Clinical manifestations of acute neurological conditions such as stroke and traumatic brain injuries (TBI), can often include low arousal, impaired respiration control, secretion retention, and dysphagia^1,2) These symptoms may result in the need for mechanical ventilation³⁾. In complex neurological presentations, a tracheostomy is required to facilitate weaning from ventilation and discharge from the intensive care unit in 14–47% of patients following a TBI⁴⁾ and up to 45% of stroke patients⁵⁾.

Decannulation, the process of removing a tracheostomy, is influenced by airway secretions, and neurological populations may present with increased secretion volumes due to dysphagia⁶⁾. Pharmacological drying agents such as glycopyrrolate reduce saliva production⁷⁾ and are commonly used in specific neurological populations, i.e. cerebral palsy⁸⁾ and amyotrophic lateral sclerosis⁹⁾. Progression towards decannulation is important to reduce the risk of complications from a long–term tracheostomy, such as dislodgement, haemorrhage, and obstruction¹⁰⁾. Prediction tools for decannulation in complex neurological conditions¹¹⁾ and clinical assessments i.e. fibreoptic endoscopy¹²⁾ have been developed, however, there is limited discussion about the use of drying agents when weaning a patient from tracheostomy⁶⁾.

The four drying agents used in this study were anti–cholinergic agents, including glycopyrronium bromide (glycopyrrolate), atropine sulphate (atropine), hyoscine butylbromide (hyoscine), and ipratropium bromide (atrovent). These agents act by blocking and inhibiting the activity of acetylcholine at muscarinic and nicotinic receptors in the central and peripheral nervous systems, reducing the activity of the parasympathetic nervous system and potentially acting on many areas of the body including the gastrointestinal tract, lungs, and urinary tract¹³⁾. Indications for the use of these medications in neurological populations can be to reduce the production of saliva by inhibiting the action of acetylcholine at the M1 and M3 muscarinic receptors in the glands⁶⁾. Other anti–cholinergic agents, including benzatropine, trihexphenidyl, and propatheline, and tricyclic antidepressants such as amitriptyline can also be used for excessive secretions in neurological populations^6,14), however these were not utilised in this study.

Caution must be taken when prescribing these anti–cholinergic medications, as there are both central and peripheral side effects that can affect patients with a neurological diagnosis. Central side effects of anti–cholinergic medications include reduced cognitive function, behavioural disturbances, reduced motor function, and altered emotions¹⁵⁾. Peripheral side effects include a dry mouth, dry eyes, tachycardia, reduced peristalsis and constipation¹⁵⁾. The use anti–cholinergic drying agents is contraindicated in patients with glaucoma, where the inhibition of muscarinic receptors in the iris can lead to complete closure of the iridocorneal angle¹⁶⁾. They must also be prescribed cautiously in patients with urinary or gastrointestinal disturbances, due to their impact on the muscarinic receptors in smooth muscle¹⁷⁾.

A better understanding of the role of pharmacological drying agents in tracheostomised patients may facilitate timely and successful decannulation. The aim of this study was 1) to investigate the use of pharmacological drying agents in the weaning of tracheostomies in neurological patients and 2) to describe the clinical criteria in patients requiring pharmacological drying agents in their weaning plan.

We hypothesised that patients requiring drying agents would be more complex therefore would have a longer tracheostomy duration, longer length of stay and likely an infratentorial brain injury.

A retrospective, observational study was conducted between 2014 to 2019 at a tertiary teaching hospital in Melbourne, Australia. It was estimated that there would be approximately 20 tracheostomised patients per year admitted under the neurosciences team. Patients were included in the study if they were admitted on the acute neurosciences ward under a stroke, neurology, or neurosurgery bedcard, and had a temporary tracheostomy inserted in the intensive care unit (ICU) to facilitate weaning from mechanical ventilation. Patients were excluded if they were tracheostomised but not managed under a stroke, neurology, or neurosurgery bedcard. Institutional quality assurance approval was obtained for the study (QA2020042).

Patient were identified through the tracheostomy database which is recorded by the ICU Nurse Consultant. Patient medical notes were searched to ensure that they met the inclusion criteria. Basic demographic (age and sex) and clinical (location of stroke, length of stay in ICU and acute ward) data were extracted. Tracheostomy and patient outcome variables included timing of tracheostomy, reason for insertion, duration of tracheostomy, rationale for use of drying agents, specific agents used, length of hospital stay and discharge destination.

For analysis, data sets were dichotomized according to use or non–use of pharmacological drying agents. Variables were summarised descriptively and presented as mean, standard deviation for parametric and median interquartile range for non–parametric data. Non–numeric data was summarised narratively. Differences between subcategories for key variables were analysed via Mann–Whitney U tests for ordinal data and chi–square tests for nominal data using IBM SPSS 28.0¹⁸⁾.

A total of 116 patients were admitted to the neurosciences ward under a stroke, neurology, or neurosurgery bedcard with a tracheostomy between 2014 and 2019. The cohort were primarily admitted for stroke (48%) or TBI (36%) with a median duration of tracheostomy insertion of 20 IQR [12–32] days and a length of stay of 44 IQR [31–58] days in hospital. Thirty–five patients were unable to have their tracheostomy removed during their acute admission. Table 1 summarises the variables of interest in the whole cohort and both drying agent and non–drying agent sub–categorisations.

Ninety–six patients did not use any drying agents in their management plan whilst being tracheostomised; 20 patients required at least one drying agent during the course of their tracheostomy wean. The drying agent cohort had a significantly longer acute length of stay (median 58, IQR [40–110] days) compared to the non–drying agent cohort (median 43, IQR [28–57] days, Mann–Whitney U statistic 612.0, z= –2.544 (corrected for ties), p= 0.011, small effect size r=0.24). The results also indicated a significantly longer tracheostomy duration of 36 [21–71] days (Mann–Whitney U statistic 221.0, z=–2.90 (corrected for ties), p=0.004, medium effect size r=0.32) for the drying agents cohort compared to the non–drying agents cohort.

With respect to location of brain injury, 47% had confirmed supratentorial lesions, however, there was no significant difference between location of brain injury and the use of drying agents (χ² test=0.218, p=0.897). All three patients with lateral medullary infarction required the use of pharmacological drying agents in their tracheostomy wean. Three patients with varied neurological diagnosis (status epilepticus, transverse myelitis, and myasthenia gravis) required drying agents in their tracheostomy wean. Thirty–five patients were unable to be decannulated during their acute stay.

At this organisation the prescription of drying agents is based on clinical expertise rather than a structured protocol. Fig. 1 depicts the potential process of prescription and evaluation of drying agent use in a flow chart. Drying agents were initiated in this neurological population for a high secretion load (55%), impaired swallowing function (5%) or a both a high secretion load and impaired swallowing function (40%). Of the 20 patients who required drying agents in their weaning process, 95% utilised glycopyrrolate, 30% utilised atropine, 10% utilised hyoscine and in one instance an atrovent nebuliser was utilised.

Drying agents were used for a range of 1–86 days on the acute ward, with a median of 14 days (IQR 6–34), and in three patients there was concurrent use of multiple agents. Drying agents were initiated after their tracheostomy wean had commenced in 45% of cases, however, 30% of patients were utilising agents prior to their tracheostomy wean or at the same time as their tracheostomy wean commencing (25%). Glycopyrrolate was the initial agent used in 90% of patients, with atropine being used first in 10% of patients. Of the patients that had only one drying agent used during the entire wean, glycopyrrolate was utilised in 70%. One patient had botulinum toxin injected to their parotid salivary glands during their tracheostomy wean. Nine patients were discharged with ongoing drying agent prescription.

In total 1,114 doses of drying agent medication were given over the duration of data collection, with 643 administrations being glycopyrrolate. Five incidences of complications were associated with the use of drying agents; all described when using glycopyrrolate. These included two occurrences of tachycardia, abdominal pain, very thick secretions, and airway loss due to a sputum plug.

The variability in the dosage and frequency of drying agent use was significant. Glycopyrrolate was the initial drying agent used in 80% of patients. The amount given ranged from 50 micrograms to 300 micrograms, and the frequency of dose ranged from once daily to four times per day. 75% of patients had an initial dose of 200 micrograms and in general this dose was weaned as the patient progressed towards decannulation. For patients using glycopyrrolate who were decannulated in this acute setting, 54% had weaned all use of drying agents prior to decannulation, and 46% had continued weaning of glycopyrrolate after their tracheostomy wean. Atropine drops (1% concentration) were used by six patients, with three patients using atropine alone, and three patients combining it with other agents. The dose ranged from one drop to four drops, and the frequency of dose was either three or four times per day. Four patients using atropine had been decannulated in their acute stay, and all had weaned their use of atropine prior to decannulation. Hyoscine butylbromide was utilised by two patients, with dosages of 300 micrograms once daily, and 150 micrograms three times a day. Both these dosages were consistent until one of these patients had botulinum toxin injections into their parotid glands, and the other patient was decannulated, after which they were gradually ceased. One patient utilised a 500 microgram Atrovent nebuliser four times per day in their initial week of tracheostomy weaning, before changing to a combination of glycopyrrolate and atropine drops.

Overall, 65% of patients requiring prescription of a drying agent in their tracheostomy wean were able to be decannulated in this tertiary acute setting. The other 35% were either transferred to a local acute hospital or a rehabilitation unit with their tracheostomy in situ. The process of clinical decision making around tracheostomy weaning and the use of pharmacological drying agents was unclear from reviewing patient files, predicated on clinician opinion rather than objective measures.

Table 2 outlines specific attributes of the cohort that had drying agents in their tracheostomy weaning plan.

Most neurological patients requiring a tracheostomy will not need to use a pharmacological drying agent in their tracheostomy wean, however patients who present with an impaired swallow or a high secretion load may require one. This study indicated that the patient cohort requiring drying agents in their tracheostomy wean demonstrated a trend towards a longer duration of tracheostomy and an extended length of stay, but no associations were found between drying agent use and injury location, functional mobility or respiratory comorbidities. Therapists should complete relevant, patient–focused objective assessments, with particular focus on secretion load and swallowing function, during tracheostomy weaning to appropriately inform decision making regarding utility of drying agent use.

Tracheostomies are an increasingly common intervention in the neurological population to assist in weaning from mechanical ventilation, early mobilisation, and transition out of ICU^1,19). Neurological tracheostomy weaning is associated with more frequent and different complications than in the general tracheostomy weaning population, where an inability to manage oral secretions is common²⁰⁾. Current research continues to explore whether early tracheostomies result in better patient outcomes in the neurological populations. Early tracheostomies have been shown to reduce the risk of ventilator–acquired pneumonia, duration of mechanical ventilation, and both ICU and hospital length of stay for severe stroke and traumatic brain injury patients when compared with late tracheostomies²¹⁾. Clinicians working in the neurological population are increasingly likely to care for patients with tracheostomies, and utilising a weaning process that includes objective patient–centred assessments to support the use of drying agents will aid this process.

The process of tracheostomy insertion and weaning can impact negatively on the quality of life of a neurological patient. Tube care can be frightening and painful, and patients may experience frustration and powerlessness at being unable to speak²²⁾. Patients with tracheostomies can struggle to have input into their care, and alternative methods of communication can be exhausting²³⁾. Progressing towards the removal of tubes and attachments, including tracheostomies, can be a high priority for patients as it signifies recovery²⁴⁾. Because of the increasing use and potential negative impact of tracheostomies on patients, it is important for clinicians to be skilled in tracheostomy assessment and weaning, including an awareness of appropriate use of pharmacological drying agents.

The current study data set did not identify an association between age, location of injury, functional mobility, primary injury, respiratory comorbidities and the use of a pharmacological drying agent in tracheostomy weaning. However, patients with an infratentorial lesion, or both supra- and infratentorial lesions were slightly less likely to successfully wean from tracheostomy, with 62% of isolated infratentorial lesions, and 66% of combined supra-and infratentorial lesions being successfully decannulated. In comparison, 77% of patients with an isolated supratentorial lesion and 50% of patients with other neurological conditions such as Guillain–Barre syndrome were able to wean from tracheostomy. This aligns with Mitton et al.²⁵⁾ who reported patients with an infratentorial lesions to be more likely to experience unsuccessful tracheostomy weaning when compared to patients with supratentorial lesions.

All patients in this cohort with lateral medullary insult required drying agents in their tracheostomy wean, likely aligning with pathological insult to key nuclei involved in swallowing function, e.g. nucleus ambiguous and the nucleus tractus solitaries²⁶⁾. Clinicians should be mindful of thoroughly assessing swallowing ability and secretion management in patients with brainstem (particularly lateral medullary pathology) and consider initiating drying agents if required.

A limitation of the study is that the data collection period pre–dated electronic medical records and was likely influenced by variation in documentation requirements and hand–writing legibility. Clinical reasoning behind the use of drying agents is discerned from the documented notes, but further in–depth investigation could be completed with clinician focus groups or interviews. The sample size of patients using drying agents (n=20) was significantly smaller than the sample of patients not requiring drying agents, and from a single organisation only. The population in this study was also heterogenous, with multiple neurological populations, comorbidities and medical complications experienced during their admission. This study did not control for these significant variables and their potential introduction of bias into the results, and this must be considered when interpreting the results of this study.

Based on the results of this study, it is recommended that the clinical reasoning process around tracheostomy weaning and the use of drying agents in the neurological populations should be supported by objective assessments. These could be aligned with the six key assessment areas when assessing readiness for decannulation as discussed by Kutsukutsa et al.²⁷⁾. Clinical stability could be assessed using a patient’s vital signs and oxygen requirements. Evaluating airway patency could involve trialling a one–way speaking valve²⁸⁾. Level of consciousness in the neurological population can be classified by the Glasgow Coma Scale²⁹⁾. Swallowing ability can be assessed with bedside clinical measures, a fibreoptic endoscopic evaluation of swallowing, or a video fluoroscopy of swallowing³⁰⁾. Cough strength can be measured using cough peak flow, with values of greater than 160L/min suggesting successful decannulation³¹⁾. Clearance of secretions is more challenging to measure and may include the frequency of suctioning as well as the color and viscosity of secretions²⁷⁾.

The use of a formalised process for decannulation can result in a faster decannulation in the acquired brain injury population³²⁾, and this could be considered depending on organisational protocols. In this institution, a neurological tracheostomy working group has been established to develop a formal process for progression towards decannulation. It includes representatives from speech therapy, physiotherapy, medical, pharmacy and the tracheostomy review teams, and will review the tracheostomy weaning process, documentation, and the use of drying agents.

In this study both impaired swallowing function and a high secretion load were the clinical indicators for commencing a drying agent, and patients needing a drying agent in their tracheostomy demonstrated a trend towards a longer length of tracheostomy and length of stay. For these patients, an objective measure such as a fibreoptic endoscopic evaluation of swallowing could be completed to provide recorded visual information on secretion management, swallowing ability, and airway sensitivity¹²⁾. An outcome measure, such as the Secretion Severity Rating Scale³³⁾ or the Marianjoy Secretion Scale³⁴⁾, could be completed simultaneously, and aid clinical decision making with respect to the utility of the pharmacological drying agents.