Improving Home Caregiver Independence With Central Line Care for Pediatric Cancer Patients

The intervention was carried out in a tertiary care, university-affiliated POSCT center from April 2016 to June 2017. We focused on external CLs: tunneled external catheters and peripherally inserted central catheters. External CLs require daily maintenance and are associated with higher CLABSI rates.^6,7,9,32  Our center manages approximately 50 unique POSCT patients with external CLs each month. Before our intervention, no formal caregiver CL training curriculum existed. Training documentation was inconsistent, nonstandardized, and based on a subjective evaluation. Inpatient nurses did caregiver CL care training before the hospital discharge associated with CL insertion. Caregivers observed nurses then performed hands-on demonstrations adhering to the bundle from established recommendations.^6,33,34  After initial teaching, no follow-up hands-on training or evaluation of the caregiver’s CL care proficiency was performed. There was no formal training for patients never admitted to our hospital.

A family-centered CL care skill learning curriculum and teach-back program to achieve caregiver CL care independence was codesigned with caregivers. The teach-back consisted of a patient or caregiver performing a return demonstration of CL care with an expert nurse coach providing feedback. We performed a current state analysis (CSA) using the capability, opportunity, and motivation framework to identify drivers of behavior change and design the intervention.³⁵  We used: (1) input from caregivers members of a CLABSI prevention committee, (2) short surveys of 79 POSCT caregivers responsible for home external CLs, and (3) informal feedback from a pilot of teach-backs. A working group focused on the family-centered QI intervention. POSCT physicians with CLABSI prevention and QI or safety expertise, an ambulatory and inpatient nurse champion, and a population health manager met weekly and engaged in bidirectional feedback at least monthly with an existing CLABSI prevention multidisciplinary committee. This committee also included infection control specialists and pediatric patient and family advisory council members. The CSA showed that caregivers were motivated to learn CL care and were knowledgeable. Less than one-half of caregivers felt comfortable with CL care even after months of experience. Caregivers wanted more and earlier hands-on opportunities, the most valued strategy for learning CL care. They desired standardized training to minimize confusion. Caregivers were interested in participating in clinic teach-backs embedded into routine visits.

We piloted incorporating teach-backs into patient’s routine care from May 2015 to March 2016 before launching the intervention. During the pilot, clinic nurses approached caregivers for participation without previous identification or scheduling. No formal structure was in place for teach-backs to occur. The percentage of teach-backs performed was tracked. Feedback was collected to understand implementation barriers. During the pilot, 13% (22 of 170) of eligible patients had a caregiver complete a teach-back. Eight had documentation of independence with CL care. Thirty percent of caregivers refused participation in the pilot. Informal feedback revealed that caregivers viewed pilot teach-backs as a “test of their ability to perform CL care,” increasing anxiety.

A key driver diagram (Fig 1) was developed. This was iteratively revised and refined through ongoing feedback from key stakeholders (primarily caregivers and nurses).

Plan-do-study-act rapid cycles were used to implement change beginning April 2016. We developed a standardized curriculum for caregiver training beginning soon after CL insertion, much earlier than the preexisting process closer to hospital discharge. Starting May 2016, caregivers watched training videos followed by hands-on practice in the hospital after observing nurses demonstrate the steps. The curriculum included a new component, an ambulatory teach-back program in clinic to consolidate the initial in-hospital training. Caregivers were introduced to the teach-back program through a “Ready for Home Kit” during discharge education. The kit contained cleaning supplies, a CL care learning aid tool, a description of the program with an expectation to participate, and a photograph of the ambulatory CL champion. Teach-backs were intended to begin at the first clinic visit after discharge and continue during subsequent visits until reaching independence. Changes included: (1) process mapping and redesign of clinic workflow to embed planned teach-backs into routine appointments (Fig 2), (2) creation of a tool to guide nurse trainers and standardize proficiency evaluation (August 2016), (3) simplification and standardization of proficiency documentation in the medical record, (4) and culture changes, such as iterative language revisions to increase partnership and participation, incentivizing nursing participation such as with a coffee gift card, and celebrating successes.

A health care communications and checklist design expert with experience in care delivery improvement helped develop 2 caregiver CL care learning tools. Patients or caregivers tested tools before production. These were developed in English, Spanish, and Arabic. Nurse trainers used tools and distributed for home use beginning December 2016.

A population health manager proactively identified and tracked eligible patients and caregiver proficiency, scheduled teach-backs, managed and analyzed data, and coordinated the project. As of April 2016, 2 nurses with expertise in POSCT care and education served as champions (inpatient and ambulatory). Champions had dedicated time to build culture, train nurses, and perform teach-backs, especially in the ambulatory setting. We developed an expectation to document in the medical record all CL care training using a standardized proficiency assessment. In person-translators assisted during CL care training for non-English speaking caregivers.

Patient or caregiver direct feedback drove change during project implementation. Feedback helped incorporate teach-backs into routine visits and develop language to increase teach-back acceptability (Supplemental Table 3). New language focused on teach-backs as an opportunity to trouble shoot difficulties while performing CL care in the home. Caregivers tested multiple prototypes of tools and provided active feedback in the design, language, and format.

We piloted 3 additional interventions to support the home environment for CL care starting September 2016: (a) outreach education to home nursing agencies to encourage standardized training; (b) incorporation of teach-backs into an existing home program supporting postdischarge medication reconciliation; and (c) home teach-back visits.

Patients with an external CL and at least 1 visit to our ambulatory clinic were eligible, identified from an existing automated CL days report. Patients coming only for a second opinion were excluded in the analysis, although they were still offered a teach-back. End-of-life patients were excluded at the discretion of the primary oncology team. The program focused on achieving independence with CL flushing, because this is a daily task. Training was offered for the more complex, less frequent tasks of changing the needleless connector and dressing, but these were not the program focus. Caregiver participation and proficiency was tracked. Caregiver CL care proficiency was evaluated on the basis of the number of prompts a nurse provided to the caregiver using the standardized tool developed: independent (0), requiring minimal assist (1), moderate assist (2–3), or unable to do (≥4). Prompts were identified either through nurse observation of an incorrect or missing step or if the caregiver asked for assistance during a step.

Caregivers of a child with a CL inserted before the intervention were also approached, regardless of CL experience. If >1 caregiver was responsible for CL care at home, at least 1 was expected to achieve independence. Attempts were made to involve all household caregivers. Patients who routinely cared for their own CL participated, along with their main caregiver who was also expected to participate.

The primary outcome measure was the proportion of eligible patients with external CLs whose caregiver had achieved independence (no prompts) in CL flushing. The proportion of eligible patients participating in the ambulatory teach-back program was the process measure, tied to the driver of increasing training opportunities. We tracked which nurse trainer performed the teach-back (ambulatory nurse champion versus clinic nurse) as another process measure. We aimed to have ≥ 25% performed by clinic nurses, indirectly assessing that teach-backs still occurred on days the champion was not present. This measure reflected the driver of engaging nurse trainers toward program sustainability. Data were analyzed and reported monthly to drive plan-do-study-act cycles.

Surveys of clinic staff (physician, advance practitioners, and nurses) a month and year after the intervention launch served as a balance measure (Supplemental Table 4). CLABSI rates were tracked starting January 2015 as a secondary outcome measure. Limited patient demographics were collected during the QI intervention period.

Descriptive statistics were used to characterize the patient population, participation in the teach-back program, proficiency with CL care, and survey responses. To assess change over time, outcome and process measures were tracked monthly by using statistical process control charts (p-chart, 3 σ). Measurement began with the pilot and continued postintervention through June 2021 to assess program sustainability. All statistical analysis was performed by using SAS version 9.4 (Cary, NC). Control charts were created by using SQCpack 7 version 7.0.20218.1. This project was deemed QI and did not require a formal review by the Harvard Cancer Center institutional review board. SQUIRE 2.0 guidelines were used to structure the manuscript.³⁶ 

Most participating patients (n = 181 of 205) had a tunneled external catheter and a caregiver whose primary language was English (Table 1). Forty five percent of patients had undergone a stem cell transplant (SCT). Ten percent experienced a CLABSI during the intervention period. The median patient age on CL insertion was 6.6 years (interquartile range [IQR] 2.8–12.4); median total time with the external CL (until end of QI period for patients still with a CL) was 5.3 months (IQR 3.0–8.5); 33% had a CL on start of the intervention.

Independence with CL care increased from a baseline (pilot) of <10% of eligible caregivers to >90% after 6 months of initiation of QI interventions (Fig 3). Overall, 83% of eligible (94% of participating, [n = 170]) reached independence with CL flushing throughout the intervention. Five (2%) required minimal assistance; 6 (3%) had no documentation of proficiency.

At baseline (pilot), <20% of caregivers received a teach-back. Within 3 months of the intervention launch, >90% of eligible patients were consistently reached to participate in the teach-back program (Fig 4). Overall, 88% of eligible patients participated in the program during the QI intervention; 1.5% refused participation. Five percent of eligible patients did not participate because their CLs were removed within 90 days of program initiation. Most other refusals were associated with changes in patient status (ie, cancer progression). Clinic nurses performed 39% of teach-backs; the champion performed 61%.

Rates of caregiver participation and independent proficiency were sustained at >90% until January 2020 (Figs 3 and 4).

The median number of teach-backs per patient was 1 (IQR 1–2, range 1–17) (Table 2). The median number of teach-backs to reach independence was 1 (IQR 1–1, range 1–6); 22 (12%) caregivers required >1 teach-back to achieve independence.

The CLABSI rate remained unchanged (0.30 per 1000 CL days) during the QI period. Before the intervention, SCT patients experienced the greatest number of ambulatory CLABSIs by disease program per year (40%). A year after program start, SCT patients experienced less CLABSIs (n = 4 in 2017) than previous years (n = 12 in 2015 and 2016).

Providing CL care with perfect compliance to best practice is an intricate task that requires ongoing training to achieve comfort and proficiency.^5,8,11,33,37  Little is known on how to best support nonmedically trained caregivers to achieve CL care independence and promote bundle adherence in the home. Incorporating caregivers into improvement effort design is key to maximizing their skill development to provide safe care. We codesigned a clinic-based teach-back program for CL care with caregivers of ambulatory POSCT patients, resulting in a sustainable >90% independent proficiency within 1 year. To our knowledge, this is the first intervention designed with caregivers to achieve independence with a complex medical task at home.

We focused on a clinic-based standardized teach-back program for several reasons. First, Rinke et al found that most parents look to nurses when modeling CL care and prefer hands-on training.⁵  Although hands-on practice and simulation have become part of routine training in medical skill development for health care professionals,^38–40  application of this approach for caregivers in the home is limited.²⁵  Second, it is currently recommended that caregivers stay in the hospital for days to receive training on medical tasks performed at home with a periodic assessment of skill and comfort.²²  Yet, extra days in the hospital are costly and may add further burden to caregivers.

We successfully codesigned a QI intervention by working with caregivers as active and passive participants. Patient-family representatives were part of our working group. Many provided direct feedback and participated in surveys to drive change. For example, feedback was used to develop new language to prevent caregivers from feeling as if they were being tested during teach-backs, a contributor of participation refusals during the pilot. The low pilot participation rate and lack of engagement by nurse trainers was also attributed to the lack of a standardized process in place, adding extra time or an additional visit to perform a teach-back. As a result, we embedded our standardized program into routine clinic visits.

The CSA was crucial in the codesign of the intervention to understand that providing more information was unlikely to provide added benefit to caregivers. Instead, hands-on training was more desirable. It allowed an understanding of how best to reach caregivers by early identification of patients and embedding teach-backs into routine visits. Before the QI intervention, a process to accomplish a clinic teach-back did not exist. The pilot period was cumbersome, demonstrated by a low participation rate. On starting the intervention, we needed to “catch-up” all patients with an external CL regardless of their caregivers’ CL care experience. This helped ensure that CL care in the home was standardized. The “catch-up phase” resulted in lower participation rates during the first months of the intervention. Once the program was well established, we saw a sustained >90% participation rate, with almost one-half of patients participating within the first 2 visits. This demonstrates the practical application of QI: the importance of understanding the current state; identifying and engaging key stakeholders to develop drivers and test change ideas; standardization of practice; the importance of measurement; and implementation of rapid cycle changes using a process improvement methodology.

Integral to our success were a standardized tool for training and evaluation of proficiency, a process to track eligible patients, and dedicated resources with specific roles. The nurse champions were agents of change: building culture, developing the program, training peers, and performing teach-backs in the inpatient and ambulatory setting. The manager role allowed continuous information analysis that drove change.

Our analysis has several limitations. There are no validated tools available to assess caregivers’ proficiency in performing CL care, which introduces bias. To minimize bias, we developed a standardized process to evaluate proficiency on the basis of a best-practice care bundle.³⁴  We also had 1 nurse perform most teach-backs. We could not determine whether caregivers would have reached independence without the teach-back program because no standardized processes to evaluate and document proficiency previously existed. We also modified the existing inpatient training to begin earlier and follow a standardized curriculum. Yet, during the CSA caregivers expressed wanting additional hands-on opportunities and at least 10% of caregivers required >1 teach-back to reach independence. This suggests that our intervention was integral in CL care skill development. We did not analyze factors associated with caregiver learning and reaching proficiency, which would provide insight into designing future interventions.

As with most QI interventions, we cannot associate unique tests of change with success in the intervention as many were implemented concurrently. Although not our primary outcome, trends toward improvement in CLABSI rates were observed specifically in the SCT population who had a decrease in the total number of ambulatory CLABSIs postintervention. SCT patients comprised the largest disease group participating in the program, suggesting that this improvement could be related to the intervention. Given the small numbers and data from previous efforts suggesting a lag between intervention and improvement in CLABSI efforts,^8,33,34  we did not expect to see a change in CLABSI rates during the study period.

Our work demonstrates that a successful intervention can be codesigned with caregivers to achieve independence in a medically complex task performed at home. This has significant implications for other populations who care for CLs at home, such as patients receiving parenteral nutrition. It also provides an example of the application of a model to train and achieve caregiver skill independence with complex tasks at home associated with a high error rate, such as medication administration.^41–43  Further analysis of the factors associated with caregiver learning and reaching proficiency in medical tasks is needed to scale and spread this work. The teach-back program has become the standard of care at our institution, supported by the introduction of an ambulatory nurse educator in our clinic and the expectation that all patients will have a caregiver participate in the teach-back program. We will continue to evaluate the impact of our intervention on ambulatory CLABSI rates and the caregiver CL care experience.

We thank Catriona Wagner, who provided medical writing assistance, and Kelly Eng, project manager. Our work would not have been possible without their assistance, and we are very grateful for their contributions.

CL

central line

CSA

current state analysis

CLABSI

central-line associated bloodstream infection

IQR

interquartile range

POSCT

pediatric oncology and stem cell transplant

QI

quality improvement

SCT

stem cell transplant