PTJ: Physical Therapy & Rehabilitation Journal | Physical Therapy, 2022;102:1–9

https://doi.org/10.1093/ptj/pzac056
Advance access publication date May 13, 2022
Original Research

SAfety and Feasibility of EArly Resistance Training After

Median Sternotomy: The SAFE-ARMS Study
Jacqueline Pengelly , PhD1 ,2 ,* , Stuart Boggett, MC-BMedSC3 , Adam Bryant, PhD4 ,
Colin Royse, MD1 ,3 ,5 ,6 , Alistair Royse, MD1 ,3 ,7 , Gavin Williams, PhD4 , Doa El-Ansary, PhD1 ,3 ,8
1 Department of Nursing and Allied Health, Swinburne University of Technology, Hawthorn, Victoria, Australia
2 Department of Exercise & Sport Science, Institute of Health and Wellbeing, Federation University Australia, Mount Helen, Victoria, Australia
3 Department of Surgery, University of Melbourne, Parkville, Victoria, Australia

Downloaded from https://academic.oup.com/ptj/article/102/7/pzac056/6585156 by guest on 15 May 2023

4 Department of Physiotherapy, University of Melbourne, Parkville, Victoria, Australia
5 Department of Anaesthesia and Pain Management, Royal Melbourne Hospital, Parkville, Victoria, Australia
6 Outcomes Research Consortium, Cleveland Clinic, Cleveland, Ohio, USA
7 Department of Cardiothoracic Surgery, Royal Melbourne Hospital, Parkville, Victoria, Australia
8 Clinical Research Institute, Westmead Private Hospital, Westmead, New South Wales, Australia

*Address all correspondence to Dr Pengelly at: [email protected]

Abstract
Objective. The purpose of this study was to determine the safety and feasibility of subacute upper limb resistance exercise
on sternal micromotion and pain and the reliability of sternal ultrasound assessment following cardiac surgery via median
sternotomy.
Methods. This experimental study used a pretest–posttest design to investigate the effects of upper limb resistance exercise
on the sternum in patients following their first cardiac surgery via median sternotomy. Six bilateral upper limb machine-based
exercises were commenced at a base resistance of 20 lb (9 kg) and progressed for each participant. Sternal micromotion
was assessed using ultrasound at the mid and lower sternum at 2, 8, and 14 weeks postsurgery. Intrarater and interrater
reliability was calculated using intraclass correlation coefficients (ICCs). Participant-reported pain was recorded at rest and
with each exercise using a visual analogue scale.
Results. Sixteen adults (n = 15 males; 71.3 [SD = 6.2] years of age) consented to participate. Twelve participants completed
the study, 2 withdrew prior to the 8-week assessment, and 2 assessments were not completed at 14 weeks due to assessor
unavailability. The highest median micromotion at the sternal edges was observed during the bicep curl (median = 1.33 mm;
range = −0.8 to 2.0 mm) in the lateral direction and the shoulder pulldown (median = 0.65 mm; range = −0.8 to 1.6 mm) in
the anterior–posterior direction. Furthermore, participants reported no increase in pain when performing any of the 6 upper
limb exercises. Interrater reliability was moderate to good for both lateral–posterior (ICC = 0.73; 95% CI = 0.58 to 0.83) and
anterior–posterior micromotion (ICC = 0.83; 95% CI = 0.73 to 0.89) of the sternal edges.
Conclusion. Bilateral upper limb resistance exercises performed on cam-based machines do not result in sternal micromotion
exceeding 2.0 mm or an increase in participant-reported pain.
Impact. Upper limb resistance training commenced as early as 2 weeks following cardiac surgery via median sternotomy
and performed within the safe limits of pain and sternal micromotion appears to be safe and may accelerate postoperative
recovery rather than muscular deconditioning.
Keywords: Cardiothoracic Surgery, Exercise, Rehabilitation, Resistance Training, Ultrasound

Received: August 19, 2021. Revised: November 22, 2021. Accepted: February 17, 2022
© The Author(s) 2022. Published by Oxford University Press on behalf of the American Physical Therapy Association.
This is an Open Access article distributed under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/
by/4.0/), which permits unrestricted reuse, distribution, and reproduction in any medium, provided the original work is properly cited.
2 Sternotomy: SAFE-ARMS Study

Introduction Methods
Median sternotomy remains the gold standard surgical Design
incision for optimal access to the heart and vasculature This was a pretest–posttest design study, nested within a
for cardiac surgery.1–3 However, patients can experience pilot randomized controlled trial,14 conducted in accor-
deficits in physical and functional recovery for several dance with the National Statement on Ethical Human
months after surgery, impacting performance of activities Research and the Australian Code for the Responsible
of daily living.4,5 A disconnect exists between routine Conduct of Research. Ethical approval was granted by
restrictive sternal precautions, such as carrying weights the Melbourne Health Human Research Ethics Committee
>4.5 kg, and emerging evidence, which poses a chal- (Application ID: 2017.266). Local governance approval was
lenge for exercise professionals to deliver evidence-based obtained prior to recruitment commencement. The study
rehabilitation.6,7 followed the CONsolidated Standards of Reporting Trials
Sternal precaution development was founded on the belief 2010 guidelines. This trial was prospectively registered
that early postoperative (1–6 weeks) arm movements increase with the Australian New Zealand Clinical Trials Registry
the mechanical forces acting in opposition to the holding (ACTRN12617001430325p).

Downloaded from https://academic.oup.com/ptj/article/102/7/pzac056/6585156 by guest on 15 May 2023

strength of wire sutures (lateral movement), thus jeopardizing Participants were in the resistance training arm of the Super-
the sternal wiring and skin integrity and resulting in sternal vised Early Resistance Training study.14 The protocol of the
dehiscence.2–8 In a study of sternal closure techniques in primary study, detailing recruitment, inclusion and exclusion
cadavers, McGregor et al1 suggested that activation of the criteria, randomization, allocation concealment and blinding,
muscles of the anterior chest may result in distractive forces has been described previously.15 Written informed participant
and micromotion of the bone edges >2.0 mm,1 which in turn consent was obtained prior to data collection.
may jeopardize the integrity of the healing sternum. However,
a recent study has shown that unweighted unilateral and Role of the Funding Source
bilateral upper limb movements, performed within 6 weeks The funders played no role in the design, conduct, or reporting
poststernotomy, were not detrimental to bone healing, were of this study.
well tolerated, decreased postsurgical pain, and resulted in
<2.0 mm of sternal micromotion, assessed using the Ster- Participants
nal Instability Scale and sternal ultrasound.9 Furthermore, Data from participants in the resistance training group of
coughing was found to result in the greatest sternal sepa- a pilot randomized controlled trial were analyzed in this
ration during the early postoperative period.9 Subsequently, study.14 Participants were adults undergoing cardiac surgery
evidence has prompted a paradigm shift towards promotion via median sternotomy at the Royal Melbourne or Melbourne
of upper limb exercise within the same limits of pain and Private hospitals between April 16, 2018, and August 31,
discomfort.10–12 2019. Eligibility criteria for the pilot randomized controlled
A recent systematic review of resistance training following trial included (1) first elective cardiac surgery via median
median sternotomy7 found that none of the 18 included sternotomy, (2) no diagnosed cognitive impairment, (3) no
studies met all American College of Sports Medicine Car- musculoskeletal conditions limiting exercise ability, (4) suf-
diac Rehabilitation resistance training guidelines in regard to ficient English to provide informed consent and complete
program length (4–6 months), intensity (40%–50% maximal surveys, and (5) able to attend the exercise testing site for
voluntary contraction), frequency (2–3 d/wk), session dura- exercise and/or testing sessions.14,15
tion (2–4 sets, 12–15 reps, 8–10 exercises), and resistance
(0.5–0.9 kg in the first 12 weeks poststernotomy).13 Despite Exercise Intervention
the American College of Sports Medicine recommendation Participants in the resistance training group completed a
that sternal stability be assessed before commencing upper twice-weekly, moderate-intensity resistance training program
limb resistance exercise, no studies reported that they assessed for 12 weeks under the supervision of an accredited exercise
sternal stability prior to resistance training commencement or physiologist. The 6 dynamic upper limb exercises performed
progression.7 on cam-based machines (ie, machines with a cam device that
It is common practice for resistance exercise to be pre- distribute the load evenly throughout the entire exercise range
scribed more conservatively than guidelines recommend, pos- of movement) that were assessed in this study consisted of
sibly attributable to the lack of evidence investigating the a seated row, shoulder pulldown, shoulder press, triceps dip,
effect of upper limb resistance exercises on sternal healing. lateral raise, and bicep curl. A single set of each exercise was
Therefore, the aims of this study were to determine whether completed to volitional fatigue (typically 10–15 repetitions).
(1) bilateral machine-based resistance training commenced 2 The exercise intervention was previously reported.14,15
weeks following cardiac surgery via median sternotomy is Each cam-based resistance machine was adjusted for each
safe and feasible, and (2) sternal ultrasound during bilateral participant to ensure optimal comfort and uniformity of
resistance training is a reliable measure of sternal micromo- the testing set-up and exercise performance (Supplementary
tion. We hypothesized that (1) the seated row would result in Materials A & B).
the greatest amount of sternal edge micromotion at 2 weeks
postoperatively, and (2) weighted upper limb exercises (eg, Outcome Measures
triceps dip, bicep curl, shoulder pulldown, shoulder press, Sternal Micromotion
lateral raise, and seated row) would not exceed 2.0 mm of Real-time sternal images were obtained using a L38v 9 cm
sternal edge separation at any time point (2, 8, and 14 weeks linear transducer (10–5 MHz) and Fujifilm SonoSite iViz
postoperatively). ultrasound (SonoSite Australasia Pty Ltd, Brookvale, NSW,
Pengelly et al 3

Australia). Sternal edge micromotion was measured using

the annotation and measurement function of the Fujifilm
Prosolv5 Synapse Cardiovascular software program (Fujifilm
Australia). Measurement of sternal edge separation and/or
overlap was recorded in the lateral (coronal plane) and ante-
rior–posterior (sagittal plane) directions in millimeters.16

Sternal Pain
Sternal pain was evaluated using the Visual Analogue Scale
because this has been reported as a valid and reliable measure
of postoperative pain.17 Participants were required to rate the
maximal amount of sternal pain they experienced at rest and
during each of the 6 upper limb tasks, with any increase in
pain from rest during the 6 upper limb resistance exercises

Downloaded from https://academic.oup.com/ptj/article/102/7/pzac056/6585156 by guest on 15 May 2023

being recorded. Pain was rated on a scale of 0 to 10, with a
rating of 0 corresponding to no sternal pain and a rating of 10
indicative of the worst possible pain.17 Pain was not recorded
during the cough because it was used as a comparator with
upper limb resistance exercise sternal micromotion.

Testing Procedure
Figure 1. Ultrasound image analysis procedure. (A) Identification of
Sternal motion testing occurred during 3 testing sessions: (1) sternal edge tips. (B) Horizontal and vertical axis. (C) Vertical
at 2 weeks, (2) at 8 weeks, and (3) at 14 weeks postoper- (anterior–posterior) vector measurement. (D) Horizontal (lateral) vector
atively. To minimize systemic errors of measurement, these measurement.
sessions were standardized and ultrasound images obtained
by a single assessor (J.P.). The assessor received training in each hemi-sterna (Fig. 1A and B) to assist with measurement
sternal ultrasound imaging by a senior physiotherapist, who of the vertical and horizontal vectors between the sternal
had developed and validated this measurement technique,16 in edges. Vertical vector measurement was taken from the right
addition to a cardiothoracic surgeon. The assessor acquired 1 sternal edge tip to the intersection of the left sternal edge inter-
year of experience obtaining sternal ultrasound images in the section (upper and lower points of axes intercepts) (Fig. 1C).
clinical setting prior to study commencement. Horizontal vector measurement was taken from the tip of
A single, 2-dimensional, 15-second ultrasound video was the left sternal edge to the intersection of the axes (Fig. 1D).
taken capturing motion at the sternal edges at rest (baseline Sternal separation that increased from rest (ie, sternal edges
measurement) during a maximal cough and 6 dynamic upper moved further apart) was denoted by a positive value, and
limb resistance exercises. Sternal ultrasound measurement sternal separation that decreased from rest (ie, sternal edges
occurred throughout the entire range of movement for the 6 moved closer together) was denoted by a negative value.
exercises.
Markings were placed on the participant’s skin with a Reliability
surgical marker at points 6 cm (mid-sternum) and 10 cm Eighty ultrasound images were randomly selected using a
(lower sternum) distal from the sternal notch. The ultrasound computer-generated sequence from Research Randomizer
transducer was placed directly onto the skin, with minimal (randomizer.org). To evaluate intrarater and interrater
pressure applied, throughout the duration of the aforemen- reliability, 2 raters independently assessed each of the 80
tioned activities. Ultrasound images were obtained at both images twice, with 1 month between each assessment.
the mid- and lower-sternum locations for each activity. The Rater 1 (J.P.) was deemed to have intermediate experience
activities were repeated if the transducer moved during the (>100 hours) completing sternal ultrasound measurement
movement to obtain an optimal image. Participants were using Prosolv5, whereas Rater 2 (S.B.) was deemed a novice
instructed to perform a maximal cough, as hard as possible, (<20 hours experience). To ensure Rater 2 was familiar with
and each of the upper limb exercises within their comfortable the study protocol, including identification of the sternal
range of motion to minimize participant burden. Although edges, image annotation, and the methodology of obtaining
the order of upper limb exercises was not randomized, the measurements, Rater 2 attended four 1-hour training sessions
order of performance was dependent on exercise machine under the guidance of Rater 1. Three fidelity checks of 10
availability at the time of the scheduled sessions. The above ultrasound images were also conducted as part of the training
procedure was repeated at each testing session for each time and analysis process, which were cross-checked with Rater 1.
point reported.
Statistical Analysis
Ultrasound Image Analysis Procedure Continuous participant demographic data (age) were reported
Each ultrasound video was viewed in real-time to determine as mean and SD, whereas categorical participant demographic
the frame in which (1) the sternal edges could clearly be data (sex assigned at birth, comorbidities, surgical procedure,
viewed in the resting position, and (2) the maximum move- and sternal closure method) were reported as the absolute
ment for the exercise occurred. Frame-by-frame video analysis number of participants and percentage of the study popula-
was used to determine the point of maximum movement. tion. Postoperative days to start of intervention was reported
Straight-line annotations representing the vertical and hori- as median and interquartile range values. Safety during each
zontal axes were drawn through the left and right edges of exercise was determined using lateral and anterior–posterior
4 Sternotomy: SAFE-ARMS Study

Downloaded from https://academic.oup.com/ptj/article/102/7/pzac056/6585156 by guest on 15 May 2023

Figure 2. Summary of participant screening and recruitment.

sternal edge micromotion values measured by the more expe- intervention commencement. A further 2 participants with-
rienced assessor (J.P.) and was compared across the 3 time drew from the study: 1 relocated out of the area and the other
points, reported as median and minimum and maximum transferred to an alternate cardiac rehabilitation program.
values. Intraclass correlation coefficient (ICC) estimates and A summary of participant screening and enrolment is
their 95% CIs were used to calculate intrarater and inter- shown in Figure 2. Additionally, data were not collected
rater reliability. Intrarater reliability was calculated for each for 2 participants at the 14-week assessment due to
of the 2 raters individually, based on a consistency, 2-way assessor unavailability, leaving 12 participants with complete
mixed-effects model. To determine the reliability of measure- data.
ments between raters, ICC interrater reliability was calculated The mean age of participants was 71.3 (SD = 6.2) years
based on a mean-rating (k = 2), absolute-agreement, 2-way and comprised of 15 (94%) males. Twelve patients (75%)
mixed-effects model. Reliability was categorized according underwent coronary artery bypass grafting and 4 (25%) had
to Koo and Li18 ICC value classifications (<0.50 = poor, a coronary artery bypass plus a valve repair/replacement.
0.50–0.75 = moderate, 0.75–0.90 = good, >0.90 = excellent). Median sternotomy closure was achieved via sternal plating
All statistical analyses were undertaken using IBM SPSS Statis- (n = 6), conventional stainless-steel wires (n = 6), and stainless-
tics version 26 (IBM Corporation, Armonk, New York, USA). steel cables (n = 4), with the sternal closure method varying
according to surgeon preference. The resistance training inter-
vention commenced 14.5 days (interquartile range = 2.5 days)
Results postoperatively. Patients were not taking any prescribed pain
Participant Demographics medications at the time of commencing the intervention. A
Of the 20 participants randomized to the resistance training summary of baseline patient demographic data is described in
group, 4 withdrew following hospital discharge but prior to Table 1.
Pengelly et al 5

Table 1. Summary of Patient Demographics at Baselinea Tab. 3; Fig. 3D). The lowest median anterior–posterior move-
Baseline Data Values ment occurred at the lower sternum during the seated row
at 2 weeks (median = 0.00 mm; −0.9 to 1.0 mm) and the
Age, y 71.3 (SD = 6.2) bicep curl at 14 weeks (median = 0.00 mm; −0.9 to 1.0 mm;
Sex assigned at birth
Tab. 3; Fig. 3D). No patients exceeded the 2-mm micromotion
Male 15 (94%)
Female 1 (6%) threshold for any upper limb exercise performed.
Comorbidities Secondary analysis of sternal micromotion for each exer-
Overweight 6 (37.5%) cise according to the sternal closure mechanism was also
Obese 8 (50%) undertaken and is reported in Supplementary Material C,
Smoking history 10 (62.5%) Supplementary Table 1, and Supplementary Figures 1, 2, 3,
Ex-smoker 9 (56%)
and 4.
Current smoker 1 (6%)
Type 2 diabetes mellitus 8 (50%)
Kidney disease 4 (25%) Sternal Pain
Atrial fibrillation 2 (12.5%) There was no increase in participant-reported sternal pain,

Downloaded from https://academic.oup.com/ptj/article/102/7/pzac056/6585156 by guest on 15 May 2023

Asthma 2 (12.5%)
from 0 at rest, during the performance of any of the weighted
Macular degeneration 2 (12.5%)
Surgical procedure upper limb exercises (Tab. 3).
Coronary artery bypass graft 12 (75%)
CABG + valve 4 (25%) Upper Limb Exercise Resistance and Repetitions
Sternal closure method At 2 weeks postoperatively, all participants completed the 6
Plating 6 (37.5%)
Cables 4 (25%)
upper limb exercises using a 20-lb weight (the lowest machine
Wires 6 (37.5%) resistance) for 10 repetitions (Supplementary Appendix 1).
Time to intervention commencement At the subsequent time points, the greatest increase in mean
Postoperative days 14.5 (IQR = 2.5) weight lifted occurred with the triceps dip at 8 weeks and the
a IQR = interquartile
shoulder pulldown at 14 weeks, with patients completing 12
range.
repetitions at an average of 81.1 lb (SD = 26.9 lb) and 90.9 lb
(SD = 50.4), respectively (Supplementary Appendix 1).

Reliability
The ICC for interrater reliability for lateral micromotion Discussion
was between moderate and good (0.73 [0.58–0.83]), whereas This study found that bilateral weighted upper limb exercises
intrarater reliability was between good and excellent for rater performed on cam-based machines that moved in a single
1 (0.90 [0.85–0.93]) and between poor and good for rater 2 plane, in a controlled environment and supervised by an exer-
(0.63 [0.48–0.74]; Tab. 2). cise physiologist, are safe. To our knowledge, this study is the
For anterior–posterior micromotion, ICC interrater relia- first to examine the effects of upper limb resistance exercises
bility was between moderate and good (0.83 [0.73–0.89]). on sternal micromotion in the early postoperative period using
Intrarater reliability was between moderate and good for rater resistance loads exceeding those required to perform many
1 (0.74 [0.62–0.82]) and between poor and moderate for rater activities of daily living.19,20 All exercises resulted in ≤2.0-
2 (0.58 [0.41–0.71]; Tab. 2). mm sternal micromotion in the lateral and coronal planes
at every time point. Sternal healing and osteosynthesis are
Sternal Edge Micromotion enhanced by early bone stability.21 It has been hypothesized
Lateral Motion at the Sternal Edges that >2.0 mm of displacement at bony edges of a fracture
site may lead to necrosis and impair healing.1,21 McGregor
Median lateral micromotion was highest during the bicep curl
et al1 applied lateral, anterior–posterior, rostral-caudal, and
(median = 1.33 mm; −0.8 to 2.0 mm) at the mid-sternum at
simulated Valsalva distractive forces to cadavers that were
14 postoperative weeks (Tab. 3; Fig. 3A), whereas the greatest
closed using varied sternal wiring techniques and reported
range in lateral movement (4.7 mm) also occurred at the
that fractures and complete dehiscence of the sternum may
mid-sternum during a cough (median = 0.81 mm; −1.5 to
occur when exercise activates the muscles of the anterior chest
3.2 mm) at 8 weeks. The lateral raise resulted in the lowest
and displacement of the bone edges is >2.0 mm.1
median movement at 0.10 mm (−1.3 to 1.2 mm) at the lower
Furthermore, there were no clinical signs and symptoms
sternum at 8 weeks (Tab. 3; Fig. 3B). No patients exceeded
of sternal complications, such as a significant increase in
the 2-mm micromotion threshold for any upper limb exercise
pain from that reported at rest, clicking or crepitus, or pal-
performed.
pable increase in sternal separation or motion during the
performance of the 6 upper limb resistance exercises. This
Anterior–Posterior Motion at the Sternal Edges finding is consistent with prior research that has reported
Anterior–posterior micromotion is depicted in Figure 3C for the safety and efficacy of active unweighted upper limb and
the mid-sternum and Figure 3D for the lower sternum. The trunk tasks in cardiac surgery patient cohorts with uncompli-
greatest sternal movement was a 0.65-mm (−0.8 to 1.6 mm) cated sternotomy and sternal instability.9,22 Collectively, this
increase in separation from rest at the mid-sternum, which evidence supports an active participatory model of cardiac
occurred during the shoulder pulldown at 14 postoperative rehabilitation that engages patients in early active and resisted
weeks (Tab. 3; Fig. 3C). The highest variability in movement exercise to reduce pain, facilitate activities of daily living, and
was 3.1 mm, which occurred at the lower sternum during the optimize recovery, rather than a restrictive postoperative pro-
lateral raise at 8 weeks (median = 0.39 mm; −1.8 to 1.3 mm; tocol, historically derived from early cadaver studies.2,16,22
6 Sternotomy: SAFE-ARMS Study

Downloaded from https://academic.oup.com/ptj/article/102/7/pzac056/6585156 by guest on 15 May 2023

Figure 3. (A) Changes in lateral motion at the sternal edges (mm) during cough, bicep curl, triceps dip, shoulder press, lateral raise, seated row, and
shoulder pulldown at the mid-sternum at 2, 8, and 14 weeks postoperatively. (B) Changes in lateral motion at the sternal edges (mm) during cough, bicep
curl, triceps dip, shoulder press, lateral raise, seated row, and shoulder pulldown at the lower sternum at 2, 8, and 14 weeks postoperatively. (C) Changes
in anterior–posterior motion at the sternal edges (mm) during cough, bicep curl, triceps dip, shoulder press, lateral raise, seated row, and shoulder
pulldown at the mid-sternum at 2, 8, and 14 weeks postoperatively. (D) Changes in anterior–posterior motion at the sternal edges (mm) during cough,
bicep curl, triceps dip, shoulder press, lateral raise, seated row, and shoulder pulldown at the lower sternum at 2, 8, and 14 weeks postoperatively.

Table 2. ICCs for Intrarater and Interrater Reliabilitya

Standard Error of
Intrarater Reliability ICC 95% CI Cronbach Alpha
Measurement
Lateral micromotion
Rater 1 0.90 0.85 to 0.93 0.95 0.46
Rater 2 0.63 0.48 to 0.74 0.77 0.87
Interrater reliability 0.73 0.58 to 0.83 0.74 0.66
Anterior–posterior micromotion
Rater 1 0.74 0.62 to 0.82 0.85 0.51
Rater 2 0.58 0.41 to 0.71 0.73 0.60
Interrater reliability 0.83 0.73 to 0.89 0.83 0.37
a ICC = intraclass correlation coefficient.

Previous literature pertaining to exercise after cardiac point, so no reduction in sternal micromotion was observed
surgery was based on the premise that exercises involving over time.
the pectoral muscles act in direct opposition to the wire This study concurred with our hypothesis that weighted
fixation8 and the observation that distraction of the skin upper limb exercises would result in sternal micromotion
was caused by bilateral end-range shoulder extension.23 measures <2 mm at every time point recorded. This was
The seated row primarily involves horizontal abduction, independent of the mode of sternal closure. Although none
extension, and retraction of the shoulder girdle; thus, we of the 6 upper limb resistance exercises exceeded 2.0 mm,
hypothesized that the moderate-resistance seated row exercise performance of a cough did at the mid- and lower sternum
would result in the greatest amount of sternal micromotion at in the first 8 weeks postoperatively for 3 participants.
2 weeks postoperatively. However, this study found that there The difference in intrarater and interrater ICC reliability,
was negligible (≤2.0 mm) sternal edge movement at any time between raters 1 and 2, indicates that when sternal ultrasound
Pengelly et al 7

Table 3. Median (Minimum to Maximum) Sternal Micromotion (mm) for Each Activity at 2, 8, and 14 Week Postsurgerya

Activity 2 Weeks (n = 16) 8 Weeks (n = 14) 14 Weeks (n = 12)

Micromotion Pain Micromotion Pain Micromotion Pain
Lateral mid-sternum
Cough 0.15 (−1.60 to 2.50) N/A 0.81 (−1.5 to 3.2) N/A 0.28 (−0.2 to 2.6) N/A
Bicep curl 0.72 (−1.4 to 1.4) 0/10 (0) 0.83 (1.5 to 3.2) 0/10 (0) 1.30 (−0.6 to 1.9) 0/10 (0)
Triceps dip 0.42 (−0.8- to 1.9) 0/10 (0) 0.65 (−1.3 to 1.7) 0/10 (0) 0.85 (−0.6 to 2.0) 0/10 (0)
Shoulder press 0.77 (−1.3 to 1.6) 0/10 (0) 0.54 (−0.9 to 1.7) 0/10 (0) 0.55 (−0.2 to 1.6) 0/10 (0)
Lateral raise 0.64 (−1.3 to 1.5) 0/10 (0) 0.58 (0.4 to 1.9) 0/10 (0) 1.05 (−0.4 to 1.7) 0/10 (0)
Seated row 0.54 (−1.8 to 1.8) 0/10 (0) 1.13 (−0.8 to 1.7) 0/10 (0) 1.08 (−0.1 to 1.6) 0/10 (0)
Shoulder pulldown 0.33 (−0.2 to 1.3) 0/10 (0) 0.49 (−1.3 to 1.5) 0/10 (0) 0.37 (−1.1 to 2.0) 0/10 (0)
Lateral lower sternum
Cough 0.63 (−1.0 to 2.9) N/A 0.74 (−0.9 to 2.4) N/A 0.74 (−1.0 to 1.6) N/A
Bicep curl 0.80 (−1.1 to 1.7) 0/10 (0) 0.68 (−0.6 to 1.8) 0/10 (0) 1.33 (−0.8 to 2.0) 0/10 (0)
Triceps dip 0.57 (−0.9 to 2.0) 0/10 (0) 0.83 (−1.6 to 2.0) 0/10 (0) 1.26 (−0.4 to 1.7) 0/10 (0)

Downloaded from https://academic.oup.com/ptj/article/102/7/pzac056/6585156 by guest on 15 May 2023

Shoulder press 0.63 (−1.5 to 1.7) 0/10 (0) 0.65 (−1.8 to 1.8) 0/10 (0) 0.94 (−1.3 to 1.3) 0/10 (0)
Lateral raise 0.32 (−0.5 to 1.5) 0/10 (0) 0.10 (−1.3 to 1.2) 0/10 (0) 1.21 (−1.0 to 1.6) 0/10 (0)
Seated row 0.67 (−1.7 to 1.6) 0/10 (0) 0.14 (−0.4 to 1.2) 0/10 (0) 0.77 (−0.3 to 1.8) 0/10 (0)
Shoulder pulldown 0.75 (0.0 to 1.4) 0/10 (0) −0.22 (−1.5 to 1.8) 0/10 (0) 0.46 (−0.3 to 1.9) 0/10 (0)
Anterior–posterior mid-sternum
Cough 0.08 (−0.5 to 0.7) N/A −0.05 (−0.6 to 1.1) N/A 0.03 (−0.1 to 0.4) N/A
Bicep curl −0.07 (−1.9 to 1.0) 0/10 (0) 0.12 (−0.8 to 0.8) 0/10 (0) −0.33 (−1.4 to 1.5) 0/10 (0)
Triceps dip 0.15 (−1.4 to 0.8) 0/10 (0) 0.15 (−0.5 to 1.3) 0/10 (0) 0.07 (−1.0 to 0.9) 0/10 (0)
Shoulder press −0.04 (−0.6 to 0.4) 0/10 (0) 0.25 (−1.4 to 1.1) 0/10 (0) −0.20 (−1.0 to .9) 0/10 (0)
Lateral raise 0.04 (−1.0 to 0.7) 0/10 (0) 0.18 (−0.5 to 1.6) 0/10 (0) 0.27 (−0.9 to 0.7) 0/10 (0)
Seated row 0.07 (−1.0 to 1.4) 0/10 (0) −0.32 (−1.4 to 1.4) 0/10 (0) 0.02 (−0.4 to 1.6) 0/10 (0)
Shoulder pulldown 0.05 (−0.6 to 1.1) 0/10 (0) −0.09 (−1.5 to 1.1) 0/10 (0) 0.65 (−0.8 to 1.6) 0/10 (0)
Anterior–posterior lower sternum
Cough 0.04 (−0.3 to 1.9) N/A −0.09 (−0.5 to 1.7) N/A 0.20 (−0.4 to 1.1) N/A
Bicep curl 0.22 (−0.8 to 1.2) 0/10 (0) 0.09 (−0.2 to 0.9) 0/10 (0) 0.00 (−0.9 to 1.0) 0/10 (0)
Triceps dip 0.08 (−0.9 to 1.0) 0/10 (0) 0.09 (−0.6 to 1.0) 0/10 (0) −0.02 (−1.3 to 1.3) 0/10 (0)
Shoulder press 0.20 (−1.0- to 1.3) 0/10 (0) −0.08 (−0.6 to 0.9) 0/10 (0) 0.15 (−0.6 to 0.8) 0/10 (0)
Lateral raise −0.04 (−1.4 to 0.9) 0/10 (0) 0.39 (−1.8 to 1.3) 0/10 (0) −0.02 (−0.9 to 0.9) 0/10 (0)
Seated row 0.00 (−0.9 to 0.1) 0/10 (0) 0.26 (−1.1 to 1.9) 0/10 (0) 0.24 (−1.2 to 0.9) 0/10 (0)
Shoulder pulldown −0.02 (−1.2 to 0.7) 0/10 (0) 0.40 (−0.9 to 1.7) 0/10 (0) −0.07 (−0.8 to 1.5) 0/10 (0)
a N/A = not assessed.

is used to assess sternal edge micromotion, greater experience analyzing ultrasound images, which may have resulted in an
with obtaining and measuring ultrasound images may increase under-estimation of the ICC reliability. To ensure reliability of
the reliability of ultrasound measurement. However, it should the time point comparisons, safety was determined using data
be noted that in this study we calculated our measures of from the more experienced assessor.
sternal separation and micromotion on an external platform Another limitation of this pilot study is that the small
to ensure a consistent methodology and to assess reliabil- sample size that completed the testing (n = 12) were predom-
ity of calculating the measures. This is not usual practice inantly men (n = 11). There are potential differences in bone
because most sonographers obtain measures in real-time at healing between men and women and particularly in those of
the time of image acquisition by using the ultrasound machine advanced age (ie, 60s vs 80s). As such, a larger sample size
calipers. is warranted to investigate the prevalence of pain or other
A strength of the study is that the exercise equipment sternal complications.
utilized for resistance training was purpose built to facilitate Although the findings of this study inform safety and
activation of desired muscle groups and stabilize other body feasibility of early postoperative resistance exercise training, it
segments by way of seatbelts. This ensured optimal body should be noted that the exercises were performed bilaterally
biomechanics during each exercise without accessory move- on specialized exercise machines. Thus, the findings may
ments and assisted injury risk reduction. The findings also not translate to unilateral upper limb resistance exercises
suggest that performing bilateral daily tasks, such as pulling or to home-based equipment or exercises. Because the
a chair out, opening doors, lifting groceries, or carrying a exercise equipment is not routinely used in broader cardiac
load of washing,19,20 could be performed safely where patient rehabilitation programs, it is important that future research
education on safe lifting techniques is provided. investigates the impact of other modes of resistance training
(eg, TheraBand and free weights) that can be used within
Limitations community and residential settings to provide options for
One limitation of the study is skin motion with respect to the exercise intervention and inform cardiac rehabilitation
transducer during performance of the 6 upper limb exercises, guidelines.
which may have contributed to error of measurement. Within This study found that cam machine-based upper limb
this study there was the potential for bias, because both ultra- resistance exercises, performed as early as 2 weeks following
sound assessors did not have the same degree of experience median sternotomy, in the absence of pain and discomfort
8 Sternotomy: SAFE-ARMS Study

are safe and feasible following cardiac surgery. Performance References

of 6 upper limb resistance exercises resulted in sternal
1. McGregor WE, Trumble DR, Magovern JA. Mechanical analysis
micromotion <2.0 mm, which is within the safe limits for
of midline sternotomy wound closure. J Thorac Cardiovasc Surg.
bone healing. Future research investigating sternal micro- 1999;117:1144–1150.
motion during upper limb exercises performed with varying 2. Robicsek F, Fokin A, Cook J, Bhatia D. Sternal instability after
modes of resistance training equipment in a larger cohort is midline sternotomy. Thorac Cardiovasc Surg. 2000;48:1–8.
warranted. 3. Deb S, Wijeysundera HC, Ko DT, Tsubota H, Hill S, Fremes SE.
Coronary artery bypass graft surgery vs percutaneous interven-
tions in coronary revascularization a systematic review. JAMA.
Author Contributions 2013;310:2086–2095.
4. LaPier TK. Functional status of patients during subacute recov-
Concept/idea/research design: J. Pengelly, A. Bryant, C. Royse, A. Royse, ery from coronary artery bypass surgery. Heart Lung. 2007;36:
G. Williams, D. El-Ansary 114–124.
Writing: J. Pengelly, A. Bryant, C. Royse, A. Royse, G. Williams, 5. LaPier TK, Wintz G, Holmes W et al. Analysis of activities
D. El-Ansary of daily living performance in patients recovering from coro-

Downloaded from https://academic.oup.com/ptj/article/102/7/pzac056/6585156 by guest on 15 May 2023

Data collection: J. Pengelly, D. El-Ansary nary artery bypass surgery. Phys Occup Ther Geriat. 2008;27:
Data analysis: J. Pengelly, S. Boggett, A. Bryant, D. El-Ansary 16–35.
Project management: J. Pengelly, A. Bryant, C. Royse, D. El-Ansary 6. Balachandran S, Lee A, Royse A, Denehy L, El-Ansary D. Upper
Providing facilities/equipment: A. Royse, D. El-Ansary limb exercise prescription following cardiac surgery via median
Providing institutional liaisons: D. El-Ansary sternotomy: a web survey. J Cardiopulm Rehabil Prev. 2014;34:
Consultation (including review of manuscript before submitting): 390–395.
A. Royse 7. Pengelly J, Pengelly M, Lin K-Y et al. Resistance training following
median sternotomy: a systematic review and meta-analysis. Heart
Lung Circ. 2019;28:1549–1559.
Acknowledgments 8. Brocki BC, Thorup CB, Andreasen JJ. Precautions related to mid-
The authors acknowledge Dr Sandy Clarke-Errey from the Statistical line sternotomy in cardiac surgery. A review of mechanical stress
Consulting Centre, University of Melbourne, for providing statistical factors leading to sternal complications. Eur J Cardiovasc Nurs.
analysis support; Kieser Australia for supporting the delivery of the 2010;9:77–84.
exercise intervention; and the patients and staff at the Royal Melbourne 9. Balachandran S, Denehy L, Lee A, Royse C, Royse A, El-Ansary
and Melbourne Private Hospitals. D. Motion at the sternal edges during upper limb and trunk tasks
in-vivo as measured by real-time ultrasound following cardiac
surgery: a three-month prospective, observational study. Heart
Lung Circ. 2019;28:1283–1291.
Funding 10. Katijjahbe MA, Granger CL, Denehy L et al. Standard restrictive
J. Pengelly is supported by a Research Training Program Stipend, sternal precautions and modified sternal precautions had similar
Swinburne University of Technology. effects in people after cardiac surgery via median sternotomy
(‘SMART’ Trial): a randomised trial. J Physiother. 2018;64:
97–106.
Ethics Approval 11. Adams J, Lotshaw A, Exum E et al. An alternative approach to
prescribing sternal precautions after median sternotomy, “Keep
Ethical approval was granted by the Melbourne Health Human Your Move in the Tube”. Proc (Bayl Univ Med Cent). 2016;29:
Research Ethics Committee (Application ID: 2017.266). 97–100.
12. El-Ansary D, LaPier TK, Adams J et al. An evidence-based perspec-
tive on movement and activity following median sternotomy. Phys
Clinical Trial Registration Ther. 2019;99:1587–1601.
13. Durstine JL, Painter P, Franklin BA, Morgan D, Pitetti KH, Roberts
This trial was prospectively registered with the Australian New Zealand
SO. Physical activity for the chronically ill and disabled. Sport Med
Clinical Trials Registry (ANZCTR) (ref. no.: ACTRN12617001430325p;
(Auckland, NZ). 2000;30:207–219.
universal trial no. [UTN] U1111-1203-2131).
14. Pengelly J, Royse C, Williams G et al. Effects of 12-week supervised
The full protocol in published in Trials (https://doi.org/10.1186/
early resistance training (SEcReT) vs aerobic-based rehabilita-
s13063-020-04558-x).
tion on cognitive recovery following cardiac surgery via median
Data from this study were presented at:
sternotomy: a pilot randomised controlled trial. Heart Lung Circ.
(1) The Australian Cardiovascular Health and Rehabilitation Asso- 2022;31:395–406.
ciation (ACRA) Virtual Annual Scientific Meeting; August 11, 15. Pengelly JMS, Royse AG, Bryant AL et al. Effects of supervised
2020. early resistance training versus standard care on cognitive recovery
(2) The American Association of Cardiovascular and Pulmonary following cardiac surgery via median sternotomy (the SEcReT
Rehabilitation (AACVPR) Annual Scientific Meeting; September study): protocol for a randomised controlled pilot study. Trials.
30, 2020; virtual. 2020;21:649.
(3) The Cardiac Society of Australia and New Zealand (CSANZ) 16. El-Ansary D, Waddington G, Adams R. Measurement of non-
Annual Scientific Meeting; December 8, 2020; virtual. physiological movement in sternal instability by ultrasound. Ann
(4) Research to Practice 2021, Exercise and Sports Science Australia; Thorac Surg. 2007;83:1513–1516.
May 7, 2021; virtual. 17. Deloach LJ, Higgins MS, Caplan AB, Stiff JL. The visual analog
scale in the immediate postoperative period: intrasubject variability
and correlation with a numeric scale. Anesth Analg. 1998;86:
102–106.
Disclosures 18. Koo TK, Li MY. A guideline of selecting and reporting intraclass
The authors completed the ICMJE Form for Disclosure of Potential correlation coefficients for reliability research. J Chiropr Med.
Conflicts of Interest and reported no conflicts of interest. 2016;15:155–163.
Pengelly et al 9

19. Adams J, Cline MJ, Hubbard M, McCullough T, Hartman J. A new 22. El-Ansary D, Waddington G, Adams R. Trunk
paradigm for post-cardiac event resistance exercise guidelines. Am stabilisation exercises reduce sternal separation in
J Cardiol. 2006;97:281–286. chronic sternal instability after cardiac surgery: a
20. Swanson LB, LaPier TK. Upper extremity forces generated during randomised cross-over trial. Aust J Physiother. 2007;53:
activities of daily living: implications for patients following ster- 255–260.
notomy. J Acute Care Phys Ther. 2014;5:70–76. 23. Ge W, Hians B, Sfara A. Noncontact measurement of the defor-
21. Fedak PWM, Kolb E, Borsato G et al. Kryptonite bone cement pre- mation of sternal skin during shoulder movements and upper
vents pathologic sternal displacement. Ann Thorac Surg. 2010;90: extremity activities restricted by sternal precautions. Phys Ther.
979–985. 2018;978:911–917.

Downloaded from https://academic.oup.com/ptj/article/102/7/pzac056/6585156 by guest on 15 May 2023