Outcomes Associated With Insertion of Indwelling Urinary Catheters by Medical Students in the Operating Room Following Implementation of a Simulation-Based Curriculum

Abstract

Purpose

Catheter-associated urinary tract infection (CAUTI) is a priority quality metric for hospitals. The impact of placement of indwelling urinary catheter (IUC) by medical students on CAUTI rates is not well known. This study examined the impact of a simulation-based medical student education curriculum on CAUTI rates at an academic medical center.

Method

Patient characteristics, procedural data, and outcome data from all operating room IUC insertions from June 2011 through December 2016 at the Northwestern University Feinberg School of Medicine were analyzed using a multivariable model to evaluate associations between CAUTI and inserting provider. Infection data before and after implementation of a simulation-based IUC competency course for medical students were compared.

Results

A total of 57,328 IUC insertions were recorded during the study period. Medical students inserted 12.6% (7,239) of IUCs. Medical students had the lowest overall rate of CAUTI among all providers during the study period (medical students: 0.05%, resident/fellows: 0.2%, attending physicians: 0.3%, advanced practice clinicians: 0.1%, nurses: 0.2%; P = .003). Further, medical student IUC placement was not associated with increased odds of CAUTI in multivariable analysis (odds ratio, 0.411; 95% confidence interval: 0.122, 1.382; P = .15). Implementation of a simulation-based curriculum for IUC insertion resulted in complete elimination of CAUTI in patients catheterized by medical students (0 in 3,471).

Conclusions

IUC insertion can be safely performed by medical students in the operating room. Simulation-based skills curricula for medical students can be effectively implemented and achieve clinically relevant improvements in patient outcomes.

Nosocomial infection constitutes a significant burden to the U.S. health care system. Approximately 1 in 25 inpatients experience a health care-associated infection each day, with the majority of infections related to indwelling devices.¹ Catheter-associated urinary tract infection (CAUTI) is one of the most common nosocomial infections worldwide.² A large-scale survey from the Centers for Disease Control and Prevention (CDC) suggested that in 2011, over 35,000 CAUTIs were diagnosed in the United States alone, representing 67.7% of all health care-associated urinary tract infections.¹ CAUTI is also shown to be a leading cause of secondary bloodstream infections, with an associated mortality rate as high as 32.8%.³ As much as 69% of CAUTI events are thought to be avoidable.² In an attempt to address CAUTI-related medical costs, estimated in excess of $131 million annually,³ the Centers for Medicare & Medicaid Services (CMS) decided in 2008 to no longer provide reimbursement for health care costs associated with CAUTI.^4–6

With the introduction of public reporting of hospital-associated infection rates through the Medicare Patient Safety Monitoring System in 2001, increased focus has been placed on efforts to reduce the frequency of these events. Large-scale efforts to reduce CAUTI rates, such as the Michigan Health and Hospital Association Keystone Center Bladder Bundle Initiative,^7,8 have resulted in a significant decrease in indwelling urinary catheter (IUC) utilization and CAUTI rates nationwide.⁹ CAUTI prevention may be approached using a variety of strategies: avoiding the catheter outright, improving placement technique, increasing the rapidity with which it is removed, and preventing reinsertion.^10,11 Guidelines from the Healthcare Infection Control Practices Advisory Committee further recommend that catheters should only be placed by trained personnel using a specific protocol with aseptic technique and that competence should be assessed.^12,13 Because of the learning curve associated with clinical skills, complication rates are known to decrease with experience.¹⁴ In many settings, these findings may indeed provide motivation to have the most experienced providers place catheters due to fear that inexperience may increase the risk of patient harm. Some institutions have gone so far as to develop policies aimed to prevent medical students from placing IUCs for this reason.¹⁵ However, these concerns must be balanced with the societal need to provide meaningful and safe technical skills education for new providers, particularly in the academic medical center environment.

Several studies have shown that educational interventions and bundled process measures targeting nurses are associated with reduction in CAUTI rates.^16–18 Over a 4-year period and 895 IUC insertions, He and colleagues showed an association between medical student placement of IUC and increased CAUTI rates,¹⁹ while Yang and colleagues demonstrated that a credentialing pathway for medical students on a surgery clerkship resulted in CAUTI rates equivalent to those of other providers.¹⁵ However, these studies analyzed a small cohort of both students and patients. We sought specifically to determine whether medical student placement of an IUC was associated with an increase in CAUTI rates using a large-scale sample in the academic medical center setting. Further, we examined the efficacy of a simulation-based IUC placement curriculum during the surgical clerkship in reducing CAUTI events.

Method

This retrospective review was performed at a large academic institution, Northwestern University Feinberg School of Medicine, that instructs approximately 180 medical students per class. The core clerkships at our institution are formatted in 8-week rotations. The surgery clerkship comprises two 4-week blocks with one rotation in abdominal/general surgery and the second in a surgical specialty. Data for this study were collected as part of a chartered institutional quality improvement project and were accordingly deemed exempt from institutional review board review. We reviewed operating room (OR) records from an institutional electronic database from June 2011 through December 2016. For each operation, IUC placement and the role of the inserting provider (attending, resident, nurse, medical student) are documented by the OR circulating nurse from a dropdown menu. From these records, we identified patients with initial IUC placement in the OR. Patient cohorts were then defined based on the role of the inserting provider (medical student, resident/fellow, attending physician, advanced practice clinician, or nurse). We excluded from the study patients with IUCs for whom the inserting provider could not be determined. The patient’s gender, age, body mass index (BMI), and American Society of Anesthesiologists (ASA) class were collected. We broadly grouped the type of surgical procedure by surgical specialty based on Current Procedural Terminology code: cardiac surgery, gynecology, otolaryngology, neurosurgery, orthopedics, plastic surgery, spine surgery, thoracic surgery, abdominal transplant surgery, urology, and vascular surgery. The length of the operative procedure as documented in the operative record was also included in the analysis. We collected the time in hours that the catheter remained in place from the postoperative clinical documentation by reviewing data contained in an institutional electronic data warehouse.

Simulation-based curriculum

Using the “just-in-time training” construct,²⁰ medical students rotating on the surgery clerkship began a mandatory structured simulation-based training on IUC insertion in June 2014. This training included interactive discussions on IUC insertion followed by a 20-minute, small-group, hands-on practice session using a low-fidelity training mannequin. Faculty were present and available to provide immediate feedback to students on insertion technique. Training sessions were conducted during the surgery clerkship orientation before the students’ first day of the rotation. Evaluations were intended as formative experiences for the students rather than summative examinations to assess mastery. After the initial implementation, training was further refined to facilitate formal verification of proficiency and time for deliberate practice. Beginning in March 2016, a formal training curriculum for IUC insertion was implemented as a component of the surgery clerkship orientation. The curriculum included a video-based didactic session using a video provided by the manufacturer of the IUC insertion kit (C. R. Bard, Inc., Covington, Georgia), which discussed the health care implications of CAUTI and demonstrated how to use the kit.²¹ The didactic session was followed by a deliberate practice session lasting 1 hour. Again, institutional-specific insertion kits and low-fidelity mannequins were used. This session was formatted to include a discussion of institutional policies regarding IUC insertion, indications for catheters placement, and the development and impact of CAUTI, followed by supervised deliberate practice with the clerkship nurse educator. Students were subsequently required to demonstrate proficiency in IUC insertion. Proficiency was assessed by independent clinical nursing educators using a standardized universal checklist created by our institution (Supplemental Digital Appendix 1, available at http://links.lww.com/ACADMED/A768). This checklist is used to qualify all providers who insert IUCs, including physicians, advanced practice providers, and nurses. Students rotating with the obstetrics and gynecology clerkship are exposed to IUC insertion; however, they receive no formal procedural training. Thus, the surgery clerkship is the first time medical students at our institution receive standardized training in IUC insertion before insertion in patients.

Outcomes

Our primary outcome of interest was the development of CAUTI within 90 days of IUC insertion in the OR. We defined CAUTI events according to standard guidelines established by the CDC.²² Infection events were defined by an independent Infection Control department through retrospective chart review using the methodology by which data are provided to CMS to conform to established standards for extramural reporting. We cross-referenced all patients with a CAUTI with OR IUC data to determine if the IUC was initially placed in the OR and, if so, the role of the inserting provider.

Statistical analyses

We designed this study to examine 2 primary end points: the rate of CAUTI associated with medical student IUC placement compared with IUC placement by other providers for the entire study period and the rates of CAUTI before and after the implementation of a medical student simulation-based IUC training curriculum in June 2014. Univariable and multivariable logistic regression models were used to examine the effects of patient sex, age, and BMI; surgical specialty; duration of surgery; and role of the IUC inserting provider. We chose variables for inclusion in logistic regression models a priori to include factors previously identified as associated with CAUTI development. We also assessed implementation of the educational intervention by categorizing patients into pre- and postintervention time cohorts based on operative date, with the preintervention insertions being assessed from June 2011 through June 2014. For patients with repeated measures but no infections, the time from surgery to their last follow-up date was used in the model. Patients with multiple encounters were included in the model with the date of initial IUC placement considered as time zero for analytic purposes. We truncated data analysis 90 days after IUC insertion in the OR if no CAUTI event occurred. Age, BMI, and operative case duration were binned into discrete categories for the models. The time IUC remained in place was left as a continuous variable. We calculated odds ratios (ORs) with 95% Wald confidence intervals, and a 2-tailed P value of < .05 was considered statistically significant. We generated all analyses using SAS statistical software, Version 9.4 of the SAS System for Windows (SAS Institute, Inc., Cary, North Carolina).

Results

This retrospective cohort study included 69,105 patients with IUC insertions in the OR from June 2011 through December 2016. The inserting provider was documented in 57,328 (83%) IUC insertions. The remaining 11,777 (17%) records were missing inserting provider documentation and were therefore excluded. For insertions in which the provider was known, medical students accounted for 12.6% (7,239) of insertions, resident/fellow physicians 53.9% (30,869), attending physicians 14.8% (8,495), advanced practice clinicians 11.2% (6,437), and nurses 7.5% (4,288) (Table 1). Patients showed a slight female predominance of 56% (32,090) versus 44% (25,238) and a median age of 57 years. The median BMI of the population was 27.5 kg/m² although more catheters were placed in patients who were either overweight (37.9%, n = 21,721) or obese (30.6%, n = 17,535). Indwelling catheters were used most frequently in general surgery procedures (16.9%, n = 9,716), followed by gynecologic surgery (16.4%, n = 9,421), orthopedic (16.1%, n = 9,241), and urologic operations (13.2%, n = 7,557). The majority of operations where IUCs were placed in the OR lasted between 1 and 2 hours, and catheters remained in place for a median of 23 (6–64.1) hours.

Table 1.

Demographic Comparison of Patients With and Without CAUTI, From a Study of Effects of a Simulation-Based Curriculum on CAUTI Rates, Northwestern University Feinberg School of Medicine, 2011–2016^a

Characteristic	No CAUTI	CAUTI	Total	P value
Patient gender, no.				.41b
Male	25,190	48	25,238
Female	32,018	72	32,090
Patient age, no.				.08
18–30	3,272	6	3,278
31–40	6,530	7	6,537
41–50	9,103	15	9,118
51–60	11,831	20	11,851
61–70	14,078	41	14,119
71–80	9,162	21	9,183
80+	3,232	10	3,242
BMI, no.				.23
< 18	721	3	724
18.1–24.9	17,313	35	17,348
25–30.9	21,684	37	21,721
31–34.9	8,195	18	8,213
35–39.9	5,193	13	5,206
> 40	4,102	14	4,116
Operative case type, no.				< .001
General surgery	9,692	24	9,716
Cardiac surgery	4,523	13	4,536
Gynecology	9,409	12	9,421
Otolaryngology	760	1	761
Neurosurgery	3,520	12	3,532
Orthopedics	9,232	9	9,241
Plastic surgery	1,768	4	1,772
Spine surgery	4,490	9	4,499
Thoracic surgery	1,478	1	1,479
Abdominal transplant surgery	3,346	7	3,353
Urology	7,546	11	7,557
Vascular surgery	1,444	17	1,461
Operative case duration, no.				< .001
> 1 hour	6,680	9	6,689
1–2 hours	14,257	18	14,275
2–3 hours	12,926	21	12,947
3–4 hours	9,739	18	9,757
4–5 hours	5,730	13	5,743
5–6 hours	3,203	7	3,210
6–7 hours	1,855	11	1,866
7–8 hours	1,175	10	1,185
> 8 hours	1,643	13	1,656
ASA class, no.				< .001
Class 1	16,842	23	16,865
Class 2	23,029	18	23,047
Class 3	14,787	65	14,852
Class 4	2,513	14	2,527
Class 5	37	0	37
Provider inserting IUC, no.				.003
Nurse	4,279	9	4,288
Medical student	7,235	4	7,239
Attending physician	8,467	28	8,495
Resident/fellow physician	30,799	70	30,869
Advanced practice clinician	6,428	9	6,437
IUC insertions: All providers, no.	57,208	120	57,328	.002b
Before simulation training	32,518	85	32,603
After simulation training	24,690	35	24,725
Time IUC remained, hours				< .001c
Median	22.92	81.83	23.00
25th percentile	5.90	45.08	6.00
75th percentile	63.70	246.98	64.10

Abbreviations: CAUTI, catheter-associated urinary tract infection; IUC, indwelling urinary catheter; BMI, body mass index; ASA, American Society of Anesthesiologists.

^aSignificance calculated with chi-square analysis, except as noted.

^bFisher’s exact test.

^cWilcoxon rank-sum test.

Medical student IUC placement resulted in CAUTI in 0.05% of patients (4 in 7,239 insertions), as compared with 0.2% (70 in 30,829 insertions) for residents/fellows, 0.3% (28 in 8,495 insertions) for attending physicians, 0.1% (9 in 6,437 insertions) for advanced practice clinicians, and 0.2% (9 in 4,288 insertions) for nurses. All 4 instances of CAUTI in IUC placement by medical students occurred in the general surgery population. Patients in whom the IUC was placed by a medical student had significantly decreased odds of CAUTI compared with nurse placements (OR: 0.26; 95% confidence interval [CI]: 0.08, 0.85; P = .03) (Table 2). We found no association between CAUTI and gender, age, or BMI in univariable analysis. CAUTI events were significantly associated with orthopedic and vascular operations, cases longer than 6 hours, ASA class of 3 or greater, and increased duration of IUC use after surgery (Table 3).

Table 2.

Univariable Analysis of Factors Associated With CAUTI Development, From a Study of Effects of a Simulation-Based Curriculum on CAUTI Rates, Northwestern University Feinberg School of Medicine, 2011–2016^a

Factor	Odds ratio	95% confidence interval	P value
Provider inserting IUC
Nurse	1 (reference)		—
Medical student	0.263	0.081, 0.854	.03
Attending physician	1.572	0.741, 3.335	.24
Resident/fellow physician	1.081	0.539, 2.165	.83
Advanced practice clinician	0.666	0.264, 1.678	.39
Time IUC remained	1.003	1.003, 1.004	< .001

Abbreviations: CAUTI, catheter-associated urinary tract infection; IUC, indwelling urinary catheter.

^aSignificance calculated with the Wald test.

Table 3.

Multivariable Analysis of Factors Associated With Development of CAUTI, From a Study of Effects of a Simulation-Based Curriculum on CAUTI Rates, Northwestern University Feinberg School of Medicine, 2011–2016^a

Factor	Odds ratio	95% Wald confidence interval	P value
Simulation training	0.48	0.316, 0.73	< .001
Patient gender
Male	1 (reference)	—	—
Female	1.596	1.067, 2.388	.02
Patient age
18–30	1 (reference)	—	—
31–40	0.581	0.193, 1.756	.34
41–50	0.76	0.289, 1.998	.58
51–60	0.717	0.281, 1.828	.49
61–70	1.047	0.429, 2.553	.92
71–80	0.796	0.309, 2.053	.64
80+	0.962	0.333, 2.776	.94
BMI
< 18	1.193	0.357, 3.987	.77
18.1–24.9	1 (reference)	—	—
25–30.9	0.823	0.512, 1.324	.42
31–34.9	1.026	0.573, 1.838	.93
35–39.9	1.06	0.552, 2.032	.86
> 40	1.227	0.642, 2.346	.54
Operative case type
General surgery	1 (reference)	—	—
Cardiac surgery	0.493	0.183, 1.33	.16
Gynecology	0.579	0.276, 1.216	.15
Otolaryngology	0.419	0.055, 3.18	.40
Neurosurgery	0.965	0.468, 1.991	.92
Orthopedics	0.488	0.217, 1.095	.08
Plastic surgery	0.826	0.282, 2.422	.73
Spine surgery	0.603	0.272, 1.335	.21
Thoracic surgery	0.169	0.023, 1.262	.08
Abdominal transplant surgery	0.327	0.129, 0.828	.02
Urology	0.545	0.252, 1.179	.12
Vascular surgery	1.802	0.911, 3.564	.09
Operative case duration
< 1 hour	1 (reference)	—	—
1–2 hours	0.908	0.388, 2.123	.82
2–3 hours	1.165	0.509, 2.666	.72
3–4 hours	1.231	0.53, 2.858	.63
4–5 hours	1.275	0.517, 3.144	.60
5–6 hours	1.215	0.429, 3.437	.71
6–7 hours	2.813	1.088, 7.272	.03
7–8 hours	4.004	1.503, 10.664	.006
> 8 hours	2.639	1.015, 6.862	.047
ASA class
Class 1	1 (reference)	—	—
Class 2	1.003	0.516, 1.951	.99
Class 3	4.788	2.838, 8.078	< .001
Class 4 and 5	5.282	2.463, 11.325	< .001
Provider inserting IUC
Nurse	1 (reference)	—	—
Medical student	0.411	0.122, 1.382	.15
Attending physician	2.684	1.187, 6.068	.02
Resident/fellow physician	1.681	0.816, 3.463	.16
Advanced practice clinician	0.732	0.225, 2.388	.61
Time IUC remained	1.003	1.002, 1.003	< .001

Abbreviations: CAUTI, catheter-associated urinary tract infection; BMI, body mass index; ASA, American Society of Anesthesiologists; IUC, indwelling urinary catheter.

^aSignificance calculated with the Wald test.

In an adjusted multivariable analysis, patients whose catheters were inserted by attending physicians had increased odds of infection compared with those whose catheters were placed by nurses (OR: 2.684; 95% CI: 1.19, 6.07; P = .02). When the catheter was placed by a medical student, CAUTI rates were comparable to nurse-placed IUC CAUTI rates (OR: 0.411; 95% CI: 0.122, 1.382; P = .15). Additionally, women were more likely to have a CAUTI (OR: 1.6; 95% CI: 1.07, 2.39; P = .02) while there remained no difference based on age or BMI (Table 3). Abdominal transplant surgery was the only case type to demonstrate a decreased risk for CAUTI compared with general surgery (P = .02). Cases lasting longer than 6 hours and patients with an ASA class of 3 or greater were associated with and increased risk of CAUTI. The duration of IUC use was also significantly associated with increased CAUTI rates (OR: 1.003; 95% CI: 1.002, 1.003; P < .0001).

The incidence of CAUTI for medical student-placed IUC decreased after the introduction of the simulation-based medical student education program. Overall raw rates for all providers dropped from a preintervention rate of 0.3% to a postintervention rate of 0.1% (P = .002) with medical student rates specifically decreasing from 0.1% (4 in 3,764) to 0% (0 in 3,471; P = .13). Additionally, no other inserting provider category experienced a significant decrease in CAUTI incidence from pre to post intervention with the exception of the nurse category (P = .004). The monthly rate of IUC placement decreased from an average of 906 cases per month in the 36 months before the intervention to 824 cases per month over the 30 months following the intervention (P ≤ .001). CAUTIs remained associated with increased catheter days pre and post intervention; however, the duration of IUC use decreased from a mean of 79.9 (10.7–93.2) hours to 27.6 (2.8–25.9) hours (P < .001) in the period following the intervention.

Discussion

Our findings demonstrate the safety of medical student IUC insertion in the OR. This study was initially motivated by institutional concern that the relative inexperience of medical students might result in patient harm when students were permitted to perform IUC placement.¹⁵ A prior single-institution study suggested that placement of an IUC by a medical student was an independent risk factor for CAUTI.¹⁹ However, fewer than 900 insertions over a 3-year period were evaluated in that study, limiting the ability to draw generalizable conclusions. Our study, by contrast, included a comprehensive analysis of over 57,000 IUC insertions within the OR in a diverse cohort of patients. At a minimum, our findings show that CAUTI risk appeared similar for medical student and nursing IUC placement.

Medical students should not be excluded from specific aspects of patient care without clear cause as this is contradictory to the mission of undergraduate medical education. Appropriate training and supervision of students are necessary to ensure that their involvement in care is safe. While there is much interest in the outcomes of procedures performed by trainees, there is a paucity of data related to higher levels in the Kirkpatrick hierarchy.²³ Our study achieves the goal of observing patient-level outcomes directly attributable to medical student activities. This study can be used as a model for research design in medical education and answers the frequent call to link educational outcomes with meaningful clinical outcomes at the patient level.^20,24,25 Our work demonstrates that with appropriate education and supervision, medical students can safely be trained to insert IUCs without increased infection risk to the patient, thus achieving meaningful educational and clinical end points.

The implementation of a simulation-based curriculum for medical students essentially eliminated CAUTI events during the study time frame, dropping from 4 infections in 3,764 medical student IUC insertions to 0 infections in 3,471 cases following the intervention. Similar work looking at simulation and patient-level outcomes has yielded affirmative results, especially when evaluating device-associated infection rates.^14,26–30 Several studies have discussed options for teaching procedural skills such as IUC placement to medical students.^31,32 There is significant institutional variance in patient safety education methodologies³³; however, checklist assessments such as those currently in use at our institution can be a valid and reliable way to evaluate clinical skills.³⁴ This simulation training program for IUC insertion successfully achieved recognized core competencies in medical education including patient safety and quality improvement^33,35 and resulted in a significant reduction in the rate of CAUTI following its implementation as a component of the medical student curriculum during the surgery clerkship.

Limitations

Specific limitations of this study should be noted that might affect the generalizability of the findings. First, these data reflect the experience at a single academic medical center and may be subject to confounding factors related to the institution itself. For example, the medical center supports ongoing quality improvement efforts aimed at CAUTI reduction, including an institution-wide initiative starting in the spring of 2014 to address the need for intraoperative IUC placement and emphasize adherence to a 2-person sterile technique. This initiative required a second provider to assess the indication for the placement of an IUC and ensure sterile technique is adhered to during the procedure. This initiative also included a sign-out protocol for catheter removal post procedure to ensure there was a definitive plan for IUC duration and indication. For OR cases less than 3 hours in duration, this institutional initiative reduced the number of unnecessary catheter placements from 14.2 to 12.9 catheters per day (9.1% change) and increased the rate of immediate postprocedure catheter removal from 18.3% to 21.0%.³⁶ The mean duration of IUC placement decreased from 80 hours to 28 hours (P < .0001). These interventions, which overlapped with the implementation of the student-level training, may have influenced the observed results, particularly those related to catheter duration. In addition, because we looked at IUC insertions exclusively in the OR setting, the overall infection rate may have been lower than that observed for catheters placed in other clinical settings,³⁷ and care should be taken when extrapolating this experience outside of the surgical setting. While mitigated by verification of proficiency before IUC insertion in patients, medical students’ experience will vary depending on the timing of their progression within the school year, which could conceivably lead to varying degrees of proficiency and confidence based on previous clerkship experience with IUC placement. Medical students may be selectively permitted to place catheters in lower-acuity patients or those with more favorable anatomy. Additionally, they may be under increased scrutiny by faculty and staff, thereby decreasing the risk of CAUTI to patients. While we attempted to control for this by including both ASA classification and BMI in our analysis, it is possible that some bias may still exist. Finally, we must acknowledge the 17% of the cohort who lacked inserting provider data during the study period. It cannot be known if these data harbor additional insights into the outcomes of interest; however, we attempted to minimize this limitation with our large sample size.

Conclusions

At a large academic institution, IUC insertion by a medical student in the OR was not associated with increased odds of CAUTI compared with other licensed providers. While the development of a CAUTI is multifactorial in nature, the implementation of a simulation-based training program for medical students before starting rotations on their surgery clerkship reduced CAUTI to zero in patients who had an IUC placed by a medical student in the OR. It is important to keep medical students actively involved in as many aspects of patient care as possible by aligning educational and clinical outcomes.^35,38,39 The successful implementation of this simulation-based medical education program demonstrates the potential of this instructional format to result in the long-term, cost-effective acquisition and retention of clinical skills. Further, this study underscores the importance of linking educational and clinical outcomes, which may serve as a model research design for future studies in medical education.