Poke Plan: An Initiative to Improve Distraction and Pain Mitigation With Venous Access in Hospitalized Children

The quality improvement (QI) project was conducted in an academic 144-bed children’s hospital within a hospital in the Midwest that has an average of 6100 children admitted annually. A multidisciplinary QI team was assembled with key stakeholders, including members from pediatric hospital medicine, pediatric critical care, child life services, and inpatient nursing leadership. A core team was formed to consistently attend monthly meetings, monitor the implementation of Plan-Do-Study-Act (PDSA) cycles, and drive the project through each intervention. Baseline data when no interventions were performed were obtained from November 2020 to March 2021. The QI project was implemented from April 2021 through February 2022 and was reviewed for sustainability to evaluate the outcomes and process measures (with no further interventions) from March 2022 to June 2022.

Using a fish bone diagram, the reasons for poor use of pain mitigation were identified. Of those identified, 2 factors that were repeatedly noted were chosen to focus on initially: lack of education and lack of standard pain management guidelines for procedures (Supplemental Fig 5). A key driver diagram was then created to outline the primary drivers that would contribute to achieving our goal of improving pain control (Fig 1). We used the PDSA cycle approach to implement and drive this change into practice.

All patients admitted to the general pediatric, pediatric intermediate care, PICU, and hematology-oncology unit who did not require emergency interventions were included. Emergency interventions were defined as patients who had to be intubated on arrival or escalated to PICU because of severe cardiorespiratory, or airway compromise, or because of behavioral or mental status changes. Additionally, patients having preexisting central or PIV were considered out of the scope of the project.

The hospital analytics team created a report that would provide real-time data each week that included patients who received PIV access after arrival at the hospital. From this report, we used systemic sampling. The data monitoring was intended to be done each week; however, the intervals varied depending on hospital census and acuity, and each observation was not strictly constant. A minimum of 1 patient from each of the 3 general pediatric units (general, pediatric intermediate care, and PICU) were required for each interval. The report acted as a sampling frame from which every fourth patient was pulled. If the patient had a PIV placed in the emergency department or an outside institution, that patient was subsequently replaced by every fifth patient.

We used statistical process control methodology to ensure process stability. The primary outcome measure was the percent utilization of pain mitigation strategies defined as the percentage of topical/systemic analgesic or anxiolytic use on all patients for PIV placement. These included topical lidocaine cream, buffered lidocaine via needle-free injection device (JFNS [J Tip]), oral sucrose and ethyl chloride (topical anesthetic freeze spray), and intranasal midazolam.^15–17  This measure was calculated by dividing the number of patients who received PIV access with analgesics or anxiolytics over each monitoring period the total number of patients who received peripheral access placed during hospitalization.

For secondary analysis, we compared the utilization of different medications and the use of child life services as well as expert resources before and after the project. Because of the intermittent availability of child life services, we could not collect all the process measures with sufficient sample sizes to create control charts so we used pre–post comparisons for this measure. Other outcomes, such as the number of attempts for each PIV placement, could not be feasibly measured because no nursing expectations existed to document that data within our electronic medical record, resulting in no baseline data for comparison.

Initial data analysis revealed that <20% of patients were receiving pain control before a needle insertion for PIV. We therefore set an achievable goal of doubling to 40% over a period of 1 year. The QI team started with biweekly meetings in the first 3 months, which were then spaced to monthly meetings. Six-week-long PDSA cycles were conducted and a total of 8 PDSA cycles were completed using data from the statistical process control charts and staff feedback to guide each cycle (Supplemental Table 2). The process was then monitored for 3 months for sustainability. This project was exempted by the institutional review board. This report is written in accordance with and adheres to the SQUIRE 2 guidelines.¹⁸ 

The initial intervention was the development and dissemination of standard educational content related to pain management in children. Mandatory nursing sessions were provided to all registered nurses by the same nursing educator to ensure consistency in information regarding the importance of controlling procedural pain for children, different pain control options, additional pain mitigation strategies, and the role of parents and child life experts. A live demonstration using different medications via different routes was provided. Each nurse also received a badge card outlining available analgesics and distraction techniques by age (Fig 2).

The second intervention was the creation of a poke plan pathway (Fig 3). The goal of the pathway was to identify patients with the potential for difficult access and early escalation to experts when required. This pathway was posted in all nursing workstations and resident workrooms.

The third intervention was training our “crisis” nurses to perform ultrasound-guided IV insertion. At our institution, crisis nurses are expert ad hoc vascular access nurses who serve in various areas of the hospital in emergency responses and/or when any need arises. This process was created as a stepwise escalation plan to prevent the burden on experts in the PICU. Although it was recognized that ultrasound use itself would not lead to improved analgesia for needle sticks, theoretically, by decreasing the number of attempts, it may lead to less patient discomfort. Additionally, we intended optimal placement in the least number of attempts while using analgesia. The training involved didactic sessions, simulated ultrasound cannulation practice, and supervised ultrasound-guided PIV insertions. Successful credentialing with ultrasound was based on both the trainers and trainees feeling comfortable with the clinical skill with a minimum of 15 successful cannulation attempts to achieve competency.

Our fourth intervention was to create a plan for peripheral access with the patient’s families in collaboration with the child life specialists and the nursing staff (Supplemental Fig 6). On admission, caregivers were requested to fill out a questionnaire about any past experiences with needle-related procedures. They were also asked to provide details on their child’s comfort position preferences, previous IV sites, the child’s support person, and options for coping and pain control. These forms were placed on all pediatric units for prompt completion on admission. Once the plans were codeveloped with families and patients, they were documented in the electronic medical record to guide subsequent procedures during the hospital stay. Additionally, parents were also asked to fill out a survey about their child’s PIV experience during their hospital stay. This survey was posted as a QR code in each room.

Project education targeted residents and faculty in multiple venues including departmental and multispecialty grand rounds, resident meetings, and hospitalists’ division meetings over a period of 6 weeks. This was also incorporated in the monthly inpatient resident orientation for pediatric, medicine-pediatric residents, and emergency medicine residents.

The team created a smart phrase so staff could easily identify patients with difficult access. We used the validated tool, Difficult Intravenous access score.^9,19,20  We also educated the nurses on improved documentation for IV access on the number of cannulation attempts and on distraction techniques.

The seventh intervention was developed because of variable nursing involvement by having weekly competitions among the 3 inpatient units. Each week, the overall percentage of patients receiving child life support, distraction, and pain mitigation on each unit was displayed in a pictorial form and positive patient stories were shared to increase engagement (Supplemental Fig 7).

The eighth intervention instituted was a modification of the order sets within the electronic health record. All existing pediatric admission order sets were updated to have preselected orders for topical lidocaine, sucrose, and JFNS (J Tip). To maintain safety, intranasal midazolam was only available on individual request.

During this period, charts were reviewed with no further interventions initiated. The comfort plan was incorporated into the electronic medical record to avoid paper charting by the child life specialists.

Standard descriptive analysis was performed with calculation of median (interquartile range) and proportion as applicable. A 2-group comparison of categorical variables was performed using a χ² test, with a P value <.05 considered statistically significant. The time series data on percentage use of pain and anxiolysis medications during PIV insertions/interval were charted on a control chart (p chart), using commercially available excel add-in (QI Macros, KnowWare international Inc, Denver, CO). The 1, 2, and 3 Sigma control limits were plotted based on the observed data. Stability of the process was analyzed using QI macro stability analysis, which is based on Nelson’s control chart rules.²¹  Statistical analysis was performed using JMP V 16.0 (SAS Institute, Cary, NC).

Of the total of 6611 admissions on pediatric floors during the study period, 3945 patients received PIV access while hospitalized. Using systematic sampling, 21.9% (883) of patient charts were reviewed (99) for baseline and 784 after intervention/sustainability. Fifty-nine percent of patients were age <5 years and 41% were age >5 years. respectively.

The statistical process control chart for the primary outcome of the use of pain mitigation strategies demonstrated a stable process with 2 centerline shifts. Immediately corresponding to our first intervention, the team saw a centerline shift to 60.3%. After that, the use of medications for PIV trended downward with interval 31 (October 5, 2021–October 11, 2021) below the 3 Sigma limits. Although a definitive cause of this special cause variation could not be identified, the ongoing efforts helped return the system to its previous state. However, once the order set with medications for PIV access was updated in cycle 8, we noticed special cause variation prompting another centerline shift to 78.9% that was sustained for the rest of the project (Fig 4).

Child life experts’ involvement increased from a baseline of 3% to 11.5% (Table 1). The utilization of charge nurses and crisis nurses increased from 3% to 11.7%. Escalation to ICU advanced practice registered nurses and physicians increased from 2% to 5.8%. Of all the patients in the implementation phase, 3.4% of the vascular accesses were performed under ultrasound guidance during the project (Table 1). According to the results of the parental surveys conducted in the study, 80% (38/47) of the parents felt their child was kept comfortable during the procedure. Of these, 74% (35/47) reported some form of pain medication use and 89% (42/47) of all these patients received some type of comforting measure during the procedure.

Through this 12-month QI initiative, we successfully exceeded our target for implementing pain mitigation techniques in children undergoing PIV access. We believe that providing ongoing audits, feedback, and visual displays in a control chart format to the nursing staff helped them become more aware of the process and improve their performance. Although these data were not analyzed at the individual provider level, it has enabled us to shift use of pain control more effectively during IV insertion.

Parents who exhibit fear of needles based on their childhood experiences may inadvertently increase their children’s anxiety by passing on their own fears. Developing a personalized treatment plan for their child was our way to alleviate this concern. Most young children required a multimodal approach for preparation of PIV, which included a combination of topical lidocaine, distraction toys, buzzy, and at times intranasal midazolam as parents provided comfort holds. Adolescents were also offered effective ways to cope with pain based on their pain self-efficacy.²²  During the project, our team frequently affirmed small but significant reminders about each medication whenever an opportunity was identified. Some of these included examples such as application of a “thick layer” of lidocaine to provide adequate topical anesthesia or an approximate “20-minute downtime” after intranasal midazolam spray to achieve optimal anxiolysis. Implementation of educational workshops with child life specialists was an essential step to lay strong groundwork for the project. Focused education was given on the use of appropriate choice of nonmedical language and coping strategies such as engaging in nonprocedural talk, coaching, or distraction.²³  For instance, comfort hold was called “a hug” and needle was phrased as a “small straw” instead of a “shot” or a “pinch.” Use of uninformative reassurance such as “procedure won’t hurt” or “it is ok” was discouraged. Although there is some emerging literature in identifying optimal pain-framing language for children and their parents, there is not enough evidence to identify what language may serve as equally informative and reassuring to them. On reflection, the authors believe that children and parents may have preferred to use an alternate term such as “comfort plan” instead of “poke plan.” However, because the latter term was used throughout the project, we have maintained its usage in the report.

Our project target was attained more rapidly than we had anticipated. Despite achieving our aim of 40% medication usage, several process gaps were identified and the team decided to continue the efforts over subsequent cycles. By addressing these gaps, we were able to reach a level of 78.9%. We had anticipated an increase in the utilization of the first tier of experts (charge nurse or crisis nurses) by the change in the work flow and quicker escalation. However, the second tier including ICU physicians and advanced practice registered nurses also increased, which was attributed to improved documentation at multiple care levels. It could be argued that the increase in pain medications use may also be a mere improvement because of better documentation. However, medications in our institution cannot be administered without an order and nurse charting; thus, we believe this increase is reliable.

During our final intervention (addition of medications to the order set and J tips made available on each unit), a special cause variation was noticed in the sustainability phase. By including prechecked medications, such as J tips, topical lidocaine cream, and sucrose, barriers to immediate access were purposefully reduced. This empowered nurses to more candidly ask for medications anytime during the hospitalization. Because of this change in practice, patients with an urgent need for vascular access (children with diabetic ketoacidosis and patients with acute respiratory failure) seemed to have benefitted from this project as they were provided with appropriate pain relief, which was a rare practice before this initiative. We also noticed an increase in the use of pain mitigation strategies for all kinds of phlebotomy procedures. Although blood draws were not measured for our project, this initiative has changed our institution’s culture toward needle insertions, and the poke plan is now considered best practice in the children’s hospital.

As with other quality improvement projects, there are limitations to this study. Because our goal was to measure the increase the use of pain mitigation medications, we did not measure their actual efficacy in decreasing pain and anxiety during PIV placement. Implementing the project during the pandemic caused multiple unforeseen challenges. Although family representation is instrumental in improving the quality of any process, early COVID-19 pandemic protocols presented barriers to their initial inclusion in planning. Future work will include family input into our quality initiatives. Strict isolation policies and our need to conserve personal protective equipment during the COVID-19 pandemic limited our ability to improve child life utilization to our goal. Additionally, all caregivers were offered the QR code to complete the survey but most opted not to participate, affecting our overall survey numbers. Although we aimed to eliminate inherent bias by sharing the survey with all caregivers, it is possible that only those who had exceptionally positive or negative experiences chose to complete the forms.

Implementation of pain mitigation strategies with distractions and medication usage during PIV access resulted in better preparation for the procedure. Continued efforts to alleviate pain and anxiety during such procedures in hospital settings would foster satisfaction in the health care system and protect children from unnecessary suffering.

The authors acknowledge all the nurses and child life services who contributed to collecting surveys during the quality improvement initiative.