Factors governing the adoption of artificial intelligence in healthcare providers.

Artificial intelligence applications are prevalent in the research lab and in startups, but relatively few have found their way into healthcare provider organizations. Adoption of AI innovations in consumer and business domains is typically much faster. While such delays are frustrating to those who believe in the potential of AI to transform healthcare, they are largely inherent in the structure and function of provider organizations. This article reviews the factors that govern adoption and explains why adoption has taken place at a slow pace. Research sources for the article include interviews with provider executives, healthcare IT professors and consultants, and AI vendor executives. The article considers differential speed of adoption in clinical vs. administrative applications, regulatory approval issues, reimbursement and return on investments in healthcare AI, data sources and integration with electronic health record systems, the need for clinical education, issues involving fit with clinical workflows, and ethical considerations. It concludes with a discussion of how provider organizations can successfully plan for organizational deployment.

Rapid identification of myocardial infarction risk associated with diabetes medications using electronic medical records.

OBJECTIVE To assess the ability to identify potential association(s) of diabetes medications with myocardial infarction using usual care clinical data obtained from the electronic medical record. RESEARCH DESIGN AND METHODS We defined a retrospective cohort of patients (n = 34,253) treated with a sulfonylurea, metformin, rosiglitazone, or pioglitazone in a single academic health care network. All patients were aged >18 years with at least one prescription for one of the medications between 1 January 2000 and 31 December 2006. The study outcome was acute myocardial infarction requiring hospitalization. We used a cumulative temporal approach to ascertain the calendar date for earliest identifiable risk associated with rosiglitazone compared with that for other therapies. RESULTS Sulfonylurea, metformin, rosiglitazone, or pioglitazone therapy was prescribed for 11,200, 12,490, 1,879, and 806 patients, respectively. A total of 1,343 myocardial infarctions were identified. After adjustment for potential myocardial infarction risk factors, the relative risk for myocardial infarction with rosiglitazone was 1.3 (95% CI 1.1-1.6) compared with sulfonylurea, 2.2 (1.6-3.1) compared with metformin, and 2.2 (1.5-3.4) compared with pioglitazone. Prospective surveillance using these data would have identified increased risk for myocardial infarction with rosiglitazone compared with metformin within 18 months of its introduction with a risk ratio of 2.1 (95% CI 1.2-3.8). CONCLUSIONS Our results are consistent with a relative adverse cardiovascular risk profile for rosiglitazone. Our use of usual care electronic data sources from a large hospital network represents an innovative approach to rapid safety signal detection that may enable more effective postmarketing drug surveillance.

Seven durable ideas.

Partners Healthcare, and its affiliated hospitals, have a long track record of accomplishments in clinical information systems implementations and research. Seven ideas have shaped the information systems strategies and tactics at Partners; centrality of processes, organizational partnerships, progressive incrementalism, agility, architecture, embedded research, and engage the field. This article reviews the ideas and discusses the rationale and steps taken to put the ideas into practice.

Too far ahead of the IT curve?

Peachtree Healthcare has major IT infrastructure problems, and CEO Max Berndt is struggling to find the right fix. He can go with a single set of systems and applications that will provide consistency across Peachtree's facilities but may not give doctors enough flexibility. Or he can choose service-oriented architecture (SOA), a modular design that will allow Peachtree to standardize incrementally and selectively but poses certain risks as a newer technology. What should he do? Four experts comment on this fictional case study, authored by John P. Glaser, CIO for Partners HealthCare System. George C. Halvorson, the chairman and CEO of Kaiser Permanente, warns against using untested methodologies such as SOA in a health care environment, where lives are at stake. He says Peachtree's management must clarify its overall IT vision before devising a plan to achieve each of its objectives. Monte Ford, the chief information officer at American Airlines, says Peachtree can gradually replace its old systems with SOA. An incremental approach, he points out, would not only minimize risk but also enhance flexibility and control, and would allow IT to shift priorities along the way. Randy Heffner, a vice president at Forrester Research who focuses on technology architectures for computer-based business systems, thinks SOA's modular approach to business design would best meet Peachtree's need for flexibility. He says that Peachtree's CIO sees SOA as a new product category but should instead view it as a methodology. John A. Kastor, a professor at the University of Maryland School of Medicine, questions the goal of standardized care. He argues that it would be difficult to persuade doctors, many of whom are fiercely independent, to follow rigid patterns in their work.

Architecture of the open-source clinical research chart from Informatics for Integrating Biology and the Bedside.

The Informatics for Integrating Biology and the Bedside (i2b2) is one of the sponsored initiatives of the NIH Roadmap National Centers for Biomedical Computing (http://www.bisti.nih.gov/ncbc/). One of the goals of i2b2 is to provide clinical investigators broadly with the software tools necessary to collect and manage project-related clinical research data in the genomics age as a cohesive entity, a software suite to construct and manage the modern clinical research chart. The i2b2 "hive" is a set of software modules called "cells" that have a common messaging protocol that allow them to interact using web services and XML messages. Each cell can be developed by independent investigators to achieve specific analytic goals, and then be integrated into the hive to enhance the functionality available in the i2b2 Hive. We have applied this architecture through several ongoing clinical studies and found it to be of high value. The current version of this software has been released into the public domain and is available at the URL-http://www.i2b2.org.

Information technology strategy: three misconceptions.

The core intent in developing an IT strategy is to ensure that there is a strong and clear relationship between IT investment decisions and the organization's overall strategies, goals, and objectives. In the course of developing an IT strategy, an organization may fall victim to three major misconceptions about IT strategy. Those misconceptions are: The IT strategy should be solely derived from a thorough review of organizational strategies and plans. The IT strategy should be dominated by a focus on defining needed application systems. The IT strategy is better if it is developed by using a rigorous methodology. These misconceptions are dangerous. While they are right, they are not completely right. Hence, a dogmatic approach embracing these misconceptions risks an incomplete IT strategy or a strategy that is not as aligned with the organization as it should be.

Concepts for building inter-organizational systems in healthcare: lessons from other industries.

A centerpiece of the nation's healthcare information technology vision is the implementation of health information exchanges (HIEs). HIEs have the significant potential to improve the quality of patient care, reduce care costs, and enhance patient safety. These kinds of exchanges are new to healthcare in the United States. While more than 200 communities are in various stages of implementing them, there is little mature experience with the factors that contribute to the successful formation and sustainability of these exchanges. The creation and management of mechanisms to support the exchange of data between organizations has been used in other industries. There are many examples of interorganizational systems that have been established in industries as diverse as banking, manufacturing, government, and retail. This paper examines some of these experiences and attempts to glean the concepts and guidance healthcare can learn from other industries.

Facilitating applied information technology research.

Most healthcare organizations need to develop an understanding of the role, maturity, and utility of new information technologies. Performing this research can present a series of challenges. For example, IT staff may not know how to obtain resources; the research may show promising results but fail to be implemented; and organizations may not develop a thoughtful portfolio of research projects. To minimize these problems, applied information technology research must be managed. This paper describes the information technology research process at Partners HealthCare. The research process is a management process designed to identify important research topics, solicit IT staff proposals to investigate topics, prioritize and fund research projects, and communicate and disseminate research results. Sample research topics include the role of RFID, the impact of Internet-based e-visits, and approaches for integrating speech and direct input for clinical information systems. The paper describes the goals and activities of the research process and the role of the research council. The research proposal review process is discussed, and an inventory of approved research projects is reviewed. The paper discusses several challenges involved in creating and managing the research process.

The New England Healthcare EDI Network.

The New England Healthcare EDI Network (NEHEN) is a collaborative of providers and payers in eastern Massachusetts that created, manages, and operates a shared insurance EDI infrastructure. NEHEN currently has 12 provider and three payer members, and supports over 1,000,000 insurance EDI transactions per month. This paper describes the philosophies that define the NEHEN business model and discusses its governance structure, technology, operational issues associated with its implementation, and its current status, along with lessons learned from the NEHEN undertaking.

Experiences with knowledge application in medical care.

IT interventions, such as provider order entry and knowledge repositories, are improving the application of medical knowledge. One health system's experiences are reviewed, including challenges faced and organizational factors that supported the process.