Simple quality improvement interventions reduce unnecessary intravascular device dwell time.

To the Editor—Intravascular devices (IVD) are a vital part of medical care. IVD-associated infections are an important cause of iatrogenic morbidity in inpatients. IVD-related bloodstream infections (BSI) prolong hospital stay and increase costs. IVD-related phlebitis is also a significant problem. The risk of IVD-associated complications correlates with IVD dwell time. The Centers for Disease Control recommends prompt removal of nonessential IVDs to reduce the rate of IVD-related BSI (category IA recommendation). Parenti et al" showed that quality improvement programs could reduce the unnecessary use of IVDs. We identified the leaving of IVDs in situ unnecessarily as a significant quality issue at Auckland City Hospital. We therefore developed lowcost interventions intended to reduce unnecessary IVD dwell time, and we assessed their effectiveness. The interventions were implemented on the 4 internal medicine wards at Auckland City Hospital, a 700-bed tertiary hospital. The first intervention was a sticker placed in every patient's clinical notes each morning. This required the medical team to indicate whether IVDs were required or should be removed, and this also required the nursing staff to contact the medical team if no indication was made. The second intervention was the daily distribution of an educational pamphlet (designed to be printed on the daily menu sheet) to every patient. This pamphlet showed a photograph of a peripheral IVD and explained the usefulness of these devices and their potential to cause infection. It requested that patients with an IVD in situ ask their doctors and nurses whether it was still required. Baseline data were gathered for 14 consecutive days beginning 7 weeks prior to the implementation of the interventions. The interventions were implemented on 14 consecutive days, during which the same types of data were collected. Each patient was assessed daily, and the number and type of IVDs in situ were recorded. Each patient assessed was counted as a patient-day. If an IVD was present in the patient, this was counted as an IVD-day. If a patient had more than 1 IVD, each device was counted as 1 IVD-day. Each IVD-day was defined at the time of review as "necessary" or "unnecessary" according to strict prespecified criteria. An IVD was deemed necessary if the patient was receiving appropriate intravenous antibiotic therapy; was receiving other intravenous medications or hydration; had an unstable condition, such as seizures or gastrointestinal bleeding, or was undergoing cardiac monitoring; or had a procedure requiring vascular access planned within the following 24 hours. If a patient had more than 1 IVD in situ, each IVD-day required a separate indication to be defined as "necessary." Because these interventions were being assessed as a quality improvement exercise, approval by the institutional review board was not considered to be required. The project was approved by the head of the Department of Internal Medicine and by the charge nurses of the wards involved. The results during the baseline and intervention periods are shown in the Table. A statistically significant reduction in the number of both total IVD-days and unnecessary IVDdays occurred during the intervention period. The percentage of patient—days on which an unnecessary IVD was in use during the intervention period was reduced by 7.8% (from 20.4% to 12.6%; P< .001). Therefore, for every 13 patientdays of intervention, 1 unnecessary IVD-day was avoided. We have shown that the introduction of 2 low-cost interventions can significantly reduce the number of unnecessary IVD-days. This would be expected to result in a reduction in the incidence of IVD-related complications, including BSI. Infection control measures such as these are also increasingly important because of the emergence of antimicrobial resistance among nosocomial pathogens. A recent meta-analysis showed the risk of IVD-related BSI associated with use of peripheral short lines (which accounted for more than 95% of the IVDs in our internal medicine wards) was 0.5 cases per 1,000 IVD-days. Thus, 1 IVD-related BSI would be prevented per 26,000 patient-days, with our interventions. The estimated cost of the interventions was US$0.10 per patient-day, which is equivalent to $2,600 per IVD-related BSI prevented. This compares favorably with the