Clinical decision support systems (CDSS)

The computerised prescribing of medicines is considered an important measure in reducing medication errors. A system  enabling doctors to prescribe drugs electronically is often referred to as computerised physician order entry (CPOE). These systems often include integrated decision support components that attempt to improve prescribers’ decisions through advice, alerts and reminders. However, if the burden of alerts is too high, alert fatigue may cause prescribers to override most alerts, in a manner that compromises the desired safety effect of integrating decision support into the CPOE systems.

Decision support can be defined as any information added by a system to assist the clinician’s decision making process. The format of this support can vary widely[1] and include:

• Reference materials (eg. textbooks, journals,medication guides).
• “Expert systems” that assist in diagnosis (eg, clinical decision rules).
•  Rule-based systems that provide clinical alerts (eg, drug-allergy and drug-drug interaction checking).
• Scoring systems (eg, mortality or infection scores).
• Data tags that add context (eg, adding an ‘L’ for low next to a laboratory value).
• Systems that assist in therapeutic decisions (eg, automatic suggestion for insulin and glucose in the case of a hyperkalaemic patient).
• Computer-based implementations of clinical pathways that facilitate and standardise entry of diagnostic and therapeutic orders.
• Algorithms and tools to enhance physician efficiency and patient satisfaction.

One of the most widely used clinical decision support systems (CDSS) is Micromedex. This company was founded in 1974 in Denver, Colorado, and is now owned by Thompson Reuters. Pharmacists at more than 3,500 US hospitals and in 83 countries rely on Micromedex to deliver evidence-based information that spans drugs, disease management, toxicology,
alternative medicines and neonatal dosing.

The company began by delivering product on microfiche and then led the field in moving to CDs in the mid-1980s. Now, the latest version, Micromedex 2.0, is web-based. Approximately 90 full-time editorial staff members are currently employed, including physicians, clinical pharmacists, nurses and other allied health professionals as well as medical librarians. All are involved in the process to create and review the content of the newly released Micromedex 2.0. Over half a million articles are reviewed, evaluated and synthesised annually to provide the best treatment recommendations and guidelines.

The editorial team is trained in the identification of relevant literature and accepted literature evaluation techniques that assess methodological rigour, appropriateness of statistical analyses, as well as clinical relevance. These literature evaluation skills, in conjunction with clinical judgement, are employed throughout the content creation and review process.
The content facilitates the practice of evidence-based medicine by helping the clinician to identify best practices and choose the most appropriate treatment plan for a specific patient. There are five steps in the creation of the content:

STEP 1: Identification of literature and topics for inclusion
Senior editorial staff consider many factors when selecting database subjects.

Topics chosen for further research are based on ongoing review of the world’s medical journals, clinical judgement and recommendations, regulatory standards and compliance, national healthcare trends, FDA approvals, editorial board suggestions, external requests, and policy changes in health and disease management from professional health organisations.

The foundation of the ongoing literature surveillance is an automated literature query designed and monitored by internal medical library staff and senior editorial staff. This literature evaluation process comprises three levels: surveillance; title and abstract analysis; and full-text analysis.

This three-step approach supports objective and systematic selection of the most important evidence based research. Selected literature is prioritised and launched into the editorial workflow for content creation.

STEP 2: Creation of content
Primary literature identified in Step 1 is forwarded to a clinical writer to assess further the appropriateness of inclusion in the databases. This additional assessment is based on accepted literature evaluation techniques and includes an analysis of methodological rigour, application of statistics and clinical relevance.

If deemed appropriate for inclusion, the information is synthesised and content is created in the Content Management System. Clinical writers adhere to internal style guides, policies and procedures to create accurate, consistent content. Once written, new content is forwarded for internal review.

STEP 3: Internal review of content
New content is reviewed by an internal senior clinical staff member. The reviewer assesses clinical accuracy and relevance as well as adherence to internal style guides, policies and procedures. Content creation and review is an iterative process in which the reviewer provides feedback to the writer. Content
may circulate through the creation and review process several times before moving to the next step. This ensures the content is of the highest quality. The senior clinical staff and internal experts meet regularly to validate content decisions.

STEP 4: Editorial Board review of content
Certain topics may undergo additional review by an
Editorial Board.

STEP 5: Final clinical review and promotion to production
An internal senior clinical staff member performs a final review of the content, taking into account any review: CDSS 27
Content is passed to production and is available for release.

Micromedex 2.0 is produced using an editorial process that remains completely unbiased and assesses methodological rigour, appropriateness of statistical analyses, as well as clinical relevance, so healthcare professionals get the right information, at the right time and in the right format. This new solution builds on the strength of the Micromedex clinical content, and adds a new approach to search and interface design. Micromedex 2.0 also incorporates compatibility with 3G smartphones and devices, including iPhone, iPod Touch and BlackBerry.

Speaking about the latest release of Micromedex, Jill Sutton, Vice President of Product Management at Thomson Reuters said: “Clinicians know we [Thomson Reuters] have the right answers, but they do not have time to spend searching for information. So we rebuilt the user interface and search functions to fit the clinician’s workflow and help them find the right answers more quickly”.

The result is that pharmacists and other clinicians can get the benefit of the content in a new package that streamlines workflow, speeds responses to clinical questions and helps them to be more efficient.

Win Eckert, Director of Product Management at Thomson Reuters says delivering its latest release of Micromedex was no easy task. We conducted ethnographic studies and informational interviews with pharmacists, nurses and MDs,” he said, “we then went through several rounds of user-focused testing and 6-7 iterations of product prototypes with 60 plus end users”.

As the gap between the amount of information available and the time allowed to process it grows wider it becomes important that readers are helped to synthesise and summarise data, supported by credible, justified statements from high-quality sources. This ensures integrity in the provision of evidencebased guidance, services and solutions, while guarding against publication bias and information of little or no relevance.

References
1. Rynja FJ et al. EJHP Practice 2005/6;12:53.