Time to occurrence and predictors of peripheral intravenous cannula-induced complications among pediatrics patients in Debre Birhan City public hospitals, Ethiopia, 2025

Abstract

Background

Peripheral intravenous cannulation is a common but complication-prone procedure in pediatric care. In Ethiopia, however, there is a critical lack of evidence on the timing and predictors of these complications, hindering the development of effective local preventive strategies.

Objective

To assess time to occurrence and predictors of peripheral intravenous cannula -induced complications among pediatric patients in Debre Birhan city public hospitals, Ethiopia, 2025.

Method

An institutional-based prospective cohort study was conducted at Debre Birhan public hospitals from March 1 to May 20, 2025. Participants were selected using a systematic random sampling technique (K = 2). The data were collected with Kobo Toolbox and analyzed using Stata-17. Descriptive statistics and Cox regression were fitted. The Cox proportional hazards assumptions were checked both graphically and statistically. Variables with p < 0.25 in bivariable cox regration were included in the multivariable Cox model. The strength of statistical association was assessed by adjusted hazard ratios and respective 95% confidence intervals.

Result

Of the 584 pediatric patients included, 35.8% (95% CI 32% − 40%) developed peripheral intravenous cannula-induced complications. The median time to develop these complications was 108 h (95% CI: 96–108). Parental education (unable to read and write) (AHR = 2.80, 95% CI: 1.70, 4.90, p < 0.001), no mask use during insertion (AHR = 2.3, 95% CI: 1.70, 3.09, p < 0.001), performing the procedure without insertion set kit (AHR = 4.34, 95% CI: 2.10, 8.93, p < 0.001), right side of cannulation (AHR = 1.96, 95% CI: 1.40, 2.72, p < 0.001), inappropriate cannula dressing (AHR = 1.45, 95% CI: 1.02, 2.37, p = 0.025) and blood administration (AHR = 1.54, 95% CI: 1.09, 2.15, p = 0.014) were predictors of time to occurrence of peripheral intravenous cannula induced complications.

Conclusion

Peripheral intravenous cannula induced complications tend to occur after a relatively prolonged dwell time. Parental education, mask use, insertion set kit, side of cannulation, cannula dressing, and blood administration were predictors for these complications. Therefore, targeted interventions like ensuring mask use, insertion kits, and proper dressing practices may reduce complications and improve pediatric outcomes.

Introduction

A peripheral intravenous cannulation (PIVC) is an insertion of short catheter into the vein in the peripheral areas of the patient [1]. It is one of the most frequently performed invasive procedures in pediatric care worldwide [2, 3]. Despite its essential role in modern healthcare, PIVC use is often accompanied by complications that can lead to adverse outcomes [4]. These complications include phlebitis [5], occlusion, infiltration and extravasation [6], not only prolong hospital stays but also increase healthcare costs and compromise patient safety [7].

Globally, pediatric patients experience a high incidence of PIVC complications [8]. This burden is likely greater in low-resource settings such as Ethiopia, where limited access to quality healthcare, shortages of essential medical supplies, and variable healthcare provider skills exacerbate the problem [9, 10]. In Ethiopia, the incidence exceeds 50% [12, 13], however, the timing and patterns of these complications are not well understood [14]. This knowledge gap hinders the development of targeted interventions, leaving vulnerable pediatric patients; who often have smaller veins, thinner skin, and reduced physiological reserves, at higher risk [15].

Peripheral intravenous cannula (PIVC) use is essential in pediatric care but is frequently associated with complications that can lead to serious adverse outcomes if not properly managed. When a cannula is not removed before/promptly after signs of complications, the risk of severe outcomes such as septicemia, pulmonary embolism, and even mortality significantly increases [16, 17]. To minimize these risks, healthcare institutions in developed countries often practice scheduled PIVC removal every 72–96 h, whereas in Ethiopia, removal is typically based on clinical indications [18–20]. Despite these strategies, the timing of PIVC-induced complications (PIVCCs) remains highly variable, most commonly occurring between 46 and 120 h after insertion [21, 22]. Several factors contribute to this variability, including patient-related characteristics such as age, weight, and hydration status, as well as cannula and procedure-related factors like catheter size and insertion techniques [23]. Furthermore, healthcare provider clinical experience, and the frequency of cannula site monitoring are significant predictors [2, 24]. With increasing efforts to enhance pediatric care in Ethiopia [25, 26], the development of effective, hospital-specific interventions requires local evidence [12]. However, research on the timing and predictors of PIVC complications in children is limited [23]. While one study found a median time of 46 h to PIVC complications, it overlooked critical factors like hand hygiene, disinfection practices, and the use of insertion kits [12]. Adding to the urgency, Debre Birhan city public hospitals serve as major referral facilities for the North Shewa zone, providing care for a high volume of pediatric patients [27]. This lack of context-specific data on the time to occurrence and predictors of PIVC complications hinders the ability to identify high-risk pediatric patients and optimize cannulation practices. Therefore, this study aims to determine the median time to occurrence and predictors of PIVC-induced complications among pediatric patients in Debre Birhan city public hospitals, Ethiopia.

Methods and materials

Study setting, period and design

An institutional-based prospective cohort study was conducted from March 1 to May 20/2025 in the pediatric wards of Debre Birhan city public hospitals. The city has eight health centers and three hospitals (two public and one private). However, only the two public hospitals; Debre Birhan Comprehensive Specialized Hospital and Debre Berhan university Hakim Gizaw Hospital, have pediatric wards (serve for age from newborn to 15 years old). These facilities admit over 6500 pediatric patients annually, nearly all of whom require intravenous (IV) cannulation. Both hospitals serve as main centers for pediatric healthcare in the zone, providing comprehensive inpatient and outpatient pediatric services. Clinically indicated removal of PIVCs is the standard practice in both hospitals. Patients were followed from cannula insertion until a complication occurred, the cannula was removed, the patient was discharged or died, or the 12-week study period ended.

Population and eligibility

The source population was all admitted children (aged 0–15 years) who received a peripheral IV cannula for treatment. From this group, a randomly selected pediatric patients who had the IV placed during the study period were considered for the study. Only the first cannula inserted during a patient’s admission was included in the study; any subsequent cannulations were not considered.

Patients were excluded if they were transferred with a functional IV cannula already in place from another facility, as the timing and procedure of its insertion could not be verified.

Sample size determination

The sample size for this study was calculated using sample size calculation formula for survival analysis [28]. The calculation considers a Z-value of 1.96 (corresponding to a 95% confidence interval, alpha = 0.05), a power of 80%, and the probability of the event (P(m) = 0.564) and the anatomical site of cannula insertion as predictor, with a crude hazard ratio (CHR) of 1.37 [12] were the primary variables requiring the largest sample size and θ denotes the natural logarithm of the hazard ratio (lnHR).

Where: m represents the number of events, P(m) is the probability of the event (PIVC complication), considering a 5% non-response rate, the final sample size is determined to be 590.

Sampling technique and procedure

A systematic random sampling technique was used to select eligible pediatric patients admitted to the pediatric and neonatal unit who required IV cannulation during the study period. Hospital records from the same three-month period in the previous year served as a reference. The total number of pediatric admissions during that period (N = 1,31 0) was divided by the desired sample size (n = 590) to determine the sampling interval (k = 1,310/590 ≈ 2.22, rounded to 2). The first participant was selected from the first two eligible pediatric patients using simple random sampling, resulting in the second patient being chosen as the starting point in both hospitals. Thereafter, every 2nd patient was included in the study. In cases of where a single neonate required two IV lines, one IV line was randomly selected before securing the IV.

Operational definitions

• Time to occurrence of complications: The time interval in hours between the IV cannula insertion and the first occurrence of any complication phlebitis, Infiltration, extravasation, and occlusion until discharge [23, 29].

• Event: The development of PIVC complications before cannula removal [2].

• Censored: Pediatric patients did not experience the event of interest, patients whose cannulas were removed due to a physician’s decision unrelated to complications (e.g., change of treatment plan), accidental removal of the canula, death and children transferred out of the facility with a PIVC in place before completing the follow-up period.

• Dwell time of the PIVC (in hours): calculated as the removal date and time (in hrs) minus the insertion date and time (in hours [30].

• Appropriate cannula site dressing: covering the cannula insertion site by V-shaped through plaster and add another plaster over it, the half edges of the cannulated hand (180 degree) but not tourniquet the site 360-degree rotation [31, 32].

• Insertion set kit for IV cannulation: a pre-packaged/preprepared set of essential medical supplies specifically prepared for peripheral intravenous cannulation [23].

Data collection procedure

The data were collected by using Kobo Toolbox with a structured, interviewer-administered questionnaire for demographic and some of clinical related characteristics. And an observational and chart review checklist to capture some clinical characteristics, drugs, administration, device and insertion related characteristics. The questionnaire and checklist were adapted from previous studies [12, 21, 29, 30, 33–37]. The data collection process was begun with obtaining assent (for age ≥ 12 years) [38, 39] and informed consent from parents or guardians, before the insertion of peripheral IV cannula.

Measurement of outcome variable

The outcome of interest was the time-to-event, defined as the development of the first PIVC-induced complication (phlebitis, infiltration, extravasation, or occlusion) necessitating removal. The measurement of this outcome was conducted through a rigorous process of prospective follow-up and standardized assessment. Each pediatric patient was closely monitored for signs of complications for the entire duration that the IV line was in-situ using an observational checklist. Following catheter insertion, each PIVC site was monitored at regular intervals (at least every 4 h), during any interaction with the IV line (e.g., medication or fluid administration), and in response to any patient or caregiver concerns. The precise time of cannula insertion, the onset of a complication, and catheter removal were meticulously noted. An event was confirmed and its time recorded only when a complication was identified and directly led to the unscheduled removal of the PIVC.

Data quality control

The interviewer-administered questionnaire was first developed in English, then translated into the local language (Amharic), and subsequently back-translated into English by independent translators to ensure consistency and accuracy. Prior to data collection, face validity was evaluated by five experts, including two public health professionals, two MSc holders in Pediatrics and Child Health Nursing, and one pediatrician. Based on their feedback, revisions were made. A pretest was conducted on 5% of the sample size (n = 30) at Debre Berhan university Hakim Gizaw Hospital. Based on the findings, necessary adjustments were made, particularly regarding the order and arrangement of checklist. Additionally, the maximum duration of cannula in-situ was estimated during the pretest phase.

Eight BSc nurses were recruited as data collectors and two MSc holders in Pediatrics and Child Health Nursing as supervisors. The principal investigator provided a two-day intensive training session covering the study protocol, data collection tools, and the standardized classification of complications using case examples and reference photographs. A pre-study assessment of inter-rater reliability demonstrated excellent agreement among data collectors, with a Cohen’s kappa of 0.88 for the classification of complication types. This consistency was further reinforced during patient follow-up through the use of a structured, pre-tested data collection checklist to minimize observer subjectivity. Continuous quality control was maintained as the supervisors routinely spot-checked the data collectors’ assessments, and the principal investigator provided ongoing supervision throughout the entire data collection process.

Data management and analysis

Data were coded, cleaned, and analyzed using Stata version 17. Descriptive statistics were used to check for missing values and to summarize frequencies, proportions, means, medians, and measures of variation. Survival analysis was conducted using life tables and Kaplan-Meier curves to estimate survival probabilities and median survival time; differences between groups were compared with the log-rank test. Multicollinearity among independent variables was assessed using the Variance Inflation Factor (VIF).

The Cox proportional hazards assumptions were checked graphically and statistically. Graphically, the assumption was assessed with log-minus-log survival plots, which showed reasonably parallel lines for categorical predictors, indicating no violation. Statistically, the Schoenfeld residuals test was applied both globally and for each covariate individually. The global test was not statistically significant (χ² = 25.94, p = 0.076), and all individual predictor tests also yielded p-values > 0.05. Together, these results confirm that the proportional hazards assumption was satisfied for the final model.

Bi-variable Cox regression was used to select candidate predictors for the multivariable model based on a p-value ≤ 0.25. Multivariable Cox proportional hazards regression was then performed to identify independent predictors of peripheral intravenous catheter complications at a 95% confidence level. A p-value < 0.05 was used to declare statistical significance.

Ethics approval and consent to participants

Ethical approval for this study was obtained from the Institutional Review Board (IRB) of Asrat Woldeyes Health Science Campus, Debre Berhan University (Ref. number-IRB-01/40/2017). Permission to conduct the study was secured from Debre Birhan city public hospital administrators before data collection commenced. Verbal informed consent was obtained from all participants. The research adhered to the principles of the Declaration of Helsinki, upholding respect for autonomy, beneficence, non-maleficence, and justice.

Results

Sociodemographic characteristics of pediatric patients and their parents

A total of 584 pediatric patients and their parents were participated in this study, yielding a response rate of 98.98%. Half of the participants were neonates 297 (50.9%). Slightly more than half were female 301 (51.5%), and resided in urban areas 309 (52.9%). The predominant parental age group being 30–35 years 304 (52.1%) and median age of 32 years (Interquartile range (IQR): 28–35). More than one third (38.5%) of parents were unable to read and write, and over half were housewives 330 (56.5%) (Table 1).

Table 1.

Socio-demographic characteristics of child -parent pair of pediatric patients at Debre Birhan City public hospitals, Ethiopia, 2025 (N = 584)

Variable	Categories	Status				Total (%)
		Event		Censored
		Frequency	%	Frequency	%	Frequency	%
Child age	Neonate (0–28 days)	105	18.0	192	32.9	297	50.9
	Post-neonatal to 5 yrs	82	14.0	116	19.9	198	33.9
	Above five years	22	3.8	67	11.5	89	15.2
Sex of child	Male	74	12.7	209	35.8	283	48.5
	Female	135	21.3	166	28.4	301	51.5
Residence	Urban	108	18.5	201	34.4	309	52.9
	Rural	101	17.3	174	29.8	275	47.1
Parent relation	Mother	197	33.7	327	56.0	524	89.7
	Father	9	1.5	47	8.0	56	9.6
	Grand parent	3	0.5	1	0.2	4	0.7
Parent age in years	Less than 20	11	1.9	0	0.0	11	1.9
	20–24	21	3.6	28	4.8	49	8.4
	25–29	42	7.2	89	15.2	131	22.4
	30–35	104	17.8	200	34.2	304	52.1
	Above 35	31	5.3	58	9.9	89	15.2
Educational status of parent	Unable to read & writ	109	18.7	118	20.2	227	38.9
	Able to read &write	23	3.9	47	8.0	70	12.0
	Primary	14	2.4	54	9.2	68	11.6
	Secondary	46	7.9	47	8.0	93	15.9
	Collage & above	17	2.9	109	18.7	126	21.6
Occupation of parent	Housewife	140	24.0	190	32.5	330	56.5
	Gov’t employee	38	6.5	116	19.9	154	26.4
	Merchant	25	4.3	42	7.2	67	11.5
	Farmer	3	0.5	10	1.7	13	2.2
	Others *	3	0.5	17	2.9	20	3.4

other* =Daily laborer, IDPs and students

Clinical, cannula insertion and device related characteristics

At admission, 249 (42.6%) of the children were hypothermic and 92 (15.8%) presented with hyperthermia. Most children 515 (88.2%) were well hydrated, and 113 (19.3%) had comorbid conditions. PIVC insertions were primarily performed by nurses 519 (88.9%). For 67.8% of the procedure insertor wore mask. Hand hygiene was inconsistently practiced, 192 (32.9%) used water with soap, 3.3% used water alone, and 206 (35.3%) did not wash hands at all before preforming the procedure (Table 2).

Table 2.

Clinical, device and insertion related characteristics of pediatric patients at Debre Birhan City public hospitals, Ethiopia, 2025 (N = 584)

Variable		Categories	Status				Total (%)
			Event		Censored
			Freq	%	Freq	%	Freq	%
Admission unit		NICU	105	18.0	192	32.9	297	50.9
		General pediatrics	104	17.8	183	31.3	287	49.1
Temperatures @ admission in oc		Normal (36.5–37.5)	65	11.1	178	30.5	243	41.6
		Hypothermia (< 36.5)	90	15.4	159	27.2	249	42.6
		Hyperthermia (> 37.5)	54	9.2	38	6.5	92	15.8
Child hydration		Hydrated	183	31.3	332	56.8	515	88.2
		Dehydrated	26	4.5	43	7.4	69	11.8
Has comorbid conditions		Yes	52	8.9	61	10.4	113	19.3
		No	157	26.9	314	53.8	471	80.7
Cannula insertion done by		Nurse	187	32	332	56.8	519	88.9
		Physician	1	0.2	3	0.5	4	0.7
		Student	21	3.6	40	6.8	61	10.4
clinician hand hygiene before insertion	Yes	Hand rub	57	9.7	110	18.8	167	28.5
		Water with soap	57	9.7	135	23.2	192	32.9
		Water	19	3.3	0	0.0	19	3.3
	No		76	13.0	130	22.3	206	35.3
Insertor wear glove		Yes	196	33.6	375	64.2	571	97.8
		No	13	2.2	0	0.0	13	2.2
Mask wear		Yes	95	16.3	301	51.5	396	67.8
		No	114	19.5	74	12.7	188	32.2
Insertion set pack use		Yes	146	25.0	309	52.9	455	77.9
		No	63	10.8	66	11.3	129	22.1
Solution used for skin preparation		Alcohol	195	33.4	371	63.5	566	96.9
		NS*	11	1.9	4	0.7	15	2.6
		None	3	0.5	0	0.0	3	0.5
Material used for tourniquet		Glove	200	34.2	343	58.7	543	93
		Tourniquet	9	1.5	32	5.5	41	7.0
Vein visibility		Yes	166	28.4	320	54.8	486	83.2
		No	55	9.4	43	7.4	98	16.8

*NS Normal saline

Cannula and drug related characteristics

Regarding the site and side of cannula insertion, the majority of cannulas were inserted into the upper limb, accounting for 428 (73.3%) of cases, and 353 (60.4%) is to the left side of the patient (left limb or the left side of the head for scalp cannulas). The majority of cannulas 576 (98.6%) were secured using non-transparent plasters and for 221 (37.8%) IV securing dressing condition is not appropriate. In terms of cannulation attempts, 399 (68.3%) of patients underwent two or fewer trials. The median number of attempts per person per procedure to secure intravenous access was 2, with an interquartile range (IQR) of 1–3 (Table 3).

Table 3.

Cannula and drug related characteristics of pediatric patients at Debre Birhan City public hospitals, Ethiopia, 2025 (N = 584)

Variable	Categories			Status				Total (%)
				Event		Censored
				Freq	%	Freq	%	Freq		%
Site of cannula insertion	Upper limb			109	18.7	319	54.6	428		73.3
	Lower limb			86	14.7	34	5.8	120		20.5
	Scalp			14	2.4	22	3.8	36		6.2
Side of cannulation	Right side			89	15.2	142	24.3	231		39.6
	Left side			120	20.5	233	39.9	353		60.4
Cannula size	Small (24&26)			154	26.4	356	61.0	510		87.3
	Large (20&22)			55	9.4	19	3.3	74		12.7
Plaster used for dressing	Non-transparent			208	35.6	368	63.0	576		98.6
	Transparent			1	0.2	7	1.0	8		1.4
Total number of trials	Two and less			129	22.1	270	46.2	399		68.3
	Above two			80	13.7	105	18.0	185		31.7
Cannula dressing condition	Appropriate			100	17.1	263	45.0	363		62.2
	Not appropriate			109	18.7	112	19.2	221		37.8
IV board use	Yes			29	5.0	95	16.3	124		21.2
	No			180	30.8	280	47.9	460		78.8
Type of infusion	Continuous			88	21.9	257	63.9	345		85.8
	Intermittent			4	1.0	1	0.2	5		1.2
	Bolus			42	10.4	10	2.5	52		12.9
Administration method	Gravitational			187	32.0	253	43.3	440		75.3
	IV pump			19	3.3	83	14.2	102		17.5
	IV push			3	0.5	39	6.7	42		7.2
Medication	Yes	Antibiotics		153	26.2	290	49.7	443	75.9
		Steroid		30	5.1	24	4.1	54	9.2
		Others*		4	0.7	20	3.4	24	4.1
	No			22	3.8	41	7.0	63		10.8
Blood given	Yes, is via same Iv		Yes	56	9.6	31	5.3	87		14.9
			No	4	0.7	5	0.8	9		1.5
	No			149	25.5	339	58.0	488		83.6

Others*=antiepileptics, diazepam, pethidine

Median survival time and incidence of PIVC complication

The cumulative incidence and incidence rate of peripheral intravenous cannula complication were 35.8% (95% CI 32% − 40%) and 5/1000 persons per hours observation (95% CI 4.44/1000–5.7/1000) respectively. The commonest PIVC complication was occlusion 71 (33.97%) followed by infiltration 65 (31.1%) and phlebitis (30.14%). The overall median time to develop peripheral intravenous cannula induced complication was 108 (95% CI: 96, 108) hours (Fig. 1).

Fig. 1.

Over all Kaplan Meier survival probability curve of PIVC induced complications among pediatric patients at Debre Birhan public hospitals, Amhara, Ethiopia, 2025 (N = 584)

Comparison of survival difference between categories of predictors

The log-rank test, supported by visual assessment of Kaplan-Meier curves, was used to evaluate survival differences between categories of categorical predictors. Significant associations with the development of PIVC complications were identified for sex, IV cannula dressing condition, IV split use, tourniquet material, and mask use during insertion (E.g. Figure 2, p-value < 0.001). No significant differences were observed for admission unit, patient age group, residence, or temperature at admission (Fig. 3, p-value = 0.3992).

Fig. 2.

KM survival difference curve for cannula dressing status among patients with PIVC at Debre Birhan public health hospitals, Amhara, Ethiopia, 2025 (N = 584)

Fig. 3.

KM survival difference curve for residence among pediatric patients with PIVC at Debre Birhan city public hospitals, Amhara, Ethiopia, 2025 (N = 584)

Predictors of time to develop peripheral intravenous cannula induced complication

All potential predictors of time to PIVC-induced complications were first assessed for compliance with the proportional hazards (PH) assumption using both graphical (log-minus-log plots) and statistical (Schoenfeld residuals) methods. Thirteen variables met the PH assumption and demonstrated significance (p < 0.25) in bivariable Cox proportional hazards regression, qualifying them for inclusion in the multivariable cox regration analysis. In the final model, six predictors were significantly associated with time to occurrence of PIVC-induced complications (p < 0.05). These predictors were: parental educational status, use of a face mask during IV cannulation, use of an insertion set kit, side of cannulation, cannula dressing condition, and blood transfusion.

The hazard of developing PIVC-induced complications among pediatric patients whose parents were unable to read and write were 2.8 (AHR = 2.8, 95% CI: 1.7, 4.9) times higher compared to those whose parents had a college and above education. Pediatric patients cannulated without mask use had a 2.3 (AHR = 2.3, 95% CI: 1.7, 3.09) times higher hazard of developing PIVC-induced complications compared to those cannulated with mask use. The hazard of developing PIVC induced complications was 4.34 (AHR = 4.34, 95% CI: 2.1, 8.93) times higher among children cannulated without using an insertion set kit compared to those where an insertion set kit was used. Children with inappropriately dressed cannulas had a 1.45 (AHR = 1.45, 95% CI: 1.02, 2.37) times higher hazard of developing complications compared to those with appropriate dressing (Table 4).

Table 4.

Bi-variable and multi-variable Cox proportional hazards regression analysis for the predictors of PIVCCs among pediatric patients, Debre Birhan public hospitals Ethiopia (N = 584)

Variable	Categories	Status		CHR (Crude hazard ratio	AHR (95%CI)	P-value
		Event	Censored
Sex of child	Male	74	209	1	1
	Female	135	166	1.38 (1.03, 1.84)	1.4 (0.87, 2.30)	0.165
Educational status of parent	Unable to read & wr	109	118	1.85 (1.1, 3.11)	2.8(1.7, 4.9)	< 0.001
	Able to read &write	23	47	2.1 (1.13, 3.97)	1.8 (0.97, 3.6)	0.062
	Primary	14	54	0.9 (0.44, 1.84)	0.82 (0.37, 1.8)	0.642
	Secondary	46	47	2.55 (1.4564.48)	1.7 (0.93, 3.07)	0.081
	College and above	17	109	1	1
Age of parents		-	-	0.96 (0.93,0.98)	0.96 (0.93, 1.01)	0.18
Weight of child		-	-	0.99 (0.97, 1.0)	0.99 (0.96, 1.02)	0.838
Hydration status	Hydrated	183	332	1	1
	Dehydrated	26	43	1.32 (0.87, 2)	1.06 (0.48, 1.56)	0.83
Mask use during canulation	Yes	95	301	1	1
	No	114	74	3.23 (2.45, 4.22)	2.3(1.7, 3.09)	< 0.001
Insertion set kit use	Yes	146	309	1	1
	No	63	66	1.6 (1.18–2.18)	4.34 (2.1, 8.93)	< 0.001
Tourniquet used	Glove	200	343	2.37 (1.20, 4.64)	1.25 (0.85, 1.82)	0.258
	Elastic torniquet	9	32	1	1
Side of cannulation	Right	89	142	1.39 (1.05, 1.83)	1.96 (1.4, 2.72)	< 0.001
	Left	120	233	1	1
Cannula dressing	Appropriate	100	263	1	1
	Not appropriate	109	112	1.58 (1.2, 2.08)	1.45 (1.02, 2.37)	0.025
IV- board split use	Yes	29	95	1
	No	180	280	2.6 (1.74, 3.88)	1.29 (0.83, 2.01)	0.257
IV administration methods	Gravitational/IV	187	253	2.26 (1.4, 3.63)	1.4 (0.96, 3.91)	0.057
	IV pump	19	83	1	1
	IV push only	3	39	0.64(0.19, 2.18)	0.36 (0.1–1.25)	0.11
Blood administered	Yes	60	36	1.578 (1.17, 2.13)	1.54 (1.09, 2.15)	0.014
	No	149	339	1	1

Discussion

The findings of this study reveal critical insights into the incidence, timing and predictors of peripheral intravenous (IV) cannula-induced complications among pediatric patients at Debre Birhan Public Hospitals, Ethiopia. The median survival time for PIVC induced complication among pediatric patients in Debre Birhan public hospitals was 108 h (95% CI: 96–108). Comparatively, it aligns with studies from high-resource settings, such as Australia (96 h) [40], Sweden (96–120 h) [34] and the United States (102 h) [41]. This similarity suggesting that the biological timeline for IV-related complications may be broadly similar across diverse populations. Whereas this median time is much longer than study done in Brazil (48 h) [42], and Debre Tabor, Ethiopia (46 h) [12]. This discrepancy may be attributed to differences in patient age groups, both the above studies conducted among under one age whereas as this study encompassed a broader pediatric population beyond neonates and infants. And also difference in the follow up period may contribute for this discrepancy [12, 42].

The high incidence of PIVC complications in this finding (cumulative incidence (35.8%) and incidence rate; 5 events per 1000 person-hours of observation) underscores the need to identify modifiable factors influencing their timing. The hazard of developing PIVC-induced complications among pediatric patients whose parents were unable to read and write were 2.8 (AHR = 2.8, 95% CI: 1.7, 4.9) times higher compared to those whose parents had a college and above education. This finding aligns with prior research indicating that low parental health literacy is a significant risk factor for adverse medical device-related outcomes in children [43].

The primary justification for this association is that parents with limited literacy often face challenges in comprehending and adhering to post-insertion care instructions. This can lead to improper handling of the PIVC, such as difficulties in maintaining hygiene protocols, recognizing early signs of complications like infiltration or phlebitis, and seeking timely clinical intervention [44–46].

This study found that the omission of mask use during pediatric IV cannulation was associated with a 2.3-fold increased hazard of PIVC-induced complications (AHR = 2.3, 95% CI: 1.7, 3.09). While our data cannot confirm direct droplet contamination, this strong association provides robust empirical support for the 2021 Infusion Therapy Standards of Practice (ITSP) which recommend mask use during all aseptic procedures [47]. The increased risk likely stems from two interrelated factors: the direct potential for respiratory droplet contamination of the sterile field, and the possibility that forgoing a mask is a visible indicator of broader lapses in aseptic technique [48]. The finding has a direct implication for local infection prevention practice. Mask wearing during PIVC insertion should be standardized as non-negotiable component of the procedure bundle, moving beyond a simple recommendation.

The present study found that the hazard of developing PIVCCs was 4.34 (AHR = 4.34, 95% CI: 2.1, 8.93) times higher among children cannulated without using an insertion set kit compared to those where an insertion kit was used. This finding is strongly aligned with global guidelines. This finding aligns with global guidelines, including the 2021 ITSP and the 2024 WHO guidelines, which recommend routine use of insertion kits as part of a bundled approach to prevent PIVC induced complications [23, 41]. These kits typically include antiseptic swabs, sterile drapes, gloves, and securement devices components essential for maintaining aseptic technique and minimizing microbial contamination. Evidence from multiple studies reinforces this practice. A study done in India reported that the use of standardized insertion kits reduced the risk of phlebitis compared to non-standardized approaches [49]. Similarly, other research has shown that pre-packaged insertion kits reduce microbial contamination by ensuring the immediate availability of all necessary aseptic materials in an organized and efficient manner [50].

In this study, the hazard of developing PIVC-induced complications was 1.96 (AHR = 1.96, 95% CI: 1.4, 2.72) times higher among pediatric patients cannulated on the right side (whether it be a limb or the right side of scalp) compared to those cannulated on the left side. This laterality effect is likely related to anatomical and behavioral factors. In right-handed children.

(for older children, approximately 90% of the population), right limbs experience greater mobility, potentially increasing occlusion and mechanical phlebitis risk [3]. In addition, the more frequent use of the right side in daily activities may lead to increased manipulation of the dressing or catheter [51]. Children with inappropriately dressed cannulas had a 1.45 (AHR = 1.45, 95% CI: 1.02, 2.37) times higher hazard of developing PIVC induced complications compared to those with appropriate dressing. This is consistent with previous studies that highlighted the critical role of appropriate cannula dressing in reducing the risk of complications [33, 52]. It may be explained by, poorly dressed cannulas may be more susceptible to external contamination, which could increase the likelihood of infections such as phlebitis. Moreover, improper dressing may contribute to discomfort or irritation, which could lead to inadvertent manipulation of the cannula by the child or caregivers, further increasing the risk of complications [14]. Children who were administered blood had a 1.54 (AHR = 1.54, 95% CI: 1.09, 2.15) times higher hazard of developing PIVC induced complications compared to those who were not administered blood. This finding aligns with other studies that have reported a significant association between blood transfusion and an increased risk of PIVC-related complications [12, 53]. This could be explained by the fact that blood product administration increases the viscosity and osmolarity of the infused fluids and triggers inflammatory responses to blood components and increase risk of complication [54].

While transparent dressings did not emerge as a significant predictor in this finding, their evidence-based benefits warrant discussion given the striking discrepancy between clinical guidelines and observed practice. Transparent dressings are permeable to water, vapor, and oxygen while serving as a barrier to contaminants and external pathogens. They allow for continuous monitoring of the insertion site without the need for removal [23]. A randomized controlled trial found that using transparent dressings to secure PIVCs significantly reduced the incidence of complications. Specifically, the incidence of phlebitis decreased to one-third of the baseline rate when transparent dressings were used [55]. And also, the use of transparent dressing is recommended by both the WHO and England national evidence-based guidelines [23, 56]. However, in the current study, for 576 (98.6%) pediatric patients, transparent dressings were not used, indicating a substantial deviation from the above evidence base recommendations.

The findings of this study offer a practical roadmap for improving pediatric care in resource-limited hospitals. By demonstrating that mask use and standardized insertion kits significantly reduce complications, we highlight that enhancing PIVC safety is not a matter of advanced technology, but of consistent access to low-cost, essential supplies and strict adherence to protocol. Hospitals can operationalize this by prioritizing the reliable procurement of these basic items and formally integrating their use into mandatory, checklist-driven insertion procedures.

Limitation of the study

Several measures were implemented to minimize the Hawthorne effect, including covert observation and integrating data collection into routine workflows. Despite this, the potential for an overestimation of adherence to standards remains and should be considered when interpreting the results. In addition, there was no microbiological confirmation of infection-related complications. This means that while clinical signs of infection were monitored, the specific pathogens responsible were not identified, which could influence the understanding of infection sources and patterns.

Conclusion

. Peripheral intravenous cannula induced complications tend to occur after a relatively prolonged dwell time. Parental education, no mask use during insertion, performing the procedure without insertion set kit, right side of cannulation, inappropriate cannula dressing, and blood administration were associated with short time to complication. Therefore, targeted interventions focusing on ensuring mask use, utilizing insertion kits, and reinforcing proper dressing practices are recommended to potentially reduce complication rates and improve pediatric outcomes. Future longitudinal or interventional studies are recommended to confirm these associations and establish causality.

Abbreviations

AHR

Adjusted hazard ratio

CI

Confidence interval

IV

Intravenous

KM

Kaplan-Meier curves

PIVC

Peripheral Intravenous Cannulation/cannula

Authors’ contributions

TM1 designed the study, performed the statistical analysis, and participated in the writing of the manuscript. ZAG2 and AH2 reviewed the manuscript for intellectual content and provided technical guidance in the statistical analysis. All authors read and approved the final manuscript.

Funding

No funding.

Data availability

Data supporting the conclusions of this article are provided within the article. The raw data supporting the conclusions of this article is available from the corresponding author upon reasonable request.

Declarations

Ethics approval and consent to participants

Ethical approval for this study was obtained from the Institutional Review Board (IRB) of Asrat Woldeyes Health Science Campus, Debre Berhan University (Ref. number-IRB-01/40/2017). Permission to conduct the study was secured from Debre Birhan city public hospital administrators before data collection commenced. Verbal informed consent was obtained from all participants. The research adhered to the principles of the Declaration of Helsinki, upholding respect for autonomy, beneficence, non-maleficence, and justice.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.