Mechanical insufflation-exsufflation for the prevention and treatment of respiratory complications in acute cervical spinal cord injury: A retrospective service evaluation

Introduction

A spinal cord injury (SCI) is a life-changing event caused by irreversible damage to the spinal cord (World Health Organisation, 2013). Approximately 45% of injuries occur in the cervical region (Singh et al., 2014). In cervical spinal cord injuries (CSCI), respiratory function may be severely compromised due to denervation of the primary inspiratory muscle, the diaphragm. As a result, CSCI results in lower forced vital capacity (FVC) and peak cough flow (PCF) measurements, severely impairing secretion clearance (Berlowitz et al., 2016).

Respiratory complications (RC) affect up to 80% of individuals with CSCI and are the leading cause of morbidity and mortality in the acute phase (Berlowitz et al., 2016). Atelectasis, pneumonia, and respiratory failure (RF) are the most commonly reported RC (Jackson & Groomes, 1994). Due to the progressive nature of paralysis, secondary to spinal shock, acute management is targeted at preventing respiratory deterioration by promoting airway clearance and reducing work of breathing (Sheel et al., 2018).

Mechanical insufflation-exsufflation (MI-E) is a cough augmentation technique utilised for those with reduced PCF (Chatwin et al., 2018). MI-E augments tidal volume by delivering a positive inspiratory pressure (insufflation), which promotes mobilisation of peripheral secretions and may improve atelectasis (Chatwin et al., 2018). Expiratory flow is enhanced by rapid switch to negative pressure on expiration (exsufflation), thereby improving cough efficiency and secretion clearance (Chatwin et al., 2018). Literature supports the instigation of treatment when PCF <160L/min (Chatwin et al., 2018).

Research of MI-E in subjects with CSCI has focused on investigating its efficacy for improving lung function (Reid et al., 2010). Both Lee et al. (2012) and Pillastrini et al. (2006) found MI-E effective in improving FVC, forced expiratory volume in 1 second and PCF. However, the role and impact of MI-E as a prophylactic adjunct to prevent RC in this population remains unknown. For those with acute CSCI, MI-E is currently implemented in line with best practice recommendations and expert opinion, rather than high quality research (Bott et al., 2009). Therefore, uncertainty exists of how best to utilise MI-E in CSCI.

Objectives of this service evaluation were to report the following in subjects admitted with traumatic CSCI:

1. The incidence of RC within the 1st 3 weeks following injury for those who did/did not receive MI-E.

2. MI-E parameters used and differences in those parameters between prophylactic and therapeutic use.

Methods

Ethical approval was not required as this study was deemed a service evaluation. The project was registered and approved by The Patient Safety, Assurance and Audit Service at the study site (application number: CE22662).

Study design

A single-centre retrospective case note review was undertaken.

Setting

The service evaluation took place at a regional major trauma and neurosurgical centre in Southwest England, U.K. Patients with suspected traumatic SCI are admitted directly from the scene or transferred from other hospitals. Those with confirmed CSCI are typically transferred to the intensive care unit for cardiorespiratory monitoring and/or support. However, some with less severe impairments may be suitable for ward-based care with outreach support. During normal working hours (Monday–Friday, 8:30–16:30), new patients are screened and assessed by a physiotherapist within 4 hours of admission. MI-E (E70 Philips Respironics, Pennsylvania, U.S.A.) is standardly used for the 1st 7 days, as advised by a tertiary spinal treatment centre. Patients who have evidence of or are symptomatic of RC receive treatment considered ‘therapeutic’ rather than ‘prophylactic’. MI-E may be used in conjunction with other respiratory interventions, for example, manual assisted cough, manual techniques, and suction. FVC is assessed daily as part of nursing observations and MI-E is discontinued providing FVC is stable and >1 litre.

Sample

Potential subjects were identified through the Trauma Audit and Research Network database with the following criteria: ‘spinal cord injury’, admitted between 1st January 2017 and 30th September 2018. Subjects were screened for eligibility using inclusion and exclusion criteria outlined in Table 1.

Table 1: Subject inclusion and exclusion criteria.

Outcome variables

Diagnostic criteria for RC utilised followed guidance specific to the study setting and supported by relevant literature (Ray et al., 2013; Davidson et al., 2016; Kalil et al., 2016; O’Driscoll et al., 2017) (Table 2).

Table 2: Diagnostic criteria for respiratory complications.

Data collection

The following data was collected for all included subjects: age, gender, neurological level of injury, International Standards for Neurological Classification of Spinal Cord Injury (ISNCSCI) and medical management of SCI, in addition to respiratory status on admission. Furthermore, data was collected pertaining to MI-E set up. The incidence of RC were collected up to 14 days following MI-E discontinuation.

Data collection was completed between January 2019 and April 2019. Data was anonymised at point of collection by successive subject number allocation. The spreadsheet was piloted on five subjects prior to commencing data collection to assess the suitability of the RC diagnostic criteria. The pilot resulted in the inclusion of an additional criterion for pneumonia (‘the need for/increasing supplementary oxygen’).

Statistical analysis

Statistical analysis was carried out using IBM Statistical Package for the Social Science, version 26. Although normality of data was assessed, non-parametric tests were used for all variables for consistency. The Mann-Whitney U test was used to analyse differences in numerical variables. For categorical data, for samples >20 Chi-Square was used and <20, Fishers Exact was used. When comparing the proportions of 2 variables, odds ratios and their confidence intervals were calculated. Additionally, two-sided p values with a significance level <0.05 were used.

Results

A total of 33 subjects met the inclusion criteria (Figure 1). For the purpose of analysis, subjects were split into 2 cohorts; those who received MI-E prophylactically and/or therapeutically (MI-E cohort, n = 15) and those who did not (noMI-E cohort, n = 18). 1 subject who initially did not receive MI-E went on to develop pneumonia and received therapeutic MI-E. This subject was included in the noMI-E cohort.

See Figure 1: Inclusion flow chart.

Cohort characteristics

Characteristics for all cases and comparative characteristics between cohorts are summarised in Table 3. There was no statistical difference found in baseline characteristics between cohorts, except for gender. There were no statistically significant differences in baseline respiratory status between cohorts (Table 4).

Table 3: Subject characteristics for all cases and comparative characteristics within each cohort.

*Denotes statistically significant p <0.05 (2-sided test).

+Age is expressed as median and interquartile range. All categorical variables are expressed as proportion.

Where there is missing data, proportions are presented out of the number recorded.

Table 4: Baseline respiratory status on admission for all cases.

All categorical variables are expressed as proportions. Where there is missing data, proportions are presented out of the number recorded.

Incidence of respiratory complications

Overall, there were 7 episodes of RC reported in 5 subjects during the evaluation period (Table 5). This included 5 recorded 0–3 weeks post-injury and 2 recorded on admission. Subjects who developed more than 1 RC in the same episode (n = 2), were counted as 1 overall complication for that individual. The incidence of RC was higher in the noMI-E cohort, although this difference was not statistically significant. None of those who received MI-E prophylactically developed a RC during the evaluation period. The subject with pneumonia on admission received MI-E therapeutically, but still went on to develop an additional complication (RF).

Table 5: Overall incidence of respiratory complications in all cases.

+This includes recorded incidences on admission and during the evaluation period.

Clinical application and device set up

The MI-E cohort was further split into those who received MI-E prophylactically (MI-Ep) and those who received MI-E therapeutically (MI-Et). The majority received treatment prophylactically (n = 29). There was variation in the clinical application and set up of MI-E between prophylactic and therapeutic use (Table 6). Some data, specifically ‘pause time’, ‘cough trak’ and ‘oscillation’, were missing in all cases.

Table 6: Comparison of MI-E application (therapeutic MI-E (MI-Et) v. prophylactic MI-E (MI-Ep)).

All numerical variables are expressed as median and interquartile range.

+If used in manual mode.

*Denotes statistically significant p <0.05 (2-sided test).

Discussion

This service evaluation aimed to describe the incidence of RC in those with traumatic CSCI, as well as detail the use of MI-E. To the authors knowledge, this is the first published service evaluation to do so in this specific population. Results suggest there is no difference in RC incidence for patients who do/do not receive MI-E. Additionally, variation in the clinical application of MI-E was observed.

The overall incidence of RC observed in this service evaluation is low compared to existing data. One study (Lemons & Wagner, 1994) found 62% (n = 39) of subjects with acute CSCI experienced RC, which is consistent with earlier studies by Fishburn et al. (1990) and Reines and Harris (1987) who reported incidences of 57% (n = 17) and 77% (n = 95) respectively. A possible explanation for differences observed might relate to the inclusion of a greater number of complete injuries in those studies. This is noteworthy since complete injuries often result in more defined impairments and little improvement potential (Harvey, 2016). The significance of this is reflected in the higher incidence of RC reported in complete injuries alone (69%) (Jackson & Groomes, 1994). Indeed, medical care improvements have seen a decline in the incidence of complete injuries (World Health Organisation, 2013), therefore the RC reported within this evaluation may reflect more accurately on the wider population.

Incidence of respiratory complications in noMI-E and MI-E cohorts

Since prophylactic MI-E is considered standard practice at the study location, it is surprising that over half of the cohort did not receive this. Despite its increasing popularity, a recent survey of MI-E use in an adult intensive care unit highlighted MI-E use in clinical practice as sporadic (Swingwood et al., 2020). Whilst there is inevitable variation in opinion and practice among clinicians, this survey highlighted lack of experience and training as barriers to MI-E’s practical use (Swingwood et al., 2020). Although clinician demographics were not recorded in this study, level of experience and confidence in device use may have influenced whether MI-E was instigated prophylactically.

Though there was no statistically significant difference in neurological level of injury and ISNCSCI classification between cohorts, data for the latter was missing for 39% (n = 7) of subjects in the noMI-E cohort. It is possible that this cohort were less neurologically impaired and consequently may have had comparatively better respiratory function. Therefore, it is plausible these subjects were deemed at a lower risk of developing RC by the treating clinician.

While the results reveal the incidence of RC were higher in the noMI-E cohort, they did not reach statistical significance. Therefore, no firm conclusions can be drawn as to whether instigating prophylactic MI-E provides additional benefit in preventing RC. Whilst none of those receiving treatment prophylactically went on to develop a RC during the evaluation period, the small sample and confounding factors (for example, pre-existing respiratory co-morbidities and associated chest wall injuries) may limit extrapolation of results.

Clinical application of MI-E

There is a paucity of research reporting commonly used pressures for prophylactic MI-E for those with SCI, therefore direct comparisons cannot be made. However, artificial lung studies suggest to achieve inspiratory volumes of normal capacity (>3.6 litres), minimum insufflation pressures of +40cmH2O are required (Volpe et al., 2018; Chatwin & Simonds, 2020). The median insufflation pressure used in this study was +30cmH2O, which may be inadequate to achieve the desired physiological effect. Therefore, the finding of no significant difference between cohorts may be unsurprising.

The variation in clinical application between MI-Ep and MI-Et was anticipated. Clinicians used greater exhalation pressures in the presence of a complication, which corresponds with existing research favouring asymmetrical pressure settings to achieve adequate expiratory flow bias for sputum clearance (Chatwin & Simonds, 2020). Given that CSCI affects inspiratory and expiratory capacity, it is surprising that the same insufflation pressures were used for MI-Ep and MI-Et. In the presence of an acute RC, lung function measurements such as FVC have shown to decline by up to 70% (Nasher et al., 2014). Therefore, higher pressures may be required to achieve adequate lung volume increases and airway clearance. Notwithstanding this, whilst MI-E pressures should be individualised (Chatwin et al., 2018), ultimately tolerability will guide MI-E prescription (Spinou, 2020).

Device settings, ‘pause time’, ‘cough trak’ and ‘oscillation’, were not recorded consistently by the treating clinician. The underutilisation of such settings may suggest sub-optimal MI-E implementation, thus limiting clinical effectiveness. However, it is possible these settings were not used as opposed to not recorded; particularly for those receiving prophylactic treatment.

Limitations

The absence of a comprehensive evaluation of potential confounders should be considered when interpreting results. The sample included only adults with a CSCI admitted to a single centre, which may affect the study’s external validity. Whilst RC are most prevalent in CSCI, those with thoracic lesions are also at risk, therefore their inclusion may have been warranted. The overall sample was small, increasing the probability of a type II error when comparing cohorts (Banerjee et al., 2009). Additionally, 7 medical records were inaccessible, therefore potentially eligible subjects may have been missed.

Larger prospective studies are required to further establish the incidence of RC in patients with CSCI who receive MI-E post-injury. Multi-site data collection may allow for a larger sample and for prescription variability.

Conclusion

Whilst the findings of this service evaluation demonstrate a low incidence of RC in those receiving MI-E following CSCI, limitations exist that may restrict the application and confound the results. Current guidelines recommend MI-E is considered for those with impaired respiratory function following SCI (Bott et al., 2009). Therefore, greater efforts are needed to ensure those at high risk (for example, complete injuries, those with pre-existing respiratory co-morbidities or additional respiratory complications) receive prophylactic treatment. There is a need to establish the incidence of RC in those with thoracic and CSCI who receive MI-E prophylactically. Emphasis should be placed on obtaining a sufficiently sized sample to produce clinically meaningful results.

Funding

This service evaluation was conducted as part of a master’s research project and received financial assistance from the Private Physiotherapy Educational Foundation to cover tuition fees.

Banerjee, A., Chitnis, U. B., Jadhav, S. L., Bhawalkar, J. S., & Chaudhury, S. (2009). Hypothesis testing, type I and type II errors. Industrial Psychiatry Journal, 18(2), 127–131. https://doi.org/10.4103/0972-6748.62274.

Berlowitz, D. J., Wadsworth, B., & Ross, J. (2016). Respiratory problems and management in people with spinal cord injury. Breathe (Sheffield, England), 12(4), 328–340. https://doi.org/10.1183/20734735.012616.

Bott, J., Blumenthal, S., Buxton, M., Ellum, S., Falconer, C., Garrod, R., Harvey, A., Hughes, T., Lincoln, M., Mikelsons, C., Potter, C., Pryor, J., Rimington, L., Sinfield, F., Thompson, C., Vaughn, P., White, J., & British Thoracic Society Physiotherapy Guideline Development Group (2009). Guidelines for the physiotherapy management of the adult, medical, spontaneously breathing patient. Thorax, 64 Suppl 1, i1–i51. https://doi.org/10.1136/thx.2008.110726.

Chatwin, M., & Simonds, A. K. (2020). Long-term mechanical insufflation-exsufflation cough assistance in neuromuscular disease: Patterns of use and lessons for application. Respiratory Care, 65(2), 135–143. https://doi.org/10.4187/respcare.06882.

Chatwin, M., Toussaint, M., Gonçalves, M. R., Sheers, N., Mellies, U., Gonzales-Bermejo, J., Sancho, J., Fauroux, B., Andersen, T., Hov, B., Nygren-Bonnier, M., Lacombe, M., Pernet, K., Kampelmacher, M., Devaux, C., Kinnett, K., Sheehan, D., Rao, F., Villanova, M., Berlowitz, D., Morrow, B. M. (2018). Airway clearance techniques in neuromuscular disorders: A state of the art review. Respiratory Medicine, 136, 98–110. https://doi.org/10.1016/j.rmed.2018.01.012.

Davidson, A. C., Banham, S., Elliott, M., Kennedy, D., Gelder, C., Glossop, A., Church, A. C., Creagh-Brown, B., Dodd, J. W., Felton, T., Foëx, B., Mansfield, L., McDonnell, L., Parker, R., Patterson, C. M., Sovani, M., Thomas, L., & BTS Standards of Care Committee Member, British Thoracic Society/Intensive Care Society Acute Hypercapnic Respiratory Failure Guideline Development Group, on behalf of the British Thoracic Society Standards of Care Committee (2016). BTS/ICS guideline for the ventilatory management of acute hypercapnic respiratory failure in adults. Thorax, 71 Suppl 2, ii1–ii35. https://doi.org/10.1136/thoraxjnl-2015-208209.

Fishburn, M. J., Marino, R. J., & Ditunno, J. F., Jr. (1990). Atelectasis and pneumonia in acute spinal cord injury. Archives of Physical Medicine and Rehabilitation, 71(3), 197–200.

Harvey L. A. (2016). Physiotherapy rehabilitation for people with spinal cord injuries. Journal of Physiotherapy, 62(1), 4–11. https://doi.org/10.1016/j.jphys.2015.11.004.

Jackson, A. B., & Groomes, T. E. (1994). Incidence of respiratory complications following spinal cord injury. Archives of Physical Medicine and Rehabilitation, 75(3), 270–275. https://doi.org/10.1016/0003-9993(94)90027-2.

Kalil, A. C., Metersky, M. L., Klompas, M., Muscedere, J., Sweeney, D. A., Palmer, L. B., Napolitano, L. M., O’Grady, N. P., Bartlett, J. G., Carratalà, J., El Solh, A. A., Ewig, S., Fey, P. D., File, T. M., Jr, Restrepo, M. I., Roberts, J. A., Waterer, G. W., Cruse, P., Knight, S. L., & Brozek, J. L. (2016). Management of adults with hospital-acquired and ventilator-associated pneumonia: 2016 clinical practice guidelines by the infectious diseases society of America and the American Thoracic Society. Clinical Infectious Diseases, 63(5), e61–e111. https://doi.org/10.1093/cid/ciw353.

Lee, Y., Kim, J., & Jin, Y. (2012). The efficacy of pulmonary rehabilitation using a mechanical in-exsufflator and feedback respiratory training for cervical cord injury patients. The Journal of Physical Therapy Science, 24(1), 88–92.

Lemons, V., & Wagner, F. (1994). Respiratory complications after cervical spinal cord injury. Spine, 19(20), 2315–2320.

Nasher, M., Pocock, A., & Ward, T. (2014). P158 evaluation of vital capacity changes in spinal injuries during episodes of sepsis. Thorax, 69, A143.

O’Driscoll, B., Howard, L., Earis, J., & Mark, V. (2017). British Thoracic Society guidelines for oxygen use in adults in healthcare and emergency settings. British Medical Journal Open Respiratory Research, 4(1), 1–20.

Pillastrini, P., Bordini, S., Bazzocchi, G., Belloni, G., & Menarini, M. (2006). Study of the effectiveness of bronchial clearance in subjects with upper spinal cord injuries: Examination of a rehabilitation programme involving mechanical insufflation and exsufflation. Spinal Cord, 44(10), 614–616. https://doi.org/10.1038/sj.sc.3101870.

Ray, K., Bodenham, A., & Paramasivam, E. (2013). Pulmonary atelectasis in anaesthesia and critical care. British Journal of Anaesthesia, 14(5), 236–245.

Reid, W. D., Brown, J. A., Konnyu, K. J., Rurak, J. M., & Sakakibara, B. M. (2010). Physiotherapy secretion removal techniques in people with spinal cord injury: A systematic review. The Journal of Spinal Cord Medicine, 33(4), 353–370. https://doi.org/10.1080/10790268.2010.11689714.

Reines, H. D., & Harris, R. C. (1987). Pulmonary complications of acute spinal cord injuries. Neurosurgery, 21(2), 193–196. https://doi.org/10.1227/00006123-198708000-00010.

Sheel, A., Welch, J., & Townson, A. (2018). Respiratory management following spinal cord injury. In J.J. Eng, R. Teasell, W. Miller, D. Wolfe, A. Townson, A. Hisieh, & S. Conncolly. Spinal Cord Injury Rehabilitation Evidence, 6, 1–72.

Singh, A., Tetreault, L., Kalsi-Ryan, S., Nouri, A., & Fehlings, M. G. (2014). Global prevalence and incidence of traumatic spinal cord injury. Clinical Epidemiology, 6, 309–331. https://doi.org/10.2147/CLEP.S68889.

Spinou, A. (2020). A review on cough augmentation techniques: Assisted inspiration, assisted expiration and their combination. Physiological Research, 69(Suppl 1), S93–S103. https://doi.org/10.33549/physiolres.934407.

Swingwood, E., Tume, L., & Cramp, F. (2020). A survey examining the use of mechanical insufflation-exsufflation on adult intensive care units across the U.K. Journal of the Intensive Care Society, 21(4), 283–289. https://doi.org/10.1177/1751143719870121.

Volpe, M. S., Naves, J. M., Ribeiro, G. G., Ruas, G., & Amato, M. (2018). Airway clearance with an optimized mechanical insufflation-exsufflation maneuver. Respiratory Care, 63(10), 1214–1222. https://doi.org/10.4187/respcare.05965.

World Health Organisation. (2013). Spinal cord injury fact sheet, number 384. http://www.who.int/mediacentre/factsheets/fs384/en/.