Presented at the Society of Healthcare Epidemiology of America 19th Annual Scientific Meeting, San Diego, CA, March 2009.
Effect of Establishing Guidelines on Appropriate Urinary Catheter Placement
Article first published online: 1 MAR 2010
© 2010 by the Society for Academic Emergency Medicine
Academic Emergency Medicine
Volume 17, Issue 3, pages 337–340, March 2010
How to Cite
Fakih, M. G., Pena, M. E., Shemes, S., Rey, J., Berriel-Cass, D., Szpunar, S. M., Savoy-Moore, R. T. and Saravolatz, L. D. (2010), Effect of Establishing Guidelines on Appropriate Urinary Catheter Placement. Academic Emergency Medicine, 17: 337–340. doi: 10.1111/j.1553-2712.2009.00677.x
The authors are grateful for the financial support by a Blue Cross Blue Shield of Michigan Foundation grant 1270.II.
- Issue published online: 1 MAR 2010
- Article first published online: 1 MAR 2010
- Received August 11, 2009; revision received September 17, 2009; accepted September 18, 2009.
- urinary catheterization;
- emergency hospital service;
- medical education
Objectives: Avoiding placement of unnecessary urinary catheters (UCs) in the emergency department (ED) affects UC utilization during hospitalization. The authors sought to evaluate the effect of establishing institutional guidelines for appropriate UC placement coupled with emergency physician (EP) education on UC utilization.
Methods: Urinary catheter utilization was measured before and after the establishment of guidelines and EP education. Data collected included the presence of a UC on ED arrival, placement of a UC in the ED, documentation of a physician order for UC placement, reasons for placement, and compliance with the guidelines. Chi-square analyses were used to study the association between pre- and postintervention time periods and catheter use.
Results: A total of 377 (15%) patients had UCs; only 151 (47%) UCs initially placed in the ED had a physician order documented. UC placement was appropriately indicated in 75.5% of patients with a documented physician order, but in only 52% of cases without a documented physician order (p < 0.001). The physician intervention was associated with an overall reduction in UC utilization from 16.4% to 13% (p = 0.018). Physicians ordered 40% fewer UCs postintervention compared to preintervention. Preintervention, a physician order for UC placement was found indicated in 72.6% patients, compared to 82.2% patients with UC placed postintervention (p = 0.21).
Conclusions: Establishing guidelines for UC placement and physician education in the ED were associated with a marked reduction in utilization. However, addressing appropriate UC utilization may require evaluating other factors such as nursing influence on utilization.
ACADEMIC EMERGENCY MEDICINE 2010; 17:337–340 © 2010 by the Society for Academic Emergency Medicine
Urinary tract infections (UTIs) account for 36% of all hospital-acquired infections, with catheter-associated UTI (CAUTI) representing the majority of these cases.1 The optimal prevention of CAUTI is not placing a urinary catheter (UC) if not indicated, or if placed, to remove it when no longer needed.2,3 Avoiding placement of unnecessary UCs in the emergency department (ED) may significantly affect UC utilization during hospitalization.4,5 We recently demonstrated that evaluating the need for UCs in the non–intensive care units significantly reduced unnecessary UC use.6 There are currently no consensus guidelines for UC placement in the ED. The purpose of this study was to evaluate the effect of establishing consensus institutional guidelines on the practice of emergency physicians (EPs) regarding the utilization of UCs.
This was a quasi-experimental study (pre–post intervention design) conducted over 21 months from January 2007 to September 2008. No patient identifiers were entered in the final database. We obtained support from key administrative and departmental leaders before starting the study and obtained institutional review board approval.
Study Setting and Population
This study was conducted in a 769-bed urban tertiary care teaching hospital with a Level 2 trauma center and an ED census of approximately 80,000 adult patients per year. The participants were attending EPs (n = 39) and resident physicians (n = 30, mostly emergency medicine, with occasional rotating residents from internal medicine, family medicine, obstetrics and gynecology, and transitional year programs).
The study was divided into two periods: preintervention and postintervention. For the preintervention period, all admissions through the ED during 5 consecutive days every quarter for four quarters (January 2007 to December 2007, total of 20 days) were sampled.
In October 2007, we established institutional guidelines for UC placement in the ED (Table 1). Approval of these guidelines was obtained after review by the EP attending staff. These guidelines were then presented in the form of a lecture (by MGF) to the EPs and resident staff in December 2007. The lecture included all of the new established indications and non-indications for UC placement. We defined an indication as a reason of UC placement that was in compliance with the new institutional guidelines. These included published recommendations for indications of UC placement,2,3 in addition to accepted indications pertinent to the ED setting. These included acute hip fracture until surgical correction, short-term use for unresponsive or severely agitated patients, and emergent pelvic ultrasound for evaluation of ectopic pregnancy. We also included severe hypoxia requiring ≥6 L/min oxygen (or 40% FIO2) and patients undergoing emergency surgery. An EP champion (MEP) addressed the UC utilization with her peers and resident physicians through direct communication and providing them with pocket cards that included the indications for utilization of UCs. We collected data postintervention on all admissions during 5 consecutive days every quarter for three quarters (January 2008 to September 2008, total of 15 days).
|Reasons for UC Utilization (n = 322)||Number (% of Total Catheters Used)|
|Compliant with guidelines (n = 203, 63.0%)|
|Non–intensive care ≥6 L/min oxygen||40 (12.4)|
|Output monitoring in intensive care||39 (12.1)|
|Emergent pelvic ultrasound||33 (10.2)|
|Neurogenic bladder||14 (4.3)|
|Emergency surgery||12 (3.7)|
|Urinary obstruction||10 (3.1)|
|Acute hip fracture||6 (1.9)|
|Urologic procedures||4 (1.2)|
|Acute mental status changes with agitation||4 (1.2)|
|Stage 3 or 4 sacral decubitus ulcers with incontinence||1 (0.3)|
|Hospice or palliative care||1 (0.3)|
|Noncompliant with guidelines (n = 119, 37.0%)|
|No clear reason||64 (19.9)|
|Oxygen supplementation <6 L/min||26 (8.1)|
|Urine specimen collection||5 (1.6)|
|Patient request||3 (0.9)|
|Output monitoring outside intensive care||2 (0.6)|
Data on UC presence on ED arrival, placement of a UC in the ED, documentation of a physician order for UC placement, reasons for placement, and compliance with the indications were collected retrospectively reviewing the ED records. Only one reason for UC use was documented per patient. All cases that did not have a clear reason for placement were reviewed by two of the investigators; those without adequate documentation were classified as “no clear reason for placement.” The same variables were collected in the preintervention and postintervention periods. Measurements included UC utilization rate ([number of patients admitted with UCs/number of patients admitted] × 100), proportion of indicated UCs ([number of indicated UCs/total number of UCs] × 100), and proportion of nonindicated UCs ([number of nonindicated UCs/total number of UCs] × 100). Our primary outcome measure was total and indicated (based on our institutional guidelines) UC utilization rate in the ED.
Chi-square analyses were used to study the association between pre- and postintervention time period and catheter use. All data analyses were performed using SPSS v.12.0 (SPSS Inc., Chicago, IL), and a p-value of 0.05 or less was considered to indicate statistical significance.
UC Utilization and Indications During the Study
Of the 2,517 sampled patients admitted, 377 (15.0%) had UCs; 55 of them were placed prior to being seen at the ED and were excluded. The reasons for UC placement in the ED during the study period are described in Table 1. Only 151 of 322 (47.0%) UCs initially placed in the ED had a physician order documented. UC placement was indicated in 114 (75.5%) cases with physician order, but in only 89 (52.0%) cases without physician order (p < 0.001, odds ratio [OR] = 2.84, 95% confidence interval [CI] = 1.76 to 4.57).
The Effect of Establishing Indications on UC Utilization
UC utilization dropped significantly after the physician intervention, from 212 of 1,421 (14.9%) preintervention to 110 of 1,041 (10.6%) postintervention (p = 0.002, OR = 1.48, 95% CI = 1.16 to 1.90). Moreover, physicians ordered fewer UCs postintervention (45 of 1,041, 4.3%) compared to preintervention (106 of 1,421, 7.5%; p = 0.002, OR = 0.56, 95% CI = 0.39 to 0.80).
There were no significant differences between the proportions of nonindicated UCs placed before (n = 80 of 238 UCs, 33.6%) and after the intervention (n = 41 of 139 UCs, 29.5%, p = 0.41). Preintervention, a physician order for UC placement was found indicated in 77 of 106 (72.6%) patients compared to 37 of 45 (82.2%) patients with UC placed postintervention (p = 0.21). No differences were seen in those cases where a UC was placed without a physician order (55 of 106 [51.9%] vs. 34 of 65 [52.3%], p = 0.96).
Urinary catheter use is associated with the majority of hospital acquired UTIs. In addition, UCs may increase patient discomfort and restrict activity.7 Immobility has been associated with much morbidity, including pressure ulcers.8 The Centers for Medicare and Medicaid Services (CMS) have included CAUTI as one of 10 conditions that are nonreimbursable if acquired during a hospitalization; this has focused attention on the utilization of UCs and the infection risk associated with their use.9 Because most of a hospital’s admissions come through the ED, it is important that the ED be targeted as the “point of entry” where efforts to reduce unnecessary UC utilization should be directed.
The best way to prevent the complications of a UC is not to insert a UC unless indicated. Preventing the placement of unnecessary UCs in the ED should help reduce the prevalence of unnecessary UCs on the general medical–surgical wards. Evaluating the need for a UC based on criteria for patients admitted through the ED, and enforcing compliance with the indications for placement, will likely reduce unnecessary UC utilization. Before we started our intervention, we developed institutional guidelines for UC utilization in the ED. These guidelines were reviewed by the EP attending staff and approved.
Our results show that establishing indications for UC placement and obtaining physician “buy-in” are associated with a reduction in UC placement in the ED. This reduction in utilization may also be in part related to having an EP champion who promoted these indications. EPs ordered 40% fewer UCs postintervention compared to preintervention. We expect that this would reduce UC utilization hospitalwide. In a hospital where about 25,000 patients are admitted annually through the ED, the reduction in utilization translates to 700 fewer UCs placed in the ED per year.
We also found that the presence of a physician order was associated with a higher compliance with indications. Notably, more than half of the UCs were placed without physician orders. About half of them did not fit any of the indications. This may reflect a difference in what EPs and nurses had considered appropriate indications for UC placement. The nursing staff was not targeted in our intervention; sharing the new institutional guidelines and educating them about the risk of inappropriate UC utilization is needed. Recognizing the factors that lead nursing to inappropriate placement of UCs may help us target the noncompliance with UC indications.
Our data were obtained by a retrospective chart review of the ED records (although the intervention was prospectively planned and executed). Second, in cases where no physician order was documented, physicians may have agreed with the nurse’s decision of placement but did not write an order. This may partly bias our results. Third, the sample was based on selecting 5 days every quarter to evaluate the utilization of UCs. This was done the first week of each quarter in the study. This may not reflect completely the total ED population admitted to the hospital. Finally, the indications that we have applied were reached after consultation with our ED staff and their acceptance of the criteria. This may impact generalizability if other EPs at the different hospitals do not share the same views of the indications.
Establishing guidelines for urinary catheter placement and physician education in the ED were associated with a marked reduction in utilization. Our data suggest that addressing appropriate urinary catheter utilization needs to be based on having clear adoptable indications used by both the EPs and the nurses. Physician education alone, although associated with a lower urinary catheter placement, may not be sufficient to improve the appropriate utilization of urinary catheters.
- 9Centers for Medicare and Medicaid Services. Hospital-acquired Conditions (Present on Admission Indicator). Available at: http://www.cms.hhs.gov/HospitalAcqCond/06_Hospital-Acquired_Conditions.asp#TopOfPage. Accessed Sep 10, 2009.