NICU Medication Errors: Identifying a Risk Profile for Medication Errors in the Neonatal Intensive Care Unit

TA Stavroudis; AD Shore; L Morlock; RW Hicks; D Bundy; MR Miller


J Perinatol. 2010;30(7):459-468. 

In This Article


Data Source: United States Pharmacopeia MEDMARX Reporting System

The United States Pharmacopeia is an independent, internationally recognized, science-based public health organization and is the official public standards-setting authority for medicines manufactured and sold in the United States.[21] In 1998, United States Pharmacopeia introduced MEDMARX as an anonymous, confidential, de-identified, Internet-accessible medication error-reporting program. Data are collected locally and stored in a central data repository. As of January 2005, MEDMARX included 616 hospitals from all 50 states.

Data and Variables

MEDMARX uses the definition of a medication error established by the National Coordinating Council for Medication Error Reporting and Prevention (NCC MERP), an independent body composed of 22 national organizations.[22] A medication error is defined as 'any preventable event that may cause or lead to inappropriate medication use or patient harm while the medication is in the control of the health care professional, patient, or consumer. Such events may be related to professional practice, health care products, procedures, and systems, including prescribing; order communication; product labeling, packaging, and nomenclature; compounding; dispensing; distribution; administration; education; monitoring; and use.'[22]

MEDMARX also uses the NCC MERP taxonomy for data collection and the Index for categorizing medication errors.[22] Good inter-rater agreement (κ=0.61) has been reported for MEDMARX users.[23] Each error report is standardized and provides information regarding the error's timing, location, error category, node, cause, medication involved, and error type. Facility characteristics captured include bed size, number of doses delivered, ownership, type of facility, and degree of pharmacist availability. Definitions for error categories[22] are listed in Table 1 . Variables used in MEDMARX are listed in Supplementary Appendix A and Table 2 . Some data fields within the error reports are single-pick lists (error category and error node), and some are multi-select pick lists, allowing users to choose and enter as many items as they believe apply (error cause, error type, and medication involved).

A key driver for the types of information that are reported in MEDMARX is the error category variable. With increasing levels of harm, increasingly more data fields are required to be completed. For example, error categories A and B, the two lowest categories of error, do not require reporting of patient age, whereas patient age is required for all categories of error that reach the patient (C and above). Patient gender is always an optional field. For these reasons, there are more error reports coded as error Category A and B with unknown age and gender. Additionally, between 1998 and 2004, MEDMARX underwent a number of changes in required data fields as the concept of error reporting matured.


We conducted a retrospective cross-sectional study on NICU medication error reports submitted to MEDMARX between 1 January 1999, and 31 December 2005. For error type, response choices were grouped into categories involving similar concepts, as defined by our research team. For example, under error type response choices of deteriorated product and expired product were combined into one category. For error cause, the pick-list choices were grouped into a modified version of the NCC MERP error cause categories developed by the research team (Table 2).[22] Medications were grouped by the American Hospital Formulary Service (AHFS) Drug Information 2005 classification system,[24] and the 2005 ISMP High-Alert Medications list[20] was used for analysis. Those medications that were not listed in AHFS or listed under multiple categories were modified based on the agent's pharmacological use in the NICU, as identified by a round table of neonatology staff. Duplicate medications listed in the same report were unduplicated.

Data analyses were conducted using Statistical Analysis Software, version 9.1.[25] The significance of associations across categorical variables was tested with the Rao–Scott modified χ2 test to account for the clustering of medication error reports by facility.[26] The study was granted exemption from the Johns Hopkins University Institutional Review Board.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.