Data mining: sophisticated forms of managed care modeling through artificial intelligence

Data mining is a recent development in computer science that combines artificial intelligence algorithms and relational databases to discover patterns automatically, without the use of traditional statistical methods. Work with data mining tools in health care is in a developmental stage that holds great promise, given the combination of demographic and diagnostic information.

Predicting readmission to the psychiatric hospital in a managed care environment: implications for quality indicators

OBJECTIVE

This study examined predictors of hospital readmission to determine whether readmissions can serve as a quality indicator for an inpatient psychiatric service.

METHOD

A series of 255 patients consecutively admitted to any of seven psychiatric hospitals in a regional managed care program were followed to determine whether they were readmitted within 6 months of discharge. Case managers assessed patients with the use of a reliable outcome management/decision support system designed for acute psychiatric services.

RESULTS

Patients with greater impairment in self-care, more severe symptoms, and more persistent illnesses were more likely to be readmitted than other patients. Suicidal patients were less likely to be readmitted. There was no evidence to suggest that poor hospital outcome or premature discharge was associated with readmission either within 30 days or within 6 months.

CONCLUSIONS

Although patients at risk for hospital admission can be identified, it does not appear that the success of the hospital intervention per se influences the likelihood of readmission. Use of readmission rates as quality indicators for hospital care providers is not recommended.

Decision support systems in health care

This paper delineates the application of decision support systems in the healthcare arena. An overview of decision support systems in the healthcare environment in the last three to five years, mostly in the USA and the UK is provided. Some reference is made to work in South Africa. Although this report endeavours to emphasise the decision support system side, some operations research issues are also addressed. This paper is divided into several sections: decision support systems definition, techniques used, the use of operations research in specific areas of healthcare, information systems, assessment of decision support systems, in-house development, decision support in South Africa and future of decision support systems.

The rigors and rewards of clinical evaluation in an alliance of academic health systems

The University HealthSystem Consortium (UHC), a 70-member alliance of academic health centers, was established in 1984 to help its members improve clinical practice and to strengthen their competitive position in the health care market. This article describes the programs of the UHC's Clinical Practice Advancement Center and its progress to date in promoting strategies to control cost, demonstrate value, and improve health care quality.

A decision-making aid for long-term care waiting list policies: modelling first-come, first-served vs. needs-based criteria

A chronic care hospital in London, Ontario, was faced with the question of whether to continue to admit applicants on a first-come, first-served basis, or to implement a needs-based admission criterion To provide information relevant to this decision, a simple computer modelling technique was used to model the waiting list under both policies. Analysis of the results indicated substantial variability in the need for care demonstrated by waiting list applicants, and individual placement priority under the two schemes. Descriptions of individuals affected by the proposed change in policy provided additional information. The information provided by the waiting list analysis contributed to the decision-making process about waiting list policy. The method used is applicable to waiting list management decisions in other institutions.

Risk management. Betting on good information

Eating potato salad left out in the sun, flying a kite in a rainstorm, sticking a fork in a light socket--we don't need an actuarial table to know that these and many other actions are downright life-threatening. Most of us grow up learning these basic lessons and think of them as common sense, not as a measure of risk. But what about having a heart attack? Contracting a fatal disease? Experiencing unusual side effects from medicine? These risks aren't so obvious. But in each situation, information is available that can identify the chances of any one of us experiencing a stroke or food poisoning. In fact, conclusions about the incidence and potential health risks are key in the daily business decisions of healthcare. Just knowing the gambles, however, is not what risk in healthcare is all about. Instead, it is about the potential costs of those risks to health planners footing the bill.

The Healthcare Administrator's Associate: an experiment in distributed healthcare information systems

The Healthcare Administrator's Associate is a collection of portable tools designed to support analysis of data retrieved via the Internet from diverse distributed healthcare information systems by means of the InfoSleuth system of distributed software agents. Development of these tools is part of an effort to enhance access to diverse and geographically distributed healthcare data in order to improve the basis upon which administrative and clinical decisions are made.

Uncertainty considerations in group decisions: staffing and ratesetting in a health promotion center

There are a variety of methods for including risk or uncertainty in Group Decision Support Systems (GDSS). Monte Carlo simulation is one method. An application of Monte Carlo simulation in Group Decisions and GDSS for diversification of services is presented in this paper. Specifically, the uncertainty associated with the finances of a Health Promotion Center are evaluated for staffing and rate setting. The distributions of outcome variables are used in decision rules.

Productive use of IT in support of FM (facilities management) solutions

Drawing on the author's personal experience within the UK National Health Service, outlines at the macro level what information is needed, and why, in facilities management. Suggests that top-level data gathering is a priority, describes the processes required, and finally considers the advantages of sharing information with competitors and others.

The importance of being earnest: reflections on an attempt to create an information-led organizational culture in the NHS

Presents a study of organizational culture in a community health-care Trust in the UK. The Trust has been involved in attempting to create what was described as an information-led culture and is concurrent with the implementation of a new information system. Describes some of the recent management initiatives which were intended to improve communications within the Trust and to provide employees with a better understanding of the changes which have been taking place. Identifies findings which suggest that there was some distance between the perceptions of management and those of employees. While the senior management team appear to have been in earnest in attempting to ensure that staff have the information and understanding they need, the results of an organizational culture survey suggest that these efforts may have been ineffective.

A conception of a support system for optimising the organisation of nursing staff from the viewpoint of the nursing care needs structure

Methods of allocating nursing staff in hospitals should be established that improve the "sufficiency rate" of existing care needs for nursing in patients. The conception of a support system for the optimisation of organising nursing staff was proposed so that the degree of meeting care needs could be maximised. In this study, the structure of care needs was defined by sampling information related to eight factors. A basic framework was developed for a support system for the optimisation of organising nursing staff by using data based on (1) rate of satisfying eight selected criteria, (2) condition factors that can affect this rate, and (3) organising nursing. The possible usage of this support system is also discussed in this paper.

Managing the gap: balancing advances in technology with advances in management practice

Expenditure on information systems is widely anticipated to lead to improved management of health care resources. Despite large investments in hardware and software, these expectations are difficult to realise. Part of the difficulty lies in the manner in which information systems are applied to, rather than integrated within, organisations. This paper considers some of the the personal and organisational issues that need to be addressed to 'manage the gap' in balancing advances in information technology with advances in management practice. The issues identified are consistent with the concept of a learning organisation dealing with environmental change.

Corporate information systems in health organisations

This paper presents an overview of the nature of corporate information systems and their applications in health organisations. It emphasises the importance of financial and human resource information in the creation of a corporate data model. The paper summarises the main features of finance and human resource systems as they are used in health organisations. It looks at a series of case studies carried out in health organisations, which were selected on the basis of their representation of different aspects of service delivery. It also discusses the theoretical and practical perspectives of the systems themselves, their roles in information management, executive and decision support, and in planning and forecasting.