Implementation of the intervention was associated with a significant reduction in average (per patient) cumulative benzodiazepine dose, from 250 mg to 96 mg (P < .001), as shown in Table 2. Average length of stay decreased from 8.0 days to 5.1 days (P < .001). Secondary safety measures were notable for a reduction in DT incidence, from 47.5% to 22.5% (P < .001), and lower rates of intubation, from 13.8% to 1.3% (P = .003). Seven-day readmission rates were 0% preintervention and 1.4% postintervention.
Discussion
We found that hospitalized patients with severe alcohol withdrawal treated with STT required fewer benzodiazepines and had a lower length of stay than patients treated with a conventional combined STT and FS regimen. Implementation of the change from the STT and FS approach to the STT approach in the SDU resulted in concerns that waiting for symptoms to appear could result in more severe withdrawal and prolonged treatment.3 To address this, the intervention included monitoring and dosing every 30 minutes, as compared to monitoring and dosing every 1 hour preintervention. In addition, a sliding-scale approach to match alcohol withdrawal score with dosage was employed in postintervention patients.
Employment of the STT protocol also resulted in decreased complications, including lower rates of DT and transfer to the ICU. This new intervention resulted in significantly decreased time required to control severe symptoms. In the preintervention phase, if a patient’s symptoms escalated despite administration of the as-needed dose of benzodiazepine, there was often a delay in administration of additional doses due to the time needed for nurses to reach a physician and subsequent placement of a new order. In the postintervention phase, the STT protocol allowed nursing staff to give benzodiazepines without delay when needed. We believe this reduced the number of calls by nursing staff to physicians requesting additional medications, and that this improved teamwork when managing these patients.
As part of the intervention, a decision was made to use the MINDS scale rather than the CIWA-Ar scale to assess withdrawal severity. This was because the CIWA-Ar has only been validated in patients with uncomplicated alcohol withdrawal syndrome and has not been researched extensively in patients requiring ICU-level care.1 MINDS assessment has proven to be reliable and reflects severity of withdrawal. Furthermore, MINDS requires less time to administer—3 to 5 minutes vs 5 to 15 minutes for the CIWA-Ar scale. CIWA-Ar, unlike MINDS, requires subjective input from the patient, which is less reliable for higher acuity patients. Our study is unique in that it focused on high-acuity patients and it showed both a significant reduction in quantity of benzodiazepines prescribed and length of stay. Previous studies on lower acuity patients in detoxification units have confirmed that STT is more effective than a FS approach.3-5 In patients of higher acuity, STT has not proven to be superior.
A key lesson learned was the need for proper education of nursing staff. Concurrent nursing audits were necessary to ensure that scoring was performed in an accurate and timely manner. In addition, it was challenging to predict which patients might develop DTs versus those requiring a brief inpatient stay. While there was initial concern that an STT protocol could result in underdosing, we found that patients had fewer DT episodes and fewer ICU transfers.