

56.

Kohli

Grover (2008), “Business Value of IT: An Essay for Expanding Research Directions to Keep Up With the Times,” Journal of the Association for Information Systems, 9 (1), 23–39.

57.

Kuperman

Gibson

(2003), “Computerized Physician Order Entry: Benefits, Costs and Issues,” Annals of Internal Medicine, 139 (1), 31–39.

58.

Prabhala

N. R.

(2007), “Self-Selection Models in Corporate Finance,” in Handbook of Corporate Finance: Empirical Corporate Finance, Vol. 1, Ch. 2, Eckbo

B. E.

, ed. North-Holland: Elsevier, 37–86.

59.

Maddala

G. S.

(1991), “A Perspective on the Use of Limited-Dependent and Qualitative Variables Models in Accounting Research,” The Accounting Review, 66 (4), 788–807.

60.

Maddala

G. S.

(1983), Limited-Dependent and Quantitative Variables in Econometrics. Cambridge, UK: Cambridge University Press.

61.

Masten

S. E.

(1993), “Transaction Costs, Mistakes, and Performance: Assessing the Importance of Governance,” Managerial and Decision Economics, 14 (2), 119–129.

62.

Mata

Fuerst

Barney

(1995), “Information Technology and Sustained Competitive Advantage: A Resource-Based Analysis,” MIS Quarterly, 19 (4), 487–505.

63.

McAfee

(2002), “The Impact of Enterprise Information Technology Adoption on Operational Performance: An Empirical Investigation,” Production Operations Management, 11 (1), 33–53.

64.

McCullough

J. S.

Casey

Moscovice

Prasad

(2010), “The Effect of Health Information Technology on Quality in U.S. Hospitals,” Health Affairs, 29 (4), 647–654.

65.

McCullough

J. S.

(2005), “The Adoption and Diffusion of Hospital Information Systems,” Doctoral Dissertation, University of Pennsylvania, Pennsylvania, PA.

66.

Mekhjian

H. S.

Kumar

R. R.

Kuehn

Bentley

T. D.

Teater

Thomas

Payne

Ahmad

(2002), “Immediate Benefits Realized Following Implementation of Physician Order Entry at an Academic Center,” Journal of the American Medical Informatics Association, 9 (1), 529–539.

67.

Mousin

Remmlinger

Weil

J. P.

(1999), “IT Integration Options for Integrated Delivery Systems,” Healthcare Financial Management, 53 (2), 46–50.

68.

National Institutes of Health (NIH) (2006), Electronic Health Records Overview. www.ncrr.nih.gov/publications/informatics/EHR.pdf.

69.

New York Times (2010), “Doctors and Hospitals Say Goals on Computerized Records Are Unrealistic,” (accessed June 7, 2010), [available at http://www.nytimes.com/2010/06/08/health/policy/08health.html?_r=0].

70.

Nolan

R. L.

(1982), “Managing Information Systems by Committee,” Harvard Business Review, 60 (4), 72–79.

71.

Ondategui-Parra

Bhagwat

Zou

Nathanson

Gill

Ros

(2005), “Use of Productivity and Financial Indicators for Monitoring Performance in Academic Radiology Departments: U.S. Nationwide Survey,” Radiology 236 (1), 214–219.

72.

Overhage

J. M.

Suico

McDonald

C. J.

(2001), “Electronic Laboratory Reporting: Barriers, Solutions and Findings,” Journal of Public Health Management Practice 7 (6), 60–66.

73.

Palacio

Harrison

J. P.

Garets

(2010), “Benchmarking Electronic Medical Records Initiatives in the US: A Conceptual Model,” Journal of Medical Systems, 34 (3), 273–279.

74.

Parente

McCullough

(2009), “Health Information Technology and Patient Safety: Evidence from Panel Data,” Health Affairs, 28 (March/April), 357–360.

75.

Pavlou

P. A.

El Sawy

O. A.

(2006), “From IT Leveraging Competence to Competitive Advantage in Turbulent Environments: The Case of New Product Development,” Information Systems Research, 17 (3), 198–227.

76.

Premkumar

Ramamurthy

Crum

(1997), “Determinants of EDI Adoption in the Transportation Industry,” European Journal of Information Systems, 6 (2), 107–121.

77.

Premkumar

Ramamurthy

Nilakanta

(1994), “Implementation of Electronic Data Interchange: An Innovation Diffusion Perspective,” Journal of Management Information Systems, 11 (2), 157–186.

78.

Preston

D. S.

Leidner

D. E.

Chen

(2008), “CIO Leadership Profiles: Implications of Matching CIO Authority and Leadership Capability on IT Impact,” MIS Quarterly Executive, 7 (2), 57–69.

79.

Preston

D. S.

Karahanna

Rowe

(2006), “Development of a Shared Understanding Between the Chief Information Officer and Top Management Team in U.S. and French Organizations: A Cross-Cultural Comparison,” IEEE Transactions on Engineering Management, 53 (2), 191–206.

80.

Queenan

C. C.

Angst

C. M.

Devaraj

(2011), “Doctors’ Orders—If They’re Electronic, Do They Improve Patient Satisfaction? A Complements/Substitutes Perspective,” Journal of Operations Management, 29 (7-8), 639–649.

81.

Raghupathi

(1997), “Health Care Information Systems,” CACM, 40 (8), 81–82.

82.

Ray

Barney

J. B.

Muhanna

W. A.

(2004), “Capabilities, Business Processes, and Competitive Advantage: Choosing the Dependent Variable in Empirical Tests of the Resource-Based View,” Strategic Management Journal, 25 (1), 23–37.

83.

Reich

B. H.

Benbasat

(2000), “Factors that Influence the Social Dimension of Alignment Between Business and Information Technology Objectives,” MIS Quarterly, 24 (1), 81–113.

84.

Samaan

Z. M.

Klein

M. D.

Mansour

M. E.

DeWitt

T. G.

(2009), “The Impact of the Electronic Health Record on an Academic Pediatric Primary Care Center,” Journal of Ambulatory Care Management, 32 (3), 180–187.

85.

Schiff

G. D.

Klass

Peterson

Shah

Bates

D. W.

(2003), “Linking Laboratory and Pharmacy: Opportunities for Reducing Errors and Improving Care,” Archives of Internal Medicine, 163 (8), 893–900.

86.

Schneller

E. S.

(2000), The Value of Group Purchasing in the Health Care Supply Chain. Tempe: School of Health Administration and Policy, Arizona State University.

87.

Shaver

J. M.

(1998), “Accounting for Endogeneity when Assessing Strategy Performance: Does Entry Mode Choice Affect FDI Survival,” Management Science, 44 (4), 571–585.

88.

Shehata

(1991), “Self-selection Bias and the Economic Consequences of Accounting Regulation: An Application of Two-Stage Switching Regression to SFAS No 2,” The Accounting Review, 66 (4), 768–787.

89.

Shukla

R. K.

Pestian

Clement

(1997), “A Comparative Analysis of Revenue and Cost-Management Strategies of Not-for-Profit and For-Profit Hospitals,” Hospital Health Service Administration, 42 (1), 117–134.

90.

Smaltz

D. H.

Sambamurthy

Agarwal

(2006), “The Antecedents of CIO Role Effectiveness in Organizations: An Empirical Study in the Health Care Sector,” IEEE Transactions on Engineering Management, 53 (2), 207–222.

91.

Sobol

M. G.

Klein

(2009), “Relation of CIO Background, IT Infrastructure, and Economic Performance,” Information and Management, 46 (5), 271–278.

92.

Tabak

Jain

(2000), “Predicting Hospital Innovativeness: An Exploration of the Perceived Organizational Context,” Journal of Applied Management Studies, 9 (2), 217–226.

93.

Tang

P. C.

Ash

J. S.

Bates

D. W.

Overhage

J. M.

, D. Z. & Sands (2006), “Personal Health Records: Definitions, Benefits, and Strategies for Overcoming Barriers to Adoption,” Journal of The American Medical Informatics Association, 13 (2), 121–126.

94.

Teo

H. H.

Wei

K. K.

Benbasat

(2003), “Predicting Intention to Adopt Interorganizational Linkages: An Institutional Perspective,” MIS Quarterly, 27 (1), 19–49.

95.

Thirumalai

Sinha

K. K.

(2012), “To Personalize or Not to Personalize Online Purchase Interactions: Implications of Self-Selection by Retailers,” Information Systems Research (forthcoming).

96.

Tierney

W. M.

McDonald

C. J.

Martin

D. K.

Rogers

M. P.

(1987), “Computerized Display of Past Test Results. Effect on Outpatient Testing,” Annals of Internal Medicine, 107 (4), 569–574.

97.

Zahra

S. A.

George

(2002), “Absorptive Capacity: A Review, Reconceptualization, and Extension,” The Academy of Management Review, 27 (2), 185–203.

98.

Zhu

Kraemer

K. L.

(2006), “The Process of Innovation Assimilation by Firms in Different Countries: A Technology Diffusion Perspective on E-Business,” Management Science, 52 (10), 1557–1576.

99.

Zhu

Kraemer

K. L.

(2005), “Post-Adoption Variations in Usage and Value of E-Business by Organizations: Cross-Country Evidence from the Retail Industry,” Information Systems Research, 16 (1), 61–84.

100.

Zhu

Kraemer

K. L.

(2003), “E-business Adoption by European Firms. A Cross Country Assessment of the Facilitators and Inhibitors,” European Journal of Information Systems, 12 (4), 251–268.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.06 MB