Feasibility of a multicomponent cognitive behavioral intervention for fear of falling after hip fracture: process evaluation of the FIT-HIP intervention

Many older adults who have sustained a hip fracture will go through an extensive and generally challenging process of rehabilitation [1, 2]. During this recovery process, a substantial number of patients will experience concerns about falling (once) again [3, 4]. This fear of falling (FoF), is defined as ‘a lasting concern about falling that leads to an individual avoiding activities that he/she remains capable of performing’. [5] Prevalence rates of up to 63% have been reported for FoF in inpatient geriatric rehabilitation after hip fracture [4]. As a consequence of the activity restriction associated with FoF, deterioration in physical functioning and a decline in social participation and quality of life can occur [3, 6]. FoF may even have more effect on functional recovery after fracture than pain and depression [7]. Hence, FoF appears to be an important risk factor for impaired recovery [3, 8, 9], which could possibly be addressed by treatment.

Patients with a recent hip fracture differ from the general population of community-dwelling older adults in that they experience a sudden impairment of their gait function and consequently become dependent in (basic) activities of daily living [2]. In the Netherlands, approximately half of all older patients with a hip fracture follow an inpatient multidisciplinary rehabilitation program after surgical repair of the fracture. These ‘geriatric rehabilitation’ services are specialized in the medical care for frail older adults [10]. Therapy is aimed at optimizing the patient’s physical condition and restoring (gait) function [11]. Physical therapy focuses on training balance and muscle strength, and practicing activities of daily living [12]. At present there are no treatment programs aimed specifically at reducing FoF after a recent hip fracture. However, for community-dwelling older adults, various evidence-based interventions have been developed to reduce FoF [13‐18]. Particularly the treatment programs that combine exercise with cognitive behavioral approaches have been found to be effective in reducing FoF [16‐18]. In the Netherlands, two of these evidence-based programs using cognitive behavioral approaches have been nationally implemented (based on ‘A Matter of Balance’) [15, 19, 20]. However, in their current format (community- or home-based), these established programs are not suitable for the therapeutic setting of inpatient geriatric rehabilitation. The cognitive behavioral approaches used in these programs were therefore adjusted to an individualized treatment program that fits the (physio) therapeutic setting within rehabilitation services. This Fear of falling InTervention in HIP fracture geriatric rehabilitation (FIT-HIP intervention) was designed to reduce FoF and consequently to improve functional outcome in inpatient geriatric rehabilitation after hip fracture [21]. However, a recent cluster randomized controlled trial evaluating the effects of the FIT-HIP intervention showed the program was not effective in reducing FoF or improving functional outcome after hip fracture [12].

The aim of this process evaluation therefore is to gain insight into factors that may have influenced the effectiveness of the intervention. Subsequently, findings from this study can provide insight into opportunities to improve both the intervention itself and its implementation in clinical practice. In this study we assessed the feasibility of the FIT-HIP intervention in clinical practice based on the following aspects of the intervention process: 1] recruitment and reach; 2] performance according to protocol (dose delivered and fidelity); 3] adherence (dose received exposure); and 4] opinion on the intervention provided by patients and facilitators (dose received satisfaction and context). These items are based on the framework of Saunders and colleagues [22, 23]. This model for process evaluation is frequently used within health care innovations and is based on the widely acknowledged principles of Steckler et al. (2002) [24].

This study assessed the feasibility of a multicomponent cognitive behavioral intervention for FoF after hip fracture, integrated in usual care in inpatient rehabilitation. To a fair degree the intervention was performed according to protocol, but cognitive restructuring and the telephonic booster were not provided to all patients. Patients rated the intervention positively and half of them reported that the intervention was (very) helpful in reducing FoF. Most facilitators were positive about the intervention and considered it feasible. However, this study also identified barriers that may have affected this feasibility, and these should be addressed to improve the intervention. Two important barriers were the limited level of FoF after enrollment (possibly related to timing of the intervention), and the fact that physiotherapists, having limited experience with such approaches, perceived cognitive restructuring as challenging.

A considerable body of evidence demonstrates that programs based on cognitive behavioral approaches (preferably combined with physical exercise) are effective to reduce FoF in older adults with fall risk [16‐18, 27]. However, despite the benefit perceived by patients, the FIT-HIP intervention was not effective in reducing FoF when compared to usual care [12]. It is therefore crucial to reflect on the intervention process, in particular cognitive restructuring as this was not administered to all patients and was considered the most challenging element for facilitators. First, the dose of cognitive restructuring within the intervention does not differ significantly from other programs [14, 28, 29], and this does not explain the absence of effect. However, in our study fewer patients received cognitive restructuring according to protocol (72.2% in the FIT-HIP study versus 83.4% in the home-based program for FoF in community-dwelling older adults) [30]. This may have contributed to the lack of effectiveness.

The fact that cognitive restructuring is perceived as challenging does not by definition imply it is not feasible in practice or not suitable for frail older adults. Literature on nurse-led programs for FoF in community-dwelling older adults confirms the finding that cognitive restructuring can be challenging for facilitators and participants, yet these programs - despite the perceived difficulties - proved to be effective [14, 15, 20]. Regarding the appropriateness of cognitive restructuring for frail older adults, facilitators in our study acknowledged that even in cases of mild cognitive impairment, this approach still had potential short-term effects (during the therapy session), enhancing the rehabilitation process.

In a broader perspective, we could question whether it is appropriate for physiotherapists to apply cognitive restructuring. In the past years, interest in incorporating a biopsychosocial approach to physiotherapy practice to enhance the rehabilitation process has increased [31]. Research illustrates that overall, physiotherapists have positive attitudes and beliefs regarding psychosocial interventions [31]. Common barriers to implementation of psychosocial interventions in clinical practice include lack of knowledge, time constraints (including the perceived need to prioritize physical care) and the scope of practice (role clarity and public perceptions of traditional physiotherapist role) [31, 32]. These factors were also identified in our study, but rather than the lack of knowledge, the facilitators mentioned a desire for more experience. The current literature concerning psychosocial interventions with physiotherapists as facilitators recommends that, in order to ensure treatment fidelity, psychologists should provide comprehensive training and mentoring to the physiotherapists, including performance feedback [32, 33]. Effectiveness of such an approach is supported by a recent study that showed positive effects of a physiotherapist-led in-home intervention to reduce FoF and activity avoidance, including cognitive restructuring and exposure therapy, in community-dwelling older adults [18]. The physiotherapists received weekly supervision by a psychologist, based on video tapes of the therapy sessions. Likewise, the ‘Step by Step intervention’ aimed at reducing FoF after hip- or pelvic fracture, performed by physiotherapists who received weekly supervision by clinical psychologists, also had favorable effects on reducing FoF [27]. In our intervention protocol the supervision by psychologists was limited to monthly team meetings and individual coaching on request. In practice this supervision occurred less frequently. This is therefore an area of attention for the future.

Reflecting on the therapy intensity in our intervention, thus comparing the individual intervention items to various effective multi-component interventions for FoF, is not straightforward, as this is not always described in detail in the available literature. Also, tailoring of the intervention to the specific needs of the patients can complicate insight in the therapy intensity. The core element of the FIT-HIP intervention is guided exposure to feared activities, which is integrated in most of the therapy sessions. In other intervention programs this element was generally limited to one or two therapy sessions [28, 33]. Only the ABLE intervention, an in-home intervention for community dwelling older adults with excessive FoF, incorporated the exposure as a more elementary part of the program [29]. To the best of our knowledge, based on the intervention protocols, all programs had comparable frequency of delivery for psychoeducation on home safety and relapse prevention. Comparable to our program, the ABLE program included psychoeducation on the background on anxiety consequences and rationale for treatment [28, 29, 33]. The only other treatment program for FoF in this specific target group, the ‘Step by Step intervention’ includes problem-solving and relaxation techniques as additional items as compared to the FIT-HIP intervention [33]. The intended therapy intensity of cognitive therapy in this program was similar to our intervention. Hence, the therapy intensity of the individual FIT-HIP intervention items, in the form of therapy frequency, does in itself not clearly explain the lack of effectivity of the FIT-HIP intervention.

Regarding the feasibility of the telephonic booster (6 weeks after discharge): this element proved to be easily forgotten, as the physiotherapist was no longer involved in the patient’s treatment after discharge. We incorporated the booster in the intervention based on lessons learned from the programs based on a ‘Matter of Balance ‘ [30, 34], and the insight that (increase in) FoF is common after discharge from geriatric rehabilitation [35]. We can, however, question whether a telephonic booster is useful for our target group, as patients who received the booster reported no benefit from this intervention element. Perhaps it would be more appropriate to extend the treatment for FoF to an ambulatory rehabilitation setting (in-home) [27, 36].

An important barrier to acknowledge is the limited level of the FoF reported after enrollment in the study (i.e. selection of the target population). Facilitators pointed out that during screening (first week of rehabilitation), patients were mainly sedentary. Once patients started the process of mobilization (i.e. walking during therapy), in clinical practice the FoF appeared to decrease. The timing of the intervention in relation to the timeline after fracture may be a relevant factor to consider in the selection of the target group. Current literature illustrates that FoF present 2–4 weeks after fracture is not associated with negative effects on long-term functional outcomes, contrary to FoF present 6–12 weeks post-fracture [7, 8]. Provided that the fear is not disproportionate and does not lead to significant avoidance behavior (activity restriction), this could imply that FoF shortly after fracture can in some cases be a normal or adaptive process which does not require treatment. Unfortunately, for this specific group of patients, it is currently unknown what a disproportionate level of FoF is as measured with established instruments such as the FES-I. We can question whether the standard cut-off values are appropriate for this target group, especially because the FES-I appears to be more closely related to functional performance than to psychological concepts such as anxiety [37]. Patients with hip fracture experience a sudden impairment of the lower body function, and a certain level of ‘caution’ in relation to an increased fall risk in the early stage of recovery after fracture, may be an appropriate response. For clinical practice it seems relevant to monitor the course of FoF. Findings from a cohort study of hip fracture patients show three distinct patterns of FoF evolving from 4 to 12 weeks after fracture; i] patients with consistently low levels of FoF; ii] patients with high levels of FoF at 4 weeks that continue to increase; iii] patients with high levels of FoF at 4 weeks which decrease at 12 weeks post-fracture [38]. It is currently unknown how these distinct trajectories relate to avoidance behavior. However, it is plausible that especially those patients that have increasing levels of FoF are more susceptible to develop activity restriction as a consequence of FoF. Accordingly this may be an important group to address by means of intervention.

Another factor to consider when screening for FoF, is the (mediating) role of anxiety (traits) in the development of maladaptive or dysfunctional fear of falling [18, 39, 40]. Findings from Bower et al. show that patients with higher scores for neuroticism were more likely to have high levels of FoF [38]. Also, the previously mentioned in-home cognitive behavioral program for FoF that was conducted by physiotherapists and showed positive effects on reducing FoF and activity restriction, was aimed at patients with disproportionate FoF; as defined as high fear and low to moderate objective fall risk and functional impairment because of FoF [29]. The majority of participants had a psychiatric disorder, most frequently a pre-existing anxiety disorder [18]. In contrast, the FIT-HIP study population reported low scores for anxiety, had a lower level of FoF at baseline (Falls Efficacy Score-International); and we excluded patients with generalized anxiety [12, 21]. It may therefore be useful to incorporate screening for more generalized anxiety symptoms and also specifically include patients with anxiety for treatment.