PubMed Health⌕ Search

PubMed · 14980159

[Computer-assisted clinical decision support systems].

Abstract

This review deals with the historical evolution and methodological basis of computerized medical decision support tools. It also reviews the verification, validation and impact evaluation methods of these systems. The acceptability of them from the health professionals depends on the pro-activity and usefulness of the systems, between other factors. The clinical impact evaluation methods are described in deep detail and the Hunt et al review of 1998 is updated.

Explore related subjects

Keep this discovery

Explore connections, maps & timelines

BibTeXRIS

Julio Bonis, Juan J Sancho, Ferran Sanz. 2004. [Computer-assisted clinical decision support systems].. https://doi.org/10.1157/13057545

Cite the original work for its findings. Save a collection to share your selection of sources.

KEEP EXPLORING

Related citations

First evaluation of the NHS direct online clinical enquiry service: a nurse-led web chat triage service for the public.

BACKGROUND: NHS Direct is a telephone triage service used by the UK public to contact a nurse for any kind of health problem. NHS Direct Online (NHSDO) extends NHS Direct, allowing the telephone to be replaced by the Internet, and introducing new opportunities for informing patients about their health. One NHSDO service under development is the Clinical Enquiry Service (CES), which uses Web chat as the communication medium. OBJECTIVE: To identify the opportunities and possible risks of such a service by exploring its safety, feasibility, and patient perceptions about using Web chat to contact a nurse. METHODS: During a six-day pilot performed in an inner-city general practice in Coventry, non-urgent patients attending their GP were asked to test the service. After filling out three Web forms, patients used a simple Web chat application to communicate with trained NHS Direct triage nurses, who responded with appropriate triage advice. All patients were seen by their GP immediately after using the Web chat service. Safety was explored by comparing the nurse triage end point with the GP's recommended end point. In order to check the feasibility of the service, we measured the duration of the chat session. Patient perceptions were measured before and after using the service through a modified Telemedicine Perception Questionnaire (TMPQ) instrument. All patients were observed by a researcher who captured any comments and, if necessary, to assisted with the process. RESULTS: A total of 25 patients (mean age 48 years; 57% female) agreed to participate in the study. An exact match between the nurse and the GP end point was found in 45% (10/22) of cases. In two cases, the CES nurse proposed a less urgent end point than the GP. The median duration of Web chat sessions was 30 minutes, twice the median for NHS Direct telephone calls for 360 patients with similar presenting problems. There was a significant improvement in patients' perception of CES after using the service (mean pre-test TMPQ score 44/60, post-test 49/60; p=0.008 (2-tailed)). Patients volunteered several potential advantages of CES, such as the ability to re-read the answers from the nurse. Patients consider CES a useful addition to regular care, but not a replacement for it. CONCLUSIONS: Based on this pilot, we can conclude that CES was sufficiently safe to continue piloting, but in order to make further judgments about safety, more tests with urgent cases should be performed. The Web chat sessions as conducted were too long and therefore too expensive to be sustainable in the NHS. However, the positive reaction from patients and the potential of CES for specific patient groups (the deaf, shy, or socially isolated) encourage us to continue with piloting such innovative communication methods with the public.

Decision Making, Computer-Assisted↗

Improving patient safety through informatics tools for shared decision making and risk communication.

This paper describes how informatics tools can support shared decision making and risk communication and thereby play an important role in enhancing patient safety. Using preference elicitation techniques and knowledge on risk communication, such tools can help patients understand their treatment options and associated short- and long-term benefits and risks, assist in the elicitation of patient preferences, and help patients and clinicians in making treatment choices with the highest likelihood of achieving desired patient outcomes. Important features of such tools are proposed, including: (1) Interactive tutorials to improve risk comprehension and prepare patients and clinicians for the decision making task; (2) choices between different presentation modes to meet patients' individual reading levels and presentation preferences; (3) risk calculations that account for individual risk profiles; (4) performance of necessary calculations to reach the actual decision; (5) automatic updates of evidence; and (6) the use of different preference-elicitation techniques.

Decision Making, Computer-Assisted↗

Computerised triage in a prostate assessment clinic.

An MS Office package has been developed to score IPSS, take a patient history, compare flows with nomograms and decide on interim management. This is based on these scores, residual volume and laboratory results. A clearly formatted GP letter is produced. The patient uses a touch screen to answer questions on the IPSS and other medical history. These questions and responses are stored in Excel spread sheets. Clinic staff then enter results of flow studies, urinalysis, U&E and PSA. Excel Visual Basic creates a detailed printout for the notes and the MS Office mail merge facility creates a summary printout, which also serves as a letter to the GP. Excel allows embedding of formulae and program code to implement the rules for management. Based on these rules, the program either generates a request for an urgent appointment in the clinic or recommends the use of either an alpha blocker (if not contraindicated by medical history) or 5 alpha reductase inhibitors in the interim period before they are reviewed in clinic. A total of 139 patients have been seen and the computer decisions compared with those of a consultant urologist. Agreement was found in 106, disagreement in 33. However, 21 of the 33 involved computer oversensitivity to flow results. We do not anticipate difficulty improving this and are investigating using an artificial neural network. Of the other 12 patients, the urologist departed from the fixed rules for IPSS, creatinine, PSA and residual urine when only one variable was slightly abnormal. To conclude, this novel user-friendly system shows great potential in the management of prostate outpatients. Some tuning is needed, with particular respect to uroflow results.

Decision Making, Computer-Assisted↗