Material and method
The One Million Global Catheters PIVC worldwide prevalence study (4, 12) was initiated by the University of Western Sydney, Australia. The aim was to examine complications and risk factors associated with the use of peripheral venous catheters worldwide. The 419 participating hospitals in 51 countries were sent a standardised data collection form developed at Griffith University, Australia (Appendix 1).
The study was approved by the Griffith University Human Research Ethics Committee (application NRS/34/13/HREC). The project was also evaluated and approved by the Regional Committee for Medical and Health Research Ethics, Central Norway (case no. 2014/951). No directly identifiable data were collected, and each patient provided verbal consent upon inclusion.
Hospitals were recruited through professional networks, health organisations, the Norwegian Ministry of Health, suppliers of intravenous equipment, and social media (4). In Norway, the directors of all regional health trusts and all hospitals were contacted by e-mail. Only St. Olavs Hospital and Levanger Hospital responded. The former is a university hospital with 800 beds and responsibility for 725 600 inhabitants in Central Norway. Levanger Hospital is a local hospital in Trøndelag county with 265 beds and responsibility for approximately 100 000 inhabitants. Data were collected in the form of a prevalence study on an arbitrarily chosen weekday in November 2014 (Levanger Hospital) and February 2015 (St. Olavs Hospital, Trondheim University Hospital). Seven surgical and medical wards (156 beds) participated at St. Olavs Hospital, along with all medical and surgical wards (66 beds and 60 beds, respectively) at Levanger Hospital.
The study population comprised all patients ≥ 18 years of age with peripheral venous catheters who were available on the wards in question on the screening day. Patients who were undergoing diagnostic examinations at the time of screening were excluded.
A total of 13 professional development nursing specialists at the two hospitals performed the catheter assessments. Prior to data collection, the English-language form (Appendix 1) was reviewed by the participating nurses to ensure common understanding of the variables.
The form was completed for each venous catheter and contained: a) background information (hospital and country, ward, prevalence date, patient age and sex), b) date and time of peripheral venous catheter insertion, reason for catheter insertion, catheter brand and size, insertion environment and anatomical placement, (c) clinical observations of the venous catheter (insertion site, condition of dressing, catheter connections), (d) catheter site assessment within the past 24 hours documented in the patient chart, and (e) intravenous fluids and medications administered on the prevalence day. Data were entered into an electronic form.
After data collection, the nurse responsible for the patient was instructed to inspect, remove or replace the venous catheter if it was causing serious clinical problems, in accordance with professional guidelines.
Descriptive statistics are presented in the form of number (n) and percentage (%) for categorical variables and mean (standard deviation, SD) for continuous variables.
A collective variable for phlebitis (yes/no) was defined as the presence of at least one of the following clinical problems: pain/tenderness on palpation, redness >1 cm from insertion site, swelling >1 cm from insertion site, purulence, palpable hard vein cord beyond IV tip, streak/red line along vein, or induration/hardness of tissues >1 cm (5). Yes/no data for 'peripheral venous catheter in use' were generated by combining the variables 'IV fluids today (none)' and 'IV medications today (none)' on the basis of information in the patient's chart/medical records. If the catheter had not been used for either infusions or medications, the variable was assigned the value 'no'.
The number of catheter days was defined as the number of days between catheter insertion and the date of screening. Peripheral venous catheters inserted on the screening day were thus scored as 0 catheter days.
We used a two-level, mixed logistic regression model with phlebitis as the dependent variable, the hypothesis variables (one at a time) as independent variables, and the patient as the random effect. The model took into account the lack of independence between multiple venous catheters in the same patient. A sensitivity analysis was performed and two patients with three peripheral venous catheters were excluded.
The analyses were performed in Stata version 15.1 and SPSS version 25.