ADC Survey Shows Some Improvements, But Unnecessary Risks Still Exist

Automated dispensing cabinets (ADCs)—perhaps more accurately referred to as automated distribution cabinets*—were introduced in the 1980s to help with medication distribution, storage, security, and retrieval documentation. Adoption of this technology started slowly, with only about half of hospitals using ADCs in 1999. However, by 2005, close to three-quarters of acute care facilities were utilizing ADCs.¹ According to more than 800 respondents to our November 2007 survey in the ISMP Medication Safety Alert! and Nurse Advise-ERR newsletters, 94% of readers told us they are utilizing ADCs in their facilities; of those, more than half (56%) are using the technology as the primary means of drug distribution.  

ADCs can decrease the amount of time before a medication is available on patient care units for administration, ensure greater security of medications—particularly narcotics—and capture drug charges more efficiently and accurately. ADCs can also reduce the risk of medication errors, but only when specific safeguards are consistently available and used. When compared to data from our 1999 survey on ADCs, it appears that more organizations are currently employing these important safeguards. Yet, as the highlighted findings below demonstrate, the improvements are incremental and not as widespread as needed to maximize the safety benefits that ADC technology offers. 

Checking processes. The requirement for a pharmacist to check ADC stock medications before they leave the pharmacy increased from 65% in 1999 to 75% in 2007. No improvement was seen between 1999 and 2007 regarding verification processes after restocking the ADCs; in both years, just 18% of respondents reported that another person verifies drug placement in the ADC. Requiring another practitioner to double-check a drug removed from an ADC via override, before pharmacy review, only increased from 10% in 1999 to 29% in 2007. These manual checking processes are important to prevent stocking and/or drug retrieval errors, similar to the 2006 events that occurred in an Indianapolis hospital when the wrong concentration of heparin vials was stocked in an ADC, leading to fatal 1,000-fold overdoses in three neonates. While these checking functions could be performed with bar-coding, just a quarter (25%) of hospitals use this technology to verify the accuracy of drugs that have been selected to restock the ADCs or to verify accuracy when placing them in the cabinets (although adoption of bar-code technology is up from 3% in 1999 to 25% in 2007). 

Pharmacist review and overrides. Profiled ADC systems are one of the most important safety enhancements that has evolved in ADCs during the last decade. This safety feature provides a direct interface between the pharmacy information system and ADCs so pharmacists can profile, screen, and approve medications before they are removed from the cabinet for administration. Pharmacy profiling supports The Joint Commission medication management standard for pharmacist review of new medication orders before administration. In 1999, only 28% of respondents reported that a pharmacist must verify orders before drugs can be removed from ADCs; but in 2007, 64% of respondents reported adopting this practice. Interestingly, fewer (56%) frontline nurses reported that pharmacy verification always or frequently occurs before removing medications from ADCs, compared to 72% of other healthcare professionals, particularly pharmacists. Further, just 59% of 2007 respondents reported that that all ADCs in their facilities are capable of profiling.

Cabinet design. An increase in 20 percentage points occurred between 1999 and 2007 regarding the user’s ability to remove only the requested drug from the ADC. Nevertheless, just 50% of respondents noted that individual compartments for each drug are always or frequently available in the ADC cabinets.

ADC stock. In both 1999 and 2007, 35% of respondents reported that they always or frequently encounter multiple concentrations of medications in ADCs. In 2007, respondents, particularly nurses, also reported that they encounter fewer ready-to-administer medications in ADCs than reported in 1999. Almost a quarter (23%) of 2007 respondents reported that non-medications are now being stored in ADCs, an increase from 15% in 1999.

Workflow and practice habits. In 2007, additional assessment items were added to the ADC survey related to workflow and practice habits. Almost a third (30%) of frontline nurses reported that they always or frequently wait in line to access the ADC, and almost half (48%) reported that the ADCs are not located in areas free from distractions. Only two-thirds (69%) of frontline nurses reported that they always or frequently remove just one patient’s medications at a time, implying that multiple patients’ medications are removed a third of the time—a practice that is known to lead to drug administration errors. When compared with responses from other healthcare professionals who completed the survey, frontline nurses reported less satisfaction with ADC workflow and reported using more workarounds to compensate for problems.

Despite the growing popularity of ADCs and the benefits this technology offers, few resources exist to guide healthcare organizations toward best practices and safest use of this technology. To address this deficit, ISMP convened a group of stakeholders in the spring of 2007 to develop ADC practice guidelines. These guidelines (draft) are currently posted on the ISMP website. The guidelines contain 12 core processes associated with safe ADC use (see Table 1 below). Please be sure to review these core processes and make plans to employ as many as possible in 2008 to reduce the risk of serious errors associated with ADC use.   

_{*Professional licensing boards have suggested calling ADCs automated distribution cabinets since pharmacists, not nurses, dispense medications.}  

Reference:

1. Pedersen CA, Schneider PJ, Scheckelhoff DJ. ASHP national survey of pharmacy practice in hospital settings: dispensing and administration. Am J Health Syst Pharm. 2006; 63:327-45.