Ultrasound-guided Peripheral Intravenous Catheter Insertion Technique

This study is a randomized prospective trial comparing US-guided PIVC insertion with standard PIVC insertion technique. This study will evaluate the predictive strength of catheter-to-vein ratio to PIVC post insertion success.

The study population is all patients presenting to the ED. The sample population will be patients requiring a PIVC for treatment or diagnostic purposes and don't have an existing PIVC (EMS or other). Inclusion and exclusion criteria focusing on most suitable participants and safety will determine the participant population. Participants will be randomized to one of two cohorts; 1) PIVC insertion based on standard technique and 2) US-guided PIVC insertion.

There will be two data collection phases in this study:

Phase one occurs in the emergency department (ED). ED staff will be notified of the study period, however to minimize bias, details of the study will not be given. Patients in this phase include those registered in the ED seeking treatment. To identify potential candidates, the investigator will perform an initial screening by communicating with ED staff. Initial screenings will be recorded without using any identifying data. When a potential participant is identified, the investigator will then coordinate with ED staff, approach the candidate and confirm that they meet inclusion and exclusion criteria. If the patient qualifies for enrollment, the investigator will review the study purpose including risks and benefits of enrolling, and have the patient sign the consent form. Randomization will occur after obtaining consent. Patients will have a PIVC inserted by any qualified Nurse or Medic. The PIVC will be placed according to the Reading Hospital Peripheral Intravenous Therapy for Adults guideline. After the PIVC (Standard or US-guided) is inserted, the investigator will then measure the vein prior to any infusion that could distort vein measurement.

All investigators involved with vein measurement will complete vein measurement training. The vein size will be measured using one of the following US machines: SonoSite Edge II or X-Porte each equipped with a 13-6 MHz, 6 cm linear probe (FUJIFILM SonoSite, Inc. Bothell, WA, USA). The investigator will then record collected data on the phase one data collection form and place the form in a secure container. A member of the research team will then collect the completed data collection forms and enter the data in a secured study database. To limit bias, investigators will not inform the ED staff of the specifics of this study and will only provide enough information for data to be collected safely and accurately. In the event that a PIVC could not be successfully inserted according to the randomized technique, the PIVC will be analyzed in an intention to treat (ITT) format and the participant will receive vascular access as per the treatment team.

Phase two of data collection is divided into two parts, and participants in this phase include those that had a PIVC inserted by either randomized method and were subsequently placed in an observation status, inpatient status or discharged home. Phase two part one occurs after the PIVC has been removed. Data collection is performed electronically and may also involve verbal follow up. Trained investigators will obtain an active list of subjects whose PIVC was removed. Each subject will have a blank phase two part one data collection form completed in its entirety when possible. Subjects who remain on an inpatient or observation status will be followed in the electronic health record (EHR). In this study post insertion failure is identified when any of the following reasons for PIVC removal has been entered into the participants EHR: leakage, drainage, phlebitis, painful, occluded, infiltrated, and catheter damage. If more than one of these reasons has been entered, the investigator will further investigate the EHR to determine the primary reason for removal. If electronic data is not sufficient and the subject is still hospitalized, the investigator will attempt to visit the subject for completion of data collection. Completed phase two part one forms will be placed in a secure container until entry into the study database. Following data entry, paper forms involved in this part will be destroyed. Subjects with a PIVC that was removed for a reason not determined as a failure will proceed into phase two part two.

Phase two part two occurs when a PIVC was removed without complication. A trained investigator will obtain a list of subjects whose PIVC was removed without complication. The investigator will determine if 48 hours has passed since the removal of the PIVC. If so, a post removal follow-up interview is performed if the patient is still admitted in the hospital or a phone interview is conducted if the patient was discharged home. Participant charts will also be reviewed for any subsequent ED or primary care physician visits related to PIVC complication(s). The investigator completes the phase two part two data collection form and places it in a secured container. A time limit of 48 hours to contact the participant has been set for this study. A member of the research team will then collect the data collection forms and enter the data in the secured study database.

Statistics and data analysis

Sample size calculation:

The baseline failure rate for post-insertion PIVCs at Reading Hospital is 30%. Our study has set an a priori improvement of an absolute percentage change of 10% as significant. Using a single tail change in improvement (alpha = 0.025) with the intervention with an 80% power, we calculated the number needed for each cohort to be 291 subjects, however we will randomize for 360 in each cohort for a total of 720 subjects to allow for attrition.

Randomization:

Using the statistical package 'R', a series of random numbers will be generated to allocate individual patients to groups. This random number series will be validated for randomness using the 'runs test' available in SPSS. A conservative P value of 0.50 will be used to determine if the sequence of numbers is random. Randomization will continue until a sequence with a p-value greater than 0.50 is obtained. Once completed, the random number sequence will be saved to an Excel file and passed to Reading Hospital researchers for processing.

Data analysis:

Data analysis will occur in two separate phases. First descriptive analysis for all variables for patients within each group. For this analysis discrete variables will be reported as count, and percent within each category. For continuous variables statistics will include mean, minimum and maximum values, standard deviation and median. Significance testing will be conducted between groups. Discrete variables will use chi-square test and continuous variables will use group or independent t-test. An a priori p-value of 0.050 will determine statistical significance between groups.

The second phase of this analysis will include inferential statistics on the comparison between groups for both primary and secondary outcomes.

The primary outcomes to be compared between the PIVC insertion techniques include:

1. Failure rate (overall and individual cause/reason).
2. Utility time
3. Catheter-to-vein ratio

For the analysis of failure rate chi-square analysis will be used between the two groups. Utility time will be analyzed by group t-test. Catheter to vein ratio will be aggregated among all patients within group and a rate analysis defined by Fleiss will be used to determine if the ratios between treatment cohorts are significantly different. A standard p-value of 0.05 will be used for these analyses.

The secondary outcomes will be:

1. Describe post removal complication rates among traditional and US-guided PIVC insertion techniques
2. Identifying a catheter-to-vein ratio that predicts complication

The complication rate will be analyzed by standard rate analysis by Fleiss. Complication rates will be calculated as the number of complication divided by the number of patients.

In order to determine the catheter to vein ratio that may predict a complication, both groups will be added together in aggregate, and a ROC curve will be calculated using the catheter to vein ratio as the independent variable and the complication (0, 1) as the dependent variable. Standard methodology to determine the cutoff point will be used, (diagonal line drawn to ROC curve then projected down to sensitivity value) to determine the cutoff point. Standard ROC descriptive statistics will be used including area under the curve, standard error of the area, 95% confidence interval of the area, p-value as well as sensitivity and specificity of the cutoff value. The null hypothesis for the ROC curve will be area is equal to 0.50.

Due to the exploratory nature of these analysis there will be no corrections for multiple comparisons.