Introduction

In 2018 over 1.2 million men were diagnosed with prostate cancer globally, with overall incidence expected to increase a further 42% by 2030 [1]. As incidence rises, advances in detection and treatment have led to improved survival rates in many countries, with Australia reporting 90.6% 10-year survival [2], the USA 98% [3], and the UK 84% [4]. Hence the prevalence of prostate cancer continues to rise: in Australia, more than 200 000 men are living with a previous diagnosis [2, 5]; 3 000 000 in the USA [6, 7], and over 300 000 men in the UK [8]. Problematically, after diagnosis and treatment many men (up to 40%) experience poorer quality of life and satisfaction with life over the long term (10 years) [9] and, even with localized disease, one in five men will experience persistent anxiety and depression 1 year after treatment [10], with distress greater in men with advanced disease [11]. Poorer long-term outcomes are associated with androgen deprivation therapy, multiple comorbidities, younger age at diagnosis, and socio-economic disadvantage [9]. Survivorship care that seeks to enhance health and well-being outcomes over both the short and longer term is therefore crucial for this patient group.

Limitations of existing survivorship guidelines include an over-reliance on expert opinion [12, 13], invisibility of the consumer voice [14, 15], a lack of translation into policy and practice [16], and, in the case of prostate cancer survivorship, gaps in knowledge [17]. Perspectives about masculinity and men’s health are notably absent [18]. Recent commentary proposes that survivorship care for men with prostate cancer needs to take into account unique disease-specific factors, both clinical and biological, as well as the subjective patient experience [12]. Given the increasing burden of prostate cancer and the lack of clear progress in the development, acceptance and delivery of quality prostate cancer survivorship care, a different approach is needed that connects current evidence, expert opinion and consumer perspectives.

The Delphi technique is a widely used method that seeks to forecast and elicit informed expert opinion and consensus in a structured and iterative approach. In health, the Delphi method has been used to develop the following: health system performance, prescribing and disease indicators [19]; clinical models of care [20, 21]; patient outcome measure sets [22]; and cancer survivorship classifications [23]. The policy Delphi [24] is appropriate for complex health policy issues and uses panel participants with a range of potentially different perspectives who are well informed and have a vested interest in the issue at hand, and specifically considers feasibility as well as importance [20, 25]. This method is cost-effective as it does not require multiple and ongoing committee meetings, avoids constraining committee processes and, through expert involvement, sets a platform for dissemination [26]. Accordingly, we undertook a policy Delphi to describe the current state of prostate cancer survivorship in Australia and New Zealand and identify survivorship domains and domain elements for inclusion in care guidelines taking into account evidence, importance, feasibility and consensus.

Participants and Methods Participants

Using purposive sampling through consultation with leading Australian (n = 46) and New Zealand (n = 1) clinical and community groups we identified 47 potential panel members who were leaders in the field with recognized authority in prostate cancer and survivorship care or who were able to represent the experiences of men with prostate cancer. Australian participants spanned six jurisdictions (New South Wales, n = 18; Queensland, n = 8; Victoria, n = 8; South Australia, n = 7; Western Australia, n = 3; and the Australian Capital Territory, n = 2). All of those invited agreed to participate. The panel included 31 nationally leading clinical, allied health, nursing and academic and community leaders and 16 consumers who had experience in the provision of support in the community (Table 1) from a range of professional and academic organizations (Table 2). Representatives from indigenous health, the LGBTIQ community, rural and regional as well as urban consumers, and partners of men with prostate cancer were included. Health professional and academic leaders had between 15 and 40 years of experience with prostate cancer, and, for men, their time since diagnosis ranged from 6 to 20 years (Table 1). Sample sizes for policy Delphi range from 10 to 30, with a maximum of 50 considered appropriate [25, 26].

Table 1. Participant characteristics. Demographic characteristics % (n) Age 18–44 years 13 (6) 45–54 years 19 (9) 55–64 years 26 (12) 65–74 years 36 (17) 75–84 years 4 (2) 85+ years 2 (1) Gender Male 66 (31) Female 34 (16) Role* Health professionals Urologist 4 (2) Medical oncologist 6 (3) Radiation oncologist 4 (2) General practitioner 4 (2) Physiotherapist 2 (1) Exercise physiologist 6 (3) Registered nurse 9 (4) Other (health professional) 19 (9) Consumers Patients 30 (15) Partners 8 (4) Family of survivors 4 (2) Time since diagnosis for survivors, years Mean (range) Patients 12 (6–20) Partners 15 (9–23) Health professional and academic leaders experience, years 16 (15–40) Table 2. Panel member organizational affiliations.

Australian and New Zealand Urogenital and Prostate Cancer Trials Group

Australia and New Zealand Urological Nurses Society

Australian Prostate Centre

Cancer Council Australia

Queensland Cancer Occupational Therapy Interest Group

Cancer Voices New South Wales

Centre for Research Excellence in Prostate Cancer Survivorship

Exercise and Sports Science Australia

Flinders Centre for Innovation in Cancer

Healthy Male

Macquarie Health

Medical Oncology Group of Australia

Peter MacCallum Cancer Centre

Primary Care Collaborative Cancer Clinical Trials Group

Prost! Exercise Group

Prostate Cancer Foundation of Australia

Prostate Cancer Foundation of New Zealand

Psychology Board of Australia

Royal Australian and New Zealand College of Radiologists

Urological Society of Australia and New Zealand

Regional and Major Urban Prostate Cancer Support Group Leadership

Procedure

A four-round policy Delphi was undertaken between 9 September 2019 and 20 February 2020, with ethical approval from the University of Technology Sydney (Approval #ETH19-3855). The first three rounds were administered using the QualtricsXM survey platform, with the final round conducted at a face-to-face meeting. Each survey was pilot-tested in advance and revised as needed. Survey response rates were 96% (n = 45) for the first survey round, 96% (n = 45) for the second round and 94% (n = 44) for the third round, and 28 panel members attended the face-to-face meeting. In Round 1, 47 people replied but only 45 had complete data. After Round 1, one of the expert panel members withdrew from the project owing to a role change.

The purpose of Round 1 was generation of ideas and views about prostate cancer survivorship. Open-ended questions invited respondents to list three words describing the current survivorship experience for men diagnosed with prostate cancer and then to outline what domains they viewed as essential for prostate cancer survivorship care. To stimulate participant reflection, full-text article links to the ASCO prostate cancer survivorship guidelines [13] and recently proposed domains for a Cancer Survivorship Care Framework [16] were provided.

In Round 2 a synthesized list of descriptors for the prostate cancer survivorship experience generated in Round 1 were provided for panel members to choose up to five words that most closely represented men’s current experience. Next, participants were asked to rate the survivorship care elements synthesized from Round 1 for importance and feasibility (1 – not important at all, to 7 – extremely important; 1 – not feasible at all, to 7 – extremely feasible). Importance was defined as the degree to which a survivorship element is of significance or value to improving prostate cancer survivorship care. Feasibility was the degree to which a survivorship element can be achieved, performed or implemented.

In Round 3 the survivorship elements from the previous rounds were thematically analysed to derive survivorship domains for the panel to consider the extent to which each domain should be included in prostate cancer survivorship guidelines (1 – not at all, to 7 – absolutely). For each domain, an open-ended question invited further commentary and suggestions for missing elements.

For Round 4, at a 1-day face-to-face meeting, participants reviewed the evidence for intervention for each survivorship domain through pre-reading and discussion of relevant systematic literature reviews [17], consumer perspectives [14], and a series of expert presentations. Participants were assigned to groups for each of the survivorship domains, each group was then asked to identify priority actions for change, and to consider feasibility of change in their designated domain to improve the survivorship experience in prostate cancer. Each group reported back to the other expert panel members on their identified priority actions. The expert panel members and two clinician-researchers (A.K., M.F.) were then given 10 votes each to vote on the top priority actions to target for change to inform future implementation. The top priorities for action were determined by the priority actions that received over 50% of the combined votes.

Analysis

Data from each round were considered verbatim and then underwent independent content and thematic analysis by three authors (J.D., A.G., S.K.C.), followed by discussion and consensus to provide synthesized data for the panel to consider in subsequent rounds. There are no universally accepted rules for consensus in the Delphi method. In the present study we determined the direction of consensus on seven-point rating scales using score categories of 6–7 as highly important/feasible/essential, 5 as moderately important/feasible/essential, 4 as neutral; 3 as less important/feasible/essential, and 1–2 as not important/feasible/essential [27]. The direction of consensus was defined as the proportion of agreement in either one agreement category (e.g. ‘highly important’), or across two contiguous categories according to the consensus rule [27] (e.g. ‘highly important’ and ‘moderately important’). The consensus rule was: high consensus – 70% in one agreement category or 80% in two contiguous categories; moderate consensus – 60% in one agreement category or 70% in two contiguous categories; low consensus – 50% in one agreement category or 60% in two contiguous categories; and no consensus – less than 50% in one agreement category or less than 60% in two contiguous categories [27, 28].

Results Experience of Prostate Cancer Survivorship

Round 1 elicited 135 words or phrases to describe the prostate cancer survivorship experience from which 30 unique words/phrases were identified. Of these, 18 were negative, four neutral, and eight positive. In Round 2, the top six words endorsed by at least 25% of the panel as best describing men’s current prostate cancer survivorship experience were: dealing with side effects; challenging; medically focused; uncoordinated; unmet needs; and anxious (Table 3).

Table 3. Endorsement of prostate cancer survivorship descriptors. Descriptors Endorsement % (n) Dealing with side effects 78 (35) Challenging 38 (17) Medically focused 33 (15) Uncoordinated 29 (13) Unmet needs 29 (13) Anxious 27 (12) Emotional 24 (11) Family relationships 22 (10) Variable 20 (9) Surveillance 18 (8) Optimistic 18 (8) Resilience 18 (8) Mostly ok 18 (8) Decision-making 16 (7) Well-being 13 (6) Confusing 13 (6) Resigned 11 (5) Distressing 11 (5) Living 11 (5) Relief 9 (4) Learning 9 (4) Positive 7 (3) Transformative 7 (3) Regret 7 (3) Burdensome 4 (2) Poor 2 (1) Isolating 2 (1) Helping 0 (1) Diminished 0 (1) Lifelong 0 (1) Prostate Cancer Survivorship Care Elements and Domains

In Round 1, participants described 310 elements of care that were synthesized through content analysis to produce 26 unique prostate cancer survivorship elements. In Round 2, most items were rated as very important with high consensus (22 items) and the remaining four items were very important to important with high consensus (Table 4). For feasibility, six items were rated as very feasible to moderately feasible with high consensus, three items as very feasible to moderately feasible with moderate consensus and the remaining 17 had either no [5] or low [12] consensus (Table 5). Through thematic analysis [29] and data consolidation [30] of these items, six domains of survivorship care were elicited: health promotion and advocacy; shared management; vigilance; personal agency; care coordination and evidence-based survivorship interventions (see Appendix 1 for domain definitions).

Table 4. Frequency (%) of response regarding the degree to which each element of survivorship is important (Round 2; N = 45). Consensus (direction)

Very important

(6–7)

n (%)

Moderately important

(5)

n (%)

Neutral

(4)

N (%)

Less important

(3)

n (%)

Not important

(1–2)

n (%)

Management of advanced symptoms High (VI) 45 (100.0) 0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0) Access to care High (VI) 45 (100.0) 0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0) Palliative care High (VI) 44 (97.8) 1 (2.2) 0 (0.0) 0 (0.0) 0 (0.0) Multidisciplinary teams High (VI) 43 (95.6) 2 (4.4) 0 (0.0) 0 (0.0) 0 (0.0) Managing physical effects High (VI) 43 (95.6) 1 (2.2) 1 (2.2) 0 (0.0) 0 (0.0) Psychosocial care High (VI) 42 (93.3) 3 (6.7) 0 (0.0) 0 (0.0) 0 (0.0) Up-to-date information High (VI) 42 (93.3) 3 (6.7) 0 (0.0) 0 (0.0) 0 (0.0) Surveillance of recurrence and second cancers High (VI) 42 (93.3) 2 (4.4) 1 (2.2) 0 (0.0) 0 (0.0) Care coordination High (VI) 42 (93.3) 3 (6.7) 0 (0.0) 0 (0.0) 0 (0.0) Shared and informed decision High (VI) 41 (91.1) 3 (6.7) 1 (2.2) 0 (0.0) 0 (0.0) Person-centred care High (VI) 41 (91.1) 3 (6.7) 1 (2.2) 0 (0.0) 0 (0.0) Family support for the patient High (VI) 41 (91.1) 3 (6.7) 1 (2.2) 0 (0.0) 0 (0.0) Prostate cancer specialist nurses High (VI) 40 (88.9) 5 (11.1) 0 (0.0) 0 (0.0) 0 (0.0) Monitoring physical effects High (VI) 39 (86.7) 4 (8.9) 2 (4.4) 0 (0.0) 0 (0.0) Exercise, physical activity and nutrition High (VI) 39 (86.7) 5 (11.1) 1 (2.2) 0 (0.0) 0 (0.0) Health promotion High (VI) 39 (86.7) 3 (6.7) 1 (2.2) 1 (2.2) 1 (2.2) Screening for psychosocial effects High (VI) 39 (86.7) 5 (11.1) 1 (2.2) 0 (0.0) 0 (0.0) Self-management High (VI) 39 (86.7) 4 (8.9) 1 (2.2) 1 (2.2) 0 (0.0) Management of comorbidities High (VI) 38 (84.4) 5 (11.1) 1 (2.2) 0 (0.0) 1 (2.2) Maintaining intimate relationships High (VI) 38 (84.4) 7 (15.6) 0 (0.0) 0 (0.0) 0 (0.0) Advocacy High (VI) 37 (82.2) 3 (6.7) 3 (6.7) 2 (4.4) 0 (0.0) Surveillance of comorbidities High (VI) 37 (82.2) 7 (15.6) 0 (0.0) 1 (2.2) 0 (0.0) Empowerment High (VI-MI) 35 (77.8) 8 (17.8) 1 (2.2) 1 (2.2) 0 (0.0) Psychosocial care of family members High (VI-MI) 30 (66.7) 13 (28.9) 2 (4.4) 0 (0.0) 0 (0.0) Financial assistance High (VI-MI) 28 (62.2) 9 (20.0) 6 (13.3) 2 (4.4) 0 (0.0) Peer support High (VI-MI) 27 (60.0) 12 (26.7) 6 (13.3) 0 (0.0) 0 (0.0) MI, moderately important; VI, very important. Table 5. Frequency (%) of response regarding the degree to which element of survivorship is feasible (Round 2; N = 45). Consensus (direction)

Very feasible

(6–7)

n (%)

Moderately feasible (5)

n (%)

Neutral

(4)

n (%)

Less feasible

(3)

n (%)

Not feasible

(1–2)

n (%)

Surveillance of recurrence and second cancers High (VF-MF) 32 (71.1) 8 (17.8) 3 (6.7) 2 (4.4) 0 (0.0) Up-to-date information High (VF-MF) 27 (60.0) 11 (24.4) 5 (11.1) 2 (4.4) 0 (0.0) Monitoring physical effects High (VF-MF) 26 (57.8) 15 (33.3) 3 (6.7) 1 (2.2) 0 (0.0) Management of advanced symptoms High (VF-MF) 25 (55.6) 15 (33.3) 3 (6.7) 1 (2.2) 1 (2.2) Palliative care High (VF-MF) 22 (48.9) 15 (33.3) 6 (13.3) 2 (4.4) 0 (0.0) Advocacy High (VF-MF) 20 (44.4) 16 (35.6) 4 (8.9) 4 (8.9) 1 (2.2) Health promotion Moderate (VF-MF) 19 (42.2) 14 (31.1) 8 (17.8) 2 (4.4) 2 (4.4) Exercise, physical activity and nutrition Low (VF-MF) 18 (40.0) 12 (26.7) 8 (17.8) 4 (8.9) 3 (6.7) Family support for the patient Low (VF-MF) 18 (40.0) 13 (28.9) 9 (20.0) 2 (4.4) 3 (6.7) Shared and informed decision Low (VF-MF) 17 (37.8) 12 (26.7) 11 (24.4) 5 (11.1) 0 (0.0) Prostate cancer specialist nurses Low (VF-MF) 17 (37.8) 10 (22.2)