A physiotherapy-led early mobilisation protocol for neurosurgical patients with external ventricular drains in intensive care: A service evaluation

Introduction

An external ventricular drain (EVD) is surgically inserted into the brain ventricles to drain cerebrospinal fluid (CSF) and is commonly seen in the neurological intensive care unit (ICU) (Muralidharan, 2015). It is used to relieve raised intra-cranial pressure (ICP) and treat acute hydrocephalus in patients with acquired brain injuries such as traumatic brain injury, intra-cerebral haemorrhage, subarachnoid haemorrhage and meningitis (Muralidharan, 2015), which would otherwise result in reduced cerebral perfusion and increased cerebral oedema (Hinson et al., 2010). This drain also enables clearing of intra-ventricular blood clots, monitoring of ICP and administering of medications.

Historically, patients with EVDs remain on bed rest for the duration of EVD placement (Moyer et al., 2017; Gaspari et al., 2018; Young et al., 2019) which averages 8 days (Albano et al., 2018). The presence of an EVD posed a major barrier to early mobilisation as there were safety concerns pertaining to bleeding and dislodgement during the process (Hale et al., 2013). In the event of EVD dislodgement, an emergency surgical procedure would be necessary for re-insertion (Gaspari et al., 2018). As the EVD is clamped to prevent inappropriate drainage of CSF during position changes, there is a potential risk of elevated ICP (Muralidharan, 2015). These factors could deter physiotherapists from encouraging mobilisation. Conversely, 8 days on bed rest could have significant negative consequences on the patient’s musculoskeletal, cardiovascular and respiratory systems, further prolonging intensive care stay.

Several studies have proposed the use of a mobilisation protocol in the ICU, with eligibility criteria that provide specific safety recommendations to identify suitable patients (Bailey et al., 2007; Morris et al., 2008; Miranda Rocha et al., 2017). However, most of these were developed and introduced in the medical ICU (Conceição et al., 2017) with only 2 mobilisation protocols devised particularly for patients with EVDs in the neurological ICU (Moyer et al., 2017; Young et al., 2019). These EVD mobilisation protocols only consisted of exclusion criteria and parameters related to the neurological system. Parameters related to cardiovascular, respiratory and musculoskeletal systems were not clearly stated. The lack of clarity on eligibility criteria could compromise safety and delay mobilisation.

In Tan Tock Seng Hospital (TTSH), Singapore, a physiotherapy-led early mobilisation protocol for neurological patients with EVDs was introduced in April 2018. The protocol contained specific eligibility criteria indicating the parameters for each aforementioned physiological system. The aim of this service evaluation was to determine whether the early mobilisation protocol, which included physiological parameters, was safe, feasible and effective in patients with EVDs.

Methods

Study design and ethical considerations

This was a retrospective service evaluation of an early mobilisation protocol for patients with EVDs in the neurological ICU. Medical records were reviewed for 2 comparable 6-month periods from October 2017 to March 2018 (pre-protocol period) and October 2019 to March 2020 (protocol period). The UCL Research Ethics Committee and Singapore’s Domain Specific Review Board granted ethics exemption and waiver of informed consent. Service evaluation registration procedures within the hospital were followed.

ICU setting

Tan Tock Seng Hospital is the 2nd largest public tertiary hospital in Singapore. In the 18-bedded neurological ICU, a typical nurse-to-patient ratio is 1:2. A blanket referral system for physiotherapy is in place, and physiotherapists are allowed to screen and review patients. On weekdays, the workload is shared among 2 permanent senior physiotherapists and 2 rotating junior physiotherapists. They provide respiratory and rehabilitation therapy for the patients. On weekends, patients are reviewed by other physiotherapists who are rostered to work for the day, mainly for respiratory interventions.

Protocol

The EVD early mobilisation protocol was devised through discussion with physiotherapists, neurosurgeons and ICU consultants. The term ‘early’ mobilisation was determined from the time at which the patients were deemed physiologically stable following a set of inclusion and exclusion criteria, as shown in Tables 1 and 2 respectively. Patients with EVDs were evaluated by physiotherapists on a daily basis to determine their eligibility for mobilisation. As for the type of mobilisation activity performed and the level of assistance given, it was dependent on physiotherapy assessment on the day of therapy.

Table 1: Inclusion criteria.

bpm = beats per minute; EVD = external ventricular drain; FiO2 = fraction of inspired oxygen; GCS = Glasgow Coma Scale; HR = heart rate; ICA = internal carotid artery; ICP = intracranial pressure; LL = lower limb; MAP = mean arterial pressure; MCA = middle cerebral artery; PEEP = positive end-expiratory pressure; RASS = Richmond Agitation Sedation Scale; RR = respiratory rate; SpO2 = peripheral capillary oxygen saturation, UL = upper limb.

Table 2: Exclusion criteria.

CSF = cerebrospinal fluid; EVD = external ventricular drain.

Mobilisation was defined as any active out-of-bed activity that consisted of sitting-on-edge-of-bed, sit-to-stand, sitting-out-of-bed, marching and ambulation. The flow chart used to guide mobilisation was modified from Moyer et al. (2017) and is illustrated in Figure 1. The physiotherapist assisted the patient in mobilisation while a nurse managed the EVD. The EVD was clamped during mobilisation and levelled to the tragus of the ear to measure ICP before session, sitting-on-edge-of-bed and after session. For higher levels of mobilisation activity, ICP was not measured but vital signs and symptoms were monitored throughout by both the nurse and physiotherapist.

See Figure 1: EVD mobilisation flow chart.

Service evaluation process

Relevant data for the studied time periods were extracted from the hospital’s electronic system by the principal investigator and were de-identified. Data were stored in an encrypted Microsoft Excel file. For patients from the pre-protocol period, eligibility for mobilisation was determined by retrospectively reviewing their hourly physiological parameters in the ICU and comparing to the protocol’s eligibility criteria. For patients deemed eligible for mobilisation, physiotherapy documentation was further scrutinised to find out whether they were, in fact, mobilised. Data entry was verified by a random check of 10% of patients with an independent physiotherapist who was not part of the project team.

Outcome measures

Outcome measures for safety, feasibility and effectiveness were collected. Safety was determined by the number and type of adverse events occurring during mobilisation. Adverse events were defined as events exceeding the established safety limits. Types of adverse events included elevated ICP (>20mmHg), EVD dislodgement or malfunction, haemodynamic events (orthostatic systolic blood pressure (BP) drop by ≥20mmHg; hypotension or hypertension depending on patient’s baseline BP range and the targeted BP range set by neurosurgeons) and adverse symptoms (such as dizziness, emesis and headache). Adverse events that led to clinical deterioration and the need for medical intervention were termed as ‘adverse events with consequences’. Feasibility was determined by the proportion of patients who were successfully mobilised in the neurological ICU. Effectiveness of the protocol was assessed using the following outcomes:

1. Time from EVD placement to first mobilisation in the hospital.

2. ICU length of stay (LOS).

3. Hospital LOS, in eligible patients.

Statistical analysis

Data were analysed using the SPSS (2013) Windows, version 25.0. Continuous variables were analysed using a 2-sample t-test for parametric data or Mann-Whitney U test for non-parametric data. Chi-Square test (n >20) or Fisher’s exact test (n <20) was used for categorical variables. Level of significance was set at 0.05 for all tests. As significant differences in patient baseline characteristics may influence the outcomes of the protocol, a correlation test was conducted. A Pearson’s correlation (r) for parametric data or Spearman’s correlation (rs) for non-parametric data was used for the correlation test.

Results

A total of 75 patients received EVDs during the pre-protocol period (Figure 2) and 42 patients underwent EVD placement in the protocol period (Figure 3). Their eligibility for mobilisation are summarised in Figures 2 and 3. The baseline characteristics of eligible patients for the 2 periods are shown in Table 3. There were similar baseline characteristics except for the Glasgow Coma Scale (GCS) on emergency department (ED) admission, which was significantly higher in the protocol group (p = 0.023).

Safety, feasibility and effectiveness of the early mobilisation protocol are shown in Table 4. After protocol implementation, the median time from EVD placement to 1st mobilisation decreased from 14 to 3.5 days (p <0.0001) in eligible patients. The median ICU and hospital length of stays were also significantly reduced from 8 to 3.5 days (p = 0.037) and 38 to 22.5 days (p = 0.030) respectively. Correlation tests between the GCS on ED admission and LOS in both ICU and hospital showed either weak correlation (rs = 0.038 to 0.070, p >0.05) or a clinically unimportant relationship (multiple outliers). Hence, the GCS on ED admission may not influence LOS.

No adverse events were recorded for the 21 mobilisation sessions in the protocol period. The highest level of mobility achieved during each session was sitting-on-edge-of-bed (47.6%), followed by marching (23.8%), sit-to-stand (19.1%) and ambulation (9.5%). Only 1 out of 25 eligible patients was mobilised in the pre-protocol period, compared to 11 out of 16 patients in the protocol period. Reasons for the remaining 5 patients not mobilised are seen in Figure 3. This contributes to a 64.8% (95% CI, 35.9–82.1%, p <0.0001) increase in the proportion of patients mobilised.

See Figure 2: Flow chart for patients’ eligibility for mobilisation in the pre-protocol period.

See Figure 3: Flow chart for patients’ eligibility for mobilisation in the protocol period.

Table 3: Eligible patient baseline characteristics.

CI = confidence interval; ED = emergency department; EVD = external ventricular drain; GCS = Glasgow Coma Scale; ICH = intra-cerebral haemorrhage; IVH = intra-ventricular haemorrhage; SAH = subarachnoid haemorrhage; TBI = traumatic brain injury.

***Level of significance p <0.05.

1Mann-Whitney U test with median (25th–75th percentile).

2Fisher’s exact test.

Table 4: Safety, feasibility and effectiveness of the early mobilisation protocol.

CI = confidence interval; EVD = external ventricular drain; ICU = intensive care unit; LOS = length of stay; SOEOB = sitting-on-edge-of-bed; SOOB = sitting-out-of-bed; STS = sit-to-stand.

***Level of significance p <0.05.

1Fisher’s exact test.

2Mann-Whitney U test with median (25th–75th percentile).

Discussion

This service evaluation demonstrated that the physiotherapy-led early mobilisation protocol in patients with EVDs was safe, feasible and effective within TTSH. The time from EVD placement to 1st mobilisation, ICU LOS and hospital LOS were significantly reduced in eligible patients. A significantly greater proportion of patients with EVDs were also mobilised in the neurological ICU after protocol implementation, with no recorded adverse events.

In line with Moyer et al. (2017) and Young et al. (2019), the time from EVD placement to 1st mobilisation, as well as ICU and hospital lengths of stay were significantly reduced. While LOS is an important outcome, it is difficult to evaluate owing to many contributing factors and is likely not dependent on physiotherapy treatments alone (Gruenberg et al., 2006). Assessment of physical function in patients with EVDs could have provided data more directly associated with the intervention and could be a better outcome measure to examine the effectiveness of early mobilisation in the ICU.

Early mobilisation in patients with EVDs was found to be safe given that no adverse events were recorded. This was in contrast to studies with reported EVD mobilisation protocols in which adverse events, ranging between 0.8% and 5.9% occurred (Moyer et al., 2017; Young et al., 2019). The robust safety profile for mobilisation of patients with EVDs in this study may be due to more cautious eligibility criteria, or differing definitions of adverse events.

A systematic review by Conceição et al. (2017), suggested that early mobilisation of critically ill patients could be carried out with reference to specific safety criteria which can be broadly categorised into cardiovascular, respiratory, neurological and orthopaedic-related. In this study, the protocol clearly stated eligibility criteria pertaining to each of the above mentioned physiological systems for initiation of early mobilisation in patients with EVDs. Having such definitive criteria eliminated any ambiguity for determining whether a patient was eligible for mobilisation. Coupled with the mobilisation flow chart that illustrated a step-by-step safety checklist, as well as the collaborative efforts between the nurse and physiotherapist in monitoring the patient’s vital signs and symptoms throughout mobilisation, adverse events were avoided.

This study reported an increased in mobilisation rate during the protocol period. Titsworth et al. (2012) reported similar finding in the population admitted into the neurological ICU using the Progressive Upright Mobility Protocol Plus. Despite differences in the methods used in quantifying mobilisation rate in both studies, they demonstrated feasibility of introducing a mobilisation protocol within critical care.

This study was not designed as research, but to evaluate a specific service. Patient numbers were small, and data were analysed retrospectively from a single centre with no concurrent control group or randomisation process. Results would not be generalisable to the wider population, although this work provides some practice sharing and learning points. The accuracy of the reported adverse events was dependent on the quality of note-taking at the time of the incident. Eligibility criteria applied retrospectively to the pre-protocol group may have flaws as the medical records may not have presented a full picture as to whether patients were appropriate for mobilisation.

The clinical implications of having a mobilisation protocol included developing greater confidence amongst physiotherapists and other healthcare professionals in mobilising patients with EVDs. Furthermore, it helped to promote a mobilisation culture within the ICU in this patient population, who traditionally would have been placed on bed rest.

Conclusion

A physiotherapy-led early mobilisation protocol for patients with EVDs enabled safe, feasible and effective mobilisation in 1 neurological ICU. Although these findings add to the current evidence supporting mobilisation among this patient population, a service evaluation cannot be generalised more widely. Future prospective, controlled research studies are warranted.

Acknowledgements

The author would like to thank the physiotherapy department and the neurocritical care team at TTSH for the support. Acknowledgements are also extended to the physiotherapists and nursing staff involved in mobilisation, as well as Dr. Vincent Ng Yew Poh and Dr. Wong Yu Lin for the collaboration in developing the protocol’s eligibility criteria.

Funding

This service evaluation was undertaken as a fulfilment of the author’s MSc research thesis. There was no specific grant from any funding agency in the public, commercial or not-for-profit sectors.

Albano, S., Berman, B., Fischberg, G., Siddiqi, J., Liu, B., Khan, Y., Zafar, A., Quadri, S. A., & Farooqui, M. (2018). Retrospective analysis of ventriculitis in external ventricular drains. Neurology Research International, 2018, 5179356. https://doi.org/10.1155/2018/5179356.

Bailey, P., Thomsen, G. E., Spuhler, V. J., Blair, R., Jewkes, J., Bezdjian, L., Veale, K., Rodriquez, L., & Hopkins, R. O. (2007). Early activity is feasible and safe in respiratory failure patients. Critical Care Medicine, 35(1), 139–145. https://doi.org/10.1097/01.CCM.0000251130.69568.87.

Conceição, T., Gonzáles, A. I., Figueiredo, F., Vieira, D., & Bündchen, D. C. (2017). Safety criteria to start early mobilization in intensive care units. Revista Brasileira de Terapia Intensiva, 29(4), 509–519. https://doi.org/10.5935/0103-507X.20170076.

Gaspari, C., Lafayette, S., Jaccoud, A., Kurtz, P., Lavradas, L. & Cavalcanti, D. (2018). Safety and feasibility of out-of-bed mobilization for patients with external ventricular drains in a neurosurgical intensive care unit. Journal of Acute Care Physical Therapy, 1.

Gruenberg, D. A., Shelton, W., Rose, S. L., Rutter, A. E., Socaris, S., & McGee, G. (2006). Factors influencing length of stay in the intensive care unit. American Association of Critical-Care Nurses, 15(5), 502–509.

Hale, C., Wong, K., Pennings, A., Rnic, A., Tobali, B., Hawke, C., Brown, J., Cott, C., Heck, C., & Duncan, C. (2013). Practice patterns of canadian physiotherapists mobilizing patients with external ventricular drains. Physiotherapy Canada, 65(4), 365–373. https://doi.org/10.3138/ptc.2012-32.

Hinson, H. E., Hanley, D. F., & Ziai, W. C. (2010). Management of intraventricular hemorrhage. Current Neurology and Neuroscience Reports, 10(2), 73–82. https://doi.org/10.1007/s11910-010-0086-6.

Miranda Rocha, A. R., Martinez, B. P., Maldaner da Silva, V. Z., & Forgiarini Junior, L. A. (2017). Early mobilization: Why, what for and how? Medicina Intensiva, 41(7), 429–436. https://doi.org/10.1016/j.medin.2016.10.003.

Morris, P. E., Goad, A., Thompson, C., Taylor, K., Harry, B., Passmore, L., Ross, A., Anderson, L., Baker, S., Sanchez, M., Penley, L., Howard, A., Dixon, L., Leach, S., Small, R., Hite, R. D., & Haponik, E. (2008). Early intensive care unit mobility therapy in the treatment of acute respiratory failure. Critical Care Medicine, 36(8), 2238–2243. https://doi.org/10.1097/CCM.0b013e318180b90e.

Moyer, M., Young, B., Wilensky, E. M., Borst, J., Pino, W., Hart, M., LoBreglio, J., Zaleski, D., Leonor, I., Kung, D., Smith, M., Zager, E., Grady, M. S., & Kumar, M. (2017). Implementation of an early mobility pathway in neurointensive care unit patients with external ventricular devices. The Journal of Neuroscience Nursing, 49(2), 102–107. https://doi.org/10.1097/JNN.0000000000000258.

Muralidharan, R. (2015). External ventricular drains: Management and complications. Surgical Neurology International, 6(Suppl 6), S271–S274. https://doi.org/10.4103/2152-7806.157620.

SPSS IBM Corp. (2013). IBM SPSS statistics for Windows, version 22.0, released 2013. https://www.ibm.com/uk-en/analytics/spss-statistics-software.

Titsworth, W. L., Hester, J., Correia, T., Reed, R., Guin, P., Archibald, L., Layon, A. J., & Mocco, J. (2012). The effect of increased mobility on morbidity in the neurointensive care unit. Journal of Neurosurgery, 116(6), 1379–1388. https://doi.org/10.3171/2012.2.JNS111881.

Young, B., Moyer, M., Pino, W., Kung, D., Zager, E., & Kumar, M. A. (2019). Safety and feasibility of early mobilization in patients with subarachnoid hemorrhage and external ventricular drain. Neurocritical Care, 31(1), 88–96. https://doi.org/10.1007/s12028-019-00670-2.