Limits...
Evaluation of internet-based clinical decision support systems.

Thomas KW, Dayton CS, Peterson MW - J. Med. Internet Res. (1999 Oct-Dec)

Bottom Line: We similarly compared residents using the computerized TB DSS to residents using a printed reference card; the residents using the computerized DSS scored significantly better (95.8% vs. 56.6% correct; p<0.001).Both systems improved physician compliance with national guidelines when tested in clinical scenarios.By providing information that is coupled to relevant activity, we expect that these widely available DSS's will serve as effective educational tools to positively impact physician behavior.

View Article: PubMed Central - HTML - PubMed

Affiliation: Division of Pulmonary, Critical Care and Occupational Medicine, University of Iowa, Iowa City, Iowa, USA.

ABSTRACT

Background: Scientifically based clinical guidelines have become increasingly used to educate physicians and improve quality of care. While individual guidelines are potentially useful, repeated studies have shown that guidelines are ineffective in changing physician behavior. The Internet has evolved as a potentially useful tool for guideline education, dissemination, and implementation because of its open standards and its ability to provide concise, relevant clinical information at the location and time of need.

Objective: Our objective was to develop and test decision support systems (DSS) based on clinical guidelines which could be delivered over the Internet for two disease models: asthma and tuberculosis (TB) preventive therapy.

Methods: Using open standards of HTML and CGI, we developed an acute asthma severity assessment DSS and a preventative tuberculosis treatment DSS based on content from national guidelines that are recognized as standards of care. Both DSS's are published on the Internet and operate through a decision algorithm developed from the parent guidelines with clinical information provided by the user at the point of clinical care. We tested the effectiveness of each DSS in influencing physician decisions using clinical scenario testing.

Results: We first validated the asthma algorithm by comparing asthma experts' decisions with the decisions reached by nonpulmonary nurses using the computerized DSS. Using the DSS, nurses scored the same as experts (89% vs. 88%; p = NS). Using the same scenario test instrument, we next compared internal medicine residents using the DSS with residents using a printed version of the National Asthma Education Program-2 guidelines. Residents using the computerized DSS scored significantly better than residents using the paper-based guidelines (92% vs. 84%; p <0.002). We similarly compared residents using the computerized TB DSS to residents using a printed reference card; the residents using the computerized DSS scored significantly better (95.8% vs. 56.6% correct; p<0.001).

Conclusions: Previous work has shown that guidelines disseminated through traditional educational interventions have minimal impact on physician behavior. Although computerized DSS have been effective in altering physician behavior, many of these systems are not widely available. We have developed two clinical DSS's based on national guidelines and published them on the Internet. Both systems improved physician compliance with national guidelines when tested in clinical scenarios. By providing information that is coupled to relevant activity, we expect that these widely available DSS's will serve as effective educational tools to positively impact physician behavior.

Show MeSH

Related in: MedlinePlus

© Copyright Policy
Related In: Results  -  Collection


getmorefigures.php?uid=PMC1761710&req=5

Mentions: We next examined the effect of the DSS on the decision-making process among similarly trained physicians. An unselected group of Internal Medicine residents was randomly divided and assigned to use either the DSS (N=11) or a printed copy of the Practical Guide for the Diagnosis and Management of Asthma [16]. The residents using the DSS performed much better than the residents using the printed guidelines (mean score 91.6%, CI 88.0-95.3% compared to 83.6%, CI 80.5-86.7%; p = 0.001) (Figure 3). There was no statistical difference between the mean scores of the expert panel and the residents using the DSS (p = 0.26). However, the residents using only the printed materials performed statistically worse than the experts (p = 0.017). Both the resident group using the computer DSS and the resident group using the printed resources completed all patient scenarios in less than 30 minutes (% minutes per case).


Evaluation of internet-based clinical decision support systems.

Thomas KW, Dayton CS, Peterson MW - J. Med. Internet Res. (1999 Oct-Dec)

© Copyright Policy
Related In: Results  -  Collection

Show All Figures
getmorefigures.php?uid=PMC1761710&req=5

Mentions: We next examined the effect of the DSS on the decision-making process among similarly trained physicians. An unselected group of Internal Medicine residents was randomly divided and assigned to use either the DSS (N=11) or a printed copy of the Practical Guide for the Diagnosis and Management of Asthma [16]. The residents using the DSS performed much better than the residents using the printed guidelines (mean score 91.6%, CI 88.0-95.3% compared to 83.6%, CI 80.5-86.7%; p = 0.001) (Figure 3). There was no statistical difference between the mean scores of the expert panel and the residents using the DSS (p = 0.26). However, the residents using only the printed materials performed statistically worse than the experts (p = 0.017). Both the resident group using the computer DSS and the resident group using the printed resources completed all patient scenarios in less than 30 minutes (% minutes per case).

Bottom Line: We similarly compared residents using the computerized TB DSS to residents using a printed reference card; the residents using the computerized DSS scored significantly better (95.8% vs. 56.6% correct; p<0.001).Both systems improved physician compliance with national guidelines when tested in clinical scenarios.By providing information that is coupled to relevant activity, we expect that these widely available DSS's will serve as effective educational tools to positively impact physician behavior.

View Article: PubMed Central - HTML - PubMed

Affiliation: Division of Pulmonary, Critical Care and Occupational Medicine, University of Iowa, Iowa City, Iowa, USA.

ABSTRACT

Background: Scientifically based clinical guidelines have become increasingly used to educate physicians and improve quality of care. While individual guidelines are potentially useful, repeated studies have shown that guidelines are ineffective in changing physician behavior. The Internet has evolved as a potentially useful tool for guideline education, dissemination, and implementation because of its open standards and its ability to provide concise, relevant clinical information at the location and time of need.

Objective: Our objective was to develop and test decision support systems (DSS) based on clinical guidelines which could be delivered over the Internet for two disease models: asthma and tuberculosis (TB) preventive therapy.

Methods: Using open standards of HTML and CGI, we developed an acute asthma severity assessment DSS and a preventative tuberculosis treatment DSS based on content from national guidelines that are recognized as standards of care. Both DSS's are published on the Internet and operate through a decision algorithm developed from the parent guidelines with clinical information provided by the user at the point of clinical care. We tested the effectiveness of each DSS in influencing physician decisions using clinical scenario testing.

Results: We first validated the asthma algorithm by comparing asthma experts' decisions with the decisions reached by nonpulmonary nurses using the computerized DSS. Using the DSS, nurses scored the same as experts (89% vs. 88%; p = NS). Using the same scenario test instrument, we next compared internal medicine residents using the DSS with residents using a printed version of the National Asthma Education Program-2 guidelines. Residents using the computerized DSS scored significantly better than residents using the paper-based guidelines (92% vs. 84%; p <0.002). We similarly compared residents using the computerized TB DSS to residents using a printed reference card; the residents using the computerized DSS scored significantly better (95.8% vs. 56.6% correct; p<0.001).

Conclusions: Previous work has shown that guidelines disseminated through traditional educational interventions have minimal impact on physician behavior. Although computerized DSS have been effective in altering physician behavior, many of these systems are not widely available. We have developed two clinical DSS's based on national guidelines and published them on the Internet. Both systems improved physician compliance with national guidelines when tested in clinical scenarios. By providing information that is coupled to relevant activity, we expect that these widely available DSS's will serve as effective educational tools to positively impact physician behavior.

Show MeSH
Related in: MedlinePlus