Skip Navigation
Skip to contents

ACC : Acute and Critical Care

OPEN ACCESS
SEARCH
Search

Articles

Page Path
HOME > Acute Crit Care > Volume 41(3); 2026 > Article
Original Article
Neurology
Peripheral administration of 11.7% hypertonic saline and phlebitis risk in acute ischemic stroke requiring hyperosmolar therapy in South Korea
Acute and Critical Care 2026;41(3):608-620.
DOI: https://doi.org/10.4266/acc.003328
Published online: August 31, 2026

1Department of Neurology, Chonnam National University Hospital, Gwangju, Korea

2Department of Nursing, Chonnam National University Hospital, Gwangju, Korea

3Department of Neurology, Chonnam National University Medical School, Gwangju, Korea

Corresponding author: Hyunsoo Kim Department of Neurology, Chonnam National University Hospital, Chonnam National University Medical School, 42 Jebong-ro, Dong-gu, Gwangju 61469, Korea Tel: +82-62-220-6171 Fax: +82-62-228-3461 E-mail: kharistma@naver.com
*These authors contributed equally to this work.
• Received: April 6, 2026   • Revised: June 1, 2026   • Accepted: June 12, 2026

© 2026 The Korean Society of Critical Care Medicine

This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/licenses/by-nc/4.0/) which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited.

  • 921 Views
  • 30 Download
prev next
  • Background
    Hypertonic saline (HTS) is used for hyperosmolar therapy in acute ischemic stroke (AIS) but is traditionally administered via central venous catheter to minimize vascular complications. Procedural burden and urgency often necessitate peripheral administration in practice. We investigated phlebitis risk associated with peripheral 11.7% HTS in AIS patients requiring hyperosmolar therapy.
  • Methods
    This retrospective study included 200 AIS patients receiving hyperosmolar therapy with mannitol (January 2022–December 2024), classified into an HTS group (n=67) receiving additional peripheral 11.7% NaCl and a no-HTS group (n=133). Phlebitis was assessed using the Infusion Nurses Society Phlebitis Scale. The primary outcome was phlebitis within 72 hours of therapy initiation. Multivariable logistic regression assessed the independent association between HTS use and phlebitis, with a secondary analysis examining frequency-dependent risk.
  • Results
    Phlebitis occurred in 31.3% (21/67) of the HTS group versus 15.8% (21/133) of the no-HTS group (P=0.016). Among those with phlebitis, most cases were mild (grade 1, 90.5%; grade 2, 9.5%), with no grade 3 or higher events. HTS was independently associated with phlebitis (adjusted OR, 2.19; 95% CI, 1.02–4.70; P=0.044) but not with mortality or hospital length of stay. Single administration was not associated with increased risk, whereas two administrations (adjusted OR, 5.43; P=0.005) and three or more (adjusted OR, 3.59; P=0.016) were associated with markedly elevated risk (P for trend=0.002).
  • Conclusions
    Peripheral 11.7% HTS was associated with increased phlebitis risk, though cases were predominantly mild. Risk increased with repeated rather than single-dose administration, suggesting caution is warranted when multiple peripheral doses are required.
Cerebral edema is a critical complication of acute ischemic stroke (AIS), contributing to elevated intracranial pressure (ICP), cerebral herniation, and early neurological deterioration [1]. Timely ICP control is therefore a fundamental therapeutic priority in the acute phase of stroke management [2,3]. Hyperosmolar therapy remains the cornerstone of medical ICP management, with mannitol and hypertonic saline (HTS) serving as the principal pharmacologic agents [4,5].
HTS exerts its therapeutic effect by generating a robust osmotic gradient that facilitates translocation of water from the cerebral parenchyma into the intravascular compartment. Owing to a reflection coefficient approaching 1.0 across the intact blood–brain barrier, HTS provides a sustained osmotic effect with reduced risk of intracerebral accumulation compared to mannitol [5-7]. Beyond osmotic pressure reduction, HTS expands intravascular volume, augments mean arterial pressure, and improves cerebral perfusion pressure, offering potential hemodynamic benefits in patients with compromised cerebral perfusion [8-10]. Several prior studies and guideline statements suggest that HTS may be considered particularly in cases where the response to mannitol is inadequate, or as a complementary agent in refractory ICP elevation [4,11,12].
Historically, HTS has been recommended for administration via central venous catheter (CVC) [4,13], based on concerns that its high osmolarity may induce endothelial injury when infused peripherally, increasing the risk of thrombophlebitis, infiltration, extravasation, and tissue necrosis [13-15]. However, in acute neurological emergencies requiring rapid ICP reduction, the time required for CVC placement and its associated procedural risks may introduce clinically significant delays in treatment. This practical constraint has prompted increasing clinical interest in whether peripheral HTS administration can be performed safely under monitored conditions.
Prior investigations into peripheral HTS safety have yielded conflicting results, with reported complication rates varying widely across studies depending on HTS concentration, infusion method, and observation window [16-21]. Studies have predominantly focused on 3% HTS administered as a continuous infusion, with limited evidence available for higher-concentration bolus formulations [20]. In clinical practice, high-concentration bolus formulations are increasingly utilized for acute ICP crises, offering more rapid pressure reduction with lower infusion volumes compared to continuous 3% HTS infusion [22]. While 23.4% NaCl is the standard high-concentration formulation in Western practice, it is not commercially available in South Korea; 11.7% NaCl, which delivers an equivalent sodium dose in twice the volume, represents the only available high-concentration formulation in Korean stroke centers. Critically, evidence regarding peripheral administration of high-concentration HTS specifically in AIS patients—and whether phlebitis risk scales with the number of administrations—has not been systematically evaluated, leaving a clinically important gap in the guidance for real-world practice. Current major guidelines continue to recommend CVC administration as the standard of care, though the supporting evidence remains low-quality or based on expert consensus [4]. Therefore, this study aimed to evaluate whether peripheral administration of 11.7% NaCl HTS independently increases phlebitis risk in AIS patients receiving hyperosmolar therapy, and to characterize the frequency-dependent nature of this risk.
The current study was approved by Institutional Review Board at Chonnam National University Hospital (No. CNUH-2026-186). A waiver for informed consent was provided because of study subject anonymity and minimal risk to the participants. The data used in this study are available upon reasonable request following the submission of a legitimate academic research proposal to be assessed by the corresponding author.
Study Subjects
This retrospective study was conducted at a single comprehensive stroke center. Consecutive patients admitted with AIS within 48 hours of symptom onset between January 2022 and December 2024 were included (n=2,512). After excluding patients who did not receive hyperosmolar therapy (n=2,275) and those younger than 19 years (n=5), 232 patients who received hyperosmolar therapy for cerebral edema management during hospitalization were identified. Patients who received HTS via CVC (n=12) or those with incomplete infusion data (n=20) were excluded, resulting in a final analytic cohort of 200 patients (Figure 1). All patients received peripheral intravenous mannitol as the primary hyperosmolar agent and were classified into two groups: those who additionally received peripheral HTS (HTS group, n=67) and those who received mannitol alone (no-HTS group, n=133).
Data Collection
Baseline demographic, clinical, and laboratory data were systematically extracted from electronic medical records. Initial National Institutes of Health Stroke Scale (NIHSS) scores were assessed by emergency department neurologists or certified stroke unit nurses at the time of admission. Pre-stroke modified Rankin Scale (mRS) scores were obtained through structured interviews with patients or their caregivers, with functional independence defined as mRS 0–1. Vascular risk factors (hypertension, diabetes mellitus, hyperlipidemia, atrial fibrillation, coronary heart disease, prior stroke, peripheral artery disease, and smoking status) and medication history (antiplatelet agents, anticoagulants, antihypertensive drugs, antidiabetic agents, and statins) were documented at admission through medical record review and patient interviews. Stroke subtypes were independently determined by two neurologists according to the Trial of Org 10172 in Acute Stroke Treatment (TOAST) classification, incorporating predefined imaging criteria [23,24]. Laboratory parameters were obtained from the first blood tests performed at admission. All patients underwent brain imaging (computed tomography or magnetic resonance imaging) within 24 hours of admission.
Hyperosmolar Treatment
All study participants received mannitol as the first-line hyperosmolar agent for cerebral edema control. Mannitol was administered at doses of 0.5–1.5 g/kg, repeated three to six times daily at the discretion of the attending physician. The decision to administer additional HTS was made by the attending neurologist based on clinical judgment, primarily guided by insufficient neurological improvement or signs of progressive cerebral edema despite mannitol therapy, including worsening level of consciousness or radiologic evidence of midline shift.
The HTS formulation used in this study was 11.7% NaCl 20 ml ampoules (11.7% Sodium Chloride inj., JW Pharmaceutical), which is commercially available in South Korea. Each dose consisted of three ampoules (total 60 ml, approximately 120 mEq Na), which is equivalent in sodium content to 23.4% NaCl 30 ml recommended in North American guidelines [4]. Since the 23.4% NaCl formulation is not commercially available in South Korea, the 11.7% NaCl formulation was used to achieve the same sodium dose. HTS was administered as a bolus via peripheral intravenous access within 15 minutes and was repeated at 4–6-hour intervals as needed. All patients underwent regular monitoring of serum sodium and electrolyte levels during HTS administration, with a target serum sodium concentration of 145–155 mEq/L. HTS was administered through peripheral venous catheters of 20 gauge or larger. Peripheral infusion sites were assessed by nursing staff at every shift change (minimum 8-hour intervals) for signs of phlebitis including erythema, edema, pain, and induration. Upon detection of suspicious findings, infusion was immediately discontinued and switched to the contralateral vein.
Definition of Phlebitis
Phlebitis was defined and graded according to the Infusion Nurses Society (INS) Phlebitis Scale, a validated 5-point classification system based on clinical signs at the infusion site [13,25]. The scale ranges from 0 (no symptoms) to 4 (severe thrombophlebitis with purulent drainage). Grade 0 indicates the absence of any clinical signs. Grade 1 is defined as erythema at the access site with or without pain. Grade 2 is characterized by pain at the access site with erythema and/or edema. Grade 3 comprises pain accompanied by erythema, streak formation, and a palpable venous cord. Grade 4, the most severe, is defined as pain with erythema, streak formation, a palpable venous cord exceeding one inch in length, and purulent drainage.
Phlebitis assessment was conducted retrospectively through comprehensive medical record review, including nursing documentation, physician notes, and photographic images of infusion sites when available. Two independent assessors—a neurologist (HK) and a neurocritical care nurse (HH)—reviewed all available documentation and assigned INS grades. All discrepancies were resolved through consensus discussion, and final grades were determined by agreement between both reviewers. For patients with documented venous ultrasonography or other imaging studies of the affected extremity, radiologic findings were incorporated into the assessment. For the primary analysis, phlebitis was defined as any event with INS grade ≥1.
Outcome
The primary outcome was the occurrence of phlebitis within 72 hours of hyperosmolar therapy initiation, defined as the development of clinical signs consistent with INS grade ≥1 following peripheral intravenous infusion. The observation window was anchored to the initiation of hyperosmolar therapy rather than to the time of the last HTS administration, in order to apply a uniform and prospectively defined observation period across all patients, given that the number and timing of administrations varied between individuals. Secondary outcomes included in-hospital mortality, favorable functional outcome at 3 months (defined as mRS score 0–3, reflecting the expected severe neurological burden in patients requiring hyperosmolar therapy), and hospital length of stay (LOS), calculated as the number of days from admission to discharge.
Statistical Analysis
Baseline characteristics were compared between the HTS and no-HTS groups. Continuous variables were expressed as mean±standard deviation and compared using Student t-test. Non-normally distributed continuous variables (initial NIHSS score and pre-stroke mRS) were expressed as median (interquartile range) and compared using the Mann-Whitney U-test. Categorical variables were presented as frequencies and percentages and compared using Fisher's exact test with Monte Carlo simulation (B=10,000).
Event rates for phlebitis and mortality were presented as the number of events, percentage, and 95% CI calculated using the exact binomial method. Between-group differences in phlebitis were assessed using the chi-square test. The distribution of phlebitis severity by INS grade was presented as patient counts and percentages for each grade. LOS was presented as median (interquartile range) and compared between groups using the Mann-Whitney U-test. In-hospital mortality was additionally analyzed using the log-rank test, with a maximum observed follow-up duration of 91 days.
Multivariable logistic regression was used for the binary outcome of phlebitis, with results expressed as odds ratios (ORs) with 95% CI. Cox proportional hazards regression was used for in-hospital mortality, with results expressed as hazard ratios (HRs) with 95% CI; the proportional hazards assumption was verified using Schoenfeld residuals. For LOS, which exhibited a right-skewed distribution, log-transformation was applied prior to analysis, and multivariable linear regression was performed; results are expressed as the regression coefficient (β) with 95% CI. Covariates in all adjusted models were selected a priori based on clinical relevance. The final set of covariates comprised age, initial NIHSS score, pre-stroke mRS score, prior antiplatelet use, reperfusion therapy, and TOAST classification.
To examine the dose-response relationship between HTS administration frequency and phlebitis risk, patients were stratified into four groups: no HTS (reference), HTS with one administration, two administrations, and three or more administrations. Phlebitis event rates are presented with exact binomial 95% CIs for each group. Crude and adjusted ORs relative to the no HTS reference group were estimated using logistic regression. The Cochran-Armitage trend test was used to assess the statistical significance of the monotonic trend in crude phlebitis rates across the four ordered groups. For the adjusted trend test, HTS administration frequency was entered as an ordinal variable (coded 0, 1, 2, ≥3) in the multivariable logistic regression model, and the corresponding p-value was derived from the Wald test of this ordinal term. In addition, among HTS patients with available administration count data (n=67), the continuous dose-response relationship between HTS administration frequency and each clinical outcome (phlebitis and LOS) was modeled using restricted cubic spline (RCS) regression with three knots placed at the 10th, 50th, and 90th percentiles of the administration frequency distribution. To identify independent predictors of phlebitis among all study participants, a multivariable logistic regression model was constructed with HTS use as the primary exposure variable and the aforementioned covariates simultaneously entered. All statistical analyses were performed using R software version 4.4 (R Foundation for Statistical Computing). A two-sided P-value <0.05 was considered statistically significant.
Baseline Characteristics
Among the 200 patients, 67 (33.5%) received additional peripheral HTS (HTS group), while 133 (66.5%) received mannitol alone (no-HTS group). The mean age was 75±11 years, and 88 (44.0%) were male. Pre-stroke functional independence (mRS 0–1) was observed in 194 patients (97.0%). Most baseline characteristics were similar between the two groups; however, the HTS group had a higher initial NIHSS score compared to the no-HTS group (median, 14 vs. 13; P=0.022) and higher prior antiplatelet use (25.4% vs. 12.8%, P=0.030). Vascular risk factors including diabetes and peripheral arterial disease, reperfusion therapy, TOAST classification, and laboratory parameters showed no significant differences between the groups (Table 1).
Phlebitis and Clinical Outcomes
Phlebitis occurred in 42 patients (21.0%), with an incidence of 31.3% (21/67) in the HTS group and 15.8% (21/133) in the no-HTS group, representing a significantly higher rate in the HTS group (P=0.016). Regarding INS grade distribution, grade 1 accounted for the majority with 38 patients (90.5%), while grade 2 occurred in 4 patients (9.5%), all exclusively in the HTS group (6.0% vs. 0.0%, P=0.012). Grade 3 or higher severe phlebitis was not observed in either group.
For secondary outcomes, in-hospital mortality was 23.9% (16/67) in the HTS group and 15.0% (20/133) in the no-HTS group, with no statistically significant difference (P=0.171). Favorable functional outcome at 3 months (mRS 0–3) also showed no difference between the groups (20.9% vs. 25.6%, P=0.490). The median LOS was 10.9 days (interquartile range [IQR], 4.6–18.6) in the HTS group and 9.1 days (IQR, 5.5–13.7) in the no-HTS group, with no significant difference (P=0.341) (Table 2).
In multivariable analysis, HTS use was independently associated with phlebitis occurrence (adjusted OR, 2.19; 95% CI, 1.02–4.70; P=0.044). In contrast, no independent association was observed with in-hospital mortality (adjusted HR, 1.48; 95% CI, 0.73–2.97; P=0.274) or LOS (adjusted β, 0.024; 95% CI, −0.262 to 0.311; P=0.867) (Table 3).
Frequency-Response Relationship
Stratified analysis by HTS administration frequency revealed phlebitis incidence rates of 15.8% in the no-HTS group, 10.0% in the single-administration group, 53.3% in the two-administration group, and 45.5% in the three-or-more-administration group. Using the no-HTS group as reference, the adjusted OR was 0.51 (95% CI, 0.11–1.75; P=0.325) for the single-administration group, showing no significant risk increase, whereas it was significantly elevated at 5.43 (95% CI, 1.65–18.22; P=0.005) for the two-administration group and 3.59 (95% CI, 1.25–10.27; P=0.016) for the three-or-more-administration group (P for trend=0.002) (Table 4).
RCS analysis demonstrated a significant association between HTS administration frequency and phlebitis risk, with evidence of non-linearity (P for overall effect=0.002, P for non-linear effect=0.028) (Figure 2A). The predicted probability of phlebitis increased steeply from the first to third administration, reaching an estimated plateau of approximately 45%–50% beyond three administrations, consistent with the threshold pattern observed in the categorical analysis. For LOS, a significant overall association with HTS administration frequency was observed (P for overall effect=0.031); however, the non-linear term did not reach statistical significance (P for non-linear=0.186) (Figure 2B), suggesting a broadly increasing trend in LOS with more frequent administrations rather than a distinctly non-linear pattern.
Independent Predictors of Phlebitis
In multivariable logistic regression analysis of all study participants, HTS use was the only independent predictor of phlebitis occurrence (adjusted OR, 2.21; 95% CI, 1.03–4.75; P=0.04). Age, sex, initial NIHSS score, pre-stroke mRS, reperfusion therapy, TOAST classification, prior antiplatelet use, hypertension, diabetes mellitus, and atrial fibrillation all showed no statistically significant association (Table 5).
In this study of AIS patients requiring hyperosmolar therapy, those who received 11.7% NaCl HTS via peripheral intravenous access had a phlebitis incidence of 31.3%, significantly higher than the 15.8% observed in the no-HTS group, and HTS use was confirmed as an independent predictor of phlebitis in multivariable analysis (adjusted OR, 2.19; P=0.044). Notably, however, a distinct dose-response relationship was observed according to administration frequency. The incidence sharply increased from the second peripheral HTS administration (adjusted OR, 5.43; P=0.005), and this trend was statistically significant (P for trend=0.002). Furthermore, the severity of phlebitis was mostly mild, with INS grade 1 accounting for the majority (90.5%), grade 2 in only 4 patients (9.5%), and no grade 3 or higher severe phlebitis or tissue necrosis observed. Additionally, HTS use showed no independent association with in-hospital mortality or LOS, indicating that the clinical harm from HTS was confined to predominantly non-severe phlebitis.
The severity distribution of phlebitis in this cohort warrants brief consideration. The majority of events were INS grade 1, which is generally reversible and managed conservatively through catheter removal and relocation, typically resolving within several days without invasive intervention. Although four grade 2 events occurred exclusively in the HTS group (6.0% vs. 0.0%, P=0.012), suggesting that HTS may occasionally contribute to phlebitis extending beyond simple erythema, no event reached grade 3 or higher and no tissue necrosis was observed. Taken together, the vascular injury associated with peripheral 11.7% HTS was predominantly mild and clinically manageable, although even low-grade events may accumulate clinical relevance when repeated peripheral administration is required.
Our observed phlebitis incidence of 31.3% is substantially higher than the pooled 6.2% phlebitis rate reported in a meta-analysis of peripheral 3% HTS studies [17], and the 3.7% complication rate reported in a bolus 3% HTS study [16]. This difference is expected given the markedly higher osmolarity of 11.7% NaCl (4,004 mOsm/L) relative to 3% NaCl (1,027 mOsm/L). Notably, even 3% HTS has been associated with phlebitis rates as high as 47% in some cohorts [21], suggesting that factors beyond osmolarity alone, such as infusion duration and administration method, may also contribute to vascular complications, though detailed comparison with our bolus protocol is not straightforward. For higher-concentration formulations, prior studies have predominantly assessed extravasation rather than phlebitis as their primary endpoint, and were conducted almost exclusively in single-dose emergency department settings. The largest such study reported an extravasation incidence of 6.4% with peripheral 23.4% NaCl across 863 administrations, which is a figure not directly comparable to our phlebitis incidence, given the differences in outcome definition, observation window, and clinical context. This lack of comparability arises because that study focused predominantly on single emergency department administrations assessed over 24 hours, whereas our study involved repeated inpatient dosing observed over 72 hours. Our findings may therefore extend rather than contradict prior evidence, by characterizing vascular complications across a broader outcome spectrum and a clinically distinct administration context [20].
The recommendation for CVC administration of HTS is based on the concern that high osmolarity solutions, which far exceed the 900 mOsm/L threshold considered safe for peripheral venous administration, may induce endothelial injury, increasing the risk of thrombophlebitis and extravasation [15]. While this concern is physiologically grounded, the requirement for CVC access introduces a practical dilemma in acute ICP emergencies: central line placement is time-consuming and carries its own procedural risks including infection and thrombosis [26]. In this context, the time-to-treatment argument is clinically important. Prior studies have demonstrated that CVC placement can introduce meaningful delays in osmotherapy initiation, and that earlier treatment is associated with better neurological outcomes in patients with acute ICP elevation [20,27]. When central access is not immediately available, peripheral administration of a single HTS dose may represent the only means of delivering timely osmotherapy. Our finding that single-dose peripheral administration was not associated with a statistically significant increase in phlebitis risk, whereas two or more administrations were associated with such an increase, may provide some quantitative context for this clinical consideration, though prospective data are needed before firm recommendations can be made.
Another key contribution of this study is the first systematic quantification of frequency-dependent phlebitis risk with peripheral high-concentration HTS. While prior studies addressed the binary question of "is peripheral administration safe," our study examined the more clinically actionable question of "how many times can it be administered peripherally." The absence of statistically significant increased risk with single administration (adjusted OR, 0.51; P=0.325) may reflect a genuine threshold below which a single hyperosmolar bolus does not cause sufficient cumulative endothelial injury to manifest as clinically detectable phlebitis. The numerically lower incidence in the single-administration group compared to the no-HTS reference (10.0% vs. 15.8%) may additionally reflect selection of patients with relatively mild cerebral edema who required only one dose, rather than a true protective effect. In contrast, the sharp risk increase from the second administration (OR, 5.43) is both statistically and clinically clear, consistent with a model of cumulative endothelial damage in which repeated hyperosmolar exposure before sufficient vascular recovery compounds rather than resets prior injury [14,28]. This non-linear dose-response relationship was further corroborated by RCS analysis, demonstrating a marked increase in phlebitis probability beginning at the second administration, consistent with the elevated rates observed in the categorical analysis for two or more administrations. The apparent decline in the adjusted OR from the two-administration group (5.43) to the three-or-more-administration group (3.59) should not be interpreted as a true reversal of the dose-response relationship. Both subgroups were small (n=15 and n=22, respectively), yielding wide and largely overlapping confidence intervals (1.65–18.22 and 1.25–10.27), such that the two point estimates are statistically indistinguishable. This instability, together with the previously described possibility of survivor bias, likely accounts for the non-monotonic pattern. The RCS analysis offers a complementary and more stable perspective, depicting a continuous rise in phlebitis probability that plateaus rather than reverses beyond the third administration. Taken together, the categorical and spline analyses are best interpreted as concordant evidence of a threshold-type increase in risk emerging at the second administration, rather than as a precisely graded dose-response across each additional dose.
We used the 11.7% NaCl formulation, which is the only commercially available high-concentration HTS product in South Korea and is equivalent to 23.4% NaCl 30 ml in sodium content (120 mEq), but differs in volume (60 ml vs. 30 ml) and osmolarity (approximately 4,004 mOsm/L vs. 8,008 mOsm/L). Theoretically, the larger volume (60 ml) may increase the dilution effect with blood in the peripheral vein, lowering the contact osmolarity on vascular endothelial cells, which could reduce phlebitis incidence compared to the 23.4% formulation [14,29]. On the other hand, the increased infusion time associated with larger volumes may prolong vascular endothelial cell exposure to the hyperosmolar environment. Since previous studies have reported that longer infusion times increase phlebitis risk, if our infusion duration was longer than that for 23.4%, this could have increased phlebitis risk [28]. Given these opposing theoretical effects, the net phlebitis risk of 11.7% NaCl relative to 23.4% NaCl remains uncertain, and our observed incidence should not be extrapolated to other formulations without further comparative data. Importantly, to our knowledge, no prior study has specifically examined the peripheral administration safety of 11.7% NaCl. As the only commercially available high-concentration HTS formulation in South Korea, this agent is used in routine clinical practice without an established evidence base for peripheral safety, further highlighting the clinical relevance of our findings.
Our study has several important limitations. First, the single-center retrospective design precludes the complete exclusion of residual confounding, and indication bias is inherent given that the decision to administer HTS rested entirely on the judgment of the attending physician. Patients in the HTS group had higher initial NIHSS scores, raising the possibility that greater neurological severity, rather than HTS itself, contributed to the observed risk of phlebitis through more intensive treatment or more frequent venous access. Although both groups received peripheral mannitol as the first-line agent, the number of mannitol administrations and the cumulative dose were not recorded and could not be included as covariates; a higher overall infusion burden in the HTS group therefore cannot be excluded. However, the primary outcome of this study was the occurrence of phlebitis rather than mortality or functional prognosis, which are the outcomes influenced by neurological severity, and in the multivariable model neurological severity was not significantly associated with the occurrence of phlebitis. Nevertheless, because of the observational design, residual or unmeasured confounding cannot be entirely excluded, and the findings of this study should therefore be interpreted as hypothesis-generating. Second, the small subgroup sizes in the two-administration (n=15) and three-or-more-administration (n=22) groups yield wide confidence intervals and unstable point estimates, limiting the precision of the frequency-stratified analysis. Third, detailed peripheral intravenous catheter-related variables— including catheter gauge, insertion site location, indwelling duration, and replacement frequency—were not systematically recorded, precluding site- and catheter-specific risk analyses. These factors are recognized contributors to phlebitis development, and their omission may have introduced additional unmeasured confounding access. Consequently, the three-or-more-administration group may have been enriched with patients with better vascular tolerance, which could potentially explain the paradoxically lower adjusted OR in that group compared to the two-administration group (3.59 vs. 5.43). Fourth, phlebitis was assessed retrospectively through medical record review rather than prospective bedside evaluation. This approach may have resulted in under-ascertainment of mild grade 1 events not formally documented in nursing records. Furthermore, inter-rater reliability was not formally quantified; however, all discrepancies were resolved through consensus between two independent reviewers with photographic and ultrasonographic evidence incorporated where available, partially mitigating this limitation. Fifth, the 72-hour observation window, while longer than the 24-hour window used in prior single-dose studies, was selected to accommodate the delayed phlebitis manifestation expected with repeated inpatient dosing. However, phlebitis events occurring beyond 72 hours, which may occur with later administrations, would not have been captured, potentially leading to underestimation of cumulative risk. Finally, misclassification of phlebitis caused by concomitant irritant medications as HTS-related cannot be entirely excluded.
Despite these limitations, several strengths of this study merit acknowledgment. First, to our knowledge, this is the first study to evaluate the peripheral administration safety of 11.7% NaCl HTS in AIS patients. Second, by evaluating phlebitis risk according to the number of administrations, this study systematically quantified frequency-dependent risk across different categories. This approach addresses a clinically specific and actionable question—how many times HTS can be safely administered peripherally—which prior binary safety studies of other HTS formulations were unable to answer.
In conclusion, our study demonstrated that peripheral administration of 11.7% high-concentration NaCl in AIS patients increases phlebitis risk; however, with single-dose administration, the risk increase is not evident, and most phlebitis that occurred was mild and clinically manageable. These findings suggest the possibility of considering single-dose high-concentration HTS bolus administration via peripheral vein as rescue therapy in emergency situations when central venous access is not immediately available. However, repeated administration of two or more doses sharply increases phlebitis risk, and in such cases, central venous access should be actively pursued. This study is hypothesis-generating in nature, and prospective multicenter studies and pragmatic randomized trials based on standardized protocols are needed to establish such clinical guidelines.
▪ Peripheral administration of 11.7% hypertonic saline was independently associated with phlebitis in acute ischemic stroke patients requiring hyperosmolar therapy, though observed cases were predominantly mild with no severe tissue injury.
▪ Phlebitis risk showed a frequency-dependent pattern: single-dose administration was not associated with a statistically significant risk increase, whereas two or more administrations were associated with markedly elevated risk.
▪ These findings may inform clinical decision-making regarding peripheral hypertonic saline use when central venous access is not immediately available, and highlight the need for prospective studies to establish definitive guidance.

CONFLICT OF INTEREST

No potential conflict of interest relevant to this article was reported.

FUNDING

This study was supported by a grant (BCRI25031) of Chonnam National University Hospital Biomedical Research Institute.

ACKNOWLEDGMENTS

None.

AUTHOR CONTRIBUTIONS

Conceptualization: BSS, HH, JJ, HK. Methodology: BSS, HK. Formal analysis: HH, HK. Data curation: BSS, HH, JHH, HK. Project administration: HK. Funding acquisition: HK. Writing - original draft: BSS, HH, HK. Writing - review & editing: HK. All authors read and agreed to the published version of the manuscript.

Figure 1.
Study flowchart.
acc-003328f1.jpg
Figure 2.
Restricted cubic spline analysis of hypertonic saline (HTS) administration frequency and clinical outcomes. Curves were estimated using restricted cubic spline models with three knots: logistic regression for phlebitis (A) and linear regression for hospital length of stay (LOS; B). Shaded areas represent 95% CIs. Rug plots along the x-axis indicate the distribution of observed administration frequencies. Overall and nonlinearity P-values are derived from likelihood ratio tests.
acc-003328f2.jpg
Table 1.
Baseline characteristics according to hypertonic saline use
Variable All (n=200) No HTS (n=133) HTS (n=67) P-value
Age (yr) 75±11 76±11 73±12 0.140
Male 88 (44.0) 56 (42.1) 32 (47.8) 0.455
Initial NIHSS score 13.0 (9.0–17.0) 13.0 (8.0–17.0) 14.0 (10.0–18.0) 0.022
Pre-stroke mRS 0.666
 mRS 0–1 194 (97.0) 128 (96.2) 66 (98.5)
 mRS 2–5 6 (3.0) 5 (3.8) 1 (1.5)
Reperfusion therapy 0.289
 None 122 (61.0) 87 (65.4) 35 (52.2)
 IVT 32 (16.0) 18 (13.5) 14 (20.9)
 EVT 30 (15.0) 19 (14.3) 11 (16.4)
 IVT+EVT 16 (8.0) 9 (6.8) 7 (10.4)
TOAST classification 0.790
 LAA 31 (15.5) 21 (15.8) 10 (14.9)
 CE 94 (47.0) 64 (48.1) 30 (44.8)
 OD 2 (1.0) 2 (1.5) 0
 UD 73 (36.5) 46 (34.6) 27 (40.3)
Vascular risk factor
 Prior stroke 23 (11.5) 15 (11.3) 8 (11.9) >0.999
 Coronary heart disease 21 (10.5) 17 (12.8) 4 (6.0) 0.220
 Peripheral artery disease 3 (1.5) 2 (1.5) 1 (1.5) >0.999
 Hypertension 123 (61.5) 86 (64.7) 37 (55.2) 0.220
 Diabetes mellitus 60 (30.0) 44 (33.1) 16 (23.9) 0.195
 Hyperlipidemia 37 (18.5) 21 (15.8) 16 (23.9) 0.180
 Atrial fibrillation 92 (46.0) 62 (46.6) 30 (44.8) 0.881
 Smoking 0.812
  Never 175 (87.5) 115 (86.5) 60 (89.6)
  Ex-smoker 19 (9.5) 13 (9.8) 6 (9.0)
  Current 6 (3.0) 5 (3.8) 1 (1.5)
Medication history
 Antiplatelet use 34 (17.0) 17 (12.8) 17 (25.4) 0.030
 Anticoagulant use 37 (18.5) 25 (18.8) 12 (17.9) >0.999
 Antihypertensive use 86 (43.0) 55 (41.4) 31 (46.3) 0.547
 Antidiabetic use 34 (17.0) 22 (16.5) 12 (17.9) 0.843
 Statin use 29 (14.5) 20 (15.0) 9 (13.4) 0.834
Laboratory findings
 WBC (×10³/µl) 9.6±3.7 9.7±3.4 9.4±4.1 0.592
 Hemoglobin (g/dl) 13.2±2.1 13.3±2.2 13.0±1.9 0.446
 Platelet (×10³/µl) 218.8±84.1 215.2±79.4 225.8±92.9 0.428
 BUN (mg/dl) 19.4±11.2 20.1±12.0 18.0±9.4 0.179
 Creatinine (mg/dl) 0.8±0.4 0.8±0.5 0.8±0.2 0.672
 LDL-C (mg/dl) 99.7±36.4 101.7±36.2 95.5±36.9 0.259
 HDL-C (mg/dl) 45.0±11.4 45.3±11.5 44.4±11.4 0.612
 Total cholesterol (mg/dl) 166.5±40.5 168.8±40.6 162.1±40.2 0.270
 Triglyceride (mg/dl) 103.1±55.4 104.1±52.0 101.3±62.0 0.752
 HbA1c (%) 6.0±1.0 6.0±1.0 6.1±1.0 0.452
 Fasting glucose (mg/dl) 144.2±47.6 144.3±48.2 144.0±46.6 0.965
 Systolic BP (mm Hg) 142.0±26.0 142.0±25.2 141.9±27.8 0.981

Values are presented as mean±standard deviation, number (%), or median (interquartile range).

HTS: hypertonic saline; NIHSS: National Institutes of Health Stroke Scale; mRS: modified Rankin Scale; IVT: intravenous thrombolysis; EVT: endovascular treatment; TOAST: Trial of Org 10172 in Acute Stroke Treatment; LAA: large artery atherosclerosis; CE: cardioembolism; OD: other determined etiology; UD: undetermined etiology; WBC: white blood cell; BUN: blood urea nitrogen; LDL-C: low-density lipoprotein cholesterol; HDL-C: high-density lipoprotein cholesterol; HbA1c: glycated hemoglobin; BP: blood pressure.

Student t-test was used for continuous variables. Fisher's exact test with Monte Carlo simulation was used for categorical variables.

Table 2.
Clinical outcomes according to hypertonic saline use
No. of events/event rate, % (95% CI) P-valuea)
All (n=200) HTS (n=67) No HTS (n=133)
Primary outcome
 Phlebitis
  All grade 42/21.0 (15.6–27.3) 21/31.3 (20.6–43.8) 21/15.8 (10.1–23.1) 0.016
  Grade 1 38/19.0 (13.8–25.1) 17/25.4 (15.5–37.5) 21/15.8 (10.0–23.1) 0.127
  Grade 2 4/2.0 (0.5–5.0) 4/6.0 (1.7–14.6) 0/0.0 (0.0–2.7) 0.012
  Grade 3–4 - - -
Secondary outcome
 In-hospital mortality 36/18.0 (12.9–24.0) 16/23.9 (14.3–35.9) 20/15.0 (9.4–22.3) 0.171
 3-Month mRS 0–3 48/24.0 (18.3–30.5) 14/20.9 (11.9–32.6) 34/25.6 (18.4–33.9) 0.490
 LOS (day) 9.5 (5.4–15.4) 10.9 (4.6–18.6) 9.1 (5.5–13.7) 0.341

Values are presented as number (%, 95% CI by exact binomial method) or median (IQR). Log-rank test was used for mortality (follow-up, 91 days).

HTS: hypertonic saline; mRS: modified Rankin Scale; LOS: hospital length of stay; IQR: interquartile range.

a)Chi-square test for phlebitis; Fisher’s exact test for 3-month mRS 0-3; Mann-Whitney U-test for LOS; Log-rank test for in-hospital mortality.

Table 3.
Association between hypertonic saline use and clinical outcomes
Outcome Statistical model Crude effect (95% CI) P-value Adjusted effect (95% CI) P-value
Phlebitis Logistic regression (OR) 2.43 (1.21 to 4.91) 0.012 2.19 (1.02 to 4.70) 0.044
Mortality Cox proportional hazards (HR) 1.65 (0.86 to 3.19) 0.134 1.48 (0.73 to 2.97) 0.274
LOS Linear regression (β)a) 0.028 (−0.238 to 0.295) 0.834 0.024 (−0.262 to 0.311) 0.867

Reference group: no-HTS. Adjusted for age, initial National Institutes of Health Stroke Scale score, pre-stroke modified Rankin Scale score, previous antiplatelet treatment, reperfusion therapy, and TOAST classification.

OR: odds ratio; HR: hazard ratio; LOS: hospital length of stay; TOAST: Trial of Org 10172 in Acute Stroke Treatment.

a)Length of stay was log-transformed prior to analysis; β represents the difference in log-transformed days between groups.

Table 4.
Association between frequency of hypertonic saline administration and phlebitis
Group Number Phlebitis n (%, 95% CI) Crude OR (95% CI) P-value Adjusted OR (95% CI) P-value
HTS (−) 133 21 (15.8, 10.0–23.1) 1.00 (reference) 1.00 (reference)
HTS (+) 1 30 3 (10.0, 2.1–26.5) 0.59 (0.13–1.88) 0.423 0.51 (0.11–1.75) 0.325
HTS (+) 2 15 8 (53.3, 26.6–78.7) 6.10 (1.99–19.19) 0.002 5.43 (1.65–18.22) 0.005
HTS (+) ≥3 22 10 (45.5, 24.4–67.8) 4.44 (1.68–11.68) 0.002 3.59 (1.25–10.27) 0.016
P for trend <0.001 0.002

Crude odds ratios were estimated using logistic regression with no-HTS as the reference group. Adjusted odds ratios were derived from multivariable logistic regression adjusting for age, initial NIHSS score, pre-stroke modified Rankin Scale score, previous antiplatelet treatment, reperfusion therapy (none/IVT/EVT/IVT+EVT), and TOAST classification. The crude P for trend was calculated using the Cochran-Armitage trend test. The adjusted P for trend was obtained by including HTS frequency as an ordinal variable (0, 1, 2, ≥3) in the multivariable logistic regression model.

OR: odds ratio; HTS: hypertonic saline; NIHSS: National Institutes of Health Stroke Scale; IVT: intravenous thrombolysis; EVT: endovascular treatment; TOAST: Trial of Org 10172 in Acute Stroke Treatment.

Table 5.
Independent association of hypertonic saline use with phlebitis
Variable Adjusted OR (95% CI) P-value
Hypertonic saline use 2.21 (1.03–4.75) 0.040
Age (per 1-year increase) 0.99 (0.95–1.03) 0.529
Male sex 1.16 (0.53–2.53) 0.709
Initial NIHSS (per 1-point increase) 1.04 (0.98–1.10) 0.225
Pre-stroke mRS (per 1-point increase) 1.11 (0.49–2.07) 0.764
Reperfusion: IVT (vs. none) 2.11 (0.79–5.45) 0.127
Reperfusion: EVT (vs. none) 1.06 (0.35–2.99) 0.909
Reperfusion: IVT+EVT (vs. none) 1.00 (0.23–3.57) 0.996
TOAST: cardioembolism 0.30 (0.07–1.16) 0.086
Prior antiplatelet use 0.51 (0.15–1.40) 0.222
Hypertension 0.89 (0.40–2.00) 0.772
Diabetes mellitus 0.94 (0.39–2.22) 0.895
Atrial fibrillation 2.18 (0.57–8.82) 0.258

Multivariable logistic regression model with phlebitis as the dependent variable. All listed covariates were entered simultaneously. Hypertonic saline use was the primary exposure variable.

OR: odds ratio; NIHSS: National Institutes of Health Stroke Scale; mRS: modified Rankin Scale; IVT: intravenous thrombolysis; EVT: endovascular treatment; TOAST: Trial of Org 10172 in Acute Stroke Treatment.

  • 1. Hacke W, Schwab S, Horn M, Spranger M, De Georgia M, von Kummer R, et al. 'Malignant' middle cerebral artery territory infarction: clinical course and prognostic signs. Arch Neurol 1996;53:309-15.ArticlePubMed
  • 2. Vahedi K, Hofmeijer J, Juettler E, Vicaut E, George B, Algra A, et al. Early decompressive surgery in malignant infarction of the middle cerebral artery: a pooled analysis of three randomised controlled trials. Lancet Neurol 2007;6:215-22.ArticlePubMed
  • 3. Simard JM, Kent TA, Chen M, Tarasov KV, Gerzanich V. Brain oedema in focal ischaemia: molecular pathophysiology and theoretical implications. Lancet Neurol 2007;6:258-68.ArticlePubMedPMC
  • 4. Cook AM, Morgan Jones G, Hawryluk GW, Mailloux P, McLaughlin D, Papangelou A, et al. Guidelines for the acute treatment of cerebral edema in neurocritical care patients. Neurocrit Care 2020;32:647-66.ArticlePubMedPMCPDF
  • 5. Ropper AH. Hyperosmolar therapy for raised intracranial pressure. N Engl J Med 2012;367:746-52.ArticlePubMed
  • 6. Bhardwaj A, Ulatowski JA. Cerebral edema: hypertonic saline solutions. Curr Treat Options Neurol 1999;1:179-88.ArticlePubMedPDF
  • 7. Mangat HS, Wu X, Gerber LM, Schwarz JT, Fakhar M, Murthy SB, et al. Hypertonic saline is superior to mannitol for the combined effect on intracranial pressure and cerebral perfusion pressure burdens in patients with severe traumatic brain injury. Neurosurgery 2020;86:221-30.ArticlePubMedPMCPDF
  • 8. Mortazavi MM, Romeo AK, Deep A, Griessenauer CJ, Shoja MM, Tubbs RS, et al. Hypertonic saline for treating raised intracranial pressure: literature review with meta-analysis. J Neurosurg 2012;116:210-21.ArticlePubMed
  • 9. Larive LL, Rhoney DH, Parker D, Coplin WM, Carhuapoma JR. Introducing hypertonic saline for cerebral edema: an academic center experience. Neurocrit Care 2004;1:435-40.ArticlePubMed
  • 10. Georgiadis AL, Suarez JI. Hypertonic saline for cerebral edema. Curr Neurol Neurosci Rep 2003;3:524-30.ArticlePubMedPDF
  • 11. Karamian A, Seifi A, Lucke-Wold B. Comparing the effects of mannitol and hypertonic saline in severe traumatic brain injury patients with elevated intracranial pressure: a systematic review and meta-analysis. Neurol Res 2024;46:883-92.ArticlePubMed
  • 12. van der Worp HB, Hofmeijer J, Jüttler E, Lal A, Michel P, Santalucia P, et al. European Stroke Organisation (ESO) guidelines on the management of space-occupying brain infarction. Eur Stroke J 2021;6:XC-CX.ArticlePubMedPMC
  • 13. Nickel B, Gorski L, Kleidon T, Kyes A, DeVries M, Keogh S, et al. Infusion therapy standards of practice, 9th edition. J Infus Nurs 2024;47:S1-285.ArticlePubMed
  • 14. Kuwahara T, Asanami S, Kawauchi Y, Kubo S. Experimental infusion phlebitis: tolerance pH of peripheral vein. J Toxicol Sci 1999;24:113-21.ArticlePubMed
  • 15. O'Grady NP, Alexander M, Burns LA, Dellinger EP, Garland J, Heard SO, et al. Guidelines for the prevention of intravascular catheter-related infections. Am J Infect Control 2011;39:S1-34.ArticlePubMed
  • 16. Khasiyev F, Hakoun A, Christopher K, Braun J, Wang F. Safety and effect on intracranial pressure of 3% hypertonic saline bolus via peripheral intravenous catheter for neurological emergencies. Neurocrit Care 2024;41:202-7.ArticlePubMedPDF
  • 17. Madieh J, Hasan B, Khamayseh I, Hrizat A, Salah T, Nayfeh T, et al. The safety of intravenous peripheral administration of 3% hypertonic saline: a systematic review and meta-analysis. Am J Med Sci 2023;366:135-42.ArticlePubMed
  • 18. Faiver L, Hensler D, Rush SC, Kashlan O, Williamson CA, Rajajee V, et al. Safety and efficacy of 23.4% sodium chloride administered via peripheral venous access for the treatment of cerebral herniation and intracranial pressure elevation. Neurocrit Care 2021;35:845-52.ArticlePubMedPDF
  • 19. Holden DN, Mucksavage JJ, Cokley JA, Kim KS, Tucker NL, Esordi MS, et al. Hypertonic saline use in neurocritical care for treating cerebral edema: a review of optimal formulation, dosing, safety, administration and storage. Am J Health Syst Pharm 2023;80:331-42.ArticlePubMedPDF
  • 20. Silva CR, Nagy J, Zullo AR, Greene RA. Safety of peripherally administered 23.4% sodium chloride. Neurocrit Care 2025;43:978-85.ArticlePubMedPDF
  • 21. Meng L, Nguyen CM, Patel S, Mlynash M, Caulfield AF. Association between continuous peripheral i.v. infusion of 3% sodium chloride injection and phlebitis in adults. Am J Health Syst Pharm 2018;75:284-91.ArticlePubMed
  • 22. Wang X, He Q, Ma L, You C. Comparison of different concentrations of hypertonic saline in patients with traumatic brain injury: evidence from direct and indirect comparisons. Injury 2022;53:3729-35.ArticlePubMed
  • 23. Ko Y, Lee S, Chung JW, Han MK, Park JM, Kang K, et al. MRI-based algorithm for acute ischemic stroke subtype classification. J Stroke 2014;16:161-72.ArticlePubMedPMC
  • 24. Adams HP, Bendixen BH, Kappelle LJ, Biller J, Love BB, Gordon DL, et al. Classification of subtype of acute ischemic stroke. definitions for use in a multicenter clinical trial. TOAST. trial of ORG 10172 in acute stroke treatment. Stroke 1993;24:35-41.ArticlePubMed
  • 25. Gorski LA, Hadaway L, Hagle ME, Broadhurst D, Clare S, Kleidon T, et al. Infusion therapy standards of practice, 8th edition. J Infus Nurs 2021;44:S1-224.ArticlePubMed
  • 26. Ouriel K. Preventing complications of central venous catheterization. N Engl J Med 2003;348:2684-6.Article
  • 27. Sekarningrum PA, Wati DK, Suwarba IM, Hartawan IN, Mahalini DS, Suparyatha IG, et al. Early mannitol administration improves clinical outcomes of pediatric patients with brain edema. Med J Indones 2018;27:244-9.ArticlePDF
  • 28. Deveau RF, Marino KK, Crowley KE, McLaughlin KC, Culbreth SE. Safety of peripherally administered 3% hypertonic saline. Am J Emerg Med 2023;63:127-31.ArticlePubMed
  • 29. Webster J, McGrail M, Marsh N, Wallis MC, Ray-Barruel G, Rickard CM, et al. Postinfusion phlebitis: incidence and risk factors. Nurs Res Pract 2015;2015:691934.ArticlePubMedPMCPDF

Figure & Data

References

    Citations

    Citations to this article as recorded by  

      • PubReader PubReader
      • ePub LinkePub Link
      • Cite
        CITE
        export Copy
        Close
        Download Citation
        Download a citation file in RIS format that can be imported by all major citation management software, including EndNote, ProCite, RefWorks, and Reference Manager.

        Format:
        • RIS — For EndNote, ProCite, RefWorks, and most other reference management software
        • BibTeX — For JabRef, BibDesk, and other BibTeX-specific software
        Include:
        • Citation for the content below
        Peripheral administration of 11.7% hypertonic saline and phlebitis risk in acute ischemic stroke requiring hyperosmolar therapy in South Korea
        Acute Crit Care. 2026;41(3):608-620.   Published online August 31, 2026
        Close
      • XML DownloadXML Download
      Figure
      • 0
      • 1
      Peripheral administration of 11.7% hypertonic saline and phlebitis risk in acute ischemic stroke requiring hyperosmolar therapy in South Korea
      Image Image
      Figure 1. Study flowchart.
      Figure 2. Restricted cubic spline analysis of hypertonic saline (HTS) administration frequency and clinical outcomes. Curves were estimated using restricted cubic spline models with three knots: logistic regression for phlebitis (A) and linear regression for hospital length of stay (LOS; B). Shaded areas represent 95% CIs. Rug plots along the x-axis indicate the distribution of observed administration frequencies. Overall and nonlinearity P-values are derived from likelihood ratio tests.
      Peripheral administration of 11.7% hypertonic saline and phlebitis risk in acute ischemic stroke requiring hyperosmolar therapy in South Korea
      Variable All (n=200) No HTS (n=133) HTS (n=67) P-value
      Age (yr) 75±11 76±11 73±12 0.140
      Male 88 (44.0) 56 (42.1) 32 (47.8) 0.455
      Initial NIHSS score 13.0 (9.0–17.0) 13.0 (8.0–17.0) 14.0 (10.0–18.0) 0.022
      Pre-stroke mRS 0.666
       mRS 0–1 194 (97.0) 128 (96.2) 66 (98.5)
       mRS 2–5 6 (3.0) 5 (3.8) 1 (1.5)
      Reperfusion therapy 0.289
       None 122 (61.0) 87 (65.4) 35 (52.2)
       IVT 32 (16.0) 18 (13.5) 14 (20.9)
       EVT 30 (15.0) 19 (14.3) 11 (16.4)
       IVT+EVT 16 (8.0) 9 (6.8) 7 (10.4)
      TOAST classification 0.790
       LAA 31 (15.5) 21 (15.8) 10 (14.9)
       CE 94 (47.0) 64 (48.1) 30 (44.8)
       OD 2 (1.0) 2 (1.5) 0
       UD 73 (36.5) 46 (34.6) 27 (40.3)
      Vascular risk factor
       Prior stroke 23 (11.5) 15 (11.3) 8 (11.9) >0.999
       Coronary heart disease 21 (10.5) 17 (12.8) 4 (6.0) 0.220
       Peripheral artery disease 3 (1.5) 2 (1.5) 1 (1.5) >0.999
       Hypertension 123 (61.5) 86 (64.7) 37 (55.2) 0.220
       Diabetes mellitus 60 (30.0) 44 (33.1) 16 (23.9) 0.195
       Hyperlipidemia 37 (18.5) 21 (15.8) 16 (23.9) 0.180
       Atrial fibrillation 92 (46.0) 62 (46.6) 30 (44.8) 0.881
       Smoking 0.812
        Never 175 (87.5) 115 (86.5) 60 (89.6)
        Ex-smoker 19 (9.5) 13 (9.8) 6 (9.0)
        Current 6 (3.0) 5 (3.8) 1 (1.5)
      Medication history
       Antiplatelet use 34 (17.0) 17 (12.8) 17 (25.4) 0.030
       Anticoagulant use 37 (18.5) 25 (18.8) 12 (17.9) >0.999
       Antihypertensive use 86 (43.0) 55 (41.4) 31 (46.3) 0.547
       Antidiabetic use 34 (17.0) 22 (16.5) 12 (17.9) 0.843
       Statin use 29 (14.5) 20 (15.0) 9 (13.4) 0.834
      Laboratory findings
       WBC (×10³/µl) 9.6±3.7 9.7±3.4 9.4±4.1 0.592
       Hemoglobin (g/dl) 13.2±2.1 13.3±2.2 13.0±1.9 0.446
       Platelet (×10³/µl) 218.8±84.1 215.2±79.4 225.8±92.9 0.428
       BUN (mg/dl) 19.4±11.2 20.1±12.0 18.0±9.4 0.179
       Creatinine (mg/dl) 0.8±0.4 0.8±0.5 0.8±0.2 0.672
       LDL-C (mg/dl) 99.7±36.4 101.7±36.2 95.5±36.9 0.259
       HDL-C (mg/dl) 45.0±11.4 45.3±11.5 44.4±11.4 0.612
       Total cholesterol (mg/dl) 166.5±40.5 168.8±40.6 162.1±40.2 0.270
       Triglyceride (mg/dl) 103.1±55.4 104.1±52.0 101.3±62.0 0.752
       HbA1c (%) 6.0±1.0 6.0±1.0 6.1±1.0 0.452
       Fasting glucose (mg/dl) 144.2±47.6 144.3±48.2 144.0±46.6 0.965
       Systolic BP (mm Hg) 142.0±26.0 142.0±25.2 141.9±27.8 0.981
      No. of events/event rate, % (95% CI) P-valuea)
      All (n=200) HTS (n=67) No HTS (n=133)
      Primary outcome
       Phlebitis
        All grade 42/21.0 (15.6–27.3) 21/31.3 (20.6–43.8) 21/15.8 (10.1–23.1) 0.016
        Grade 1 38/19.0 (13.8–25.1) 17/25.4 (15.5–37.5) 21/15.8 (10.0–23.1) 0.127
        Grade 2 4/2.0 (0.5–5.0) 4/6.0 (1.7–14.6) 0/0.0 (0.0–2.7) 0.012
        Grade 3–4 - - -
      Secondary outcome
       In-hospital mortality 36/18.0 (12.9–24.0) 16/23.9 (14.3–35.9) 20/15.0 (9.4–22.3) 0.171
       3-Month mRS 0–3 48/24.0 (18.3–30.5) 14/20.9 (11.9–32.6) 34/25.6 (18.4–33.9) 0.490
       LOS (day) 9.5 (5.4–15.4) 10.9 (4.6–18.6) 9.1 (5.5–13.7) 0.341
      Outcome Statistical model Crude effect (95% CI) P-value Adjusted effect (95% CI) P-value
      Phlebitis Logistic regression (OR) 2.43 (1.21 to 4.91) 0.012 2.19 (1.02 to 4.70) 0.044
      Mortality Cox proportional hazards (HR) 1.65 (0.86 to 3.19) 0.134 1.48 (0.73 to 2.97) 0.274
      LOS Linear regression (β)a) 0.028 (−0.238 to 0.295) 0.834 0.024 (−0.262 to 0.311) 0.867
      Group Number Phlebitis n (%, 95% CI) Crude OR (95% CI) P-value Adjusted OR (95% CI) P-value
      HTS (−) 133 21 (15.8, 10.0–23.1) 1.00 (reference) 1.00 (reference)
      HTS (+) 1 30 3 (10.0, 2.1–26.5) 0.59 (0.13–1.88) 0.423 0.51 (0.11–1.75) 0.325
      HTS (+) 2 15 8 (53.3, 26.6–78.7) 6.10 (1.99–19.19) 0.002 5.43 (1.65–18.22) 0.005
      HTS (+) ≥3 22 10 (45.5, 24.4–67.8) 4.44 (1.68–11.68) 0.002 3.59 (1.25–10.27) 0.016
      P for trend <0.001 0.002
      Variable Adjusted OR (95% CI) P-value
      Hypertonic saline use 2.21 (1.03–4.75) 0.040
      Age (per 1-year increase) 0.99 (0.95–1.03) 0.529
      Male sex 1.16 (0.53–2.53) 0.709
      Initial NIHSS (per 1-point increase) 1.04 (0.98–1.10) 0.225
      Pre-stroke mRS (per 1-point increase) 1.11 (0.49–2.07) 0.764
      Reperfusion: IVT (vs. none) 2.11 (0.79–5.45) 0.127
      Reperfusion: EVT (vs. none) 1.06 (0.35–2.99) 0.909
      Reperfusion: IVT+EVT (vs. none) 1.00 (0.23–3.57) 0.996
      TOAST: cardioembolism 0.30 (0.07–1.16) 0.086
      Prior antiplatelet use 0.51 (0.15–1.40) 0.222
      Hypertension 0.89 (0.40–2.00) 0.772
      Diabetes mellitus 0.94 (0.39–2.22) 0.895
      Atrial fibrillation 2.18 (0.57–8.82) 0.258
      Table 1. Baseline characteristics according to hypertonic saline use

      Values are presented as mean±standard deviation, number (%), or median (interquartile range).

      HTS: hypertonic saline; NIHSS: National Institutes of Health Stroke Scale; mRS: modified Rankin Scale; IVT: intravenous thrombolysis; EVT: endovascular treatment; TOAST: Trial of Org 10172 in Acute Stroke Treatment; LAA: large artery atherosclerosis; CE: cardioembolism; OD: other determined etiology; UD: undetermined etiology; WBC: white blood cell; BUN: blood urea nitrogen; LDL-C: low-density lipoprotein cholesterol; HDL-C: high-density lipoprotein cholesterol; HbA1c: glycated hemoglobin; BP: blood pressure.

      Student t-test was used for continuous variables. Fisher's exact test with Monte Carlo simulation was used for categorical variables.

      Table 2. Clinical outcomes according to hypertonic saline use

      Values are presented as number (%, 95% CI by exact binomial method) or median (IQR). Log-rank test was used for mortality (follow-up, 91 days).

      HTS: hypertonic saline; mRS: modified Rankin Scale; LOS: hospital length of stay; IQR: interquartile range.

      Chi-square test for phlebitis; Fisher’s exact test for 3-month mRS 0-3; Mann-Whitney U-test for LOS; Log-rank test for in-hospital mortality.

      Table 3. Association between hypertonic saline use and clinical outcomes

      Reference group: no-HTS. Adjusted for age, initial National Institutes of Health Stroke Scale score, pre-stroke modified Rankin Scale score, previous antiplatelet treatment, reperfusion therapy, and TOAST classification.

      OR: odds ratio; HR: hazard ratio; LOS: hospital length of stay; TOAST: Trial of Org 10172 in Acute Stroke Treatment.

      Length of stay was log-transformed prior to analysis; β represents the difference in log-transformed days between groups.

      Table 4. Association between frequency of hypertonic saline administration and phlebitis

      Crude odds ratios were estimated using logistic regression with no-HTS as the reference group. Adjusted odds ratios were derived from multivariable logistic regression adjusting for age, initial NIHSS score, pre-stroke modified Rankin Scale score, previous antiplatelet treatment, reperfusion therapy (none/IVT/EVT/IVT+EVT), and TOAST classification. The crude P for trend was calculated using the Cochran-Armitage trend test. The adjusted P for trend was obtained by including HTS frequency as an ordinal variable (0, 1, 2, ≥3) in the multivariable logistic regression model.

      OR: odds ratio; HTS: hypertonic saline; NIHSS: National Institutes of Health Stroke Scale; IVT: intravenous thrombolysis; EVT: endovascular treatment; TOAST: Trial of Org 10172 in Acute Stroke Treatment.

      Table 5. Independent association of hypertonic saline use with phlebitis

      Multivariable logistic regression model with phlebitis as the dependent variable. All listed covariates were entered simultaneously. Hypertonic saline use was the primary exposure variable.

      OR: odds ratio; NIHSS: National Institutes of Health Stroke Scale; mRS: modified Rankin Scale; IVT: intravenous thrombolysis; EVT: endovascular treatment; TOAST: Trial of Org 10172 in Acute Stroke Treatment.


      ACC : Acute and Critical Care
      TOP