Use of a real-time, algorithm-driven, publicly displayed, automated signal to improve insulin prescribing practices.

AIM: The clinical andon board (CAB) is a novel electronic surveillance and communication system, which alerts providers to and prompts treatment of dysglycemia. This investigation was designed to determine the CAB's effectiveness in supporting adherence to standardized evidence-based protocols, as well as improving glycemic control. METHODS: This study was a retrospective pre/post analysis of insulin orders and blood glucose values. We used a Student's t-test for continuous variables and Chi2 for all other variables. This study included patients 18 years or older admitted to the hospital medical service as an inpatient with a length of stay greater than 24 h and less than 90 days. We used Pearson's correlation coefficient to evaluate the relationship between CAB and blood glucose. RESULTS: The rate of compliance in prescribing basal insulin for patient with diabetes increased from 56% to 77% (p < 0.001). Similarly, compliance rates for prescribing correctional insulin in patients without diabetes increased from 15% to 37% (p < 0.001). Performance on the CAB was linearly related to blood glucose (p = 0.004), and there was a small statistically (not clinically) significant improvement in mean blood glucose values. CONCLUSION: This approach is effective in alerting and engaging providers to prescribe insulin in a standardized manner with potential to improve glycemic control.

Effect of a Real-Time Electronic Dashboard on a Rapid Response System.

A rapid response system (RRS) may have limited effectiveness when inpatient providers fail to recognize signs of early patient decompensation. We evaluated the impact of an electronic medical record (EMR)-based alerting dashboard on outcomes associated with RRS activation. We used a repeated treatment study in which the dashboard display was successively turned on and off each week for ten 2-week cycles over a 20-week period on the inpatient acute care wards of an academic medical center. The Rapid Response Team (RRT) dashboard displayed all hospital patients in a single view ranked by severity score, updated in real time. The dashboard could be seen within the EMR by any provider, including RRT members. The primary outcomes were the incidence rate ratio (IRR) of all RRT activations, unexpected ICU transfers, cardiopulmonary arrests and deaths on general medical-surgical wards (wards). We conducted an exploratory analysis of first RRT activations. There were 6736 eligible admissions during the 20-week study period. There was no change in overall RRT activations (IRR = 1.14, p = 0.07), but a significant increase in first RRT activations (IRR = 1.20, p = 0.04). There were no significant differences in unexpected ICU transfers (IRR = 1.15, p = 0.25), cardiopulmonary arrests on general wards (IRR = 1.46, p = 0.43), or deaths on general wards (IRR = 0.96, p = 0.89). The introduction of the RRT dashboard was associated with increased initial RRT activations but not overall activations, unexpected ICU transfers, cardiopulmonary arrests, or death. The RRT dashboard is a novel tool to help providers recognize patient decompensation and may improve initial RRT notification.

Effectiveness of a multi-component quality improvement intervention on rates of hyperglycaemia.

PURPOSE: To evaluate the effectiveness of a multifaceted, hospital-wide glycaemic control quality improvement programme. METHODS: The quality improvement intervention comprised three components, derived through root cause analysis: standardising and simplifying care (including evidence-based order sets), increasing visibility (through provider access to clinical data and direct feedback) and educational outreach (directed at the entire institution). Effectiveness was determined at a single urban acute care hospital through time-series analysis with statistical process control charts. Primary outcomes included rate of hyperglycaemia and rate of hypoglycaemia. RESULTS: The study included 70 992 hospital admissions for 50 404 patients, with 3 35 645 patient days. The hyperglycaemia ratio decreased 25.2% from 14.1% to 10.5% (95% CI 3.3 to 3.9 percentage points, p<0.001). The ratio of patient days with highly elevated blood glucose (>299 mg/dL) decreased 31.8% from 4.8% to 3.3% (95% CI 1.4 to 1.7 percentage points, p<0.001). Hypoglycaemia ratio decreased from 5.2% to 4.6% (95% CI 0.27 to 0.89 percentage points, p<0.001) in patients with diabetes, but increased in patients without diabetes from 1.2% to 1.7% (95% CI 0.46 to 0.70 percentage points, p<0.001). CONCLUSIONS: We demonstrate improved hospital-wide glycaemic control after a multifaceted quality improvement intervention in the context of strong institutional commitment, national mentorship and Lean management.

Inbox Messaging: an effective tool for minimizing non-urgent paging related interruptions in hospital medicine provider workflow.

Communication is one of the foundations on which safe, high quality care is built.1, 3, 6, 17, 20 The nature of hospital medicine requires that nurses and providers be efficient and effective in communicating with multiple disciplines.17 This need for timely communication must continually be balanced with the need to minimize interruptions in workflow.1,2 3,4,6,7,9,13,15,17,18 Interruptions not only lead to distraction, they also add inefficiency to the care process and have been shown to contribute to an increased risk of medical error.2,3,4,7,17,18 A major source of interruptions are pagers that emit an audible tone with each message received.3,9,10,17,18 This interruptive nature makes pagers a less-than-ideal tool for communicating non-urgent (address within one hour) messages received.3,9,10,17,18 In addition to increasing interruptions, pagers do not facilitate closed loop communication, another feature that has been shown to improve safety.14,17,25 Inbox Messaging is intended to provide a less disruptive closed-loop method of communication for non-urgent messages. Inbox Messaging is an interface within the electronic health record (EHR) that functions similarly to e-mail. A multi-disciplinary communication workgroup identified this interface as having potential to not only decrease interruptions, but to also facilitate closed-loop communication. Inbox is currently utilized between the hours of 0700 and 1800 for non-urgent nurse-provider communication about patients on the hospital medicine service. The number of RN non-urgent pages per day was 103 (SD=19, n=97) prior to the Inbox intervention, with a significant decrease (p<.001) during follow-up to 38 (SD=14, n=354) pages per day. At the same time, the number of messages per day increased from 0 to 80 (SD=20, n=354) messages during follow-up. As desired, the mean number of RN urgent pages was unchanged from 13 per day to 13 per day (p=.52). Cerner Inbox Messaging decreases the frequency of non-urgent pager-related interruptions in workflow.

Making background work visible: opportunities to address patient information needs in the hospital.

Despite growing use of patient-facing technologies such as patient portals to address information needs for outpatients, we understand little about how patients manage information and use information technologies in an inpatient context. Based on hospital observations and responses to an online questionnaire from previously hospitalized patients and caregivers, we describe information workspace that patients have available to them in the hospital and the information items that patients and caregivers rate as important and difficult to access or manage while hospitalized. We found that patients and caregivers desired information-such as the plan of care and the schedule of activities-that is difficult to access as needed in a hospital setting. Within this study, we describe the various tools and approaches that patients and caregivers use to help monitor their care as well as illuminate gaps in information needs not typically captured by traditional patient portals.

Building an automated problem list based on natural language processing: lessons learned in the early phase of development.

Detailed problem lists that comply with JCAHO requirements are important components of electronic health records. Besides improving continuity of care electronic problem lists could serve as foundation infrastructure for clinical trial recruitment, research, biosurveillance and billing informatics modules. However, physicians rarely maintain problem lists. Our team is building a system using MetaMap and UMLS to automatically populate the problem list. We report our early results evaluating the application. Three physicians generated gold standard problem lists for 100 cardiology ambulatory progress notes. Our application had 88% sensitivity and 66% precision using a non-modified UMLS dataset. The systemâs misses concentrated in the group of ambiguous problem list entries (Chi-square=27.12 p<0.0001). In addition to the explicit entries, the notes included 10% implicit entry candidates. MetaMap and UMLS are readily applicable to automate the problem list. Ambiguity in medical documents has consequences for performance evaluation of automated systems.