Design

A novel, anonymous, online survey was designed to explore current practice amongst physiotherapists treating patients following UAS in general surgical wards in Australian hospitals. Closed questions were predominately used incorporating a ranking system and matrix scale; open questions and free text boxes were also included. The survey was comprised of seven (7) sections, with a total of fifty-five (55) questions aiming to investigate the assessment tools and interventions commonly used by physiotherapists in the UAS cohort.

For the purpose of this study PPC was defined as “an identifiable disease or dysfunction that is clinically relevant and adversely affects the clinical course” [1]. Likewise, risk factors that prompt PPCs in patients post-UAS were classified as per those of Souza Possa et al. [4] including advanced age, smoking history and impacts of surgery. Early mobilisation tasks were examined, which included walking away from the bedside (greater than five metres) in conjunction with upright positions, sitting out of bed and stairs [1, 10]. Definitions of frequently used terms included in the matrix styled questions were provided and designed to guide participants to answer accurately (see Additional file 1). Other definitions of commonly used terms were also supplied to participants.

Ethics approval

The study was approved by the Human Research Ethics Committee (HREC) at The University of Notre Dame Australia (015151F) and by relevant local institutional HRECs.

Piloting

Prior to use, four independent individuals piloted the survey for readability and face validity; three were experienced cardiorespiratory physiotherapists with a particular interest and knowledge regarding UAS, the fourth was a health professional in a different field of work. Piloting identified any unanticipated problems and ambiguity within the instructions and questions and recognised time commitments required to complete the survey, allowing modification prior to dissemination. Minor amendments only were identified requiring adjustment for enhancing clarity on a couple of questions, plus some formatting adjustments of the online tool to enhance presentation / readability, and the online survey was finalised for distribution in February 2016.

Participants

Participants were included if they were qualified physiotherapists treating patients following UAS in an Australian hospital.

Recruitment

All Australian hospitals that conducted general surgery were identified via publicly accessible websites. Hospitals were contacted to establish if UAS was performed and whether the facility provided a physiotherapy service to patients undergoing UAS.

Hospital contact details were distributed between researchers by a random sequence generator in order to avoid bias. Phone calls were directed to the Physiotherapy Head of Department who provided further contact details and/or email addresses of physiotherapists. An outline of the study’s objectives was discussed during the phone calls, emphasising the necessity for physiotherapists treating patients undergoing UAS to be involved. Participants were encouraged to forward on the email to other relevant clinicians, increasing response rate via the snowball effect.

Sample size

As this study was descriptive and did not test any hypotheses, no sample size calculations were undertaken. A purposeful sample of convenience was used. All hospitals performing UAS were targeted resulting in a total population of 189 Australian hospitals being contacted. Not all facilities performed UAS or had physiotherapists treating these patients, whilst others did not want to provide contact details due to security and/or confidentiality reasons. As a result, contact details of 178 physiotherapists were retrieved.

Distribution

An invitation to participate with the anonymous survey was distributed by email in February 2016 using the online survey tool Qualtrics (http://www.qualtrics.com). Voluntarily opening and completing the survey implied consent.

Participants were given up to 3 months to complete the survey; follow up emails were sent to participants prompting completion of the survey at four, 6 and 8 weeks after the initial email was sent, aiming to optimise response rate.

Data analysis

Data was anonymously collected via the Qualtrics Q-Lite Package then exported into Microsoft Excel (Version 1.23.1 for Mac) for analysis. Categorical data were expressed in terms of count, frequency and proportions, primarily reporting percentages and means, specifically clarifying the total responses (n). Means were used to decipher the matrix styled questions. A 5-point Likert scale was used, aligning with “never” to “always”. For ranking styled questions, respondents were limited to three responses, allowing greater control of bias. Quantitative content analysis was used to determine patterns of participant responses to open ended questions, these were organised using a category matrix describing post-operative treatment goals [11]. Frequency counts were undertaken and expressed as numbers and percentages.

Demographics

Screening patients for pre-existing and post-operative risk factors prior to commencing treatment

Screening of patients prior to UAS by physiotherapists was not performed routinely; 51% respondents (n = 29/57) reporting this was “never” part of their practice, with a further 37% (n = 21/57) responding that they “rarely” screened / assessed patients prior to their UAS.

From a list provided to them, physiotherapists were asked to specify the frequently used parameters they utilise to assist them in identifying pre-existing risk factors that put patients at greater risk of PPC development. ‘Advanced age’ and ‘smoking history’ were both recognised by 98% (n = 54/55) of respondents as pre-existing risk factors; 94% (n = 52/55) of respondents noted that ‘pre-morbid respiratory conditions’ are “always” a risk factor, with 91% (n = 50/55) agreeing that ‘pre-existing heart conditions’ were also a factor to acknowledge. Screening tools “never” used include ‘neurological conditions’ (53%, n = 29/55) and an ‘American Society of Anesthesiology (ASA) score greater than 2’ (42%, n = 23/55).

Post-operatively, physiotherapists commonly use ‘chest x-ray’ (81%, n = 21/26), ‘auscultation’ (77%, n = 20/26) and ‘decreased saturation of oxygen (SpO₂)’ (77%, n = 20/26) as parameters to screen for high priority patients. ‘Surgical duration’ was rated by 58% (n = 32/55) as “sometimes” and “often” by 12% (n = 12/55) as being used to identify a patient at risk of PPCs. However, respondents did not rate ‘sputum classification’ (46%, n = 12/26) or ‘high temperature’ (50%, n = 13/26) as commonly used post-operative screening tools.

Treatment

Physiotherapists were asked to comment on their primary goals and/or foci for their management of patients over consecutive days post-UAS. Recurring statements indicated the majority of patients were not seen on day zero (day of surgery). On average, 94% (n = 51/54) of physiotherapists in general surgical wards treated their UAS patients once daily, with 93% (n = 51/55) initiating intervention on day one post-operatively.

Outcome measures

Barriers to treatment

Screening patients for pre-existing and post-operative risk factors prior to commencing treatment

Respiratory intervention

Results from this survey suggest that physiotherapists are currently implementing respiratory interventions into their practice; more than half combining chest treatments and mobilisation as their standard practice. This is despite recent evidence supporting the use of mobilisation as a standalone treatment, concluding that the addition of DBEx and coughing provides no additional benefit [9].

The majority of respondents specified an aim to perform routine chest treatment on day one, with results indicating that physiotherapists universally prescribed DBEx and supported coughs. It is evident that positioning and TEEs are also favourable interventions, with over half of clinicians consistently implementing them. These results are reflective of Hanekom et al. [1] who recommended that respiratory interventions are warranted for patients post-UAS. These recommendations are purely based on clinical experience, as the current literature remains somewhat inconclusive. A further study found adherence to mobilisation and chest therapy was effective at reducing the incidence of atelectasis to 0% [4], but continued research is necessary to validate this claim.

Despite conflicting evidence, clinical experience may be the primary driver behind why physiotherapists continue to use chest treatment as standard practice and not as per required. Although clinical experience is not necessarily unreliable, it needs to be acknowledged as a potential factor in resistance to change and should be integrated with evidence from high quality studies to promote best practice for patients undergoing UAS.

Further research to clarify the role of standard respiratory interventions and translation of evidence-based practice within UAS has the potential to encourage physiotherapists to agree and consistently implement interventions that are validated and most beneficial to this patient population, whilst best utilising valuable physiotherapy resources. That being said, there was no indication throughout this study that physiotherapists were using respiratory techniques as a standalone treatment, as mobilisation was universally accepted as the optimal choice of treatment in this patient population.

Mobilisation

Hanekom et al. [1] reported early mobilisation to be a beneficial intervention for patients following UAS. This is further validated by Silva and colleagues [9] emphasising the benefits of early mobilisation away from the bedside when performed at sufficient intensities, whilst Haines et al. [10] established that delaying early mobilisation caused an increase in PPCs.

This study demonstrated a positive link between a majority of milestones clinicians expected patients to achieve post-operatively and the physiotherapy treatment actually delivered over consecutive days. Following day one, physiotherapists indicated that when their patients were medically stable, providing no limitations to physiotherapy management, they mobilised their patients away from the bedside. This is in line with physiotherapists’ expectations and primary focus that from day one onwards, all patients should be mobilising away from the bedside. These findings are similar to the pre-existing literature of Silva et al. [9] and Haines et al. [10] concluding that the implementation of mobilisation alone provided an adequate reduction in PPC rates. This suggests recent literature is being translated into current practice as Australian physiotherapists demonstrated an awareness of mobilising away from the bedside as an effective treatment post-operatively.

Stair climbing was not necessarily being prescribed as an intervention despite physiotherapists expecting patients to achieve it as a milestone. Respondents also indicated that cycle pedals are an uncommon intervention post-operatively despite Bhatt and colleagues [6] determining that early aerobic exercise through the use of cycle pedals halved the rate of respiratory infection and length of stay. This was, however, a small single-centre study that needs to be validated prior to translation into standard physiotherapy practice in upper abdominal surgery patients.

Overall, this study’s results suggest that physiotherapists are implementing early mobilisation and that it reflects recent literature within this patient population. Despite all physiotherapists identifying early mobilisation as the primary focus of treatment, a small percentage of physiotherapists acknowledged that it was not implemented on every occasion, suggesting that barriers to ideal treatment exist.

Barriers to treatment

Findings of this survey indicate that a variety of patient-dependent factors limit the commencement of physiotherapy treatment post-operatively. ‘Pain’ was the most prominent barrier reported, followed by ‘fatigue’ and ‘patient readiness’, all having the capacity to reduce mobility and hence increasing the risk of a PPC. Similarly, ‘non-compliance’ was an evident barrier to treatment despite not previously being recognised as a post-operative risk factor for PPC development and delayed mobility.

The barriers identified by physiotherapists in this study are reflective of those previously reported by Browning et al. [16], in particular ‘availability of staff’ and ‘assistance to mobilise’ were both found to affect the amount of ‘uptime’ patients receive post-operatively following UAS. ‘Patient condition’ and ‘patient compliance’ were also reported as factors impacting the commencement and frequency of treatment, especially mobility, in this study.

Strong, validated evidence could give physiotherapists the opportunity to become more autonomous in the prescription of interventions post-operatively, assisting them to overcome external barriers such as ‘physician instruction’. There could be additional value in finding ways to support knowledge translation beyond physiotherapy cohorts and outwards to the wider field of the multi-disciplinary team to enhance physiotherapy management of patients post-UAS.

Barriers to treatment was not a focus of this study, therefore as a consequence of these incidental findings, it is unclear as to whether barriers such as pain and fatigue limit the efficiency and desired outcome of physiotherapy interventions or if it prevents the commencement of treatment completely. These findings provide avenue for further investigation into the impact of these barriers on commencing treatment and the strategies physiotherapists use to overcome them, creating the foundation of future studies to discover ways to facilitate treatment.

Limitations

Despite piloting, the length of the survey was the primary limitation of this study, with only two-thirds of those commencing survey completing (38/57). Not all questions were universally answered, with some respondents commenting that various questions were not applicable and/or repetitive, despite such issues not being apparent with piloting. Additionally, respondents may have perceived the questions differently to what was intended, again despite face validity being a focus of the piloting. Not all question responses were mandated in order to proceed through the survey, possibly accounting for varied response rates per question; providing opportunities for ‘not applicable’ answer options may be appropriate for future projects. The survey was anonymous potentially allowing multiple people to contribute from one facility. Likewise, the rotational nature of physiotherapy jobs may have hindered the response rate. Nevertheless, reminder emails worked to increase response rates to an adequate number with desirable representation Australia wide. Useful information was obtained throughout the survey making it a reflective summary of current practice in the UAS population.

Recommendations for future work

Continued research is necessary to determine whether the addition of respiratory interventions to early mobilisation confers any additional benefit to mobilisation alone in preventing PPCs. Further discussion is necessary to establish whether formalising an agreed minimal dataset of core screening tools could be a potential solution for prioritising resources. Also, further investigation into the barriers to treatment need to be completed. This has the potential to reduce PPC rates, improve patient-related outcomes and encourage the appropriate use of physiotherapy resources.