Adherence engineering in a central line dressing change

Infections associated with Central Line Dressing Changes (CLDC) represent a significant cost in healthcare and human life, with approximately 250,000 cases a year and costs of up to $29,000 per single episode (OʼGrady et al., 2002; Shannon et al., 2006). The introduction of a kit designed in accordance with adherence engineering (AE) to direct behaviors during a CLDC could result in improved patient safety, increased efficiency, and lower costs of healthcare delivery (Drews, 2013). This kit could also be more effectively utilized by novice users in a home healthcare setting. The present study examined 39 novice participants performing a CLDC, 19 with the AE kit and 20 with a traditional "jumble kit," under the hypothesis that the AE participants would complete more steps of the CLDC correctly, and do so significantly faster than the jumble kit participants. The study also aimed to investigate the relationships between time spent on the education portion and CLDC performance, and use of training materials and performance. While the AE participants did complete the procedure significantly faster, there was no difference found in performance between the two groups, nor was there any correlation found between use of training materials and performance. A moderate (though significant) correlation was found for time spent on education and performance outcome. These results indicate that more research is needed to determine how best to adapt the AE kit for novice users; suggestions for future experiments are discussed.

ADHERENCE ENGINEERING IN A CENTRAL LINE DRESSING CHANGE by Danielle N. Zimmerman A Senior Thesis Submitted to the Faculty of The University of Utah In Partial Fulfillment of the Requirements for the Honors Degree in Bachelor of Science In The Department of Psychology Approved: ____________________ Frank Drews Supervisor ____________________ Carol Sansone Chair, Department of Psychology ____________________ Jeanine Stefanucci Department Honors Advisor ____________________ Dr. Sylvia D. Torti Dean, Honors College May 2014   ABSTRACT Infections associated with Central Line Dressing Changes (CLDC) represent a significant cost in healthcare and human life, with approximately 250,000 cases a year and costs of up to $29,000 per single episode (OʼGrady et al., 2002; Shannon et al., 2006). The introduction of a kit designed in accordance with adherence engineering (AE) to direct behaviors during a CLDC could result in improved patient safety, increased efficiency, and lower costs of healthcare delivery (Drews, 2013). This kit could also be more effectively utilized by novice users in a home healthcare setting. The present study examined 39 novice participants performing a CLDC, 19 with the AE kit and 20 with a traditional “jumble kit,” under the hypothesis that the AE participants would complete more steps of the CLDC correctly, and do so significantly faster than the jumble kit participants. The study also aimed to investigate the relationships between time spent on the education portion and CLDC performance, and use of training materials and performance. While the AE participants did complete the procedure significantly faster, there was no difference found in performance between the two groups, nor was there any correlation found between use of training materials and performance. A moderate (though significant) correlation was found for time spent on education and performance outcome. These results indicate that more research is needed to determine how best to adapt the AE kit for novice users; suggestions for future experiments are discussed. ii     TABLE OF CONTENTS ABSTRACT ii INTRODUCTION 1 METHODS 7 RESULTS 9 DISCUSSION 10 REFERENCES 13 iii     1   INTRODUCTION Health care is a field that requires both stringent protocol adherence, and constant monitoring for maximum efficiency in systems and users. Recently examined means of improving efficiency such as checklists (Drews, 2013) and ease of access to supplies (Boyce & Pittet, 2002) that contribute to the overall understanding of health care effectiveness are not enough in themselves to fully address the issue, particularly for more complex systems. Many recommendations are also designed specifically for hospital and intensive care unit environments (e.g. Shannon et al., 2006; Pronovost et al., 2006; Mermel, 2000; O’Grady et al., 2002), neglecting the needs of inexperienced users in informal home health care settings. Adherence engineering may provide a comprehensive solution for a specific problem within health care – the high infection rate associated with Central Line Dressing Changes (CLDC) – by guiding user behavior and cognition within a system that utilizes multiple devices, and where the need for protocol adherence is crucial (Drews, 2013).   The Dangers of Central Line Dressing Changes A Central Line is a catheter placed through the arm, neck, or groin of an extremely ill patient requiring long term or forceful treatment, in which the smaller veins associated with a traditional IV would be unable to withstand pharmaceutical intervention (such as chemotherapy or aggressive antibiotics) (Drews, 2013). The greatest risk posed by a Central Line is the high risk of infection, which is most often contracted while the dressing for the Central Line is being changed. In the United States, Central Line Associated Bloodstream Infections (CLABSI) occur in approximately 250,000 patients     2   per year, with a mortality rate of nearly 25% (OʼGrady et al., 2002; Center for Disease, 2011). The healthcare costs associated with CLABSI can reach $29,000 per episode (Shannon et al., 2006). The source of most CLABSI is the accumulation of microbes surrounding catheter hubs and on the skin around the insertion site (Mermel, 2000), through which infection can easily spread to the rest of the body, and so these areas are targeted in the present study. The following guidelines from Pronovost, et al. (2006), O’Grady et al. (2002), and Mermel (2000) are also emphasized: -Education about the use of intravascular catheters, procedure for maintenance of such, and seriousness of infection risk (via informational PowerPoint during the educational portion of the experiment) -Use of only upper-body (arm) insertion site (which has lower risk of infection) -Use of Chlorahexadine as an antiseptic (found by Mermel to be the most efficient choice) -Disinfection of catheter hubs -Importance of hand hygiene and clean and/or sterile gloves -Contact with the assess port with only sterile devices Another critical problem associated with Central Line Dressing Changes (CLDC) is the supplies necessary. In hospital settings, nurses must either gather the needed items prior to beginning the procedure (which can be as high as 27, and so forgetting an item is both likely and costly) or rely on commercially available “jumble kits,” which include the most commonly used items in a simple plastic bag (Drews, 2013) (See Figure 1). The lack of structure displayed in jumble kits require sorting out of the items first, as well as     3   reliance on user cognition; there are no active defenses against error (Bakdash & Drews, 2012). A system to address the shortcomings of traditional jumble kits and the risks of infection was modified for the present study. Adherence Engineering Adherence engineering (AE) refers to a conceptual framework put forth by the Center for Human Factors in Patient Safety with the goal of increasing protocol adherence, and working off the belief that some behaviors are externally shaped and so can be positively affected by equipment and systems design, particularly in situations where traditional human factors application is insufficient (Drews, 2013). The seven principles of AE are as follows: -Create object affordances (the environment/object guides the necessary action) -Provide task intrinsic guidance (the task directs its own use) -Implement nudging (toward correct choices) -Select and implement smart defaults (the most practical default values are provided) -Provide feedback -Reduce cognitive effort -Reduce physical effort A system utilizing AE will include all of the above principles, and will likely have a positive impact on behavior. For the present study, a kit designed in accordance with AE to safely and efficiently perform a CLDC was compared to traditional jumble kits. The AE kit utilizes the 27 most commonly used items used in a CLDC and breaks the     4   procedure down into a 26 step process (42 with substeps), guiding the user through a labeled set of two plastic packets. The packets, contained within one plastic bag the same size as those associated with jumble kits, unroll on any flat surface and contain all the necessary items in individual pouches, with the user working from left to right following the numbered instructions above each pouch (See Figures 2 & 3). With one packet for use during the non-sterile steps of a CLDC and one for the sterile steps, sterility is less easily violated. The kit is also single-use, so used items are thrown away rather than retained and used again (violating sterility). Home Health Care and Novice Users User feedback from assessing the AE kit in Drews (2013) proved to be extremely positive, yet this population was limited to nurses. For the present study, only novice users, with no prior medical training or knowledge of central lines were included, with the goal of targeting informal caregivers in home health care. Informal caregivers refer to non-paid friends and family members who typically do not have any health care training, yet perform medically oriented tasks. Estimates place this number at about 15% of the American population (Committee on the Role of Human Factors, 2011). Swedberg, Chiriac, Törnkvist, and Hylander (2013) identified a competence gap, defined as professed lack of proficiency in expected tasks, as the largest barrier preventing health care assistants with no training or experience from providing adequate health care. Informal caregivers are also more likely to misperceive info, make incorrect judgments, respond incorrectly, and not understand the implications of the information they do receive (Committee on the Role of Human Factors).     5   Naturally, special considerations must then be made when tailoring tasks and systems to this population, which typically has minimal preparation and training (one in three informal caregivers who assist with dressing changes receives no medical training for the task at all; Committee on the Role of Human Factors, 2011). Layout of the AE kit and the included supplies were not altered, but changes were made to the labeling of the kit to better suit the target audience. Icons, while initially included to provide visual cues and reduce cognitive effort (Drews, 2013), were deemed either too complex or unnecessary, and thus not useful. Whenever possible, language was simplified and streamlined, and the amount of information given per label was kept at a minimum (Tables 1 & 2 show finished products). Changes were also made to the educational portion. The informational PowerPoint was reduced in slide number, and slides were simplified when possible. The importance of sterility and the risk of infection were highly emphasized with statistics, greater detail on how to maintain sterility, and an informational video demonstrating how to put on sterile gloves. Present Study For the present study, the above modifications were made to the AE kit with the purpose of addressing the needs of novice users, while keeping in line with the principles of AE to ensure the kit is as intuitive for the user and effective at completing the task (i.e. a CLDC) as possible. The primary question of interest was then whether participants using the AE kit would in fact perform more steps of the CLDC correctly (suggesting greater efficiency) than participants using the jumble kit. The following predictions were also made:     6   1. AE participants will perform the CLDC more quickly than the jumble participants (further suggesting ease of use of the kit). 2. There will be a correlation for participants of both conditions between time spent on the education portion and CLDC performance (demonstrating that novice users can improve their performance in healthcare related tasks through self-education, lessening the competence gap; Committee on the Role of Human Factors, 2011). 3. There will be a correlation for participants of both conditions between use of training materials (optional, as discussed in the next section) and CLDC performance above 85% correct (demonstrating again the effects of self-education, and of simulation; Orledge, Phillips, Murray, & Lerant, 2012).     7   METHOD Participants Participants were 40 undergraduate students recruited using an online Psychology participant pool in exchange for course credit. Only individuals with no medical training of any kind were eligible for participation. One participant was excluded from data analysis due to receiving a faulty AE kit and being unable to perform the procedure correctly. Procedure The first part of the experiment was the educational portion. After filling out a brief questionnaire to verify their lack of medical training and knowledge about Central Lines, participants read through an informational PowerPoint explaining how to perform a CLDC on their own. They were allowed as much time to study the PowerPoint as they desired, ranging from 13 to 57 minutes, with an average of 28.56 minutes. Several items used during the procedure were provided for participants to acquaint themselves with during this portion. These included a catheter line, a Needleless Injection Site cap (NIS), a syringe (filled with sterile saline solution), a Site-Scrub™, a Dual-Cap™, a WingGuard®, a Chlorhexidine Gluconate scrub (CHG), a BioPatch®, and a dressing (specific items varied slightly between participants based on supplies availability). Participants were not required, but encouraged to use these items to follow along with the PowerPoint instructions. Of the 35 participants for whom these data were available, 24 utilized the training supplies. For the second part of the experiment, participants performed a CLDC, using either the AE kit or the jumble kit (conditions were alternated to attempt to provide an     8   even number of each). The procedure was performed on an artificial arm used for medical training and filmed for coding. Participants in the jumble kit condition were given a two-page set of instructions (based on information from the PowerPoint) to use during the procedure to keep the amount of information provided equal to the AE kit’s instructive labels. Coding Videos encompassed the entire CLDC procedure, ending when the participant declared himself/herself to be finished. Videos were coded via checklist, in which the coder checked whether or not each step was performed and whether it was performed correctly (Figure 4). With substeps, the total amount of steps was 42.     9   RESULTS Each of the 156 coding checklists (four coders were used for each of the 39 participants) was entered into Excel, with one spreadsheet per participant. For each coding checklist, a 1 was entered to designate a step was completed and a blank designated the step was not completed, for a total of the 168 cells for the procedure for each participant (with one cell per step of the 42 steps x 4 coding checklists). The percentage of steps performed correctly per participant was determined by taking the mean of the four coding sheets for each of the 42 steps. Overall, jumble kit participants performed 69.25% of the steps correctly, or approximately 29-30 steps. The AE participants performed 71.20% of the steps correctly, also approximately 29-30 steps. The difference was not significant (p = .122). A one directional t-test was used to analyze the data for time spent performing the procedure. AE participants did perform the CLDC significantly faster (26.2 minutes, SD = 5.48 mins) than the jumble participants (29.42 minutes, SD = 4.98 mins), t(38) = 1.94; p = .03). Yet, no significant correlation was found for time spent performing the procedure and percentage of steps correct (Pearson’s r = 0.0637, df =33), making this finding more difficult to interpret. However, time spent on the education portion and percentage of steps correctly performed were significantly correlated (Pearson’s r = 0.357, p = 0.03, df = 32). Of the 35 participants for whom information about training supplies use was available, 11 performed at least 80% of the steps correctly. Seven of these participants used the training supplies and four did not, suggesting a slight, though unclear and insignificant advantage of using the supplies on performance outcome.     10   DISCUSSION The primary hypothesis – AE participants’ performance on a CLDC will be significantly better than the jumble participants,’ as demonstrated by number of steps correctly performed – was not supported. This represents a failure to adequately modify the existing AE kit for novice users. In addition, there was no added performance benefit with utilization of the training supplies during the education portion, suggesting that exposure to materials prior to task performance had little to no impact on performance. However, in the present study, participants were not required to use the supplies, and so it is possible that they did not feel motivated to either use them at all (as 11 participants did not) or to utilize them fully as learning tools. Additionally, it should be noted that use of the supplies was determined by whether or not the supplies were opened from their packages, but this did not definitively indicate how the supplies were used by participants. There was a moderate correlation found between time spent on the education portion and performance outcome, with participants in both conditions who spent more time studying the informational PowerPoint performing more steps correctly. This is a promising outcome for informal caregivers, as self-teaching (i.e. teaching themselves to perform a task through a PowerPoint presentation) could provide a low-cost way for this population (which often receives no training at all) to learn about performing a medical procedure at their own pace, on their own time, and even in their own environment (via home computer). Finally, the AE participants were found to perform the CLDC significantly faster than the jumble participants, but there was no correlation between time spent performing     11   the procedure and percentage of steps correct. It is then unclear as to whether or not the AE kit actually aided performance by allowing participants to complete a CLDC in less time. The shortened amount of time to perform the procedure may better demonstrate the AE kit’s ability to reduce cognitive and physical effort (by requiring less time to set up the supplies and plan for the implementation of the task; Drews, 2013). Implications for Adherence Engineering Addressing the needs of users who do not have formal healthcare training in future experiments might be accomplished by conducting usability trials, in which the CLDC is broken down into smaller sections (e.g., sterile and non-sterile) to determine potential problem areas. The large number of individual steps in an entire CLDC makes it difficult to closely examine the procedure, especially among novice users, who may blend together steps or skip them entirely. Despite the simplicity and readability of the labels on the AE kit, participants sometimes chose to ignore them altogether, preferring to rely on their own memory and the layout of the kit and supplies to guide them through the procedure. While this speaks well of the kit’s object affordances and task intrinsic design (Drews, 2013), it also points to a need to develop the labels in the kit further. Future experiments may find it helpful to make the labeling more salient, perhaps by printing in non-white colors or increasing the size of the wording. Implications for Informal Caregivers While acquainting oneself with the items to be used in a task may or may not help improve performance, it appears that spending longer amounts of time learning about the     12   task through self-guided learning can be beneficial for certain outcomes. Although there is likely a limit on how long informal caregivers can spend learning about a task, this is still a hopeful sign that novice users can be adequately taught about important medical procedures they will have to perform through informal means. According to the present study, the area in which informal caregivers have the most trouble performing a CLDC (and perhaps other medical procedures as well) is in maintaining sterility. None of the 39 participants managed to keep the area and supplies (notably their gloves) sterile throughout the second half of the procedure. Common mistakes included opening sterile items (or the sterile package in the AE condition) before sterile gloving and contaminating the items, failure to put on sterile gloves the proper way (which results in their contamination), and dropping or placing sterile items on contaminated surfaces. Because the risk for infection is so high and the consequences of such are so dire, this is without a doubt the area in which future researchers on this subject should direct most of their efforts.     13   REFERENCES Alonso-Echanove, J., Edwards, J. R., Richards, M. J., Brennan, P., Venezia, R. A., Keen, J., et al. (2003). Effect of nurse staffing and antimicrobial-impregnated central venous catheters on the risk for bloodstream infections in intensive care units. Infection Control and Hospital Epidemiology, 24(12), 916-925. Argani, C. H., Eichelberger, M., Deering, S., & Satin, A.J. (2012). The case for simulation as part of a comprehensive patient safety program. American Journal of Obstetrics & Gynecology, 206(6), 451-455. Bakdash, J.Z. and Drews, F. A. (2012). Using knowledge in the world to improve patient safety: Designing affordances in health care equipment to specify a sequential checklist. Human Factors and Ergonomics in Manufacturing & Service Industries, 22, 7-20. Boyce, J. M., & Pittet, D. (2002). Guideline for hand hygiene in health-care settings: recommendations of the Healthcare Infection Control Practices Advisory Committee and the HICPAC/SHEA/APIC/IDSA Hand Hygiene Task Force. Infection control and hospital epidemiology, 23(S12), S3-S40. Center for Disease Control and Prevention (2011). Vital Signs: Central-line associated blood stream infections-United States, 2001, 2008, and 2009. Morbidity and Mortality Weekly Report, 60(08), 243-248 Committee on the Role of Human Factors in Home Health Care. (2011). Health care comes home: The human factors. Washington, DC: National Academies Press. Drews, F.A. (2013). Adherence Engineering: A new approach to increasing adherence     14   to protocols. Ergonomics in Design, 21(4). 19-25. Drews, F.A., Wallace, J., Bernuzillo, J., Markewitz, B., and Samore, M. (2012). Protocol adherence in the Intensive Care Unit. Human Factors and Ergonomics in Manufacturing & Service Industries, 22, 21-31. Gurses, A. P., Seidl, K. L., Vaidya, V., Bochicchio, G., Harris, A. D., Hebden, J., et al. (2008). Systems ambiguity and guideline compliance: a qualitative study of how intensive care units follow evidence-based guidelines to reduce healthcareassociated infections. Qualitative and Safety in Health Care, 17(5), 351-359. Mermel, L. A. (2000). Prevention of intravascular catheter-related infections. Annals Of Internal Medicine, 132(5), 391-402. OʼGrady, N. P., Alexander, M., Dellinger, E. P., Gerberding, J. L., Heard, S. O., Maki, D. G., et al. (2002). Guidelines for the prevention of intravascular catheter-related infections. Infection Control and Hospital Epidemiology, 23(12), 759-769. Orledge, J., Phillips, W. J., Murray, W. B., & Lerant, A. (2012). The use of simulation in healthcare: From systems issues, to team building, to task training, to education and high stakes examinations. Current Opinion in Critical Care, 18(4), 326-332. Pronovost, P., Needham, D., Berenholtz, S., Sinopli, D., Chu, H., Cosgrove, S., et al. (2006). An intervention to decrease catheter-related bloodstream infections in the ICU. The New England Journal of Medicine, 355(26), 2725-2732. Shannon, R. P., Frndak, D., Grunden, N., Lloyd, J. C., Herbert, C., Patel, B., et al. (2006) Using real-time problem solving to eliminate central line infections. Joint Commission Journal on Quality and Patient Safety, 32, (9) 479-487. Swedberg, L., Chiriac, E., Törnkvist, L., & Hylander, I. (2013). From risky to safer home     15   care: Health care assistants striving to overcome a lack of training, supervision, and support. International Journal Of Qualitative Studies On Health And Well-Being, 8.     16   Figure 1. Jumble kit items Figure 2. Non-sterile AE package, shown unrolled, with instructive labels     17   Figure 3. Sterile AE package, shown unrolled on sterile draping, with instructive labels     18   Table 1 Non-Sterile AE package labels 1. a. Put on mask b. Mask Patient 2. a. Wash/sanitize ands b. Put on exam gloves 3. a. Attach new syringe to each new NIS and push through 3-4 drops liquid; leave syringe attached b. Clamp catheters 4. a. Remove old NIS b. Scrub the HUB w/ SiteScrub for 10 sec; allow to dry for 10 sec 5. a. Attach new NIS/syringe to HUB b. Unclamp catheter; inject 5-10 mL saline from syringe c. Remove syringe only d. Cover NIS w/ light blue DualCap e. Re-clamp catheters 6. a. Remove old dressing w/ alcohol wipe/stick, pulling toward insertion site b. Remove WingGuard and BioPatch, without dislodging catheter c. Remove gloves and continue to Sterile portion     19   Table 2 Sterile AE package labels 1. a. Use non-dominant hand & gauze to stabilize catheter b. Scrub insertion site, and area covered by dressing & catheter w/ CHG c. Scrub 30 sec, dry 30 sec 2. a. Attach new WingGuard to catheter b. Remove adhesive backing & place on patient 3. Apply BioPatch - Printed side visible - Directly over insertion site - Slit parallel to catheter 4. Apply dressing, larger portion first - Catheter should exit gap in dressing - Smaller portion overlapping larger - Cut-outs facing one another, surrounding catheter w/ no skin visible 5. Date and initial dressing 6. Cover catheter & dressing w/ netting     20   Figure 4. Coding checklist   PARTICIPANT: DATE: KIT TYPE RA: AE JUMBLE Task Performed? Y 1.DON MASK - verbal OK 2. MASK PATIENT - verbal 3. SANITIZE HANDS - verbal is OK 4. Don Exam Gloves -Note if sterile used 5a. Attach 1st Syringe to new NIS 5b. Prime 1st NIS - Push 3 or 4 drops out of syringe through cap, verbal is OK 6a. Attach 2nd Syringe to 2nd new NIS 6b. 2nd NIS - Push 3 or 4 drops out of syringe through cap, verbal is OK 7. Clamp Catheters - If already clamped OK, THIS STEP CAN COME BEFORE OR AFTER PRIMING (STEP 5), count as correct 8. 1st catheter: Remove Cap and NIS, scrub the hub w SiteScrub 9a. Screw on syringe and NIS to clean hub 9b. Unclamp catheter 9c. Flush syringe into arm (verbal ok) 9d. Re-clamp catheter 10. 2nd catheter: Remove Cap and NIS, scrub the hub (must use new SiteScrub, otherwise incorrect) 11a. Screw on syringe and NIS to clean hub   N NOTES   21   11b. Unclamp catheter 11c. Flush syringe (verbal ok) 11d. Re-clamp catheter 12. 1st NIS - Remove Syringe and attach light Blue cap to NIS- NIS must stay on catheter. Can be completed immediately after flushing & be considered correct 13. 2nd NIS - Remove Syringe and attach light blue cap to NIS- NIS must stay on catheter 14. Remove old dressing using alcohol Swab Stick, pulling toward insertion site. Catheter MUST NOT be pulled out, please note if this occurs. 15. Remove BioPatch 16. Remove WingGuard 17. Remove old gloves 18. Open Sterile Glove package w/o contaminating 19. Sterilize hands, verbal OK 20. Put gloves on without violating sterility 21. Gloves remain sterile throughout procedure 22a. CHG, in Jumble pack opened appropriately 22b. Squeeze wings so cartridge breaks 22c. Scrub around insertion site and entire area to be covered by dressing 23a. WingGuard (WG) in Jumble pack opened appropriately 23b. Attach to patient 23c. Attach WG to catheter, can be done before attaching to patient     22   24a. Biopatch, in Jumble pack opened appropriately 24b. Blue side up 24c. Directly on top of insertion site 25a. Apply Transparent Dressing: Larger portion must go on 1st, transparent window can not be violated 25b. Larger, transparent portion: covers insertion site, bio patch, WG; catheter exits through gap 25c. Place smaller portion underneath catheter lines, overlapping larger, transparent dressing, no skin visible 26. Date and Initial TOTAL PARTICIPANT KIT TYPE AE RA: JUMBLE