Lessons learned from implementing service-oriented clinical decision support at four sites: A qualitative study.

OBJECTIVE: To identify challenges, lessons learned and best practices for service-oriented clinical decision support, based on the results of the Clinical Decision Support Consortium, a multi-site study which developed, implemented and evaluated clinical decision support services in a diverse range of electronic health records. METHODS: Ethnographic investigation using the rapid assessment process, a procedure for agile qualitative data collection and analysis, including clinical observation, system demonstrations and analysis and 91 interviews. RESULTS: We identified challenges and lessons learned in eight dimensions: (1) hardware and software computing infrastructure, (2) clinical content, (3) human-computer interface, (4) people, (5) workflow and communication, (6) internal organizational policies, procedures, environment and culture, (7) external rules, regulations, and pressures and (8) system measurement and monitoring. Key challenges included performance issues (particularly related to data retrieval), differences in terminologies used across sites, workflow variability and the need for a legal framework. DISCUSSION: Based on the challenges and lessons learned, we identified eight best practices for developers and implementers of service-oriented clinical decision support: (1) optimize performance, or make asynchronous calls, (2) be liberal in what you accept (particularly for terminology), (3) foster clinical transparency, (4) develop a legal framework, (5) support a flexible front-end, (6) dedicate human resources, (7) support peer-to-peer communication, (8) improve standards. CONCLUSION: The Clinical Decision Support Consortium successfully developed a clinical decision support service and implemented it in four different electronic health records and four diverse clinical sites; however, the process was arduous. The lessons identified by the Consortium may be useful for other developers and implementers of clinical decision support services.

A multi-layered framework for disseminating knowledge for computer-based decision support.

BACKGROUND: There are several challenges in encoding guideline knowledge in a form that is portable to different clinical sites, including the heterogeneity of clinical decision support (CDS) tools, of patient data representations, and of workflows. METHODS: We have developed a multi-layered knowledge representation framework for structuring guideline recommendations for implementation in a variety of CDS contexts. In this framework, guideline recommendations are increasingly structured through four layers, successively transforming a narrative text recommendation into input for a CDS system. We have used this framework to implement rules for a CDS service based on three guidelines. We also conducted a preliminary evaluation, where we asked CDS experts at four institutions to rate the implementability of six recommendations from the three guidelines. CONCLUSION: The experience in using the framework and the preliminary evaluation indicate that this approach has promise in creating structured knowledge, to implement in CDS systems, that is usable across organizations.

Fall prevention in acute care hospitals: a randomized trial.

CONTEXT: Falls cause injury and death for persons of all ages, but risk of falls increases markedly with age. Hospitalization further increases risk, yet no evidence exists to support short-stay hospital-based fall prevention strategies to reduce patient falls. OBJECTIVE: To investigate whether a fall prevention tool kit (FPTK) using health information technology (HIT) decreases patient falls in hospitals. DESIGN, SETTING, AND PATIENTS: Cluster randomized study conducted January 1, 2009, through June 30, 2009, comparing patient fall rates in 4 urban US hospitals in units that received usual care (4 units and 5104 patients) or the intervention (4 units and 5160 patients). INTERVENTION: The FPTK integrated existing communication and workflow patterns into the HIT application. Based on a valid fall risk assessment scale completed by a nurse, the FPTK software tailored fall prevention interventions to address patients' specific determinants of fall risk. The FPTK produced bed posters composed of brief text with an accompanying icon, patient education handouts, and plans of care, all communicating patient-specific alerts to key stakeholders. MAIN OUTCOME MEASURES: The primary outcome was patient falls per 1000 patient-days adjusted for site and patient care unit. A secondary outcome was fall-related injuries. RESULTS: During the 6-month intervention period, the number of patients with falls differed between control (n = 87) and intervention (n = 67) units (P=.02). Site-adjusted fall rates were significantly higher in control units (4.18 [95% confidence interval {CI}, 3.45-5.06] per 1000 patient-days) than in intervention units (3.15 [95% CI, 2.54-3.90] per 1000 patient-days; P = .04). The FPTK was found to be particularly effective with patients aged 65 years or older (adjusted rate difference, 2.08 [95% CI, 0.61-3.56] per 1000 patient-days; P = .003). No significant effect was noted in fall-related injuries. CONCLUSION: The use of a fall prevention tool kit in hospital units compared with usual care significantly reduced rate of falls. TRIAL REGISTRATION: clinicaltrials.gov Identifier: NCT00675935.

Self-reported familiarity with acute respiratory infection guidelines and antibiotic prescribing in primary care.

OBJECTIVE: Familiarity with guidelines is generally thought to be associated with guideline implementation, adherence and improved quality of care. We sought to determine if self-reported familiarity with acute respiratory infection (ARI) antibiotic treatment guidelines was associated with reduced or more appropriate antibiotic prescribing for ARIs in primary care. PARTICIPANTS: and MAIN OUTCOME MEASURES: We surveyed primary care clinicians about their familiarity with ARI antibiotic treatment guidelines and linked responses to administrative diagnostic and prescribing data for non-pneumonia ARI visits. RESULTS: Sixty-five percent of clinicians responded to the survey question about guideline familiarity. There were 208 survey respondents who had ARI patient visits during the study period. Respondents reported being 'not at all' (7%), 'somewhat' (30%), 'moderately' (45%) or 'extremely' (18%) familiar with the guidelines. After dichotomizing responses, compared with clinicians who reported being less familiar with the guidelines, clinicians who reported being more familiar with the guidelines had higher rates of antibiotic prescribing for all ARIs combined (46% versus 38%; n = 11 164; P < 0.0001), for antibiotic-appropriate diagnoses (69% versus 59%; n = 3213; P < 0.0001) and for non-antibiotic appropriate diagnoses (38% versus 28%; n = 7951; P < 0.0001). After adjusting for potential confounders, self-reported guideline familiarity was an independent predictor of increased antibiotic prescribing (odds ratio, 1.36; 95% confidence interval, 1.25-1.48). CONCLUSIONS: Self-reported familiarity with an ARI antibiotic treatment guideline was, seemingly paradoxically, associated with increased antibiotic prescribing. Self-reported familiarity with guidelines should not be assumed to be associated with consistent guideline adherence or higher quality of care.

Electronic health record feedback to improve antibiotic prescribing for acute respiratory infections.

OBJECTIVE: To examine whether the Acute Respiratory Infection (ARI) Quality Dashboard, an electronic health record (EHR)-based feedback system, changed antibiotic prescribing. STUDY DESIGN: Cluster randomized, controlled trial. METHODS: We randomly assigned 27 primary care practices to receive the ARI Quality Dashboard or usual care. The primary outcome was the intent-to-intervene antibiotic prescribing rate for ARI visits. We also compared antibiotic prescribing between ARI Quality Dashboard users and nonusers. RESULTS: During the 9-month intervention, there was no difference between intervention and control practices in antibiotic prescribing for all ARI visits (47% vs 47%; P = .87), antibiotic-appropriate ARI visits (65% vs 64%; P = .68), or non­antibiotic-appropriate ARI visits (38% vs 40%; P = .70). Among the 258 intervention clinicians, 72 (28%) used the ARI Quality Dashboard at least once. These clinicians had a lower overall ARI antibiotic prescribing rate (42% vs 50% for nonusers; P = .02). This difference was due to less antibiotic prescribing for non-antibiotic-appropriate ARIs (32% vs 43%; P = .004), including nonstreptococcal pharyngitis (31% vs 41%; P = .01) and nonspecific upper respiratory infections (19% vs 34%; P = .01). CONCLUSIONS: The ARI Quality Dashboard was not associated with an overall change in antibiotic prescribing for ARIs, although when used, it was associated with improved antibiotic prescribing. EHR-based quality reporting, as part of "meaningful use," may not improve care in the absence of other changes to primary care practice.

Effects of documentation-based decision support on chronic disease management.

OBJECTIVE: To evaluate whether a new documentation-based clinical decision support system (CDSS) is effective in addressing deficiencies in the care of patients with coronary artery disease (CAD) and diabetes mellitus (DM). STUDY DESIGN: Controlled trial randomized by physician. METHODS: We assigned primary care physicians (PCPs) in 10 ambulatory practices to usual care or the CAD/DM Smart Form for 9 months. The primary outcome was the proportion of deficiencies in care that were addressed within 30 days after a patient visit. RESULTS: The Smart Form was used for 5.6% of eligible patients. In the intention-to-treat analysis, patients of intervention PCPs had a greater proportion of deficiencies addressed within 30 days of a visit compared with controls (11.4% vs 10.1%, adjusted and clustered odds ratio =1.14; 95% confidence interval, 1.02-1.28; P = .02). Differences were more pronounced in the "on-treatment" analysis: 17.0% of deficiencies were addressed after visits in which the Smart Form was used compared with 10.6% of deficiencies after visits in which it was not used (P <.001). Measures that improved included documentation of smoking status and prescription of antiplatelet agents when appropriate. CONCLUSIONS: Overall use of the CAD/DM Smart Form was low, and improvements in management were modest. When used, documentation-based decision support shows promise, and future studies should focus on refining such tools, integrating them into current electronic health record platforms, and promoting their use, perhaps through organizational changes to primary care practices.

Documentation-based clinical decision support to improve antibiotic prescribing for acute respiratory infections in primary care: a cluster randomised controlled trial.

BACKGROUND AND OBJECTIVE: Clinical guidelines discourage antibiotic prescribing for many acute respiratory infections (ARIs), especially for non-antibiotic appropriate diagnoses. Electronic health record (EHR)-based clinical decision support has the potential to improve antibiotic prescribing for ARIs. METHODS: We randomly assigned 27 primary care clinics to receive an EHR-integrated, documentation-based clinical decision support system for the care of patients with ARIs - the ARI Smart Form - or to offer usual care. The primary outcome was the antibiotic prescribing rate for ARIs in an intent-to-intervene analysis based on administrative diagnoses. RESULTS: During the intervention period, patients made 21 961 ARI visits to study clinics. Intervention clinicians used the ARI Smart Form in 6% of 11 954 ARI visits. The antibiotic prescribing rate in the intervention clinics was 39% versus 43% in the control clinics (odds ratio (OR), 0.8; 95% confidence interval (CI), 0.6-1.2, adjusted for clustering by clinic). For antibiotic appropriate ARI diagnoses, the antibiotic prescribing rate was 54% in the intervention clinics and 59% in the control clinics (OR, 0.8; 95% CI, 0.5-1.3). For non-antibiotic appropriate diagnoses, the antibiotic prescribing rate was 32% in the intervention clinics and 34% in the control clinics (OR, 0.9; 95% CI, 0.6-1.4). When the ARI Smart Form was used, based on diagnoses entered on the form, the antibiotic prescribing rate was 49% overall, 88% for antibiotic appropriate diagnoses and 27% for non-antibiotic appropriate diagnoses. In an as-used analysis, the ARI Smart Form was associated with a lower antibiotic prescribing rate for acute bronchitis (OR, 0.5; 95% CI, 0.3-0.8). CONCLUSIONS: The ARI Smart Form neither reduced overall antibiotic prescribing nor significantly improved the appropriateness of antibiotic prescribing for ARIs, but it was not widely used. When used, the ARI Smart Form may improve diagnostic accuracy compared to administrative diagnoses and may reduce antibiotic prescribing for certain diagnoses.

Improving management of chronic diseases with documentation-based clinical decision support: results of a pilot study.

Clinical Decision Support Systems (CDSS) have the potential to improve patient care. We developed the Coronary Artery Disease and Diabetes Mellitus (CAD/DM) Smart Form as a documentation-based application that provides decision support for the management of chronic diseases. Results of a pilot study suggest that the CAD/DM Smart Form has the potential to improve patient care.

An effort to improve electronic health record medication list accuracy between visits: patients' and physicians' response.

OBJECTIVE: To evaluate the efficacy of a secure web-based patient portal called Patient Gateway (PG) in producing more accurate medication lists in the electronic health record (EHR), and whether sending primary care physicians (PCPs) a clinical message updating them on the information their patients provided caused physicians to update the EHR medication list. METHODS: We compared the medication list accuracy of 84 patients using PG with that of 79 who were not. Patient-reported medication discrepancies were noted in the EHR in a clinical note by research staff and a message was sent to the participants' PCPs notifying them of the updated information. RESULTS: Participants were taking 665 medications according to the EHR, and reported 273 additional medications. A lower percentage of PG users' drug regimens (54% versus 61%, p=0.07) were reported to be correct than those of PG non-users, although PG users took significantly more medications than their non-user counterparts (5.0 versus 3.1 medications, p=0.0001). Providing patient-reported information in a clinical note and sending a clinical message to the primary care doctor did not result in PCPs updating their patients' EHR medication lists. CONCLUSIONS: Medication lists in EHRs were frequently inaccurate and most frequently overlooked over-the-counter (OTC) and non-prescription drugs. Patients using a secure portal had just as many discrepancies between medication lists and self-report as those who did not, and notifying physicians of discrepancies via e-mail had no effect.

Do physicians take action on high risk family history information provided by patients outside of a clinic visit?

Clinically relevant family history information is frequently missing or not readily available in electronic health records. Improving the availability of family history information is important for optimum care of many patients. Family history information on five conditions was collected in a survey from 163 primary care patients. Overall, 53% of patients had no family history information in the electronic health record (EHR) either on the patient's problem list or within a templated family history note. New information provided by patients resulted in an increase in the patient's risk level for 32% of patients with a positive family history of breast cancer, 40% for coronary artery disease, 50% for colon cancer, 74% for diabetes, and 95% each for osteoporosis and glaucoma. Informing physicians of new family history information outside of a clinic visit through an electronic clinical message and note in the EHR was not sufficient to achieve recommended follow-up care. Better tools need to be developed to facilitate the collection of family history information and to support clinical decision-making and action.

Clinical decision support to improve antibiotic prescribing for acute respiratory infections: results of a pilot study.

Acute Respiratory Infections (ARIs) are the number one reason for antibiotic prescribing in the United States, and much antibiotic prescribing for ARIs is inappropriate. We designed an electronic health record-integrated, documentation-based clinical decision support system for the care of patients with ARIs, the ARI Smart Form. To evaluate the ARI Smart Form and assess the feasibility of performing a larger trial, we conducted a pilot study with 10 clinicians who used the ARI Smart Form with 26 patients. Clinicians prescribed antibiotics to 6 of 6 patients with antibiotic-appropriate diagnoses and to 3 of 20 (15%) patients with antibiotic-inappropriate diagnoses. The average duration of use of the ARI Smart Form was 7.5 (SD+/-4.5) minutes. Eight of 10 respondents reported that the ARI Smart Form was either time-neutral or timesaving. The ARI Smart Form requires further evaluation but has the potential to improve workflow and reduce inappropriate antibiotic prescribing.

Improving electronic health record (EHR) accuracy and increasing compliance with health maintenance clinical guidelines through patient access and input.

BACKGROUND: Health maintenance is crucial for preventing morbidity and premature mortality, but many patients do not receive preventive services at recommended intervals. One reason for this is the lack of up-to-date information accurately reflecting patients' history. Electronic health records (EHRs) can be useful, but are often incomplete. Patient input has the potential to improve the accuracy of this information. In this study, we assessed the current state of EHR completeness for preventive services and the added value of patient reported information. METHODS: Participants were sent a survey, pre-populated with health maintenance procedure information from their EHRs. They were asked to review this information and indicate whether it was accurate or if they had a procedure done more recently. Of 1098 patients recruited from a primary care practice, 163 returned the survey. When a patient reported a more recent test than was noted in the EHR, researchers updated the EHR to reflect the additional information. Data were also gathered from the EHR 6 months after surveys were completed to analyze whether providing due test information encouraged patients to get tested and vaccinated. A review of medical records was performed on a control group to analyze differences in adherence to preventive guidelines between those that were notified of their overdue status and those who were not notified. RESULTS: The EHR was frequently incomplete when compared to patient report. In particular, many patients were misidentified as being overdue for health maintenance procedures when they had obtained them in other places. Showing patients their information resulted in little impact on overall adherence. However, with the cumulative effects of additional patient-reported procedures and procedures performed after the survey, intervention patients had higher documented adherence rates for every procedure than the control group. CONCLUSIONS: Health maintenance data in EHRs were often incomplete. Patients were often able to provide useful information, demonstrating the value of patient contributions in keeping records up-to-date.

Barriers to electronic health record use during patient visits.

The effectiveness of electronic health record (EHR)-based clinical decision support is limited when clinicians do not interact with the EHR during patient visits. To assess EHR use during ambulatory visits and determine barriers to such use, we performed a cross-sectional survey of 501 primary care clinicians. Of 225 respondents, 53 (24%) never or only sometimes used any EHR functionality during patient visits. Non-physician clinicians (e.g., nurse practitioners) were marginally more likely to be EHR non-users than physicians (39% versus 21%, respectively; p = .05). The most commonly reported barriers to using the EHR during patient visits were loss of eye contact with patients (62%), falling behind schedule (52%), computers being too slow (49%), inability to type quickly enough (32%), feeling that using the computer in front of the patient is rude (31%), and preferring to write long prose notes (28%). EHR developers and healthcare system leaders must address social, workflow, technical, and professional barriers if clinicians are to use EHRs in the presence of patients and realize the full potential of ambulatory clinical decision support.

How accurate is information that patients contribute to their Electronic Health Record?

Increased patient interaction with medical records and the advent of personal health records (PHRs) may increase patients' ability to contribute valid information to their Electronic Medical Record (EHR) medical record. Patient input through a secure connection, whether it be a patient portal or PHR, will integrate many aspects of a patient's health and may help lessen the information gap between patients and providers. Patient reported data should be considered a viable method of enhancing documentation but will not likely be as complete and accurate as more comprehensive data-exchange between providers.