• Open Access

The use of haematopoietic stem cell transplantation in Fanconi anaemia patients: a survey of decision making among families in the US and Canada

Authors

  • Sadie P. Hutson PhD, RN, WHNP, BC,

    Corresponding author
    1. College of Nursing, University of Tennessee, Knoxville, TN, USA
    2. Clinical Genetics Branch, Division of Cancer Epidemiology and Genetics, Department of Health and Human Services (DHHS), National Cancer Institute (NCI), National Institutes of Health (NIH), Rockville, MD, USA
    • Correspondence

      Sadie P. Hutson, PhD, RN, WHNP, BC

      University of Tennessee Knoxville

      College of Nursing

      1200 Volunteer Blvd., Room 161

      Knoxville

      TN 37996

      USA

      E-mail: shutson@utk.edu

    Search for more papers by this author
  • Paul K. J. Han MD, MA, MPH,

    1. Center for Outcomes Research and Evaluation, Maine Medical Center Research Institute, Scarborough, ME, USA
    Search for more papers by this author
  • Jada G. Hamilton PhD, MPH,

    1. Cancer Prevention Fellowship Program, Center for Cancer Training, NCI, NIH, DHHS, Rockville, MD, USA
    Search for more papers by this author
  • Sean C. Rife MA,

    1. Department of Psychology, Kent State University, Kent, OH, USA
    Search for more papers by this author
  • Mohamad M. Al-Rahawan MD, MPH,

    1. Department of Pediatrics, University of Illinois College of Medicine at Peoria, Peoria, IL, USA
    Search for more papers by this author
  • Richard P. Moser PhD,

    1. Science of Research and Technology Branch, Division of Cancer Control and Population Sciences, NCI, NIH, DHHS, Rockville, MD, USA
    Search for more papers by this author
  • Seth P. Duty RN, BSN,

    1. College of Nursing, East Tennessee State University, Johnson City, TN, USA
    Search for more papers by this author
  • Sheeba Anand MPH,

    1. Quillen College of Medicine, East Tennessee State University, Johnson City, TN, USA
    Search for more papers by this author
  • Blanche P. Alter MD, MPH

    1. Clinical Genetics Branch, Division of Cancer Epidemiology and Genetics, Department of Health and Human Services (DHHS), National Cancer Institute (NCI), National Institutes of Health (NIH), Rockville, MD, USA
    Search for more papers by this author

Abstract

Background

Fanconi anaemia (FA) is a rare genetic disorder associated with bone marrow failure (BMF), congenital anomalies and cancer susceptibility. Stem cell transplantation (SCT) offers a potential cure for BMF or leukaemia, but incurs substantial risks. Little is known about factors influencing SCT decision making.

Objective

The study objective was to explore factors influencing patients' with FA and family members' decision making about SCT.

Design

Using a mixed-methods exploratory design, we surveyed US and Canadian patients with FA and family members who were offered SCT.

Main variables studied

Closed-ended survey items measured respondents' beliefs about the necessity, risks and concerns regarding SCT; multivariable logistic regression was used to examine the association between these factors and the decision to undergo SCT. Open-ended survey items measured respondents' perceptions of factors important to the SCT decision; qualitative analysis was used to identify emergent themes.

Results

The decision to undergo SCT was significantly associated with greater perceived necessity (OR = 2.81, = 0.004) and lower concern about harms of SCT (OR = 0.31, = 0.03). Qualitative analysis revealed a perceived lack of choice among respondents regarding the use of SCT, which was related to physician influence and respondent concerns about patients' quality of life.

Conclusions

Overall, study results emphasize the importance of the delicate interplay between provider recommendation of a medical procedure and patient/parental perceptions and decision making. Findings can help providers understand the need to acknowledge family members' perceptions of SCT decision making and offer a comprehensive discussion of the necessity, risks, benefits and potential outcomes.

Introduction

Fanconi anaemia (FA) is a complex, rare, primarily autosomal recessive disorder associated with a broad array of congenital and developmental abnormalities and a high risk of aplastic anaemia and haematological and solid tumour malignancies.[1, 2] Patients with FA have >50% cumulative incidence of severe bone marrow failure (BMF) before age 20 and >10% incidence of acute myelogenous leukaemia (AML) by age 30.[1, 3]

Treatment options for FA are limited, particularly in the context of AML, and are complicated by toxicity of chemotherapeutic agents due to patients' abnormal DNA repair in all tissues.[4] Androgens may be used to treat aplastic anaemia; however, hematopoietic stem cell transplantation (SCT, usually bone marrow transplantation, BMT) is the only cure for BMF.[5] If an FA-negative, human leucocyte antigen (HLA)-matched sibling donor is available, SCT is considered the first choice for treating BMF in patients with FA.[6] Patients with mismatched (usually unrelated) donors are at higher risk of mortality due to transplant-related complications.[5] Matched sibling cord blood transplantation has also been used successfully following pre-implantation genetic diagnosis to rule out FA and, by definition, provides a tissue-matched donor.[7]

It is preferable to use SCT before complications develop from chronic transfusions or infections, as these may increase the risk of poor outcomes. Improvement in SCT-conditioning regimens and transplant techniques over the past several years has made SCT safer and more successful and has resulted in more patients surviving to adulthood.[8] For example, the use of fludarabine-based conditioning regimens along with lower radiation doses has reduced graft failure from 10 to 1% and improved long-term survival rates to more than 80% with matched unrelated donors.[9]

However, SCT remains a risky and distressing procedure for patients and their families,[10] and the decision to undergo SCT is difficult.[8] Patients with FA are more vulnerable to the toxicity of the conditioning regimen, facing potential complications such as failure to engraft, acute and chronic graft vs. host disease (GVHD), long-term morbidity including growth failure, cognitive defects, endocrine dysfunction and secondary malignancy (usually squamous cell carcinoma).[11] Other issues complicating decisions about SCT include the appropriate timing of transplant (weighing the benefits of early SCT against practical challenges including identifying an appropriate donor), the effects of SCT on patients' quality of life,[12, 13] and ethical issues including the selection of appropriate decision-makers, rights of minors and use of experimental therapies. Further, only a handful of institutions are considered expert in SCT for patients with FA; this requires families to consider temporary relocation to the city in which transplantation will occur.

To facilitate informed and shared decision making about SCT, clinicians must elicit and understand patients' and family members' disease- and treatment-related beliefs, concerns and preferences.[14] Such factors have emerged from limited research on BMT treatment decisions for other hematological disorders and include fear of treatment consequences, expectations of benefit and trust in physicians.[15, 16] However, no prior research has investigated the extent to which these and other potentially important factors, including perceptions of the benefits and risks of SCT, specifically influence SCT decisions in patients with FA and families. This knowledge gap represents a potentially important barrier to shared decision making in this vulnerable patient population, as well as in other genetic disorders.

The purpose of our study was to explore factors influencing patients' with FA and family members' decision making about SCT. Our reason for choosing SCT as the primary outcome variable is because it is the only known cure for the haematological complications of FA. We were specifically interested in three factors related to perceptions of SCT: perceived necessity (cognitive beliefs regarding the necessity of SCT for maintaining health), perceived risk of harms (cognitive beliefs regarding the likelihood of adverse effects of SCT) and concerns about harms (emotional worries regarding the adverse effects of SCT). These factors were selected because previous empirical research has confirmed their relevance to decision making about other medical treatments.[17-19] For example, studies of medication adherence in chronic disease have shown that patients' perceptions of the necessity of the medication and of the likelihood of potential harms, as well as the level of concern about these harms, are major determinants of medication adherence.[17, 18] Furthermore, for many patients, FA is a chronic disease that they have been managing for considerable time prior to making a transplant decision, typically with medications.

It is important to note, however, that treatment decisions in FA are more complicated than many medication decisions. Among patients for whom a HLA-matched sibling donor does not exist, they must decide to either pursue an unrelated donor transplant or androgen (i.e. medication) therapy which can also be beneficial for severe cytopenias. However, the decision to use androgens is challenging because only approximately half of patients will respond to androgen therapy and many of these patients later become refractory to treatment. There is also uncertainty regarding the time of response and potential treatment side-effects (e.g. virilization, premature epiphyseal closure, mood swings, hypertension, transaminitis, cholestatic jaundice, peliosis hepatis and increased risk of liver tumours).[1]

While the decision to undergo SCT in FA is likely more complicated than many medication decisions in chronic illness, psychological factors known to influence medication adherence may be nonetheless influential and deserve exploration. This is especially true because there are no existing theoretical models regarding the determinants of transplant decisions in haematological diseases such as FA. As an initial step towards developing such a model, therefore, we conducted an exploratory study using research on chronic illness medication adherence to identify potential determinants of the transplant decision. Specifically, we reasoned that factors known to influence chronic illness medication decisions might also be important in decision making about SCT in patients with FA, given the poor prognosis of FA, the high risk of iatrogenic harm from SCT and the complexity of the decision surrounding androgen therapy. We hypothesized that patients with FA and families would be more likely to choose SCT when they perceived greater necessity of SCT, perceived lower risks and had fewer concerns about harms of SCT.

We conducted an exploratory mixed-methods survey investigation to understand patients' with FA and their families' self-reported perceptions of the necessity and risks of SCT, and their level of concern about its harms. We used quantitative analysis of closed-ended survey items to examine the extent to which these psychosocial factors were related to patients' with FA and family members' decision to undergo SCT, and qualitative analysis of open-ended items to gain additional insights about how these factors might influence decision making.

Materials and methods

Study design and respondents

We initiated this cross-sectional survey in November 2007. Family support organizations in the US (Fanconi Anaemia Research Fund) and Canada (Fanconi Canada) mailed surveys to 905 members of registered families representing 451 patients. Non-responders from the first mailing were sent a second questionnaire 2 months later. We queried mothers, fathers, adult patients (18 years or older at the time of the survey) and adult patients' spouses/significant others separately. The study was approved by the National Cancer Institute Institutional Review Board (NCI protocol 02-C-0052) and is registered in www.clinicaltrials.gov (NCT00027274).

Survey

The survey consisted of 55 items assessing self-reported demographics, details about patients' FA diagnosis, medical treatments including SCT (age at SCT, transplant type, donor and degree of HLA-match), health outcomes such as cancer and patients' decision making about SCT.

Perceived necessity of SCT

Perceived necessity of SCT (hereafter referred to as perceived necessity) was measured using a five-item scale adapted from the perceived necessity subscale of the Beliefs about Medicines Questionnaire (BMQ).[17] Exemplary questions asked respondents the extent to which they believed the patient's future health would depend on undergoing SCT and that without SCT the patient would be very ill. Responses to each item were made on a four-point scale ranging from 1 = ‘not at all’ to 4 = ‘a great deal.’ The mean of these items was computed to produce a total scale score, with higher values indicating greater perceived necessity. This scale demonstrated good internal consistency (Cronbach's α = 0.79).

Concern about harms of SCT

Concern about harms of SCT (hereafter referred to as concern) was measured using a 10-item scale adapted from the concerns subscale of the BMQ to make it more specific to the potential harms of SCT. Respondents indicated how worried or concerned they were about issues including physical side-effects of transplant and conflicts with religious or ethical beliefs. Responses were made on the same four-point scale mentioned above. Higher values on the total scale score derived from the means indicated greater concern (Cronbach's α = 0.83).

Perceived risk of harms of SCT

Perceived risk of harms of SCT (hereafter referred to as perceived risk) was assessed with an 11-item scale designed by us to ascertain perceptions of the likelihood of different potential harms–for example, post-transplant cancer, infection and death–associated with SCT. Respondents were asked how likely they thought it was that the patient would experience each of these harms, using a four-point scale ranging from 1 = ‘very unlikely’ to 4 = ‘very likely.’ Higher values of the mean total score indicated greater perceived risk (Cronbach's α = 0.84).

Decision to undergo SCT

Decision to undergo SCT was the main dependent variable for the quantitative analyses and was assessed with a single item which asked ‘did the patient undergo transplant?’ Responses were coded as 0 = ‘no’ and 1 = ‘yes’.

Factors influencing the SCT decision

Factors influencing the SCT decision were assessed using two open-ended, descriptive questions as follows: (i) In your own words, what was your major reason for choosing or not choosing transplant?, (ii) Please explain your understanding of why you/he/she/they made the transplant decision.

Data analysis: quantitative

Data for each study variable (perceived necessity, perceived risk, concern and decision to undergo SCT) were examined for missing values using SPSS 14.0 (IBM Corporation, Armonk, NY, USA) Missing Values statistical analysis software. Missing values were not differentially related to specific variables. We excluded from analysis surveys with ≥20% missing items for any component scale variables. Next, we computed descriptive statistics and bivariate correlations among study variables. We then used multivariable logistic regression to model the association between the independent variables of perceived necessity, perceived risk, and concern and the dependent variable of the decision to undergo SCT. Because it has been theorized that cognitive and emotional variables may interact to influence decision making,[20, 21] we e explored potential moderating effects between the independent variables by fitting models that included interaction terms, using mean-centred predictors to reduce potential multicollinearity.[22] All analyses were adjusted to correct for the effects of potential confounding variables that were available for all patients regardless of their SCT decision, including patients' vital status at the time of the survey, patient sex, country of birth, age at FA diagnosis and the major reason for SCT. All statistical tests were two-tailed with α = 0.05.

Because the sampling strategy allowed for responses from multiple decision-makers per patient (e.g. patient, mother and father), the data were correlated at the family unit level, thereby violating the assumption of independence between cases. Thus, all quantitative analyses and standard error estimates were corrected to account for the correlated data with the use of SPSS 14.0 Complex Samples statistical analysis software, which uses Taylor linearization to compute accurate standard errors for statistical testing. This technique allowed for the inclusion of all decision-makers per patient, maximizing the size of the analytic sample while protecting against inaccurate variance estimates.

Data analysis: qualitative

Qualitative, descriptive responses to the two open-ended questions were transcribed verbatim from the surveys and analysed using line-by-line qualitative content analysis (NVivo 7.0; QSR International, Doncaster, Vic., Australia[23]), consisting of detailed examination of words and phrases to generate rich descriptions and to identify thematic categories.[24] Two authors (SPH and SPD) performed open and axial coding, consistent with prescribed qualitative techniques.[24] During axial coding, the authors worked independently, then jointly, to identify relationships between themes. Disagreements were resolved through discussion. The result of this process yielded one major theme with two subthemes.

Results

We received 398 questionnaire responses from the 905 individuals in North America who were sent a survey (response rate 44%). Two hundred and fourteen questionnaires were respondents for patients who had been offered SCT and were eligible for the present study. After the exclusion of 14 cases with substantial missing data, the analytic sample for the quantitative analyses consisted of 200 respondents (patients with FA and parents combined, who provided complete responses about their decision-making experiences regarding 139 individual patients with FA). The qualitative analyses included 189 respondents who provided responses to the two open-ended questions regarding 134 patients.

Quantitative findings

Demographic characteristics

Respondent characteristics are shown in Table 1. The majority of survey respondents (87%) reported that the patient with FA had received a transplant. Most respondents were mothers (59%) or fathers (36%) of patients with FA. Mothers (84%) and fathers (76%) were also most frequently identified as persons responsible for the SCT decision (respondents could endorse more than one responsible person). Demographic and medical characteristics of the 139 patients with FA are shown in Table 2; no significant differences emerged between those who chose SCT and those who did not. The majority of patients were alive at the time of the survey (65%). The most frequent reason for being offered transplant was aplastic anaemia (68% overall, 70% of those who received SCT and 55% of those who did not receive SCT). More than one-third of those transplanted reported complications including acute GVHD (25%) and chronic GVHD (16%).

Table 1. Respondent characteristics (n = 200)
 Patient had SCTPatient did not have SCTTotal 
 (n = 174)(= 26)(= 200) 
 (column%)(column%)(column%) P a
  1. a

    Pearson chi-squared significance tests used to compute P-values. Dashes indicate a P-value was not computed due to inadequate cell size.

  2. b

    Column percentages do not total 100% because respondents could endorse more than one person as responsible for the decision.

Respondent's relationship to patient
Mother98 (56)19 (73)117 (59)0.11
Father65 (37)7 (27)72 (36)0.30
Self10 (6)0 (0)10 (5)0.37
Other1 (1)0 (0)1 (1)
Person responsible for SCT decisionb
Mother146 (84)21 (81)167 (84)0.74
Father135 (78)17 (65)152 (76)0.40
Self/Patient20 (12)4 (15)24 (12)0.48
Other14 (8)2 (8)16 (8)0.97
Missing2 (1)2 (8)4 (2)
Table 2. Patient characteristics (= 139)
 Patient had SCTPatient did not have SCTTotal 
 (n = 117)(= 22)(= 139) 
 (column%)n (column%)n (column%) P a
  1. Dashes indicate that either data were not collected for a characteristic or that a P-value was not computed due to inadequate cell size.

  2. AGVHD = acute graft vs. host disease; CGVHD = chronic graft vs. host disease.

  3. a

    Pearson chi-squared significance tests were used to compute P-values unless otherwise noted.

  4. b

    Age in years.

  5. c

    t-test used to compute P-value.

Alive at time of survey (yes)77 (66)13 (59)90 (65)0.55
Male/Female59 : 5815 : 774 : 650.13
Country of birth   0.17
US96 (82)21 (96)117 (84) 
Canada or other21 (18)1 (5)22 (16) 
Median age at diagnosis of FAb4.9 (range: 0.0–18.3)5.0 (range: 0.1–12.1)4.9 (range: 0.0–18.3)0.99c
Median age at transplantb8.5 (range: 1.7–33.5)
Major reason for offer of transplant
Aplastic anaemia82 (70)12 (55)94 (68)0.15
Myelodysplastic syndrome8 (7)2 (9)10 (7)0.97
Leukaemia6 (5)1 (5)7 (5)0.86
Unknown0 (0)1 (5)1 (1)
Other21 (18)6 (27)27 (19)0.23
Donor type
Sibling46 (39)
Parent7 (6)
Unrelated adult50 (43)
Unrelated cord12 (10)
Unknown2 (2)
Source of stem cells
Cord24 (21)
Bone marrow81 (69)
Mobilized peripheral blood10 (9)
Unknown2 (2)
Degree of HLA matching
Matched sibling (6/6–12/12)39 (33)
Partial-match sibling (5/6– 1/12)2 (2)
Partial-match parent (5/6–11/12)4 (3)
Matched unrelated donor (6/6–12/12)32 (27)
Partial-match unrelated donor (5/6–11/12)23 (20)
Poor match (≤5/6–11/12)11 (9)
Unknown6 (5)
Transplant-related complications
AGVHD29 (25)
CGVHD19 (16)

Perceived necessity, risk and concern

Descriptive statistics and bivariate correlations between study variables are shown in Table 3. These analyses demonstrated that perceived necessity, perceived risk and concern were significantly correlated with one another (with the exception of perceived necessity and concern, which were not significantly associated). Furthermore, each variable was significantly correlated with the decision to undergo SCT (all  0.01). Specifically, greater perceived necessity, lower perceived risk and lower concern were each associated with the decision to undergo SCT. However, in multivariable logistic regression analyses, only greater perceived necessity (OR=2.81, 95% CI=1.41–5.61) and lower concern about harms of transplant (OR = 0.31, 95% CI = 0.11–0.91) were associated with the decision to undergo SCT (Table 4). No potential confounding variables were associated with the SCT decision, and the same pattern of findings was observed regardless of whether a patient's mother or father completed the survey (results not shown; sample sizes for patient (= 10) and other (= 1) respondents were too small to allow for such comparisons).

Table 3. Correlation coefficients (r) and descriptive statistics for study variables (= 200)
 1234
  1. Numbers in the column and row headers are the same. M, mean, SD, standard deviation.

  2. a

     0.01.

  3. b

     0.001.

1. Perceived necessity of SCT−0.20a−0.070.31b
2. Perceived risk of harms of SCT 0.65b−0.22a
3. Concern about harms of SCT  −0.23b
4. Decision to undergo SCT   
M (SD)3.57 (0.63)2.84 (0.59)2.81 (0.62)0.87 (0.34)
Median3.802.822.841.00
Score Range1–41–41–40–1
Table 4. Multivariable logistic regression model of psychosocial variables associated with the decision to undergo SCT (= 200)
Independent variableOdds ratio95% CI P
  1. Analyses adjusted for whether the patient was alive at time of survey, patient sex, country of birth, age at FA diagnosis and major reason patient was offered SCT; ps ≥ 0.10 for these variables in the tested model.

Perceived necessity of SCT2.81(1.41–5.61)0.004
Perceived risk of harms of SCT0.69(0.24–2.00)0.50
Concern about harms of SCT0.31(0.11–0.91)0.03

Moderation analyses testing for possible interactions between perceived necessity and perceived risk, perceived necessity and concern, and concern and perceived risk in predicting the decision to undergo SCT showed no significant interactions (results not shown). Thus, it appears that perceived necessity and concern are the primary correlates of the decision to have SCT, and their potential effects are not influenced by the levels of the other psychosocial variables examined in this model.

Qualitative findings

One major theme emerged from the qualitative data: Lack of Perceived Choice regarding the Use of SCT in FA. Additionally, data yielded two contextual subthemes: (i) physician influence and (ii) quality of life. These themes and subthemes are illustrated below by selected quotations from study respondents.

I) Lack of Perceived Choice regarding the Use of SCT in FA.

We asked respondents to provide additional details regarding the major reason why SCT was or was not chosen. This theme was apparent in both groups of respondents: patients/families who underwent SCT, and those who were offered but did not undergo SCT.

One mother whose child underwent transplant stated:

We had no choice. If we did not go for transplant, we would have lost our daughter regardless. With transplant we at least had a chance. Was it a great chance? Not at all, but we had to try. It was [one] of the worst decisions [we have] ever had to endure. (Patient-Transplanted/Deceased)

Another mother whose child did not undergo transplant expressed:

None of the specialists we contacted believed she could survive a transplant and if we wanted to have one we would have had to decide very early (immediately) to avoid the effects of multiple transfusions. We felt we would have her with us longer by not pushing forward with transplant. There was really no choice. (Patient- Untransplanted/Deceased)

These quotations corroborate the dominant quantitative finding of our study – that is, the strong association between the perceived necessity of SCT among patients and family members and the decision about whether to undergo SCT. Many patients with FA and families simply saw no options other than the choice they made, and therefore did not engage in further consideration of the risks and benefits of SCT.

Physician Influence

These qualitative data provided further insights on the origins of respondents' perceptions of the necessity of SCT and the lack of decisional options. An important subtheme that emerged was the importance of physician influence in the decision-making process. For those who decided to undergo transplant, the advice of their physician(s) helped solidify the decision to undergo SCT. One mother shared:

Her counts were failing and we had to decide on treatment. Several options were offered to us. Six months prior we had attended a FA summer camp…several physicians were recommending transplant as a first response. We thought it was best to take her to transplant early. (Patient- Transplanted/Alive)

Another father stated:

Once we understood [our daughter] wouldn't get better on her own, we decided we would do anything to give her a chance at a full life. Her doctor reassured us that he would make the same decision if it were his child and stated ‘it'll be ok…we can do this.’ (Patient- Transplanted/Alive)

Physician influence is known to be powerful in medical decision making, and it is not surprising that this is the case in decisions about SCT in patients with FA. The quotations illustrate that for many patients with FA and families, the physician's role was a driving force in SCT decisions.

Similarly, physician influence was strong for those individuals who were offered but did not undergo SCT. One mother wrote:

I chose not to take [my son] to transplant due to the fact that he did not have a donor that was a good enough match and because of the severity of the leukemia. Doctors said he had less than 10% chance to make it through pre-transplant. (Patient- Untransplanted/Deceased)

In this case, the physicians apparently emphasized the low likelihood of success of the procedure and the high likelihood of harm; this information was pivotal. The data in these quotations do not have sufficient detail to fully explore the extent of directiveness of counselling efforts. However, they do illustrate the great influence of the physicians' recommendations on the decision to undergo or forego SCT and suggest that these recommendations may be a source of patients' and families' perceptions of the necessity of the procedure.

Quality of life

Choosing whether or not to undergo SCT for reasons related to quality of life was another prevalent subtheme. Many families noted that the SCT decision was made with the hope for improved quality of life for the patient post-transplant. One father remarked:

We agreed that in the face of suddenly diagnosed and rapidly advancing AML, on top of fundamental bone marrow failure due to FA, this (transplant) was the [patient's] only chance. She was a gifted outstanding student who had great quality of life and enormous potential if the disorder could be remediated by BMT. We (mother, father, and patient) did not believe in capitulation to the disorder if transplant, even against long odds, had a colourable [sic] chance of success. (Patient- Transplanted/Deceased)

As this quotation illustrates, many individuals decide to undergo transplant hoping that quality of life will be improved. In contrast, other families felt that quality of life might not be improved by undergoing SCT. Consistent with our quantitative findings regarding the role of patients' and family members' concerns about harms of SCT in decision making, many respondents expressed a belief that SCT had the potential to increase suffering and worsen patients' quality of life. As one mother wrote:

I did not proceed with a transplant because I didn't think my daughter would survive. I was also terrified that she would suffer mercilessly. I witnessed transplants go very wrong on the unit and kids died often. I worry everyday that I didn't make the right decisions for my precious child and miss her so much. She ended up suffering with a lung infection and at least she would have had a chance, even if it was a smaller one, with the BMT. On other days I believe she had 3 more years of life by not having the BMT. I suppose I will always be haunted by this. (Patient- Untransplanted/Deceased)

The passages eloquently depict the turmoil these families experience while trying to make a treatment decision. Many wonder if the decisions they made concerning treatment were the best for the patient; many will never know. These findings shed light on yet another set of factors influencing patients' with FA and families' perceptions of the necessity of SCT. The qualitative data suggest that perceptions of the impact of SCT on patients' quality of life are critically related to perceptions of the necessity of SCT and may potentially attenuate these perceptions or moderate their influence on the SCT decision.

Discussion

This is the first study to explore psychosocial factors associated with the decision to undergo SCT in individuals with FA. Taken together, the triangulated quantitative and qualitative findings of our study highlight important challenges presented by this complex decision. The quantitative results convincingly indicate that, among the psychosocial variables assessed in the present study, perceived necessity is most strongly associated with respondents' decisions regarding the use of SCT; those who perceived SCT as necessary were more likely to choose transplant than were those who did not. Furthermore, concern about potential harms was also associated with the decision to undergo SCT; those respondents who experienced less concern were more likely to choose transplant than were those who experienced more concern about harms from SCT. These findings are notably consistent with studies of medication adherence.[17, 18] Although greater perceived risk of harms of SCT was negatively associated with the decision to undergo SCT in bivariate analyses, it was not significantly associated in multivariable analyses adjusting for all psychosocial variables. It is possible that the effects of perceived risk on the decision to undergo SCT are mediated by concern or perceived necessity; however, exploratory analyses (results not shown) did not support this explanation. Nonetheless, these findings suggest that the effects of perceived risk are weaker and likely confounded by these other variables.

The qualitative results validate and extend the quantitative findings by suggesting possible mechanisms by which perceptions of necessity influence SCT decisions. The qualitative data demonstrate a marked lack of perceived choice among many patients and family members regarding the use of SCT in FA; this finding helps make sense of the strong quantitative association between perceived necessity and the decision to either undergo or not undergo SCT. These findings suggest that families make decisions about SCT with the perception that SCT is so necessary that a real choice does not exist. This raises important ethical questions that cannot be answered by the current study, because each patient's and family's unique clinical circumstance, and the extent to which other options were presented to them is unknown. Some families may truly have had no other viable choices due to characteristics of their own disease and treatment history. Further, it is possible that these reports regarding perceived lack of choice reflect post-decisional cognitive processes, which we are unable to decipher given our choice of data collection method. Still, the qualitative approach used in this study represents a significant strength.

For other patients and families, however, one can ask whether their perceived choice was indeed the only choice. It is also possible that respondents' perceived lack of options may have contributed to psychological distress regarding the treatment decision. Our quantitative results showed that concern about SCT was associated with the decision to undergo SCT; however, the extent to which this concern may have been influenced by their perceived lack of options is unknown. Our findings raise the need for future research to ascertain the extent of shared decision making about SCT among patients with FA and families and to develop a better understanding about the origins and effects of patients' perceptions of limited treatment options. High perceived necessity of SCT and low concern about its harms might be justified if these responses resulted from a thoughtful process of shared decision making; however, they might be unjustified if they arose from a lack of information or adequate deliberation. This problem also applies to decisions about numerous other diseases and conditions of unclear benefit for individual patients – ranging from prostate-specific antigen (PSA) screening to coronary revascularization procedures for chronic stable angina. In such cases, the ethical responsibility falls on clinicians to ensure that adequate deliberation about all available treatment options occurs.

These problems are all the more important given our qualitative findings of physicians' significant influence in decision making, which also call for caution in the way physicians approach patients with FA and families and involve them in the decision-making process. Because of their great potential influence on patients' decisions, health-care providers must take extra care to elicit, understand, and respect patients' with FA and families' perceptions, concerns, preferences, and values and to encourage their participation in the decision-making process.[14] In the context of a complex disorder such as FA, this is a difficult undertaking. More research is needed to better understand and to develop ways of facilitating shared decision making about SCT among patients with FA. This is particularly true given our study's other findings regarding the difficulties that patients with FA and families experience in weighing the substantial but uncertain effects of SCT on patients' quality of life.

Our study has several strengths. It represents the first to explore SCT decision making among patients with FA and families, an area of particular interest as a result of the state of uncertainty in which treatment choices must be made. We had a 44% response rate among respondents from North America, suggesting that many FA family members have unmet psychosocial needs regarding the decision-making process for SCT. Insights into individual or family-level decision-making factors are important to elucidate, as they tend to be more amenable to interventions, such as trials of shared decision making. Our results provide the foundation for future research with families and health-care providers to more fully explore decision-making perceptions and avenues for improving the process by which a treatment decision is reached.

One limitation to the present study is the small number of respondents who decided against SCT (26 respondents representing 22 patients), which limits the power to detect true differences and potential interactions or mediational relationships between the psychosocial variables included in our model and as such, introduces bias. Statistical adjustments were made for the effects of multiple medical and demographic variables; however, these data were self-reported, limited in scope, and may not have fully captured the factors that could influence post-decision reports of the SCT decision-making process. Future studies may benefit from abstracting such data from additional sources including patient medical records. Furthermore, data were not available regarding when the offer of SCT was made to patients and their families. This is an important limitation, because the timing of the SCT decision may have influenced responses. There may be greater recall bias for those respondents who made decisions long ago. Further, a cohort effect may exist, influencing perceptions of necessity, risk and concern about SCT due to temporal improvements in the efficacy of SCT for treating FA; those offered SCT in recent years are likely to have higher odds of survival, even with unrelated donors, than those offered SCT in the past. Nonetheless, the magnitude of the quantitative results, as well as their consistency with the findings of the qualitative analyses, support the validity of our conclusions. However, to fully understand the predictors of SCT decisions in this population, future studies should include detailed measures tailored to the unique experiences of the study population, and recruit more respondents who have not undergone the procedure. Moreover, next steps should include the investigation of factors identified in previous work related to BMT decision making (fear of treatment consequences, expectations of benefit, and trust in physicians,[15, 16]) which we did not directly address in the present study.

Another area for future research is to examine how the influence of various decision-making factors may differ between patients with FA and their parents or surrogate decision-makers. People may make different medical decisions when choosing for themselves vs. others, and the parental perspective in particular may increase preferences for risk-reducing active treatments.[25] More work is needed to examine how patient vs. parental perspectives influence decisions about SCT and other interventions in patients with FA and their families.

Overall, our study results emphasize the importance of the delicate interplay between provider recommendation of a medical procedure and patient/parental perceptions and decision making. Providers must recognize that their discussion of necessity and risks of SCT may be critical to the ultimate decision to choose or refuse the procedure. Thus, our findings can help providers understand the need to better acknowledge family members' perceptions of the SCT decision-making process and offer a more comprehensive discussion of the necessity, risks, benefits and potential outcomes.

Acknowledgements

We are extremely grateful to the Fanconi Anaemia Research Fund and Fanconi Canada for mailing questionnaires, and to all participating families. We would also like to acknowledge the College of Nursing at East Tennessee State University for assisting with database construction and resources for qualitative analyses.

Funding supported

This research was supported in part by the Intramural Research Program of the NIH and NCI in Rockville, MD, USA (SPH, PKJH, BPA, JGH, RPM).

Conflicts of interest

We have no conflicts of interests or financial interests to disclose.

Ancillary