Expert consensus on the minimum clinical standards of practice for Nigerian physiotherapists working in intensive care units: A modified Delphi study

Background Management of patients in intensive care units (ICUs) needs staff with a recommended level of expertise and experience owing to the life-threatening nature of illnesses, injuries and complications that these patients present with. There are no specific guidelines governing physiotherapy practice in ICUs in Nigeria. Hence, there is a need to have expert consensus on the minimum clinical standard of practice for physiotherapists working in ICUs as a first step to proposing/developing guidelines in the future. Objectives To assess the expert consensus on the minimum clinical standard of practice for physiotherapists working in ICUs in Nigeria. Methods Physiotherapists with working experience in Nigerian ICUs were purposively recruited into the present study using a modified Delphi technique. A questionnaire comprising 222 question items on the role of physiotherapy in critical care was adopted and administered to the participants over three rounds of Delphi procedure (online). Participants checked either ‘essential’, ‘not essential’ or ‘unsure’ for each question item. For each question item to be considered ‘essential’ or ‘not essential’, a consensus agreement ≥70% had to be met. Questions without consensus were further modified by providing definition or clarification and presented in subsequent rounds. Data were analysed descriptively. Results We recruited 26 expert physiotherapists who consented to the study and completed the first round of the study. The majority of the physiotherapists (n=24) remained in the study after the third round. A total of 178 question items were adjudged to be ‘essential’ after the first round, and a further 15 and three additional items were subsequently adjudged to be as ‘essential’ after modifying the outstanding question items during the second and third rounds, respectively. No consensus was reached for 24 items. None of the question items were ranked as ‘not essential’ after all the rounds. Conclusion Expert consensus was achieved for a substantial number of question items regarding knowledge and skills for assessment, condition and treatment items of the questionnaire by experienced critical care physiotherapists in Nigeria.

Intensive care units (ICUs) are specially staffed and equipped hospital wards for management of patients with life-threatening illnesses, injuries or complications. However, the range to which different hospitals provide services to critically ill patients depends on the skills, expertise, facilities and clinical specialties available in the hospitals. [1] Physiotherapy is one of the fundamental interventions administered to patients in ICU. [2] The major goals of physiotherapy in the ICU include maintaining/restoring the general patient's functional capacity, and restoring respiratory and physical independence. Physiotherapy also helps to decrease the risks associated with stay in the ICU such as acquired muscle weakness, physical deconditioning and poor quality of life. [3] Moreover, the positive impact of physiotherapy in the management of patients whose conditions require critical care is well documented [4,5] and noted to improve survival rates. [6] There is moderate-to-strong evidence to support the role of physiotherapy for managing critically ill patients. [7] The ICUs in resource-restricted settings have limited infrastructure, materials and human resources. [8] In the UK, just as in other developed countries, physiotherapy is provided for 24 hours/day and 7 days a week (including on-call and public holidays) for patients in the ICUs. [7] In Nigeria, ICU patients are managed by physiotherapists every day of the week, but no evidence exists to support whether this is actually instituted in the ICUs standard of practice or guidelines. Consequently, some fresh graduate physiotherapists may start their first on-call service with less or no previous hands-on experience or training in managing critically ill patients in some ICUs. [9] Some centres compensate for this inadequacy by organising in-house/local critical care programmes/ICU workshops or mentoring to solve the problems of novice physiotherapists working in the ICU. Nevertheless, a bachelor's degree remains the least requirement for physiotherapists to work in ICUs in Nigeria. In short, the quality of care provided in the ICU largely depends on the skills of the attending physiotherapist.
Previous modified Delphi studies by Skinner et al. [10] in Australia, Twose et al. [11] in the UK and Takahashi et al [12] in Japan identified 132, 107 and 199, respectively, question items that are considered 'essential' for physiotherapists working in critical care units, and form a 222-item questionnaire developed by Skinner et al. [10] A similar study using the same questionnaire is necessary in Nigeria to identify and possibly suggest ways to standardise the competencies of physiotherapy practice in ICUs. As no standards of practice currently exist for the training of physiotherapists working in critical care in Nigeria and adopting an existing questionnaire has precedence in Delphi methodology. [13] Using the questionnaire developed by Skinner et al. [10] offered us the opportunity to contextualise our findings using a global point of view. Moreover, ICU patients require the best care possible, irrespective of the setting. The absence of national treatment guidelines indicates that most ICU interventions in resource-limited settings are often based upon treatment guidelines adopted from developed countries or international stakeholders.
The development of critical care in resource-poor settings relies on service improvements, including leveraging human resources through training, a focus on sustainable technology, continuous analyses of cost effectiveness and sharing of context-specific best practices. [14] Therefore, the present study is designed to explore the consensus of experienced physiotherapists regarding the minimum clinical standards of practice that physiotherapists working in critical care in our environment should possess. Moreover, the findings from this study could help in focusing future treatment guidelines, postgraduate education and ICU-related training of critical care physiotherapists in Nigeria and other similar countries.

Method
The present study followed a guide for Conducting and REporting of DElphi Studies (CREDES). [15] Ethics Ethical approval was sought and obtained from the Research and Ethical Committee of the College of Health Sciences, Bayero University, Kano (ref. no. NHREC/06/12/19/22). The ethical protocols of the Declaration of Helsinki including the ethical principles of informed consent, privacy and confidentiality of data provided were followed in the Delphi rounds.

Design
A modified Delphi technique was used to seek consensus on the minimum standards of practice for physiotherapists working in ICUs in Nigeria. Delphi techniques using online surveys are easily accessible and were the most appropriate for this study because they allow participants from across Nigeria to take part in the study from start to finish. [16,17] In addition, the Delphi technique allows for formation of consensus or exploration of a field beyond existing knowledge. It can be adapted to the particular requirements of the research question, and it takes the form of open and exploratory questions to standardised confirmatory approaches. [15] In the present study, the basic Delphi technique was modified to allow for question clarification, addition of new items and eventual agreement of the items by participants. Furthermore, in the second and third rounds of the Delphi study, additional definitions with examples (where possible) were provided for some terminologies where consensus was not reached in the first round to further enhance understanding of the terms.

Respondents
The prospective participants for this study were recruited using purposive sampling, alongside a snowballing method via advertisement through the official communication channels of the physiotherapy practice regulatory board and professional associations in Nigeria: namely (i) Medical Rehabilitation Therapist Registration Board of Nigeria (MRTB); (ii) Nigeria Society of Physiotherapy; and (iii) Association of Clinical and Academic Physiotherapists of Nigeria. Participants' recruitment was also made via social media platforms, particularly WhatsApp and Facebook groups of professional associations (e.g. Nigerian physiotherapists). Eligibility criteria were only communicated to participants after they had accepted an invitation to participate in the present study.
Sample size was not fixed so that we would be able to recruit as many participants as possible who met the eligibility criteria. The eligibility criteria for participation included the following: (i) physiotherapists with a minimum of 5 years' working experience, three of which must be in a senior role within the critical care setting; (ii) physiotherapists involved in the supervision or teaching of physiotherapy staff working on-call or completing emergency duty; and (iii) academic physiotherapy staff involved in the provision of entry-level cardio-respiratory physiotherapy with at least two articles published in the area of critical care.

Questionnaire
The questionnaire utilised in the present study was adopted from Skinner et al. [10] The questionnaire was structured and planned content-wise to be as extensive as possible across the physiotherapy role in critical care. [10] We did not pilot the questionnaire because it had been previously used in multiple studies [11] and we also aimed to contextualise the opinions of experienced physiotherapists in comparison with those in previous studies. Twose et al. [11] also adopted and used the same questionnaire without piloting. The questionnaire further highlighted that the motive of the present study was to determine the minimum standard of clinical practice that should be expected from physiotherapists to qualify them to work autonomously and safely with patients in critical care settings. The questionnaire consisted of 222 question items. For each question item, participants were asked to either check 'essential' , 'not essential' , or 'unsure' option. Respondents were also asked to submit additional items that were not previously included if they thought it necessary and essential for inclusion in the first round.

Procedure
Three rounds of questionnaire administration were sent to the study participants between 10 August 2020 and 2 October 2020, with each round lasting an average of 2 weeks. Delbecq et al. [18] recommended that 2 weeks is enough for Delphi participants to attempt each round. The study questionnaires were administered using Google forms via WhatsApp or email. The participants' information sheets were included in the invitation message. Reminders (phone calls, SMS and WhatsApp messages) were also sent to non-responders 7 days and a day before the deadline. Participants were also assured of the anonymity of all information provided so that they would not be afraid to admit their knowledge or lack of knowledge on any question item. Moreover, the participants were not asked to provide their names or any identifiers that might link them to any information provided. Demographic variables were only collected during round one, and were analysed separately by a blinded author (ASD). On completion of each round, participants were sent a personalised message thanking them and requesting for their co-operation in the subsequent rounds.

Data analysis
Data were analysed using descriptive statistics. Since demographic data were collected in the first round only, the data were summarised as a mean with standard deviation (SD), frequencies and percentages using Microsoft Excel. For each question item, a consensus to determine if such a question item was either 'essential' or 'not essential' was based on a consensus agreement of ≥70%. A study by the original developers of the instrument recommended a threshold of 70%. [10] Therefore, this threshold was used in the present study to allow for comparison of the study findings later. The percentage of 'unsure' responses was also taken into consideration to determine whether consensus was reached. Specifically, consensus for each item was calculated by subtracting the number of 'unsure' responses from the percentage of 'essential' responses. For example, if an item ranked 74% as 'essential' and the percentage of 'unsure' responses was 9%, then the percentage of 'unsure' responses was removed from that of 'essential' . For any item to be considered as 'essential' or 'not essential' , consensus had to be >70% in any of the three rounds.

Results
A panel of 47 experts (experienced physiotherapists working in Nigerian ICUs) responded to our invitation to participate in the Delphi study, and 32 potential participants met the inclusion criteria and were invited to participate in the first round of the present study. However, 81.3% (n=26) of the participants completed the consent forms and participated in the first round of the present study. Three-quarters of the participants (75%; n=24) remained by the end of the third round ( Fig. 1).
The participants comprised physiotherapy clinicians (n=24) and academics (n=2). The participants were drawn from all the six geopolitical regions of Nigeria (North East (n=4), North Central (n=5), North West (n=9), South East (n=4), South-South (n=1) and South West (n=3)). The majority of the participants (61.5%) had 10 years' or more ICU work-related experience. A substantial number of them (n=19) were staff of federal tertiary hospitals ( Table 1).
The first round of the study consisted of 222 items, of which 178 were considered as 'essential' by the respondents. One additional item was suggested by one participant; however, it was not added in the second round because it was a duplicate of an existing question item. None of the question items qualified to be ranked as 'not essential' in the first round, and consensus was not reached on 17 items on account of 'unsure' responses. In the end, the remaining 44 items were presented again in the second round.
Following the second round, 17 additional items were ranked as 'essential' and consensus was reached for 27 items. Still, no item was ranked as 'not essential' , and no additional item was suggested by the participants in the second round. The remaining 27 question items were further presented in the third round. Consensus was only reached for 3 question items that were categorised as 'essential' . Consequently, consensus was not reached for the remaining 24 items. Just like in rounds 1 and 2, no question item was ranked as 'not essential' in the third round.
Participants contacted to participate in the study, N=47 Excluded/dropped, n=21 Not in meeting study requirement, n=12 Not practising in Nigeria, n=1 Later contacted, no response, n=5 Declined consent, n=3 Agreed and eligible. Participated in round 1, n=26 Did not respond, n=1 Participated in round 2, n=25 Did not respond, n=1

Fig. 1. Flowchart of the recruitment and Delphi process.
Overall, the participants reached consensus on 197 items as 'essential' after three rounds. While no items were considered 'not essential' , no consensus was reached for 24 items. Detailed breakdown of the question items that reached consensus following the Delphi rounds are presented in Tables 2, 3 and 4. The question items that did not reach consensus after the third round are presented in Table 5.

Discussion
The present study aimed to assess the consensus of experienced physiotherapists working in ICUs in Nigeria with the view of determining a consensus for minimum standards of clinical practice. The present study is important in our environment because of the growing concerns about the variability of skills, level of qualifications, postgraduate experiences and clinical practice acumen of physiotherapists treating critically ill patients. Earlier studies highlighted varying standards of education, changing the on-call services, reduction in workforce capacity and irregularity in staff entry level. [19,20] We considered physiotherapists with ≥5 years of work experience in critical care settings or emergency on-call services to be experts and those with <5 years of work experience as novices based on information from previous studies. [21,22] We utilised several means including snowballing sampling technique to reach as many physiotherapists as possible to participate in the present study. We found that 197 items of knowledge and skills were judged to be 'essential' as a minimum standard of clinical practice in critical care settings following three rounds of online Delphi survey. Consensus was not reached to classify any question item from the questionnaire as 'not essential' for clinical practice. The results of the present study are similar to those of Takahashi et al. [12] who reported consensus for 199 items from a partially modified version of the Skinner et al. [10] questionnaire. Nevertheless, our findings are different from those in Australia [10] and the UK, [11] where consensus was reached for fewer items as 'essential' and 'not essential' . We think that the experts in our study had comparatively lower experience and exposure in ICUs as they were more likely to have received little or no training, and they were practicing in a resourcelimited setting. It is important to note that the scope of practice of physiotherapists in the ICU differ across countries. [21,23] We had a good response rate, considering the low number of physiotherapists with experience in ICU clinical practice in Nigeria. Our study also recorded very low dropout rates between rounds. Twose et al. [11] reported significantly higher dropout rates in their study (from 80% in round 1 to 65% in the final round). However, this is still within the accepted range for Delphi studies. [24] It must also be stated that several reminder messages were sent to non-responders via phone calls in addition to emphasising the need to observe the deadline. [25] In addition, sending the questionnaire via WhatsApp was helpful as most participants had smartphones. Overall, the number of participants in our study was small compared with other studies, but we considered them as having the best critical care experience in view of the strict inclusion criteria.
Participants in the present study ranked more items as 'essential' , and no items as 'non-essential' , unlike other studies that recorded    Ventilator hyperinflation via an endotracheal tube or tracheostomy 46.2* 92 As a minimum standard, a physiotherapist can appropriately request/coordinate the following: Titration of inotropes to achieve physiotherapy goals 82.6 As a minimum standard, a physiotherapist is aware: Of key literature that guides evidence-based physiotherapy practice in critical care settings 96.2 As a minimum standard, a physiotherapist can accurately interpret readings from clinical monitoring including: Central venous pressure 100 As a minimum standard, a physiotherapist can accurately interpret findings from laboratory investigations including: Renal function tests e.g. urea and creatinine 100 Sputum cultures 96.  several items from the same questionnaire (Australia, [10] New Zealand, [10] Japan [12] and UK [11] ). Van Aswegen et al. [26] reported consensus was achieved on knowledge of normal integrated anatomy and physiology, knowledge of and skill to conduct a holistic assessment of an ICU patient, knowledge and skill of clinical reasoning, and knowledge of physiotherapy techniques by physiotherapists working in critical care units in South Africa.
Another reason for the high number of 'essential' question items reaching a definite consensus after the final round was because we modified the questions not reaching consensus by providing definitions and examples during the second and third rounds. Therefore, the additional consensus obtained for these items that were modified could mean that our study participants may not have been conversant with some of the question items as presented in the Skinner et al. [10] questionnaire. Specifically, most of the items not reaching consensus appear to require intensive training and high clinical skills. Therefore, it is not surprising that items on intubation/extubation of patients, interpreting measurements in ICUs such as haemofiltration, pulmonary arterial catheter measurements, C-reactive protein as well as performing cuff volume and/or pressure test on an endotracheal tube or tracheostomy, among others, did not reach consensus. The skills and knowledge that a physiotherapist needs to deliver services independently in critical care need to be streamlined. [27] Moreover, there is lack of standardised programmes for ICU physiotherapist education in our environment. [28] In the present Delphi study, the participants appeared to agree with the majority of the items on the questionnaire, with no tangible additional items suggested across the rounds. This may be because the requirements for practice are less strict in terms of scope and responsibilities of physiotherapists in Nigeria compared with many advanced countries. [29,30] We also observed that some question items such as the Feldenkrais technique, which is not common in our critical care settings, reached consensus after the third round, after alternative definitions and examples were provided. Furthermore, items like bronchial lavage, which is a diagnostic method of the lower respiratory system in which a bronchoscope is passed into the lungs with a measured amount of fluid introduced and then collected for examination, reached consensus in the first round without necessitating alternative definitions. This technique is typically outside the physiotherapist's scope of practice, not popular in our environment and requires high technological and technical expertise. Hence, we cannot at the moment explain why the participants must have checked 'essential' for this item. Nevertheless, it is important that we could still draw attention to the consensus opinion of participants of the present study using this kind of question items. Moreover, more physiotherapists working in the critical care setting are increasingly having the opportunity to acquire further knowledge and skills from online ICU rehabilitation workshops delivered by people from all parts of the world. Our study offers a good first step in approaching institutions with possible items that could form a curriculum for training of physiotherapists who will work in ICUs in Nigeria. The existing curriculum in undergraduate physiotherapy programmes in Nigeria does not provide for teaching and clinical practice experience in most of the items checked as 'essential' . Therefore, the use of graduate physiotherapists without additional specialised training in the ICU at postgraduate level may not be desirable.

Study limitations
The instrument from Skinner et al. [10] was not locally validated prior to administration in the present study. It was noted that the questionnaire was quite detailed, with many question items covering all aspects of critical care. Nevertheless, the study results are valid because we used a consensus to arrive at the items selected eventually. Participants were given a chance in the first round of the Delphi to provide additional inputs to the question items by means of open-ended questions requiring them to suggest additional questions. However, no further suggestions were made and this may mean that they were satisfied with the instrument or they had response exhaustion due to the many questions, as observed in previous Delphi studies. [31] In the present study, the questionnaire comprised many questions and required ~30 minutes to complete. Hence, participants were given up to 2 weeks to complete the questionnaire for each round. The critical care capacity, resources and manpower are relatively low and limited in Nigeria compared with developed countries, [32] so the number of physiotherapists who were available for recruitment was relatively small. Future studies should focus on testing the knowledge of physiotherapists working in the ICU contained in the 'essential items' reaching consensus with a view to ascertaining specific areas of need in this environment.

Conclusion
Expert consensus was achieved for a substantial number of questions on knowledge/skills of assessment, condition and treatment. These items could be considered as 'essential' minimum standards of clinical practice for physiotherapists working in ICUs, based on the opinion of experienced physiotherapists in Nigeria. The results of this present study need to be further validated by appropriate authorities to support development of training programmes and curricula for critical care physiotherapy specialisation in Nigeria with a view to reducing clinical practice variability and achieving acceptable quality in patient management.