Safety and tolerability study of clonidine for sedation in mechanically ventilated adult patients

Abstract

Background

Clonidine is used as an adjunct sedative in mechanically ventilated ICU patients, but prospective data on the safety and tolerability of higher intravenous doses remain limited.

Methods

This secondary analysis used data from a prospective pharmacokinetic study in 32 critically ill adults. Patients were consecutively enrolled into four dose groups (0, 600, 1200 or 1800 μg/day, 8 per group) and received continuous intravenous clonidine added to standard sedation for up to seven days. Half of the patients in the clonidine groups underwent a four-hour loading phase during which the infusion rate was tripled. Serious adverse events (SAEs), a predefined set of adverse events (AEs), and outcome parameters were recorded. The relatedness of SAEs to clonidine was assessed using the WHO-UMC causality classification system. Patients were monitored for 48 h after clonidine discontinuation.

Results

Ten SAEs were reported (3, 3, 3 and 1 in the 0, 600, 1200 and 1800 μg/day dose groups, respectively; Bayes factor (BF) 0.23). Seven SAEs were considered unrelated to clonidine, while three were judged possibly related. No SAEs occurred during the loading phase. No dose-dependent relationship was observed between clonidine dose and bradycardia or hypotension (BF 0.39 and 0.40, respectively). Pleural effusions were more frequent in the control group than in the clonidine groups (BF 12.9). Days on sedation, duration of mechanical ventilation, days in ICU and hospital mortality did not differ between groups. No clinically meaningful tachycardia or hypertension suggestive of clonidine withdrawal was observed.

Conclusion

Clonidine was well tolerated within the studied dose range when used as an adjunct to standard sedation in our cohort of non-bradycardic, mechanically ventilated, critically ill adults. No clear dose-dependent increase in predefined adverse events was observed within the studied dose range. These findings provide support for further prospective evaluation of clonidine at doses up to 1800 µg/day in future efficacy studies.

Trial registration

ClinicalTrials.gov NCT02466373.

Supplementary Information

The online version contains supplementary material available at https://doi.org/10.1186/s44158-026-00447-y.

Background

Clonidine was originally developed in the 1960s as an antihypertensive agent. It crosses the blood–brain barrier and acts on brainstem nuclei that regulate sympathetic activity [1]. By stimulating central α2-adrenoceptors, it reduces sympathetic tone and enhances vagal activity, lowering blood pressure and heart rate [2]. Clonidine also decreases general arousal and produces sedation, analgesia and anxiolysis [3, 4].

When clonidine is used to treat hypertension, common adverse reactions include sleepiness, drowsiness and weakness, particularly at higher doses [5]. However, in settings where central depression is desired, such as during invasive procedures [6, 7] or mechanical ventilation [8] this relationship is reversed: sedation becomes the therapeutic effect, while bradycardia and hypotension become potential adverse reactions. Using clonidine for its sedative effects is off-label in most countries.

In some intensive care units (ICUs), intravenous clonidine is used as an adjunct to sedation and analgesia [9], with the aim of reducing requirements for other sedatives or potentially decreasing delirium [10, 11]. However, this practice has evolved without formal dosing studies, resulting in substantial variability. In Dutch ICUs, doses varied from 240 to 2400 μg/day [9], while systematic reviews of clonidine for ICU sedation reported intravenous doses ranging from 50 to 3360 μg/day, with inconsistent use of loading doses [8, 12].

This variability may reflect the absence of formal dose finding studies. For new sedative agents, phase 1 trials typically use escalating doses in closely monitored patients to identify the maximum tolerated dose, defined as the highest dose that does not produce harmful side effects. Because clonidine was not initially developed as a sedative, no such dose–response evaluation exists in critically ill adults.

The sedative effects of clonidine seem to occur at higher plasma concentrations than those needed for blood pressure control. For hypertension, plasma concentrations around 1 µg/L are generally sufficient, [13], typically achieved with oral doses of 200–600 µg/day [14]. Substantially higher exposures are necessary to produce meaningful sedation. In mechanically ventilated children, plasma concentrations of 2 µg/L have been suggested as a target for adequate sedation [15]. Likewise, in a controlled study of healthy adults, Hall and colleagues found that clonidine concentrations averaging 1.96 µg/L produced noticeable sedation, yet still light enough that participants could be awakened by calling their name [3].

Given the absence of consensus on appropriate dosing for sedation and the potential for cardiovascular adverse effects at higher exposures, dose-finding studies of clonidine for sedation in adult ICU patients remain necessary. In the present study, incremental doses of clonidine were administered as adjunctive sedation in mechanically ventilated adults, to evaluate safety and tolerability within a predefined dosing range. The dose levels were based on previously published pharmacokinetic and clinical data [8, 12].

Methods

Study design

This study represents a planned secondary analysis of a prospective dose-escalation project, of which the population pharmacokinetic analysis was reported previously by Cloesmeijer et al. [12]. The project was registered in ClinicalTrials.gov (NCT02466373) and EudraCT (2015–001699-23), approved by the Medical Ethics Review Committee (METC Isala Zwolle), and conducted in accordance with the principles of the World Medical Association Declaration of Helsinki.

The original project was an open-label exploratory dose-escalation study. Mechanically ventilated patients received predefined escalating doses of clonidine and underwent continuous monitoring as part of standard ICU care during a nine-day observation period that included the wash-out phase. Frequent haemodynamic and neurocognitive assessments, plasma concentrations and predefined adverse events (AEs) were collected concurrently. Conducting these investigations within a single prospective study allowed several complementary research questions to be addressed while avoiding repeated exposure of critically ill patients to the study intervention.

Whereas the previously published population pharmacokinetic analysis [12] included only the 24 clonidine-treated patients, the present analysis includes all 32 enrolled patients. The 0 µg/day dose group served as a control because the substantial burden of AEs in critically ill patients would otherwise have obscured interpretation of dose-related differences.

Participants

Between January 2017 and April 2018, 32 mechanically ventilated patients were enrolled in the mixed medical-surgical ICU of Deventer Hospital. Eligible patients were expected to remain ventilated for at least three days and were included after obtaining informed proxy consent. At the time of inclusion, all patients were receiving standard sedation, consisting of continuous intravenous morphine and midazolam, with propofol available as an alternative. Sedation was interrupted daily and restarted as clinically indicated. In cases of hyperactive delirium with restlessness or hallucinations, haloperidol could be prescribed at the discretion of the attending physician.

Exclusion criteria included age under 18 years, neurotrauma, post-anoxic coma, clonidine use within 96 h prior to inclusion, known clonidine intolerance, bradycardia, severe hypotension, hypertensive emergency, pregnancy or lactation, epilepsy, liver cirrhosis, recent or acute myocardial infarction, severe heart failure, second- or third-degree atrioventricular block without a permanent pacemaker, and renal failure. The study cohort was identical to that described previously [12].

Study procedure

Patients were consecutively assigned to four dose groups, each comprising eight subjects. These groups received either 0 µg/day (control), or continuous intravenous clonidine infusions at 600, 1200 or 1800 μg/day, corresponding to infusion rates of 25, 50 and 75 μg/h, respectively.

Clonidine was discontinued after a maximum of 7 days, 48 h prior to anticipated ICU discharge, or earlier at the discretion of the treating physician. To detect possible signs of clonidine withdrawal, follow-up continued from the start of clonidine infusion until 48 h after its discontinuation. In patients who did not receive clonidine, the time of study inclusion was taken as the presumed start of the ‘study drug’.

In the last four patients of each dose group, the clonidine infusion started with a loading dose in which 50% of the total daily dose was infused over four hours. This resulted in temporary infusion rates of 75, 150 and 225 μg/h, respectively, during the loading period.

Clonidine hydrochloride (Catapresan®, 150 µg/mL, 1-mL ampoules; Boehringer Ingelheim, Dutch registration number RVG 06055) was diluted with 0.9% sodium chloride to prepare patient-specific infusion syringes. The hospital pharmacy made syringes containing clonidine concentrations of 12.5, 25.0, and 37.5 µg/mL for the 600, 1200 and 1800 µg/day dose groups, respectively. For the loading dose, 50-mL syringes containing 24 mL of diluted clonidine were used, whereas syringes containing 48 mL of diluted clonidine were used for maintenance infusions. Patients allocated to the 0 µg/day group did not receive placebo infusions. Clonidine was administered intravenously via a dedicated lumen of either a peripheral or central venous catheter at a constant infusion rate of 2 mL/h during maintenance and 6 mL/h during the loading phase.

Data collection

At baseline, demographic and clinical characteristics were recorded, including age, sex, height, weight, admission diagnosis, and comorbidities as defined by the Charlson Comorbidity Index (Additional file 1) [16]. Severity of illness was assessed using the Acute Physiology and Chronic Health Evaluation (APACHE) IV score (Additional file 2) [17]. After initiation of clonidine infusion, the following assessments were performed three times daily: Richmond agitation and sedation scale (Additional file 3) [18], confusion assessment method score for ICU patients (CAM-ICU; Additional file 4) [19], mechanical ventilation status (hours with or without ventilation), and hemodynamic parameters including heart rate (HR) and systolic, diastolic and mean arterial blood pressures (SAP, DAP and MAP, respectively). Medication administration was also recorded three times daily, including doses of midazolam, morphine, fentanyl, propofol, haloperidol and other antipsychotics. Once daily, (serious) adverse events were documented.

Serious adverse events

A serious adverse event (SAE) was defined as “any untoward medical occurrence in a patient or trial subject, which does not have a causal relationship with the treatment, and is fatal, and/or is life-threatening for the subject, and/or makes hospital admission or an extension of the admission necessary, and/or causes persistent or significant invalidity or work disability, and/or manifests itself in a congenital abnormality or malformation, and/or could, according to the person that carries out the research, have developed to a serious undesired medical event, but was however prevented due to premature interference” [20]. SAEs were reported to the principal investigator by the attending physicians, and subsequently forwarded to the sponsor and to the competent authority in the Netherlands, the Central Committee on Research Involving Human Subjects (CCMO) in The Hague, in accordance with national regulations. The potential relationship of each SAE to the investigational treatment was initially assessed by the principal investigator using clinical judgement, consistent with Good Clinical Practice. In a second step, on behalf of the sponsor, two clinical pharmacologists (MZ and MEA) reviewed all individual SAE narratives, and performed a structured causality assessment using the WHO-UMC system [21], which classifies causality as certain, probable/likely, possible or unlikely, handling disagreements by consensus.

Adverse events

Investigators screened the electronic medical records for adverse events. Given the large number of monitored parameters and the complex physiology in ICU patients, a list of 45 predefined potential adverse events, previously used in related research [22], was employed. This list covered electrolyte, gastrointestinal, cardiovascular, pulmonary, renal, haematological and other disturbances (Additional file 5). If an AE occurred multiple times in the same patient, it was recorded only once.

Data handling and statistics

Data were collected using Microsoft Access, and analysed using Bayesian methods in R (version 4.5.0; R foundation for Statistical Computing, Vienna, Austria). Continuous baseline parameters, adverse events and endpoint measures were evaluated across the four clonidine dose groups using Bayesian ANOVA; binomial values were evaluated using Bayesian contingency tables analysis.

Bayesian statistics were chosen due to the small group sizes, and because the aim was to explore potential dose–response relationships and quantify their strength. The strength of support for the alternative hypothesis (i.e. that values are drawn from different populations) over the null hypothesis (i.e. that values originate from the same population) was expressed as Bayes factor (BF). A BF around 1 indicates no preference between hypotheses; values > 3 suggest increasing support for the alternative hypothesis, while values < 1/3 suggest increasing support for the null hypothesis.

Results

Study cohort

Baseline characteristics

The study included 32 patients (21 men and 11 women) with a median age of 67 years. The median APACHE IV score was 81, with scores evenly distributed across clonidine dose groups. Intravenous sedation was required for study inclusion. Midazolam was the predominant sedative at the time of the study, while propofol was administered in 37.5% of patients, usually with the intent to reduce benzodiazepine exposure. There were no statistically supported differences between dose groups in the proportion of patients receiving sedative or antipsychotic medications at baseline. Most patients (23 out of 32) received noradrenaline, without strong support for differences in noradrenaline requirements between groups at baseline (BF 2.92). Dobutamine was not used in any of the patients. Beta-blocker use was evenly distributed across dose groups. There were no apparent differences in admission diagnoses or comorbidities. Patient characteristics are summarised in Table 1.

Table 1.

Baseline characteristics, sedative medication, admission diagnosis, and comorbidity (Charlson Comorbidity Index) of 32 patients receiving add-on sedation with varying doses of clonidine for a maximum of seven days. Data are presented as number (n) or median [interquartile range]. Differences between groups were analysed using Bayesian ANOVA for continuous variables and contingency tables BF for binomial values

Patient characteristics	All patients n = 32	No Clonidine n = 8	Clonidine 600 µg/day n = 8	Clonidine 1200 µg/day n = 8	Clonidine 1800 µg/day n = 8	Bayes factor
At baseline
Sex (male/female)	21/11	5/3	7/1	4/4	5/3	0.65
Age (years)	67 [54–75]	59 [51–75]	66 [60–75]	68 [58–79]	65 [55–77]	0.19
Body weight (kg)	83 [72–90]	77 [69–90]	86 [73–102]	90 [82–94]	72 [66–84]	0.75
Height (cm)	174 [168–184]	174 [168–178]	185 [171–188]	170 [166–180]	171 [161–184]	0.58
Body mass index (kg/m2)	27 [23–31]	27 [22–30]	26 [23–31]	30 [26–34]	25 [22–31]	0.82
APACHE IV	81 [66–101]	86 [70–103]	85 [68–99]	90 [73–110]	74 [56–98]	0.22
Sedative medication
Midazolam	21	5	6	7	3	1.63
Propofol	12	4	3	2	3	0.30
Morphine	26	5	7	7	7	0.67
Haloperidol	13	5	3	2	3	0.54
Vasoactive medication
Noradrenaline	23	5	6	8	4	2.92
Dobutamine	0	0	0	0	0	1.00
Beta-blocker (metoprolol)	7	1	2	2	2	0.32
Admission diagnosis
Pneumonia (bacterial)	10	4	3	1	2
Pneumonia (viral)	1	0	0	0	1
Pneumonia (other)	2	1	0	1	0
Emphysema/bronchitis	4	0	2	1	1
Sepsis (abdominal)	5	3	1	1	0
Sepsis (unknown cause)	2	0	0	1	1
Abdominal aortic aneurysm	3	0	1	1	1
Surgery for GI perforation/rupture	3	0	1	1	1
Congestive heart failure	1	0	0	1	0
Other	1	0	0	0	1
Comorbidity
Chronic pulmonary disease	9	1	3	3	3
Tumour without metastases	3	1	1	0	2
Myocardial infarction	2	1	1	0	2
Peripheral vascular disease	2	0	1	0	2
Diabetes, no end-organ failure	3	1	0	3	0
Connective tissue disease	1	1	0	1	0
Lymphoma	1	0	0	1	0
Cerebrovascular disease	1	1	0	1	0
Peptic ulcer disease	1	1	0	0	0
Leukaemia	1	1	0	0	0
Tumour with metastases	1	1	0	0	0

Primary outcome parameters

Serious adverse events

During the study period, the principal investigator reported ten SAEs to the regulatory authority (Table 2). None of these occurred during the clonidine loading phase. WHO-UMC causality assessment indicated that seven were unlikely related to clonidine and three were possibly related; no events met criteria for a suspected unexpected serious adverse drug reactions (SUSAR).

Table 2.

Serious adverse events by clonidine dose group. Causal relationship to study medication was assessed by two clinical pharmacologists

Clonidine dose group	Serious adverse event	SAE criterion	WHO-UMC causality assessment
0 µg/day (controls)	Pleural empyema complicating pneumococcal pneumonia, multiple chest drainage required	Extended hospital admission, survived to discharge	Causality unlikely (received no study medication)
	Pleural empyema following subphrenic abscess complicating liver surgery, drainage and reintubation required	Life-threatening complication, resulting in death	Causality unlikely (received no study medication)
	Recurrence of acute myeloid leukemia, presenting with pneumococcal pneumonia	Life-threatening complication, survived to discharge	Causality unlikely (received no study medication)
600 µg/day	Spontaneous caecal perforation during COPD exacerbation, twelve hours after starting study drug	Life-threatening complication, survived to discharge	Causality possible
	Reintubation several hours after extubation in pneumococcal pneumonia	Extended hospital admission, survived to discharge	Causality unlikely
	Bowel necrosis complicating acute aortic surgery, symptoms started before study drug administration	Life-threatening complication, resulting in death	Causality unlikely
1200 µg/day	Reintubation one day after extubation in COPD exacerbation	Extended hospital admission, survived to discharge	Causality possible
	Abdominal dehiscence after emergency laparotomy for perforated diverticulitis	Extended hospital admission, survived to discharge	Causality unlikely
	Multiple organ failure complicating acute pancreatitis	Life-threatening complication, resulting in death	Causality unlikely
1800 µg/day	Cardiac arrest; brady-arrhythmia during central line placement	Life-threatening, survived to discharge	Causality possible

Abbreviations: COPD chronic obstructive pulmonary disease, SAE Serious adverse event, WHO-UMC World Health Organisation-Uppsala Medical Centre causality assessment system

Bayesian contingency table analysis provide strong evidence in favour of no difference in the total number of SAEs between clonidine dose groups (BF 0.23), suggesting that SAE frequency was independent of clonidine dose. Two of the 32 patients required transfer to a tertiary academic ICU for escalation of care.

Taken together, these findings suggest that dose-limiting toxicity was not observed within the predefined dose-escalation range.

Adverse events

During the nine-day observation period, 349 predefined AE categories were reported in 32 patients, with a median of 11 [IQR 9.5–14] categories per patient. The complete list is provided in additional file 6 in the supplementary information files, while Table 3 shows a selection of the most relevant AEs. Frequent AEs were hypotension (MAP ≤ 65 mmHg), tachycardia (HR > 100/min), arrhythmia (other than brady- or tachycardia), hypoxaemia (arterial PO₂ ≤ 9.0 kPa), pleural effusion (radiologist or attending physician report), hypercapnia (arterial PCO₂ ≥ 6.2 kPa), and delirium (positive CAM-ICU score). Of these, only pleural effusion showed a non-uniform distribution, occurring more often in the eight control patients than in the 24 patients receiving clonidine (BF 12.9). Bradycardia (HR < 60/min) was reported as an AE seven times. With higher clonidine dose, tachycardia appeared less frequent and bradycardia more frequent, but BFs did not support these trends as meaningful. Clonidine infusion was never interrupted for safety concerns. A full list of AEs and their occurrence by dose group is shown in the Additional file 6.

Table 3.

A selection of predefined adverse events, rebound phenomena, treatment duration and mortality by dose group in 32 patients receiving add-on sedation with varying doses of clonidine for up to seven days. The complete list of predefined adverse events by dose group is presented online in the Supplemental information, file 6. Values in parentheses indicate events occurring in the control group who did not receive clonidine. Deep sedation was defined as days with a Richmond Agitation and Sedation Scale score of −4 or −5. Data are presented as n (number) or median [interquartile range]. Differences between groups were analysed using Bayesian ANOVA for continuous variables and Bayesian contingency-table analysis for categorical variables

Outcome parameter	All patients n = 32	No clonidine n = 8	Clonidine 600 µg/day n = 8	Clonidine 1200 µg/day n = 8	Clonidine 1800 µg/day n = 8	Bayes factor
Hypoxaemia	24	5	7	7	5	0.26
Hypercapnia	16	4	5	3	4	0.20
Pleural effusion	20	8	3	6	3	12.9
Atelectasis	12	2	3	5	2	0.46
Delirium	16	6	3	4	3	0.49
Hypotension	28	7	8	8	5	0.40
Hypertension	6	2	1	2	1	0.09
Tachycardia	24	8	7	5	4	1.69
Bradycardia	7	0	2	2	3	0.39
Arrhythmia	13	2	5	3	3	0.36
Rebound  hypertension	3	(0)	1	0	2	0.11
Rebound  tachycardia	9	(2)	3	3	3	0.13
Days in ICU	13 [9–21]	13 [10–28]	13 [9–20]	20 [8–24]	11 [8–17]	0.35
Days on ventilation	8 [5–12]	8 [7–20]	9 [8–14]	7 [3–14]	5 [5–10]	0.29
Days on deep sedation	5 [3–8]	7 [2–9]	8 [4–9]	5 [3–6]	3 [2–4]	1.25
In-hospital mortality	6	2	2	2	0	0.88

Apart from bradycardia reported as an AE, a database search identified eight HR measurements below 50 bpm out of 783 observations. All instances of severe bradycardia occurred outside the loading phase. Seven of these came from two unique patients and were associated with concomitant β-blocker use.

Rebound phenomena after clonidine withdrawal

We specifically assessed for rebound tachycardia and hypertension during the two days following clonidine discontinuation, as these were potential adverse effects related to α2-agonist withdrawal. Tachycardia was defined as HR > 100 bpm, and hypertension as SAP ≥ 180 mmHg or DAP ≥ 110 mmHg.

Mild rebound tachycardia (peaking at 110 bpm) occurred three times in each clonidine dose group. Mild rebound hypertension (maximum SAP 185 mmHg) was observed once in the 600 µg/day group and twice in the 1800 µg/day group. For comparison, among patients who received no clonidine, two tachycardic episodes and no hypertension were recorded during the final two days of observation. Bayesian contingency table analysis indicated strong support for no difference between clonidine dose groups (Table 3).

Treatment duration and mortality

Patients had a median ICU stay of 13 days, with no apparent differences between the clonidine dose groups. Two patients died during the study follow-up period (reported as SAE), and two additional patients died in the hospital after the follow-up period ended (not reported as SAE). Treatment duration and mortality are presented in Table 3.

Concomitant medication

The daily use of noradrenaline, metoprolol, midazolam, propofol, morphine, and haloperidol was recorded throughout the nine days of follow-up (Additional file 7). There were no discernible differences in the consumption patterns of these agents between the clonidine dose groups.

Discussion

The main finding of this exploratory safety, tolerability and dose-finding study of clonidine used as adjunctive sedation in mechanically ventilated, critically ill adults is that no evidence of dose-limiting toxicity was observed within the investigated dose range up to 1800 μg/day. We found no evidence of an association between clonidine exposure and the occurrence of AEs or SAEs. Specifically, the incidence of hemodynamic events commonly attributed to clonidine, such as hypotension and bradycardia, did not differ significantly across dose levels. In the clonidine arms, half of the patients received a four-hour loading dose, with temporary infusion rates up to 225 μg/h. Even during these periods of relatively high drug exposure, no increase in bradycardia, hypotension, or other safety events was observed, and the infusions were never interrupted for safety reasons. A signal suggesting a potential protective effect against pleural effusion was found; however, in the absence of a plausible pharmacological mechanism, this finding is most likely attributable to chance. Overall, clonidine was well tolerated at doses up to 1800 µg/day in this cohort. Although this exploratory study cannot exclude uncommon adverse events, these findings provide support for further evaluation in future efficacy trials.

The question arises whether the upper limit of 1800 μg/day was sufficiently high to detect dose-related side effects. At the time the study was designed, this dose range was consistent with previously published data in adult ICU patients. Spies et al. evaluated a combination of flunitrazepam and clonidine to reduce withdrawal symptoms in alcohol-dependent postoperative ICU patients and reported a median clonidine dose of 0.83 μg/kg/h (approximately 1490 μg/day for a 75 kg person), with occasional use of higher doses but without an increased incidence of cardiac events [10]. Other studies have used doses above 1800 mcg/day, but typically in younger, less critically ill populations [11], or they reported increased rates of bradycardia and hypotension at higher infusion rates [23]. In our study, clonidine was intended as an adjunct to standard sedation rather than a sole sedative. During protocol development, patient safety was given priority. Given the high severity of illness in our cohort, we limited the maximum dose of clonidine to 1800 μg/day to reduce the risk of hemodynamic complications.

Although we anticipated that patients receiving clonidine would require less additional sedation, and perhaps fewer antipsychotics, this was not observed in our cohort. This contrasts with findings from the recently published A2B trial, in which patients receiving clonidine at doses of 1245 to 1820 µg/day demonstrated markedly reduced propofol requirements [24]. Several factors may explain this discrepancy. Most importantly, the A2B trial employed a strongly protocolised sedation strategy targeting mostly light sedation, whereas sedation management in our study was left to the discretion of the attending clinical team. In addition, our dose groups were small, whereas the A2B trial randomised more than 400 patients to clonidine, increasing the power to detect differences in sedative requirements.

We had also hypothesized that higher clonidine doses might be associated with greater noradrenaline requirements and reduced beta-blocker use, given its sympatholytic action. However, neither of these patterns emerged. With a median APACHE IV score of 81, this was a relatively ill cohort, at substantial risk of hemodynamic instability. Baseline characteristics were well balanced across the four clonidine dose groups, as were the duration of sedation, ventilation, and ICU stay, and hospital mortality, indicating that the groups were clinically comparable at the start of the study and moved through the ICU trajectory in a comparable fashion. Together, these findings suggest that the absence of dose–response relationships in AEs and SAEs was not masked by compensatory hemodynamic management. In other words, the lack of differences in AE/SAE rates across dose levels cannot be attributed to increased vasopressor support or altered beta-blockade in the higher dose groups, making this explanation less likely.

Our findings also differed from those of the A2B trial regarding the incidence of bradycardia. We defined bradycardia as a HR below 60 bpm and encountered it as an AE in 7 of 24 patients (29%). In contrast, Walsh et al. reported severe bradycardia (HR < 50 bpm) in 33% of patients, which frequently prompted dose reduction or discontinuation of clonidine [24]. For comparison, we assessed severe bradycardia (HR < 50 bpm) in our cohort and identified it in only three patients (8%), mainly in the context of concomitant beta-blocker therapy. In our study, bradycardia was never a reason to interrupt clonidine infusion. This discrepancy with the A2B results is not readily explained by differences in population characteristics, sedation targets, or clonidine exposure patterns, while concomitant beta-blocker use in the A2B cohort was not reported. Importantly, a low incidence of bradycardic or hypotensive adverse events at higher clonidine exposures does not exclude the presence of dose-dependent hemodynamic effects.

Before the trial began, there was particular concern regarding potential rebound tachycardia and rebound hypertension after clonidine discontinuation, prompting us to extend haemodynamic monitoring for 48 h following drug withdrawal. Despite this proactive surveillance, neither rebound tachycardia nor rebound hypertension was observed to a clinically meaningful extent. Observed fluctuations were mild, transient and comparable across dose groups. These findings suggest that the clonidine regimens used in our study were not associated with pronounced withdrawal phenomena in our cohort of 32 patients.

To our knowledge, this study represents the first systematic dose-escalation evaluation of clonidine for sedation in mechanically ventilated, critically ill adults, conducted with detailed safety monitoring. Dose-finding studies are rare in ICU settings, but the high density of data points and the precision afforded by invasive and continuous monitoring are clear advantages; there were almost no missing observations in our database. The study design enabled us to assess both acute haemodynamic responses and potential withdrawal phenomena after discontinuation. Because adverse events can be difficult to classify in critically ill patients, the use of predefined criteria further enhanced the internal validity of the findings. The combination of a limited sample size, escalating dose levels, and highly granular safety observations resembles a phase 1 design. At the same time, the present analysis originated from a pharmacokinetic study rather than a formally designated drug-development trial, and the findings should therefore be interpreted within that context.

Limitations include the single-centre design and the exclusion of patients with severe cardiovascular instability, both of which may limit generalisability to broader ICU populations. The sample size was not powered to detect uncommon but potentially important adverse events. For example, tachycardia occurred less frequently and bradycardia more frequently at higher clonidine exposures, but these numerical trends were not supported by the statistical analyses. Consequently, limited statistical power cannot be distinguished from true absence of effect, and the possibility of type II error should be considered when interpreting negative findings. Another limitation is that patients contributed different observation times because ICU discharge frequently occurred before completion of the planned nine-day follow-up. The contingency table analyses used to compare individual adverse events between dose groups did not account for these differences in exposure time. Consequently, these comparisons should be regarded as exploratory. Furthermore, clonidine was used strictly as an adjunct rather than a sole sedative. This approach may enhance real-world relevance but may limit conclusions about clonidine as a single agent. Sedation intensity was not protocolised, which may reduce comparability with modern sedation trials. Despite these constraints, the safety findings were internally consistent across dose groups and indicate that further prospective evaluation of clonidine within the studied dose range is justified.

Subsequent efficacy-oriented trials will need to establish the clinical utility and optimal dosing of clonidine, as the highest tolerated dose is not necessarily the optimal dose in clinical practice. Clonidine, for example, has been associated with reduced delirium in some studies [11], but this potential benefit can only be meaningfully evaluated if doses are sufficient to produce a pharmacological effect. Future efficacy studies should therefore consider including doses that are expected to result in pharmacologically relevant drug exposure. In the context of the present findings, doses of up to 1800 μg/day appear appropriate for further evaluation, although lower doses may ultimately provide comparable efficacy with improved tolerability. The present study provides a foundation for such investigations by defining a dose range that appears suitable for prospective efficacy studies.

Conclusion

In this exploratory safety and tolerability study, clonidine was well tolerated at doses up to 1800 μg/day when used as an adjunct to standard sedation in our cohort of mechanically ventilated, critically ill adults. No evidence of dose-limiting toxicity was observed within the studied dose range, although the sample size was insufficient to exclude uncommon adverse events. Hemodynamic adverse events, including hypotension and bradycardia, did not appear to increase with clonidine dose and did not require treatment interruption. Rebound tachycardia and hypertension after drug discontinuation were infrequent and clinically mild. These findings provide support for further prospective evaluation of clonidine at doses up to 1800 µg/day in future efficacy studies.

Abbreviations

AE

Adverse event

APACHE IV

Acute physiology and chronic health evaluation, version IV

BF

Bayes factor

CAM-ICU

Confusion assessment method for the ICU

CCMO

Central committee on research involving human subjects in the Netherlands

DAP

Diastolic arterial blood pressure

HR

Heart rate

ICU

Intensive care unit

MAP

Mean arterial blood pressure

SAE

Serious adverse event

SAP

Systolic arterial blood pressure

Authors’ contributions

All authors contributed to the study conception and design. Data collection was done by HvdO, AKG, PN, CB, and AdM. Data analysis was done by EK, HvdO, MA, and MZ. The first draft of the manuscript was written by HvdO, EK, and MA. All authors read and approved the final manuscript.

Funding

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Data availability

The datasets generated and analysed during the current study contain individual patient data and are not publicly available due to ethical and privacy restrictions. De-identified data may be made available from the corresponding author upon reasonable request and subject to approval by the institutional review board and data transfer agreement.

Declarations

Ethics approval and consent to participate

This study was conducted in accordance with the World Medical Association Declaration of Helsinki. The Medical Ethics Committee of Isala Hospitals, Zwolle, Netherlands, approved the study protocol.

Consent for publication

Each subject or his or her legal representative gave written informed consent for participation in the study and publication of the results prior to study enrolment.

Competing interests

The authors declare no competing interests.