Context

A quality improvement study was conducted from January 2016 to December 2019 in a pediatric unit of the "Hospital Infantil De México Federico Gómez." This tertiary care public teaching hospital serves low-income and uninsured populations in Mexico City. This unit comprises 30 beds for patients across various pediatric subspecialties, including gastroenterology (9 beds), cardiology (7 beds), neurology (4 beds), rheumatology (2 beds), pulmonology (2 beds), and general pediatrics (6 beds). This unit was selected for the quality improvement study due to significantly higher CLABSI rates than the rest of the hospital (5.71 CLABSIs/1000 catheter days [CD] vs. 1.92 CLABSIs/1000 CD, respectively).

All patients aged 1 month to 18 years who had a CVC placed for at least 1 day during their stay in the pediatric unit were included. Subjects were prospectively assessed daily to identify those who developed CLABSI. Data collected for each case included sex, age, inpatient days, and CD.

Information on each HCAI was gathered through active epidemiological surveillance, as described by de la Rosa et al.²⁰ Trained nurses visited the unit 3 times a week to identify signs suggestive of CLABSI, with validation from an infectious disease physician and the medical and nursing chiefs of the department of hospital epidemiology (DHE).

During active surveillance, incidental education was provided for asepsis or antisepsis omissions. The preventive measures, or bundles, included catheter replacement for malfunctioning catheters, removal of unnecessary catheters, catheter port cleaning, use of transparent dressings, regular dressing changes, and full-barrier precautions during catheter insertion.

Adherence to this bundle was documented at over 95%. In addition, the hospital maintained a strong and ongoing HH program established in 2013²⁰. All these measures remained unchanged during the intervention.

Interventions

Five nurses trained in hospital epidemiology and infection control, along with two infectious diseases specialists from the DHE, reviewed the free checklist provided on the official website of the ANTT for medication and intravenous (IV) fluid preparation and administration. This checklist was adapted to the hospital’s context and resources. Table 1 contains the adapted version of the checklist. The entire unit nursing staff received theoretical training followed by one-on-one hands-on training for the morning shift nurses. Training began with the morning shift when the largest number of IV medications and fluids were administered. Subsequently, all shifts received training. Practical training included on-site assessments with immediate feedback, continuing until each nurse demonstrated error-free preparation and administration of IV fluids and drugs. Personalized training for all staff lasted approximately 1 month, with the entire training process taking 4 months. After training, a single rotating nurse was assigned to prepare and administer IV solutions and medications, with weekly covert assessments and one-to-one feedback.

Weekly meetings were conducted with the ANTT nurses and DHE staff throughout the intervention. PDSA cycles were executed to address any factors that led to non-adherence, including infrastructure, material availability, and staff-related concerns. Action was taken accordingly to solve problems and barriers. The primary aim was to reduce CLABSI rates through ANTT implementation, as detailed in table 2. Each PDSA cycle measured CLABSI rates and ANTT adherence as outcome measures.

Study of the interventions

To evaluate adherence to the ANTT, monitoring was performed using the checklist in table 1. Each step was scored as completed or not, resulting in a score from 0 to 100% each time the technique was monitored. Immediate feedback was provided to each nurse. Monitoring was solely carried out during the morning shift due to insufficient personnel in the other shifts.

CLABSI definitions from the Centers for Disease Control and Prevention and the National Health-care Safety Network were employed⁴. Monthly CLABSI incidence rates/1000 CD were calculated monthly for each PDSA cycle.

Statistical analysis

Measures of central tendency and dispersion were used for quantitative variables, while frequencies were employed for qualitative data. The Chi-square test for proportions was used to compare ANTT adherence during PDSA cycles, and a comparison of two rates was utilized to assess CLABSI rate differences at the study´s start and end. SPSS version 21 was used for statistical analysis, and the standards for quality improvement reporting excellence 2.0 guidelines were followed.

Ethical considerations

As a quality improvement program, no informed consent or ethics committee approval was necessary, according to hospital regulations. Subject confidentiality was maintained throughout the study.

Summary

This study demonstrates the outcomes of adapting and implementing the ANTT through a quality improvement program with PDSA cycles, which reduced CLABSI rates.

Throughout the PDSA cycles, ANTT adherence improved, variability decreased, and the CLABSI rate was reduced. PDSA cycles were crucial for identifying, addressing, and resolving issues related to the implementation and adoption of the ANTT in our context, from increasing HH moments within the technique due to the resources available within our hospital to creating awareness so that this program could be carried out. These areas of improvement are discussed in depth below.

Interpretation

We consider that the quality improvement program was successful for several reasons:

First, our hospital lacked a standardized technique for medication and solution preparation. Before the study, the same nurse attended to patients with various conditions and prepared their medications. Implementing ANTT with a single nurse responsible for all preparations likely contributed to the reduction in CLABSIs. The eight moments of HH performed during the ANTT steps, the duration of asepsis, and the vascular access management likely influenced the study results.

During the PDSA cycles, we found that theoretical training alone was insufficient for nurses to carry out ANTT medication preparation. We adjusted the strategy to provide practical and individualized training, which was crucial for implementation. Initially, we believed that theoretical training, based on nurses’ experience, would suffice; however, practical training instilled confidence in carrying out the technique. Sonoiki et al. observed that nurses with more experience may feel anxious when learning new techniques with²¹ which they are not familiar. For this reason, the WHO recommends providing training in these new techniques, both theoretical and practical²². The Health and Social Care Act 2012 suggests monitoring adherence to the technique, which we continue to do²³.

Through PDSA cycles, we recognized the importance of monitoring to assess adherence to checklist components. Even after 6 PDSA cycles, we found that material collection and port disinfection had the lowest adherence. All checklist items were dichotomous, so if all materials were not collected or if the port was not disinfected for eight full seconds, the checklist point was not achieved. Similar issues with disinfection time have been observed in studies like ours¹³,¹⁶,²⁴. We need to explore new strategies to improve adherence to these items.

One of the main challenges in implementing improvement strategies is ensuring that the staff understands the rationale behind the changes. Initially, when aiming to reduce the CLABSI rate, the staff perceived it as an increased workload. This issue was identified during a PDSA cycle. To address this and improve adherence to the ANTT, we implemented the "let everyone know about the bacteremia" program, informing all staff when a CLABSI was diagnosed. After the program started, if a new CLABSI occurred, the nurses in the ward became concerned about why it had happened and sought help from the DHE. As described by Gould et al., it is crucial for patient safety that the staffs understand the purpose of implementing ANTT²⁵.

An important finding in the study was reducing the CLABSI rate since the first PDSA cycle. This reduction may have occurred for several reasons. Before implementing the "let everyone know about bacteremia" program, weekly meetings were held in the ward to identify and discuss bacteremia cases, raising awareness of the increasing CLABSI rate in 2015. Simultaneously, preparations for ANTT implementation began, including clinic room adaptation, restricted access, a reinforced HH) program, and bacteremia prevention classes, which may have contributed to the decrease in the CLABSI rate. In addition, we observed that pediatric nursing staff generally respond quickly to changes when they understand the improvements in patient health.

Despite the initial improvement in the CLABSI rate during the first PDSA cycle, we continued with subsequent cycles due to the variability in technique adherence. Specific technique steps needed improvement to increase ANTT adherence and reduce process variability. We did not consider the initial improvement in CLABSI rates to be a success; we were aiming for increased ANTT adherence with low variability and sustained low CLABSI rates, which we achieved after 6 PDSA cycles.

The study limitations primarily stem from not measuring all variables that can increase CLABSI risks, such as total parenteral nutrition (TPN), multi-lumen devices, chemotherapy treatment, immunosuppression, prolonged use of antibiotics, prophylactic antibiotics²⁶-²⁸, untrained rotating personnel, and patient comorbidities. A large percentage of the patients in that unit had gastrointestinal ailments and commonly had TPN and comorbidities. It is unlikely that patients in that unit received chemotherapy treatment, as they are in other hospital units. The rest of the risk factors were not measured because they were beyond the scope of the study, yet the CLABSI rate decreased after ANTT implementation.

Another limitation was the initial 7-day period where ANTT was not followed for medication preparation. This was due to resistance to change and inadequate nursing staff. The nursing staff, with the union, filed complaints as they felt an increased workload. After addressing these issues with the union, they supported the change and helped emphasize its importance for patient care.

Adherence to ANTT was only measured during the morning shift, potentially overestimating it. There was not enough staff to measure the evening and night shifts. However, despite the lack of measurement in other shifts, the CLABSI rate decreased.