Customized Reference Ranges for Laboratory Values Decrease False Positive Alerts in Intensive Care Unit Patients

Background Traditional electronic medical record (EMR) interfaces mark laboratory tests as abnormal based on standard reference ranges derived from healthy, middle-aged adults. This yields many false positive alerts with subsequent alert-fatigue when applied to complex populations like hospitalized, critically ill patients. Novel EMR interfaces using adjusted reference ranges customized for specific patient populations may ameliorate this problem. Objective To compare accuracy of abnormal laboratory value indicators in a novel vs traditional EMR interface. Methods Laboratory data from intensive care unit (ICU) patients consecutively admitted during a two-day period were recorded. For each patient, available laboratory results and the problem list were sent to two mutually blinded critical care experts, who marked the values about which they would like to be alerted. All disagreements were resolved by an independent super-reviewer. Based on this gold standard, we calculated and compared the sensitivity, specificity, positive and negative predictive values (PPV, NPV) of customized vs traditional abnormal value indicators. Results Thirty seven patients with a total of 1341 laboratory results were included. Experts’ agreement was fair (kappa = 0.39). Compared to the traditional EMR, custom abnormal laboratory value indicators had similar sensitivity (77% vs 85%, P = 0.22) and NPV (97.1% vs 98.6%, P = 0.06) but higher specificity (79% vs 61%, P<0.001) and PPV (28% vs 11%, P<0.001). Conclusions Reference ranges for laboratory values customized for an ICU population decrease false positive alerts. Disagreement among clinicians about which laboratory values should be indicated as abnormal limits the development of customized reference ranges.

[1]  M. Chambrin,et al.  Multicentric study of monitoring alarms in the adult intensive care unit (ICU): a descriptive analysis , 1999, Intensive Care Medicine.

[2]  T M O'Carroll,et al.  Survey of alarms in an intensive therapy unit * , 1986, Anaesthesia.

[3]  Brian W. Pickering,et al.  Data Utilization for Medical Decision Making at the Time of Patient Admission to ICU* , 2013, Critical care medicine.

[4]  V. Herasevich,et al.  Novel Representation of Clinical Information in the ICU , 2010, Thrombosis and Haemostasis.

[5]  M. Brigden,et al.  Problems in Interpreting Laboratory Tests: What do Unexpected Results Mean? , 2000, Postgraduate medicine.

[6]  P. Carpenter,et al.  The electronic medical record: perspective from Mayo Clinic. , 1994, International journal of bio-medical computing.

[7]  K. Shojania,et al.  National Patient Safety Goals. , 2006, Bulletin of the American College of Surgeons.

[8]  L. Lind,et al.  Reference Values for 27 Clinical Chemistry Tests in 70-Year-Old Males and Females , 2009, Gerontology.

[9]  V. Herasevich,et al.  Informatics infrastructure for syndrome surveillance, decision support, reporting, and modeling of critical illness. , 2010, Mayo Clinic proceedings.

[10]  S. Lawless.  Crying wolf: False alarms in a pediatric intensive care unit , 1994, Critical care medicine.

[11]  E. Ivers,et al.  Early Goal-Directed Therapy in the Treatment of Severe Sepsis and Septic Shock , 2001 .

[12]  V. Herasevich,et al.  The effect of two different electronic health record user interfaces on intensive care provider task load, errors of cognition, and performance* , 2011, Critical care medicine.

[13]  Mark I. Hwang,et al.  Information dimension, information overload and decision quality , 1999, J. Inf. Sci..