Use of the Theoretical Domains Framework to Develop an Intervention to Improve Physical Therapist Management of the Risk of Falls After Discharge
- S. Thomas, PT, PhD, International Centre for Allied Health Evidence (iCAHE) and the Sansom Institute, University of South Australia, Adelaide, Australia, and Physiotherapy Department, Flinders Medical Centre, Adelaide, Australia.
- S. Mackintosh, PT, PhD, International Centre for Allied Health Evidence (iCAHE) and the Sansom Institute, University of South Australia, and School of Health Sciences, University of South Australia.
- Address all correspondence to Dr Thomas at: susie.thomas{at}health.sa.gov.au.
Abstract
Background and Purpose Older adults have an increased risk of falls after discharge from the hospital. Guidelines to manage this risk of falls are well documented but are not commonly implemented. The aim of this case report is to describe the novel approach of using the Theoretical Domains Framework (TDF) to develop an intervention to change the clinical behavior of physical therapists.
Case Description This project had 4 phases: identifying the evidence-practice gap, identifying barriers and enablers that needed to be addressed, identifying behavior change techniques to overcome the barriers, and determining outcome measures for evaluating behavior change.
Outcomes The evidence-practice gap was represented by the outcome that few patients who had undergone surgery for hip fracture were recognized as having a risk of falls or had a documented referral to a community agency for follow-up regarding the prevention of falls. Project aims aligned with best practice guidelines were established; 12 of the 14 TDF domains were considered to be relevant to behaviors in the project, and 6 behavior change strategies were implemented. Primary outcome measures included the proportion of patients who had documentation of the risk of falls and were referred for a comprehensive assessment of the risk of falls after discharge from the hospital.
Discussion A systematic approach involving the TDF was useful for designing a multifaceted intervention to improve physical therapist management of the risk of falls after discharge of patients from an acute care setting in South Australia, Australia. This framework enabled the identification of targeted intervention strategies that were likely to influence health care professional behavior. Early case note audit results indicated that positive changes were being made to reduce the evidence-practice gap.
Immediately after discharge from the hospital, the risk of falls and fall injuries for older adults is high.1,2 Up to 40% of patients fall at least once in the 6 months immediately after discharge, with 54% of these falls resulting in injury.2 Cohorts of patients with a high risk of falls include patients with a surgical repair of a hip fracture (53% of patients fall within 6 months)3 and patients with stroke (50% of patients fall within 6 months).4 Although there is a growing body of literature focusing on strategies to prevent falls in the hospital5 as well as extensive literature for community settings,6 there has not been such a strong focus on reducing the risk of falls during transitions between settings (eg, when patients transfer from the hospital back to the community or to a subsequent facility, such as a rehabilitation facility, other acute care hospital site, or a residential care facility).
Best practice guidelines for the prevention of falls and harm from falls in older people have been developed in Australia,7 and the National Safety and Quality Health Service Standards in Australia now address the prevention of falls and harm from falls.8 Both of these place emphasis on implementing prevention strategies to reduce the risk of falls after discharge from the hospital, including recommendations to identify risk early in a patient's admission, to take action to increase the proportion of at-risk patients undergoing a comprehensive assessment of the risk of falls, to refer patients at risk to appropriate services as part of the discharge process, and to educate patients and caregivers about this risk and strategies for the prevention of falls.
A study investigating physical therapist treatment of patients admitted to the hospital after a hip fracture indicated that management of the risk of falls was given little attention.9 This finding raised concerns about an evidence-practice gap because up to 90% of hip fractures occur as the result of a fall,10 and there is an increased risk of repeat falls within the first 6 months after surgery for a hip fracture.3
The purpose of this case report is to describe a systematic approach for improving the identification and treatment of patients at risk of falls after discharge from an acute care hospital setting in South Australia, Australia.
The Theoretical Domains Framework (TDF) was used to develop interventions aiming to change clinical practice behavior and improve the uptake of evidence into practice.11,12 The TDF targets behavior change in health care professionals and comprises 14 domains that encompass factors that are likely to influence health care professional behavior change: knowledge; skills; social/professional role and identity; beliefs about capabilities; optimism; beliefs about consequences; reinforcement; intentions; goals; memory, attention, and decision processes; environmental context and resources; social influences; emotion; and behavioral regulation (Tab. 1).12 These domains also can be mapped to a component within a “behavior change wheel” (BCW) on the basis of the concept that capability, opportunity, and motivation interact to generate behavior.12,13 From here, the link can be made to relevant behavior change strategies13 that are most likely to overcome the identified barriers and enhance the enablers of changes in practice. Behavior change strategies that are relevant, feasible, and acceptable in local settings can be chosen, and specific interventions can be devised.
Fourteen Domains of the Theoretical Domains Framework and 84 Associated Component Constructs12
Case Description
A 4-step method described by French et al14 (Appendix 1) was used to identify barriers and enablers for bridging the evidence-practice gap and to identify behavior change strategies most likely to improve the identification and treatment of patients at risk of falls after discharge from the hospital.
Identifying the Problem (Step 1)
Setting.
The setting was the orthopedic ward of a 588-bed acute care teaching hospital in South Australia. This ward admits, on average, 5 to 10 new patients with hip fracture per week. The ward is staffed by 2 full-time-equivalent physical therapists, but inclusive of weekend staff, up to 20 different physical therapists as well as 25 to 30 physical therapist students per year can assess and treat the patients on this ward.
Target groups.
The target groups were clinical staff responsible for the assessment and treatment of patients who had undergone surgery for hip fracture: the Southern Community Falls Prevention Team, a team of health care professionals based in the community and responsible for the coordination of services for the prevention of falls in older people who were living in southern metropolitan Adelaide, South Australia, Australia, and had been identified as having a risk of falls; executive managers in the hospital who had an interest in the management of falls and were responsible for resource allocation; and patients who had been admitted with a hip fracture as well as their partners/caregivers (consumers).
Data sources.
A mixed-methods approach involving both quantitative and qualitative data collection methods to identify current practice and potential barriers to practice change was used. This approach included focus groups with physical therapists responsible for the assessment and treatment of patients who were admitted to the site and who were at risk of falls after discharge; in-person interviews and regular meetings with key stakeholders; reviews of routinely collected hospital data; audits of case notes and in-person interviews with patients who had undergone surgery for hip fracture; and phone interviews with patients referred to the Southern Community Falls Prevention Team after discharge from the hospital. The information obtained was synthesized to allow mapping of current processes, and feedback from key stakeholders was sought to determine accuracy and rigor. The process map was then compared with best practice guidelines7 and Standard 10 (relevant to falls) of the National Safety and Quality Health Service Standards.8
Assessing the Problem (Step 2)
The TDF (Tab. 1) was applied retrospectively to the barriers and enablers that had been identified in step 1, in which a detailed gap analysis was undertaken. This analysis allowed TDF domains to be linked to each barrier and enabler and all aspects influencing clinical practice behavior to be identified (Tab. 2).
Barriers to Behavior Change, Relevant Theoretical Domains, Components of the COM-B System, Intervention Functions, and Behavior Change Strategiesa
Forming Possible Solutions (Step 3)
A BCW13 was used to identify potential intervention components that were most likely to overcome the modifiable barriers and enhance the enablers (Tab. 2). Factors such as feasibility, local relevance, and acceptability of the chosen interventions to the site were considered in the selection of behavior change strategies, along with relevant previously examined implementation interventions for acute care hospital settings.
A project governance committee comprising representatives from the physical therapy department at the hospital was established. The members were all directly involved with the proposed practice change, had relevant knowledge regarding the project design, or had a keen interest in promoting practice change. The committee was chaired by the chief investigator (S.T.) and performed tasks such as reviewing and providing feedback on identified barriers and enablers, identifying and testing implementation strategies for dissemination, and adapting education materials. The committee met once per month during the first 18 months of the project.
Outcome
Steps 1 through 3 ran for 11 months from March 2012 until February 2013.
Identifying the Problem (Step 1)
A case note audit9 revealed that only 10% of patients who had been admitted to the site for surgery for hip fracture were identified as having a risk of falls and that only 8% had a documented referral to a community agency for follow-up regarding the prevention of falls. Key approaches to reducing the risk of falls after discharge from the hospital were not being routinely implemented—in particular, a lack of identification of the risk of falls; a lack of assessment and management of the risk of falls; a lack of a rehabilitation approach that addressed fall risk factors; and a lack of follow-up to review and advance prescribed interventions. Therefore, the initial focus of the project was improving the identification and management of the risk of falls after discharge for patients who had been admitted with a fractured hip.
From subsequent process mapping (including interviews with staff and patients), it became apparent that the scope needed to be expanded to include all patients who were admitted to the site and who were at risk of falls after discharge. Therefore, the interventions needed to be tailored with regard to applicability and feasibility for all wards in the hospital involved in managing the risk of falls after discharge and not just the orthopedic ward. Two key health professional groups were identified as target groups for the practice change. The first group was physical therapists who were responsible for assessing and treating patients with an identified risk of falls, and the second was the Southern Community Falls Prevention Team. Finally, the project needed to focus on improving the communication and process links between acute and primary care settings—that is, improving the transition of care provided to patients at risk of falls after discharge from the site.
Step 1 ensured that project aims were clarified and aligned with both the Australian best practice guidelines for the prevention of falls and harm from falls in older people7 and the National Safety and Quality Health Service Standards.8
Three main issues were identified and became the focus for the project. First, patients who were at risk of falls after discharge from the hospital were not routinely being identified (Appendix 2). There was no defined process to identify patients who were potentially at risk. The fact that the determination of which patients needed follow-up in the community relied on physical therapist discretion may have accounted for a large degree of variability. Second, documentation of high-quality clinical transfer of information relating to patients who were identified as having a risk of falls after discharge from the hospital was not occurring routinely or consistently (Appendix 2). Baseline data collection revealed that the only way in which community agencies could obtain information about patients to assist in the processes of triage and service matching was to access an electronic discharge summary. If this summary was absent or of poor quality, the likelihood of patients receiving appropriate management of the risk of falls after discharge from the hospital was reduced. Third, patients were not involved in the process of identifying their risk of falls or planning for the management of this risk (Appendix 3).
On the basis of these data, project aims were redefined and are summarized in Appendix 4.
Assessing the Problem (Step 2)
Table 2 shows the barriers and enablers that were identified in step 1 of the project. The identified theoretical domains with which barriers and enablers were most frequently associated were environmental context and resources, knowledge, and social/professional role and identity. Twelve of the 14 theoretical domains were considered to be relevant to behaviors in the project; intentions and behavioral regulation were the 2 domains that were not identified.
Forming Possible Solutions (Step 3)
A multifaceted intervention was designed to overcome the modifiable barriers and enhance the enablers while also considering what was likely to be feasible and relevant at the site and acceptable to the staff members who would be expected to implement the changes. The chosen behavior change strategies were presented and pilot tested with key stakeholders (physical therapists, Southern Community Falls Prevention Team, and consumers) and modified on the basis of feedback. The final behavior change strategies selected (Tab. 2) were education sessions for physical therapists about guideline recommendations and the consequences of failing to meet guideline recommendations; development of a “pathway” to guide the identification and management of the risk of falls; modification of an existing standardized initial assessment pro forma to prompt identification of the risk of falls and encourage documentation and appropriate action after identification; development of standardized processes for high-quality transfer of information (including paper and electronic options) to community service providers; and dissemination of the “Don't Fall For It. Falls Can Be Prevented!” booklet.15 This booklet provides easily understandable written information for consumers to take with them after discharge and assists staff in discussions about the risk of falls with patients.
Mechanisms of Ensuring Sustainability
While the multifaceted intervention was being implemented, several strategies were used to promote project sustainability. “Snapshot audits” of the intervention strategies were conducted over a 3-month period, and feedback about outcomes was given to staff. This approach enabled the identification and modification of processes that were difficult to establish, allowed positive behavior to be reinforced, and kept the project at the forefront of the minds of the physical therapists who were changing their practice.
A “fall committee” was established within the physical therapy department; this committee was responsible for governing ongoing audits and providing feedback to staff, modifying processes in line with environmental or organizational changes (eg, consideration of the move to an electronic medical records system), maintaining lines of communication with primary care settings, and establishing lines of reporting from the committee up to executive-level management at the site. The committee agreed on a name for the project—the “Healthy Hips Project”—to build a brand and make the project easily identifiable.
Intervention processes were incorporated into standardized operating procedures for the physical therapy department and made widely available; this step included building the processes into the department's orientation manual to ensure that new staff members were made aware of the processes as soon as possible.
Finally, an allocation of staff time was incorporated into the staff roster to ensure that a specific staff member was always allocated to overseeing the processes of identification and referral to community services on an ongoing basis and that this position was provided with “cover” during periods of leave.
Evaluating the Selected Intervention (Step 4)
Measures of change were identified, and the primary outcomes selected were the proportion of patients with documentation of the risk of falls and the proportion of patients referred for a comprehensive assessment of the risk of falls after discharge from the hospital. Secondary outcomes were the percentage of time that the transfer of information about the risk of falls from the primary site met a standardized set of criteria for transfer quality and a change in the awareness of consumers of their own risk of falls and the management strategies put in place to mitigate this risk. Through the use of a time series approach to data collection, outcomes were measured at 3 points in time both before and after implementation of the behavior change strategies. The overall success of the implemented multifaceted intervention and project outcomes will be reported in a separate article upon project completion.
Discussion
A systematic approach in which the TDF11,12 and the BCW12,13 were used to underpin a theory-informed multifaceted intervention to improve management of the risk of falls after discharge from an acute care setting was undertaken. The use of the TDF to inform behavior change is a relatively new approach that was recently examined to inform the design of implementation interventions.16–19 Previous studies focused primarily on changing the behavior of the medical profession (physicians/doctors); this approach has never been used to inform an intervention to improve management of the risk of falls in a retrospective manner. It is too early to determine whether the behavior change strategies have been successful; however, we have already been able to demonstrate that this approach was useful for designing a targeted intervention.
Although many resources have been directed at improving the quality and safety of health care by increasing the uptake and speed of knowledge translation,20,21 behavior change strategies continue to have variable effects.22 One of the benefits of using the TDF in our project was that the structured framework allowed us to make decisions in a systematic manner. By using the TDF with a 4-step method,14 we were not only directed from identifying targeted behaviors to designing behavior change strategies but also given confidence in the intervention choices that were made. This approach enabled an analysis of possible influences on the behaviors being displayed at the site and provided a mechanism for determining and prioritizing intervention choices.12 It was previously recommended that interventions should be multifaceted23 but should not deviate too far from current practice to increase the likelihood of uptake. The use of the TDF allowed the selection of behavior change strategies that addressed the domains that were most frequently identified, thus prioritizing strategies on the basis of the likelihood of success.
An advantage of applying the TDF in a retrospective manner was that it saved time. The 2-year project funding did not allow for a more thorough step 1 and step 2, in which interviews and focus groups could be structured around the TDF, as previously described in the literature.17,24 Although the use of the TDF has been recommended, this approach has also been reported to require considerable time and resources.14 A prospective approach ensures that all domains are addressed with the target groups, as members of the groups are prompted to consider whether their behaviors occur in a manner relevant to the domains, but such an approach was considered too time-consuming for this project.
Another advantage of a structured approach is that at project completion, it will be possible to systematically analyze, through retrospective review, what has worked and why. It will be possible to consider whether the chosen behavior change strategies actually addressed the influences on behavior appropriately and whether barriers or enablers that were present at the beginning of the project continued to exist as the environment or culture changed. A review of the literature on strategies for improving management of the risk of falls in acute care hospital settings revealed that many investigators failed to use any sort of framework or systematic approach in their project design.25–29 This situation made it impossible to replicate successful strategies described in the literature. This scenario is not uncommon; other systematic reviews revealed that many investigators in knowledge translation studies failed to use theoretical or modeling research to frame their research design.30
Some behavior change strategies that were identified as potentially appropriate at the site could not be implemented because of the environmental context. The data suggested that a “falls coordinator” role within the physical therapy department or even within allied health at the site would be beneficial for overseeing the ongoing success of processes not only for the risk of falls after discharge but also for consideration of the inpatient risk of falls. This role would require only a part-time position (0.2 full-time equivalent) to provide ongoing education and resources to staff. However, discussions with management were not successful in establishing this role; consequently, it was decided that this role will be built into existing positions. It is not clear whether this strategy will be successful because the role could end up being “lost” within the positions, and the engagement of staff members who take on the role may not be as extensive as that of someone specifically appointed to the role.
The overall success of the project will not be known until final data collection is complete. However, it is anticipated that the data will be unique because they address an intervention to improve management of the risk of falls across care settings, an area of practice that has not been considered. In the past, interventions focused on improving management of the risk of falls in acute care settings or once the patient has transitioned to another facility5 or back to the community.6 The transition itself has been neglected.
Limitations
The data from the project may not be relevant to larger-scale implementation interventions. Our approach was tried at 1 acute care setting and focused primarily on 1 health care professional group. Therefore, settings in which multidisciplinary teams need to be considered or multiple sites are involved may not be able to follow the approach described. Several other elements in the care pathway for management of the risk of falls were identified as deviating from best practice care; however, given project limitations (resources and time frames), only areas considered likely to achieve the greatest patient and hospital outcomes without deviating too far from current practice were chosen. Finally, final data collection is not yet complete. Therefore, although the data provide some insight into how useful the TDF was in the design and implementation of the project, it is not yet possible to comment on the contribution of the TDF to the overall success of the project.
Conclusion
A systematic approach involving the TDF and the BCW was useful in the design of an intervention to improve management of the risk of falls after discharge of patients from an acute care setting in South Australia. It not only provided a framework to assist in identifying barriers and enablers influencing behaviors at the site but also allowed these behaviors to be linked to potentially successful behavior change strategies. Including the opinions of consumers in the data collection phase was a vital source of information for driving behavior change strategies. The application of a theoretical framework will enable us, at project completion, to determine which strategies were successful in this particular setting and why.
Appendix 1.
Steps for Developing a Theory-Informed Implementation Intervention to Change Clinical Behavior14
Appendix 2.
Case Note Audit Data Relevant to Identification of the Risk of Falls and Quality of Clinical Transfer
Appendix 3.
Summary of Telephone and In-Person Consumer Interview Baseline Data
Appendix 4.
Redefined Project Aims After Step 1 of the Methodological Process
Footnotes
Both authors provided concept/idea/project design, writing, data analysis, fund procurement, facilities/equipment, and consultation (including review of manuscript before submission). Dr Thomas provided data collection, project management, patients, and institutional liaisons. The authors thank the Southern Community Falls Prevention Team for their support and encouragement of this project and the Physiotherapy Department TRIP Governance Committee for providing feedback and support regarding project design and interventions.
Ethics approval was obtained from the Southern Adelaide Clinical Human Research Ethics Committee (Research Application Number 229.12) and the University of South Australia Human Research Ethics Committee (Protocol Number 0000030397).
An oral presentation of this project was given at the South Australian Rehabilitation Research Forum; March 15, 2013; Glenelg, South Australia, Australia.
Dr Thomas is supported by a National Health and Medical Research Council (NHMRC) Translating Research Into Practice (TRIP) Fellowship (2012–2015).
- Received September 1, 2013.
- Accepted July 4, 2014.
- © 2014 American Physical Therapy Association