- Email: [email protected]
Using computerized clinical decision support for latent tuberculosis infection screening
Brief Reports
Using Computerized Clinical Decision Support for Latent Tuberculosis Infection Screening Andy W. Steele, MD, MPH, Sheri Eisert, PhD, Art Davidson, MD, MPH, Taylor Sandison, MD, Pat Lyons, Nedra Garrett, Patricia Gabow, MD, Eduardo Ortiz, MD, MPH Background: The Centers for Disease Control and Prevention (CDC) has published guidelines recommending screening high-risk groups for latent tuberculosis infection (LTBI). The goal of this study was to determine the impact of computerized clinical decision support and guided web-based documentation on screening rates for LTBI. Design:
Nonrandomized, prospective, intervention study.
Setting and Participants were 8463 patients seen at two primary care, outpatient, public community participants: health center clinics in late 2002 and early 2003. Intervention: The CDC’s LTBI guidelines were encoded into a computerized clinical decision support system that provided an alert recommending further assessment of LTBI risk if certain guideline criteria were met (birth in a high-risk TB country and aged ⬍40). A guided web-based documentation tool was provided to facilitate appropriate adherence to the LTBI screening guideline and to promote accurate documentation and evaluation. Baseline data were collected for 15 weeks and study-phase data were collected for 12 weeks. Main outcome measures:
Appropriate LTBI screening according to CDC guidelines based on chart review.
Results:
Among 4135 patients registering during the post-intervention phase, 73% had at least one CDC-defined risk factor, and 610 met the alert criteria (birth in a high-risk TB country and aged ⬍40 years) for potential screening for LTBI. Adherence with the LTBI screening guideline improved significantly from 8.9% at baseline to 25.2% during the study phase (183% increase, p ⬍ 0.001).
Conclusions: This study demonstrated that computerized, clinical decision support using alerts and guided web-based documentation increased screening of high-risk patients for LTBI. This type of technology could lead to an improvement in LTBI screening in the United States and also holds promise for improved care for other preventive and chronic conditions. (Am J Prev Med 2005;28(3):281–284) © 2005 American Journal of Preventive Medicine
Background
T
uberculosis (TB) remains a major disease in the United States and in the world. Given the estimated 2 million deaths annually, tuberculosis is the second leading infectious cause of death worldwide behind human immunodeficiency virus (HIV).1 Among U.S.-born people, there was a 62% decrease in From Information Services, Denver Health (Steele), Health Services Research, Denver Health (Eisert), Public Health, Denver Health (Davidson), Department of Medicine, Denver Health (Gabow), and University of Colorado Health Sciences Center (Sandison), Denver, Colorado; Siemens Medical Solutions USA, Inc. (Lyons), Malvern, Pennsylvania; Information Resource Management Office, Centers for Disease Control and Prevention (Garrett), Atlanta, Georgia; and Veterans Administration Medical Center (Ortiz), Washington DC Address correspondence and reprint requests to: Andy Steele, MD, MPH, Director, Medical Informatics, Denver Health (1932), 660 Bannock St., Denver CO 80218. E-mail: [email protected].
the number of reported TB cases from 1992 to 2002. In contrast, there was a 5% increase in the reported TB cases among foreign-born people during the same time period.2 With over 50% of new cases occurring in people born outside of the United States, both the Centers for Disease Control and Prevention (CDC) and the Institute of Medicine have advocated targeted latent tuberculosis infection (LTBI) screening to include screening high-risk groups, such as foreign-born people, those with HIV infection, and the homeless, rather than community-wide screening.3,4 The increased use of computerized clinical decision support for the management of tuberculosis treatment has been recommended by some.5 Electronic medical record systems with decision support capabilities provide a potentially powerful mechanism for the dissemination and integration of guidelines at the point of
Am J Prev Med 2005;28(3) © 2005 American Journal of Preventive Medicine • Published by Elsevier Inc.
0749-3797/05/$–see front matter doi:10.1016/j.amepre.2004.12.012
281
care.6 – 8 The purpose of this study was to evaluate the effects of computerized clinical decision support on the screening of LTBI in a public healthcare setting.
Methods A computerized clinical decision support system at two outpatient primary care clinics in Denver CO was used to assess screening for latent tuberculosis infection following the CDC guideline: “Targeted Tuberculin Testing and Treatment of Latent Tuberculosis.”4 In collaboration with the Public Health Practice Program Office of the CDC and Siemens Medical Solutions USA, Inc., the CDC LTBI screening guidelines were encoded into a computerized clinical decision support system using a rules engine to alert clinic staff of the potential need for screening for LTBI. Providers utilized a guided web-based data entry tool to document appropriate LTBI screening.
Patients and Setting This study took place at two community health center clinics at Denver Health (DH), an integrated public healthcare system serving as Colorado’s principal safety-net institution. Annually, DH serves approximately 150,000 individuals, where 40% are uninsured and 70% are racial/ethnic minorities, many of whom are born outside of the United States. All registered patients were eligible for the intervention.
Data and Timeframe Baseline results of the computerized clinical decision support system application were collected for 15 weeks between October 17, 2002 and January 29, 2003. Data from after the intervention were collected for 12 weeks between February 6, 2003 and April 30, 2003.
Analytical Approach This study consisted of a nonrandomized implementation analysis for CDC LTBI guideline adherence. A random sample chart audit was conducted at both clinics during the study period. For continuous variables, t -tests were used to assess statistically significant differences before and after the study phase. Chi-square analysis was used for categorical variables. Analysis was completed in July 2003.
Study Phase The implementation consisted of two components: (1) automated generation of a paper alert to clinic staff based on computerized rules interpretation of the CDC LTBI guideline; and (2) creation of a guided web-based data entry tool within the current electronic medical record to facilitate appropriate LTBI screening, which was used by providers and clinical staff in response to the alert. During the patient registration process, a computerized rule that contained the encoded CDC LTBI guideline processed data collected during patient registration (date of birth, country of birth, patient medical record number, homeless status), and clinical data collected from billing databases (International Classification of Diseases-9 [ICD-9]
282
codes) from previous outpatient encounters. Country codes were a required field during registration, and were coded by selecting the appropriate two-letter International Organization for Standardization classification. If specific rules criteria were met (i.e., the patient was born in a high-risk TB country and aged ⬍40 years), a paper alert was automatically printed with the patient’s registration paperwork. This alert was placed on top of the patient’s chart and served to remind clinic staff that this person needed further assessment for potential LTBI screening. The paper alert was used, since not all providers used a computer during all patient interactions. Although all components of the LTBI guideline were assessed, reminders were generated for just one of the CDC-identified risk factors: patients born in countries with a high rate of TB prevalence. The computerized clinical decision support application had the flexibility to alert on any available set of parameters, but only one risk factor was selected with an age limitation of ⬍40 years, since a baseline operational impact analysis indicated that as many as 75% of patients would trigger the rule if the alert was based on all CDC risk factors. To further reduce this operational burden, and yet focus on the highest-risk patients, patients aged ⬍40 years were selected since they were more likely to be recent immigrants to the United States. In addition, there was concern that over-alerting providers would lead to a lack of compliance with guidelines for those at highest risk. Provider staff were expected to document their assessments within the existing electronic medical record using a guided web-based application. Nursing and medical staff assistants had responsibility for placement and subsequent interpretation of the skin test purified protein derivative (PPD). The structured format of the documentation tool facilitated the collection of appropriate information identified in the CDC guidelines.
Primary Outcome Measures The overall outcome measure was appropriate adherence with CDC LTBI screening guidelines, which was defined by the presence of documentation regarding the following: Interpretation of a PPD performed as part of screening for LTBI either during the study period or previously, or previous active or latent TB infection, or absence of LTBI screening risk factors due to further assessment, or patient refusal to undergo LTBI screening. The provider adherence rates were determined from a random chart audit conducted during the baseline and study periods. A total of 249 charts were reviewed based on sample size and power estimates—146 during the baseline period and 103 during the study period.
Results There were 4328 and 4135 unique adult registrations during the baseline and study phases, respectively. The average age of this group was 49 years, 64% were female, 71% were Hispanic, and 50% were uninsured. Seventy-four percent of the baseline patients (3213 of 4328), and 73% of the study period patients (3034 of
American Journal of Preventive Medicine, Volume 28, Number 3
4135) had at least one LTBI risk factor that qualified for screening by CDC guidelines. Country of birth was the most common risk factor, accounting for 39% of the patients registered. Of these, 97% were born in Mexico. Clinical risk factors were present in 49% of the patients, the most prevalent of which were diabetes (23%), hematologic disorders (17%), chronic alcoholism (13%), immunosuppression (10%), and chronic liver disease (8%), with homelessness (5%) and injection drug use (2%) being less common. There were no statistically significant differences between the baseline and study period groups for the above demographic and clinical variables. Among the 4328 and 4135 patients that were seen during the baseline and study phases, respectively, 683 (16%) and 610 (15%) patients met the study intervention criteria of being born in a high-risk TB country and aged ⬍40 years based on electronic query of electronic medical record data. Compared to the average clinic patient, the study groups were more likely to be Hispanic (94% vs 75%, p ⬍ 0.01) and uninsured (90% vs 62%, p ⬍ 0.01). The rule processed appropriately, and generated a paper alert in 608 of 610 patients (99.7%). The two patients (0.3%) for whom the rule did not generate an alert were attributed to dropped transactions, system downtime, or other unexplained technologic problems. Among these 608 patients in the 12-week intervention period when alerts were generated, providers utilized the guided web-based assessment for 156 (25.6%) of the patients. Among the 156 high-risk patients who had some web-based LTBI assessment, providers completed appropriate screening in 105 (67%), which represents 17% of the original 608 patients. A PPD was placed on 87 (83%), and the PPD was read within 48 to 72 hours in 67 (77%) of those skin tested. Among the 67 patients who returned for PPD skin test reading, 30 (45%) patients had a positive PPD by CDC criteria. Patients with positive PPDs were further evaluated and treated as deemed appropriate. Of 249 charts reviewed, higher LTBI guideline adherence was achieved after implementation of the intervention. Adherence increased 183% from a baseline of 8.9% to 25.2% during the study period with a p value of ⬍0.001 (Figure 1).
Discussion This study demonstrated successful application of computerized clinical decision support to adapt a national clinical guideline to local needs in a safety-net institution outpatient setting. This study is the first to demonstrate that computerized clinical decision support systems can improve screening for latent tuberculosis infection. It was surprising to observe that ⬎70% of the patients seen in the safety-net setting had at least one CDC-defined LTBI risk factor. In addition, the high
None or partial LTBI screening Complete LTBI screening
100% 80% 60%
77 133
40% 25.2%
20% 0%
8.9%
13
Baseline
26
Study phase
Figure 1. Provider adherence with latent tuberculosis infection (LTBI) screening guidelines (p ⬍ 0.001).
PPD-positive rate of 45% among those tested highlights the importance of screening in this select population. This study addressed only LTBI screening, while future studies will need also to focus on LTBI treatment to fully understand the overall clinical impact. In addition, providers identified additional work required with web-based assessment and documentation as a barrier to screening patients. Finally, given the limited duration of evaluation during the study phase, it will be important to assess the long-term sustainability of this type of intervention. Ultimately, if LTBI screening were improved, one would expect fewer conversions to active tuberculosis. This in turn would lead to decreased transmission of disease to close contacts. The increasing use of electronic medical records provides a new and potentially powerful mechanism for the dissemination and integration of various types of guidelines at the point of care with both individual and public health benefits. This research was performed by Denver Health through the support of the Agency for Healthcare Research and Quality (AHRQ) (contract 290-00-0014, task order 3). The authors of this article are solely responsible for its contents. No statements or views in this article should be construed as endorsements or positions of AHRQ, the U.S. Department of Health and Human Services, the Department of Veterans Affairs, the Centers for Disease Control and Prevention, Siemens Medical Solutions USA, Inc., or the federal government. No financial conflict of interest was reported by the authors of this paper.
References 1. Dye C, Scheele S, Dolin P, Pathania V, Raviglione MC. Consensus statement. Global burden of tuberculosis: estimated incidence, prevalence, and mortality by country. WHO Global Surveillance and Monitoring Project. JAMA 1999;282:677– 86.
Am J Prev Med 2005;28(3)
283
What This Study Adds . . . The Centers for Disease Control and Prevention recommends screening high-risk patients for latent tuberculosis infection (LTBI). Few studies have evaluated the use of computerized clinical decision support (CDSS) to improve screening rates. This study evaluated the impact of CDSS and guided web-based documentation on screening rates for LTBI, and demonstrated a significant improvement in screening rates for LTBI. This type of technology holds promise for improved screening for LTBI and other health conditions.
284
2. Centers for Disease Control and Prevention. Reported tuberculosis in the United States, 2002. Atlanta GA: U.S. Department of Health and Human Services, September 2003. 3. Centers for Disease Control and Prevention. Screening for tuberculosis and tuberculosis infection in high-risk populations. Recommendations of the Advisory Council for the Elimination of Tuberculosis. MMWR Recomm Rep 1995;44:19 –34. 4. Centers for Disease Control and Prevention. Targeted tuberculin testing and treatment of latent tuberculosis infection. MMWR Morb Mortal Wkly Rep 2000;49:1–51. 5. Tanser F, Wilkinson D. Spatial implications of the tuberculosis DOTS strategy in rural South Africa: a novel application of geographical information system and global positioning system technologies. Trop Med Int Health 1999;4:634 – 8. 6. Hunt DL, Haynes RB, Hanna SE, Smith K. Effects of computer-based clinical decision support systems on physician performance and patient outcomes: a systematic review. JAMA 1998;280:1339 – 46. 7. Pestotnik SL, Classen DC, Evans RS, Burke JP. Implementing antibiotic practice guidelines through computer-assisted decision support: clinical and financial outcomes. Ann Intern Med 1996;124:884 –90. 8. Garrett N, Yasnoff W. Disseminating public health practice guidelines in electronic medical record systems. J Public Health Manag Pract 2002;8:1–10.
American Journal of Preventive Medicine, Volume 28, Number 3
Using Computerized Clinical Decision Support for Latent Tuberculosis Infection Screening Andy W. Steele, MD, MPH, Sheri Eisert, PhD, Art Davidson, MD, MPH, Taylor Sandison, MD, Pat Lyons, Nedra Garrett, Patricia Gabow, MD, Eduardo Ortiz, MD, MPH Background: The Centers for Disease Control and Prevention (CDC) has published guidelines recommending screening high-risk groups for latent tuberculosis infection (LTBI). The goal of this study was to determine the impact of computerized clinical decision support and guided web-based documentation on screening rates for LTBI. Design:
Nonrandomized, prospective, intervention study.
Setting and Participants were 8463 patients seen at two primary care, outpatient, public community participants: health center clinics in late 2002 and early 2003. Intervention: The CDC’s LTBI guidelines were encoded into a computerized clinical decision support system that provided an alert recommending further assessment of LTBI risk if certain guideline criteria were met (birth in a high-risk TB country and aged ⬍40). A guided web-based documentation tool was provided to facilitate appropriate adherence to the LTBI screening guideline and to promote accurate documentation and evaluation. Baseline data were collected for 15 weeks and study-phase data were collected for 12 weeks. Main outcome measures:
Appropriate LTBI screening according to CDC guidelines based on chart review.
Results:
Among 4135 patients registering during the post-intervention phase, 73% had at least one CDC-defined risk factor, and 610 met the alert criteria (birth in a high-risk TB country and aged ⬍40 years) for potential screening for LTBI. Adherence with the LTBI screening guideline improved significantly from 8.9% at baseline to 25.2% during the study phase (183% increase, p ⬍ 0.001).
Conclusions: This study demonstrated that computerized, clinical decision support using alerts and guided web-based documentation increased screening of high-risk patients for LTBI. This type of technology could lead to an improvement in LTBI screening in the United States and also holds promise for improved care for other preventive and chronic conditions. (Am J Prev Med 2005;28(3):281–284) © 2005 American Journal of Preventive Medicine
Background
T
uberculosis (TB) remains a major disease in the United States and in the world. Given the estimated 2 million deaths annually, tuberculosis is the second leading infectious cause of death worldwide behind human immunodeficiency virus (HIV).1 Among U.S.-born people, there was a 62% decrease in From Information Services, Denver Health (Steele), Health Services Research, Denver Health (Eisert), Public Health, Denver Health (Davidson), Department of Medicine, Denver Health (Gabow), and University of Colorado Health Sciences Center (Sandison), Denver, Colorado; Siemens Medical Solutions USA, Inc. (Lyons), Malvern, Pennsylvania; Information Resource Management Office, Centers for Disease Control and Prevention (Garrett), Atlanta, Georgia; and Veterans Administration Medical Center (Ortiz), Washington DC Address correspondence and reprint requests to: Andy Steele, MD, MPH, Director, Medical Informatics, Denver Health (1932), 660 Bannock St., Denver CO 80218. E-mail: [email protected].
the number of reported TB cases from 1992 to 2002. In contrast, there was a 5% increase in the reported TB cases among foreign-born people during the same time period.2 With over 50% of new cases occurring in people born outside of the United States, both the Centers for Disease Control and Prevention (CDC) and the Institute of Medicine have advocated targeted latent tuberculosis infection (LTBI) screening to include screening high-risk groups, such as foreign-born people, those with HIV infection, and the homeless, rather than community-wide screening.3,4 The increased use of computerized clinical decision support for the management of tuberculosis treatment has been recommended by some.5 Electronic medical record systems with decision support capabilities provide a potentially powerful mechanism for the dissemination and integration of guidelines at the point of
Am J Prev Med 2005;28(3) © 2005 American Journal of Preventive Medicine • Published by Elsevier Inc.
0749-3797/05/$–see front matter doi:10.1016/j.amepre.2004.12.012
281
care.6 – 8 The purpose of this study was to evaluate the effects of computerized clinical decision support on the screening of LTBI in a public healthcare setting.
Methods A computerized clinical decision support system at two outpatient primary care clinics in Denver CO was used to assess screening for latent tuberculosis infection following the CDC guideline: “Targeted Tuberculin Testing and Treatment of Latent Tuberculosis.”4 In collaboration with the Public Health Practice Program Office of the CDC and Siemens Medical Solutions USA, Inc., the CDC LTBI screening guidelines were encoded into a computerized clinical decision support system using a rules engine to alert clinic staff of the potential need for screening for LTBI. Providers utilized a guided web-based data entry tool to document appropriate LTBI screening.
Patients and Setting This study took place at two community health center clinics at Denver Health (DH), an integrated public healthcare system serving as Colorado’s principal safety-net institution. Annually, DH serves approximately 150,000 individuals, where 40% are uninsured and 70% are racial/ethnic minorities, many of whom are born outside of the United States. All registered patients were eligible for the intervention.
Data and Timeframe Baseline results of the computerized clinical decision support system application were collected for 15 weeks between October 17, 2002 and January 29, 2003. Data from after the intervention were collected for 12 weeks between February 6, 2003 and April 30, 2003.
Analytical Approach This study consisted of a nonrandomized implementation analysis for CDC LTBI guideline adherence. A random sample chart audit was conducted at both clinics during the study period. For continuous variables, t -tests were used to assess statistically significant differences before and after the study phase. Chi-square analysis was used for categorical variables. Analysis was completed in July 2003.
Study Phase The implementation consisted of two components: (1) automated generation of a paper alert to clinic staff based on computerized rules interpretation of the CDC LTBI guideline; and (2) creation of a guided web-based data entry tool within the current electronic medical record to facilitate appropriate LTBI screening, which was used by providers and clinical staff in response to the alert. During the patient registration process, a computerized rule that contained the encoded CDC LTBI guideline processed data collected during patient registration (date of birth, country of birth, patient medical record number, homeless status), and clinical data collected from billing databases (International Classification of Diseases-9 [ICD-9]
282
codes) from previous outpatient encounters. Country codes were a required field during registration, and were coded by selecting the appropriate two-letter International Organization for Standardization classification. If specific rules criteria were met (i.e., the patient was born in a high-risk TB country and aged ⬍40 years), a paper alert was automatically printed with the patient’s registration paperwork. This alert was placed on top of the patient’s chart and served to remind clinic staff that this person needed further assessment for potential LTBI screening. The paper alert was used, since not all providers used a computer during all patient interactions. Although all components of the LTBI guideline were assessed, reminders were generated for just one of the CDC-identified risk factors: patients born in countries with a high rate of TB prevalence. The computerized clinical decision support application had the flexibility to alert on any available set of parameters, but only one risk factor was selected with an age limitation of ⬍40 years, since a baseline operational impact analysis indicated that as many as 75% of patients would trigger the rule if the alert was based on all CDC risk factors. To further reduce this operational burden, and yet focus on the highest-risk patients, patients aged ⬍40 years were selected since they were more likely to be recent immigrants to the United States. In addition, there was concern that over-alerting providers would lead to a lack of compliance with guidelines for those at highest risk. Provider staff were expected to document their assessments within the existing electronic medical record using a guided web-based application. Nursing and medical staff assistants had responsibility for placement and subsequent interpretation of the skin test purified protein derivative (PPD). The structured format of the documentation tool facilitated the collection of appropriate information identified in the CDC guidelines.
Primary Outcome Measures The overall outcome measure was appropriate adherence with CDC LTBI screening guidelines, which was defined by the presence of documentation regarding the following: Interpretation of a PPD performed as part of screening for LTBI either during the study period or previously, or previous active or latent TB infection, or absence of LTBI screening risk factors due to further assessment, or patient refusal to undergo LTBI screening. The provider adherence rates were determined from a random chart audit conducted during the baseline and study periods. A total of 249 charts were reviewed based on sample size and power estimates—146 during the baseline period and 103 during the study period.
Results There were 4328 and 4135 unique adult registrations during the baseline and study phases, respectively. The average age of this group was 49 years, 64% were female, 71% were Hispanic, and 50% were uninsured. Seventy-four percent of the baseline patients (3213 of 4328), and 73% of the study period patients (3034 of
American Journal of Preventive Medicine, Volume 28, Number 3
4135) had at least one LTBI risk factor that qualified for screening by CDC guidelines. Country of birth was the most common risk factor, accounting for 39% of the patients registered. Of these, 97% were born in Mexico. Clinical risk factors were present in 49% of the patients, the most prevalent of which were diabetes (23%), hematologic disorders (17%), chronic alcoholism (13%), immunosuppression (10%), and chronic liver disease (8%), with homelessness (5%) and injection drug use (2%) being less common. There were no statistically significant differences between the baseline and study period groups for the above demographic and clinical variables. Among the 4328 and 4135 patients that were seen during the baseline and study phases, respectively, 683 (16%) and 610 (15%) patients met the study intervention criteria of being born in a high-risk TB country and aged ⬍40 years based on electronic query of electronic medical record data. Compared to the average clinic patient, the study groups were more likely to be Hispanic (94% vs 75%, p ⬍ 0.01) and uninsured (90% vs 62%, p ⬍ 0.01). The rule processed appropriately, and generated a paper alert in 608 of 610 patients (99.7%). The two patients (0.3%) for whom the rule did not generate an alert were attributed to dropped transactions, system downtime, or other unexplained technologic problems. Among these 608 patients in the 12-week intervention period when alerts were generated, providers utilized the guided web-based assessment for 156 (25.6%) of the patients. Among the 156 high-risk patients who had some web-based LTBI assessment, providers completed appropriate screening in 105 (67%), which represents 17% of the original 608 patients. A PPD was placed on 87 (83%), and the PPD was read within 48 to 72 hours in 67 (77%) of those skin tested. Among the 67 patients who returned for PPD skin test reading, 30 (45%) patients had a positive PPD by CDC criteria. Patients with positive PPDs were further evaluated and treated as deemed appropriate. Of 249 charts reviewed, higher LTBI guideline adherence was achieved after implementation of the intervention. Adherence increased 183% from a baseline of 8.9% to 25.2% during the study period with a p value of ⬍0.001 (Figure 1).
Discussion This study demonstrated successful application of computerized clinical decision support to adapt a national clinical guideline to local needs in a safety-net institution outpatient setting. This study is the first to demonstrate that computerized clinical decision support systems can improve screening for latent tuberculosis infection. It was surprising to observe that ⬎70% of the patients seen in the safety-net setting had at least one CDC-defined LTBI risk factor. In addition, the high
None or partial LTBI screening Complete LTBI screening
100% 80% 60%
77 133
40% 25.2%
20% 0%
8.9%
13
Baseline
26
Study phase
Figure 1. Provider adherence with latent tuberculosis infection (LTBI) screening guidelines (p ⬍ 0.001).
PPD-positive rate of 45% among those tested highlights the importance of screening in this select population. This study addressed only LTBI screening, while future studies will need also to focus on LTBI treatment to fully understand the overall clinical impact. In addition, providers identified additional work required with web-based assessment and documentation as a barrier to screening patients. Finally, given the limited duration of evaluation during the study phase, it will be important to assess the long-term sustainability of this type of intervention. Ultimately, if LTBI screening were improved, one would expect fewer conversions to active tuberculosis. This in turn would lead to decreased transmission of disease to close contacts. The increasing use of electronic medical records provides a new and potentially powerful mechanism for the dissemination and integration of various types of guidelines at the point of care with both individual and public health benefits. This research was performed by Denver Health through the support of the Agency for Healthcare Research and Quality (AHRQ) (contract 290-00-0014, task order 3). The authors of this article are solely responsible for its contents. No statements or views in this article should be construed as endorsements or positions of AHRQ, the U.S. Department of Health and Human Services, the Department of Veterans Affairs, the Centers for Disease Control and Prevention, Siemens Medical Solutions USA, Inc., or the federal government. No financial conflict of interest was reported by the authors of this paper.
References 1. Dye C, Scheele S, Dolin P, Pathania V, Raviglione MC. Consensus statement. Global burden of tuberculosis: estimated incidence, prevalence, and mortality by country. WHO Global Surveillance and Monitoring Project. JAMA 1999;282:677– 86.
Am J Prev Med 2005;28(3)
283
What This Study Adds . . . The Centers for Disease Control and Prevention recommends screening high-risk patients for latent tuberculosis infection (LTBI). Few studies have evaluated the use of computerized clinical decision support (CDSS) to improve screening rates. This study evaluated the impact of CDSS and guided web-based documentation on screening rates for LTBI, and demonstrated a significant improvement in screening rates for LTBI. This type of technology holds promise for improved screening for LTBI and other health conditions.
284
2. Centers for Disease Control and Prevention. Reported tuberculosis in the United States, 2002. Atlanta GA: U.S. Department of Health and Human Services, September 2003. 3. Centers for Disease Control and Prevention. Screening for tuberculosis and tuberculosis infection in high-risk populations. Recommendations of the Advisory Council for the Elimination of Tuberculosis. MMWR Recomm Rep 1995;44:19 –34. 4. Centers for Disease Control and Prevention. Targeted tuberculin testing and treatment of latent tuberculosis infection. MMWR Morb Mortal Wkly Rep 2000;49:1–51. 5. Tanser F, Wilkinson D. Spatial implications of the tuberculosis DOTS strategy in rural South Africa: a novel application of geographical information system and global positioning system technologies. Trop Med Int Health 1999;4:634 – 8. 6. Hunt DL, Haynes RB, Hanna SE, Smith K. Effects of computer-based clinical decision support systems on physician performance and patient outcomes: a systematic review. JAMA 1998;280:1339 – 46. 7. Pestotnik SL, Classen DC, Evans RS, Burke JP. Implementing antibiotic practice guidelines through computer-assisted decision support: clinical and financial outcomes. Ann Intern Med 1996;124:884 –90. 8. Garrett N, Yasnoff W. Disseminating public health practice guidelines in electronic medical record systems. J Public Health Manag Pract 2002;8:1–10.
American Journal of Preventive Medicine, Volume 28, Number 3