Qualitative Evaluation of a Novel Educational Tool to Communicate Individualized Hip Fracture Prognostic Information to Patients and Surrogates: My Hip Fracture (My-HF)

Introduction Mortality and morbidity are high for older adults after hip fracture (HF), but patients and surrogate decision makers (SDMs) are typically unaware of the poor prognosis. We developed a novel educational tool, My Hip Fracture (My-HF), to provide patients and SDMs of patients hospitalized with acute HF individualized estimates of their post-HF prognosis. We conducted initial usability testing of My-HF in a sample of patients with HF and SDMs. Materials and Methods My-HF provides information about: 1) anatomy and risk factors for HF; 2) Hip fracture treatment received; 3) individualized predicted risk of adverse events and 4) anticipated discharge trajectory. We conducted a qualitative usability study using a convenience sample of hospitalized, post-operative patients with acute HF or SDMs of patients who lacked decision-making capacity. We used semi-structured interviews to obtain feedback. Thematic analysis was used to identify themes and concepts. Results We conducted interviews with 8 patients and 9 SDMs (mean age of interviewees 70.1 years, 41% female). My-HF was generally well received. Thematic analysis identified legibility and visual appeal, comprehension, numeracy, utility and reflection as prominent themes. Most respondents found My-HF to be useful in improving their understanding of HF and as a potential mechanism for sharing information with other care team members (including family and professionals). Suggestions for improvement of legibility, presentation of the individualized prognosis information and content were identified. Discussion Patients and SDMs are generally accepting of My-HF and found it useful for communicating individualized prognostic information. Feedback identified areas for improvement for future iterations of the tool. Conclusion My-HF presents a means of addressing the gap in understanding of prognosis post-HF as a part of patient-centered care. Further evaluation will be needed to assess the impact of My-HF on patient and SDM reported outcomes as we transition from a paper to smart-phone enabled web application.


Introduction
Low impact hip fracture (HF) remains a major cause of morbidity and mortality for older adults worldwide, despite improvements in care over the past 20 years. [1][2][3][4] Not only are patients with HF subject to a 3-to 8-fold increase in mortality in the first 30 days, but survivors also have increased mortality rates up to 10 years post-fracture compared to peers who did not experience a fracture. 5,6 Further, frail elders who survive fracture typically have persistently reduced function even after intensive rehabilitation. [7][8][9] Prior research suggests that patients and surrogate decision makers (SDMs) have limited understanding of the seriousness of HF, including limited awareness of risk of death and limited functional recovery. [10][11][12][13] This discrepancy is concerning because it violates foundational principles of patient-centered care. 14,15 Additionally, lack of understanding of illness prognosis and severity can be associated with frustration, anxiety and decision regret. 16,17 There is a growing body of literature suggesting the need for earlier introduction of palliative care and advanced care planning (ACP) for frail elders with an array of conditions, including HF. 18,19 Lack of understanding of HF prognosis has significant implications for patient and SDM readiness to engage in ACP. 20 Recognizing a gap in current HF care, we assembled a multi-disciplinary team to develop a novel educational tool (My Hip Fracture [My-HF]) to improve communication about HF treatment and convey individualized prognostic information to patients and SDMs of patients hospitalized with acute HF. We describe the development of My-HF and the results of qualitative usability testing conducted in a sample of patients with HF and their SDMs.

My-HF Description
We convened a multi-disciplinary team of geriatricians, orthopaedic surgeons, hospitalists, physiotherapists, palliative care specialists and medical decision scientists to develop My-HF over a 3-year period. First, we conducted a foundational study to explore patient and SDM understanding of HF along with gaps in knowledge. 10 Based upon our findings and input from our team, an initial paperbased draft of My-HF was built, recognizing that the paper instrument would ultimately be converted to a smart-phone enabled web-app for eventual widespread evaluation. Key information was presented in 4 discrete sections: 1) hip anatomy and HF risk factors; 2) primary HF treatment; 3) individualized predicted risk of major adverse event within 30 days and discharge to post-acute care and 4) anticipated post-HF discharge trajectory. Drafts were circulated amongst team members, reviewed and refined iteratively though in-person meetings and email until a suitable draft was developed [see Supplemental File 1].
The individualized predicted risk section [see Supplemental File 1, Section 3] was designed to facilitate clinicians' communication of complication rates for individual patients, easily calculated based upon demographic and clinical risk factors using the American College of Surgeons National Surgical Quality Improvement Program (ACS-NSQIP) risk calculator [21][22][23] The risk calculator is publicly available and commonly used in the surgical setting, but the output is not designed for patients. We designed My-HF to facilitate communication of individualized risk of 2 key adverse outcomes in a userfriendly, patient-centered fashion: 1) death or major complication within 30 days of surgery and 2) discharge to post-acute care (rehabilitation or long-term care). Major complication included cardiac arrest, myocardial infarction, pneumonia, renal failure/progressive renal insufficiency, pulmonary embolism, deep vein thrombosis, return to operating room, surgical site infection, sepsis, unplanned intubation and wound disruption-all as defined by the ACS-NSQIP. We considered several numerical and graphical options for presenting risk, ultimately choosing percentage risk represented numerically and on a thermometer bar. This technique incorporates both numerical estimate and visual aid, both of which enhance comprehension. 24,25 Because the goal of this study was to assess the usability and acceptability of My-HF, patients and SDMs were not provided with their individualized risk estimates. Our tool was designed for a grade 6 reading level.

Study Population
We assessed the usability of our paper-based instrument through semi-structured interviews with a convenience sample of patients aged ≥65 years, hospitalized with acute isolated HF on the orthopaedic service in a major Toronto teaching hospital between May and December 2019. Patients were excluded if: 1) unable to provide informed consent and their SDM was unavailable or 2) unable to communicate in English and lacked an English-speaking SDM. For patients who had cognitive impairment (identified as a Mini-Cog score of <3 26 or documented severe dementia), psychiatric illness, or a significant language barrier, we sought SDM participation reflecting that in clinical practice My-HF would be administered to SDMs when patients lack capacity. This study was approved by the research ethics board (IRB), and all methods were performed in accordance these guidelines and with the Declaration of Helsinki. Informed written consent was obtained from all participants prior to study participation.

Data Collection
Patients and SDMs participated in semi-structured, audio recorded, in-person interviews conducted by research team members (AK or CV) to explore reaction to My-HF using a structured interview guide [see Supplemental File 2]. We collected basic participant demographics: age, sex, educational attainment and relationship to patient (for SDMs). Field notes were reviewed and discussed, and the interview guide was iteratively revised to explore emerging themes. Data collection continued until no new themes or concepts arose and thematic saturation was achieved. 27 Interviews were conducted in hospital setting, prior to patient discharge; duration was 15-45 minutes.

Data Analysis
We used a constructivist framework to guide analysis. 28,29 Interview recordings were transcribed verbatim and analyzed using MaxQDA qualitative software (VERBI Software, Berlin, Germany). Inductive thematic analysis was used to identify themes and concepts. All investigators participated in developing an initial coding scheme and identify themes, and 2 team members (AK and CV) subsequently refined code definitions and hierarchies by coding 4 representative interviews. Differences in coding were reconciled by consensus. The remaining interviews were coded by CV. Coding and emerging themes were discussed and updated at regular team meetings. Finally, we re-grouped transcripts by participant type (patient vs SDM) and re-read them to see if themes varied by role. 28
We identified 5 themes with related subthemes described below with exemplar quotes. The source for each quote (patients [P], SDM [SDM], interviewer [I]) is shown.

Theme 1: Legibility and Visual Appeal
Most participants felt that the lay-out, colours and images were clear and legible; however, some issues with font size and coloured backgrounds were identified. Patients, particularly those who did not have their corrective lenses, were more likely than SDMs to identify concerns. 'Anything for people with [eye-related diseases], anything white and black is the best' (P 06)

Theme 2: Comprehension
Terminology. Most participants found the tool easy to understand and that explanations of medical terminology improved clarity. Certain medical terms (hemiarthroplasty, arthroplasty, comfort care, long-term care, acute care, rehabilitation and palliative care) were cited as confusing by multiple subjects. Images. Respondents felt most images improved understanding, but 2 were identified as unclear (Table 2 and Supplemental File 1).

Theme 3: Numerical Understanding
Most participants had a general understanding that in Section 3 (Supplemental File 1), a higher percentage implied greater risk of an adverse outcome. A few participants had a very good understanding of risk as presented.
'Okay. So if it's 95%, only 1 out of 20 cases will not experience a complication. So very likely, very high risk'. (SDM 05) However, many participants made at least 1 error in interpretation (Table 3); errors were commonly identified

Theme 4: Utility
While most participants found My-HF useful, some found it to be of more limited value (Table 4). Most participants felt that the tool was useful in providing an opportunity to review or consolidate knowledge and as a mechanism for sharing information about their HF with friends, family, and other health care providers. Some thought the information would facilitate conversations with family or their care team about ACP or motivate behaviour change. Some felt the tool, or aspects of the tool, were less useful because: 1) information interpreted as not useful or confusing; 2) information was already known; or 3) information was discordant with what respondents desired.
Suggestions for improvement: Split composite endpoint into the probabilities for individual components (e.g., death separated from non-death outcomes).
'75%, I don't know if it's death or the-so I would definitely need which one-like, for me, each picture needs specific numbers, each one of them'. (SDM 07) Remove section 3B 'need for rehab or LTC after acute care' altogether as most patients know their discharge destination by this time and applying a probability was perceived as confusing. A deeper discussion with their health care team about the My-HF individualized prediction information, particularly if the prediction was worrisome.
'No. I'd want somebody to explain it to me if it was that high. If it was like 50%, then written information would be fine. But anything above that, I think you need someone to explain it to you'. (P 03) Some participants suggested that My-HF results would best be reviewed prior to surgery, others after or that timing was less important.
'You could get it before or after. Before, you're really not going to know too much. But after, no, I think it would be very helpful either way'. (P 03)

Theme 5: Reflection
Affect. Some participants felt that presentation of individualized prognosis data would produce negative emotions-such as sadness, worry, or fear-in themselves and/or others. Others had a positive response, expressing appreciation for inclusion of individualized risk and these sensitive topics. 'To me, that's a little bit not depressing, but-I don't know. Some people take it differently. I don't know. I wouldn't like that'. (P 02) 'I mean, if you're telling me 'Okay, it's time to move your spouse into palliative care-' yeah, I mean it's a fact of life. I don't think there's any issues with that at all'. (SDM 08) Applicability to Self. Participants frequently reflected on how and if the information presented related to their current situation. Participants who were more accepting commonly noted that My-HF was useful precisely because 'it applies to me', while respondents who felt the information did not apply to them more commonly reported inutility.
'I: Do you find that information helpful or useful? P: Yeah, it applies to me" (P 09)  'I mean, obviously, the risk of death is of more of concern than the risk of a urinary tract infection which can be treated, right'? (P 11) Dislike percentage referring to discharge destination 'So I'm just not sure that that gives me, as a percentage, any useful information. If anything, it makes me think 'Well, Christ, you should know yes or no.' What kind of care do I need'? (SDM 08) P-Patient, SDM-surrogate decision maker.

Discussion
We assessed the usability and acceptability of My-HF in a sample of patients and SDMs of patients hospitalized with acute HF. Respondents reported My-HF to be useful in improving their understanding of HF and as a mechanism for sharing information with family and their health care team. Participants also viewed the inclusion of individualized risk estimates as valuable in starting conversations about prognosis and prompting reflection about goals of care. My-HF should be considered in the context of current care of older adults with low impact HF. Many recent efforts to improve HF care have focused on the medical and surgical aspects of treatment. Efforts have included initiatives to reduce the delay in time to surgical repair, defining the optimal type of repair (e.g., hemiarthroplasty vs total hip arthroplasty) and choice of anesthesia (regional vs general), and incorporating and evaluating surgical comanagement between orthopaedic surgeons, geriatricians, and hospitalists. [30][31][32][33] There has been far less attention to addressing the gap in patient and caregiver understanding of HF prognosis. 10,11 While the internet provides many publicly available resources that discuss HF treatment and recovery, 34,35 these resources are not tailored to the prognosis or risk of the individual patient. Alternatively, the ACS-NSQIP risk calculator provides individualized information regarding prognosis but is not designed for patients or SDMs with limited medical knowledge or numeracy. 36 The importance of both individualized information and tailoring to numeracy and literacy is wellestablished in the patient education literature. [37][38][39] The lack of tailored HF educational materials belies a general lack of attention to patient-centred care for a condition that is often misconstrued as solely mechanical and fixed with surgery. The acute care of HF frequently overlooks the increased mortality risk that persists in seniors month-to-years after the inciting fracture is repaired. Likewise, despite a growing movement to view HF as an opportunity to discuss prognosis and ACP, communication of the increased risk of adverse outcomes has received limited attention. 40,41 Patient-centred HF care should involve a partnership between the care team and patients and SDMs that begins with conveying treatment and prognosis information in ways that they can understand. 14 Patient education is foundational to patients (and SDMs) having the information needed to actively engage in their care, including acute illness management and ACP. For those who are not ready to engage in care decisions, including ACP, patient education (or consciousness raising) interventions are important in early stages of behaviour change to progress patients and SDMs towards engagement. 42,43 Additionally, evidence from other diseases and conditions suggests that patient education about illness prognosis is associated with decreased anxiety, regret and even improvement in overall quality of life measures. 16,17 Thus, the development of My-HF represents a novel effort to integrate a well-studied risk prediction model (ACS-NSQIP) into a state-of-the art HF educational tool to bridge the gap between the clinical care team's and the patient's (or SDM's) understanding of HF trajectory.
It is important to comment on the specific feedback we received from patients and SDMs in our testing. Despite careful attention to literacy and numeracy during initial development of My-HF, patients and SDMs cited difficulty with medical jargon, unfamiliar terms (e.g., comfort care), and with certain graphics. This feedback reinforces the need for usability testing and will guide changes in the next iteration of our tool. Many patients and SDMs had difficulty understanding aspects of the individualized prognostic information including the interpretation of the composite endpoint and how the numerical risk applied to their personal circumstance. Difficulty with probabilistic risk interpretation is consistent with prior research, even among highly educated samples. 44,45 In reviewing My-HF, participants requested a reference to help in their understanding, which has been previously shown to improve patient understanding of numerical data. 46 Adding anchors or interactive simulated experiences to future versions of My-HF should help enhance understanding of risk prediction. While the provision of individualized risk estimates was viewed favorably, some participants were surprised by the range of possible adverse outcomes, especially death. Our finding that patients and SDMs are accepting of individualized HF prognostic information, even if somewhat alarmed, is consistent with studies from other conditions. 47,48 Limitations First, our study was conducted in a single academic medical center and should be generalized with care. Second, health literacy and numeracy of participants were not formally evaluated, but this mimics how My-HF would be used in actual real-world clinical practice. Finally, our study asked participants to review My-HF as a hypothetical clinical tool and did not provide them with their personal individualized prognostic information, consistent with the goals of our usability testing. Future studies are required to evaluate the impact of My-HF when providing patients and SDMs with their actual prognostic information and assessing the impact of My-HF on outcomes including understanding, satisfaction and readiness to engage in advanced care planning.

Next Steps
We are revising My-HF based upon the feedback we received. Several of the issues identified by participants were inherent to our paper-based tool and will be easily addressed as we convert our paper instrument to a smart-phone enabled web application. For example, customizable settings will allow users to see only the treatment and discharge information applicable to them and adjust font size and background for the visually impaired. Medical jargon will be removed or clarified. Specifically, 'Palliative Care' will be removed as a disposition destination and replaced by 'Comfort Care' with options of home palliative support or palliative care unit so as to avoid perpetuating misconceptions of palliative care and life-sustaining care as mutually exclusive. The application will also have functionality that allows users to click on terms or words that are confusing to allow for greater explanation. Our web-app will directly integrate our HF prediction model into the interface and also allow for alternative presentations of numerical risk-such as the use of simulated experiences to convey probabilistic information. 49 Finally, it is important to make decision tools that are easy to use at bedside, and while well received, a paper-based tool would prove too cumbersome for day-to-day clinical use. After converting My-HF into a web-app we will conduct a pilot randomized trial in preparation for an anticipated definitive large multi-centre randomized controlled trial to evaluate the impact of My-HF on an array of patient reported outcome and experience measures (PROMs and PREMs).

Conclusion
In conclusion, patients and SDMs found My-HF to be useful for communicating HF risk and individualized prognostic information. After conversion to a web-app we will assess the effect of My-HF on patient and SDM experience measures including satisfaction, regret, and readiness to engage in advanced care planning.

Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the Mount Sinai Hospital Resident Research Grant. Dr Cram is supported in part by a grant from the US National Institute of Aging [R01AG058878] and also receives support from a salary award through the University of Toronto Faculty of Medicine. The funding source had no role in the design, analysis, and interpretation of the data, review or approval of the manuscript or decision to submit for publication.

Ethical Approval
This study received ethics approval from the Sinai Health System research ethics board (IRB) Approval number 18-0309-E

Informed Consent
All participants provided informed written and verbal consent prior to participation in the study.

Data Availability
The datasets generated and/or analysed during the current study are not publicly available to protect the confidentiality of participants. De-identified data may be made available from the corresponding author upon request.

Supplemental Material
Supplemental material for this article is available online.