Simulation-based paediatric intensive care unit central venous line maintenance bundle training

Simulation-based paediatric intensive care unit central venous line maintenance bundle training

Intensive and Critical Care Nursing (2015) 31, 44—50 Available online at www.sciencedirect.com ScienceDirect journal homepage: www.elsevier.com/iccn...

647KB Sizes 0 Downloads 17 Views

Intensive and Critical Care Nursing (2015) 31, 44—50

Available online at www.sciencedirect.com

ScienceDirect journal homepage: www.elsevier.com/iccn

Simulation-based paediatric intensive care unit central venous line maintenance bundle training夽 Kiran B. Hebbar a,b,∗, Charlene Cunningham a, Courtney McCracken b, Pradip Kamat a,b, James D. Fortenberry a,b a b

Children’s Healthcare of Atlanta at Egleston, United States Emory University School of Medicine, United States

Accepted 20 October 2014

KEYWORDS Bundle; Blood stream infection; Central line; Paediatric; Simulation

Summary Objective: Research has demonstrated that additional reduction in paediatric catheter associated blood stream infection (CA-BSI) rates can be achieved through improving compliance with maintenance bundle care for central venous lines. Our objective was to improve maintenance bundle compliance rates and nursing competency surrounding central venous line (CVL) care in our paediatric intensive care unit (PICU). Methods: A multidisciplinary team developed a bedside simulation-based training programme to improve compliance with standard PICU CVL maintenance bundle. We then performed a randomised comparison study comparing a standard CVL bundle training process for bedside PICU nurses in a control group (CG) to an intervention group (IG) receiving bedside training to simulate a CVL dressing change and maintenance bundle followed by intermittent training refreshers. Groups were assessed for compliance with prescribed components of the CVL bundle maintenance (CVL score). Results: At baseline the CG and IG had similar mean CVL scores (p = 0.725). At twelve months mean CVL bundle compliance score in the IG was significantly higher than in the CG (p < 0.0001). The largest CVL score increase for IG occurred between zero and three months. Coincidentally,

Abbreviations: CA-BSI, catheter associated blood stream infection; CVL, central venous line; IG, intervention group; CG, control group. ClinicalTrials.gov Identifier: NCT01960556. ∗ Corresponding author at: Division of Critical Care Medicine, Department of Pediatrics, Emory University and Children’s Healthcare of Atlanta at Egleston, 1405 Clifton Road NE, Atlanta, GA 30322, United States. Tel.: +1 404 785 6135; fax: +1 404 785 6233. E-mail address: [email protected] (K.B. Hebbar). 夽

http://dx.doi.org/10.1016/j.iccn.2014.10.003 0964-3397/© 2014 Elsevier Ltd. All rights reserved.

Simulation CVL education and associated decrease CA-BSI

45

CA-BSI rates in the Egleston PICU significantly decreased from 1.9 ± 2.2 BSIs per 1000/CVL days, prior to the study, to 0.6 ± 1.6 BSIs per 1000/CVL days following implementation of the intervention (p = 0.034). Conclusions: Bedside simulation based training in CVL dressing change is associated with improved compliance with CVL maintenance bundle practice. Enhanced CVL maintenance bundle practice could contribute to reduction in CA-BSI rates. © 2014 Elsevier Ltd. All rights reserved.

Implications for Clinical Practice • Previous studies demonstrated that additional reduction in paediatric catheter associated blood stream infection rates could be achieved through improving compliance with maintenance bundle care for central lines. Maintaining competence and performance of central venous line care can be challenging. • Simulation based training initiative makes a significant impact on maintenance bundle care performance. • Simulation may be a rapid and effective method for educating bedside nurses on a key factor in the fight against catheter associated blood stream infections (CA-BSI). Further studies are needed. • Bedside simulation helped contribute to reduction in catheter associated blood stream infections in a busy paediatric intensive care unit.

What is known on this subject? • Previous studies have demonstrated that additional reduction in paediatric catheter associated blood stream infection rates could be achieved through improving compliance with maintenance bundle care for central lines. • Maintaining competence and performance of central venous line care can be challenging. What this study adds • The simulation based educational initiative can significantly impact upon maintenance bundle care performance and contribute to reduction in catheter associated blood stream infections in busy paediatric intensive care units.

Introduction Catheter associated blood stream infections (CA-BSI) remain a source of significant morbidity and mortality in adult and paediatric critical care units (Burke, 2003; Pittet et al., 1994; Siempos et al., 2009; Wylie et al., 2010). These infections with Staphylococcus aureus, Streptococcus species and other skin flora are a significant source of morbidity and mortality. In addition to the impact on patient care and outcome, CA-BSIs add significant cost to a patient’s hospitalisation (Dominguez et al., 2001; Margolin et al., 2011; Nowak et al., 2010). Miller et al. demonstrated that insertion bundles alone were not adequate to consistently lower CA-BSI rates, and that focus on maintenance bundles, which are checklists of tasks needed for proper catheter care and dressing change, were essential to drive change (Miller et al., 2010). After adjusting for region and PICU demographics, the only significant predictor of CA-BSI rate decrease was maintenance-bundle compliance (RR: 0.41 [95% CI: 0.20—0.85]; p = 0.017). While it has been established that central venous line (CVL) maintenance bundles are pivotal to achieving and maintaining low CA-BSI rates,

training and retention of proper bundle compliance remains a challenge. Simulation training has been a rapidly growing tool for healthcare quality to promote clinical competency and allow reflective thinking to improve skill acquisition (Gaba, 2007; Haskvitz and Koop, 2004; Long, 2004). We prospectively compared this simulation-based programme to a standard training approach. Our purpose was to improve compliance with CVL maintenance bundle procedure. We hypothesised that implementation of bedside simulationbased training of CVL dressing changes by registered nurses would improve CVL maintenance bundle compliance.

Methods Process improvement and development In 2010, our PICU quality committee performed a retrospective review of 10 identified CA-BSIs, of which six (60%) demonstrated significant issues with CVL maintenance bundle compliance. Dressing change training was identified as a key opportunity for improving compliance. Beginning in 2010, a revised nursing training programme was developed which included didactic training on the maintenance bundle, followed by competency assessment, post-test evaluation and required observation of a dressing change on an annual basis. Simultaneously, in cooperation with the Children’s Healthcare of Atlanta Simulation Center and its medical director (KBH), unit leadership developed an alternative training approach utilising an intensive bedside-based simulation teaching method. We made use of a low fidelity mannequin with a central line in place. Several studies have identified skill deterioration within six months for technical and basic skill principles among paramedics and respiratory care personnel (Long, 2004; Skelton and McSwain, 1977; Tuttle et al., 2007; Zautcke et al., 1987). Therefore we elected to build the simulation based training strategy to provide systematic refreshes at three, six and twelve months.

46

Implementation and trial After building a bedside simulation-based training programme, we conducted a prospective cohort comparison study of traditional based and simulation based central venous line (CVL) care training to evaluate knowledge retention and ‘‘bundle’’ compliance in the paediatric intensive care unit (PICU) at Children’s Healthcare of Atlanta at Egleston. The PICU is a 30 bed multidisciplinary medical—surgical ICU providing tertiary and quaternary care for patients 0—21 years of age. Patients receive care for all standard paediatric critical illnesses, as well as solid organ transplantation, continuous renal replacement therapies and extracorporeal life support. Institutional review board (IRB) approval was obtained. All nursing participants provided informed consent. Each subject was assigned a unique study number when entered into a database. Demographic information on the subjects (nurses), their recent exposure to CVL dressing care and CVL dressing change training was also collected. The study was performed between December 2010 and December 2011. Any nurse assigned to the PICU on a regular basis was eligible to participate in the study, including full-time, part-time and occasional (prn) PICU nursing staff, as well as central staffing office (ICU float nurses) and traveller nurses contracted to the PICU for a minimum of twelve weeks. The PICU nursing group was ‘‘simulation-’’naïve’’, having no exposure to scenarios or simulation-based teaching prior to the start of this study. Nurse participation was randomised to either the simulation-based group (intervention group-IG) or the standard CVL dressing group (control groupCG) using random name draws. The IG performed a central line dressing change with CVL standard prepackaged dressing change kits on a low fidelity mannequin with a subclavian venous central catheter in place. The mannequin was placed on a cart and taken to the bedside as the nurses worked in the PICU. Nurses in the CG completed standard training, as described above, with a demonstration, self-study poster and test. Both the IG and the CG observed a CVL dressing change on a low-fidelity mannequin with a temporary, non-tunnelled CVL at a patient’s bedside, in an unoccupied patient room or training classroom located in the PICU depending on acuity and room availability. The investigators served as the second nurse assistant, but did not provide instructive comments during the dressing change. Following the skill demonstration, the nurse in the CG returned to clinical duties. Nurses in the IG additionally performed the dressing change with an instructor who reviewed all points in the checklist. An investigator assessed the demonstrated procedure for participants in each group by calculating a CVL maintenance bundle score (CVL score). The CVL score was devised from the standard 17 point CVL dressing change checklist utilized in our unit and based on CDC guidelines (Table 1). Scores could range from zero if no components were correctly observed, to 17 if all were observed. Scores were entered into a secured de-identified database. Both groups were reassessed twelve months after the initial observation for bundle score compliance. Only the intervention group was reassessed three months, six months and twelve months after the initial dressing change (see Fig. 1).

K.B. Hebbar et al. Table 1 Paediatric intensive care unit (PICU) central venous line (CVL) maintenance bundle checklist used for clinical care for study evaluation. CVL Bundle Checklist 17 points Indications for CVL dressing change Obtain supplies Clean area for sterile supplies Mask all participants Create sterile field w/dressing kit, avoid crossover Hand hygiene, clean gloves, dressing removal Instructions for assistant: Speak up Repeat hand hygiene Don sterile gloves Scrub w/CHG: document measured scrub time Scrub w/back and forth action Dry time: until surface dry Biopatch application Skin protectant application, avoiding hardware & insertion; allow to dry Sterile application of tegaderm Seal line exit site. NO CHEVRON Date, time, initial dressing. Document in epic

Time points for bundle compliance score reassessment were chosen based on previous findings suggesting ‘‘skill decay’’ following simulation training was noted at three months (Skelton and McSwain, 1977; Zautcke et al., 1987). CA-BSI rates for the PICU were tabulated as part of ongoing system-based tracking of CA-BSI for all units and were compared over the study time period.

Statistical analysis All analyses were conducted using SAS 9.2 (Cary, NC). Statistical significance was assessed at the 0.05 level. Descriptive statistics of subjects’ demographics were calculated for all variables of interest (e.g., age, years of experience, years in the ICU, years in the PICU and number of dressing changes in the last month). The participants’ ages were age was classified based on ≤35 years and >35 years. Years of experience and years in the ICU or PICU were classified into groups based on ≤5 years and >5 years. Because subjects were being followed repeatedly over the study period, repeated measures analysis of variance models were used to assess the impact of intervention group (IG vs. CG), study phase (baseline, three months, six months and one year) and other potential confounding factors while adjusting for multiple observations made on the same participant. Interaction terms were included in all the models and the Tukey—Kramer multiple comparison procedure was used to control the type-I error rate when making pair wise comparisons between or within study groups.

Inter-rater reliability Inter-rater reliability was assessed using percentage agreement. Since multiple raters assessed compliance, inter-rater

Simulation CVL education and associated decrease CA-BSI

47

Figure 1 CVL bundle checklist scores throughout study period by study group. (A = Control Group, B = Intervention Group). Significant improvement from baseline and compared to control in maintenance bundle checklist scores was seen at three months and increased over the year.

agreement was established prior to randomisation to ensure that study staffs’ monitoring the CVL checklist were similar. Using the 17-item bundle checklist, items received a score of 1 if the two raters were in agreement (i.e., they both observed the item being performed or both did not observe the item being performed); otherwise they received a score of 0. For each observed simulation, agreement scores could range from 0 (no agreement on any checklist item) to 17 (agreement on all 17 checklist items). Percent agreement was calculated by dividing the agreement score by 17 and multiplying by 100, so that percent agreement ranged from 0% to 100%.

Results Eighty nurses were enrolled and consented to participate in the study; however one did not complete the entire study, leaving 39 in the IG and 40 in CG. randomisation was successful, as the groups were well matched in terms of age and years of experience (Table 2). Average total time spent per each simulation intervention was 10 minutes. Overall, agreement between the two raters was excellent. Six out of the nine observed simulations resulted in perfect agreement (i.e., an agreement score of 17) between the two raters. Amongst raters,

Table 2 Demographic characteristics of PICU registered nurses participating in the study based on study group. The two intervention group (IG) and control group (CG), were well matched in terms of age, overall years of experience, years in the ICU or PICU and average number of dressing changes performed within the month prior to study initiation. Demographics

Level

Control Group CG (N = 40)

Intervention Group IG (N = 39)

Age group (years)

≤35 >35 ≤5 >5 ≤5 >5 ≤5 >5 —

28 (70.0%) 12 (30.0%) 17 (42.5%) 23 (57.5%) 22 (56.4%) 17 (43.6%) 24 (60.0%) 16 (40.0%) 2.6 ± 2.0

23 (59.0%) 16 (41.0%) 12 (30.8%) 27 (69.2%) 14 (35.9%) 25 (64.1%) 16 (41.0%) 23 (59.0%) 2.9 ± 2.4

Years of experience Years on ICU Years in PICU Number of dressing changes in last month

P value

0.306 0.280 0.069 0.092 0.594

48

K.B. Hebbar et al.

Figure 2 Catheter-associated bloodstream infection (CA-BSI) rates per 1000 CVL days from November 2010 to July 2012 and CVL Checklist Scores for Intervention Group. Significant reduction in PICU CA-BSI rates was seen during the study period, correlating with improved CVL checklist scores. Estimated cost saving during this time was $500,000.

inter-rater reliability, measured using percent agreement, was high (98%) (p < 0.001 and 95% CI [96—100%]). At baseline, the CG and IG had similar mean CVL scores (10.9 ± 2.2 vs. 10.4 ± 2.2; p = 0.725) (Fig. 1). However, at the end of the study (twelve months), the mean CVL score from the IG was significantly higher than that of the CG (15.8 ± 1.1 vs. 13.2 ± 2.1; p < 0.0001) (Fig. 1). For the IG, a significant increase in CVL scores from baseline was found at three months, six months and one year (p < 0.0001 for all comparisons with baseline) (Fig. 2). The largest CVL score increase for IG occurred between zero and three months. The increasing trend in CVL score continued through the end of the study; however, CVL scores at three, six and twelve months were not significantly different from one another, indicating knowledge retention. No significant association was found between CVL score and any RN demographic (data not shown). Intra-study and post study period PICU BSI rates significantly decreased from a mean of 1.9 ± 2.2 BSIs per 1000/CVL days for the six months pre-study to 0.6 ± 1.6 BSIs per 1000/CVL days for the last six months and immediately following the study (p = 0.034). These rates remained low following the study (Fig. 2).

Discussion Maintenance care of CVL and bundle compliance significantly impacts CA-BSI rates (Dominguez et al., 2001;

O’Grady et al., 2002). However, adhering to CDC recommended guidelines and maintaining competence can be a challenge for seasoned or novice nursing staff. While many studies have been done to demonstrate the importance of simulation training in the initial placement of CVLs, little work has been performed to assess the impact of the role of simulation in impacting ongoing CVL maintenance (Barsuk et al., 2009; Britt et al., 2009; Moureau et al., 2013; Scholtz et al., 2013). Our study demonstrates that implementation of a simulation-based model with refresher sessions effectively improved maintenance bundle checklist scores and compliance. We were able to perform this training at the RN’s patient bedside, in an efficient time frame. Challenges exist to implementing simulation-based learning (high or low fidelity), including finding time for staff to participate outside of clinical work hours, motivating staff to attend, the cost of scheduling extra training sessions and educator time. Our training approach in this study proved to be efficient, making use of staff time while during clinical work and more importantly impactful CVL maintenance knowledge and psychometric skill application. CA-BSI rates fell significantly during the study period coincidentally with study implementation. Based on estimated costs associated with treating a single CA-BSI, this reduction of CA-BSI rates can be estimated to have resulted in approximately $500,000 U.S. dollars in cost avoidance over the study period (Nowak et al., 2010). Decrease in PICU CA-BSI rates was consistent with a general trend nationally

Simulation CVL education and associated decrease CA-BSI in BSI rates over the same period (Malpiedi et al., 2012). The association of improved training results with decreased CA-BSI rates is intriguing, but of an uncertain relationship, as the study design was neither developed nor powered to test a hypothesis that BSI rates would be impacted by the intervention. It is important to note that no single factor, such as improved maintenance bundle compliance, can fully account for the decline in BSI rates. Multiple interventions aimed at reducing CA-BSIs were undertaken during the time period, including standardising chlorhexidine (CHG) preparation time, improving physician insertion techniques with enhanced use of bedside ultrasonography and decreasing use of intravenous central hyperalimentation. While the improved rates are quite encouraging, separating the effect of these interventions from improved CVL maintenance training would be difficult. Limitations of the study also include the lack of blinding and the relatively small sample size. We elected to use a low fidelity mannequin with a CVL in place. Such simulators provide for sustainability, standardisation, portability and skill specificity (Beaubien and Baker, 2004; Britt et al., 2009). Simulation is especially effective in developing skills in procedures that require eye-hand coordination and ambidextrous manoeuvres (Grantcharov et al., 2004). Some of the advantages of simulation in healthcare training and quality are that it facilitates deliberate practice with feedback, exposure to uncommon events, reproducibility and opportunity for assessment of learners (Barsuk et al., 2009; Britt et al., 2009). In general adults learn best when they can immediately apply what they have learned. Traditional teaching methods, in which an instructor imparts facts to the student in a unidirectional model, have been found less effective in adult learning because of the need for adults to make sense of what they experience or observe (Fanning and Gaba, 2007; Grant and Marriage, 2012). Studies show simulation based training imparts a significantly better transfer of skills to applied clinical practice (Andreatta et al., 2011).

Conclusions Bedside simulation-based RN training of a PICU CVL maintenance bundle was associated with significantly improved compliance scores compared to standard training practice, with evidence of skill knowledge retention. Implementation of enhanced CVL maintenance approaches, such as improved dressing change training and knowledge retention, could be associated with improved CLABSI rates in concert with other interventions. It is uncertain whether implementation of simulation bundle training could directly improve CA-BSI rates. Further study is needed to clinically translate these results.

Funding The authors have no sources of funding to declare.

Conflict of interest The authors have no conflicts of interest to disclose.

49

Authors’ contribution Kiran B. Hebbar conceptualised and designed the study, drafted the initial manuscript, and approved the final manuscript as submitted. Charlene C. Cunningham designed the data collection instruments, and coordinated and supervised data collection for the entire study period, critically reviewed the manuscript, and approved the final manuscript as submitted. Courtney McCracken assisted in the study methodology, carried out the initial statistical analyses, developed the figures and tables, reviewed and revised the manuscript, and has approved the final manuscript as submitted. Pradip Kamat, James D. Fortenberry carried out the initial analyses, reviewed and revised the manuscript and figures, and approved the final manuscript as submitted.

Acknowledgements We are grateful for the contributions of the Egleston PICU nurses, for their willingness to participate and embrace this training culture.

References Andreatta P, Chen Y, Marsh M, Cho K. Simulation-based training improves applied clinical placement of ultrasound-guided PICCs. Support Care Cancer 2011;19(4):539—43. Barsuk J, McGaghie W, Cohen E, Balachandran J, Wayne D. Use of simulation-based mastery learning to improve the quality of central venous catheter placement in a medical intensive care unit. J Hosp Med 2009;4(7):397—403. Beaubien J, Baker D. The use of simulation for training teamwork skills in health care: how low can you go? Qual Saf Health Care 2004;13(Suppl. 1):i51—6. Britt R, Novosel T, Britt L, Sullivan M. The impact of central line simulation before the ICU experience. Am J Surg 2009;197(4):533—6. Burke J. Infection control: a problem for patient safety. N Engl J Med 2003;348(7):651—6. Dominguez T, Chalom R, Costarino A. The impact of adverse patient occurrences on hospital costs in the pediatric intensive care unit. Crit Care Med 2001;29(1):169—74. Fanning R, Gaba D. The role of debriefing in simulation-based learning. Simul Healthc 2007;2(2):115—25. Gaba D. The future vision of simulation in healthcare. Simul Healthc 2007;2(2):126—35. Grant D, Marriage S. Training using medical simulation. Arch Dis Child 2012;97(3):255—9. Grantcharov T, Kristiansen V, Bendix J, Bardram L, Rosenberg J, Funch-Jensen P. Randomized clinical trial of virtual reality simulation for laparoscopic skills training. Br J Surg 2004;91(2):146—50. Haskvitz L, Koop EC. Students struggling in clinical? A new role for the patient simulator. J Nurs Educ 2004;43(4):181—4. Long K. Preparing nurses for the 21st century: reenvisioning nursing education and practice. J Prof Nurs 2004;20(2):82—8. Malpiedi P, Peterson K, Soe M, Edwards J, Scott R, Wise M, et al. National and state healthcare-associated infection standardized infection ratio report; 2012. Margolin F, Robinson W, D’Andrea G, Doyle A. Getting to zero: reducing rate of CLABSI in community based hospitals; 2011, http:// www.leapfroggroup.org/media/file/Final GettingToZero.pdf [accessed 17.03.14].

50 Miller M, Griswold M, Harris J, Yenokyan G, Huskins W, Moss M, et al. Decreasing PICU catheter-associated bloodstream infections: NACHRI’s quality transformation efforts. Pediatrics 2010;125(2):206—13. Moureau N, Lamperti M, Kelly L, Dawson R, Elbarbary M, van Boxtel A, et al. Evidence-based consensus on the insertion of central venous access devices: definition of minimal requirements for training. Br J Anaesth 2013;110(3):347—56. Nowak J, Brilli R, Lake M, Sparling K, Butcher J, Schulte M, et al. Reducing catheter-associated bloodstream infections in the pediatric intensive care unit: business case for quality improvement. Pediatr Crit Care Med 2010;11(5):579—87. O’Grady N, Alexander M, Dellinger E, Gerberding J, Heard S, Maki D, et al. Guidelines for the prevention of intravascular catheter-related infections. Infect Control Hosp Epidemiol 2002;23(12):759—69. Pittet D, Tarara D, Wenzel R. Nosocomial bloodstream infection in critically ill patients: excess length of stay, extra costs, and attributable mortality. J Am Med Assoc 1994;271(20):1598—601, 25.

K.B. Hebbar et al. Scholtz A, Monachino A, Nishisaki A, Nadkarni V, Lengetti E. Central venous catheter dress rehearsals: translating simulation training to patient care and outcomes. Simul Healthc 2013;8(5):341—9. Siempos I, Kopterides P, Tsangaris I, Dimopoulou I, Armaganidis A. Impact of catheter-related bloodstream infections on the mortality of critically ill patients: a meta-analysis. Crit Care Med 2009;37(7):2283—9. Skelton M, McSwain N. A study of cognitive and technical skill deterioration among trained paramedics. J Am Coll Emerg Phys 1977;6(10):436—8. Tuttle R, Cohen M, Augustine A. Utilizing simulation technology for competency skills assessment and a comparison of traditional methods of training to simulation-based training. Respir Care 2007;52(3):263—70. Wylie M, Graham D, Potter-Bynoe G, Kleinman M, Randolph A, Costello J, et al. Risk factors for central line-associated bloodstream infection in pediatric intensive care units. Infect Control Hosp Epidemiol 2010;31(10):1049—56. Zautcke J, Lee R, Ethington N. Paramedic skill decay. J Emerg Med 1987;5(6):505—12.