How do Information Technology Choices Made by an Institution Alter Consumer Participation in the Health Care System?
Teresa ZAYAS-CABÁN, MS, PhD(c)a, Jenna L. MARQUARD, MS, PhD(s)a, Patricia Flatley BRENNAN, RN, PhD, FAAN, FACMIa,b
aDepartment of Industrial and Systems Engineering, bSchool of Nursing University of Wisconsin-Madison, Madison, WI, USA
Abstract. As information technology (IT) is applied within increasingly complex organizational environments, its design must consider and account for this richness. The health care industry illustrates these complexities through a wide range of organizations that operate interdependently, but differ in expectations regarding participation and in their ability to posit or respond to information queries. This article summarizes results of a discrete event simulation that models health information access within a community and how institutions changing modes of health information delivery to patients can affect the patients’ ability to obtain information.
Keywords. Information technology, organizations, health care, decision-making
1. Introduction
As information technology (IT) is applied within increasingly complex organizational environments, its design must consider and account for this richness. This challenge is particularly true for the US health care industry, where care is provided by many different institutions (hospitals, clinics, physician offices) with which patients may need to interact to receive the full range of care. Patients play a particularly important role in contemporary health care, not only in reporting their symptoms accurately but also in following recovery advice and health promotion instructions provided from health care professionals. Thus, designing information technology solutions for contemporary health care requires explicit consideration of (1) the complex, multi-organizational context of health care; (2) the shifting of health care work from professional-delivered to a patient-clinician partnership and (3) the need to manage information not only within a single organization but across organizations and directly into the patients’ home.
2. Background
Health care is composed of an amalgam of organizations that have different levels of resources. Because of differences in size, types of services provided, resources, infrastructure, geographic location, expenditures, and payment schemes, health care organizations have varying incentives and motivations to adopt new information technology (Ash & Bates, 2005). These differences can also affect each organization’s ability to interact using different modes and volumes of communication.
Most IT design in health care emphasizes improving the information technology available in hospitals and clinics to help insure better quality, efficient, and safe care (Overhage, Evans, & Marchibroda, 2004). There are also emerging movements towards the use of regional health information organizations, electronic health records (EHRs), and personal health records (Detmer, 2003; IOM, 2002; Yasnoff et al., 2004). All of these applications will require a high level of integration between health care organizations and, between the health care organization and the home. This study capitalized on an existing industry-funded project, Health@Home, to examine transmission of health information from institutions to the home. The Health@Home project had two key aims: (1) documenting the health information resources of a community and (2) assessing the information technology readiness of 49 households within that community.
Traditional information technology design approaches draw from organizational behavior models, and are of little use in advising the institution-home care information exchange problem. Many focus on an internal organization that is part of a “fixed” system and include a presumption of rules and boundaries. Macroergonomics and sociotechnical systems theories are more broad, addition attention to interdependencies and networks into systems design. Karsh and Holden (2004) describe IT implementation in health care that has employed a sociotechnical systems framework. Trist suggests that the study of community networks, which he describes as “fluid and unbound” (Trist, 1981, page 56) will yield further expansion to understanding diffusion of innovations. Marjchrzak (1992) advises focusing on relationships between suppliers (health care organizations) and consumers (patients) in the product and/or work process design. Because IT should support and fit within current care practices (Berg, 1999), and because in contemporary health care these practices are community- rather than institution-centered, IT design strategies must extend beyond a given institution’s boundaries into consumers’ homes.
Simulation is a powerful tool that has been proven to help decision-makers realize the potential effects of alternative system configurations. It affords developers the opportunity to explore a system and change parameters without incurring the risk and costs associated with implementation. This article presents a discrete event simulation, built using data from the Health@Home projects, as well as data from related literature, to model health information access within a community and how institutions changing modes of health information delivery to patients can affect the patients’ ability to obtain this information.
3. Methods
3.1 Simulation
A discrete event simulation model was created using ArenaÒ simulation software. This simulation explores what information is provided by direct health care providers in a rural Midwestern community and patients’ capacity to receive information. The simulation is illustrated in Figure 1 and models the “health information flow” process. The model is created through a set of sub models that, in this case, depicts how the information moves from providers to patients.
Figure 1—Simulation Model
In the model, packets of health information are generated from each provider in the community on a per month basis. This creation takes place in the “Provider Information Creation” sub model. These packets of health information are then each assigned to verbal, written or electronic “Transmission Type” status. Depending on the “Transmission Type” assigned to the information, the packets are sent through different pathways in the “Information Filtering” sub model. During this sub model, information packets move through potential failure modes, continuing towards the patient or failing to reach them due to one of these failure modes. The “Patient” and “Failure to Reach Patient” sub models distinguish what information reaches the patient and what information does not, as well at why failed information did not reach the patient.
3.2 Assumptions
For the purposes of the simulation, several assumptions were made. The model does not include the time or cost of information creation, transfer or reception and there are only three types of transmissions used in the model: (1) verbal, either in person or by telephone; (2) written, either through publications or mail; and (3) electronic, either via Internet or e-mail. Also, the provider always gives the patient information, but each patient is provided only one type of educational material (i.e. no redundancy in transmission types given to the patient). Finally, health information was limited to patient education materials.
3.3 Data
Data for the model came from the results of the Health@Home projects. These data determined how many providers would be included in the model. The model, after data reduction, included 29 health information-providing organizations. The number of patients served per month by each organization was used to generate the number of information packets created per month (1 per patient). Health@Home results also reported format rankings for each of the transmission types (verbal, written, electronic), which were used to assign percentages of differing transmission types per each organization.
The Health@Home data were also used to determine patients’ electronic reception capability. This was calculated by Internet access in the home only (51% have internet in the home) and by access across home, school, work and library (63% access rate). Additional data needed to describe potential transmission failures came from the relevant literature. Table 1 lists all the failure rates used in the simulation and their data sources (V= verbal, W=written, and E=electronic).
Table 1—Failure Rates and Data Sources
Type of Transmission Failure / Used For Transmission Types / Failure Percentage Used / Failure Percentages Found / Source(s) of this estimate /Forgetting and Information processing Capability / V / 55% / 50-68% / (Scheitel, Boland, Wollan, & Silverstein, 1996)
Language Barrier / V, WHC, E / 8.1% / 8.1% / (U.S. Census Bureau, 2000)
Literacy and Information Processing CapabilitySedation and Mental State / HCW, E / 35%(65%)5% / · 26-60%
· 21-48%
39%None / · (Williams et al., 1995)
· (Davis, Michielutte, Askov, Williams, & Weiss, 1988)
· (Wisconsin Literacy, 2003)Estimate
Forgetting and Information processing Capability / 55% / 50-68% / (Scheitel, Boland, Wollan, & Silverstein, 1996)
Lost or Never Read / HCW, E / 20% / None / Estimate
Network Interception or DropLiteracy and Information Processing Capability / E / 1%(99%)35%(65%) / 1%26-60%
· 21-48%
· 39% / (Kalyanakrishnan, Iyer, & Patel, 1997)(Williams et al., 1995)
(Davis, Michielutte, Askov, Williams, & Weiss, 1988)
(Wisconsin Literacy, 2003)
Incompatible Reception Capacity / E / 40%(60%) / · 37-49%
· 38% / · Health@Home Data
· (Pew Internet & American Life Project, 2005)
Sedation and Mental StateNetwork Interception or Drop / V, HCW, E / 5%1%(99%) / None1% / Estimate(Kalyanakrishnan, Iyer, & Patel, 1997)
4. Results
The simulation model was run under five sets of initial conditions, as follows: (1) community providers supplying information as they are now; (2) the largest five patient servers (>1000 patients per month) changing all written transmissions to electronic (3) all providers using only verbal transmission; (4) all providers using only hard-copy transmission; and (5) all providers using only electronic transmission.
Table 2 – Original Initial Conditions and with Largest Information Providers Switching from Written to Electronic Transmissions
Conditions
/ Original Conditions /Largest Providers Electronic Instead of Written
Outcome
/ Count / Percent /Count
/Percent
Successfully Reaches Patient / 84015 / 41.27% / 74237 / 36.47%Forgetting and Information Processing Failure / 65339 / 32.10% / 65153 / 32.01%
Language Barrier Failure / 7702 / 3.78% / 6820 / 3.35%
Literacy and Information Processing Failure / 23079 / 11.34% / 17150 / 8.43%
Lost or Never Read Failure / 16740 / 8.22% / 12323 / 6.05%
Network Interception or Drop Failure / 73 / 0.04% / 590 / 0.29%
Incompatible Reception Capacity Failure / 2289 / 1.12% / 23358 / 11.47%
Sedation and Mental State Failure / 4319 / 2.12% / 3925 / 1.93%
Total
/ 203556 / 100% / 203556 / 100%Table 3 – Transmissions Only in Verbal, Written, or Electronic Form
Conditions
/Only Verbal
/ Only Written / Only ElectronicOutcome
/Count
/Percent
/ Count / Percent / Count / PercentSuccessfully Reaches Patient / 79696 / 39.16% / 92444 / 45.43% / 55155 / 27.10%
Forgetting and Information Processing Failure / 112484 / 55.28%
Language Barrier Failure / 7271 / 3.57% / 8393 / 4.12% / 5115 / 2.51%
Literacy and Information Processing Failure / 56975 / 28.00% / 33782 / 16.60%
Lost or Never Read Failure / 40997 / 20.15% / 24044 / 11.81%
Network Interception or Drop Failure / 2068 / 1.02%
Incompatible Reception Capacity Failure / 80469 / 39.53%
Sedation and Mental State Failure / 4105 / 2.02% / 4747 / 2.33% / 2923 / 1.44%
Total
/ 203556 / 100% / 203556 / 100% / 203556 / 100%As seen in Tables 2 and 3, information successfully reaching patients decreases as organizations transfer to electronic transmission. Written transmission is the most successful form of communication in terms of percent of sent packets reaching patients.
5. Discussion and Conclusions
This simple simulation performed well in understanding the types of potential failures and rates of failure for differing health information transmission types. The results show the impact of a movement toward electronic communication on patient reception of the information, even when this push comes from only a few large providers. The simulation also worked to systematically pull together disparate sources of information. Where studies often look at patient education from a single transmission type and limited failure types, this work provided a more cohesive look at the patient education process.
With appropriate data, a simulation model may be useful in modeling health information flow to patients. Such a model can help an organization forecast the consequences of their current health information distribution practices.
6. Future Work
The simulation model could be enhanced in several ways. Data transmission types estimates were based on percentages; future work could look toward using distributions. This model should also explore redundancy in transmission types because it increases the probability that patients receive information. Finally, there is a need to explore the costs and times associated with the creation, transfer and reception of these transmission types in order to understand issues such as maximizing the probability of reaching a patient given time or cost constraints.
7. Acknowledgements
Teresa Zayas-Cabán and Jenna L. Marquard are funded by the National Science Foundation Graduate Research Fellowship Program. Support for the Advanced Technologies for Health@Home project was provided by the Intel Corporation Research Group, P. Brennan, PI.
8. References
Ash, J. S., & Bates, D. W. (2005). Factors and Forces Affecting EHR System Adoption: Report of a 2004 ACMI Discussion. Journal of the American Medical Informatics Association, 12(1), 8-12.
Berg, M. (1999). Patient care information systems and health care work: A sociotechnical approach. International Journal of Medical Informatics, 55(2), 87-101.
Davis, T. C., Michielutte, R., Askov, E. N., Williams, M. V., & Weiss, B. D. (1988). Practical assessment of adult literacy in health care. Health Education & Behavior, 25(5), 613-624.
Detmer, D. (2003). Building the national health information infrastructure or personal health, health care services, public health and research. BMC Medical Informatics and Decision Making, 3(1).
IOM. (2002). Fostering rapid advances in healthcare: learning from system demonstrations. Washington, DC: National Academy Press.
Kalyanakrishnan, M., Iyer, R. K., & Patel, J. (1997, 22-25 Sept.). Reliability of Internet hosts: a case study from the end user's perspective. Paper presented at the Sixth International Conference on Computer Communications and Networks, Las Vegas, NV.
Karsh, B.-T., & Holden, R. J. (2004). New Technology Implementation in Health Care. In P. Carayon (Ed.), Handbook of Human Factors and Ergonomics in Patient Safety. Accepted.
Majchrzak, A. (1992). Management of technological and organizational change. In G. Salvendy (Ed.), Handbook of Industrial Engineering (pp. 767-797). New York: John Wiley and Sons.