Development of the AMPDECIDE Decision Aid to Facilitate Shared Decision Making in Patients Facing Amputation Secondary to Chronic Limb Threatening Ischemia

Abstract

Introduction:

We developed a patient decision aid to enhance patient participation in amputation level decision making when there is a choice between a transmetatarsal or transtibial amputation.

Methods:

In accordance with International Patient Decision Aid Standards, we developed an amputation level patient decision aid for patients who are being considered for either a transmetatarsal or transtibial amputation, incorporating qualitative literature data, quantitative literature data, qualitative provider and patient interviews, expert panel input and iterative patient feedback.

Results:

The rapid qualitative literature review and qualitative interviews identified five domains outcome priority domains important to patients facing amputation secondary to chronic limb threatening ischemia: 1) the ability to walk, 2) healing and risk for reamputation, 3) rehabilitation program intensity, 4) ease of prosthetic use, and 5) limb length after amputation. The rapid quantitative review identified only two domains with adequate evidence comparing differences in outcomes between the two amputation levels: mobility and reamputation. Patient, surgeon, rehabilitation and decision aid expert feedback allowed us to integrate critical facets of the decision including addressing the emotional context of loss of limb, fear and anxiety as an obstacle to decision making, shaping the decision in the context of remaining life years, and how to facilitate patient knowledge of value tradeoffs.

Conclusions:

Amputation level choice is associated with significant outcome trade-offs. The AMPDECIDE patient decision aid can facilitate acknowledgment of patient fears, enhance knowledge of amputation level outcomes, assist patients in determining their personal outcome priorities, and facilitate shared amputation level decision making.

Introduction

Lower extremity amputation (LEA) is associated with a reduction in quality of life,¹ loss of mobility, as well as, a significant risk of mortality.² Choosing an amputation level for each individual patient is complex and influenced by factors such as limb perfusion, medical comorbidities, and surgeon experience.³ The overarching goal is to provide patients with the highest likelihood of healing, survival, and optimal functional outcome.⁴ However, these are competing goals that vary with amputation level; a transmetatarsal (TM) amputation may provide better functional outcome if healed, although it may reduce the likelihood of primary healing compared with a transtibial (TT) amputation, necessitating ongoing wound care or reamputation. The relative priority of each of these outcomes may differ between patients depending on their personal outcome preferences.

Shared decision making (SDM) is increasingly being advocated in the context of complex clinical care decisions,⁴ including vascular surgery interventions.^5,6 Patient decision aids (PtDAs) enhance the decisional process by enabling patients’ personal values and outcome priorities to be incorporated into the decision-making process. They have been shown to improve many dimensions of the decisional process including, improved patient knowledge, decreased decisional conflict, more accurate perception of risks, and increased satisfaction with the decision.⁷

The aim of this work was to develop a novel amputation level patient decision aid (named AMPDECIDE) for patients with chronic limb threatening ischemia (CLTI) facing either a TM or TT amputation. Our goal for the AMPDECIDE PtDA is to facilitate SDM by informing patients about the risks and benefits of each amputation level, and to assist them in better understanding their values and priorities so that they can participate in the amputation level decision.

Methods

Development team

The core development team consisted of a project coordinator, a vascular surgeon, a physical medicine and rehabilitation physician, an implementation scientist and decision aid expert, an epidemiologist and a research psychologist (authors of this manuscript). In addition, a national representative group of clinicians, implementation scientists and decision aid experts, termed an Expert Panel (Table 1), reviewed the PtDA and provided iterative feedback.

Table 1 –

Expert Input (including core team members n = 26).

Primary role	N (%)
Physical medicine and rehabilitation physician	4 (15.4%)
Vascular surgeon	3 (11.5%)
Podiatric surgeon	1 (3.8%)
Rehabilitation psychologist	1 (3.8%)
Epidemiologist	1 (3.8%)
Health services/implementation science	4 (15.4%)
Research psychologist	1 (3.8%)
Physical therapist	2 (7.7%)
Prosthetist	7 (26.9%)
Decision aid expert consultants	2 (7.7%)

Development framework

Development was guided by the Ottawa Decision Support Framework,⁸ as well as, International Patient Decision Aid Standards standards.

Literature reviews

A rapid review of the qualitative literature was performed to determine the outcome domains of importance to patients who have had a LEA secondary to CLTI. Existing studies along with input from clinical subject matter experts and librarians were used to define search terms and strategies.^1,9 Included articles were published between January 2010 and June 2019 and met the following criteria: 1) used qualitative research methods, 2) focused on patient values who had undergone a TT, TM, or partial foot amputation secondary to CLTI, 3) were conducted in a high-income country, and 4) were written in English. We extracted the study purpose, sample characteristics, and main themes from each article that met our inclusion criteria (see Appendix 2 for more details about literature reviews). Domains were structured to ensure that they encompassed all factors relevant to the decision and that each factor was independent.

A similar rapid review was performed to compare the quantitative differences in outcomes of importance. Included articles were published between December 2015 and July 2019 and met the following criteria: defined comparative outcome risks of TT and TM amputations: 1) used a randomized control trial or a cohort study design, 2) included unilateral incident amputations caused by CLTI, 3) presented data on an outcome of interest derived from the qualitative review and were within 5 y postamputation, 4) were written in English. The study purpose, sample characteristics, outcome measures, and risks (%) were extracted for each outcome.

Qualitative needs assessment

To help ensure that the decision aid informed important patient decisional needs, a qualitative needs assessment was performed including a sample of both patients and providers from different clinical settings and geographical regions. The patient sample included individuals with TM or TT amputation, as well as those who had TM amputation with a subsequent revision. Providers were physicians and surgeons involved in the care of CLTI patients undergoing amputation (Fig. 1). Interviews focused on patient experiences of amputation and factors that contribute to amputation level decisions. An iterative, team-based deductive-inductive approach to conventional content analysis was used to identify themes pertinent to the PtDA.^10,11

Fig. 1 –

AMPDECIDE transmetatarsal/transtibial patient decision aid development strategy.

Decision aid development process

Patients who had undergone a prior TM or TT amputation in the past 12 mo were recruited from VA Eastern Colorado Health Care System and VA Puget Sound Health Care System (VAPSHCS) to participate (Table 2). A cohort of consecutive eligible patients was recruited at each medical facility between February 2020 and April 2021. Potential participants were screened by clinical staff for eligibility. Those that were eligible and interested were referred to the study coordinator at each site. Participants viewed the most current version of the PtDA (both paper and online versions). They were interviewed either in person or by telephone using a semistructured interview guide. Interviews contained both open-ended questions and focused questions assessing the acceptability of the PtDA. A think aloud technique¹² was also used. Participants verbalized their thoughts as they interacted with the PtDA while the interviewer took detailed notes about any difficulties, they experienced negotiating the aid or questions they had about information presented. The interviewer probed for ideas on ways to improve the decision aid. Patient demographics, comorbidities and pertinent amputation history were collected (Table 2), and participants also completed The Acceptability Questionnaire, a 5-item survey used to quantitatively assess clarity, and freedom from bias or coerciveness in the version of the aid viewed.¹³ All aspects of the study were conducted at VAPSHCS and VA Eastern Colorado Health Care System with approval from their Institutional Review Boards and participant consent.

Table 2 –

Demographic information of enrolled patient participants (n = 18).

Patient characteristics	Mean (SD)/n (%)
Age (mean, SD)	61.78 (9.11)
Race
White	16 (88.9%)
Black/African American	1 (5.6%)
Other	1 (5.6%)
Ethnicity
Not Hispanic or Latino	17 (94.4%)
Hispanic or Latino	1 (5.6%)
Education
Some high school	1 (5.6%)
High school	6 (33.3%)
Some college	5 (27.8%)
Graduated college	6 (33.3%)
Marital status
Single (never married)	2 (11.1%)
Married/domestic partnership	8 (44.4%)
Separated	1 (5.6%)
Divorced	5 (27.8%)
Widowed	2 (11.1%)
Living status
At home with spouse or family or friend	12 (66.7%)
At home alone	6 (33.3%)
Smoking status
Nonsmoker	12 (66.7%)
Currently smoking	6 (33.3%)
Patient status
Preop	2 (11.1%)
Postop	16 (88.9%)
Amputation level*
Mid-foot (transmetatarsal)	5 (31.3%)
Below-knee (transtibial)	11 (68.8%)

^*Excludes participants enrolled preamputation.

The patient decision aid development process

The first versions of the PtDA were developed by the core development team on paper, allowing for rapid, iterative edits to streamline the content and language. Eleven iterations were completed, at which point additional feedback was solicited from the Expert Panel (Table 1). After incorporating this feedback, it was shown to patients starting with version 12. After every 2-3 interviews, the decision aid was further modified based upon patient feedback. Feedback from the Expert Panel was again solicited on paper version 14, and again on version 17. Development of the online version began after paper version 21. Twelve subsequent iterations of the on-line version were completed following a similar methodology. The final AMPDE-CIDE PtDA can be seen in Appendix 1A and at www.ampdecide.org. Patient and provider videotaped segments were included in the online PtDA to provide visual/auditory modes of information presentation as an alternative to text. Patient segments included individuals with TT and TM amputations and focused on their amputation experience while provider segments included brief discussions of the advantages and disadvantages of each level and how patients might balance these outcome differences based upon their own outcome priorities.

Results

Qualitative literature review

Twenty articles (including n = 289 patients with lower limb amputation) were identified that met study inclusion criteria. Studies varied considerably as they addressed different amputation research questions and were performed across varied time horizons postamputation, spanning 1-wk to 2 y postamputation. Five final domains were identified as outcome priorities for CLTI patients: 1) the ability to walk, 2) healing and risk for reamputation, 3) rehabilitation program intensity, 4) ease of prosthetic use, and 5) the importance of limb length preservation after amputation.

Review of the quantitative literature

Two manuscripts were identified that met inclusion criteria, with a combined sample of n = 5460.^13,14 Quantitative data to define outcome differences between the two amputation levels was only available to support two of the five domains patients identified as important in the qualitative review: the ability to walk and the risk for reamputation. Outcomes from these two studies were included in the decision aid.

Qualitative needs assessment

Twenty-two patients who had undergone a TM or TT amputation in the past 12 mo and 21 providers completed qualitative interviews between March and June of 2019 (see Tables 3 and 4). Patients were identified via the VA corporate data warehouse. Patients were purposively sampled to represent varied geography, and we oversampled women and people of color to try to ensure diverse experiences were represented. We identified providers who typically participate in amputation level decisions (vascular and podiatric surgeons and physiatrists) and reached out to these individuals via VA service line ListServs. Saturation was verified when both qualitative coders noted no more themes were being identified and the larger investigator team concurred. An iterative, team based deductive-inductive content analysis was used to identify themes related to the amputation experience and amputation level decisions. Two analysts independently coded the same transcripts and discussed points of divergence and convergence. Emergent themes were discussed in analytic meetings and developed iteratively through conversations among the analytic and larger study teams. ATLAS.ti (version 8.2) was used for data management throughout data analysis. Several important findings were identified from the qualitative interviews and literature reviews: 1) mortality is an important overarching construct that may have wide-reaching implications for amputation-level decision-making; providers communicated its importance but also reflected that it is something that is rarely addressed in provider-patient discussions, and there was concern that it may have an adverse effect on patient rehabilitation and psychological status. Patients, in contrast, had varying levels of comfort with mortality discussions; 2) future mobility and wound healing are competing priorities and their relative importance differs between patients, and also between patients and providers; 3) many factors identified as being important by both patients and providers did not vary by amputation level; 4) both patients and providers reported a desire to engage in, and the importance of SDM¹⁵; and 5) providers and patients reported that SDM occurred infrequently.

Table 3 –

Qualitative needs assessment patient participant characteristics (n = 22).

Variable	N (%)
Age
40-49	1 (4.55)
50-59	4 (18.18)
60-69	5 (22.73)
70-79	11 (50.00)
80-89	1 (4.55)
Race
White	13 (59.09)
Black	7 (31.82)
Latinx	2 (9.09)
Setting
Emergency department	10 (45.45)
Clinical	12 (54.55)
Amputation level
Transtibial with or without revision	12 (54.55)
Transmetatarsal without revision	5 (22.73)
Transmetatarsal with any revision	2 (22.73)
Region
North Atlantic	6 (27.27)
Southeast	4 (18.18)
Midwest	4 (18.18)
Continental	4 (18.18)
Pacific	4 (18.18)

Table 4 –

Qualitative needs assessment provider participant characteristics (n = 21).

Variable	N (%)
Specialty
Vascular surgery	10 (47.61)
Podiatric surgery	5 (23.80)
Physical medicine and rehabilitation physician	6 (28.57)
Gender
Female	9 (42.86)
Male	12 (57.14)
Region
North Atlantic	0 (0)
Southeast	6 (28.57)
Midwest	3 (14.29)
Continental	2 (9.52)
Pacific	10 (47.61)

Important considerations that were identified and addressed during the development of the patient decision aid

Emotional context of the decision

Patient interviews highlighted that the need for amputation is a highly emotional experience,¹⁶ and that it was important to acknowledge this within the decision aid. Patient feedback in an early beta version indicated that the aid was too “dry” and lacked the emotional context of the amputation experience. In response, patient quotes from the qualitative needs assessment were included. Subsequent feedback indicated that this recognition and validation was welcome.

Importance of framing the decision within the context of remaining life

The 1-y mortality after incident LEA is very high in this patient population, but there is little difference between TM and TT amputation levels.² Provider qualitative interviews suggested the discussion of mortality risk was difficult, but important to include in the PtDA,¹⁷ “I wouldn’t say that I relish that opportunity, but I’m fairly matter of fact. I recognize the mortality risks for our patients, and I try to sort of inform them about the implications of amputation and what that means in terms of long-term survival.” The expert panel and core development team also felt it was important to include. Although difficult, an understanding of the high mortality rate may alter patient outcome priorities, and therefore change their amputation level choice. Its inclusion in the PtDA may encourage this challenging conversation, and was therefore included (Appendix 1E).

Including personal risks versus population risks

The core development team has previously created and validated an amputation level decision support tool (DST) that enables the calculation of an individual patient’s 1 y risk of mortality, reamputation and mobility at each LEA level.^13,14,18,19 As part of the development process, we queried patients and clinicians about their preferences for incorporating personalized, or average risks in the decision aid. Physicians preferred the presentation of average risks, as personalized risk aids were several years from developedn, and an average risk aid allowed the patient to have more timely access to the decision aid. Patients communicated that if they were to learn of their personalized outcomes, they would prefer this to take place during a personal patient/physician visit, “This is good information, but other people, my wife, my doctor, should see this with me.” After considering this feedback, the expert panel decided to proceed with average risks for this first edition of the aid, but to consider it for possible inclusion in future modifications.

Value trade-offs

An important aspect of a PtDA that distinguishes it from an educational document is helping patients explore value trade-offs.²⁰ The AMPDECIDE PtDA encourages patients to consider the pros and cons of each amputation level. Similar to other decision aids,²¹ the online version of the aid includes an interactive slider bar that highlights the relative outcome advantages of each amputation level.

The importance of other patient informational needs

Our qualitative interviews suggested that patients experience anxiety when they are informed about the need for amputation. This is driven by uncertainty about future pain, the importance of social support, possible changes in relationships, and identity; whether they may experience a loss of independence, as well as a possible need to modify their living environment.²² While an understanding of these factors may not change an amputation level decision, they are integral to improving the patients understanding of how their lives might be altered by amputation, and they would also be valuable in prompting them to ask important questions during discussions with their provider. Thus, the final PtDA includes a short description of each domain (Appendix 1H).

Finalizing the patient decision aid

The online decision aid (https://www.ampdecide.org/decisionaid/tm-tt/) was deemed final after no novel suggestions were presented during patient interviews, expert panel feedback and core development team meetings. The final aid is written at a below-8th grade level, according to both the Simple Measure of Gobbledygook and the Flesh-Kincaid Grade Level tests. At the end of the development process, 18 patients completed the measure of acceptability (Table 5); The majority (39%) reported the right amount of information while nearly equal numbers indicated that the aid presented too little (28%) and too much (22%) information, reflecting the varied knowledge level needs of patients undergoing amputation. A majority (83%) found it to be “completely balanced” in its presentation of information about the two options and 74% indicated that they would “definitely recommend” the aid.

Table 5 –

Patient reported acceptability of decision aid (n = 18).

Acceptability responses	N (%)
1. How would you rate the amount of information in the educational materials?
A little less than I needed	5 (27.8%)
About the right amount of information	7 (38.9%)
A little more information than I needed	4 (22.2%)
A lot more information than I needed	2 (11.1%)
2. How balanced was the information about getting a partial foot amputation vs below knee amputation?
A little slanted toward choosing partial foot amputation	3 (16.7%)
Completely balanced	15 (83.3%)
3. Did the educational materials present one option as the best overall choice?
Yes, the educational materials favored another option	5 (27.8%)
No, the educational materials were neutral and balanced	13 (72.2%)
4. How helpful were the educational materials in helping you make a decision about treatment options?
Not helpful	2 (11.1%)
A little helpful	3 (16.7%)
Somewhat helpful	4 (22.2%)
Very helpful	9 (50%)
5. Would you recommend the educational materials to other people who are facing the same decision?
I would probably recommend them	4 (22.2%)
I would definitely recommend them	14 (77.8%)

Implementation considerations

The need for amputation may be discussed in multiple clinical environments and with differing levels of patient acuity. While few patients expressed difficulties navigating through the online tool, many expressed that they would also want it available as a paper copy. To address this, the developed online aid allows for it to be printed. Both patients and providers reported that it should be made available when the possible need for amputation was first discussed in an outpatient environment, but also when an amputation level decision was more imminent, in either an emergency department or after admission to hospital. Some patients expressed a desire to review the aid with a member of their care team, while others preferred to share it with friends and family. The flexible format of the AMPDECIDE PtDA allows for the possibility of all these options.

Discussion

We have developed the AMPDECIDE PtDA to facilitate SDM in patients whose clinical presentation allows for consideration of either a TM or a TT amputation. Few PtDAs are available to facilitate SDM in vascular surgical conditions.^23–26 There is a perceived need for improvement in SDM, and acknowledgment of the valuable role PtDAs may play in improving SDM in these clinical conditions.^6,27,28 In the specific context of amputation level decisions, there is a prevailing patient perception that they are not involved in amputation decisions at the time of informed consent.¹⁷ However, they desire greater involvement in amputation decisions.^29–31 The primary purpose of developing this PtDA is to better inform patients about available options, their risks and benefits, and to facilitate their conceptualization of their own outcome values and priorities, therefore providing a foundation for further patient/surgeon SDM. It can also function both as an informational tool for family members, and as a tool to facilitate patient/family discussions.

The process of developing this PtDA identified key issues involved in amputation level decision making, as well as concepts that may be important to the broader context of PtDA development. These include: (1) how best to communicate and deal with the challenges of the emotionally laden and complex life-changing consequences of amputation with its impact on the multiple dimensions that affect quality of life; (2) the role of decisions in the context of limited life expectancy, particularly when the intervention does not specifically affect longevity; and (3) how the inclusion of general information may be used as a strategy to reduce anxiety and enhance SDM participation.

The loss of a portion of one’s body is highly emotionally charged. Amputation adversely affects body image and selfidentity, as well as a fear of loss of social roles, and sexuality.³² The loss of the ability to walk,¹ with its associated loss of independence and social engagement results in further distress.²³ Feedback from patients during the development process reinforced the importance of acknowledging these concerns in the decision aid. They specifically desired language that would support patients and help address these emotional concerns. This was incorporated early in the decision aid through patient quotes, structured language, and patient and provider video clips.

An additional challenge in the development of this PtDA was whether, and how best to include mortality risk. The choice of surgical procedure does not significantly affect mortality risk, but the overall population risk is extremely high.² Incorporating mortality risk in the PtDA was viewed as important because it allows patients to balance the differing risks of mobility outcome and failure of primary healing with the subsequent need for additional amputation surgery, in the context of “how they wish to spend the rest of their lives.”

Increasingly, the development of predictive models allow for the incorporation of patient specific risks in decision aids. The value of, and patient interest in, learning about their specific risks varies with the specific decision being considered.³³ Incorporating average risks may not adequately reflect individual patient risks, and therefore may distort its effect on decision making. Our parallel development of a DST that enables patient specific risks to be calculated, which has been integrated into a combined DST/PtDA website (www.ampdecide.org), made the incorporation of individualized risks plausible.¹³ Patient feedback suggested they were interested in learning their individualized risks but preferred that it occur when their surgeon was present to discuss any concerns. This information emphasizes the importance of developing the technological ability to transfer the DST output into the decision aid as well as overcoming the challenges associated with how best to incorporate the decision aid in patient care flow. This will enable surgeons to individualize patient discussions based upon their interest in learning about their patient specific, or average outcome risks.

Anxiety has been shown to be an important source of impaired decision-making.³⁴ To reduce anxiety, an informational component was included, which was concordant with a patient perceived need for additional information.³⁰ The topics included pain, the importance of social support, changes in relationships and identity, experiencing a loss of independence and possible modifications to the living environment.^23,32 This information was placed after the values clarification page, to ensure that it did not encumber the key decisional elements with excessive information.

A number of potential limitations are noteworthy. The PtDA was developed using a Veteran cohort consisting of predominantly white males which may affect its generalizability to other populations. However, the rapid review of the qualitative and quantitative literature which formed one of the foundations of the decision aid, incorporated both the broader US and international populations. Second, patient feedback incorporated into the iterative development was from individuals who had already faced the decision. Their feedback provided important insights from those who have adapted to the amputation experience. Future modifications may be necessary after it is evaluated in a patient cohort facing the decision. Third, this aid only addresses the TM/TT amputation decision. We have subsequently developed and conducted a preliminary evaluation of a TT/Transfemoral decision aid to address this additional amputation level decision (www.ampdecide.org/decisionaid/tt-tf). Future studies will include a larger scale efficacy trial as well as evaluating implementation strategies that will maximize dissemination of these valuable clinical tools.