Printer Alarm for Notification of Time-Sensitive Results

The Joint Commission recognized improved reporting of critical values as a National Patient Safety Goal. Institution-wide definition and reporting of such values may not fully serve the needs of patients in all units, however. We propose to develop and evaluate a mechanism for unit-specific handling of time-sensitive laboratory values in the emergency department (ED). The objectives of this project were the following: 1) to create a unit-specific method for defining and reporting time-sensitive laboratory values, 2) to determine the incidence of such results among ED patients, and 3) to measure the time-to-reaction by providers for such results. We hypothesize that our intervention will decrease the delay until clinically indicated responses are ordered. We conducted a pre-post intervention study in the ED of an urban level 1 trauma center with 60,000 annual visits. We created a rule within our ED information system that flags when a bicarbonate ≤ 18 mEq/L, a lactate ≥ 3 mmol/L, or an international normalized ratio (INR) ≥ 1.4 is reported. This routes a notification to a dedicated printer, where a pressure sensor sounds an alarm when paper lies in the output tray. Providers were asked to fill out a brief survey for each notification. We measured the effect of notifications for elevated lactate on time-to-order-entry for intravenous (IV) fluids or antibiotics, when indicated. We analyzed all ED visits from October 1, 2011 to July 31, 2012, and the printer alarm was deployed on January 11, 2012. We described the data by counts, percentages, medians, and interquartile ranges. We tested for differences between groups using the chi-square test and the Wilcoxon rank-sum test. Elevated lactate values occurred during 199 visits before the intervention and 423 visits after. Providers reported already being aware of the result before printer alarm notification 17% of the time. Providers ordered some clinical response in 174 cases (87.4%) before the intervention and 361 cases (85.3%) after; these rates did not differ significantly (p = 0.563). Before the intervention, IV fluids were ordered after 59.3% of elevated lactate results, antibiotics after 7.5%, and both after 20.6%. After the intervention, IV fluids were ordered for 70.7% of notifications, antibiotics for 12.1%, and both for 2.6%. The rates for each type of response differed before and after the intervention (p < 0.001). The median (IQR) of time-to-order-entry in minutes was 58 (23, 102) for the control group and 48 (23, 77) for the printer alarm group (p = 0.029). We implemented a novel mechanism for unit-specific reporting of time-sensitive laboratory values using widely available technology. This co-existed with, and did not replace, institution-wide protocols for critical value handling. Notifications for elevated lactate results were generated frequently and decreased the time-to-order-entry for IV fluids or antibiotics. The printer alarm intervention did not change the frequency of clinical responses, although it did affect the type of response; the significance of this remains unclear. Further work will evaluate the effect of the printer alarm on clinical responses to abnormal bicarbonate and INR results.