The safety of mobilisation and its effect on

haemodynamic and respiratory status of intensive

care patients

Kathy Stiller, Anna C. Phillips, and Paul Lambert

This study investigated the safety of mobilising acutely ill in-patients, in particularthe effect of mobilisation on their haemodynamic and respiratory parameters.Thirty one patients in an intensive care unit (ICU) deemed suitable formobilisation, based on a comprehensive screening process, received 69 mobilisationtreatments in total. These treatments most often included sitting on the edge of thebed and standing. Outcome measures including heart rate, systolic and diastolicblood pressure, and percutaneous saturation of oxygen, were measured prior to,during and after mobilisation. Additionally, any deterioration in clinical status, andintervention required for it, was noted. On the majority of occasions (91.3%), pre-treatment data from patients indicated marginal cardiac and=or respiratory reserve.During mobilisation, significant increases were seen in heart rate and bloodpressure, while percutaneous oxygen saturation decreased (not significantly). Thesechanges were generally of small magnitude and did not require any specificintervention. On three of the 69 occasions of mobilisation (4.3%), clinical statusdeteriorated, requiring intervention. For all three patients involved, this was a fall inoxygen saturation, requiring a temporary increase in the inspired fraction ofoxygen to stabilise respiratory status. Although mobilisation resulted in significantincreases in heart rate and blood pressure and a non-significant fall in percutaneousoxygen saturation, the ICU patients in this study deemed suitable for mobilisationwere able to be safely mobilised.

INTRODUCTION

The physiotherapy management of acutely illpatients who are in an intensive care unit (ICU)often incorporates some form of mobilisation.The aims of mobilisation for these patientsinclude increasing lung volumes, improving

ventilation=perfusion matching, providing agravitational stimulus to restore normal fluiddistribution in the body, reducing the effects ofimmobility, and maintaining or improvingfunction and fitness (Bishop, 1996; Dean, 1994;Dean and Ross, 1992a, 1992b; Stiller and Phillips,2003). Although mobilisation is deemed an

Kathy Stiller, B App Sc, PhD, Senior Physiotherapist, Physiotherapy Department, Royal AdelaideHospital, North Terrace, Adelaide, SA 5000, Australia. E-mail: Kstiller@mail.rah.sa.gov.auAnna C. Phillips, B App Sc, Senior Physiotherapist, Physiotherapy Department, Royal AdelaideHospital, North Terrace, Adelaide, SA 5000, Australia.Paul Lambert, B App Sc, Senior Physiotherapist, Physiotherapy Department, Royal AdelaideHospital, North Terrace, Adelaide, SA 5000, Australia.

Accepted for publication 22 March 2004.

Physiotherapy Theory and Practice, 20: 175�185, 2004Copyright # Taylor & Francis Inc.ISSN: 0959-3985 print=1521-0510 onlineDOI: 10.1080/09593980490487474

essential part of the physiotherapy manage-ment of acutely ill in-patients, a literaturesearch (of Medline and CINAHL databases)revealed that there is no published clinicalresearch assessing the overall safety of mobilis-ing these patients, nor the effect of mobilisationon haemodynamic and respiratory status. Thisis an important omission as there is the poten-tial for adverse side effects, especially given theborderline cardiorespiratory function of acutelyill patients. Thus, the aim of this pilot study wasto document the safety of mobilisation foracutely ill patients in ICU, in particular thehaemodynamic and respiratory responses andthe occurrence of any adverse side effectsduring mobilisation.

METHOD

A prospective study was done over threeseparate two week periods at the RoyalAdelaide Hospital (RAH) ICU and includedall patients where mobilisation formed part ofthe patient’s physiotherapy management.These distinct time periods were selected toensure that a wider selection of patients wasincluded in the study sample than would havebeen the case if one continuous time periodwas used. For the purposes of this study,mobilisation was defined as moving from lyingto sitting on the edge of the bed, sitting tostanding, a standing transfer from the edge ofthe bed to a chair, or walking. Patients under-going passive forms of mobilisation, such aspositioning upright in bed and mechanicaltransfers from bed to chair were not studied.The mobilisation task selected was based on thepatient’s general clinical status and ability.Prior to mobilising any patient, a comprehen-sive range of factors including medical back-ground, cardiovascular reserve, respiratoryreserve and other relevant factors, were takeninto consideration to assess whether mobilisa-tion was safe to proceed. This screening pro-cess is based on that described by Stiller andPhillips (2003) and is summarised in flow chartformat in Figure 1.

Background pre-treatment data wererecorded, including descriptive information(e.g., primary diagnosis, major past medicalhistory, days post-admission to ICU, intuba-tion=ventilation status), haematological data(e.g., haemoglobin, platelet count, white cellcount), body temperature and weight. In add-ition, for those patients with an arterial line,the ratio of partial pressure of oxygen inarterial blood to the inspired fraction of oxygen(PaO2=FIO2 ratio) was calculated from themost recent arterial blood gas (ABG) as anindication of oxygenation and respiratoryreserve (see Table 1).

For the purposes of this study, outcomemeasures were selected that were easily acces-sible in a clinical setting. Heart rate (HR) wasrecorded from the electrocardiograph (ECG)monitor, after ensuring that a satisfactory tra-cing was present. In addition to recording theabsolute value of HR, HR was also expressed asa percentage of the age predicted maximumHR (where the age predicted maximum HRequals 220 minus age, in years) to give anindication of cardiac reserve (Franklin, Whaley,and Howley, 2000; McArdle, Katch, and Katch,1996; Stiller and Phillips, 2003). The cardiacrhythm was observed on the ECG tracing andany arrhythmias documented. Systolic and dia-stolic blood pressure (BP) were recorded froman invasive arterial line or, for those patientswithout an arterial line, from an oscillometricsphygmomanometer. Arterial lines were cali-brated on a daily basis according to the RAHICU protocol. Before invasive BP measure-ments were recorded it was ensured that asatisfactory tracing was obtained. Percutaneousoxygen saturation (SpO2) was recorded using apulse oximeter with a finger or ear probe, afterensuring that a satisfactory tracing was estab-lished and that the HR on the oximeter wassimilar to that seen on the ECG. In addition tothese objective parameters, patient appearancewas documented and the following noted:conscious state, respiratory pattern, pallor,flushing, sweating, clamminess, cyanosis, visibleor patient reported signs of pain, discomfort orfatigue. Any deterioration in the patient’s con-dition during the mobilisation treatment was

176 K. STILLER ET AL.

Fig. 1 Overview of safety issues prior to mobilizing acutely ill in-patients (from Stiller and Phillips, 2003).

MOBILISATION AND INTENSIVE CARE PATIENTS 177

recorded and any intervention required in itsmanagement was noted. These outcome mea-sures were recorded during a baseline periodjust prior to the mobilisation treatment, duringeach mobility task (within the first 30 seconds ofcompletion of the task), and within one minuteof completion of the entire mobilisation treat-ment when the patient had been returned to aresting position.

Interval data from the different timeperiods were compared using the repeatedmeasures analysis of variance test. When asignificant time effect was found, paired t testswere used to identify which time periods weresignificantly different. Probability values of lessthan 0.05 were considered significant.

RESULTS

A total of 160 patients were in the RAH ICUduring the study period, with 31 patients(19.3%) receiving mobilisation as part of theirphysiotherapy management. The 160 patientsreceived a total of 425 physiotherapy assess-ments=treatments over the study period, with69 (16.2%) of these treatments includingmobilisation. There were 129 patients who wereexcluded from the study as they did not receivemobilisation as part of their physiotherapymanagement. There were a variety of reasonsfor exclusion such as reduced conscious state,unstable cardiovascular and=or respiratorystatus, or other precluding factors (e.g., spinalor pelvic fracture; see Figure 1).

Table 1 provides descriptive informationand background biochemical and haematolo-gical data for the 31 patients included in thestudy. As can be seen in this table, many of thepatients included in the study had limited

Table 1Background data for the 31 patients

Sex � n (%)Male 18 (58.1%)Female 13 (41.9%)

Age (years)Mean�SD 57� 15Range 20�81

Primary diagnosis � n (%)Medical 15 (48.4%)Surgical 12 (38.7%)Trauma 4 (12.9%)

Past medical history � n (%)Nil relevant 12 (38.7%)Hypertension 8 (25.8%)Obesity 4 (12.9%)Chronic obstructive pulmonarydisease

4 (12.9%)

Ischaemic heart disease 3 (9.7%)Symptoms pre-treatment n (%)Nil 24 (77.4%)Shortness of breath 6 (19.4%)Restless 1 (3.2%)

Previous mobility n (%)Independent 31 (100%)

Days post-admission to ICUMean�SD 29� 19.6Range 1�71

Intubation and ventilationstatus n (%)Not intubated, spontaneouslyventilating

18 (58.1%)

Tracheostomy, spontaneouslyventilating

6 (19.4%)

Tracheostomy, assistedventilation

7 (22.6%)

Haemoglobin (g=dL)Mean�SD 9.1� 1.6Range 7.0�15.8

Platelet count (cells=mm3)Mean�SD 301� 170Range 42�742

White cell count (cells=mm3)Mean�SD 10,500�3,600Range 4,400�20,100

Body temperature( + Celsius)Mean�SD 37.2�0.5Range 36.0�38.2

Blood glucose (mmol=L)Mean�SD 7.3� 2.5Range 4.0�13.6

Weight (kg)Mean�SD 83� 34Range 35�160

Pre-treatment PaO2=FIO2

ratio (n ¼ 65)Mean�SD 263� 112Range 124�587100�200: n (%) 19 (29.2%)201�300: n (%) 27 (41.5%)>300: n (%) 19 (29.2%)

Pre-treatment heart rate (n¼ 69)Mean�SD (bpm) 94.1� 14.5Range (bpm) 69�133

<50% age predictedmaximum: n (%)

14 (20.3%)

50 � 70% age predictedmaximum: n (%)

47 (68.1%)

71 � 80% age predictedmaximum: n (%)

7 (10.1%)

>80% age predictedmaximum: n (%)

1 (1.4%)

178 K. STILLER ET AL.

cardiac reserve at rest, as indicated by the pre-treatment HR being more than 50 per cent ofthe age predicted maximum on 55 of the 69occasions (79.7%) of mobilisation (Stiller andPhillips, 2003; see Figure 1). Marginal respira-tory reserve at rest was also evident for somepatients, in that the pre-treatment PaO2=FIO2

ratio was less than 300 on 46 of 65 occasions(70.7%) in patients with available ABGs (Stillerand Phillips, 2003, see Figure 1). In total, 63 ofthe 69 mobilisation treatments (91.3%) wereperformed with patients who had marginalcardiac and=or respiratory reserve at rest(i.e., pre-treatment HR more than 50% agepredicted maximum and=or PaO2=FIO2 lessthan 300).

The 69 mobilisation treatments received bythe 31 patients involved:

� sitting on the edge of the bed on 39occasions

� sitting on the edge of the bed and standingon 19 occasions

� sitting on the edge of the bed and standingtransfer to a chair on 10 occasions and

� sitting on the edge of the bed, standing andwalking on one occasion.

For the purposes of presenting the results (seeTable 2 and Figures 2, 3, and 4), the firstmobility task refers to the first activity duringthe mobilisation treatment (in all cases this wassitting on the edge of the bed). For thosepatients who progressed beyond sitting on theedge of the bed, the second mobility task refersto the second activity during mobilisation (i.e.,standing or standing transfer). As only onepatient progressed beyond sitting on the edgeof the bed and standing (i.e., to walking),data from the walking component of themobilisation treatment were not analysed.

Figures 2 to 4 show the mean (SD) valuesfor HR, BP and SpO2 over the 69 occasions ofmobilisation at the different time intervals.A significant change was seen over time forHR (absolute and percentage age predictedmaximum HR) and for systolic and diastolic BP(p< 0.001). Using paired t tests, HR (absoluteand percentage age predicted maximum HR)significantly increased from pre-treatment to thefirst mobility task (p< 0.001). A further sig-nificant increase was seen from the first to thesecond mobility task (p< 0.001). Post-treatment,HR was still significantly increased from thepre-treatment level (p< 0.001). Similar changes

Table 2Changes in haemodynamic and respiratory data during the 69 occasions of mobilisation for the 31 patients

Pre-treatment to firstmobility task

First to secondmobility task

Pre-treatment topost-treatment

Heart rate n¼ 69 n¼28 n¼69Fall 12 (17.4%) 6 (21.4%) 13 (18.8%)No change 4 (5.8%) 0 4 (5.8%)Increase 53 (76.8%) 22 (78.6%) 52 (75.4%)

Systolic BP n¼ 61 n¼22 n¼63Fall�20 mmHg 1 (1.6%) 3 (13.6%) 19 (30.2%)Fall>20 mmHg 8 (13.1%) 5 (22.7%) 1 (1.6%)No change 0 4 (18.2%) 4 (6.3%)Increase 52 (85.2%) 10 (45.5%) 39 (61.9%)

Diastolic BP n¼ 61 n¼22 n¼63Fall�10 mmHg 4 (6.6%) 8 (36.4%) 10 (15.9%)Fall>10 mmHg 0 2 (9.1%) 3 (4.8%)No change 2 (3.3%) 4 (18.2%) 5 (7.9%)Increase 55 (90.2%) 8 (36.4%) 45 (71.4%)

SpO2 n¼ 69 n¼26 n¼69Fall<4% 24 (34.8%) 5 (19.2%) 17 (24.6%)Fall�4% 10 (14.5%) 3 (11.5%) 7 (10.1%)No change 20 (29.0%) 7 (26.9%) 19 (27.5%)Increase 15 (21.7%) 11 (42.3%) 26 (37.7%)

MOBILISATION AND INTENSIVE CARE PATIENTS 179

were seen for systolic and diastolic BP, exceptthat BP during the second mobility task was notsignificantly different from the first mobilitytask. Although the changes seen in HR and BP

over time were statistically significant, the mag-nitude of these changes was generally small interms of absolute and relative values, with mostvalues changing by approximately 10 per cent or

Fig. 2 Heart rate response to mobilisation (means, error bars represent SD).

Fig. 3 Blood pressure response to mobilisation (means, error bars represent SD).

180 K. STILLER ET AL.

less. A fall in SpO2 was seen during mobilisation,but this was not sufficient to achieve statisticalsignificance (p¼ 0.44).

Table 2 provides the distribution of responsesseen during mobilisation. Although the majorityof patients showed the expected response in HRand BP (i.e., an increase during mobilisation),there were a number of patients where HR andBP fell during mobilisation.

The most common change in patientappearance seen during the mobilisation treat-ment was an alteration in respiratory pattern(e.g., increased respiratory rate and=orincreased use of accessory muscles of respira-tion), which was seen on 10 occasions (14.5%).On two of the 69 occasions of mobilisation(2.9%), patients reported dizziness during themobilisation treatment, but this was notaccompanied by orthostatic hypotension, nordid it limit mobilisation or require any directmedical intervention. Cardiac arrhythmias werenoted pre-treatment on eight occasions(11.6%)—four cases of atrial fibrillation andfour cases of occasional premature ventricularcontractions. In each instance the cardiacarrhythmia had been present for some timeand did not require medication, nor did itaffect haemodynamic stability. Thereforemobilisation was deemed safe to proceed (seeFigure 1). No change in the severity or

frequency of the arrhythmias was noted duringmobilisation.

On three of the 69 occasions of mobilisa-tion (4.3%), a deterioration in a patient’s con-dition occurred that required specific inter-vention. The salient features of the threepatients who deteriorated during mobilisationare provided in Table 3. In all three cases, thedeterioration was a fall in SpO2. In One case(patient 1 in Table 3), the desaturation occur-red during the first occasion on which mobili-sation was attempted. For the other twopatients, the destauration occurred on thesecond occasion of mobilisation. For two of thepatients the desaturation occurred while sittingon the edge of the bed (patients 2 and 3, Table3) and the other patient (patient 1) desatu-rated when proceeding from sitting to stand-ing. In each case the patient required atemporary increase in FIO2, which resulted inan improvement in SpO2 and did not necessi-tate further intervention or termination of themobilisation treatment. To determine if therewere any features that were able to predictpatients likely to deteriorate during mobilisa-tion, the data from the three patients whosecondition deteriorated during mobilisationwere further reviewed. All three patients (seeTable 3) had pre-treatment HRs that weremore than 60 per cent of their age predicted

Fig. 4 Percutaneous saturation response to mobilisation (means, error bars represent SD).

MOBILISATION AND INTENSIVE CARE PATIENTS 181

maximum, suggesting particularly limited car-diac reserve at rest (see Figure 1). However, 13other patients had pre-treatment HRs higherthan 60 per cent of their age predicted max-imum and did not demonstrate any adverseeffects during mobilisation. Only one of thethree patients (patient 2) had a pre-treatmentBP that would be considered abnormal. How-ever, her BP had been stable at this low level,without inotropic assistance, for several daysand she required no treatment for the hypo-tension, therefore mobilisation was deemedsafe to proceed (see Figure 1). As can be seenfrom Table 3, this patient showed a markedincrease in BP during mobilisation. However,this measurement may be inaccurate due tothe position of the arterial line. As far asoxygenation is concerned, patient 2 had parti-cularly marginal respiratory reserve pre-treat-ment (i.e., a PaO2=FIO2 ratio of 145; see Figure1). However, six other patients had a ratio lessthan 145 and did not deteriorate duringmobilisaton. As far as pre-treatment SpO2 isconcerned, there were only two of the 69occasions of mobilisation when this was lessthan 90 per cent, one of which was for patient 1who then went on to desaturate during mobi-lisation. Therefore, while a pre-treatment SpO2

of less than 90 per cent seemed to be predictiveof desaturation during mobilisation, it is not

possible to draw firm conclusions from thissmall patient sample. No other factors wereable to be identified that could predict thosepatients who deteriorated during mobilisation.Despite the marginal cardiovascular and=orrespiratory reserve of these three patients (seeFigure 1), mobilisation was undertaken as itsperceived benefits were thought to outweighthe potential risks.

DISCUSSION

This study found that mobilisation wasassociated with significant increases in HR,systolic and diastolic BP, and a decrease inSpO2. Although the changes were statisticallysignificant for HR and BP, the magnitude ofthe changes was of minor clinical importance.There were only three episodes of major clin-ical importance (4.3%) when specific inter-vention was required during mobilisation tostabilise haemodynamic and=or respiratorystatus, with all three patients respondingquickly to minimal intervention. Thus, mobili-sation was well tolerated in those patientsdeemed suitable for mobilisation, even thoughtheir pre-treatment data suggested limited car-diac and=or respiratory reserve (see Figure 1).

Table 3Characteristics of the three patients who deteriorated during mobilisation

Patient 1 Patient 2 Patient 3

Sex=age (years) F=70 F=62 M=74Primary diagnosis Exacerbation

COPDPost-operativerespiratory failure

Respiratoryfailure

Past medical history COPD, malnutrition Nil CLLWeight (kg) 35 40 75Symptoms SOB pre-treatment Nil NilDays post-admission 1 56 23Intubation status Not intubated Tracheostomy TracheostomyVentilation status Nasal speculae Pressure support Pressure supportPre-treatment PaO2=FIO2 232 145 291Pre-treatment HR (% age pred max) 66.0 60.8 66.4Highest HR during mobilisation(% age pred max)

73.3 55.1 69.2

Pre-treatment BP (mmHg) 145/65 95/51 146/45BP during mobilisation 150/64 189/100 158/64Pre-treatment SpO2 87 92 97Lowest SpO2 during mobilisation 78 88 87

182 K. STILLER ET AL.

Although a low incidence of problemsduring mobilisation was found in this study, it isimperative to stress that a comprehensivescreening process (Stiller and Phillips, 2003)was used to select suitable patients for mobili-sation. Additionally, appropriate precautionswere taken prior to, during and after mobili-sation. The screening process presented in flowchart format in Figure 1 is a simplified versionof the guidelines published by Stiller andPhillips (2003). In the current study, despitethe majority of patients showing limited pre-treatment cardiac and=or respiratory reserveaccording to the flow chart (see Figure 1), theperceived benefits of mobilisation weredeemed to outweigh the perceived risks. As isexplained more fully in the complete set ofguidelines (Stiller and Phillips, 2003), theparameters shown in Figure 1 are not intendedto be contraindications to mobilisation orinterpreted in isolation, but instead should beused in conjunction with sound clinical judge-ment. Experienced clinicians are often able todiscern which patients will tolerate mobilisa-tion despite marginal cardiac and=or respira-tory reserve at rest. This relies on the ability ofthe experienced clinician to take into accountthe more objective parameters (see Figure 1)and also to observe and interpret more sub-jective factors, such as patient appearance,conscious state and level of pain and fatigue.For example, as noted by Stiller and Phillips(2003), patient appearance (e.g., facialexpression, cyanosis, pallor, flush, clamminess,sweatiness, anxiety) can provide the discerningclinician with essential information regardinghow well a patient will tolerate mobilisation—information that may not be evident with othermeasures. With experience, clinicians can syn-thesise all the information available and dis-criminate between patients who will or will nottolerate active mobilisation, despite marginalreserve. The low incidence of problems in thisstudy suggests that this screening process,which includes clinical judgment, can assist inthe identification of patients who will toleratemobilisation. Additionally, this overview ofsafety issues prior to commencing mobilisationwas able to highlight those patients likely to

have potential problems and help identifywhich systems were likely to be challengedduring mobilisation.

The haemodynamic responses that mostpatients showed during mobilisation were asanticipated, in that there was a progressiveincrease in HR during mobilisation and a returnto near baseline levels at the completion of themobilisation treatment (Franklin et al, 2000;McArdle et al, 1996; Selwyn and Braunwald,2001). Blood pressure (diastolic and systolic)showed a similar response to HR. However it wasevident that the increases in BP seen forpatients with invasive arterial lines often seemedexcessive. This is likely to reflect the inaccuracyof invasive BP measurement when the arterialline is moved from the position in which it hasbeen calibrated. The haemodynamic responsesseen during mobilisation in this study weresimilar to those reported during respiratoryphysiotherapy treatment (Cohen, Horiuchi,Kemper, and Weissman, 1996; Klein et al, 1988;Weissman et al, 1984) and other routine ICUactivities (e.g, movement of the body and limbs,physical examination; Weissman et al, 1984).

It was anticipated that oxygenation of theseacutely ill patients would improve duringmobilisation, due to the expected beneficialeffects of the upright position on lung volumesand ventilation=perfusion distribution (Dean,1985; Dean and Ross, 1992a, 1992b; Ross andDean, 1992; Wong, 1999). Instead, in thissample of acutely ill patients, SpO2 decreasedduring the first mobility task and showed afurther decrease during the second mobilitytask. This fall in SpO2 most likely reflects that,despite the theoretical benefits, the patients’cardiorespiratory systems could not meet theincreased oxygen demand imposed by themobilisation treatment. However, these decrea-ses did not achieve statistical significance, nor,as a fall in SpO2 of four per cent or more isusually required to be considered clinically sig-nificant (Franklin et al, 2000; see Figure 1),would they be considered clinically significant.

It could be argued that the three patientswhose condition deteriorated during mobilisa-tion should not have been mobilised at all basedon their pre-treatment data (see Figure 1).

MOBILISATION AND INTENSIVE CARE PATIENTS 183

However, for all three patients it was thoughtthat the potential benefits of mobilisation out-weighed the potential risks. Furthermore, eventhough all three patients desaturated duringmobilisation, they quickly recovered once FIO2

was increased, which vindicated the decisionto perform mobilisation. Hypothetically, ifincreasing the FIO2 had not improved SpO2,appropriate interventions may have includedterminating the mobilisation treatment and, itnecessary, increasing the level of ventilatorysupport. In a similar way that pre-oxygenationprior to suction has been shown to preventsuction induced hypoxaemia (Chulay, 1988;Ciesla, 1996; Mancinelli-Van Atta and Beck,1992), it is possible that increasing FIO2 prior tomobilisation may be beneficial for patients withmarginal oxygenation.

Further research should be undertakenwith similar patient groups to confirm thefindings of this study. This may help to identifyfactors that predict which patients are likely todeteriorate during mobilisation. It may also behelpful in future research to measure oxyge-nation during mobilisation using parametersobtained from ABGs, such as the PaO2=FIO2

ratio, as this takes into account the FIO2 andthus more accurately reflects oxygenation andthe underlying respiratory reserve (Stiller andPhillips, 2003). However, the frequent mea-surement of ABGs is often impractical in theclinical setting, whereas SpO2, by virtue ofbeing a non-invasive measurement, providesinstantaneous feedback to the clinician. Addi-tionally, this study only measured patients for ashort time after the completion of the mobili-sation treatment, and longer term effects ofmobilisation could be investigated. Althoughrandomised controlled studies would moreclearly establish the role of mobilisation in therecovery of acutely ill patients, it may be diffi-cult to withhold mobilisation from an ethicalviewpoint.

CONCLUSION

This study found that mobilisation of acutely illICU patients resulted in a significant increase

in HR and BP, and a fall in SpO2. Althoughsome changes were statistically significant, themagnitude of the changes was of little clinicalimportance. There were only three episodesthat were deemed of major clinical importance(4.3%), in that specific intervention wasrequired. Although most patients demon-strated marginal cardiac and=or respiratoryfunction pre-treatment, mobilisation was welltolerated in this patient sample. Thus, ifappropriate screening procedures and pre-cautions are taken prior to and during mobili-sation, acutely ill ICU patients deemed suitablefor mobilisation can be safely mobilised with-out major deterioration in their clinical status.

Acknowledgments

Special thanks to Naomi Haensel, Shane Patman,and Louise Wiles for their helpful comments.

ReferencesBishop KL 1996 Pulmonary Rehabilitation in the IntensiveCare Unit. In: Fishman AP (ed) Pulmonary Rehabilita-tion pp 725�738. Marcel Dekker Inc, New York.

Chulay M 1998 Arterial blood gas changes with a hyper-inflation and hyperoxygenation suctioning interventionin critically ill patients. Heart Lung 17: 654�661.

Ciesla ND 1996 Chest physical therapy for patients in theintensive care unit. Physical Therapy 76: 609�625.

Cohen D, Horiuchi K, Kemper M, Weissman C 1996 Mod-ulating effects of propofol on metabolic and cardio-pulmonary responses to stressful intensive care unitprocedures. Critical Care Medicine 24: 612�617

Dean E 1985 Effect of body position on pulmonary func-tion. Physical Therapy 65: 613�618

Dean E 1994 Oxygen transport: a physiologically-basedconceptual framework for the practice of cardio-pulmonary physiotherapy. Physiotherapy 80: 347�355

Dean E, Ross J 1992a Discordance between cardio-pulmonary physiology ad physical therapy: toward arational basis for practice. Chest 101: 1694�1698

Dean E, Ross J 1992b Oxygen transport. The basis forcontemporary cardiopulmonary physical therapy and itsoptimization with body positioning and mobilization.Physical Therapy Practice 1: 34�44

Franklin BA, Whaley MH, Howley ET 2000 ACSM’s guide-lines for exercise testing and prescription, 6th edn,Philadelphia, Lippincott Williams and Wilkins

Klein P, Kemper M, Weissman C, Rosenbaum SH,Askanazi J, Hyman Al 1988 Attenuation of the hemo-dynamic responses to chest physical therapy. Chest 93:38�42

Mancinelli-Van Atta J, Beck SL 1992 Preventing hypoxemiaand hemodynamic compromise related to endotrachealsuctioning. American Journal of Critical Care 1: 62�79

184 K. STILLER ET AL.

McArdle WD, Katch Fl, Katch VL 1996 Exercise Physiology,4th edn, Baltimore, Williams and Wilkins

Ross J, Dean E 1992 Body Positioning. In: Zadai C (ed)Clinics in Physical Therapy. Pulmonary Management inPhysical Therapy pp 79�98. New York, ChurchillLivingstone Inc.

Selwyn AP, Braunwald E 2001 Ischemic Heart Disease. In:Braunwald E, Fauci AS, Kasper DL, Hauser SL, Longo DL,Jameson JL (eds) Harrison’s Principles of InternalMedicine, pp 1399�1410, 15th ed. New York, McGraw-Hill

Stiller K, Phillips A 2003 Safety aspects o mobilising acutelyill in-patients. Physiotherapy Theory and Practice 19:239�257

Weissman C, KemperM, DamaskMC, Askanazi J, Hyman Al,Kinney JM 1984 Effect of routine intensive care interac-tions on metabolic rate. Chest 86: 815�818

Wong WP 1999 Use of body positioning in the mechanicallyventilated patient with acute respiratory failure: applica-tion of Sackett’s rules of evidence. Physiotherapy Theoryand Practice 15: 25�41

MOBILISATION AND INTENSIVE CARE PATIENTS 185